A Textbook of Postpartum Hemorrhage

Color profile: Generic CMYK printer profile
Composite Default screen
A Textbook of
POSTPARTUM
HEMORRHAGE
A comprehensive guide to evaluation, management
and surgical intervention
This copy is presented free of charge by courtesy of

a generous grant from the
Society of Obstetricians and Gynecologists of Canada
Société des obstétriciens et gynécologues du Canada
(SOGC)
as part of their ongoing commitment to international
women’s health and medical education
1
Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #W.vp
25 September 2006 14:18:53
HRH The Princess Royal
This book is dedicated to all women who have died
from postpartum hemorrhage and to those who strive to
prevent and overcome it.
A Textbook of
POSTPARTUM
HEMORRHAGE
A comprehensive guide to evaluation, management
and surgical intervention
Edited by
Christopher B-Lynch FRCS, FRCOG, D.Univ
Milton Keynes General Hospital NHS Foundation Trust, Oxford Regional Health Authority, UK
Louis G. Keith MD, PhD

Northwestern University, Chicago, USA
André B. Lalonde MD, FRCSC, FRCOG

Executive Vice-President of the Society of Obstetricians and Gynaecologists of Canada
Mahantesh Karoshi MBBS, MD

Epsom and St. Helier University Hospitals NHS Trust, UK
With a Special Message from HRH The Princess Royal

and a Foreword by Arnaldo Acosta, President of the International Federation
of Gynecology & Obstetrics, FIGO, and André B. Lalonde, Co-Chair of
FIGO Save The Mother and Newborn Health Committee
All the authors and editors have contributed to this publication on a strictly charitable basis
and have received no remuneration of any kind for their contributions.
This book, and other elements in the

programme, are being produced and
published on a not-for-profit, charitable
basis by the proprietors of Sapiens
Publishing in loving and grateful memory
of their daughter ABIGAIL BLOOMER,
who worked alongside them in medical
publishing for a number of years and who
sadly died at the early age of 31 from
breast cancer in December 2001. Abigail
was especially involved in publishing on
women’s health issues. She is greatly missed.
Abigail Bloomer, 1970 –2001
The book, which is available through the A CD-ROM of the book, designed as a
normal commercial channels in the Western resource for lecturers and teachers, is being
World, is being provided free to a large produced. In addition, an aide-mémoire
number of selected physicians in developing poster on surgical techniques and also
countries and at a special low price to practical Guidelines for Immediate Action
members of all national obstetric and leaflet /wallchart for midwives and clinical
gynecological societies worldwide. The whole assistants are both being published. All
book is also available entirely free of charge these items are being made available free of
on the internet, via the Publishers’ website, charge, on a selective basis. Please contact
where it may be read or downloaded chapter the Publishers at the address below for
by chapter by anyone at any time. further information.
Published by
Sapiens Publishing, Duncow, Kirkmahoe, Dumfriesshire DG1 1TA, UK
email: info@sapienspublishing.com
www.sapienspublishing.com
Copyright © 2006 Sapiens Publishing | First published October 2006
Although protected by copyright, this book or selected parts of it may be freely copied for educational or charitable purposes.
However, neither the whole book nor any part of it may be copied for any kind of commercial purpose.
CIP data or a catalogue record for this book is available from the British Library.
ISBN: 978-0-9552282-3-0
Typeset by AMA DataSet Limited, Preston, Lancashire, UK

Printed and bound by Tien Wah Press Pte. Ltd., Singapore
Contents
Special Message from HRH The Princess Royal iii

The Editors xi
The Contributors xiii
Foreword xxiii
A. Acosta and A. B. Lalonde
Editors’ Foreword xxv
C. B-Lynch, L. G. Keith, A. B. Lalonde and M. Karoshi
Maternal mortality in the developing world – and the special challenge in Africa xxvii
L. G. Sambo
Section I: Demographic and logistical considerations

1 Postpartum hemorrhage today: living in the shadow of the Taj Mahal 2
A. B. Lalonde, B.-A. Daviss, A. Acosta and K. Herschderfer
2 Definitions and classifications 11
A. Coker and R. Oliver
3 Vital statistics: an overview 17
M. J. Cameron and S. C. Robson
4 Pitfalls in assessing blood loss and decision to transfer 35
B. S. Kodkany and R. J. Derman
5 Assessing and replenishing lost volume 45
J. G. L. Cockings and C. S. Waldmann
6 Blood loss: accuracy of visual estimation 55
A. Patel, R. Walia and D. Patel
Section II: Causation

7 Doppler evaluation of hemodynamic changes in uterine blood flow 58
G. Urban, P. Tortoli, S. Ricci, M. Paidas, P. Vergani and P. Patrizio
8 Pathophysiology of postpartum hemorrhage and third stage of labor 62
R.-U. Khan and H. El-Refaey
vii
1
Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #T.vp
04 September 2006 14:10:32
POSTPARTUM HEMORRHAGE
9 Obstetric trauma 70
D. G. Evans and C. B-Lynch
10 Placental abnormalities 80
M. K. Mehasseb and J. C. Konje
Section III: General preventive measures

11 Active management of the third stage of labor 98
W. Prendiville and M. O’Connell
12 Misoprostol: theory and practice 114
M. B. Bellad and S. Goudar
13 Labor ward drills 127
M. Tipples and S. Paterson Brown
14 Non-pneumatic anti-shock garment 136
S. Miller and P. Hensleigh
15 Special circumstances: Jehovah’s Witnesses, those who refuse blood transfusion 147
and/or consent
J. C. W. Marsh, M. O. Elebute and D. H. Bevan
Section IV: Special preventive measures: misoprostol in action

16 Misoprostol in practice 156
M. Potts
17 Management of postpartum hemorrhage at the community level 158
N. Prata
18 Oral misoprostol for prevention of postpartum hemorrhage by paramedical 160
workers in India (an ICMR Task Force Study)
N. Chandhiok
19 Overview of misoprostol studies in postpartum hemorrhage 162
A. Hemmerling
Section V: Hospital preparation

20 Resuscitation 170
M. J. Cowen
21 Equipment tray for postpartum hemorrhage 179
T. F. Baskett
22 Building hospital systems for managing major obstetric hemorrhage 183
D. W. Skupski, G. S. Eglinton, I. P. Lowenwirt and F. I. Weinbaum
Section VI: Therapy for non-atonic conditions

23 Bleeding from the lower genital tract 194
A. Duncan and C. von Widekind
24 Adherent placenta: new management options 203
G. Kayem, T. Schmitz, V. Tsatsaris, F. Goffinet and D. Cabrol
viii
2
04 September 2006 14:10:32
Contents
25 Acquired and congenital hemostatic disorders in pregnancy and the puerperium 209
R. V. Ganchev and C. A. Ludlam
26 The use of recombinant factor VIIa 233
S. Sobieszczyk and G. H. Breborowicz
Section VII: Therapy for atony

27 Standard medical therapy 256
F. Breathnach and M. Geary
28 Internal uterine tamponade 263
D. Danso and P. W. Reginald
29 The balloon internal uterine tamponade as a diagnostic test 268
S. Ferrazzani, L. Guariglia and C. Dell’Aquila
30 Embolization 277
K. Choji and T. Shimizu
31 Conservative surgical management 287
C. B-Lynch
32 Internal iliac (hypogastric) artery ligation 299
C. B-Lynch, L. G. Keith and W. B. Campbell
33 The pelvic pressure pack 308
G. A. Dildy III
34 Peripartum hysterectomy 312
T. F. Baskett
35 The management of secondary postpartum hemorrhage 316
K. M. Groom and T. Z. Jacobson
Section VIII: Consequences of postpartum hemorrhage

36 Pathology of the uterus 326
P. Kelehan and E. E. Mooney
37 Severe acute maternal morbidity 339
A. Vais and S. Bewley
38 Sheehan’s syndrome 353
L. Haddock
39 An institutional experience 362
C. T. Wohlmuth, J. Gumbs and J. Quebral-Ivie
40 Familial consequences 372
V. Walvekar and A. Virkud
41 Litigation: an international perspective 376
K. J. Dalton
Section IX: Special experiences and unusual circumstances

42 The obstetrician confronts postpartum hemorrhage 390
M. E. Setchell
ix
3
04 September 2006 14:10:32
43 The midwife confronts postpartum hemorrhage 396

A. M. Ward
44 Sepsis and postpartum hemorrhage 404
B. Das and S. Clark
45 The single-unit transfusion in the bled-out obstetric patient 408
V. Nama, M. Karoshi, M. Wac, L. G. Keith and S. A. Mujeeb
46 Out-of-hospital deliveries 413
E. Sheiner, I. Ohel and A. Hadar
47 Intraoperative autologous blood transfusion 421
S. Catling and D. Thomas
48 Treating hemorrhage from secondary abdominal pregnancy: then and now 427
N. A. Dastur, A. E. Dastur and P. D. Tank
Section X: National experiences

49 Combating postpartum hemorrhage in India: moving forward 434
D. S. Shah, H. Divakar and T. Meghal
50 Eliminating mortality: lessons from Lublin Province in Poland 442
J. Oleszczuk, B. Leszczynska-Gorzelak, D. Szymula, M. Grzechnik, G. Pietras,
J. Bartosiewicz and J. J. Oleszczuk
51 Postpartum hemorrhage in Iran 447
P. Rezai, A. Beigi and A. Jamal
52 Postpartum hemorrhage and maternal mortality in Nigeria 451
I. A. O. Ujah and I. S. Ejeh
53 Postpartum hemorrhage in Asian countries: available data and interpretation 453
S. J. Duthie
Index 462
4
04 September 2006 14:10:32
The Editors
Christopher B-Lynch, MA(Oxon), MBBS, Louis G. Keith, MD, PhD, FACOG, FACS,
LRCP, MRCS, FRCS, FRCOG, MAE, FICS, FICOG (hon)
MCIArb QDR, D.Univ Professor Emeritus of Obstetrics and
Consultant Obstetrician & Gynaecological Gynecology
Surgeon Former Head, Section of Undergraduate
Professor (visiting) Cranfield University Education and Medical Student Affairs
(Health Faculty) The Feinberg School of Medicine
Bedfordshire, UK Northwestern University
Milton Keynes General Hospital NHS 680 Lake Shore Drive, Suite 1015
Foundation Trust (Oxford Deanery) Chicago
Standing Way, Eaglestone Illinois 60611, USA
Milton Keynes e-mail: lgk395@northwestern.edu
Bucks MK6 5LD, UK
e-mail: christopherbl@aol.com
André B. Lalonde, MD, FRCSC, FRCOG, Mahantesh Karoshi, MBBS, MD

FSOGC, FACS, MSc Specialist Registrar
Co-Chair, FIGO Save the Mother and Epsom and St. Helier University Hospitals
Newborn Health Committee NHS Trust
Executive Vice-President, The Society of Epsom
Obstetricians and Gynaecologists of Canada Surrey, KT18 7EG, UK
780 Echo Drive e-mail: m.karoshi@btinternet.com
Ottawa ON K1S 5R7, Canada
e-mail: alalonde@sogc.com
xi
5
Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #V.vp
12 September 2006 11:17:42
6
04 September 2006 14:10:32
The Contributors
A. Acosta, MD D. H. Bevan, FRCP, FRCPath

President, FIGO Department of Haematology
Eligio Ayala, 1263-7e Piso St. George’s Hospital
Asunción Cranmer Terrace
Paraguay London SW17 ORE
e-mail: aaacosta@telesurf.com.py UK
J. Bartosiewicz, MD, PhD
Department of Obstetrics and Perinatology S. Bewley, MD, MA, FRCOG
Medical University in Lublin Guy’s & St. Thomas’ NHS Foundation
20-856 Lublin Trust
Poland Women’s Services
10th Floor North Wing
T. F. Baskett, MB, FRCS(C), FRCS (Ed),
St. Thomas’ Hospital
FRCOG
Lambeth Palace Road
Professor, Department of Obstetrics and
London SE1 7EH
Gynaecology
UK
Dalhousie University
e-mail: susan.bewley@gstt.nhs.uk
Halifax, Nova Scotia
B3H 6R8 Canada
e-mail: tbaskett@ns.sympatico.ca C. B-Lynch, MA(Oxon), MBBS, LRCP,
MRCS, FRCS, FRCOG, MAE, MCIArb
A. Beigi, MD
QDR, D.Univ
Associate Professor of Obstetrics & Gynecology
Consultant Obstetrician & Gynaecological
Arash Hospital
Surgeon
East 196 Street, 3rd Tehranpars Square
Professor (visiting) Cranfield University
Tehran
(Health Faculty)
Iran
Bedfordshire
e-mail: beigi-a@yahoo.com
and
M. B. Bellad, MD
Associate Professor of Obstetrics and Milton Keynes General Hospital NHS
Gynecology & Data Coordinator Foundation Trust (Oxford Deanery)
Global Network for Women’s and Children’s Standing Way
Health Research Eaglestone
J. N. Medical College, Belgaum Milton Keynes
590 010 Karnataka Bucks MK6 5LD
India UK
e-mail: mbbellad@jnmc.edu e-mail: christopherbl@aol.com
xiii
F. Breathnach, MRCOG, MRCPI, K. Choji

DipGUMed, DCH Consultant in Interventional Radiology
Clinical Lecturer and Research Fellow Milton Keynes General Hospital NHS
Royal College of Surgeons of Ireland Foundation Trust (Oxford Deanery)
Rotunda Hospital, Parnell Square Standing Way
Dublin 1 Eaglestone
Ireland Milton Keynes
e-mail: fbreathnach@eircom.net Bucks MK6 5LD
UK
G. H. Breborowicz, MD, PhD
e-mail: drchoji@hotmail.com
Director of Perinatology and Gynecology
University of Medical Sciences Poznañ S. Clark, MSc, MD, MRCP, MRCPath,
ul. Polna 33 DTMH
60-535 Poznañ Consultant Microbiologist
Poland Milton Keynes General Hospital NHS
e-mail: gbrebor@gpsk.am.poznan.pl Foundation Trust (Oxford Deanery)
Standing Way
D. Cabrol, MD, PhD
Eaglestone
Department of Obstetrics and Gynecology
Milton Keynes
Port Royal Maternity
Bucks MK5 6LD
123 Boulevard de Port-Royal
UK
75014 Paris
e-mail: sheila.clark@mkgeneral.nhs.uk
France
e-mail: dominique.cabrol@cch.ap-hop-paris.fr J. G. L. Cockings, MBBS, FRCA, FANZCA,
FJFICM
M. J. Cameron, BMSc, MB, ChB, MRCOG
Department of Intensive Care Medicine
Women’s Services, Leazes Wing
Royal Berkshire Hospital
Royal Victoria Infirmary
London Road
Newcastle upon Tyne
Reading
UK
Berkshire RG1 5AN
e-mail: martin.cameron@nuth.nhs.uk
UK
W. B. Campbell, MS, FRCP, FRCS e-mail: jerome.cockings@royalberkshire.nhs.uk
Consultant Vascular Surgeon
A. Coker, MBBS, MRCOG
Royal Devon and Exeter Hospital and
Consultant Obstetrician & Gynaecologist
Peninsula Medical School
Harold Wood Hospital
Exeter EX2 5DW
Romford
UK
Essex
e-mail: bruce.campbell@nice.org.uk
UK
S. Catling, BA(Cantab), FRCA e-mail: yemicoker@aol.com
Department of Anaesthesia and Intensive Care
M. J. Cowen
Morriston and Singleton Hospitals
Consultant Anaesthetist
Swansea NHS Trust, Swansea
Milton Keynes General Hospital NHS
UK
Foundation Trust (Oxford Deanery)
N. Chandhiok, DGO, FICMCH Standing Way
Indian Council of Medical Research Eaglestone
Ansari Nagar Milton Keynes
New Delhi 110029 Bucks MK5 6LD
India UK
e-mail: n_chandhiok@hotmail.com e-mail: mike.cowen@mkgeneral.nhs.uk
xiv
The Contributors
K. J. Dalton, LLM, PhD, FRCOG, FCLM, C. Dell’Aquila

DFMS Department of Obstetrics and Gynecology
Department of Obstetrics and Gynaecology Catholic University of Sacred Heart
University of Cambridge Rome
Addenbrooke’s Hospital (Box 223) Italy
Cambridge CB2 2QQ
R. J. Derman, MD, MPH
UK
Principal Investigator (USA)
e-mail: kjd5@cam.ac.uk
JNMC–UMKC Women’s and Children’s
D. Danso, MBBS, MRCOG Research Unit
Milton Keynes General Hospital NHS Global Network for Women’s and Children’s
Foundation Trust (Oxford Deanery) Health Research
Standing Way, Eaglestone University of Missouri at Kansas City, School
Milton Keynes of Medicine
Bucks MK6 5LD 2411, Holmes Street
UK Kansas City
e-mail:d_danso@yahoo.com Missouri 64108-2792
USA
B. Das
e-mail: dermanr@umkc.edu
Milton Keynes General Hospital NHS
Foundation Trust (Oxford Deanery) G. A. Dildy III, MD
Standing Way, Eaglestone Department of Obstetrics & Gynecology
Milton Keynes Louisiana State University School of Medicine
Bucks MK5 6DD New Orleans, Louisiana
UK and
2 Ashley Court, Park City
A. E. Dastur, MD, FICOG, FCPS, DGO,
Utah 84060
DFP, FICMU, ATMF(USA)
USA
Honorary Professor Emeritus and Dean
e-mail: Gary.Dildy@HCAhealthcare.com
Seth G S Medical College and Nowrosjee
Wadia Maternity Hospital H. Divakar, MD, FICOG, DGO, FICMCH,
Mumbai PGDMLE
India Consultant Obstetrician & Gynecologist
e-mail: naynadastur@gmail.com Divakars Speciality Hospital
220, 9th cross 2nd phase JP nagar
N. A. Dastur, MD, FICOG, DGO, DFP
Bangalore 560078
Honorary Obstetrician and Gynecologist
India
Grant Medical College
e-mail: hemadivakar246@hotmail.com
Seth G S Medical College and Nowrosjee
Wadia Maternity Hospital A. Duncan, MBBS, BSc, FRCS, FRCSE,
Mumbai FRCOG
India Consultant Gynaecological Surgeon
e-mail: aedastur@bom3.vsnl.net.in Northampton General Hospital
Northampton NN1 5BD
B.-A. Daviss, RM, BJ, MA
UK
Project Manager for FIGO Safe Motherhood
e-mail: alistair.duncan@ngh.nhs.uk
and Newborn Health
and FIGO/ICM Postpartum Hemorrhage S. J. Duthie, FRCOG
Initiative Consultant Obstetrician and Gynaecologist
36 Glen Avenue Blackpool Victoria Hospital
Ottawa ON K1S 2Z7 Blackpool FY3 8NR
Canada UK
e-mail: betty-anne@rogers.com e-mail: sjd757@aol.com
xv
G. S. Eglinton, MD R. V. Ganchev, MB, ChB

Associate Professor, Obstetrics and Specialist Registrar in Haematology
Gynecology Department of Clinical and Laboratory
Weill Medical College of Cornell University Haematology
Chairman, Department of Obstetrics and Royal Infirmary of Edinburgh
Gynecology 51 Little France Crescent
New York Hospital Medical Center of Edinburgh EH16 4SA
Queens UK
Flushing e-mail: ruslan.ganchev@luht.scot.nhs.uk
New York
M. Geary, MD, MRCOG, FRCPI, DCH
USA
Rotunda Hospital, Parnell Square
I. S. Ejeh, FRCOG Dublin 1
81 Division Hospital Nigerian Army Ireland
Ikoyi e-mail: mgeary@rotunda.ie
Lagos
F. Goffinet, MD, PhD
Nigeria
M. O. Elebute, MD, MRCP, FRCPath Port Royal Maternity
Department of Haematology 123 Boulevard de Port-Royal
St. George’s Hospital 75014 Paris
Cranmer Terrace France
London SW17 ORE
S. Goudar, MD, MHPE
and Professor of Physiology & Research
Coordinator
National Blood Service
Global Network for Women’s and Children’s
Tooting
Health Research
London
J. N. Medical College, Belgaum
UK
590010 Karnataka
H. El-Refaey, MD, MRCOG India
Consultant and Senior Lecturer e-mail: sgoudar@jnmc.edu
Imperial College School of Medicine
K. M. Groom, MB, BS, BSc
Department of Academic Obstetrics and
Registrar in Obstetrics and Gynaecology
Gynaecology
University of Auckland
Chelsea and Westminster Hospital
New Zealand
369 Fulham Road
London SW11 9NH M. Grzechnik, MD
UK Department of Obstetrics and Perinatology
e-mail: hazem@mac.com Medical University in Lublin, Jaczewskiego 8
20-856 Lublin
D. G. Evans, MBE, MB.Bch(Wales), FRCS,
Poland
FRCOG
Consultant Emeritus L. Guariglia
St. Bartholemew’s Hospital, Department of Obstetrics and Gynecology
University of London Catholic University of Sacred Heart
London Rome
UK Italy
e-mail: evans@cranley115.fsnet.co.uk
J. Gumbs, MD
S. Ferrazzani White Memorial Medical Center
Department of Obstetrics and Gynecology Department of Obstetrics and Gynecology
Catholic University of Sacred Heart 1720 Cesar Chavez Avenue
Rome Los Angeles, CA 90033
Italy USA
xvi
The Contributors
A. Hadar, MD A. Jamal, MD
Department of Obstetrics & Gynecology Associate Professor of Obstetrics & Gynecology
Soroka University Medical Center Shariati Hospital
Faculty of Health Science North Kargar Avenue
Ben Gurion University of the Negev Tehran
Beer-Sheva Iran
Israel e-mail: af-jamal@yahoo.com
L. Haddock, MD, MACP, FACE M. Karoshi, MBBS, MD
Professor Emeritus Specialist Registrar
Department of Medicine Epsom and St. Helier University Hospitals
Section of Endocrinology and Metabolism NHS Trust
University of Puerto Rico School of Medicine Epsom
GPO Box 5067 Surrey KT18 7EG
San Juan UK
Puerto Rico 00936-5067 e-mail: m.karoshi@btinternet.com
e-mail: haddoc8@attglobal.net
G. Kayem, MD
A. Hemmerling, MD, MPH Department of Obstetrics and Gynecology
Bixby Fellow of Population and International Port Royal Maternity
Health 123 Boulevard de Port-Royal
University of California, Berkeley 75014 Paris
School of Public Health France
513 Warren Hall
L. G. Keith, MD, PhD, FACOG, FACS, FICS,
Berkeley
FICOG (hon)
CA 94720-6390
Professor Emeritus of Obstetrics and
USA
Gynecology
e-mail: anke@berkeley.edu
Former Head, Section of Undergraduate
P. Hensleigh, MD, PhD Education and Medical Student Affairs
Professor Emeritus Stanford University The Feinberg School of Medicine
Department of Obstetrics and Gynecology Northwestern University
810 Allardice Way 680 Lake Shore Drive, Suite 1015
Stanford Chicago
CA 94305 Illinois 60611
USA USA
e-mail: paulhens@stanford.edu e-mail: lgk395@northwestern.edu
K. Herschderfer, RM P. Kelehan, FRCPath
Secretary General, International Confederation Consultant Histopathologist
of Midwives Department of Pathology & Laboratory
Eisenhowerlaan 138 Medicine
2517 KN The Hague National Maternity Hospital, Holles St
The Netherlands Dublin 2
e-mail: k.herschderfer@internationalmidwives. Ireland
org e-mail: pkelehan@nmh.ie
T. Z. Jacobson, MA, MRCOG, FRANZCOG R.-U. Khan, MRCOG
Consultant and Senior Lecturer in Obstetrics Imperial College School of Medicine
and Gynaecology Department of Academic Obstetrics and
Middlemore Hospital Gynaecology
Private Bag 93311 Chelsea and Westminster Hospital
Auckland 369 Fulham Road
New Zealand London SW11 9NH
e-mail: tal.jacobson@virgin.net UK
xvii
B. S. Kodkany, MD, DGO, FICS, FICOG C. A. Ludlam, PhD, FRCP, FRCPath

Senior Foreign Investigator (India) Professor of Haematology and Coagulation
JNMC–UMKC Women’s and Children’s Medicine
Research Unit Director of Haemophilia and Thrombosis
Global Network for Women’s and Children’s Centre
Health Research Royal Infirmary of Edinburgh
Chief Investigator, Human Reproduction 51 Little France Crescent
Research Collaborating Centre-215 Edinburgh EH16 4SA
Indian Council of Medical Research (New UK
Delhi) & Jawaharlal Nehru Medical College e-mail: christopher.ludlam@ed.ac.uk
Belgaum
J. C. W. Marsh, MD, FRCP, FRCPath
590010 Karnataka
Department of Haematology
India
St. George’s Hospital, Cranmer Terrace
e-mail: drkodkany@jnmc.edu
London SW17 ORE
J. C. Konje, MD, FMCOG(Nig), MRCOG UK
Professor of Obstetrics and Gynaecology e-mail: jmarsh@sgul.ac.uk
Reproductive Sciences Section T. Meghal, MD, DNB, DGO
Department of Cancer Studies and Molecular Junior Consultant
Medicine Divakars Speciality Hospital
University of Leicester Bangalore
Robert Kilpatrick Clinical Sciences Building India
Leicester Royal Infirmary
Leicester LE2 7LX M. K. Mehasseb, MSc, MRCOG,
UK MD(Egypt)
e-mail: jck4@le.ac.uk Clinical Research Fellow
Reproductive Sciences Section
A. B. Lalonde, MD, FRCSC, FRCOG, Department of Cancer Studies and Molecular
FSOGC, FACS, MSc Medicine
Co-Chair, FIGO Save the Mother and University of Leicester
Newborn Health Committee Robert Kilpatrick Clinical Sciences Building
Executive Vice-President, SOGC Leicester Royal Infirmary
780 Echo Drive Leicester LE2 7LX
Ottawa ON K1S 5R7 UK
Canada
e-mail: alalonde@sogc.com S. Miller, CNM, PhD
Director, Safe Motherhood Programs
B. Leszczynska-Gorzelak, MD, PhD Women’s Global Health Imperative
Department of Obstetrics and Perinatology Department of Obstetrics, Gynecology, and
Medical University in Lublin Reproductive Sciences
Jaczewskiego 8 University of California, San Francisco
20-856 Lublin 402 San Francisco Blvd.
Poland San Anselmo, CA 94960
USA
I. P. Lowenwirt, MD
e-mail: smiller@hotmail.com
Clinical Assistant Professor of
Anesthesiology E. E. Mooney, FRCPath
Weill Medical College of Cornell University Consultant Histopathologist
Associate Chairman and Director of Department of Pathology & Laboratory
Obstetric Anesthesiology Medicine
New York Hospital Medical Center of Queens National Maternity Hospital, Holles St
Flushing Dublin 2
New York Ireland
USA e-mail: emooney@nmh.ie
xviii
The Contributors
S. A. Mujeeb, MBBS, MPhil M. Paidas, MD

Associate Professor, Blood Bank Center for Thrombosis and Hemostasis in
Jinnah Postgraduate Medical Center Women’s Health
Karachi 75510 Yale University, New Haven
Pakistan CT 06511-6110
e-mail: smujeeb@super.net.pk USA
V. Nama, MD, MRCOG A. Patel, MD
Clinical Research Fellow and Jr Lecturer Director of Reproductive Medicine
St. George’s Medical School John H. Stroger Jr Hospital of Cook County
London 1900 W. Polk Street, 5th Floor
UK Chicago
e-mail: vnama@sgul.ac.uk Illinois 60612
USA
M. O’Connell
e-mail: ashleshapatel16@yahoo.com
Senior Lecturer
University College Dublin D. Patel, MD
Coombe Hospital Pritzker School of Medicine
Dublin 8 Chicago
Ireland Illinois
USA
I. Ohel, MD
Department of Obstetrics & Gynecology S. Paterson Brown, FRCS, FRCOG
Soroka University Medical Center Consultant Obstetrician and Gynaecologist
Faculty of Health Science Queen Charlotte’s and Chelsea Hospital
Ben Gurion University of the Negev London W12 OHS
Beer-Sheva UK
Israel e-mail: spaterson-brown@hhnt.nhs.uk
J. Oleszczuk, MD, PhD P. Patrizio, MD, MBE
Department of Obstetrics and Perinatology Director Yale Fertility Center and Medical
Medical University in Lublin Practice Director
Jaczewskiego 8 Reproductive Endocrinology and Infertility
20-856 Lublin Yale University, 150 Sargent Drive
Poland New Haven
e-mail: henryka_oleszczuk@yahoo.com CT 06511-6110
USA
J. J. Oleszczuk, MD, PhD
e-mail: pasquale.patrizio@yale.edu
Department of Obstetrics and Perinatology
Medical University in Lublin G. Pietras, MD, PhD
Jaczewskiego 8 Department of Obstetrics and Perinatology
20-856 Lublin Medical University in Lublin
Poland 20-856 Lublin
Poland
and
M. Potts, MB, BChir, PhD, FRCOG
Center for Study of Multiple Birth
Bixby Professor
Chicago
University of California, Berkeley
USA
School of Public Health
R. Oliver, MBBS, MRCOG 518 Warren Hall
Specialist Registrar Berkeley
Whipps Cross Hospital CA 94720-7360
Leytonstone USA
UK e-mail: potts@berkeley.edu
xix
N. Prata, MD, MSC S. C. Robson, MD, MRCOG

Assistant Adjunct Professor School of Surgical & Reproductive Sciences
University of California, Berkeley (Obstetrics & Gynaecology)
School of Public Health 3rd Floor, Leech Building, Medical School
314 Warren Hall University of Newcastle upon Tyne
Berkeley Newcastle upon Tyne NE2 4HH
CA 94720-6390 UK
USA e-mail: s.c.robson@ncl.ac.uk
e-mail: ndola@berkeley.edu
L. G. Sambo, MD
W. Prendiville, MAO, FRCOG Specialist in Public Health
Associate Professor of Obstetrics & Regional Director
Gynaecology WHO Regional Office for Africa
Royal College of Surgeons in Ireland Cite du Djoue
Coombe Hospital PO Box 06
Dublin 8 Brazzaville
Ireland Congo
e-mail: wprendiville@rcsi.ie email: sambol@afro.who.int
J. Quebral-Ivie, MD T. Schmitz, MD
White Memorial Medical Center Department of Obstetrics and Gynecology
Department of Obstetrics and Gynecology Port Royal Maternity
1720 Cesar Chavez Avenue 123 Boulevard de Port-Royal
Los Angeles 75014 Paris
CA 90033 France
USA M. E. Setchell, CVO, MA, FRCSEng,
P. W. Reginald, MD(London), FRCOG FRCOG
Heatherwood & Wexham Park NHS Trust Consultant Obstetrician & Gynaecologist
Slough The Whittington Hospital
Berkshire London N19
UK UK
e-mail: rajreg@aol.com Surgeon-Gynaecologist to HM The Queen
e-mail: m.setchell@thelondonclinic.co.uk
P. Rezai, MD
3946 Lizette Lane D. S. Shah, MD, FCPS, FICS, FICOG,
Glenview DGO, DFP, FICMCH
IL 60025 Program Director, Melbourne IVF, Mumbai
USA Honorary Professor & Consultant Obstetrician
e-mail: pedramrezai@hotmail.com & Gynecologist
Breach Candy Hospital, Jaslok Hospital and
S. Ricci, PhD Research Centre
Researcher, University of Florence Sir HN Hospital and Research Centre
Department of Electronics and Kwality House
Telecommunications Above Chinese Room Restaurant
Via Santa Marta 3 Kemps Corner
Florence Mumbai 400026
Italy India
e-mail: stefano.ricci@unifi.it e-mail: durushah@hotmail.com
xx
The Contributors
E. Sheiner, MD P. D. Tank, MD, DNBE, FCPS, DGO, DFP,

Department of Obstetrics and Gynecology MNAMS, MRCOG(UK)
Soroka University Medical Center Honorary Clinical Associate
PO Box 151 Seth G S Medical College and Nowrosjee
Beer-Sheva Wadia Maternity Hospital
Israel Mumbai
e-mail: sheiner@bgu.ac.il India
e-mail: pariktank@fastmail.co.uk
T. Shimizu
Department of Radiology D. Thomas, MBChB, FRCA
Hokkaido University Hospital Department of Anaesthesia and Intensive Care
N14 W5 Kita-ku Morriston and Singleton Hospitals
Sapporo Swansea NHS Trust
Japan UK
e-mail: tshim@radi.med.hokudai.ac.jp e-mail: dafydd.thomas@swansea-tr.wales.nhs.
uk
D. W. Skupski, MD
Associate Professor, Obstetrics and M. Tipples, FRCSEd, MRCOG
Gynecology Senior Registrar
Weill Medical College of Cornell Queen Charlotte’s and Chelsea Hospital
University Du Cane Road
Associate Chairman and Director of London W12 OHS
Maternal Fetal Medicine UK
P. Tortoli, PhD
New York Hospital Medical Center of
Department of Electronics and
Queens
Telecommunications
56–45 Main Street
Via Santa Marta 3
Room M-372
Florence
Flushing
Italy
NY 11355
USA V. Tsatsaris, MD, PhD
e-mail:dwskupsk@med.cornell.edu Department of Obstetrics and Gynecology
Port Royal Maternity
S. Sobieszczyk, MD, PhD
123 Boulevard de Port-Royal
Internal Medicine Consultant
75014 Paris
Consultant in Anesthesiology and Intensive
France
Care
Laboratory Director of Cardiovascular and I. A. O. Ujah, FMCOG, FICS, MNI
Hemostasis Disorders Department of Obstetrics & Gynecology
Department of Perinatology and University of Jos
Gynecology Nigeria
University of Medical Science e-mail: iaoujah@yahoo.com
ul. Polna 33
G. Urban, MD
60-535 Poznañ
Post-Doctoral Fellow, Yale University School
Poland
of Medicine
D. Szymula, MD PhD Department of Obstetrics & Gynecology
Department of Obstetrics and Perinatology 333 Cedar St. LCI Building, Room 808A
Medical University in Lublin New Haven
Jaczewskiego 8 CT 06512
20-856 Lublin USA
Poland e-mail: gabriele.urban@yale.edu
xxi
A. Vais, MB, BS, BSc C. S. Waldmann, MA, MB, BChir, FRCA,

Guy’s & St. Thomas’ NHS Foundation Trust EDIC
Women’s Services, 10th Floor North Wing Department of Intensive Care Medicine
St. Thomas’ Hospital Royal Berkshire Hospital, London Road
Lambeth Palace Road Reading
London SE1 7EH Berkshire RG1 5AN
UK UK
e-mail: cswald@aol.com
P. Vergani
Department of Obstetrics and Gynecology V. Walvekar, MD, DGO, FCPS, FICOG
University of Milan 62, Suraiya Apartments
Monza Pochkhanwala Road, Worli
Italy Mumbai 400025
A. Virkud, MD, DGO, FCPS, FICOG India
37B, Shalan Building e-mail: vandanawalvekar@hotmail.com
Gamdevi Road A. M. Ward, RM, RN
Mumbai 400007 Master of Midwifery
India 32 Moorcroft
e-mail: ajitvirkud@hotmail.com Dewsbury
C. von Widekind, MD MRCOG West Yorkshire WF13 4ER
Consultant Obstetrician & Gynaecologist UK
Northampton General Hospital e-mail: Anne_Maria.Ward@leedsth.nhs.uk
Northampton NN1 5BD F. I. Weinbaum, MD
UK Senior Vice President for Medical Affairs and
e-mail: clemens.vonwidekind@ngh.nhs.uk Quality Improvement
M. Wac, MD New York Hospital Medical Center of Queens
Resident Physician Flushing
St. Joseph Hospital New York
Chicago USA
Illinois
C. T. Wohlmuth, MD
USA
White Memorial Medical Center
R. Walia Department of Obstetrics and Gynecology
John H. Stroger Jr. Hospital of Cook County 1720 Cesar Chavez Avenue
1900 W. Polk Street, 5th Floor Los Angeles
Chicago CA 90033
Illinois 60612 USA
USA e-mail: Wohlmuct@ah.org
The surgical illustrations (in color) in this book have been prepared by Philip Wilson, Medical
and Scientific Illustrator, Old Coulsdon, Surrey, UK.
xxii
Foreword
This book, launched at the Presidential Symposium at the XVIII World Congress of the
International Federation of Gynecology and Obstetrics (FIGO) in 2006 in Kuala
Lumpur, Malaysia, marks an important day in the battle for improved treatment poss-
ibilities for postpartum hemorrhage. Beginning in October 2003 at the World Congress
in Santiago, Chile, FIGO launched an international effort with other partners and
donors to develop strategies to prevent postpartum hemorrhage and, in the cases where
it still occurred, to identify effective medical and surgical treatments.
There is no controversy about the need for prevention and treatment of postpartum
hemorrhage. Recent evidence from the World Health Organization strongly suggested
that deaths due to postpartum hemorrhage were underestimated and could reach as
high as 40% of all maternal mortality in some African countries as well as South Africa,
South-East Asia and Latin America. Indeed, postpartum hemorrhage is the cause of
close to 50% of maternal mortality in Guatemala and Afghanistan.
In the last 25 years, the world has seen only minimal progress in low-resource
countries to reduce the incidence of postpartum hemorrhage and the resulting maternal
mortality and morbidity. This new book is part of a world-wide effort to prevent
postpartum hemorrhage and offer new perspectives in the medical and surgical treat-
ment options. The first and foremost problem that is addressed is the lack of definition
and the difficulty in assessing blood loss during delivery. Amazing as it may sound,
blood loss is most often underestimated, and the clinical evaluation is very inaccurate.
In the second section of the book, causation is presented, from basic physiology to
obstetric trauma. The third section discusses the entire issue of prevention, with an
excellent description of the active management of the third stage of labor and how this
initiative, undertaken by a number of partners, including the International Confedera-
tion of Midwives (ICM), FIGO and the Prevention of Postpartum Haemorrhage Initia-
tive (POPPHI), is beginning to make a difference in many countries. Misoprostol and
its use for prevention and treatment are outlined and this will bring the discussion to the
forefront on why this low-cost alternative is not more widely used.
Lacerations and trauma following spontaneous instrument delivery are highlighted
as well as the management of adherent placenta. Coagulation disorders and dis-
seminated intravascular coagulation are fully reviewed, with a strong chapter on therapy
with factor VIIa. Therapy for uterine atony is well illustrated, with a proposal for a
postpartum hemorrhage tray in order for midwives and physicians to be prepared for
this emergency. In addition, the B-Lynch brace suture is described by the surgeon who
first demonstrated its utility. Other new procedures, such as intrauterine tamponade
and conservative surgical therapy, are also proposed in detail, with a call for more
clinical research about these possible low-cost-effective therapies.
xxiii
17
04 September 2006 14:10:34
The consequences of postpartum hemorrhage for maternal mortality and morbidity

are presented, along with the experiences of Poland, Iran, India and Africa with post-
partum hemorrhage.
The FIGO Board, at its December 2005 meeting, approved a motion that calls on all
its member countries to delegate to nurses and midwives the use of oxytocin without a
doctor’s prescription. FIGO has also requested its member countries to work with other
health professionals and regulators to ensure that misoprostol is approved for inclusion
on the essential drug list. FIGO further developed a questionnaire to evaluate the use of
oxytocin and misoprostol as well as the management of the third stage of labor.
Postpartum hemorrhage continues to threaten women’s lives in low- and high-
resource countries and is the most important cause of maternal mortality in low-
resource countries. The majority of cases of postpartum hemorrhage are due to uterine
atony and the prevention and treatment are simple, using low-cost technology that
unfortunately is not applied.
ICM and FIGO have joined international agencies, in particular USAID, to train
and implement active management of the third stage of labor as well as propose
low-cost technologies to be used by health-care professionals. FIGO and ICM are
calling upon UN agencies and major donors from high-resource countries to support
the ongoing effort to reduce maternal mortality due to postpartum hemorrhage. As part
of this effort, we heartily endorse this book which is the first-ever compendium of
information on a topic that is of vital interest to the medical professions world-wide.
Arnaldo Acosta, MD André B. Lalonde, MD

President, International Federation of Co-chair FIGO, Save the Mother and
Gynecology and Obstetrics (FIGO) Newborn Health Committee
Executive Vice-President, The Society
of Obstetricians and Gynaecologists
of Canada (SOGC)
xxiv
18
06 September 2006 16:53:01
Editors’ Foreword
The genesis of this book is unusual in that it was not conceived by any Institution,
Organization or Medical College. Rather, it came forth from a perceived need by three
individual physicians working in separate institutions (LK, MK and CBL). They
decided to initiate the effort and act as Editors. The fourth editor (AL) joined later
when he was informed of the project.
Each author who was asked to contribute enthusiastically agreed to do so, often
at very short notice in view of the desire to distribute this volume at the XVIII World
Congress of the International Federation of Gynecology and Obstetrics (FIGO) in
Kuala Lumpur, Malaysia in early November, 2006. Moreover, every author agreed to
write with no remuneration whatsoever, having been informed from the outset that the
entire project was charitable in nature. With this in mind, decisions were eventually
made to make the book available on the internet and to make translation rights by any of
the national member societies of FIGO possible at a nominal charge.
The Editors were extremely fortunate to find a like-minded publisher, David
Bloomer of Sapiens Publishing in the United Kingdom, who undertook this project in
memory of his late daughter, Abigail, who died well before her time of that other
scourge of women, namely breast cancer.
The careful reader will undoubtedly note certain duplications in the text. These
could have been removed if the book were meant to be read seriatim from cover to
cover. However, the Editors are of the opinion that many readers will focus on specific
chapters or series of chapters, at least in the beginning, so the duplications were left to
stand. We apologize for any inconvenience this may cause an individual reader.
As with any dynamic topic, last-minute changes had to be made to accommodate
‘late-breaking news’, foremost of which was that which came to light at the Interna-
tional Congress on the Prevention of Post Partum Hemorrhage, held in Goa, India, July
12–16, 2006. Although the manuscript was already in pages, a new section on Specific
preventive measures was added to reflect evidence which was not known when the
chapters were commissioned. Credit for the incorporation of these changes into the
already completely set book goes to Mrs Jean Wright, a senior and experienced editor
par excellence, who has worked with David Bloomer for many years. Credit and many
heartfelt thanks are also due to Mrs Nora Horner, private secretary to Mr B-Lynch, who
managed all the e-mail traffic between the Editors and authors connected with the
transmission of all manuscripts in electronic format, a formidable task to say the least.
Finally, the Editors join David Bloomer in thanking HRH The Princess Royal for
writing a Special Message for this volume and for speaking at the launch held at
Chandos House, the Royal Society of Medicine, London, on October 11, 2006.
xxv
19
04 September 2006 14:10:34
We sincerely hope that our effort helps all levels of health-care practitioners to work
more effectively to save women’s lives when postpartum hemorrhage occurs. If that
happens but once, then our combined efforts will have been worthwhile.
Christopher B-Lynch Louis G. Keith

Editor-in-Chief Senior Editor
Milton Keynes (Oxford Deanery), UK Chicago, USA
André B. Lalonde Mahantesh Karoshi
Editor Associate Editor
Ottawa, Canada Epsom, UK
September, 2006
xxvi
20
06 September 2006 14:26:25
Maternal mortality in the developing world

– and the special challenge in Africa
The publication of this textbook marks a great milestone in our joint efforts at achieving
the Millennium Development Goal 5, which aims to reduce maternal mortality by
three-quarters, a daunting task indeed for most developing countries. The African
region, especially Eastern and Western Africa, has the highest ratio of women dying as a
result of pregnancy or childbirth in the world, estimated at an average of 1000 per
100 000 live births. Half of all the maternal deaths world-wide occur in the African
Region, a region that accounts for only 12% of the world’s population and about 20%
of births. A lifetime risk of 1 in 3000, as is the case in high-income countries, represents
a low risk of dying from pregnancy and childbirth, while 1 in 100 is considered a
high risk. In Sub-Saharan Africa, for every 16 women, one will die of pregnancy and
childbirth-related conditions. It is obvious that a lifetime risk of 1 in 16 is extremely
high. This is a direct result of the defects in the social, cultural and economic status of
women as well as inadequacies in existing health systems. Available evidence shows that
the causes of maternal death include obstructive labor (8%), hypertension disorders
(12%), abortion (13%), sepsis (15%), indirect causes such as malaria, anemia
and HIV/AIDS (20%) and other direct causes (ectopic pregnancy, embolism and
anesthesia-related problems (7%)). Postpartum hemorrhage accounts for 25% of the
etiology of maternal mortality; however, it accounts for a third or more maternal deaths
in certain countries of the Region. Malaria in pregnancy predisposes women to a
number of complications including anemia, which places them at a higher risk of
mortality from hemorrhage. Only an estimated 46% of the deliveries in the African
Region are assisted by qualified health personnel.
The above situation makes reduction of maternal mortality one of the priorities of the
WHO Regional Office for Africa – and we are making strenuous efforts towards improv-
ing maternal health and reducing maternal mortality.
This book provides evidence-based practical guidance to build the capacity of health
providers in the management of postpartum hemorrhage in both pre-service and in-
service training of health providers. As developing countries gear up to increase the
availability of qualified personnel with the necessary skills to provide safe motherhood
services, drawing on the guidance that this book provides will further urge us to
promote safer practices.
The gap between the mortality rates from postpartum hemorrhage in developed
countries and developing countries underscores the need for effectiveness and
timeliness of the health systems in responding to pregnancy and childbirth-related
complications. There is also an urgent need to find low-cost interventions that can
xxvii
21
04 September 2006 14:10:34
be implemented in resource-poor settings. This text book provides such options.

I encourage practitioners, tutors and trainers to apply this book at all levels of care,
along with other interventions, such as community participation, to increase the timely
and appropriate utilization of health services.
Finally, I would like to assure readers that WHO will continue to work hand-in-hand
with all partners, including the research and academic communities, to promote the
implementation of evidence-based interventions to improve maternal health in the world.
Dr Luis Gomes Sambo

Regional Director
World Health Organisation Regional Office for Africa
xxviii
22
04 September 2006 14:10:34
Section I
Demographic and logistical
considerations
23
30 August 2006 14:19:06
1
POSTPARTUM HEMORRHAGE TODAY: LIVING IN THE
SHADOW OF THE TAJ MAHAL
A. B. Lalonde, B.-A. Daviss, A. Acosta and K. Herschderfer
‘Women are not dying because of a disease we cannot treat. They are dying because societies have yet
to make the decision that their lives are worth saving.’
Mamoud Fathalla, President of the International Federation of Gynecology and Obstetrics (FIGO),
World Congress, Copenhagen 1997
INTRODUCTION
to characterize these deaths is to say that one
The wife of the Shah Jahan of India, the woman dies every minute of every hour of every
Empress Mumtaz, had 14 children and died day.
after her last childbirth of a postpartum hemor- Most of the deaths and disabilities attributed
rhage in 1630. So great was the Shah Jahan’s to childbirth are avoidable, because the medical
love for his wife that he built the world’s most solutions are well known. Indeed, 99% of
beautiful tomb in her memory – the Taj Mahal1. maternal deaths occur in developing countries
Far away and to the north, another country was that have an inadequate transport system, lim-
taking a different approach: in 1663, the Swed- ited access to skilled care-givers, and poor emer-
ish Collegium Medicum was established. The gency obstetric services4. It is axiomatic that
Swedish clergy created an information system each and every mother and newborn require
that by 1749 provided the first national vital care that is close to where they live, respectful
statistics registry in Europe; by 1757, a national of their culture, and provided by persons
training was approved for midwives in all with enough skill to act immediately should
parishes of Sweden. The resulting infrastructure an unpredictable complication occur. The
– a comprehensive community midwifery sys- challenge that remains internationally is not
tem, with physician back-up expertise and an technological but strategic and organizational4.
outcome reporting system – is today considered Postpartum hemorrhage is the most common
responsible for reducing the maternal mortality cause of maternal mortality and accounts for
in Sweden from 900 to 230 per 100 000 live one-quarter of the maternal deaths world-
births in the years between 1751 and 19002. To wide5. The optimal solution for the vast major-
this day, Sweden enjoys the lowest maternal ity, if not all, of these tragedies is prevention,
mortalities in the world. both before the birth, by assuring that women
In 2006, each nation must decide whether it are sufficiently healthy to withstand postpartum
is going to build monuments to hardship and hemorrhage should it occur, and at the time of
suffering or take the steps to avoid it. Although a the birth, by the use of physiological or active
full 10 years remain until the target date of management of labor, a management strategy
2015, it is already predicted that the Millen- that unfortunately is dependent on circum-
nium Development Goal (MDG) number 5 to stances and the availability of oxytocics. To
reduce maternal mortality (MM) by 75% will their credit, the International Confederation of
not be reached. Maternal mortality is currently Midwives (ICM) as well as the International
estimated at 529 000 deaths per year, a number Federation of Gynecology and Obstetrics
that translates into a global ratio of 400 mater- (FIGO) are engaging their membership in a
nal deaths per 100 000 live births3. Another way world-wide campaign to address this travesty.
24
30 August 2006 14:19:06
Postpartum hemorrhage today
DEFINITION AND INCIDENCE POSTPARTUM HEMORRHAGE:

WHEN, WHY AND WHERE
The World Health Organization (WHO) has
examined studies on postpartum hemorrhage Sixty percent of all pregnancy-related maternal
published between 1997 and 2002 in order to deaths occur during the postpartum period and
arrive at more precise definitions of postpartum one source suggests 45% of them occur in the
hemorrhage and its incidence6. Available first 24 h after delivery9.
resources – data from 50 countries, 116 studies The risk of dying from postpartum hemor-
and 155 unique data sets – were reported to be rhage depends not only on the amount and rate
poor in quality. Definitions of postpartum hem- of blood loss but also the health status of the
orrhage were lacking in 58% of the published woman10. Poverty, lifestyle, malnutrition, and
studies and, in the population-based surveys of women’s lack of decision-making power to con-
medium quality, the prevalence ranged from a trol their own reproductive health are some of
low of 0.55% of deliveries in Qatar to a high the broad issues that have unfortunately come
of 17.5% in Honduras. Preliminary findings to be accepted as inevitable and unchangeable.
suggest that excessive bleeding was reported In a busy urban maternity hospital, in the coun-
between 0.84% and 19.80% of the time, but the try where the Taj Mahal acts as a testament to
majority of studies were reported as low in qual- contravention of this problem, nurses in a labor
ity and had problems defining and diagnosing ward may not complete patient case notes for
postpartum hemorrhage. low-caste women, depriving them of the safe-
One of the major problems plaguing the guards of other women3. But India’s problems
research is how to measure postpartum are merely a symbolic representation of a
hemorrhage with accuracy. Published data problem that faces both high- and low-resource
are scant, and an adequate and accurate countries3,4,11. The insidious reality about hav-
gold-standard method is lacking. Clinical visual ing a postpartum hemorrhage is that two-thirds
estimation of blood loss is not reliable7. As of the women who experience it have no identi-
is often the case, necessity becomes the fiable clinical risk factors such as multiple births
mother of invention. In the rural areas of or fibroids12. In this regard, postpartum hemor-
Tanzania, the use of ‘Kanga’ has been rhage is a veritable equal-opportunity occur-
adopted as a valid instrument tool8. Convenient rence. However, it is not an equal-opportunity
because it is produced and sold locally, the killer because it is the poor, malnourished,
pre-cut Kanga is a standard-sized rectangle unhealthy woman who delivers away from
(100 cm × 155 cm) of local cotton fabric. When medical care who will die from it, whereas
three to four soaked Kangas are observed at a those who are fortunate enough to deliver in a
delivery, the trained traditional birth attendant well-supplied and staffed medical facility most
(TBA) is entrusted to transfer patients to a likely will survive three delays at the actual time
health center. of birth: delay in the decision to recognize a
Even when a good measurement methodol- complication and seek help; delay in accessing
ogy is in place, there is still difficulty in defining transportation to reach a medical facility,
postpartum hemorrhage simply as blood loss and, finally, delay in receiving adequate and
greater than 500 ml because it fails to take into comprehensive care upon arrival.
account predisposing health factors that are About 95% of maternal deaths in 2000
reflected in such a definition. Since the quantity were equally distributed between Asia
of blood loss is less often important than the (253 000) and sub-Saharan Africa (251 000)13,
actual effect that it has on the laboring woman, but the risks are higher in Africa because it
it has been suggested that the definition take has a smaller population than Asia. For
into account any blood loss that causes a major decades, sub-Saharan Africa has been the
physiological change, such as low blood pres- region with the highest maternal mortality
sure, which threatens the woman’s life. These ratio in the world, at over 900/100 000 live
issues are discussed in greater detail in Chapters births. In this region, the numbers of births
2–6. attended by skilled health personnel and life
25
30 August 2006 14:19:06
expectancy at birth strongly correlate with EXISTING EVIDENCE FOR

maternal mortality. PREVENTION OF HEMORRHAGE
As an example, the increased ability to
In September 2004, Litch provided a summary
measure maternal mortality in Afghanistan
of the evidence base for the active management
has revealed a heretofore suspected but uncon-
of the third stage of labor17. The following
firmed reality. The Center for Disease Control
excerpt summarizes these data:
and Prevention’s retrospective cohort study of
women of reproductive age in four selected dis- ‘From 1988 to 1998, four large, randomized,
tricts in four provinces reported an astounding controlled studies conducted in well-resourced
maternal mortality of 1900 per 100 000 live maternity hospitals (two in the UK, one in the
births14. Another group of authors, working in United Arab Emirates and one in Ireland)
compared the effects of active and expectant
the same country, describes reasons for such a
management of the third stage of labor. In all
high maternal mortality ratio in the Province of four studies, active management was associated
Herat: with a decrease in postpartum hemorrhage and
‘. . . conditions for individual and community the length of third stage of labor . . . A Cochrane
health often depend on the protection and Library systematic review and meta-analysis also
promotion of human rights. The findings of concluded that active management of the third
this study identify a number of human rights stage in the setting of a maternity hospital was
factors that contribute to preventable maternal superior to expectant management in reducing
deaths in Herat Province. These include access blood loss, incidence of postpartum hemorrhage
to and quality of health services, adequate food, and duration of the third stage. It was also associ-
shelter, and clean water, and denial of individual ated with reduced postpartum anemia, decreased
freedoms such as freely entering into marriage, need for blood transfusion, and less use of
access to birth control methods and possibly additional therapeutic uterotonic drugs’17.
control over the number and spacing of one’s To a certain extent, the same caveat holds for
children’15. the usage of prostaglandins where at least two
In many other countries, hemorrhage accounts Cochrane Reviews have addressed the issue of
for more than half of the maternal deaths, this drug as a choice for use in active manage-
rather than the quarter of maternal mortality ment. A review in 2003 suggests rectal miso-
usually cited world-wide. For example, in prostol 800 µg may be a useful ‘first-line’ drug
Indonesia it has been reported at 43%, in for the treatment of primary postpartum hemor-
the Philippines at 53%, and in Guatemala at rhage, but that further randomized controlled
53%4. trials are required to identify the best drug com-
Within given countries, certain populations binations, route, and dose for the treatment of
are also at increased risk. In Latin America, postpartum hemorrhage. In 2004, a review says
for example, the Pan American Health ‘Neither intramuscular prostaglandins nor miso-
Organization (PAHO) has identified reasons prostol are preferable to conventional injectable
why maternal mortality is higher among the uterotonics as part of the active management of
indigenous populations: the third stage of labor, especially for low-risk
women. Future research on prostaglandin use
(1) The professional teams in charge of mater- after birth should focus on the treatment of
nity care underrate or are ignorant of postpartum hemorrhage rather than prevention
traditional cultural practices; where they seem to be more promising’18. How-
(2) The health team and pregnant women ever, this review should be read in the context
often communicate poorly, a principal that many countries do not have the infra-
factor behind the low maternity coverage; structural elements to provide uterotonics.
Even a WHO multicenter, randomized
(3) Public policies for consensus building and trial left some issues unresolved. This study
intercultural dialogue on maternal health concluded that 10 IU oxytocin (intravenous or
are in conflict over objectives and goals and intramuscular) was preferable to 600 µg oral
the allocation of resources16. misoprostol in the active management of the
26
30 August 2006 14:19:06
third stage of labor in hospital settings where reduced the frequency of severe postpartum
active management was the norm19. The hemorrhage24. Both studies state these promis-
possible troubling ‘secondary effect’ of oxytocin ing results suggest increased safety of deliveries
on manual removal of the placenta needs using misoprostol even when attended by
clarification, however, as a 2004 Cochrane practitioners not considered by the WHO/ICM/
Review suggested that, with prophylactic use of FIGO definition to be ‘skilled’. Further discus-
oxytocin, ‘the risk of manual removal of the sion of ongoing field work with misoprostol is
placenta may be increased’20. In high-resource provided in Section IV.
countries, where embolism rather than post- An even more promising alternative method
partum hemorrhage is the major cause of to deal with postpartum hemorrhage was under-
maternal mortality, hemorrhage requiring taken in Indonesia, where 1811 women were
hysterectomy is considered one of the most offered counselling about the prevention of
life-threatening conditions experienced by postpartum hemorrhage and use of miso-
women during the perinatal period21. Retained prostol by trained and supervised volunteers.
placenta represents a serious complication This study demonstrated that misoprostol was
requiring manual removal and such a ‘second- safely used in a self-directed manner among
ary outcome’ could be as critical to consider study participants who had home deliveries in
when deciding on third-stage management pro- the intervention area25.
tocols. Because the picture is not yet entirely Although misoprostol is available in most
clear, practitioners should continually update countries in Asia and the Americas, there are
themselves as to available options, and health- restrictions to its use in many countries resulting
care agencies and government planning units from the fear that it will be used as an
should be equally vigilant about what is the best abortifacient. There is no access to this agent in
approach considering the available resources. most of Africa and much of the Middle East and
Thus, although the literature suggests that only three countries have approved the obstetric
active management using the standard oxytocics use of it: Brazil, Egypt and France26. Given the
can reduce postpartum hemorrhage by 40%22, potential benefits of misoprostol to the major
this methodology is far from ideal for use in goal of the MDG #5 (maternal mortality), and
low-resource countries where the lethal post- the fact that the WHO has added it to its list of
partum hemorrhages are occurring, and where ‘essential medicines’27, there appears to be a
many births take place away from medical role for FIGO, ICM and the research commu-
facilities and are supervised solely by traditional nity in closing the gaps on research as well as the
birth attendants who do not have access to barriers to availability of this medication.
medications or the right to use them.
The WHO study did not investigate whether
ONGOING INITIATIVES TO PREVENT
misoprostol was better than placebo. Two
recent trials with misoprostol, however, suggest
favorable results for the use of this agent in Every child-bearing woman is potentially at
low-resource countries. One was a field inter- risk for postpartum hemorrhage, but biological/
vention trial in Tanzania after home births that physiological considerations are only a part of
demonstrated that implementing the use of the picture. Broader issues suggest that heath-
1000 µg of rectal misoprostol administered by care workers should assume more of an attitude
TBAs to women with 500 ml or more blood loss of service and responsibility in the larger public
decreased referral and need of further treatment health issues, empowering women to seek help
when compared to a non-intervention group23. because the health-care culture is acceptable to
The second trial was a randomized, double- them. With respect to indigenous populations
blind, placebo-controlled trial that took place and minority groups forgotten or subjugated by
among women attended by midwives at local a dominant culture, more sensitive approaches
health centers in Guinea-Bissau. Here it that respect pregnancy and birth as a social and
was concluded that routine administration of cultural rather than a medical act and incorpo-
600 µg of sublingual misoprostol after delivery rating traditional practitioners, e.g. the ‘partera’
27
30 August 2006 14:19:06
in Central America, into the health-care team, ● Administration of uterotonic agents,

are an important step forward. It is crucial that
● Controlled cord traction, and
physicians, midwives, and nurses work with
communities and women’s groups to bridge ● Uterine massage after delivery of the pla-
existing gaps in care. centa, as appropriate.
An international group including the ICM,
Every attendant at birth needs to have the
FIGO members, researchers and experts met
knowledge, skills and critical judgement
in Ottawa, Canada, in August 2003 to craft the
required to carry out active management of the
Ottawa Statement on prevention of postpartum
third stage of labor and access to appropriate
hemorrhage and offer new options for its treat-
supplies and equipment.
ment. At the last World Congress of FIGO in
Chile in 2003, President Arnaldo Acosta
announced that FIGO, in partnership with How to use uterotonic agents
ICM, would launch an initiative that would
promote active management of the third stage ● Within 1 minute of the delivery of the baby,
of labor (AMTSL) to prevent postpartum hem- palpate the abdomen to rule out the presence
orrhage and increase the knowledge of nurses, of an additional baby(s) and administer
midwives and physicians in the medical and oxytocin 10 units intramuscularly. Oxytocin
surgical treatment of postpartum hemorrhage. is preferred over other uterotonic drugs
Both FIGO and ICM are collaborating with the because it is effective 2–3 minutes after
Program for Appropriate Technology for Health injection, and has minimal side-effects so
(PATH) to conduct a project: Prevention of that it can be used on all women.
Post Partum Hemorrhage Initiative (POPPHI), ● If oxytocin is not available, other uterotonics
launched in October 2004. The program has can be used such as: ergometrine 0.2 mg
created tool kits and educational modules for intramuscularly, syntometrine (1 ampoule)
implementation of the AMTSL. POPPHI is also intramuscularly or misoprostol 400–600 µg
providing small grants to countries for FIGO orally. Oral administration of misoprostol
and ICM members to collaborate on scaling up should be reserved for situations when safe
the use of AMTSL. These initiatives have been administration and/or appropriate storage
prompted in large part by the fact that past conditions for injectable oxytocin and ergot
efforts have not decreased maternal mortality alkaloids are not possible.
and morbidity substantially. Postpartum hem-
orrhage prevention and treatment procedures ● Uterotonics require proper storage:
are well known and are proven to be scientifi- – Ergometrine: 2–8°C and protect from
cally beneficial but not readily available to light and from freezing
health workers and pregnant women. – Misoprostol: room temperature, in a
The following Joint Statement and Action closed container
Plan was launched in 2004 by ICM/FIGO. – Oxytocin: 15–30°C, protect from freezing
● Counselling on the side-effects of these drugs
JOINT ICM/FIGO STATEMENT AND should be given.
ACTION PLAN Warning! Do not give ergometrine or synto-
Management of third-stage labor should be metrine (because it contains ergometrine) to
offered to women since it reduces the incidence women with pre-eclampsia, eclampsia or high
of postpartum hemorrhage due to uterine atony. blood pressure.
Active management of the third stage of
labor consists of interventions designed to facili-
How to perform controlled cord traction
tate the delivery of the placenta by increasing
uterine contractions and to prevent postpartum ● Clamp the cord close to the perineum (once
hemorrhage by averting uterine atony. The pulsation stops in a healthy newborn) and
usual components include: hold in one hand.
28
30 August 2006 14:19:07
● Place the other hand just above the woman’s In all of the above actions, explain the proce-
pubic bone and stabilize the uterus by apply- dures and actions to the woman and her family.
ing counter-pressure during controlled cord Continue to provide support and reassurance
traction. throughout.
● Keep slight tension on the cord and await a
strong uterine contraction (2–3 minutes). IMPORTANT CHANGES TO
● With the strong uterine contraction, encour- CONSIDER IN ACTIVE MANAGEMENT
age the mother to push and very gently PROTOCOLS
pull downward on the cord to deliver the As the evidence suggesting immediate cord
placenta. Continue to apply counter-pressure clamping can reduce the quantity of red blood
to the uterus. cells an infant receives at birth and result in
● If the placenta does not descend during potential short-term and long-term problems,
30–40 seconds of controlled cord traction, do and because prior concerns about polycythemia
not continue to pull on the cord: have not been documented28, the collaborative
– Gently hold the cord and wait until the ICM/FIGO group decided not to include early
uterus is well contracted again; cord clamping in the active management
– With the next contraction, repeat con- protocol. This decision means that the present
trolled cord traction with counter-pressure. definition of active management promulgated
by ICM/FIGO differs from that described in the
Never apply cord traction (pull) without apply- early literature.
ing counter-traction (push) above the pubic FIGO now also advises that, in the absence
bone on a well-contracted uterus. of oxytocin or misoprostol at delivery, skilled
● As the placenta delivers, hold the placenta in birth attendants should use physiologic man-
two hands and gently turn it until the mem- agement of the third stage to avoid over-
branes are twisted. Slowly pull to complete exertion through cord traction until the
the delivery. uterus has contracted and the placenta has
begun being expelled. This is best described as
● If the membranes tear, gently examine the allowing the mother to expel her own placenta
upper vagina and cervix wearing sterile/ without interference from the practitioner.
disinfected gloves and use a sponge forceps
to remove any pieces of membrane that are
present. THE ROLE OF NATIONAL
PROFESSIONAL ORGANIZATIONS
● Look carefully at the placenta to be sure none
of it is missing. If a portion of the maternal The following points outline the ten key action
surface is missing or there are torn mem- imperatives that are being promoted world-wide
branes with vessels, suspect retained placenta by FIGO/ICM to prevent postpartum hemor-
fragments and take appropriate action27. rhage and manage postpartum hemorrhage
when it occurs.
(1) Disseminate and secure support for the
How to perform uterine massage
joint statement from UN agencies, and
● Immediately massage the fundus of the international and national organizations.
uterus until the uterus is contracted.
(2) Recommend that this Global Initiative
● Palpate for a contracted uterus every 15 min- on the Prevention of Postpartum Hemor-
utes and repeat uterine massage as needed rhage be integrated into the curriculum of
during the first 2 hours. medical, midwifery and nursing schools.
● Ensure that the uterus does not become (3) Work toward the goal of offering
relaxed (soft) or ‘boggy’ after you stop uterotonic drugs for prophylactic treat-
uterine massage. ment of postpartum hemorrhage to every
29
30 August 2006 14:19:07
mother giving birth anywhere in the (4) Dealing with the legislative and other barri-
world. ers that impede the prevention and treat-
ment of postpartum hemorrhage, including
(4) Ensure that every skilled attendant at a
dealing with poverty and malnutrition as
birth will have uterotonic drugs and know
well as the incorporation of active manage-
how to administer them.
ment of third stage into pre-service and
(5) Ensure that every hospital birthing unit in-service curricula for all skilled birth
and every birth center will have uterotonic attendants;
drugs and a protocol to prevent and
(5) Incorporation of active management of the
manage postpartum hemorrhage.
third stage of labor in national standards
(6) Give adequate training to every skilled and clinical guidelines, as appropriate;
attendant that attends births (including
(6) Working with national pharmaceutical
doctors, nurses and midwives) in uterine
regulatory agencies, policy-makers and
massage, bimanual compression, and
donors to assure that adequate supplies
manual removal of the placenta.
of uterotonics and injection equipment are
(7) Make the use of new simple medical and available.
surgical therapies available to skilled atten-
dants, including the use of intravenous
CONCLUSION
infusion, tamponade balloons, and shock
pants29 (see Chapters 5, 14, 21 and 28). Tourists flock to the Taj Mahal, largely unaware
how often around the world the event symbol-
(8) Provide every doctor who can perform a
ized by this monument still occurs in the
laparotomy and basic clinical officers who
shadows of a woman’s blood-soaked dirt floor,
are responsible for the surgical manage-
or when a desperate husband’s rough cart is
ment at the peripheral hospital level, with
dragged over poor roads and fails to arrive in
surgical training to perform ‘simple con-
time, or in the sad eyes of a basic health-unit
servative surgery’, including compression
nurse. Governments have been slow to priori-
sutures and sequential devascularization
tize women’s health and donor countries have
(see Chapter 31).
not shown sufficient commitment to dealing
(9) Make blood transfusion facilities with with maternal mortality. This is in a context in
secure blood supplies available in centers which there is supposed recognition that pov-
that provide comprehensive health care erty reduction and education are the keys to
(see Chapter 45). good health – that there is no health without
education and no education without health30.
(10) Make definitive surgery (hysterectomy)
To address the issue of postpartum hemor-
and modern clotting factors (recombinant
rhage, ICM and FIGO have launched a world-
factor VIIa) available in level III (tertiary
wide initiative to promote the offer of active
care) hospitals (see Chapter 26).
management to all women. Further research is
National professional associations also have an needed about the benefits of misoprostol, the
important and collaborative roles to play in the secondary side-effects of oxytocin, the anti-
following areas: shock garment, and the balloon tamponade
in preventing and treating postpartum hemor-
(1) Advocacy for skilled care at birth; rhage. Both organizations need the support of
governments, donors and the public to support
(2) Public education about the need for
the campaign that will produce results in
adequate prevention and treatment of
addressing Millennium Development Goal
postpartum hemorrhage;
number 5. We respectfully request the profes-
(3) Publication of the statement in national sional associations to join the ICM/FIGO
midwifery, obstetric and medical journals, coalition to prevent and treat postpartum
newsletters and websites; hemorrhage by working with their Ministries of
30
30 August 2006 14:19:07
Health on the broader issues of poverty, nutri- 9. Li XF, Fortney JA, Kotelchuck M, Glover LH.
tion, status of women, and access to medication The postpartum period: the key to maternal
and education, while they adopt the low-cost mortality. Int J Gynaecol Obstet 1996;52:1–10
medico-surgical approaches we have discussed 10. Coombs CA, Murphy EZ, Laros RK. Factors
associated with postpartum hemorrhage with
in this chapter. Since a good community/
vaginal birth. Obstet Gynecol 1991;77:69–76
national infrastructure designed in Sweden in
11. Kane TT, El-Kady AA, Saleh S, Hage M,
the 1700s still represents a respectable solution Stanback J, Potter L. Maternal mortality in Giza,
to our millennium goal to save mothers, it Egypt: magnitude, causes and prevention. Stud
appears to be time to act upon the answers that Fam Plann 1992; 23:45–57
have been staring us in the face for some time. 12. http://www.mnh.jhpiego.org/best/pphactmng.
asp
13. Maternal Mortality in 2000: Estimates Devel-
ACKNOWLEDGEMENT oped by WHO, UNICEF, UNFPA. Geneva:
This chapter has been modified, at the request World Health Organization, 2004
of the Editors of this book and with the 14. Bartlett LA, Mawji S, Whitehead S, et al. Where
giving birth is a forecast of death: maternal
permission of the journal Editors, from an
mortality in four districts of Afghanistan,
article by Lalonde A, Daviss BA, Acosta A,
1999–2002. Lancet 2005;365:864–70
Herschderfer K. International Journal of 15. Physicians for Human Rights. Maternal Mortal-
Gynaecology and Obstetrics 2006;94:September. ity in Heart Province, Afghanistan, 2002. www.
phrusa.org/research/afghanistan/maternal_
mortality
References
16. Maxine S, Rojas R, PAHO/WHO. Maternal and
1. Taj Mahal History and Pictures at http://www. child mortality among the indigenous peoples of
indianchild.com/taj_mahal.htm the Americas. Healing our Spirit Worldwide 2004;
2. Hogberg U. The decline in maternal mortality in 2:1–3
Sweden: the role of community midwifery. Am J 17. Litch JA. Summary of the evidence base for
Pub Health 2004;94:1312–19 active management of the third stage of labor.
3. World Health Organization. The World Report In Preventing Postpartum Hemorrhage: A Toolkit
2005. Attending to 136 million births, every year. for Providers. PATH (Program for Appropriate
2005. Make every mother and child count. Geneva: Technology for Health) 2004:B-2
the World Health Organization, 2005:61 18. Gülmezoglu AM, Forna F, Villar J, Hofmeyr GJ.
4. Abou Zahr C. Antepartum and postpartum Prostaglandins for prevention of postpartum
haemorrhage. In Murray CJL, Lopez AD, eds. haemorrhage. Cochrane Database of Systematic
Health Dimensions of Sex and Reproduction. Reviews 2004;1
Boston: Harvard University Press, 1998:172–81 19. Gulmezoglu AM, Villar J, Ngoc NT, et al. WHO
5. World Health Organization. The World Report multicentre randomised trial of misoprostol in
2005. Attending to 136 million births, every year. the management of the third stage of labour.
2005. Make every mother and child count. Geneva: Lancet 2001;358:689–95
the World Health Organization, 2005:62–3 20. Elbourne DR, Prendiville WJ, Carroli G, Wood
6. Gulmezoglu AM. Postpartum haemorrhage J, McDonald S. Prophylactic use of oxytocin in
(1997–2002). Monitoring and Evaluation the third stage of labour (Cochrane Review).
Department of Reproductive Health and Cochrane Library 2004;3:6
Research, 25–26 May 2004. Geneva: WHO, 21. Chalmers B, Wu Wen S. Perinatal care in
2004 Canada. BMC Women’s Health 2004;4:3
7. Razvi K, Chua S, Arulkumaran S, Ratnam SS. A 22. Prendiville WJ, Elbourne D, McDonald S.
comparison between visual estimation and labo- Active vs. expectant management in the third
ratory determination of blood loss during the 3rd stage of labour. In The Cochrane Library, Issue
stage of labour. Aust NZ J Obstet Gynaecol 3, 2003. Oxford: Update Software
1996;36:152-4 23. Prata N, Mbaruku G, Campbell M, Potts M,
8. Prata N, Mbaruku G, Campbell M. Using the Vahidnia F. Controlling postpartum hemorrhage
kanga to measure postpartum blood loss. Int J after home births in Tanzania. Int J Gynaecol
Gynaecol Obstet 2005;89:49–50 Obstet 2005;90:51–5
31
30 August 2006 14:19:07
24. Lars H, Cardoso P, Nielsen B, Vdiman L, 27. Gibson L. WHO puts abortifacients on its
Nielsen J, Aaby P. Effect of sublingual miso- essential drug list. Br Med J 2005;331:68
prostol on severe postpartum haemorrhage in 28. Mercer J. Current best evidence: a review of the
a primary health centre in Guinea-Bissau: literature on umbilical cord clamping. J Midwif
randomised double blind clinical trial. Br Med J Women’s Health 2001;46:402–13
2005;331:723–8 29. http://crhrp.ucsf.edu/research/researchareas/
25. Sanghvi H, Wiknjosastro G, Chanpong G, safe_motherhood.html
Fishel J, Ahmed S, Zulkarnain M. Prevention of 30. Sachs JD. Macroeconomics and Health: Investing
postpartum hemorrhage in West Java, Indonesia. in Health for Economic Development. Geneva:
Baltimore: JHPIEGO Brown’s Wharf, 2004 World Health Organization, 2001
26. PATH. Misoprostol use in obstetrics and
gynecology. Outlook 2005;21
10
32
30 August 2006 14:19:07
2
DEFINITIONS AND CLASSIFICATIONS
A. Coker and R. Oliver
INTRODUCTION clinical outcome. Therefore, a prognostic classi-

fication will guide the degree of aggressiveness
Conventionally, the term ‘postpartum hemor-
of the intervention, especially as management
rhage’ is applied to pregnancies beyond 20
may involve more than one clinical specialty.
weeks gestation. Although bleeding at an earlier
It will also help to decide on the optimal site
gestational age may have a similar etiology and
for subsequent care, for example in a high-
management to postpartum hemorrhage, these
dependency unit or intensive care unit, if such
are usually referred to as spontaneous
exist in the hospital.
miscarriages.
The third reason is to allow effective
There has been no significant change in the
communication based on standardization of the
definitions or classification over the past 50
estimate of the degree of hemorrhage, thus stan-
years; this does not reflect the advances made
dardizing differing management options. The
in medical and surgical treatment over this
initial assessment is usually made by the staff
period1. A widely used definition currently is
available on site, and these are often relatively
that proposed by the World Health Organiza-
junior medical or midwifery personnel. They, in
tion (WHO) in 1990 as ‘any blood loss from the
turn, have to assess the severity of bleeding and
genital tract during delivery above 500 ml’2.
summon help or assistance as required. Thus, a
The average blood loss during a normal
standardized easily applicable working classifi-
vaginal delivery has been estimated at 500 ml;
cation facilitates effective communication and
however, around 5% of women would lose
obviates inter-observer variation.
greater than 1000 ml during a vaginal birth3–6.
Cesarean deliveries are associated with an aver-
age estimated blood loss of 1000 ml7. There is, CLASSIFICATIONS IN USE
therefore, a degree of overlap in the acceptable
Conventional temporal classification
range of blood loss for vaginal and Cesarean
deliveries. Traditionally, the classification of postpartum
hemorrhage has been based on the timing of the
onset of bleeding in relation to the delivery.
PURPOSE OF CLASSIFICATION
Hemorrhage within the first 24 h of vaginal
Classification of postpartum hemorrhage is delivery is termed either early or primary
desirable for the following reasons. First, due to postpartum hemorrhage, whereas bleeding
the rapidity of disease progression, there is an occurring afterwards, but within 12 weeks of
overriding clinical need to determine the most delivery, is termed late or secondary postpartum
suitable line of management. The urgency of hemorrhage8.
intervention depends on the rate of the patient’s Secondary postpartum hemorrhage is less
decline or deterioration. common than primary postpartum hemorrhage,
The second reason for classification is to affecting 1–3% of all deliveries. In both cases,
assess the prognosis. This may help to deter- the true blood loss is often underestimated due
mine the immediate, medium and long-term to the difficulty with visual quantitation9,10.
11
33
30 August 2006 14:19:07
Classification based on quantification of Table 1 Benedetti’s classification of hemorrhage15

blood loss
Hemorrhage Acute blood loss Percentage
Amount of blood lost class (ml) lost
Blood loss at delivery is estimated using various 1 900 15

methods. These range from the less modern 2 1200–1500 20–25
methods of counting blood-soaked pieces of 3 1800–2100 30–35
cloth or ‘kangas’ used by traditional birth 4 2400 40
attendants in rural settings, to more modern
techniques such as calculating the blood loss
by subtraction after weighing all swabs using
sensitive weighing scales11.
The International Statistical Classification signs of volume deficit and will not require any
of Diseases and Related Health Problems, acute treatment.
Tenth Revision, Australian Modification
(ICD-10-AM) describes postpartum hemor- Class 2 A blood loss of 1200–1500 ml will
rhage as a blood loss of 500 ml or more for a begin to manifest clinical signs, such as a rise in
vaginal delivery and 750 ml or more in associa- pulse and respiratory rate. There may also be
tion with Cesarean delivery12. recordable blood pressure changes, but not the
classic cold clammy extremities.
Change in hematocrit Class 3 These are patients in whom the blood
The American College of Obstetricians and loss is sufficient to cause overt hypotension. The
Gynecologists advocates the definitions of either blood loss is usually around 1800–2100 ml.
a 10% change in hematocrit between the ante- There are signs of tachycardia (120–160 bpm),
natal and postpartum periods, or a need for cold clammy extremities and tachypnea.
erythrocyte transfusion13.
Class 4 This is commonly described as massive
obstetric hemorrhage. When the volume loss
Rapidity of blood loss
exceeds 40%, profound shock ensues and the
In attempts to overcome these inconsistencies, blood pressure and pulse are not easily record-
the classification of postpartum hemorrhage has able. Immediate and urgent volume therapy is
also been based on the rapidity of blood loss. necessary, as this quantity of blood loss can
Severe hemorrhage has been classified as blood be fatal secondary to circulatory collapse and
loss > 150 ml/min (within 20 min, causing loss cardiac arrest.
of more than 50% of blood volume) or a sudden
blood loss > 1500–2000 ml (uterine atony; loss
Classification based on causative factors
of 25–35% of blood volume)14.
The causes of postpartum hemorrhage can also
form a basis of classification (Table 2).
Volume deficit
A form of standardized classification described
Causes of primary postpartum hemorrhage
by Benedetti considers four classes of hemor-
rhage15 (Table 1). The class of hemorrhage Primary postpartum hemorrhage is traditionally
reflects the volume deficit, and this is not neces- considered as a disorder of one or more of the
sarily the same as the volume of blood loss. four processes: uterine atony, retained clots or
placental debris, genital lesions or trauma, and
Class 1 The average 60 kg pregnant woman disorders of coagulation. An aide memoire is the
has a blood volume of 6000 ml at 30 weeks ges- four Ts: tonus, tissue, trauma and thrombin.
tation. A volume loss of less than 900 ml in such Uterine atony alone accounts for 75–90% of
a woman will rarely lead to any symptoms and cases of postpartum hemorrhage.
12
34
30 August 2006 14:19:07
Definitions and classifications
Table 2 Classification of postpartum hemorrhage (PPH) according to causative factors
Causes of primary PPH

Tonus (uterine atony)
Uterine overdistention: multiparity, polyhydramnios, macrosomia
Uterine relaxants: nifedipine, magnesium, beta-mimetics, indomethacin, nitric oxide donors
Rapid or prolonged labor
Oxytoxics to induce labor
Chorioamnionitis
Halogenated anesthetics
Fibroid uterus
Tissue
Impediment to uterine contraction/retraction: multiple fibroids, retained placenta
Placental abnormality: placenta accreta, succenturiate lobe
Prior uterine surgery: myomectomy, classical or lower segment Cesarean section
Obstructed labor
Prolonged third stage of labor
Excessive traction on the cord
Trauma
Vulvovaginal injury
Episiotomy/tears
Macrosomia
Precipitous delivery
Thrombin (coagulopathy)
Acquired during pregnancy: thrombocytopenia of HELLP syndrome, DIC (eclampsia, intrauterine fetal
death, septicemia, placenta abruptio, amniotic fluid embolism), pregnancy-induced hypertension, sepsis
Hereditary: Von Willebrand’s disease
Anticoagulant therapy: valve replacement, patients on absolute bedrest
Causes of secondary PPH
Uterine infection
Retained placental fragments
Abnormal involution of placental site
Adapted from Wac et al. Female Patient 2005;30:19
Classification based on clinical signs and represent the current clinical situation. To a
symptoms certain extent, any classification is of limited use
to a clinician faced with active and continuous
Any bleeding that results in or could result
bleeding.
in hemodynamic instability, if untreated,
The change in hematocrit depends on the
is considered as postpartum hemorrhage
timing of the test and the amount of fluid
(Table 3).
resuscitation previously administered16. It could
also be affected by extraneous factors such as
prepartum hemoconcentration, which may exist
PITFALLS OF CURRENT
in conditions such as pre-eclampsia.
CLASSIFICATIONS
Where the diagnosis is made by a clinical
The drawbacks of a classification based solely estimate of blood loss, there is often signifi-
on blood loss or hematocrit include the fact that cant underestimation. The WHO definition of
this is a retrospective assessment and may not 500 ml is increasingly becoming irrelevant, as
13
35
30 August 2006 14:19:07
Table 3 Symptoms related to blood loss with postpartum hemorrhage
Blood loss
Blood pressure
% ml (mmHg) Signs and symptoms
10–15 500–1000 normal palpitations, dizziness, tachycardia

15–25 1000–1500 slightly low weakness, sweating, tachycardia
25–35 1500–2000 70–80 restlessness, pallor, oliguria
35–45 2000–3000 50–70 collapse, air hunger, anuria
Adapted from Bonnar J. Baillieres Best Pract Res Clin Obstet Gynaecol 2000;14:1
most healthy mothers in the developed world volume loss and the clinical consequences of
can cope with a blood loss of less than 500 ml such loss. The recorded parameters should be
without any hemodynamic compromise. easily measurable and reproducible. This will
Classifications based on the need for blood help in providing an accurate and consistent
transfusion alone are also of limited value assessment of loss, which can readily be com-
as the practice of blood transfusion varies municated and incorporated into most labor
widely according to local circumstances and ward protocols.
attitudes to transfusion of both patients and
physicians17.
PROPOSED CLASSIFICATION
The clinical application of such a classifica-
tion may, in addition, be limited because of The 500 ml limit as defined by WHO2 should
inherent individual differences in response be considered as an alert line; the action line is
to blood loss. Hemodynamic compensation then reached when the vital functions of the
depends on the initial hemoglobin levels prior to woman are endangered. In healthy women, this
onset of bleeding, and these vary among healthy usually occurs after the blood loss has exceeded
individuals. For these reasons, reliance on a 1000 ml.
classification solely based on the amount of We propose a classification (Table 4)
blood loss and without consideration of clinical wherein the volume loss is assessed in conjunc-
signs and symptoms may lead to inconsistency tion with clinical signs and symptoms. We pro-
with management. pose this classification as being mainly useful in
fully equipped hospitals and obstetric units. It is
not being proposed for full implementation in
NEED FOR A CLINICAL AND
areas which are resource-poor.
PROGNOSTIC CLASSIFICATION
Our adaptation of a previously described
Universally, guidelines on the management of classification15 will fulfil most of these criteria.
postpartum hemorrhage have reiterated the This guideline adopts a practical approach
importance of accurate estimation of blood loss, whereby a perceived loss of 500–1000 ml (in
and the clinical condition of the hemorrhaging the absence of clinical signs of cardiovascular
patient. This was further emphasized in the instability) prompts basic measures of monitor-
1988–1990 Confidential Enquiries into Mater- ing and readiness for resuscitation (alert line),
nal Deaths in the United Kingdom (CEMD)18 whereas a perceived loss of > 1000 ml or a
and reiterated in the 1991–1993 report as a smaller loss associated with clinical signs of
list of six bullet points, the first being ‘accurate shock (hypotension, tachycardia, tachypnea,
estimation of blood loss’19. oliguria or delayed peripheral capillary fill-
The ideal classification of postpartum hem- ing) prompts a full protocol of measures to
orrhage should take into consideration both the resuscitate, monitor and arrest bleeding.
14
36
30 August 2006 14:19:08
Definitions and classifications
Table 4 Proposed classification. Adapted from Benedetti15
Hemorrhage class Estimated blood loss (ml) Blood volume loss (%) Clinical signs and symptoms
0 (normal loss) < 500 < 10 none

ALERT LINE
1 500–1000 < 15 minimal
ACTION LINE
2 1200–1500 20–25 â urine output
á pulse rate
á respiratory rate
postural hypotension
narrow pulse pressure
3 1800–2100 30–35 hypotension
tachycardia
cold clammy
tachypnea
4 > 2400 > 40 profound shock
Need observation ± replacement therapy

Replacement therapy and oxytocics
Urgent active management
Critical active management (50% mortality if not managed actively)
References 8. Alexander J, Thomas P, Sanghera J. Treatments

for secondary postpartum haemorrhage.
1. El-Refaey H, Rodeck C. Post partum haemor- Cochrane Database of Systematic Reviews,
rhage: definitions, medical and surgical manage- 2005, Issue 3
ment. A time for change. Br Med Bull 2003;67: 9. Gahres EE, Albert SN, Dodek SM. Intrapartum
205–17 blood loss measured with Cr 51-tagged erythro-
2. World Health Organization. The Prevention and cytes. Obstet Gynecol 1962;19:455–62
Management of Postpartum Haemorrhage. Report 10. Newton M, Mosey LM, Egli GE, Gifford WB,
of a Technical Working Group, Geneva, 3–6 Hull CT. Blood loss during and immediately
July, 1989. Unpublished document. WHO/ after delivery. Obstet Gynecol 1961;17:9–18
MCH/90.7. Geneva: World Health Organiza- 11. Prata N, Mbaruku G, Campbell M. Using the
tion, 1990 kanga to measure post partum blood loss. Int J
3. Pritchard JA, Baldwin RM, Dickey JC, Wiggins Gynaecol Obstet 2005;89:49–50
KM. Blood volume changes in pregnancy and 12. National Centre for Classification in Health.
the puerperium. Am J Obstet Gynecol 1962;84: Australian Coding Standards. The International
1271–82 Statistical Classification of Diseases and Related
4. Newton M. Postpartum hemorrhage. Am J Health Problems, Tenth Revision, Australian
Obstet Gynecol 1966;94:711–17 Modification (ICD-10-AM). Sydney, Australia,
5. De Leeuw NK, Lowenstein L, Tucker EC, 2002
Dayal S. Correlation of red cell loss at delivery 13. American College of Gynecologists and Obstetri-
with changes in red cell mass. Am J Obstet cians. Quality Assurance in Obstetrics and Gynecol-
Gynecol 1968;100:1092–101 ogy. Washington DC: American College of
6. Letsky E. The haematological system. In Hytten Obstetricians and Gynecologists, 1989
F, Chamberlain G, eds. Clinical Physiology in 14. Sobieszczyk S, Breborowicz GH. Management
Obstetrics, 2nd edn. Oxford: Blackwell, 1991: recommendations for postpartum hemorrhage.
2–75 Arch Perinatal Med 2004;10:1
7. Baskett TF, ed. Complications of the third stage 15. Benedetti T. Obstetric haemorrhage. In Gabbe
of labour. In Essential Management of Obstetrical SG, Niebyl JR, Simpson JL, eds. A Pocket
Emergencies, 3rd edn. Bristol, UK: Clinical Press, Companion to Obstetrics, 4th edn. New York:
1999:196–201 Churchill Livingstone, 2002:Ch 17
15
37
30 August 2006 14:19:08
16. Cunningham FG, Gant NF, Leveno KJ, et al., 18. Hibbard BM. Report on Confidential Enquiries
eds. Conduct of normal labor and delivery. into Maternal Deaths in the United Kingdom,
In Williams Obstetrics, 21st edn. New York: 1988–1990. London: Her Majesty’s Stationery
McGraw-Hill, 2001:320–5 Office, 1994
17. Schuurmans N, MacKinnon C, Lane C, Etches 19. Anonymous. Report on Confidential Enquiries
D. Prevention and management of postpartum into Maternal Deaths in the United Kingdom,
haemorrhage. J Soc Obstet Gynaecol Canada 1991–1993. London: Her Majesty’s Stationery
2000;22:271–81 Office, 1996
16
38
30 August 2006 14:19:08
3
VITAL STATISTICS: AN OVERVIEW
M. J. Cameron and S. C. Robson
INTRODUCTION in significant morbidity, one might argue that

the classic definition of primary postpartum
Postpartum hemorrhage constitutes a major
hemorrhage is clinically inappropriate and
cause of maternal mortality, particularly in the
should be revised to identify a group of women
developing world, and of maternal morbidity in
who become ‘ill’ and at real risk of morbidity
both the developed and the developing world.
after the hemorrhage. If the classic definition
This chapter describes the incidence of primary
were to be changed, definitions of any event
postpartum hemorrhage, the difficulties in
leading to severe obstetric morbidity could then
reporting epidemiological data on primary
be based on ‘pathophysiology’, ‘management’
postpartum hemorrhage and the etiology and
or a combination of both parameters3. The
precipitating factors for primary postpartum
problem with using a management-based defi-
hemorrhage. Because of its broad scope, this dis-
nition of hemorrhage, such as number of units
cussion will invariably include several points that
of blood transfused, is that it can only be used
are discussed in greater detail elsewhere. Regard-
retrospectively and is of no value to the clinician
less, these statistics should provide additional
attempting to treat the primary postpartum
insights as many derive from secondary analyses.
hemorrhage. Further, such a definition is likely
to be highly influenced by local practitioner/
hospital beliefs about when to transfuse as well
DEFINING POSTPARTUM
as the local facilities available for transfusion
HEMORRHAGE
(see Chapter 45). Consequently, according to a
The traditional definition of primary postpartum recent UK position, it may be better to think of
hemorrhage used in most textbooks of obstetrics the term ‘significant obstetric hemorrhage’,
is a visually estimated blood loss of 500 ml or using a definition of loss of more than 1000 ml
more within the first 24 h after delivery1. Sec- or more than 1500 ml, rather than define
ondary postpartum hemorrhage is generally primary postpartum hemorrhage as > 500 ml
defined as ‘excessive bleeding’ from the genital blood loss4.
tract after 24 h and up to 6 weeks post-delivery In the average non-pregnant adult, circulat-
(see Chapter 2). As such, this latter definition ing blood represents a total of 7% of body
only contains quantification of the time period weight, or approximately 5 liters. Loss
rather than the extent of blood loss. However, of 30–40% of the circulating volume
according to older and commonly quoted data, (1500–2000 ml) results in tachycardia, tachyp-
measured blood loss during a vaginal delivery nea, a measurable fall in systolic blood pressure
averages 500 ml whereas that during a Cesarean and alterations in mental state5. Therefore,
section averages 1000 ml2. Thus, the ‘classic’ the concept of defining a ‘significant primary
definition of primary postpartum hemorrhage is postpartum hemorrhage’ as one resulting in a
in reality a reflection of the almost universal ten- blood loss of 1500 ml or more is meritorious as
dency to underestimate delivery blood loss (see this reflects the point when physiological com-
below and Chapters 4 and 6). pensatory mechanisms begin to fail. Whether
Because a loss of 500 ml at delivery for most this concept will find universal acceptance
women in the developed world does not result remains to be seen, however.
17
39
30 August 2006 14:19:08
DIFFICULTIES OF COMPARING blood pigment converted to acid or alkaline

STUDIES hematin and the concentration then compared
in a colorimeter with the patient’s own venous
Two key factors must be considered when com-
blood8. Alternatively, volumetric methods
paring published studies of primary postpartum
involve labelling the woman’s plasma or
hemorrhage: first, the method used to deter-
erythrocytes with dyes or radioactive substances
mine blood loss, and, second, the method of
and then calculating the reduction in blood
managing the third stage of labor. In addition,
volume. Unfortunately, both techniques require
confounding represents a potential problem in
expertise and are more time-consuming and
case-control studies that examine risk factors for
expensive to perform than simple measurement
primary postpartum hemorrhage.
of blood loss.
Visual estimation has long been considered
to be unreliable, but only recently have
Determining blood loss: estimating versus
data proven this to be the case. Duthie and
measuring
colleagues compared visual estimation and
Accurate measurement of blood loss at delivery measured blood loss using the alkaline-hematin
is possible but must be planned for in advance method during normal delivery in 37 primi-
(see also Chapter 4). The most obvious is gravid and 25 multigravid women. These inves-
collection of blood into receptacles and direct tigators found that, for both groups, the mean
measurement. This can be combined with a estimated blood loss (261 ml and 220 ml,
gravimetric procedure which depends upon respectively) was significantly lower than the
converting the increase in weight of sponges and mean measured blood loss (401 ml and 319 ml,
linen into milliliters of blood on a ml/g basis. respectively)9. This observation is consistent
Gulmezoglu and Hofmeyr recently proposed with studies of simulated scenarios that suggest
a method for directly measuring blood loss midwives and doctors underestimate blood
objectively which does not interfere with routine loss at delivery by 30–50%10. Importantly, esti-
care6. They suggest ‘after delivery of the baby, mates are particularly unreliable for very
the amniotic fluid is allowed to drain away and small and very large amounts of blood11 (see
amniotic fluid-soaked bed linen is covered with Chapter 6).
a dry disposable ‘linen saver’. A low-profile, Reported rates of postpartum hemorrhage
wedge-shaped plastic ‘fracture bedpan’ is also differ widely depending on the method of
slipped under the woman’s buttocks for blood measuring blood loss. Older studies that directly
collection, with blood and clots decanted into a measured blood loss reported rates of primary
measuring cylinder. Weighing of blood-soaked postpartum hemorrhage (> 500 ml) of between
swabs and linen savers occurs, with the known 22% and 29%12,13 compared to rates of 5–8%
dry weight subtracted and calculated volume with visual estimation. More recently, Prasert-
added to that from the bedpan.’ They particu- charoensuk and colleagues compared visual
larly recommend this method for all future trials estimation with direct measurement in 228
of interventions to reduce primary postpartum women who had a spontaneous vaginal deliv-
hemorrhage. Strand and colleagues suggested a ery14. The incidences of postpartum hemor-
novel method with a combination of a plastic rhage > 500 ml and > 1000 ml were 5.7% and
sheet and a bucket below a cholera bed on 0.44%, respectively by visual estimation,
which the woman rested during postpartum whereas direct measurements showed inci-
observation7. The BRASSS-V collection drape dences of 27.63% and 3.51%, respectively.
and instructions for use are described in These differences are five and seven times
Chapter 4. As with any direct measurement of higher, respectively. The authors concluded
blood loss, contamination with amniotic fluid that visual estimation underestimated the
and urine is not uncommon. incidence of postpartum hemorrhage by 89%.
Laboratory-based methods for measuring Razvi and colleagues conducted a similar
blood loss include photometric techniques, prospective study and showed a similar degree
whereby sanitary protection is collected and of underestimation15.
18
40
30 August 2006 14:19:08
Vital statistics
Conduct of third stage of labor higher rates of age-related medical diseases,

such as diabetes mellitus, which could be the
Active management of the third stage (AMTSL)
‘true’ risk factors for postpartum hemorrhage.
involves early clamping of the umbilical cord
Methods used to control confounders
before pulsations have stopped, controlled cord
include:
traction using the Brandt–Andrews technique
and the use of prophylactic uterotonics such as (1) Restriction – in the example cited in the
syntocinon or syntometrine, usually with the preceding paragraph, women with diabetes
delivery of the fetal anterior shoulder (see also mellitus could be excluded. However,
Chapter 11). In contrast, expectant or ‘physio- restriction limits the external validity of the
logical’ third stage involves late clamping of the findings and reduces the sample size.
cord after pulsations have stopped, waiting for
(2) Matching – here, if diabetes mellitus is
spontaneous separation of the placenta from the
deemed a confounder, then for every
uterine wall and avoidance of synthetic utero-
woman recruited with diabetes mellitus
tonics. Nipple stimulation has been used to pro-
who has a postpartum hemorrhage, she is
mote the release of endogenous oxytocin and
matched to a control with diabetes mellitus.
reduce the length and amount of bleeding at the
third stage of labor16, but is not part of active or (3) Stratification – can be thought of as post hoc
expectant management. A meta-analysis of five restriction performed at the analysis phase.
randomized, controlled trials (involving over
Multivariable analysis is a statistical tool for
6000 women) indicates that active management
determining the relative contributions of
results in a reduction in maternal blood loss at
different causes to a single event or outcome19.
delivery and a reduction in the risks of post-
Epidemiological studies that use multivariable
partum hemorrhage, defined as an estimated
statistical methods are much more likely to
blood loss > 500 ml (relative risk (RR) 0.38,
eliminate confounders. For readers who require
95% confidence interval (CI) 0.32–0.46), severe
further information about the problems of
postpartum hemorrhage, defined as an estimated
epidemiological studies, please refer to Grimes
blood loss ≥ 1000 ml (RR 0.33, 95% CI
and Schultz and Mamdani and colleagues20,21.
0.21–0.51) and prolonged third stage17.
Clearly, the reported incidence of post-
partum hemorrhage in any population is influ-
INCIDENCE OF PRIMARY
enced by the conduct of the third stage. As
active management is less widely practiced in
the developing world, this must be considered Denominator data
when making international comparisons of
Studies that attempt to quantify the incidence
postpartum hemorrhage rates.
and impact of postpartum hemorrhage need
a denominator value over a time period to
calculate rates. Common denominators used
CONFOUNDING FACTORS IN
to calculate maternal mortality and morbidity
EPIDEMIOLOGICAL STUDIES
rates22 are illustrated in Table 1.
Confounding is a potential problem in epi- Developed countries, including the United
demiologic studies exploring risk factors. A Kingdom, have the advantage of accurate
confounder is associated with the risk factor denominator data, including both livebirths and
and causally related to the outcome. Thus, a stillbirths. Consequently, the UK Confidential
researcher may attempt to relate an exposure to Enquiries into Maternal Deaths have used
an outcome, but actually measures the effect of maternities for denominator data because this
a third factor, the confounding variable18. As an enables establishment of a more detailed picture
example, parity, particularly grand multiparity, of maternal death rates. However, for many
is generally considered a risk factor for primary countries, particularly in the developing world,
postpartum hemorrhage. However, grand no process of stillbirth (or even livebirth) regis-
multiparas tend to be older and therefore have tration exists. Denominator data are, therefore,
19
41
30 August 2006 14:19:09
Table 1 Denominators used in calculating maternal mortality and morbidity
Denominator Definition Advantages and disadvantages

Livebirths Number of pregnancies that result in a live- Easier to collect than maternities
birth at any gestation
Maternities Number of pregnancies that result in a live- Includes the majority of women at risk from
birth at any gestation or stillbirths occurring death from obstetric causes but requires
at or after 24 weeks of completed gestation infrastructure for notification of stillbirths
and required to be notified by law
Women aged Number of women of reproductive age in a Lacks rigor of confining rate to women who
15–44 years given population were pregnant
Enables comparison with other causes of
death
likely to be based on livebirths, rather than The United Kingdom

maternities. Indeed, in some countries even
A triennial report on confidential enquiries into
livebirth data collection may not be reliable.
maternal death has been published since 1985,
As a result, it is often extremely difficult to
with reports for England and Wales commenc-
compare maternal mortality and morbidity from
ing in 1952. Direct deaths are reported that
different geographic areas.
result from obstetric complications of the preg-
nant state (pregnancy, labor and puerperium up
Maternal mortality to 42 days), from interventions, omissions,
One method of attempting to quantify the incorrect treatment or from a chain of events
magnitude of postpartum hemorrhage is to look resulting from any of the above. Obstetric hem-
at its contribution to maternal deaths around orrhage (comprising placental abruption, pla-
the world, and in a particular country over time. centa previa and postpartum hemorrhage) is
Trends over time within one country are an one example of direct deaths23. In the
important audit tool in examining the care of 2000–2002 triennium, there were 106 direct
women with postpartum hemorrhage, as can maternal deaths. Seventeen (16%) were attrib-
be seen from the UK Confidential Enquiries uted to obstetric hemorrhage with ten (9.4%)
into Maternal Deaths. However, differences attributed principally to postpartum hemor-
between countries often reflect differences in rhage. Since the UK-wide triennium report
health-care provision, general economic pros- began in 1985, 83 deaths from obstetric hemor-
perity and geographic and climactic conditions rhage have been recorded, of which half (41
that affect access to obstetric care. women) was caused by postpartum hemor-
rhage, resulting in a death rate for postpartum
hemorrhage of 3.1 per million maternities. Cal-
Global picture
culated death rates for postpartum hemorrhage
The WHO estimates that obstetric hemorrhage for each triennium are shown in Table 2.
complicates 10.5% of all livebirths in the Although at first glance there appears to be a
world, with an estimated 13 795 000 women marked increase in postpartum hemorrhage in
experiencing this complication in 200022. the last triennial report compared to the one
Around 132 000 maternal deaths are directly that immediately preceded it, two patients had
attributable to hemorrhage, comprising 28% of no contact at all with health services and two
all direct deaths. In comparison, the following patients refused blood products that would
numbers relate to other conditions: 79 000 probably have saved their lives. Excluding
deaths from sepsis, 63 000 deaths from pre- these four deaths results in a rate per million
eclampsia/eclampsia, 69 000 from abortion and maternities comparable to the reports published
42 000 from obstructed labor. between 1985 and 1996.
20
42
30 August 2006 14:19:09
Vital statistics
Table 2 Maternal mortality from postpartum hem- complications. Of the 2519 maternal deaths that
orrhage in UK (extrapolated from CEMACH23) resulted in a livebirth and the 275 maternal
deaths resulting in stillbirth, 2.7% and 21.1%,
Postpartum Total Rate per
respectively, were considered to be a direct result
hemorrhage maternities million
Triennium (n) (n) maternities from obstetric hemorrhage. Unfortunately, no
separate data were provided about postpartum
1985–87 6 2 268 766 2.6 hemorrhage. Comparison with the 1987–1990
1988–90 11 2 360 309 4.6 data shows a reduction in the percentage of
1991–93 8 2 315 204 3.4 maternal deaths from pregnancy-related hemor-
1994–96 5 2 197 640 2.2 rhage from 28.7% to 17%25.
1997–99 1 2 123 614 0.4
2000–02 10 1 997 472 5.0
France
A confidential enquiry into maternal deaths in
Of the eight women who sought care in the five of the 22 administrative areas of France
2000–2002 cohort and ultimately died from found that five deaths from 39 obstetric causes
postpartum hemorrhage, elements of sub- were due to postpartum hemorrhage26; impli-
standard care were present in seven (88%) cating postpartum hemorrhage in 13% of the
including: obstetric deaths. No denominator data were
collected, and therefore it is not possible to
(1) Organizational problems – including inap-
estimate rates.
propriate booking at hospitals with inade-
quate blood transfusion and intensive care
facilities; Africa
(2) Poor quality of resuscitation – including Bouvier-Colle and colleagues performed a
inadequate transfusion of blood and blood population-based survey of pregnant women
products; from seven West African areas from 1994 to
199627. Overall, 55 women died from direct or
(3) Equipment failure, e.g. malfunctioning of
indirect obstetric causes among 17 694 live
specimen transport system;
births. Hemorrhage accounted for 17 deaths
(4) Inadequate staffing of recovery areas; (31%), with delivery hemorrhage (third stage)
and post-delivery hemorrhage (retention of pla-
(5) Failure to recognize or treat antenatal medi-
centa) accounting for six and four deaths, respec-
cal conditions, e.g. inherited bleeding disor-
tively. This equates to a maternal mortality rate
ders;
of 565 per 1 000 000 livebirths, a rate approxi-
(6) Failure of senior staff to attend; mately 100-fold higher compared to the UK.
Another study in South Africa, involving one
(7) Concerns about the quality of surgical treat-
tertiary center, reported a maternal mortality
ment given.
rate of 1710 per 1 000 000 livebirths during
The recognition of these diverse elements pro- the period 1986–1992, with 25% of deaths
vides a blue-print to health-care authorities to attributed to obstetric hemorrhage28. Within
institute remedial action (see Chapter 22). this setting, hemorrhage was the leading cause
of death.
United States of America
Maternal morbidity
The Center for Disease Control (CDC) con-
ducted a pregnancy-related mortality survey in Because maternal death in the developed world
the USA between 1991 and 199924. Hemorrhage is a rare event, clinicians have attempted to
in pregnancy was responsible for 17% of quantify significant morbidity, which is often
maternal deaths, although this figure includes labelled as a maternal adverse event or a near
hemorrhage from first-trimester pregnancy miss (see Chapter 37). Studies have generally
21
43
30 August 2006 14:19:09
included massive obstetric hemorrhage as one of cases. However, this study did not include
indicator of severe maternal morbidity. As with thromboembolic disease, which is the leading
mortality, comparisons between studies are cause of direct maternal deaths in the UK.
often difficult because of variations in definition
of ‘massive obstetric hemorrhage’. Both ante-
Canada
natal and intrapartum bleeding are sometimes
included within the definition of ‘obstetric Wu Wen and colleagues conducted a retrospec-
hemorrhage’. tive cohort study of severe maternal morbidity
involving 2 548 824 women who gave birth in
Canadian Hospitals over a 10-year period from
Scotland
1991, using information on hospital discharges
The Scottish Programme for Clinical Effective- compiled by the Canadian Institute for Health
ness in Reproductive Health (SPCERH) con- Information30. Their criteria for severe maternal
ducted a prospective investigation into 14 morbidity included postpartum hemorrhage
severe maternal morbidity categories for all requiring hysterectomy or transfusion. Their
maternity units in Scotland in 20033. Within overall rate of all severe maternal morbidity was
this audit, major obstetric hemorrhage was 4.38 per 1000 deliveries. Overall rates for severe
defined as estimated blood loss ≥ 2500 ml, or postpartum hemorrhage in the 10-year time
transfusion of ≥ 5 units of blood or the need for frame are illustrated in Table 3 along with time
fresh frozen plasma or cryoprecipitate. Of the analysis for rates at the beginning and end of the
375 events, 176 (46%) were reported to be study. Within this study, rates for postpartum
related to obstetric hemorrhage. Because some hemorrhage requiring transfusion halved (RR
patients experienced more than one morbid 0.5, CI 0.44–0.55), but hysterectomy rates for
event, major obstetric hemorrhage occurred in postpartum hemorrhage almost doubled (RR
65% of ‘near-miss patients’ (176/270). Using a 1.76, CI 1.48–2.08). Because the definition of
denominator of 50 157 livebirths, the authors postpartum hemorrhage was based on manage-
calculated a rate of major obstetric hemorrhage ment rather than pathophysiology, it is difficult
of 3.5/1000 births (CI 3.0–4.1). Of the 176 to tease out whether the temporal change
cases notified to the investigators, full disclosure reflects a true reduction in the incidence of
of data was obtained in 152 cases; 70% of the postpartum hemorrhage or simply a change in
cases were due to primary postpartum hemor- clinical management.
rhage, 26% to intrapartum hemorrhage and
17% to antepartum hemorrhage with some
Africa
women falling into more than one category.
Filippi and colleagues conducted prospective
and retrospective data extraction on near-miss
England
obstetric events in nine referral hospitals in
In the South East Thames region, 19 maternity three countries (Benin, Cote d’Ivoire, and
units participated in a 1-year study between Morocco)31. Obstetric hemorrhage was defined
1997 and 1998 to determine the incidence of as hemorrhage leading to clinical shock,
severe obstetric morbidity29. Severe obstetric emergency hysterectomy and blood transfusion.
hemorrhage was defined as estimated blood loss The incidence of near-miss cases varied widely
> 1500 ml or a peripartum fall in hemoglobin between hospitals. Most of the women were
concentration of ≥ 40 g/l or the need for an already in a critical condition on arrival, with
acute transfusion of 4 or more units of blood. two-thirds being referred from another facility.
There were 588 cases of severe obstetric mor- The study identified a total of 507 cases of late
bidity among 48 856 women delivered over the pregnancy obstetric hemorrhage (i.e. previa,
year, giving an incidence of 12/1000 deliveries. abruption and other non-classified hemorrhage
Hemorrhage was the leading cause of obstetric and postpartum hemorrhage) from 33 478
morbidity at 6.7 (CI 6.0–7.5) occurrences per deliveries, representing a near-miss late obstet-
1000 deliveries, representing nearly two-thirds ric hemorrhage rate of 15.1/1000 deliveries. In
22
44
30 August 2006 14:19:10
Vital statistics
Table 3 Postpartum hemorrhage (PPH) rates in Canada 1991–2000. Adapted from Wu Wen30
Number Rate per 1000 Rate per 1000 Rate per 1000 Relative
of cases deliveries deliveries deliveries risk
(1991–2000) (95% CI) (1991–1993) (1998–2000) (95% CI)*
PPH requiring 2317 0.91 (0.87–0.95) 1.27 0.63 0.5 (0.44–0.55)

transfusion
PPH requiring 892 0.35 (0.33–0.37) 0.26 0.46 1.76 (1.48–2.08)
hysterectomy
*The 1991–1993 period was the reference period
total there were 266 cases of postpartum hemor- Table 4 Rates per 1000 births for near misses plus
rhage, representing a near-miss postpartum maternal deaths from severe postpartum hemor-
hemorrhage rate of 7.9/1000 deliveries. rhage in Pretoria. Adapted from Pattinson et al.33
Prual and colleagues examined severe mater-
1997–99 2000 2001 2002
nal morbidity from direct obstetric causes in
West Africa between 1994 and 199632. A severe Rate/1000 births 0.96 1.37 2.38 2.28
obstetric event was defined as prepartum,
peripartum or postpartum hemorrhage leading
to blood transfusion, or hospitalization for more Table 5 The Four Ts of postpartum hemorrhage
than 4 days or to hysterectomy. A total of 1307 (from ALSO34)
severe maternal morbidity events were identi-
Tone – uterine atony
fied, with obstetric hemorrhage representing
Trauma – of any part of the genital tract, inverted
the largest group involving 601 cases, 342 of uterus
which were postpartum hemorrhage. The near- Tissue – retained placenta, invasive placenta
miss obstetric hemorrhage rate was 30.5 (CI Thrombin – coagulopathy
28.1–33.0)/1000 live births and the near-miss
postpartum hemorrhage rate was 17.4 (CI
15.6–19.3)/1000 live births.
The Pretoria region of South Africa has Uterine atony
used the same definition of ‘near miss’ for
Uterine atony, the most common cause of
over 5 years, allowing comparison of temporal
postpartum hemorrhage, is reported in 70%
changes33. Rates per 1000 births for near misses
of cases34. It can occur after normal vaginal
plus maternal deaths over 5 years from severe
delivery, instrumental vaginal delivery and
postpartum hemorrhage are shown in Table 4.
abdominal delivery. A large cohort study found
These rates are not dissimilar to those in
an incidence of uterine atony after primary
Canada or the UK.
Cesarean section of 1416/23 390 (6%)35. Multi-
ple linear regression analysis demonstrates the
following factors as being independently associ-
ETIOLOGY AND PRECIPITATING ated with risk of uterine atony: multiple gesta-
FACTORS tion (odds ratio (OR) 2.40, 95% CI 1.95–2.93),
Hispanic race (OR 2.21, 95% CI 1.90–2.57),
Causes of primary postpartum induced or augmented labor for > 18 h (OR
hemorrhage 2.23, 95% CI 1.92–2.60), infant birth weight
In recent years, individual authors and aca- > 4500 g (OR 2.05, 95% CI 1.53–2.69), and
demic groups have used the Four Ts pneu- clinically diagnosed chorioamnionitis (OR 1.80,
monic to provide a simplistic categorization of 95% CI 1.55–2.09).
the causes of postpartum hemorrhage. This is Surprisingly, it is much more difficult to find
shown in Table 534. comparable studies of risk factors for uterine
23
45
30 August 2006 14:19:10
atony in women achieving vaginal delivery. A show that retained placenta is associated with
single center, case-control study from Pakistan increased blood loss and increased need for
reporting on women who had either assisted or blood transfusion. Stones and colleagues
non-assisted vaginal delivery found only two reported that retained placenta had a RR of 5.15
factors had a strong association with uterine (99% CI 3.36–7.87) for blood loss ≥ 1000 ml
atony: gestational diabetes mellitus (OR 7.6, within the first 24 h of delivery44. Bais and col-
95% CI 6.9–9.0) and prolonged second stage leagues found an incidence of 1.8% for retained
of labor in multiparas (OR 4.0, 95% CI placenta in Holland38. Using multiple regression,
3.1–5.0)36. They found no association with these authors determined that retained placenta
high parity, age, pre-eclampsia, augmentation was associated with an OR of 7.83 (95% CI
of labor, antenatal anemia and a history of poor 3.78–16.22) and 11.73 (95% CI 5.67–24.1) for
maternal or perinatal outcomes. postpartum hemorrhage of ≥ 500 ml and
postpartum hemorrhage ≥ 1000 ml, respectively.
In addition, retained placenta was found to have
Trauma
an OR of 21.7 (95% CI 8.9–53.2) for red cell
Trauma is reported to be the primary cause of transfusion in this Dutch cohort.
postpartum hemorrhage in 20% of cases34 (see Tanberg and colleagues reported an inci-
also Chapter 9). Genital tract trauma at delivery dence of retained placentas of 0.6% in a large
is associated with an odds ratio of 1.7 (95% CI Norwegian cohort of 24 750 deliveries and
1.4–2.1) for postpartum hemorrhage (measured showed that hemoglobin fell by a mean of
blood loss > 1000 ml)37. Similar results were 3.4 g/dl in the retained placental group com-
found in a Dutch study with a reported OR of pared to no fall in the controls45. In addition,
1.82 (CI 1.01–3.28) for postpartum hemor- blood transfusion was required in 10% of the
rhage (≥ 1000 ml) with perineal trauma ≥ first- retained placental group but only 0.5% of the
degree tears38. Trauma to the broad ligament, control group. A similar incidence of retained
uterine rupture, cervical and vaginal tears and placenta was found in a Saudi Arabian case–
perineal tears are all associated with increased control study which demonstrated increased
blood loss at normal vaginal delivery. blood loss in women with a retained placenta
Inversion of the uterus is a rare cause of (mean 437 ml) compared with controls (mean
postpartum hemorrhage (see Chapter 9). The 263 ml)46. A large study from Aberdeen of over
incidence of inversion varies from 1 in 1584 36 000 women reported postpartum hemor-
deliveries in Pakistan39 to around 1 in 25 000 rhage in 21.3% of women with retained pla-
deliveries in the USA, UK and Norway40. Blood centa compared to 3.5% in vaginal deliveries
loss at delivery with a uterine inversion is usually without retained placenta47. Both studies con-
at least 1000 ml41, with 65% of uterine inver- firmed that women with a history of retained
sions being complicated by postpartum hemor- placenta have an increased risk of recurrence
rhage and 47.5% requiring blood transfusion in in subsequent pregnancies46,47. In the study by
a large series of 40 cases42. Adelusi and colleagues, 6.1% of the patients
with retained placenta had a prior history of
retained placenta, compared to none in their
Tissue
control group of normal vaginal deliveries46.
Retained placenta accounts for approximately Placental accreta is a rare and serious compli-
10% of all cases of postpartum hemorrhage34. cation, occurring in about 0.001–0.05% of all
Effective uterine contraction to aid hemostasis deliveries48,49. Makhseed and colleagues found
requires complete expulsion of the placenta. an increasing risk for accreta with increasing
Most retained placentas can be removed manu- numbers of Cesarean sections (OR 4.11, 95%
ally, but rarely the conditions of placenta per- CI 0.83–19.34) after one previous Cesarean
creta, increta, and accreta may be responsible for section and an OR of 30.25 (95% CI 9.9–92.4)
placental retention (see Chapters 24 and 36). after two previous Cesarean sections, compared
Retained placenta occurs after 0.5–3% of deliv- with no previous Cesarean section. Kastner
eries43. Several case–control and cohort studies and colleagues found that placenta accreta was
24
46
30 August 2006 14:19:10
Vital statistics
implicated in 49% of their 48 cases of emer- (95% CI 1.3–7.3) for postpartum hemorrhage
gency hysterectomy50. Zaki and co-workers (defined as visual estimation of ≥ 600 ml)54.
found an incidence of 0.05% of placenta accreta Ijaiya and co-workers in Nigeria found that the
in a population of 23 000 women49. They found risk of postpartum hemorrhage in women > 35
that rates of postpartum hemorrhage and emer- years was two-fold higher compared to women
gency hysterectomy were higher in the accreta < 25 years, although no consideration of con-
group compared to the placenta previa group founding was made in this study55. Rates of
undergoing Cesarean section. Postpartum hem- obstetric hysterectomy have also been reported
orrhage occurred in 91.7% of the accreta group to increase with age; Okogbenin and colleagues
compared to 18.4% of the previa group (OR in Nigeria reported an increase from 0.1% at 20
48.9, 95% CI 5.93–403.25), whereas 50% of years to 0.7% at ≥ 40 years56. However, others
accreta cases required emergency hysterectomy have found no relationship between delaying
compared to 2% in the previa group (OR 48, childbirth and postpartum hemorrhage57.
95% CI 7.93–290.48). Within the accreta
group, 75% of patients had a previous history of
Ethnicity
Cesarean section, compared to 27.5% in the
previa group (OR 7.9, 95% CI 1.98–31.34). Several studies have examined whether ethnic-
ity is a factor for postpartum hemorrhage.
Magann and co-workers, using a definition of
Thrombin postpartum hemorrhage of measured blood loss
Disorders of the clotting cascade and platelet > 1000 ml and/or need for transfusion37, found
dysfunction are the cause of postpartum hemor- Asian race to be a risk factor (OR 1.8, 95%
rhage in 1% of cases34. Known associations with CI 1.4–2.2)). Other studies have observed
coagulation failure include placental abruption, similar findings in Asians58 (OR 1.73, 95% CI
pre-eclampsia, septicemia and intrauterine 1.20–2.49) and Hispanic races (OR 1.66, 95%
sepsis (see Chapter 44), retained dead fetus, CI 1.02–2.69)58 (OR for hematocrit < 26%,
amniotic fluid embolus, incompatible blood 3.99, 95% CI 0.59–9.26)59.
transfusion, abortion with hypertonic saline and
existing coagulation abnormalities4,51,52 (see Body mass index
Chapter 25).
Women who are obese have higher rates of
intrapartum and postpartum complications.
ANTENATAL RISK FACTORS FOR Usha and colleagues performed a population-
PRIMARY POSTPARTUM based observational study of 60 167 deliveries
HEMORRHAGE in South Glamorgan, UK; women with a body
mass index (BMI) > 30 had an OR of 1.5 (95%
Age
CI 1.2–1.8) for blood loss > 500 ml, compared
Increasing maternal age appears to be an inde- to women with a BMI of 20–3060. Stones and
pendent risk factor for postpartum hemorrhage. colleagues reported a RR for major obstetric
In Japan, Ohkuchi and colleagues studied hemorrhage of 1.64 (95% CI 1.24–2.17) when
10 053 consecutive women who delivered a the BMI was 27+44.
singleton infant53. Excessive blood loss (≥ 90th
centile) was defined separately for vaginal and
Parity
Cesarean deliveries (615 ml and 1531 ml,
respectively). On multivariate analysis, age ≥ 35 Although grand multiparity has traditionally
years was an independent risk factor for post- been thought of as risk factor for postpartum
partum hemorrhage in vaginal deliveries (OR hemorrhage, Stones and colleagues and
1.5, 95% CI 1.2–1.9) and Cesarean deliveries Selo-Ojeme did not demonstrate any relation
(OR 1.8, 95% CI 1.2–2.7). In Nigeria, Tsu between grand multiparity and major obstetric
reported that advanced maternal age (≥ 35 hemorrhage44,61. This observation was con-
years) was associated with an adjusted RR of 3.0 firmed in a large Australian study which used
25
47
30 August 2006 14:19:10
multivariate logistic regression analysis and Prolonged pregnancy

found no association between grand multiparity
A large Danish cohort study compared a post-
(≥ five previous births) and postpartum hemor-
term group (gestational age ≥ 42 weeks or
rhage (> 500 ml)62. Tsu reported an association
more) of 77 956 singleton deliveries and a term
with low parity (0–1 previous birth) with
group of 34 140 singleton spontaneous deliver-
adjusted RR without intrapartum factors of
ies69. Adjusted odds ratio for postpartum
1.7 (95% CI 1.1–2.7) and adjusted RR with
hemorrhage was 1.37 (95% CI 1.28–1.46),
intrapartum factors of 1.5 (95% CI 0.95–2.5)
suggesting an association between prolonged
but not with grand multiparity (defined as five
pregnancy and postpartum hemorrhage.
or more births)54. Ohkuchi also found primi-
parity to be associated with excessive blood loss
at vaginal delivery (OR 1.6, 95% CI 1.4–1.9)53. Fetal macrosomia
Studies from Pakistan63 and Nigeria55 have
reported an association between grand multi- Several studies confirm that fetal macrosomia is
parity and postpartum hemorrhage, but both associated with postpartum hemorrhage. Jolly
studies failed to account for other confounding and colleagues examined 350 311 completed
factors such as maternal age. singleton pregnancies in London70. Linear
regression analysis suggested that a birth weight
> 4 kg was better at predicting maternal mor-
bidity than birth weight > 90th centile. Post-
Other medical conditions partum hemorrhage was increased in women
Several medical conditions are associated with with fetal macrosomia (OR 2.01; 95% CI
postpartum hemorrhage. Women with type II 1.93–2.10). In a large cohort of 146 526
diabetes mellitus have an increased incidence of mother–infant pairs in California, Stotland and
postpartum hemorrhage of > 500 ml (34%) co-workers also demonstrated an adjusted OR
compared to the non-diabetic population for postpartum hemorrhage of 1.69 (95% CI
(6%)64,65. Connective tissue disorders such as 1.58–1.82) in infants of 4000–4499 g compared
Marfans and Ehlers-Danlos syndrome have also to 2.15 (95% CI 1.86–2.48) and 2.03 (95% CI
been associated with postpartum hemor- 1.33–3.09) with weights of 4500–4999 g and
rhage66,67. Blood loss at delivery is also ≥ 5000 g, respectively71. In Nigeria, a case–
increased with inherited coagulopathies52. The control study of 351 infants weighing > 4 kg
most common inherited hemorrhagic disorder with 6563 term infants found an incidence
is von Willebrand’s disease, with a reported of postpartum hemorrhage of 8.3% and
prevalence of between 1 and 3%. Most (70%) 2.1%, respectively72. Bais and colleagues, in
have Type 1 disease characterized by low their Dutch study, also demonstrated an
plasma levels of factor VIII, von Willebrand fac- increase in risk for postpartum hemorrhage
tor antigen, and von Willebrand factor activity. (≥ 500 ml) and severe postpartum hemorrhage
Less common inherited bleeding disorders (≥ 1000 ml) with infants with weights ≥ 4 kg
include carriage of hemophilia A (factor VIII (OR 2.11, 95% CI 1.62–2.76 and 2.55, 95%
deficiency) or hemophilia B (factor IX defi- CI 1.5–4.18)38.
ciency) and factor XI deficiency. In their review,
Economaides and colleagues suggest that the
Multiple pregnancies
risks of primary postpartum hemorrhage in
patients with von Willebrand’s disease, factor Epidemiological studies suggest twins and
XI deficiency, and carriers of hemophilia are higher-order pregnancies are at increased risk for
22%, 16%, and 18.5%, respectively, compared postpartum hemorrhage. Walker and co-workers
with 5% in the general obstetric population52. conducted a retrospective cohort study involving
James also reviewed the numerous case series 165 188 singleton pregnancies and 44 674 multi-
and the more limited case–control studies of ple pregnancies in Canada73. Multiple pregnan-
women with bleeding disorders and came to cies were associated with an increased risk for
similar conclusions68 (see Chapter 25). postpartum hemorrhage (RR 1.88, 95% CI
26
48
30 August 2006 14:19:11
Vital statistics
1.81–1.95), hysterectomy (RR 2.29, 95% CI excess blood loss at the time of vaginal and
1.66–3.16) and blood transfusion (RR 1.67, Cesarean delivery, respectively53. This study
95% CI 1.13–2.46). Several other studies have also reported that placenta previa was associ-
estimated the RR of postpartum hemorrhage ated with an OR of 6.3 (95% CI 4.0–9.9) for
associated with multiple pregnancies to be excessive blood loss at Cesarean delivery.
between 3.0 and 4.544,58,74. Bais and colleagues,
in a Dutch population-based cohort study of
3464 women, used multiple regression analysis Previous history of postpartum
and found that the OR for postpartum hemor- hemorrhage
rhage ≥ 500 ml for multiple pregnancy was 2.6 Magann and colleagues found previous post-
(95% CI 1.06–-6.39)38. Albrecht and co-workers partum hemorrhage to be associated with
conducted a retrospective review of 57 triplet an increased risk for subsequent postpartum
deliveries and found an incidence of 12.3% for hemorrhage (OR 2.2, 95% CI 1.7–2.9)37.
postpartum hemorrhage requiring transfusion75,
and a case series of 71 quadruplet pregnancies
conducted by Collins and colleagues estimated Previous Cesarean delivery
that the frequency of postpartum hemorrhage The Japanese study demonstrated an odds ratio
and transfusion to be 21% (95% CI 11–31%) of 3.1 (95% CI 2.1–4.4) for excessive blood loss
and 13% 95% CI 5–21%), respectively76. at vaginal delivery in women with a previous
Magann and colleagues demonstrated an OR for Cesarean section53.
postpartum hemorrhage of 2.2 (95% CI 1.5–3.2)
in multiple pregnancies37, and Stones and col-
leagues showed a relative risk of 4.46 (95% CI INTRAPARTUM RISK FACTORS
3.01–6.61) for obstetric hemorrhage with FOR PRIMARY POSTPARTUM
multiple pregnancies44. HEMORRHAGE
Induction of labor
Fibroids
Meta-analysis of trials of induction of labor at or
Obstetric textbooks suggest that leiomyomas beyond term indicates that induction does not
can be a cause of postpartum hemorrhage. This increase Cesarean section or operative vaginal
is mainly based on case reports77, but one delivery rates78. However, this meta-analysis did
cohort study of 10 000 women in Japan found not examine blood loss at delivery. Epidemio-
that women with leiomyomas had an OR of 1.9 logical studies suggest a link between induction
(95% CI 1.2–3.1) and 3.6 (95% CI 2.0–6.3) for of labor and postpartum hemorrhage. Brinsden
excessive blood loss at vaginal and Cesarean and colleagues reviewed 3674 normal deliveries
delivery, respectively53. and found that the incidence of postpartum
hemorrhage was increased after induction of
labor79; among primipara, the incidence was
Antepartum hemorrhage
nearly twice that of spontaneous labor, even
Antepartum hemorrhage has been linked to when only normal deliveries were considered.
postpartum hemorrhage risk with an OR of 1.8 The study of Magann and colleagues suggested
(95% CI 1.3–2.3)37. Stones and co-workers an OR of 1.5 (95% CI 1.2–1.7) for postpartum
found a RR for major obstetric hemorrhage hemorrhage after induction of labor37 and Bais
(> 1000 ml) of 12.6 (95% CI 7.61–20.9), 13.1 and co-workers found an OR of 1.74 (95% CI
(95% CI 7.47–23) and 11.3 (95% CI 1.06–2.87) for severe postpartum hemorrhage
3.36–38.1) for proven abruption, previa with of > 1000 ml after induction of labor38.
bleeding, and previa with no bleeding, respec- Tylleskar and colleagues performed a pro-
tively44. Ohkuchi and colleagues, in their spective, randomized, control trial of term
10 000 women, demonstrated that a low-lying induction of labor with amniotomy plus
placenta was associated with odds ratios of 4.4 oxytocin versus waiting for spontaneous labor
(95% CI 2.2–8.6) and 3.3 (95% CI 1.4–7.9) for in 84 women and found no difference in the
27
49
06 September 2006 16:51:58
amount of bleeding at the third stage80. A CI 0.22–12.19), or with oxytocin (two trials,
Cochrane review81 of amniotomy versus vaginal 245 women, RR 0.51, 95% CI 0.16–1.66).
prostaglandin for induction of labor reported Finally, a review of vaginal PGE2 for induction
no difference in postpartum hemorrhage rates. of labor suggested an increased risk of post-
Another Cochrane82 review of amniotomy plus partum hemorrhage compared to placebo88
intravenous oxytocin included only one (eight studies, 3437 women, RR 1.44, 95% CI
placebo-controlled trial, but no data on post- 1.01–2.05).
partum hemorrhage were reported. This review
compared amniotomy plus intravenous oxy-
tocin against vaginal prostaglandin (two trials, Duration of labor
160 women) and found a higher rate of First stage
postpartum hemorrhage in the amniotomy/
oxytocin group (13.8% vs. 2.5% respectively, Compared with the second stage of labor, lim-
RR 5.5, 95% CI 1.26–24.07)82. ited evidence is available regarding the influence
A review of intravenous oxytocin alone for of the duration of the first stage of labor on
cervical ripening83 found no difference in postpartum hemorrhage89. Magann and col-
postpartum hemorrhage rates compared to the leagues defined a prolonged first stage of labor
placebo/expectant management group (three as a latent phase of > 20 h in nulliparous and
trials, 2611 women; RR 1.24, 95% CI > 14 h in multiparous and/or an active phase of
0.85–1.81) or vaginal PGE2 (four trials, 2792 < 1.2 cm per hour in nulliparous and < 1.4 cm
women; RR 1.02, 95% CI 0.75–-1.4). Use of in multiparous patients37. These investigators
mechanical methods to induce labor84 was not found an OR of 1.6 for prolonged first stage of
associated with any difference in postpartum labor but the 95% CI ranged from 1 to 1.6.
hemorrhage rates when compared to placebo
(one study, 240 women, RR 0.46, 95% CI
Second stage
0.09–2.31), prostaglandin vaginal PGE2 (one
study, 60 women, RR 3.0, 95% CI 0.33–27.24), Several large studies have explored the relation-
intracervical PGE2 (three studies, 3339 women, ship between the length of the second stage
RR 0.91, 95% CI 0.40–2.11), misoprostol (one and adverse maternal and neonatal outcomes.
study, 248 women, RR 2.34, 95% CI Cohen analyzed obstetric data from 4403
0.46–11.85) or to oxytocinon alone (one study, nulliparas and found an increase in postpartum
60 patients, RR 1.0, 95% CI 0.22–4.56). hemorrhage rate after more than 3 h in the
Meta-analysis85 of trials of membrane sweep- second stage90. He attributed this to the
ing for induction of labor found a reduction in increased need for mid-forceps delivery. A large
postpartum hemorrhage compared to no inter- retrospective study involving 25 069 women in
vention (three trials, 278 women, RR 0.31, 95% spontaneous labor at term with a cephalic pre-
CI 0.11–0.89). A review of oral misoprostol for sentation found that second-stage duration had
induction of labor86 did not include any trial a significant independent association with the
that compared this agent with placebo. How- risk of postpartum hemorrhage91. A more recent
ever, one trial reported in this review, involving retrospective cohort study of 15 759 nulliparous
692 women and using PGE2 in the control arm, term, cephalic singleton births in San Francisco
found no difference in postpartum hemorrhage divided the second stage of labor into 1-h inter-
rate (RR 0.98, 95% CI 0.73–1.31). Other vals92. Postpartum hemorrhage was defined as
reviews of induction of labor methods have estimated blood loss of > 500 ml after vaginal
reported no difference in postpartum hemor- delivery or > 1000 ml after Cesarean delivery.
rhage rates between vaginal misoprostol when The frequency of postpartum hemorrhage
compared to placebo (two trials, 107 women, increased from 7.1% when the second stage
RR 0.91, 95% CI 0.13–6.37)87, vaginal prosta- lasted 0–1 h to 30.9% when it lasted > 4 h. The
glandins (five trials, 1002 women, RR 0.88, risk for postpartum hemorrhage with a second
95% CI 0.63–1.22), intracervical prosta- stage of > 3 h remained statistically significant
glandins (two trials, 172 women, RR 1.62, 95% when controlled for confounders (including
28
50
30 August 2006 14:19:12
Vital statistics
operative vaginal delivery, episiotomy, birth 95% CI 4.6–8.2). Using receiver operating char-
weight and fetal position) (OR 1.48, 95% CI acteristic (ROC) curves, the best predictor for
1.24–1.78). Myles and colleagues examined postpartum hemorrhage was a third stage of
6791 cephalic singleton births and found that ≥ 18 min97. Similarly, a Dutch population-based
the incidence of postpartum hemorrhage was cohort study of 3464 nulliparous women sugges-
2.3% in women experiencing a second stage ted that a third stage of ≥ 30 min was associated
< 2 h compared to 6.2% in women with a with a blood loss of ≥ 500 ml (OR 2.61, 95% CI
longer second stage93. Janni and co-workers 1.83–3.72) and ≥ 1000 ml (OR 4.90, 95%
compared 952 women with a singleton cephalic CI 2.89–8.32)38. Blood loss was determined by
pregnancy after 34 weeks’ gestation with a ‘nor- a combination of measurement and visual
mal’ second stage to 248 women with a second estimation.
stage > 2 h94. The median difference between
intrapartum and postpartum hemoglobin levels
Analgesia
was lower in the normal group (−0.79 g/dl)
compared to the prolonged second-stage group A retrospective case–control study involving
(−1.84 g/dl) Multivariate binary logistic regres- 1056 and 6261 women with and without epi-
sion confirmed duration of the second stage as dural analgesia, respectively, found that use of
an independent predictor of postpartum hemor- epidural analgesia was associated with intrapar-
rhage (RR 2.3, 95% CI 1.6–3.3). Magann and tum hemorrhage > 500 ml98. Magann and col-
colleagues also found an OR of 1.6 (95% CI leagues also found an OR of 1.3 for postpartum
1.1–2.1) for prolonged second stage37. hemorrhage with epidural analgesia, but the 95%
CI extended from 1 to 1.637. However, if Cesar-
ean delivery is required, regional analgesia is
Third stage
superior to general anesthesia in reducing blood
Strong evidence indicates that, despite the use of loss, according to evidence from one random-
active management, prolongation of the third ized, controlled trial involving 341 women99.
stage of labor increases the risk for postpartum
hemorrhage. Combs and colleagues studied
Delivery method
12 979 singleton, vaginal deliveries and found
that the median duration of the third stage was The NICE guideline of the UK on Cesarean sec-
6 min (interquartile range 4–10 min)95. The tion examined maternal morbidity in a compari-
incidence of postpartum hemorrhage and blood son of planned Cesarean section with planned
transfusion remaining constant until the third vaginal birth from available randomized, con-
stage reached 30 min (3.3% of deliveries). There- trolled trials on an intention-to-treat basis100.
after, it increased progressively, reaching a pla- For maternal obstetric hemorrhage (defined as
teau at 75 min95. Dombrowski and colleagues blood loss > 1000 ml), an absolute risk of 0.5%
studied the third stage in 45 852 singleton deliv- for planned Cesarean section and 0.7% for vagi-
eries ≥ 20 weeks’ gestation96. Postpartum hem- nal birth (RR 0.8, 95% CI 0.4–4.4) was
orrhage was defined as an estimated blood loss reported, suggesting there is no difference in risk.
≥ 500 ml. At all gestational ages, the frequency Magann and colleagues examined the inci-
of postpartum hemorrhage increased with in- dence and risk factors for postpartum hemor-
creasing duration of the third stage, reaching the rhage in 1844 elective Cesarean sections and
peak at 40 min. Magann and colleagues per- 2933 non-elective Cesarean sections101. Two cri-
formed a prospective observational study of 6588 teria were used to define postpartum hemor-
vaginal deliveries97. Postpartum hemorrhage was rhage: measured blood loss > 1000 ml and/or
defined as a blood loss > 1000 ml or hemodyna- need for blood transfusion and measured blood
mic instability requiring blood transfusion. Post- loss > 1500 ml and/or need for blood trans-
partum hemorrhage risk was significant (and fusion. Six percent of all Cesarean deliveries
increased in a dose-related fashion with time) at were complicated by a blood loss > 1000 ml.
10 min (OR 2.1, 95% CI 1.6–2.6), 20 min (OR The postpartum hemorrhage rates for elective
4.3, 95% CI 3.3–5.5) and at 30 min (OR 6.2, Cesarean section (blood loss > 1000 ml –
29
51
30 August 2006 14:19:12
4.84%, blood loss > 1500 ml – 1.9%) were use of a mid-line or mediolateral episiotomy is
lower than for non-elective Cesarean delivery associated with increased blood loss and post-
(6.75% and 3.04%, respectively). During the partum hemorrhage in most studies, with blood
4-year period of this study, there were 13 868 loss increases of between 300 and 600 ml com-
vaginal deliveries with a postpartum hemorrhage pared with no episiotomy105,106. Stones and
rate of 5.15% (blood loss > 1000 ml) and 2.4% colleagues reported a relative risk of 2.06 (95%
(blood loss > 1500 ml)101. No data on operative CI 1.36–3.11) for postpartum hemorrhage
vaginal delivery rate were reported. Although the when episiotomy occurred44. Bais and co-
postpartum hemorrhage rate was higher in workers reported similar results with an OR of
women undergoing non-elective Cesarean deliv- 2.18 (95% CI 1.68–-2.81)38, and Combs and
ery than after vaginal delivery, the difference in colleagues reported that a mediolateral episio-
rate for elective Cesarean delivery was not statis- tomy is associated with an odds ratio of 4.67
tically significant different. Using linear regres- (95% CI 2.59–-8.43) for postpartum hemor-
sion, risk factors for postpartum hemorrhage at rhage58. However, one recent randomized, con-
elective Cesarean delivery were leiomyomas, pla- trolled trial of the use of episiotomy when
centa previa, preterm birth and general anesthe- perineal tears appear imminent suggested no
sia. For non-elective Cesarean delivery, risk difference in postpartum hemorrhage rates107.
factors were blood disorders, retained placenta,
antepartum transfusion, antepartum/intra-
partum hemorrhage, placenta previa, general Chorioamnionitis
anesthesia, and macrosomia.
Several studies have reported an increased risk
Combs and colleagues performed a case–
for postpartum hemorrhage in the presence
control study involving 3052 Cesarean deliver-
of chorioamnionitis, ORs ranging from 1.3
ies102. They reported a postpartum hemorrhage
(95% CI 1.1–1.7) at vaginal birth37 to 2.69
incidence (based on fall in hematocrit and/or
(95% CI 1.44–5.03) at Cesarean section102 (see
need for blood transfusion) of 6.4% for Cesar-
Chapter 44).
ean delivery, similar to Magann and colleagues.
However, Combs and colleagues did not differ-
entiate elective from non-elective deliveries.
CONCLUSIONS
This group also examined 9598 vaginal
deliveries and found an overall incidence of Postpartum hemorrhage remains an extremely
postpartum hemorrhage of 3.9%58. Using important cause of maternal mortality and mor-
multiple linear regression, they reported an bidity throughout the world. Sadly substandard
adjusted OR of 1.66 (95% CI 1.06–2.60) for care continues to contribute to mortality and
forceps or vacuum extraction use, suggesting morbidity from postpartum hemorrhage, regard-
that operative vaginal delivery is associated with less of the country in which death takes place.
postpartum hemorrhage. In addition, the use of Major obstetric hemorrhage complicates
sequential instruments (forceps after unsuccess- around 10% of live births and is responsible
ful vacuum extraction) to achieve vaginal for 28% of direct deaths, globally. Marked dif-
delivery is a further risk factor (OR 1.9, 95% ferences exist between countries; in the UK
CI 1.1–3.2)37 or relative risk of 1.6 (95% CI, there are five deaths per million maternities,
1.3–2.0)103 for postpartum hemorrhage. whereas the figure is 100 times higher in parts of
Africa. Severe obstetric hemorrhage is increas-
ingly used as a measure of quality of health care
Episiotomy
in women. In the UK, severe obstetric hemor-
A Cochrane review argues for restrictive use rhage occurs in three to seven cases per 1000
of episiotomy because this policy is associated livebirths, with postpartum hemorrhage impli-
with fewer complications104. Surprisingly, this cated in 70% of cases. In contrast, rates as high
meta-analysis does not address the question of as 30.5 per 1000 livebirths are reported in parts
postpartum hemorrhage incidence with episio- of Africa, with postpartum hemorrhage rates of
tomy. Iatrogenic trauma by the indiscriminate 17.4 per 1000.
30
52
30 August 2006 14:19:12
Vital statistics
References 13. Hill JA, Fadel HE, Nelson MC, Nelson RM,
Nelson GH. Blood loss at vaginal delivery.
1. Park EH, Sachs BP. Postpartum hemorrhage South Med J 1986;79:188–192.
and other problems of the third stage. In James 14. Prasertcharoensuk W, Swadpanich U, Lumbi-
DK, Steer PJ, Weiner CP, Gonik B, eds. High ganon P. Accuracy of the blood loss estimation
Risk Pregnancy: Management Options. London: in the third stage of labor. Int J Gynaecol Obstet
WB Saunders, 1999;1231–46 2000;71:69–70
2. Pritchard JA, Baldwin RM, Dickey JC, et al. 15. Razvi K, Chua S, Arulkumaran S, Ratnam SS.
Red blood cell loss and changes in apparent A comparison between visual estimation and
blood volume during and following vaginal laboratory determination of blood loss during
delivery, caesarean section and caesarean the third stage of labour. Aust N Z J Obstet
section plus total hysterectomy. Am J Obstet Gynaecol 1996;36:152–4
Gynecol 1962;84:1271 16. Irons DW, Sriskandabalan P, Bullough CH. A
3. Brace V, Penney GC. Scottish Confidential simple alternative to parenteral oxytocics for the
Audit of Severe Maternal Morbidity: First third stage of labor. Int J Gynaecol Obstet 1994;
Annual Report 2003. 22, 5–31. 2005. 46:15–18
Aberdeen, Scottish Programme for Clinical 17. Prendiville WJ, Elbourne D, McDonald S.
Effectiveness in Reproductive Health Active versus expectant management in the
4. Griffiths D, Howell C. Massive obstetric third stage of labour (update of Cochrane Data-
haemorrhage. In Johanson R, Cox C, Grady K, base Syst Rev. 2000;(2):CD000007; PMID:
Howell C, eds. Managing Obstetric Emergencies 10796082). (Review). Cochrane Database of
and Trauma (MOET) course manual. London: Systematic Reviews 2000;CD000007
RCOG Press, 2003:151–62 18. Grimes DA, Schulz KF. Bias and causal
5. Grady K, Cox C. Shock. In Johanson R, Cox associations in observational research. Lancet
C, Grady K, Howell C, eds. Managing Obstetric 2002;359:248–52
Emergencies and Trauma. London: RCOG 19. Katz MH. Multivariable analysis: A practical
Press, 2003:81–90 guide for clinicians. Cambridge: Cambridge
6. Gulmezoglu AM, Hofmeyr GJ. Prevention and University Press, 1999
treatment of postpartum haemorrhage. In 20. Grimes DA, Schulz KF. Clinical research
MacLean AB, Neilson J, eds. Maternal in obstetrics and gynecology: a Baedeker for
Morbidity and Mortality. London: RCOG Press, busy clinicians. Obstet Gynecol Survey 2002;57:
2002:241–51 S35–S53
7. Strand RT, da Silva F, Bergstrom S. Use of 21. Mamdani M, Sykora K, Li P, et al. Reader’s
cholera beds in the delivery room: a simple and guide to critical appraisal of cohort studies. 2.
appropriate method for direct measurement of Assessing potential for confounding. BMJ
postpartum bleeding. Trop Doctor 2003;33: 2005;330:960–2
215–16 22. Anonymous. Introduction. In Lewis G, ed.
8. Chua S, Ho LM, Vanaja K, Nordstrom L, Roy Why Mothers Die 2000–2002. London: RCOG,
AC, Arulkumaran S. Validation of a laboratory 2004:1–24
method of measuring postpartum blood loss. 23. Hall M. Haemorrhage. In Lewis G, ed. Why
Gynecol Obstet Invest 1998;46:31–3 Mothers Die 2000–2002. London: RCOG Press,
9. Duthie SJ, Ven D, Yung GL, Guang DZ, Chan 2004:86–93
SY, Ma HK. Discrepancy between laboratory 24. Chang J, Elam-Evans LD, Berg CJ, et al.
determination and visual estimation of blood Pregnancy-related mortality surveillance,
loss during normal delivery. Eur J Obstet United States, 1991–1999. CDC. http//www.
Gynecol Reprod Biol 1991;38:119–24 cdc.gov/mmwr/preview/mmwrhtml/ss5202a1.
10. Glover P. Blood loss at delivery: how accurate is htm 52(SS02), 1–8. 2003
your estimation? Aust J Midwifery 2003;16:21–4 25. Berg CJ, Atrash HK, Koonin LM, Tucker M.
11. Higgins PG. Measuring nurses’ accuracy of Pregnancy-related mortality in the United
estimating blood loss. J Adv Nursing 1982;7: States, 1987–1990. Obstet Gynecol 1996;88:
157–62 161–7
12. Newton M, Mosey LM, Egli GE, Gifford WB, 26. Bouvier-Colle MH, Varnoux N, Breart G.
Hull CT. Blood loss during and immediately Maternal deaths and substandard care: the
after delivery. Obstet Gynecol 1961;17:9–18 results of a confidential survey in France.
31
53
30 August 2006 14:19:12
Medical Experts Committee. Eur J Obstet 39. Hussain M, Jabeen T, Liaquat N, Noorani K,
Gynecol Reproduct Biol 1995;58:3–7 Bhutta SZ. Acute puerperal uterine inversion.
27. Bouvier-Colle MH, Ouedraogo C, Dumont A, J Coll Phys Surg–Pakistan 2004;14:215–17
et al. Maternal mortality in West Africa. Rates, 40. Milenkovic M, Kahn J. Inversion of the uterus:
causes and substandard care from a prospective a serious complication at childbirth. Acta Obstet
survey. Acta Obstet Gynecol Scand 2001;80: Gynecol Scand 2005;84:95–6
113–19 41. Beringer RM, Patteril M. Puerperal uterine
28. Spies CA, Bam RH, Cronje HS, Schoon MG, inversion and shock. Br J Anaesthes 2004;
Wiid M, Niemand I. Maternal deaths in 92:439–41
Bloemfontein, South Africa, 1986–1992. South 42. Baskett TF. Acute uterine inversion: a review of
Afri Med J 1995;85:753–5 40 cases. J Obstet Gynaecol Can 2002;24:953–6
29. Waterstone M, Bewley S, Wolfe C. Incidence 43. Weeks AD, Mirembe FM. The retained pla-
and predictors of severe obstetric morbidity: centa – new insights into an old problem. Eur J
case–control study. BMJ 2001;322:1089–93 Obstet Gynecol Reprod Biol 2002;102:109–10
30. Wen SW, Huang L, Liston R, et al. Severe 44. Stones RW, Paterson CM, Saunders NJ. Risk
maternal morbidity in Canada, 1991–2001. factors for major obstetric haemorrhage. Eur J
Can Med Assoc J 2005;173:759–64 Obstet Gynecol Reprod Biol 1993;48:15–18
31. Filippi V, Ronsmans C, Gohou V, et al. 45. Tandberg A, Albrechtsen S, Iversen OE.
Maternity wards or emergency obstetric rooms? Manual removal of the placenta. Incidence and
Incidence of near-miss events in African clinical significance. Acta Obstet Gynecol Scand
hospitals. Acta Obstet Gynecol Scand 2005;84: 1999;78:33–6
11–16 46. Adelusi B, Soltan MH, Chowdhury N,
32. Prual A, Bouvier-Colle MH, de Bernis L, Breart Kangave D. Risk of retained placenta: multi-
G. Severe maternal morbidity from direct variate approach. Acta Obstet Gynecol Scand
obstetric causes in West Africa: incidence and 1997;76:414–18
case fatality rates. Bull WHO 2000;78: 593–602 47. Hall MH, Halliwell R, Carr-Hill R. Concomi-
33. Pattinson RC, Hall M. Near misses: a useful tant and repeated happenings of complications
adjunct to maternal death enquiries. Br Med of the third stage of labour. Br J Obstet Gynaecol
Bull 2003;67:231–43 1985;92:732–8
34. Anderson J, Etches D, Smith D. Postpartum 48. Makhseed M, el-Tomi N, Moussa M. A retro-
haemorrhage. In Damos JR, Eisinger SH, eds. spective analysis of pathological placental
Advanced Life Support in Obstetrics (ALSO) implantation – site and penetration. Int J
provider course manual. Kansas: American Gynaecol Obstet 1994;47:127–34
Academy of Family Physicians, 2000:1–15 49. Zaki ZM, Bahar AM, Ali ME, Albar HA,
35. Rouse DJ, Leindecker S, Landon M, et al. The Gerais MA. Risk factors and morbidity in
MFMU Cesarean Registry: uterine atony after patients with placenta previa accreta compared
primary cesarean delivery. Am J Obstet Gynecol to placenta previa non-accreta. Acta Obstet
2005;193:1056–60 Gynecol Scand 1998;77:391–4
36. Feerasta SH, Motiei A, Motiwala S, Zuberi 50. Kastner ES, Figueroa R, Garry D, Maulik D.
NF. Uterine atony at a tertiary care hospital in Emergency peripartum hysterectomy: experi-
Pakistan: a risk factor analysis. J Pak Med Assoc ence at a community teaching hospital. Obstet
2000;50:132–6 Gynecol 2002;99:971–5
37. Magann EF, Evans S, Hutchinson M, Collins 51. Walker ID, Walker JJ, Colvin BT, Letsky EA,
R, Howard BC, Morrison JC. Postpartum Rivers R, Stevens R. Investigation and manage-
hemorrhage after vaginal birth: an analysis of ment of haemorrhagic disorders in pregnancy.
risk factors. S Med J 2005;98:419–22 Haemostasis and Thrombosis Task Force.
38. Bais JM, Eskes M, Pel M, Bonsel GJ, Bleker J Clin Pathol 1994;47:100–8
OP. Postpartum haemorrhage in nulliparous 52. Economides DL, Kadir RA, Lee CA. Inherited
women: incidence and risk factors in low and bleeding disorders in obstetrics and gynaecol-
high risk women. A Dutch population-based ogy. Br J Obstet Gynaecol 1999;106:5–13
cohort study on standard (> or = 500 ml) and 53. Ohkuchi A, Onagawa T, Usui R, et al. Effect of
severe (> or = 1000 ml) postpartum haemor- maternal age on blood loss during parturition:
rhage. Eur J Obstet Gynecol Reprod Biol a retrospective multivariate analysis of 10,053
2004;115:166–72 cases. J Perinat Med 2003;31:209–15
32
54
30 August 2006 14:19:13
Vital statistics
54. Tsu VD. Postpartum haemorrhage in Zimba- of bleeding disorders. Haemophilia 2005;11:
bwe: a risk factor analysis. Br J Obstet Gynaecol 295–307
1993;100:327–33 69. Olesen AW, Westergaard JG, Olsen J. Perinatal
55. Ijaiya MA, Aboyeji AP, Abubakar D. Analysis and maternal complications related to post-
of 348 consecutive cases of primary postpartum term delivery: a national register-based study,
haemorrhage at a tertiary hospital in Nigeria. 1978–1993. Am J Obstet Gynecol 2003;189:
J Obstet Gynaecol 2003;23:374–7 222–7
56. Okogbenin SA, Gharoro EP, Otoide VO, 70. Jolly MC, Sebire NJ, Harris JP, Regan L,
Okonta PI. Obstetric hysterectomy: fifteen Robinson S. Risk factors for macrosomia and
years’ experience in a Nigerian tertiary centre. its clinical consequences: a study of 350,311
J Obstet Gynaecol 2003;23:356–9 pregnancies. Eur J Obstet Gynecol Reprod Biol
57. Roberts CL, Algert CS, March LM. Delayed 2003;111:9–14
childbearing – are there any risks? Med J Aust 71. Stotland NE, Caughey AB, Breed EM, Escobar
1994;160:539–44 GJ. Risk factors and obstetric complications
58. Combs CA, Murphy EL, Laros RK Jr. Factors associated with macrosomia. Int J Gynaecol
associated with postpartum hemorrhage with Obstet 2004;87:220–6
vaginal birth. Obstet Gynecol 1991;77:69–76 72. Fakeye O. The incidence, sociobiological fac-
59. Petersen LA, Lindner DS, Kleiber CM, tors and obstetric complications associated with
Zimmerman MB, Hinton AT, Yankowitz J. large infants at Ilorin, Nigeria. Int J Gynaecol
Factors that predict low hematocrit levels in the Obstet 1988;27:343–7
postpartum patient after vaginal delivery. Am 73. Walker MC, Murphy KE, Pan S, Yang Q, Wen
J Obstet Gynecol 2002;186:737–44 SW. Adverse maternal outcomes in multifetal
60. Usha KT, Hemmadi S, Bethel J, Evans J. pregnancies. Br J Obstet Gynaecol 2004;111:
Outcome of pregnancy in a woman with an 1294–6
increased body mass index. Br J Obstet Gynaecol 74. Klapholz H. Blood transfusion in contemporary
2005;112:768–72 obstetric practice. Obstet Gynecol 1990;75:
61. Selo-Ojeme DO, Okonofua FE. Risk factors 940–3
for primary postpartum haemorrhage. A case 75. Albrecht JL, Tomich PG. The maternal and
control study. Arch Gynecol Obstet 1997;259: neonatal outcome of triplet gestations. Am J
179–87 Obstet Gynecol 1996;174:1551–6
62. Humphrey MD. Is grand multiparity an inde- 76. Collins MS, Bleyl JA. Seventy-one quadruplet
pendent predictor of pregnancy risk? A retro- pregnancies: management and outcome. Am J
spective observational study. Med J Aust 2003; Obstet Gynecol 1990;162:1384–91
179:294–6 77. Akrivis C, Varras M, Bellou A, Kitsiou E,
63. Munim S, Rahbar MH, Rizvi M, Mushtaq N. Stefanaki S, Antoniou N. Primary postpartum
The effect of grandmultiparity on pregnancy haemorrhage due to a large submucosal non-
related complications: the Aga Khan University pedunculated uterine leiomyoma: a case report
experience. J Pak Med Assoc 2000;50:54–8 and review of the literature. Clin Exp Obstet
64. Dunne F, Brydon P, Smith K, Gee H. Gynecol 2003;30:156–8
Pregnancy in women with Type 2 diabetes: 12 78. Crowley P. Interventions for preventing or
years outcome data 1990–2002. Diabet Med improving the outcome of delivery at or beyond
2003;20:734–8 term (Review). Cochrane Database of Systematic
65. Dunne F. Type 2 diabetes and pregnancy. Reviews 2000;CD000170
Semin Fetal Neonat Med 2005;10:333–9 79. Brinsden PR, Clark AD. Postpartum haemor-
66. Rahman J, Rahman FZ, Rahman W, rhage after induced and spontaneous labour. Br
al-Suleiman SA, Rahman MS. Obstetric and Med J 1978;2:855–6
gynecologic complications in women with Mar- 80. Tylleskar J, Finnstrom O, Leijon I, Hedenskog
fan syndrome. J Reprod Med 2003;48: 723–8 S, Ryden G. Spontaneous labor and elective
67. Lind J, Wallenburg HC. Pregnancy and the induction – a prospective randomized study.
Ehlers-Danlos syndrome: a retrospective study I. Effects on mother and fetus. Acta Obstet
in a Dutch population. Acta Obstet Gynecol Gynecol Scand 1979;58:513–18
Scand 2002;81:293–300 81. Bricker L, Luckas M. Amniotomy alone for
68. James AH. More than menorrhagia: a review of induction of labour. (Review). Cochrane Data-
the obstetric and gynaecological manifestations base of Systematic Reviews 2000;CD002862
33
55
30 August 2006 14:19:13
82. Howarth GR, Botha DJ. Amniotomy plus of labor on maternal and fetal outcome. Acta
intravenous oxytocin for induction of labour. Obstet Gynecol Scand 2002;81:214–21
(Review). Cochrane Database of Systematic 95. Combs CA, Laros RK Jr. Prolonged third stage
Reviews 2001;CD003250 of labor: morbidity and risk factors. Obstet
83. Kelly AJ, Tan B. Intravenous oxytocin alone Gynecol 1991;77:863–7
for cervical ripening and induction of labour. 96. Dombrowski MP, Bottoms SF, Saleh AA,
(Review). Cochrane Database of Systematic Hurd WW, Romero R. Third stage of labor:
Reviews 2001;CD003246 analysis of duration and clinical practice. Am J
84. Boulvain M, Kelly A, Lohse C, Stan C, Irion O. Obstet Gynecol 1995;172:1279–84
Mechanical methods for induction of labour. 97. Magann EF, Evans S, Chauhan SP, Lanneau
(Review). Cochrane Database of Systematic G, Fisk AD, Morrison JC. The length of the
Reviews 2001;CD001233 third stage of labor and the risk of postpartum
85. Boulvain M, Stan C, Irion O. Membrane hemorrhage. Obstet Gynecol 2005;105:290–3
sweeping for induction of labour.[update of 98. Ploeckinger B, Ulm MR, Chalubinski K,
Cochrane Database Syst Rev. 2001;(2): Gruber W. Epidural anaesthesia in labour:
CD000451; PMID: 11405964]. (Review). influence on surgical delivery rates, intrapartum
Cochrane Database of Systematic Reviews fever and blood loss. Gynecol Obstet Invest
2001;CD000451 1995;39:24–7
86. Alfirevic Z. Oral misoprostol for induction 99. Lertakyamanee J, Chinachoti T, Tritrakarn T,
of labour.[update of Cochrane Database Syst Muangkasem J, Somboonnanonda A, Kolatat
Rev. 2000;(4):CD001338; PMID: 11034716]. T. Comparison of general and regional anesthe-
(Review). Cochrane Database of Systematic sia for cesarean section: success rate, blood loss
Reviews 2001;CD001338 and satisfaction from a randomized trial. J Med
87. Hofmeyr GJ, Gulmezoglu AM. Vaginal miso- Assoc Thailand 1999;82:672–80
prostol for cervical ripening and induction 100. Anonymous. Women – centred care. In
of labour.[update of Cochrane Database Syst National Collaborating Centre for Women’s
Rev. 2001;(3):CD000941; PMID: 11686970]. and Children’s Health, ed. Caesarean Section.
(Review). Cochrane Database of Systematic London: RCOG Press, 2004:20–5
Reviews 1905;CD000941 101. Magann EF, Evans S, Hutchinson M, Collins
88. Kelly AJ, Kavanagh J, Thomas J. Vaginal R, Lanneau G, Morrison JC. Postpartum
prostaglandin (PGE2 and PGF2a) for induc- hemorrhage after cesarean delivery: an analysis
tion of labour at term.[update of Cochrane of risk factors. S Med J 2005;98:681–5
Database Syst Rev. 2001;(2):CD003101; 102. Combs CA, Murphy EL, Laros RK Jr. Factors
PMID: 11406078]. (Review). Cochrane associated with hemorrhage in cesarean deliver-
Database of Systematic Reviews 2001;CD003101 ies. Obstet Gynecol 1991;77:77–82
89. Mahon TR, Chazotte C, Cohen WR. Short 103. Gardella C, Taylor M, Benedetti T, Hitti J,
labor: characteristics and outcome. Obstet Critchlow C. The effect of sequential use of
Gynecol 1994;84:47–51 vacuum and forceps for assisted vaginal delivery
90. Cohen WR. Influence of the duration of second on neonatal and maternal outcomes. Am J
stage labor on perinatal outcome and puerperal Obstet Gynecol 2001;185:896–902
morbidity. Obstet Gynecol 1977;49:266–9 104. Carroli G, Belizan J. Episiotomy for vaginal
91. Saunders NS, Paterson CM, Wadsworth J. birth. (Review). Cochrane Database of Systematic
Neonatal and maternal morbidity in relation to Reviews 2000;CD000081
the length of the second stage of labour. Br J 105. Myers–Helfgott MG, Helfgott AW. Routine
Obstet Gynaecol 1992;99:381–5 use of episiotomy in modern obstetrics. Should
92. Cheng YW, Hopkins LM, Caughey AB. How it be performed?. Obstet Gynecol Clin N Am
long is too long: Does a prolonged second stage 1999;26:305–25
of labor in nulliparous women affect maternal 106. House MJ, Cario G, Jones MH. Episiotomy
and neonatal outcomes? Am J Obstet Gynecol and the perineum: A random controlled trial. J
2004;191:933–8 Obstet Gynaecol 1986;7:107–10
93. Myles TD, Santolaya J. Maternal and neonatal 107. Dannecker C, Hillemanns P, Strauss A,
outcomes in patients with a prolonged second Hasbargen U, Hepp H, Anthuber C.
stage of labor. Obstet Gynecol 2003;102:52–8 Episiotomy and perineal tears presumed to be
94. Janni W, Schiessl B, Peschers U, et al. The imminent: randomized controlled trial. Acta
prognostic impact of a prolonged second stage Obstet Gynecol Scand 2004;83:364–8
34
56
30 August 2006 14:19:13
4
PITFALLS IN ASSESSING BLOOD LOSS
AND DECISION TO TRANSFER
B. S. Kodkany and R. J. Derman
INTRODUCTION hemoglobin to acid hematin or cyanmethemo-

globin, which in turn was measured by a
Pregnancy and childbirth involve health risks,
colorimeter; plasma volume determinations
even for women without any pre-existing health
before and after delivery using radioactive tracer
problems1–7. Obstetric hemorrhage is the single
elements; and, finally, measuring blood loss by
most important cause of maternal death. Of
using 51Cr-tagged erythrocytes.
great importance is the inaccurate assessment of
None of these methods was ever adopted in
blood loss that may result in significant adverse
clinical practice because of their complicated
sequelae. Underestimation leads to delayed
nature or due to the effort, expense and time
treatment and overestimation to unnecessary
required to obtain results before beginning
and costly interventions. It is axiomatic that
interventions. Thus, visual estimation, inaccu-
postpartum hemorrhage occurs unpredictably
rate as it may be, continues to be used clinically.
and no parturient is immune from it. Simply
Published studies, in which investigators care-
stated, postpartum hemorrhage is an equal
fully quantified blood loss after delivery, repeat-
opportunity killer8. Unlike uterine rupture
edly indicate that clinical estimates of blood
which can precede death by 24 h and
loss are notoriously unreliable, with a tendency
antepartum hemorrhage which may lead to
to underestimate the incidence of postpartum
death in half that time, postpartum hemorrhage
hemorrhage by 30–50%1. As a result, numerous
can be lethal in as little as 2 h.
authorities have advocated a more objective
The common definitions of postpartum
approach to the diagnosis of postpartum
hemorrhage are described in Chapter 2. Tradi-
hemorrhage. Although many studies address
tionally, blood loss after delivery is visually
this issue, accurate measurement of blood loss
estimated, with wide variations in accuracy.
by an ideal method remains a gray area.
The importance of accurately measuring vaginal
blood loss at delivery was stressed by Williams
as early as 19199. The birth attendant grossly
NORMAL BLOOD LOSS DURING
makes a quantitative estimate; however, the
DELIVERY
associated amount of loss is often far greater
than appreciated by visual estimation alone10. Investigators report a range of average blood
In the past, quantitative methods for estimat- loss during vaginal delivery. For example, at the
ing vaginal blood loss included direct collection low end it has been reported as 343 ml in 1000
of blood into bedpans or plastic bags; gravi- consecutive term vaginal deliveries, 339 ml and
metric methods wherein pads were weighed 490 ml, respectively, in two separate studies of
before and after use and the difference in the 100 and 123 patients using the acid hematin
weight used to determine the amount of blood spectrophotometric method, and a 450-ml
lost; determination of changes in blood indices average blood loss in 123 deliveries using
before and after delivery; the acid hematin chromium-labeled red blood cells10–13. Despite
method, by which blood in the sponges and such variations, it is now generally accepted
pads was mixed with a solution that converted that the average blood loss during delivery is
35
57
30 August 2006 14:19:13
between 400 and 500 ml, whereas most (2) In primiparae, forceps delivery is associated
Cesarean births loose about 1000 ml14. Unfor- with greater blood loss than spontaneous
tunately, these values are reflective of hospital- delivery; this is related to the episiotomies
based data, primarily among women in the and other injuries to the genital tract;
developed world.
(3) Patients with an episiotomy and a laceration
lose significantly more blood than those
PHYSIOLOGICAL ADAPTATIONS IN without such insult. Episiotomies contrib-
PREGNANCY ute 154 ml to the average blood loss25.
However, forceps delivery does not appear
Antepartum adaptations for physiologic blood
to contribute to blood loss per se; any excess
loss at delivery include a 42% increase in plasma
bleeding in this instance is due to the
volume and a 24% increase in red blood cell
episiotomy that is almost always required.
volume by the third trimester15. Women who
develop pre-eclampsia either experience little or
no expansion over non-pregnant levels or lose DIAGNOSIS OF POSTPARTUM
during the third trimester what gain had been HEMORRHAGE
accrued early in gestation16. In severe pre-
Over the years, different methods have been
eclampsia, the blood volume frequently fails to
used for estimation of blood loss; these can be
expand and is similar to that in a non-pregnant
classified as clinical or quantitative methods and
woman17. Hemoconcentration is a hallmark
are delineated below.
of eclampsia with increased sensitivity to
even normal blood loss at delivery18. Women
so afflicted are relatively less prepared to Clinical methods
withstand blood loss and may develop life-
Clinical estimation remains the primary means
threatening hypovolemia with smaller amounts
to diagnose the extent of bleeding and to direct
of hemorrhage16.
interventional therapy in obstetric practice.
Progressively complicated deliveries are
Examples include internal hemorrhage due to
accompanied by greater degrees of blood
ruptured tubal pregnancy, ruptured uterus, and
loss: vaginal delivery (500 ml), Cesarean
the concealed variety of abruptio placentae. The
section (1000 ml), repeat Cesarean section
classification of hemorrhage can be based on
plus hysterectomy (1500 ml), and emergency
a graded physiological response to the loss of
hysterectomy (3500 ml)19–21.
circulating blood volume (Table 1)26,27. This
Some of the factors leading to increased
scheme has worked well in the initial manage-
blood loss in the third stage of labor are as
ment of trauma patients. Knowing that the
follows22–24:
blood volume of a pregnant woman is 8.5–9%
(1) Mean vaginal blood loss is higher in of her weight, one is able to quickly approx-
multiparae than in primiparae; imate blood loss based on changes in pulse,
Table 1 Classes of hemorrhage
Class I Class II Class III Class IV
% Blood loss 15 20–25 30–35 40

Pulse (beats/min) normal 100 120 140
Systolic blood pressure normal normal 70–80 60
(mmHg)
Mean arterial pressure 80–90 80–90 50–70 50
(mmHg)
Tissue perfusion postural peripheral pallor, restlessness, collapse, anuria,
hypotension vasoconstriction oliguria air hunger
36
58
30 August 2006 14:19:13
Assessment of blood loss and decision to transfer
systolic blood pressure and mean arterial

pressure. Thus, the failure to respond to the
initial administration of 3000 ml of crystalloid
would suggest a Class II hemorrhage with loss
greater than 20–30% of the total blood volume
or acute ongoing bleeding26,27. A systolic blood
pressure below 100 mmHg and a pulse rate
above 100 beats/min are late signs of depleted
blood volume and indicate commencing failure
of compensatory mechanisms28, whereas acute
blood loss might not be reflected by a decrease
in hematocrit or hemoglobin level for 4 h
or more26,27. The importance of diagnosis at a Figure 1 Soakage characteristics of 10 × 10 cm
Class I stage cannot be too strongly emphasized pads
as women can progress into Class II rapidly.
At level III, unless intervention is rapid
and appropriate, women may progress to
irreversible shock.
Quantitative methods
Visual assessment
The standard method of observation used for
the measurement of blood loss is relatively
straightforward and requires no expenditure8.
Despite its inaccuracy and variation from one
care-giver to the next, birth attendants correlate Figure 2 Blood drained into a fixed collecting
it with clinical signs. A review of the records of container
32 799 deliveries at a large municipal hospital
during the decade of 1963–1972 found an inci-
dence of postpartum hemorrhage of 4.7/1000 totaling up the blood in the container, in the
live births or 0.47%. This was extremely low sponges and secondary blood spillage on the
compared to stated rates in the literature, and delivery table, garments and floor. How often
the author concluded that many cases of post- such calculation is utilized is unknown, but
partum hemorrhage were not recorded due to failure to do so undoubtedly contributes to
underestimation of blood loss29. underestimation.
The accuracy of this method can be
improved by standardization and training. The
Direct collection of blood into bedpan or plastic bags
observer needs to be trained in determining the
blood loss using a single collecting container This approach was used in the World Health
and fixed-sized gauze pads of size 10 × 10 cm. Organization (WHO) multicenter, randomized
Simulated scenarios with known measured trial of misoprostol in the management of the
blood volume need to be created and calibrated third stage of labor30. In this trial, blood loss was
visually (see Figure 1). measured from the time of delivery until the
Another method of calculation is by allowing mother was transferred to postnatal care. Imme-
blood to drain into a fixed collecting container diately after the cord was clamped and cut, the
(Figure 2) for estimation at the end of 1 h. blood collection was started by passing a flat
Blood losses on the delivery table, garments and bedpan under the buttocks of a woman deliver-
floor should also be assessed. At the end of 1 h, ing in a bed or putting in place an unsoiled sheet
the total amount of blood lost is estimated by for a woman delivering on a delivery table.
37
59
30 August 2006 14:19:18
Blood collection and measurement continued Postpartum hemoglobin

11.2
until the third stage of the labor was completed
and the woman was transferred to the postnatal 11.0
ward. This period was generally up to 1 h
Hemoglobin (g/dl)
10.8
postpartum. At that time, the collected blood
was poured into a standard measuring jar 10.6
provided by WHO and its volume measured. 10.4
To simplify the procedure for measurement 10.2
of blood loss, any available small gauze swabs Visual
soaked with blood were put into the measuring 10.0
BRASSS-V
jar and included in the measurement together 9.8 Both groups
with the blood and clots. A validity study was
9.6
performed before the trial to assess the effect of 0 1 2 3 5
adding the gauze swabs on the estimation Postpartum day
of blood loss and was found to result in an
approximately 10% increase in the blood loss Figure 3 Postpartum hemoglobin changes
measurement.
and let stand for 15 min. One ml of the
Gravimetric method filtrate is diluted 10 times in 5% sodium
hydroxide and left to stand for another
This method involves weighing sponges before 15 min. The optical density (OD) is read
and after use. The difference in weight provides with a spectrophotometer at 550 nm at
a rough estimate of blood loss. 30 min after the addition of sodium
hydroxide to the sample;
Determination of changes in hematocrit and (3) Calculations
hemoglobin
OD sample × 2000 × 10 ml Blood volume
The changes in values before and after delivery =
OD blood standard × 100 loss
of the hematocrit and hemoglobin levels provide
quantitative measurements of blood loss, as
Plasma volume changes
depicted in Figure 3.
The plasma volume can be determined before
and after delivery using radioactive tracer
Acid hematin method
elements.
This method is based on collected blood being
mixed with a standardized solution which
converts hemoglobin to acid hematin or cyan- Measurement of tagged erythrocytes
methemoglobin. This in turn can be measured Blood loss can be measured by using
by a spectrophotometer or colorimeter. Spec- 51Cr-tagged erythrocytes13.
trophotometric analysis can be performed by
the methods described below9,31:
Failures of each method
(1) Preparation of standard Two milliliters of
peripheral blood are collected pre-delivery. Visual assessment
The blood standard is prepared with 0.1 ml
The major advantage of this method is that it
of the patient’s peripheral blood in 9.9 ml of
is a real-time assessment and enables the birth
5% sodium hydroxide solution. The optical
attendant to correlate findings, on an individu-
density (OD) is read at 550 nm after
alized basis, with the clinical presentation.
30 min;
However, significant differences between
(2) Preparation of sample The collected sample clinical estimates and actual measurements
is added to 2 liters of 5% sodium hydroxide have been consistently demonstrated in several
38
60
06 September 2006 16:50:19
studies28. The most common error is under- Use of blood indices and spectrophotometric
estimation of blood lost, with an average error measurement of hemoglobin
of 46% when estimates at the time of delivery
The first study reporting on measurement of
are compared with more precise measurements.
blood loss during surgical procedures employed
As might be expected, observers tend to give
the colorimetric technique, which required that
median or average estimate of blood loss. When
hemoglobin be washed from surgical materials
losses were large, they were most often under-
in a blender and measured in a colorimeter38.
estimated and, when the losses were less than
Clearly, this is impractical in obstetric practice.
average, they tended to be overestimated11.
Routine hematocrit determination, on the other
hand, is possible if the equipment is available.
Standardized visual estimation However, routine postpartum hematocrits are
unnecessary in clinically stable patients with an
In an attempt to rectify this error, the use of a estimated blood loss of less than 500 ml. After
standardized visual estimation can be employed delivery associated with an average blood loss,
as a simple method to be routinely practiced in the hematocrit drops moderately for 3–4 days,
low-resource setting, albeit based on training followed by an increase. The peak drop may be
the providers and standardization of the pads appreciated on day 2 or day 3 postpartum39. By
(size and quality) used during delivery. The days 5–7, the postpartum hematocrit will be
accuracy of estimated blood loss is not depend- similar to the prelabor hematocrit15. Should
ent upon age or the clinical experience of the the postpartum hematocrit be lower than the
provider32–35. Teaching this tool significantly prelabor hematocrit, the blood loss may have
reduced the error in blood loss estimation for been larger than appreciated40.
inexperienced as well as experienced clinicians.
Of particular clinical importance is a reduction
in underestimation of blood loss in the face of Plasma volume changes and measurement of
greater degrees of measured blood loss; this has tagged erythrocytes
the strongest potential to reduce hemorrhage-
Blood volume estimation using dye-dilution or
related morbidity and mortality36.
radioisotope dilution techniques is more diffi-
cult and requires special equipment and serial
measurements41,42. Measurement of erythrocytes
Collection in pan or plastic bags
appears to be more consistent than estimates of
The errors in estimating blood loss arise from plasma volume secondary to physiological hemo-
failure to collect or note all the blood in stained dilution causing a fluid overload of approxi-
linen, incomplete extraction from the collection mately 1080–1680 ml in pregnancy14. Significant
device, ignoring maternal blood within the cardiovascular changes occur immediately post-
placenta (approximately 153 ml), confusion partum. The cardiac output remains elevated for
related to the mixing of blood contaminated 24 h, blood pressure declines initially and then
with amniotic fluid and urine, and technical stabilizes on postpartum day 2. Maternal physio-
inaccuracies associated with transfer of the logical changes of hemodilution lead to reduced
collection to a measuring device. hemoglobin and hematocrit values, reflecting
the importance of timing of the measurement43.
In the majority of patients44, no single timed
Gravimetric methods hemoglobin or hematocrit determination in the
The gravimetric method requires the weighing first 24 h postpartum will detect the peak.
of materials such as soaked pads on a scale and
subtracting the known weights of these materi-
BRASSS-V DRAPE: BLOOD LOSS
als to determine the blood loss37. Inaccuracies
COLLECTION TOOL
can arise at several steps in this procedure,
including lack of international standardization A randomized, placebo-controlled trial to test
of size and weight of gauze, sponges and pads. the use of oral misoprostol was conducted to
39
61
30 August 2006 14:19:18
reduce the incidence of acute postpartum

hemorrhage and hence maternal morbidity and
mortality in women delivering in rural villages
(away from major hospitals) within Belgaum
District, Karnataka, India. The intervention
was delivered by local health-care workers. A
critical component of this trial was the develop-
ment of a specially designed low-cost ‘calibrated
plastic blood collection drape’ that would objec-
tively measure the amount of blood collected
in the immediate postpartum period. The
BRASSS-V drape was developed by the
NICHD-funded Global Network UMKC/
JNMC/UIC collaborative team to specifically
estimate postpartum blood loss45,46. (The name
‘BRASSS-V’ was coined by adding the first let-
ter of the names of the seven collaborators who
developed the drape.) The drape has a cali-
brated and funneled collecting pouch, incorpo-
rated within a plastic sheet that is placed under
the buttocks of the patient immediately after the
delivery of the baby. The upper end of the sheet
has a belt, which is loosely tied around the
woman’s abdomen to optimize blood collection,
Figure 4 BRASSS-V blood collection drape with
particularly for deliveries performed on the floor
calibrated receptacle
or on a flat surface at homes or in rural primitive
health posts. This simple tool not only has
the potential for a more accurate detection of
postpartum blood loss, but we hypothesize that
this approach will lead to earlier interventions,
with an ultimate goal of decreasing maternal
morbidity and mortality due to postpartum
hemorrhage. Since most developing countries
use some form of under-buttock sheet, either at
home, in the health center or in hospitals, drape
substitution is acceptable and relatively simple.
The BRASSS-V calibrated drape used for
objective estimation of blood loss is shown in
Figures 4 and 5.
Results of three studies conducted at JNMC,
Belgaum, Karnataka, India4,7 strongly suggest
that the BRASSS-V drape is an accurate and
practical tool to measure blood loss occurring in
the third stage of labor. While, among women Figure 5 Collection of blood using BRASSS-V
with little blood loss, the ranges of blood loss blood collection drape with calibrated receptacle
were similar in both visual and drape assess-
ment, the actual visual assessment amount was in the literature, whereas differences between
considerably less compared with the calibrated the drape and spectrophotometry values were
drape values (Table 2 and Figure 6). This found to be 37.15 ml, with the drape having the
observation further underscores the inaccuracy higher value (an average error of 16.1%). The
of the visual estimation method as described drape measured blood loss equally and as
40
62
30 August 2006 14:19:22
Table 2 Distribution of blood loss
Blood loss (ml)
Visual Drape All cases

(n = 61) (n = 62) (n = 123)
Mean ± standard deviation 203.11 ± 147.49 302.82 ± 173.28 253.37 ± 168.86

Range 50–950 50–975 50–975
50 47 Table 3 Comparison between drape-measured and

45 spectrometrically analyzed blood loss
40
No. of cases
35 Blood loss (ml)

30 29
25
25 Drape-measured Spectrometry
20
15 12 Mean ± standard 225 ± 96.10 187.84 ± 61.79
10 8 deviation
5 2 Range 100–350 93.19–285.98
0
< 250 ml 250–500 ml > 500 ml
Visual Drape
diagnosis of postpartum hemorrhage; it also led
Figure 6 Number of cases detected for specific to earlier transfer from rural areas to the higher
blood loss (p < 0.01). The calibrated drape more facility. The women who delivered at home and
accurately determined true blood loss when their family members also appreciated the use-
≥ 250 ml and more accurately estimated overall
fulness of the drape for easy disposal of body
levels
fluids after birth45.
A similar approach has been used in another
efficiently as gold-standard spectrophotometry recently reported study48. A plastic collecting
(Pearson’s correlation coefficient of 0.928; bag put under the pelvis of the mother just after
p = 0.01, Table 3). delivery can serve as a quantitative and objective
Use of the drape diagnosed postpartum method of measuring blood loss. The study goal
hemorrhage four times as often as the visual was to assess sensitivity, specificity, positive
estimate. A larger validation study is presently predictive value and negative predictive value,
underway at the University of Missouri at including correlation between the bag’s volume
Kansas City School of Medicine. In addition, and hemoglobin and hematocrit variation. The
the drape is being tested in a number of inter- authors conclude that the collecting pelvis bag is
national settings including Tibet, Vietnam, a rapid and precise procedure with which to
Egypt, Ecuador, Brazil and Argentina. Based on diagnose postpartum hemorrhage in the deliv-
the Indian experience, it appears to have great ery room. It also enables a visual and quantita-
potential for training delivery attendants to tive non-subjective estimation of blood loss.
determine postpartum blood loss in an accurate Because of its simplicity and very low cost, the
and timely manner. The drape, apart from pelvis collecting bag may have applicability as a
being an objective tool for measurement of routine preventive measure.
postpartum blood loss, also provided a hygienic Accurate measurement of blood loss at deliv-
delivery surface while permitting early manage- ery as a means of early detection of postpartum
ment and referral. Residents and nurses in hemorrhage is necessary for several reasons, not
hospital settings and the nurse midwives who the least of which is the fact that oxytocic
used the BRASSS-V drape during home deliv- agents, while an important component for
ery all found it to be a very useful tool to addressing the third stage of labor, do not
measure blood loss after delivery and for early address many factors related to postpartum
41
63
30 August 2006 14:19:22
2 8 mortality. As other quantitative methods

employed have both practical and technical lim-
itations, the employment of simple tools, such
as the BRASSS-V under-buttock blood collec-
tion drape with a calibrated receptacle, can be
effectively employed for objectively assessing
the blood loss. It is likely to be of great utility to
the midwife/birth attendant and thus help to
ensure more timely and accurate patient man-
agement. Having identified excessive blood loss,
corrective measures can be taken at the earliest
time, thus improving outcomes associated with
postpartum hemorrhage.
ACKNOWLEDGEMENTS
Our sincere thanks to Dr Shivaprasad S.
123 Goudar, Professor of Physiology & Research,
Coordinator Global Network for Women’s and
Visual Drape Total
Children’s Health Research Site 8, and Dr
Kamal Patil, Associate Professor of Obstetrics &
Figure 7 Number of cases of postpartum Gynecology, JNMC for invaluable assistance
hemorrhage (PPH) detected for specific blood loss
in the preparation of this manuscript. We also
(p < 0.01). The calibrated drape diagnosed PPH at
acknowledge the contribution of Dr Kuldeep
a rate four times that of the visual estimate method
Wagh and Dr B. V. Laxmi, residents in
the Department of Obstetrics & Gynecology,
hemorrhage in resource-poor areas. Trauma of JNMC for participating in the validation study
the birth canal during delivery and retained and to Dr A. Patel for her contributions to the
placental fragments are important causes of design of the BRASSS-V drape.
postpartum hemorrhage and may occur more
often than previously reported. Visual assess-
ment of blood loss in the presence of a References
contracted uterus may diagnose traumatic post-
1. Starr A. The Safe Motherhood Agenda: Priorities for
partum hemorrhage late and therefore result
the Next Decade. New York: Inter-agency Group
in delayed referrals. In India and many other for Safe Motherhood, Family Care International,
developing nations, at least half of all births take 1997
place in rural areas. Most of these deliveries are 2. Reduction of maternal mortality. A joint WHO/
conducted by indigenous health-care providers UNFPA/UNICEF/World Bank Statement.
such as dais (traditional birth attendants) or 1999http://www.who.int/reproductive-health/
auxiliary nurse midwives having varying levels publications/reduction_of_maternal_mortality/
of training. Blood loss appears to be commonly reduction_of_maternal_mortality_contents.htm
underestimated, as visual assessment is the only 3. Abou Zahr C. Antepartum and postpartum hem-
means available to the birth attendant to make orrhage. In Murray CJL, Lopez AD, eds. Health
Dimensions of Sex and Reproduction. Boston:
this diagnosis. The clinical symptoms of blood
Harvard University Press, 1998
loss (low blood pressure, fast pulse, pallor and
4. Berg CJ, Atrash HK, Koonin LM, Tucker M.
sweating, signs of hypovolemia and impending Pregnancy-related mortality in United States,
shock) are often the primary indicators for inter- 1987–1990. Obstet Gynecol 1996;88:161–7
vention. However, relying on the onset of such 5. Hogberg U, Innala E, Sandstorm A. Maternal
symptoms may lead to delayed intervention, mortality in Sweden, 1980–1988. Obstet Gynecol
resulting in increased rates of morbidity and 1994;84:240–4
42
64
30 August 2006 14:19:23
6. Razum O, Jahn A, Blettner M, Reitmaier P. of uterine arteries. Am J Obstet Gynecol 1952;64:

Trends in maternal mortality ratio among 1143–8
women of German and non-German nationality 22. Combs CA, Murphy EL, Laros RK Jr. Factors
in west Germany, 1980–1996. Int J Epidemiol associated with hemorrhage in cesarean deliver-
1999;28:919–24 ies. Obstet Gynecol 1991;77:77–82
7. Dildy GA. Postpartum Hemorrhage. Washington, 23. Calkins LA. Factors governing blood loss in the
DC: American College of Obstetricians and third stage of labor. Am J Obstet Gynecol 1929;
Gynecologists, 1998 17:578
8. Maine D. Safe Motherhood Programs: Options 24. Hill JA, Fadel HE, Nelson MC, Nelson RM,
and Issues. Columbia University: Center for Nelson GH. Blood loss at vaginal delivery. South
Population & Family Health, 1993:42 Med J 1986;79:188–92
9. Williams JW. The tolerance of freshly delivered 25. Qubil LD, Saski A. Episiotomy blood loss. Am J
women to excessive loss of blood. Am J Obstet Obstet Gynecol 1947;54:51
Gynecol 1919;90:1 26. Spoerel WE, Heagy FC. The use of blood vol-
10. Duthie SJ, Ven D, Yung GL, Guang DZ, Chan ume determination for the evaluation of blood
SY, Ma HK. Discrepancy between laboratory loss during operation. Can J Surg 1962;5:25–32
determination and visual estimation of blood loss 27. Arulkumaran S, Symonds IB, Fowlie A. Massive
during normal delivery. Eur J Obstet Gynecol obstetric hemorrhage. In Oxford Handbook of
Reprod Biol 1990;38:119–24 Obstetrics & Gynaecology. Oxford: Oxford
11. Newton M, Mosey IM, Egli GE, Gifford WB, University Press, 2003:399
Hull CT. Blood loss during and immediately 28. Brant HA. Precise estimation of postpartum
after delivery. Obstet Gynecol 1961;17:9–18 haemorrhage: difficulties and importance. Br
12. Newton M. Postpartum hemorrhage. Am J Med J 1967;1:398–400
Obstet Gynecol 1966;94:711–16 29. Hester JD. Postpartum hemorrhage, and
13. Gahres EE, Albert SN, Dodek SM. Intrapartum re-evaluation of uterine packing. Obstet Gynecol
blood loss measured with Cr51-tagged erythro- 1975;45:501–4
cytes. Obstet Gynecol 1962;19:455–62 30. Gulmezoglu AM, Villar J, Ngoc NT, et al. WHO
14. Nelson GH, Ashford CB, Williamson R. Method Multicentre randomized trial of misoprostol in
for calculating blood loss at vaginal delivery. the management of the third stage of labour.
South Med J 1981;74:550–2 Lancet 2001;358:689–95
15. Chesley LC. Plasma and red cell volumes 31. Chua S, Ho LM, Vanaja K, Nordstrom L, Roy
during pregnancy. Am J Obstet Gynecol 1972; AC, Arulkumaran S. Validation of a laboratory
112:440–50 method of measuring postpartum blood loss.
16. Knuppel RA, Hatangadi SB. Acute hyper- Gynecol Obstet Invest 1998;46:31–3
tension related to hemorrhage in obstetric 32. Dildy GA, Paine AR, George NC, Velasco C.
patients. Obstet Gynecol Clin N Am 1995;22: Estimating blood loss: can teaching significantly
111–29 improve visual estimation? Obstet Gynecol 2004;
17. Gabbe SG, Niebyl JR, Simpson JL, eds. Obstet- 104:601–6
rics: Normal and Problem Pregnancies, 4th edn. 33. Grant JM. Treating postpartum haemorrhage.
Churchill Livingstone, 2001 Br J Obstet Gynaecol 1997;104:vii
18. Cunningham FG, Gilstrap LC, Gant NF, et al., 34. Patton K, Funk DL, McErlean M, Bartfield JM.
eds. Williams Obstetrics, 21st edn. McGraw-Hill, Accuracy of estimation of external blood loss by
2001 EMS personnel. J Trauma 2001;50:13–20
19. Pritchard JA, Baldwin RM, Dickey JC, et al. 35. Meiser A, Casagranda O, Skipka G, Laubenthal
Blood volume changes in pregnancy and the H. Quantification of blood loss. How precise
puerperium. II. Red blood cell loss and change is visual estimation and what does its accuracy
in apparent blood volume during and following depend on? Anaesthesist 2001;50:13–20
vaginal delivery, cesarean section, and cesarean 36. Luegenbiehl DL, Debra L. Improving visual
section plus total hysterectomy. Am J Obstet estimation of blood volume on peripads. MCN
Gynecol 1962;84:1272–82 Am J Matern Child Nurs 1997;22:294–8
20. Clark SL, Yeh SY, Phelan JP, et al. Emergency 37. Buchman MI. Blood loss during gynecological
hysterectomy for obstetric hemorrhage. Obstet operations. Am J Obstet Gynecol 1953;65:53–64
Gynecol 1984;64:376–80 38. Gatch WD, Little WD. Amount of blood lost
21. Waters EG. Surgical management of postpartum during some of the more common operations.
hemorrhage with particular reference to ligation JAMA 1924;83:1075–6
43
65
30 August 2006 14:19:23
39. Maruta S. The observation of the maternal 44. Nelson GH. Consideration of blood loss at
haemodynamics during labour and cesarean delivery as a percentage of estimated blood
section. Nippon Sanka Fujionka Gakkai Zasshi volume. Am J Obstet Gynecol 1980;138:1117
1982;34:776–84 45. Kodkany BS, Derman RJ, Goudar SS, et al.
40. Pritchard JA, Baldwin RM, Dickey JC, Wiggins Initiating a novel therapy in preventing
KM. Blood volume changes in pregnancy and postpartum hemorrhage in rural India: a
the puerperium. Am J Obstet Gynecol 1962;84: joint collaboration between the United States
1271 and India. Int J Fertil Women Med 2004;49:
41. Quinlivan WLG, Brock JA, Sullivan H. Blood 91–6
volume changes and blood loss associated with 46. Geller SE, Patel A, Naik VA, et al. Conduct-
labor. Correlation of changes in blood volume ing international collaborative research in devel-
measured by 131I-albumin and Evans blue dye, oping nations. Int J Gynaecol Obstet 2004;87:
with measured blood loss. Am J Obstet Gynecol 267–71
1970;6:843–9 47. Patel A, Goudar SS, Geller SE, et al. Drape esti-
42. Ueland K. Maternal cardiovascular dynamics. mation versus visual assessment for estimating
VII. Intra-partum blood volume changes. Am J postpartum hemorrhage. Int J Gynaecol Obstet
Obstet Gynecol 1976;126:671–7 2006;93:220–4
43. Robson SC, Boys RJ, Hunter S, Dunlop W. 48. Tourne G, Collet F, Lasnier P, Seffert P.
Maternal hemodynamics after normal delivery Usefulness of a collecting bag for the diagnosis of
and delivery complicated by postpartum hemor- post-partum hemorrhage. J Gynecol Obstet Biol
rhage. Obstet Gynecol 1989;74:234–9 Reprod (Paris) 2004;33:229–34
44
66
30 August 2006 14:19:23
5
ASSESSING AND REPLENISHING LOST VOLUME
J. G. L. Cockings and C. S. Waldmann
INTRODUCTION (20–30%) increase in red cell volume. Cardiac

output increases to 50% above pre-pregnancy
Classically, shock is defined as a state of inade-
levels by the 24th week. Systemic blood pres-
quate tissue perfusion for the metabolic needs of
sure is more variable in healthy uncomplicated
the patient. This state of inadequate blood flow
pregnancy, with a small fall in the first and sec-
may manifest clinically as tachycardia, pallor,
ond trimesters, but a return to pre-pregnancy
oliguria, the development of lactic acidosis and
levels by the third. Resting heart rate increases
altered mental status.
progressively in the first and second trimesters
Shock is either hypovolemic, cardiogenic,
to 15–20 beats per minute above pre-pregnant
anaphylactic or cytotoxic. Hypovolemic shock
levels. In addition to these changes, other
classically associated with postpartum hemor-
changes also take place in the autoregulation of
rhage is due to loss of circulating blood volume.
intravascular volume and the circulation, both
Hypotension is often present in severe cases, but
of which affects the body’s response to blood
is a late sign and is a poor guide to the volume of
loss. Examples include a blunted response to
blood lost, as pregnancy is accompanied by an
angiotensin II, which may in part be due to
alteration of cardiovascular physiology and the
an increased production of nitric oxide3, a
response to blood loss and its management may
decreased tolerance to postural changes and an
differ to the non-pregnant situation. Maternal
increased cardiac noradrenaline turnover4,5.
blood volume increases, total red cell mass also
Circulating volume, clinical signs of hypo-
increases but to a lesser extent, systemic vascu-
volemia and the body’s ability to compensate for
lar resistance is reduced, and cardiac output
volume loss are also all affected by pregnancy-
becomes more dependent on body position.
related diseases and their treatment, the effects
Massive postpartum hemorrhage accounts for
of which continue on into the early postpartum
35% of obstetric admissions to intensive care in
period. Pre-eclampsia, for example, causes a
the UK1,2. These patients demand rapid assess-
contracted effective arterial blood volume com-
ment and judicious replenishment of lost circu-
pared with the normal peripartum state. Vascu-
lating volume, albeit within the context of the
lar reactivity is increased, and widely used drugs
compensatory effects of hypovolemic shock and
such as hydralazine and magnesium compro-
the physiological changes seen in late pregnancy.
mise the body’s ability to produce compensa-
tory vasoconstriction in the face of hemorrhage.
Indeed, it appears that there is a failure to
PHYSIOLOGY
increase plasma volume and reduce systemic
The normal circulating blood volume for a vascular resistance in pre-eclampsia, due to
healthy non-pregnant adult is 70 ml/kg, or 7.5% inadequate trophoblastic invasion into the spiral
of body weight. Cardiac output is 4–6 l/min, arteries of the uterus5. Pre-eclamptic patients
and the non-pregnant adult systemic vascular thus have an increased tendency to develop pul-
resistance is 10–15 mmHg/l/min (900–1200 monary edema during volume replacement due
dyne.s/cm5). Maternal blood volume increases to many factors, including increased capillary
during pregnancy to 40% above baseline by the permeability, hypoalbuminemia and left ven-
30th week, with an accompanying but smaller tricular dysfunction6.
45
67
30 August 2006 14:19:23
Normal delivery results in predictable losses period have not been compiled, but the signs
of 300–500 ml blood volume for vaginal deliver- follow a similar pattern.
ies and 750–1000 ml for Cesarean section births The most important principle in the treat-
(see Chapter 4). However, in addition to blood ment of postpartum hemorrhage is early recog-
lost from the body, a substantial amount of nition and prompt correction of lost circulating
blood is also redirected into the systemic circu- volume, together with simultaneous medical
lation, often referred to as the autotransfusion and/or surgical intervention to prevent further
effect. This results in an increase in cardiac out- loss. Early recognition of life-threatening physi-
put by as much as 80%. The effect persists in ological derangements can be improved by the
uncomplicated patients, gradually returning to use of early-warning scoring systems.
non-pregnant levels at 2–3 weeks5. Recording physiological observations at
regular intervals has long been routine practice
in hospitals. Early-warning scores derived from
ASSESSMENT OF CIRCULATING
simple routine physiological recordings can
BLOOD VOLUME
identify patients with greater risk of critical
Young healthy adults can compensate for the illness and mortality. Such scores can be used
loss of large volumes from the circulation with to flag the early but sometimes subtle signs of
few obvious external signs. Accurate assessment concealed but largely compensated hemorrhage
of blood loss can be difficult for the experienced in the early postpartum patient and have been
as well as the inexperienced examiner, as recently recommended for use by the Confiden-
described in Chapter 4. tial Enquiry into Maternal and Child Health
In cases of hemorrhage symptoms often pre- report of 20047. These scores use the physiolog-
cede signs. These include unexplained anxiety ical parameters most likely to detect impending
and restlessness, the feeling of breathlessness life-threatening compromise. These usually
(with or without an increased respiratory rate), comprise respiratory rate, heart rate, systolic
and a sensation of being cold or generally blood pressure, temperature and mental aware-
unwell. For healthy, non-pregnant adults, hypo- ness. Each variable is assigned a weighted score
volemia and associated signs can be divided into and the total score is the sum of these. This
four stages (Table 1). These range from the allows a trigger value for ward staff to call for
largely undetectable stage 1 with less than 15% assistance from intensive care or other senior
loss of volume, to the severe life-threatening staff. Such systems have been shown to be
stage when more than 40% has been lost. reproducible and effective at predicting the
Unfortunately, comparable tables for early and likelihood of progressing on to critical illness.
late pregnancy and the immediate postpartum They are well suited to the early detection of the
Table 1 Stages of shock
Classification Class 1 Class 2 Class 3 Class 4
Blood loss 10–15% 15–30% 30–40% > 40%

(% volume lost)
Conscious state Alert, mild anxious and agitated or drowsy, confused or
thirst restless confused unconscious
Respiratory rate normal mildly elevated raised raised
Complexion normal pale pale marked pallor or gray
Extremities normal cool pale and cool cold
Capillary refill normal slow (> 2 s) slow (> 2 s) minimal or absent
Pulse rate normal normal elevated fast but thready
Systolic blood pressure normal normal normal or slightly low hypotensive
Urine output normal reduced reduced oligoanuric
Modified from Baskett PJF. ABC of major trauma. Management of hypovolaemic shock. BMJ 1990;300:
1453–7
46
68
30 August 2006 14:19:24
Assessing and replenishing lost volume
often subtle signs of unappreciated blood loss to the fundamental areas of vital function, the
and can be easily introduced. Altered normal conscious state and airway protection, the ade-
physiology in late pregnancy and the early quacy of respiratory function, oxygenation and
postpartum period demands that these scores, circulation. In particular, the following should
usually derived from general surgical or medical be assessed and documented:
patients, be modified for this population as
(1) Early stages of shock are associated with
shown in Table 2.
restlessness and agitation, sometimes with
Once the possibility of intravascular deple-
a heightened sense of thirst, but these
tion has been raised, a prompt clinical assess-
progress to drowsiness when around 30% of
ment is urgent, as the clinical condition of the
blood volume is lost. Loss of consciousness
patient can change rapidly. Clinical assessment,
is a very late sign, with significant risk of
in association with non-invasive and invasive
imminent death.
monitoring where appropriate, must be made
by senior clinicians (if available), with special (2) Tachypnea is an early sign, partly driven
attention to repeated assessment at frequent initially by the anxiety, but is an independ-
intervals to detect the problem as early as ent sign, and the respiratory rate increases
possible. If senior clinicians are not available, with progressive blood loss and will usually
they should be notified as described in the exceed 20 breaths/min when 30% of blood
protocols in Chapters 22 and 50. volume is lost.
Clinical examination is performed simulta-
(3) Oxygenation becomes harder to assess
neously with incident-related history taking.
clinically as peripheral pallor becomes more
This history may elicit the more obvious
marked, and the pulse oximeter becomes
features of shock such as overt blood loss
less reliable as peripheral perfusion
and pain, but may also elicit the more subtle
becomes weaker.
features such as general malaise, anxiety and
restlessness, a poorly defined sense of doom and (4) A fall in the jugular venous pressure occurs
breathlessness. Physical examination is directed reasonably early, but is partly compensated
Table 2 Modified early obstetric warning system. Reproduced with permission by Dr R Jones, Consultant
Anaesthetist, Royal Berkshire Hospital, UK, from unpublished work in progress
Score
3 2 1 0 1 2 3
Respiratory rate (bpm) <8 9–18 19–25 26–30 > 30

Pulse rate (bpm) < 40 40–50 51–100 101–110 111–129 > 129
Systolic blood pressure < 70 71–80 81–100 101–164 165–200 > 200
(mmHg)
Diastolic blood pressure < 95 95–104 > 105
(mmHg)
Conscious level unresponsive responds responds alert irritated
to pain to voice
Urine hourly (ml/h) 0 < 30 < 45 > 45
or in 24 h (< 720 ml) (< 1000 ml) (> 1000 ml)
Final score = sum of individual scores at any one time

Action:
Score 0 or 1 Repeat observations when appropriate for clinical scenario
Score 2 Inform midwife in charge, repeat in 15 min
Score 3 Inform midwife in charge, obstetric registrar and duty anesthetist
Score ≥ 4 As above but the consultant obstetrician should be informed
Consider informing duty consultant anesthetist and intensive care team
47
69
30 August 2006 14:19:24
for by a reduction on venous capacitance. two-dimensional ultrasound imaging should be

However, the jugular veins can be hard to the preferred method, with the evidence stron-
visualize reliably in postpartum women. gest for the internal jugular route8. In the
healthy, non-pregnant adult, the systemic venous
(5) A more reliable indication of hypovolemia
capacity is 3–4 liters, or 75% of circulating vol-
from the central venous pressure is the
ume. If the tone of the venous capacitance ves-
poor increase observed following volume
sels did not change as volume was lost from the
administration.
circulation, the central venous pressure would
(6) The pulse rate increases after around fall quickly and early, with early compromise of
15–20% of blood volume has been lost, but the cardiac output. However, as blood volume is
this sign can be unreliable as a sinus tachy- lost, the tone in these venous capacitance vessels
cardia is physiological in late pregnancy and increases, moving blood centrally, and main-
in the early postpartum period. taining central venous pressure.
Confusion surrounding the concept of
(7) Capillary refill is slowed after 15% of blood
venous return can be dispelled if it is thought of
volume is lost and is almost completely
in terms of right atrial pressure rather than an
absent when 40% of volume is lost.
increased flow of blood to the right atrium. As
(8) Blood pressure is well maintained, despite a blood is lost, the volume in the venous capaci-
falling cardiac output and tissue perfusion, tance vessels is reduced and the tone in these
until over 30–40% of circulating volume is vessels increases. The central venous pressure
lost. falls progressively, but to a lesser degree due to
the compensatory increase in this venous tone.
Figure 1 shows the relationship between venous
MANAGEMENT
capacitance and central venous pressure during
When the compensating mechanisms maintain- acute blood loss and immediate replacement.
ing the blood pressure have been exhausted, As blood is lost from the circulation, the patient
the blood pressure can fall dramatically. At this follows the line A to B (Figure 1). The central
point, shock is advanced and the risk of immi- venous pressure falls slowly at first, then more
nent death is significant. Once significant blood steeply as the extent of blood loss increases.
loss has been recognized, volume replacement As volume is returned to the circulation, the
has begun via large-bore peripheral access, patient will follow first the line B to C and then
medical therapies have been used and found
ineffective, and surgical intervention has been
organized, other methods to more accurately Central
venous Hypothetical
assess volume status and adequacy of the circu- lines of
pressure D
lation should be used to aid clinical assessment. constant
The first and simplest of these is invasive C venous
measurement of central venous pressure. A tone
central venous catheter can be placed in any A
central vein, but it should be remembered that,
in hypovolemia, identification of a central vein B Volume of
may be difficult without the use of ultrasound. venous
The internal jugular vein is the preferred site capacitance
in this situation, as the femoral vein is relatively vessels
inaccessible, and the subclavian route may have
a higher risk of complications in late pregnancy
and the early postpartum period, especially if
inserted under urgent conditions. Figure 1 Central venous pressure and venous
The National Institute for Clinical Excel- capacitance during blood loss and replacement.
lence (NICE) has recently issued guidance Modified from Bradley RD. Studies in Acute Heart
stating that cannulation of central veins using Failure. London: Edward Arnold, 1977:11
48
70
30 August 2006 14:19:24
C to D, rather than simply returning from B to intervention or the patient’s compensation

A. This is due to a number of factors and is to blood loss. Acute blood loss alone will not
addressed in a later section in this chapter. change the concentration of hemoglobin in the
The ability to palpate a peripheral pulse is a blood left in the system. The concentration is
good sign, but correlates poorly to any specific reduced only when the lost volume is replaced
arterial pressure. Loss of peripheral pulses is by internal fluid shifts or external fluids are
a late preterminal sign of hypovolemic shock. added to the system that are low in hemoglobin.
Arterial pressure is most simply measured using If left untreated, however, acute blood loss
a sphygmomanometer. This is familiar to all results in a fall in hemoglobin concentration
clinicians and uses a Riva Rocci cuff placed after about 4–6 h due to internal compensatory
around the upper arm. The correct size of cuff fluid shifts. In contrast, intravenous administra-
must be chosen, because, if the arm is too large tion of fluid will dilute the hemoglobin more
for the chosen cuff, the estimated reading will quickly.
be falsely high. Further information regarding the circula-
Frequently, arterial blood pressure is mea- tion status can be gained by more invasive or
sured non-invasively by means of an automated complex techniques. These include the pulmo-
version of this technique. Unfortunately, such nary artery catheter, pulse contour analysis, and
methods become inaccurate and unreliable the esophageal Doppler technique. The mea-
when the blood pressure deviates significantly surement of cardiac output by means of the
from normal, especially in times of poor periph- thermodilution technique using a pulmonary
eral perfusion and hypotension9, which are the artery catheter was first described by Bradley
very conditions seen in marked hypovolemia and Branthwaite in 196811 and subsequently
due to postpartum hemorrhage. Automated popularized by Swann, Ganz and co-workers12.
non-invasive blood pressure devices tend to Despite being the gold standard for invasive
over-estimate hypotension and under-estimate assessment of cardiac output and left atrial
hypertension. They can repeatedly cycle in an pressure for several decades, the technique has
attempt to measure when marked hypotension become less favored of late amid gathering evi-
is present, which can delay its early recognition. dence that the risks may outweigh the benefits
Finally, they can sometimes give a totally erro- in many circumstances13. The benefit of the
neous reading suggesting an adequate blood pulmonary artery catheter is that it provides reli-
pressure when, in reality, the blood pressure is able measurements even in the face of changing
absent or unmeasurable in the face of marked body position and is equally effective in the
hypovolemic shock, thereby giving a false sense awake conscious patient and the sedated or
of security9. anesthetized patient receiving artificial ventila-
The most reliable method of arterial blood tion. The risks include those associated with the
pressure measurement under abnormal or insertion of a large-bore cannula into a central
rapidly changing conditions is by means of an vein, which is higher risk in the volume-deplete
intra-arterial cannula and direct measurement patient suffering massive blood loss, even with
via a transducer. This method has the added the aid of the two-dimensional ultrasound tech-
advantage of providing easy access for blood nique. The risks also relate to risks of infection,
samples for arterial blood gas analysis and other cardiac arrhythmias, pulmonary artery damage
blood tests. Although such systems commonly and lung injury. Review of the use of the
used heparinized saline, the use of heparin is pulmonary artery catheter in the obstetric
unnecessary10, and the use of plain saline allows population14 shows that most of the experience
samples to be taken for coagulation tests with- has been in patients with pre-eclampsia and
out fear of contamination and erroneous results. eclampsia rather than massive postpartum
Hemoglobin estimates indicate the con- hemorrhage.
centration of hemoglobin in the sample. The Fortunately, pulse contour analysis is a
simplicity of this statement only reinforces the realistic alternative to the pulmonary artery
point that the concentration of hemoglobin fol- catheter15. The system requires only standard
lowing acute blood loss reflects either medical peripheral arterial and central venous cannulae.
49
71
30 August 2006 14:19:24
A common example of this technique uses car- hemorrhage and postpartum hemorrhage in
diac output estimated initially and periodically particular. This protocol should be familiar to
thereafter by the lithium dilution technique all, easily accessible and followed (see Chapters
(LiDCO Ltd., Cambridge, UK16). The pres- 13 and 22).
sure waveform is analyzed using this static The principal underlying aim of volume
cardiac output measurement as a reference to replacement during and following massive post-
estimate a stroke volume. Changes to the pulse partum hemorrhage is restoration and mainte-
waveform and heart rate from this point are nance of tissue perfusion to all body organs in
used to estimate stroke volume and cardiac out- order to maintain cellular function and viability.
put on a continuous display. Calibration is per- Although the initial focus is on restoration of
formed by periodic re-assessment of the cardiac the common clinical indicators of shock, the
output using the lithium dilution technique. clinician must proceed further. Even if all con-
Limitations relate to issues of non-linearity or ventionally used criteria resolve, shock may still
aortic compliance, how closely the radial arterial be present on a cellular, tissue or organ basis18.
pulse waveform resonance relates to the proxi- Compensated shock is the term often used to
mal aortic waveform and, therefore, stroke vol- describe the state where conventional hemo-
ume, the common problems of damped arterial dynamic parameters have been returned to
waveforms, drift between static measurements normal, despite persisting occult tissue hypo-
of cardiac output, and problems associated with perfusion, typically in the splanchnic bed. In
poor transmission of pulse waves in severe spite of adequate volume replacement, patients
arrhythmias17. Despite these concerns, this may develop multiple organ dysfunction with its
technique can provide a continuous idea of associated morbidity and mortality.
cardiac output, systemic blood pressure and Replenishing the lost volume must take place
vascular resistance in any patient with central simultaneously with control of the bleeding.
venous and peripheral arterial access and is well Medical and surgical attempts to control bleed-
suited for monitoring the postpartum patient ing must not be delayed by prolonged volume
who has undergone massive hemorrhage and is resuscitation in the face of ongoing blood loss.
undergoing resuscitation. Other pulse contour Volume resuscitation should be aimed at resto-
analysis systems are also commercially available ration of circulating blood volume as well as
that use alternative methods for measurement returning the oxygen-carrying capacity and
of the reference cardiac output, such as the hemostatic functions to an effective, albeit
PiCCO system (Pulsion Medical Systems, subnormal, level.
Munich, Germany) which employs a trans- Initial volume replacement enhances right
pulmonary thermodilution measurement from atrial filling and improves cardiac output. As
an axillary or femoral artery. shock develops with blood loss, venous tone
increases as described above. Volume adminis-
tration should be rapid, but titrated to the right
REPLENISHING LOST VOLUME
atrial filling pressure. Initially, right atrial filling
Replacing lost circulating volume should com- pressure may be restored by a smaller volume
mence as soon as significant bleeding is recog- than that lost due to the reduced capacity in the
nized and, ideally, before the signs of significant venous capacitance vessels. Indeed, immediate
hypovolemia have developed. Important initial rapid re-infusion of the entire volume lost may
measures are to simultaneously provide supple- provoke fluid overload if the tone in the capaci-
mental oxygen, ensure multiple large-bore peri- tance vessels did not decrease as rapidly. This
pheral intravenous access, undertake an initial can be appreciated from Figure 1; here, rapid
rapid clinical assessment and summon senior volume replacement occurs along the line B to
members of assistance from anesthesia, inten- C. Central venous pressure rises despite the
sive care, surgical and hematology departments intravascular compartment remaining depleted.
when these individuals are available. Each insti- As the venous tone relaxes following resuscita-
tution should have a rapid response protocol tion, further volume administration can occur
in place for the management of massive with a central venous pressure falling toward
50
72
30 August 2006 14:19:25
normal as the volume in the venous capacitance in mortality in intensive care patients requiring
becomes replete (line C to D). Thus, volume volume expansion whether this expansion was
administration should be rapid but infused in made with saline or albumin22. Colloids expand
discrete volume challenges, with the effect on the intravascular space preferentially, whereas
the right atrial filling pressure, systemic blood crystalloids quickly become distributed through-
pressure and other hemodynamic variables out the extracellular space. Saline has the disad-
being monitored. Commonly, 250–500 ml of vantage of hyperchloremia, which causes a
either a crystalloid or a colloid is administered dilutional or hyperchloremic acidosis23,24. The
over a period of 10–20 min as the urgency use of crystalloids is not associated with anaphy-
dictates (a patient with life-threatening class 4 laxis, whereas colloids such as the gelatins can
shock will receive 2–3 liters more quickly, produce severe life-threatening reactions,
but, even then, the principles of monitoring the although this is less common with hydroxyethyl
hemodynamic variables during the infusion of starch25. Crystalloids have minimal effect on co-
fluid remain). Simple measures of tissue under- agulation other than a dilutional effect, although
perfusion, which may persist after apparent saline infusions may have a procoagulant effect26.
restoration of global hemodynamics, include the Overall, crystalloids have a lower cost and lower
base deficit and serum lactate. Efforts to mea- incidence of side-effects, but the colloids have
sure and enhance tissue perfusion should con- several theoretical advantages regarding tissue
tinue until all such parameters return to normal. edema and oxygen delivery to the tissues.
More specific measures to monitor tissue perfu- Despite intense debate and research interest,
sion, including tissue oxygen tension devices19 neither crystalloids nor colloids have been
and gastric tonometry20, are not widely used. shown to be superior to one another regarding
The best fluid to use for volume expansion in survival outcome from hemorrhagic shock.
hemorrhagic shock remains a matter of debate. It is essential that a protocol be available
Both crystalloid and colloid are effective, but for the use of blood products in instances of
each has advantages and disadvantages21 (Table massive bleeding. In the UK, the responsibility
3). One recent large study showed no difference for maintaining such a protocol lies with the
Table 3 Intravenous fluids
Type of fluid Advantages Disadvantages

Crystalloids
Saline cheap; easily available; long history of use produces a hyperchloremic acidosis;
small procoagulant effect
Hartmann’s no risk of anaphylaxis; minimal direct effect on mildly hypotonic
the base deficit; easily available
5% dextrose no place in acute expansion of the intravascular hypotonic; no significant expansion of
space the vascular space; rapid distribution to
intracellular and extracellular spaces
Hypertonic rapid expansion of the intravascular space in insufficient data; uncertainty regarding
saline excess of the volume infused; possible beneficial possible adverse effects such as on the
effects on red cell and endothelial edema and immune system
capillary blood flow
Colloids
Gelatins largely remains in the intravascular space for risk of anaphylaxis; no clear survival
2–4 h advantage over crystalloids
4% human more physiological than gelatins; remains expensive; no clear survival advantage
albumin predominantly in the intravascular space for 12 h over crystalloids
Hydroxyethyl remains in the intravascular space for 12–24 h risk of coagulopathy, renal injury and
starch reticulo-endothelial accumulation
51
73
30 August 2006 14:19:25
Hospital Blood Transfusion Committee, a Recently, small-volume hypertonic resuscita-

multidisciplinary committee that all hospitals tion has been advocated for hemorrhagic shock.
must by law ensure is in place and answerable The concept is that a relatively small infused
to the hospital executive. It is unacceptable to volume will cause much larger expansion of
have situations where the laboratory insists on the circulation by drawing water into the intra-
blood samples being sent for blood count and vascular compartment. There is evidence that
coagulation studies before any blood products there may be beneficial effects of endothelial
are issued; the on-call hematology consultant and red cell edema and capillary flow, but there
should be actively involved and aid with the use are concerns regarding other potentially adverse
of blood, fresh frozen plasma, platelets, cryo- effects such as that on the immune system32.
precipitate and the use of the new recombinant This latter concern has not been shown to be a
activated Factor VII. problem in clinical practice33.
The Trendelenburg position is often used in Maintenance of the hemoglobin concentra-
the management of the hypotensive patient, but tion is essential to maintain oxygen-carrying
its benefit has been questioned. The concept is capacity and delivery to the tissues. Titration of
to displace blood from the lower limbs centrally, fluid and blood products to an exact hemoglo-
to increase preload and enhance cardiac output bin level in a rapidly bleeding patient is difficult.
as a temporary measure until adequate blood A hemoglobin level of 7–8 g/dl appears an
volume can be restored. However, there is little appropriate threshold for transfusion in the
proof that this theoretical benefit transpires in intensive-care population, with possible benefit
practice. Sibbald and colleagues in 197927 for a higher level of 9 g/dl for those with
showed that, in hypotensive patients, the ischemic heart disease34. It is logical to aim at
Trendelenburg position did not significantly the high end of the target range when resuscitat-
increase preload, but did increase afterload and ing from hemorrhagic shock as there is a ten-
blood pressure at the expense of cardiac output. dency to drift down. A target of 10 g/dl has been
A recent review of available data concludes that suggested as a reasonable goal in the actively
the Trendelenburg position ‘is probably not a bleeding patient35.
good position for resuscitation of patients who Coagulation disorders are both predisposing
are hypotensive’28. factors for, and consequences of, massive post-
The conventional approach to severe hemor- partum hemorrhage. A bleeding diathesis from
rhage, where the endpoint is euvolemia with a coagulopathy, thrombocytopenia or platelet
restoration of a normal blood pressure, heart dysfunction may result from pre-existing dis-
rate and cardiac output, has been questioned in ease, a pregnancy-acquired disorder, such as
the out-of-hospital trauma setting29. Although eclampsia, or treatment, such as aspirin. Mas-
not based on evidence from the obstetric popu- sive blood loss also creates both a coagulopathy
lation, the physiological rationale may still be and thrombocytopenia through dilution and
applicable. Falling blood pressure and cardiac consumption. These issues and their manage-
output, together with increased sympathetic ment are discussed in detail in Chapters 25 and
tone and release of endogenous catecholamines, 26.
reduce the rate of blood loss. Restoration of
these parameters without control of the bleed-
SUMMARY
ing will increase the total volume of blood loss,
increasing the degree of coagulopathy, reducing Rapid assessment of the presence of occult
oxygen-carrying capacity and ensuing multiple bleeding or intravascular volume depletion is
organ dysfunction. Low-volume fluid resuscita- essential. The body can compensate for blood
tion for hemorrhagic shock may be a possibil- loss such that, by the time obvious clinical signs
ity30 and the evidence suggests that volume are present, a significant volume can already
resuscitation should be deliberately limited to be lost and tissues already in a state of hypo-
the minimum required to sustain vital organ perfusion. Normal physiological adaptations in
function until the bleeding has been arrested, late pregnancy that persist into the postpartum
such as by surgery29,31. period can make recognition and quantification
52
74
30 August 2006 14:19:25
of intravascular loss difficult, and can render 6. Benedetti TJ, Kates R, Williams V. Hemo-
the body less capable of withstanding massive dynamic observations in severe preeclampsia
blood loss. This can be further complicated by complicated by pulmonary edema. Am J Obstet
pregnancy-related disease such as pre-eclampsia Gynecol 1985;152:330–4
7. Lewis G, Drife J. Why women die 2000–2002.
and its treatment, and modalities such as
Confidential Enquiry into Maternal and Child
hydralazine and magnesium.
Health. London: Royal College of Obstetricians
Assessment of both the degree of loss and the and Gynaecologists, 2004
response to volume replacement require clinical 8. National Institute for Clinical Excellence.
skills, invasive hemodynamic monitoring and Guidance on the use of ultrasound locating
the early involvement of senior clinicians. There devices for placing central venous catheters.
is no one correct fluid to use. It is usual to use Technology Appraisal Guidance 49. London: NHS
a combination of crystalloids or colloids and Publishers, 2002
blood products to maintain a hemoglobin con- 9. Cockings JGL. The Australian Incident
centration of near 10 g/dl during the actively Monitoring Study. Blood pressure monitoring –
bleeding period (7–9 g/dl is probably safe once applications and limitations: an analysis of 2000
incident reports. Anaesth Intensive Care 1993;21:
the active bleeding has been stopped). Coagulo-
565–9
pathies and thrombocytopenia also need to be
10. Gamby A, Bennett J. A feasibility study of the
corrected with appropriate transfusion products use of non-heparinised 0.9% sodium chloride for
and with active involvement of the hematolo- transduced arterial and venous lines. Intensive
gists. There may be a place for limited volume Critical Care Nursing 1995;11:148–50
expansion before the bleeding has been stopped 11. Branthwaite MA, Bradley RD. Measurement
surgically to reduce the volume lost, but this of cardiac output by thermal dilution in man.
must not be at the cost of demonstrable organ J Applied Physiol 1968;24:434
ischemia. 12. Swan HJ, Ganz W, Forrester J, Marcu H,
Prompt recognition, close monitoring of Diamond G, Chonette D. Catheterization of
volume status, rapid arrest of the bleeding and the heart in man with use of a flow-directed
balloon-tipped catheter. N Engl J Med 1970;283:
adequate volume resuscitation are all required
447–51
but when used together can reduce mortality
13. Harvey S, Harrison DA, Singer M, et al. Assess-
from postpartum hemorrhage. ment of the clinical effectiveness of pulmonary
artery catheters in management of patients
in intensive care (PAC-Man): a randomised
controlled trial. Lancet 2005;366:472–7
References
14. Nolan TE, Wakefield ML, Devoe LD. Invasive
1. Lewis G, Drife J. Report on Confidential Enquiries hemodynamic monitoring in obstetrics. A critical
into Maternal Deaths in the United Kingdom review of its indications, benefits, complications,
1997–1999. London: Royal College of and alternatives. Chest 1992;101:1429–33
Obstetricians and Gynaecologists, 2001 15. Godje O, Hoke K, Goetz AE. Reliability of a new
2. Umo-Etuk J, Jumley J, Holdcroft A. Critically ill algorithm for continuous cardiac output deter-
parturient women and admissions to intensive mination by pulse-contour analysis during
care: a 5-year review. Int J Obstet Anesth 1996; hemodynamic instability. Crit Care Med 2002;
5:79–84 30:52–8
3. Wong AYH, Kulandavelu S, Whiteley KH, Qu 16. Linton RA, Band DM, Haire KM. A new
D, Lowell-Langille B. Maternal cardiovascular method of measuring cardiac output in man
changes during pregnancy and postpartum in using lithium dilution. Br J Anaesth 1993;71:
mice. Am J Physiol Heart Circ Physiol 2002;282: 262–6
H918–25 17. Van Lieshout JJ, Wesseling KH. Editorial II:
4. Cohen WR, Galen LH, Vega-Rich M, Young JB. Continuous cardiac output by pulse contour
Cardiac sympathetic activity during rat preg- analysis. Br J Anaesth 2001;86:467–8
nancy. Metabolism 1988;37:771–7 18. Dabrowski GP, Steinberg SM, Ferrara JJ, Flint
5. Fujitani S, Baldisseri MR. Hemodynamic assess- LM. A critical assessment of endpoints of shock
ment in a pregnant and peripartum patient. Crit resuscitation. Surg Clin North Am 2000;80:
Care Med 2005;33(Suppl.):S354–61 825–44
53
75
30 August 2006 14:19:25
19. Huang YC. Monitoring oxygen delivery in the 27. Sibbald WJ, Paterson NA, Holliday RL,
critically ill. Chest 2005;128(5 Suppl 2):554–60S Baskerville J. The Trendelenburg position:
20. Totapally BR, Fakioglu H, Torbati D, Wolfsdorf hemodynamic effects in hypotensive and
J. Esophageal capnography during hemorrhagic normotensive patients. Crit Care Med 1979;7:
shock and after resuscitation in rats. Crit Care 218–24
2003;7:19–20 28. Bridges N, Jarquin-Valdivia AA. Use of the
21. Boldt J. Fluid choice for resuscitation of the trendelenburg position as the resuscitiation
trauma patient: a review of the physiological, position: to T or not to T? Am J Crit Care 2005;
pharmacological, and clinical evidence. Can J 14:364–8
Anaesth 2004;51:500–13 29. National Institute for Clinical Excellence. Pre-
22. Finfer S, The SAFE Study Investigators. A com- hospital initiation of fluid replacement therapy
parison of albumin and saline for fluid resuscita- in trauma: Technology appraisal guidance 74.
tion in the intensive care unit. N Engl J Med London: NHS Publishers, 2004
2004;350:2247–56 30. Stern SA. Low-volume fluid resuscitation for
23. Walters JH, Gottlieb A, Schoenwald P, Popovich presumed hemorrhagic shock: helpful or harm-
MJ, Sprung J, Nelson DR. Normal saline versus ful. Curr Opin Crit Care 2001;7:422–30
lactated Ringer’s solution for intraoperative fluid 31. Kreimeier U, Prueckner S, Peter K. Permissive
management in patients undergoing abdominal hypotension. Schweiz Med Wochenschr 2000;130:
aortic aneurysm repair: an outcome study. 1516–24
Anesth Analg 2001;93:817–22 32. Rocha-e-Silva. Small volume hypertonic resusci-
24. Scheingraber S, Rehm M, Sehmisch C, Finsterer tation of circulatory shock. Clinics 2005;60:
U. Rapid saline infusion produces hyperchlor- 159–72
aemic acidosis in patients undergoing gynaeco- 33. Kolsen-Petersen JA. Immune effect of hyper-
logical surgery. Anesthesiology 1999;90: 1265–70 tonic saline: fact or fiction? Acta Anaesthesiol
25. Laxenaire MC, Charpentier C, Feldman L. Scand 2004;48:667–78
Anaphylactoid reactions to colloid plasma sub- 34. Herbert PC, Wells G, Blajchman MA, et al. A
stitutes: incidence, risk factors, mechanisms. A multicenter, randomized, controlled clinical trial
French multicenter prospective study. Ann Fr of transfusion requirements in critical care.
Anesth Reanim 1994;13:301–10 Transfusion Requirements in Critical Care
26. Ruttmann TG, James MF, Aronson I. In vivo Investigators, Canadian Critical Care Trials
investigation into the effects of haemodilution Group. N Engl J Med 1999;340:409–17
with hydroxyethyl starch (200/0.5) and normal 35. Gutierrez G, Reines HD, Wulf-Gutierrez ME.
saline on coagulation. Br J Anaesth 1998;80: Clinical review: hemorrhagic shock. Crit Care
612–16 2004;8:373–81
54
76
30 August 2006 14:19:25
6
BLOOD LOSS: ACCURACY OF VISUAL ESTIMATION
A. Patel, R. Walia and D. Patel
As discussed in numerous other chapters, clini- predetermined simulated blood losses. Grape
cal estimation of blood loss is notoriously inac- jelly and pomegranate juice were used to
curate. The degree of inaccuracy varies greatly, simulate clots and blood. Each station had a
with many studies demonstrating that visual measured amount, ranging from 50 to 4000 ml.
estimates range from 30 to 50% of actual Simulated blood quantities were placed on sani-
losses1–3. Of great importance, this inaccuracy tary pads, delivery pads, basins and drapes and
increases with increasing blood loss2. When on the floor. This study was approved by the
translated into clinical practice, underestima- Institution Review Board.
tion may delay or deter identification and A total of 49 participants completed the skills
diagnosis of postpartum hemorrhage. This session. Participants included medical students,
circumstance may result in an unplanned physician assistants, nurses, obstetric and
obstetric emergency, with catastrophic out- gynecologic residents and attending staff. The
comes. To overcome this potential problem, results of the study are depicted in Figure 1.
multidisciplinary drills to highlight the nature The findings clearly document the inaccurate
of the problem may be of help, particularly in estimation of blood estimation, as well as the fact
training programs. that the accuracy of the estimate decreased with
We designed a labor ward drill to provide increasing blood volume. This was particularly
obstetric care teams an opportunity to assess true above 1000 ml. Of interest, the under-
their blood loss-assessing skills. A multi-station buttocks absorbent delivery pad was most decep-
blood loss simulation was designed with seven tive for estimating. In general, underestimates
stations which created opportunities to assess were similar for liquid and clots, but the 4000 ml
Overestimation
100% > 100%

Percent of participants
80% 61 99%
1 60%
60%
Correct estimation
40%
20%
0%
50 120 250 500 1000 2500 4000 Underestimation
Simulated blood volume (ml)
Figure 1 Accuracy of blood volume estimate
55
77
30 August 2006 14:19:39
station consisted entirely of ‘clots’ and was ACKNOWLEDGEMENT

most underestimated by the vast majority of
This chapter was provided by the authors, at the
participants.
request of the Editors, as a summary of their
This training program was enlightening for
article entitled ‘Accuracy of visual estimation of
participants to understand the limitations of the
blood loss’ which is to be published in The Inter-
visual assessment of blood loss. Repetitive inter-
national Journal of Gynaecology and Obstetrics as a
val sessions may aid individuals to increase
Supplement to Volume 94, 2006.
accuracy or to develop a personal blood loss
assessment coefficient to anticipate levels of
underestimation. Such a coefficient would be References
comparable to a golf handicap and of great use
1. Chua S, Ho LM, Vanaja K, Nordstrom L, Roy
to individuals who regularly are called upon to AC, Arulkumaran S. Validation of a laboratory
assess blood loss in a variety of situations. method of measuring postpartum blood loss.
Future studies could expand on this experiment Gynecol Obstet Invest 1998;46:31–3
with larger numbers and under more varied 2. Duthie SJ, Ven D, Yung GL, Guang DZ, Chan
conditions, of which the quality and quantity of SY, Ma HK. Discrepancy between laboratory
atmospheric lighting is most important. This determination and visual estimation of blood loss
information may be informative in the ongoing during normal delivery. Eur J Obstet Gynecol
education of labor and delivery room staff in Reprod Biol 1991;38:119–24
drills and other attempts to simulate real-time 3. Razvi K, Chua S, Arulkumaran S, Ratnam SS.
A comparison between visual estimation and lab-
emergency situations.
oratory determination of blood loss during the
third stage of labor. Aust N Z J Obstet Gynaecol
1996;36:152–4
56
78
04 September 2006 11:56:56
Section II
Causation
79
30 August 2006 14:19:39
7
DOPPLER EVALUATION OF HEMODYNAMIC CHANGES
IN UTERINE BLOOD FLOW
G. Urban, P. Tortoli, S. Ricci, M. Paidas, P. Vergani and P. Patrizio
INTRODUCTION Several studies have demonstrated a progres-

sive drop in impedance in all the compartments
The main uterine artery and its branches are
of the uterine circulation, from the main arteries
derivatives of the hypogastric artery. At the level
to the spiral arteries, as pregnancy advances2–4.
of the internal cervical os, the uterine artery
The impedance of the spiral arteries decreases
bifurcates into the cervical and corporal branches.
and blood flow velocities increase between the
At the uterotubal junction, the corporal branch
5th and 7th weeks of gestation. During that
turns laterally and upward toward the ovary
period, the hemodynamic status of the uterine
where it establish anastomoses with the ovarian
and arcuate arteries remains unchanged; it
artery, forming an arterial arcade that provides
is only after the 8th week of gestation that
perfusion to the upper aspect of the uterine cor-
a decrease in impedance and an increase in
pus. The blood flow takes a linear course in the
absolute flow velocities are detectable. This
hypogastric artery, then turns into a serpentine
delay between the changes in the spiral and
course in the uterine artery, and finally regains a
uterine arteries may represent the magnitude
linear course throughout the gestation, as the
of the increase of placental volume and spiral
uterus gradually increases in size.
arterial involvement, which is needed to
Approximately eight to ten arcuate arteries
effect appropriate and supportive uterine
originate from each uterine branch and enve-
hemodynamics5.
lope both the anterior and the posterior walls of
the uterus for about one-third of the thickness
of the myometrium1. These arteries take a tor-
INTRAPARTUM DOPPLER
tuous course and establish anastomoses with the
VELOCIMETRY
corresponding arteries from the contralateral
side in the midline of uterine myometrium. Fleischer and colleagues6 assessed 12 normal
The radial arteries arise from the arcuate parturients throughout labor with a continuous
arteries and are directed inward toward the wave Doppler unit to assess intrapartum
uterine mucosa. The total number is undefined changes in uterine and umbilical artery wave-
and most likely is dependent on parity and forms during labor. Each patient served as her
human biodiversity. own control. In the latent-phase of labor and
In the past, conventional Doppler techniques with intact membranes, as well as in the active
have been used unsuccessfully to study uterine phase after rupture of membranes or during
hemodynamic patterns in an attempt to provide oxytocin stimulation, no significant changes
diagnostic clues for the management of were noted in umbilical artery systolic/diastolic
postpartum hemorrhage. Recently, however, (S/D) ratios before, during or after a uterine
advances in signal acquisition and processing contraction. The uterine artery end-diastolic
have allowed precise and reproducible analysis flow velocity fell progressively during uterine
of velocity profile patterns and other variables contractions, reaching 0 when the uterine pres-
such as wall distension and shear rate at specific sure exceeded 35 mmHg. Despite intrauterine
sites of the uterine circulation. pressure of > 60 mmHg, the diastolic notch did
58
80
30 August 2006 14:19:39
Doppler evaluation of hemodynamic changes in uterine blood flow
not appear. This study demonstrated that, at flow through the arcuate artery is one-eighth (or
term, umbilical artery velocity waveforms do perhaps one-tenth, depending on the anatomic
not change over a wide range of uterine pres- variants) of the flow in the uterine artery, which
sures. Changes seen in uterine artery waveforms is three-quarters of the flow in the hypogastric
suggested that the end-diastolic component is artery at the start of the menstrual cycle. This
primarily determined by changes in the arcuate flow increases by one-third until ovulation (Fig-
and spiral arteries, both of which are affected ure 1) and remains constant until menstruation.
during uterine contractions. By way of comparison, after conception and in
early gestation, this flow increases until the end
of the second trimester, after which it remains
MULTIGATE SPECTRAL DOPPLER
stable throughout labor, at which time the arcu-
ANALYSIS
ate artery flow is one-fifth of the uterine, or
In our studies, we used a new ultrasound almost double the flow before pregnancy. In the
method to investigate blood flow, called multi- first and second stages of labor, this flow is
gate spectral Doppler analysis (MSDA), a tech- markedly reduced, if not totally discontinued,
nology that overcomes the limitations related to by compressive action of the uterine contrac-
the use of a single sample volume7. With this tions. During uterine contractions, the myo-
method, 256 small sample volumes are aligned metrial fibers also obliterate the flow in the
along an ultrasound scan line that intercepts the radial arteries, reflecting the fact that they are
blood vessel, and the Doppler data from each tributaries of the arcuate arteries (see above),
sample volume are independently analyzed to wherein flow stops until the end of contraction
produce a high-resolution flow profile. This and then rises to reach a steady-state flow until
non-standard method has been implemented in the next contraction ensues. During each con-
a system based on a proprietary electronic board traction, the placental lacunar space is com-
connected to a commercial ultrasound machine pressed, thus pumping the blood to the fetal
(Aloka SSD1400) and a personal computer. circulation throughout the umbilical vein. The
The board, which is installed on a PCI slot of a compression of the radial arteries during each
host PC, samples the I/Q signals and processes contraction acts as a valvular mechanism, avoid-
the data in an on-board DSP to carry out the ing reverse flow in the uterine circulation while
velocity profile. The profile is finally transferred directing the flow to the fetus. After delivery
in real-time to the PC to be displayed on the of the placenta, the resistance in the radial and
monitor. The hypogastric, uterine and arcuate spiral arteries decreases abruptly, being close to
arteries were investigated in women in labor ‘0’. As a consequence, there is an open flow of
before epidural anesthesia, after at least 1 h blood in the uterine cavity, which is contained
postpartum, and in women before pregnancy. by the compression caused by the prolonged
To our knowledge, no group of investigators uterine contractions. At this stage, the arcuate
had previously considered flow rates in terms and radial flow is almost absent. The absence
of the capacity to sustain a life-threatening of flow through the radial and spiral arteries
postpartum hemorrhage. facilitates the clotting mechanism in the
This Doppler evaluation shows essentially endometrial bed.
how we define bidimensional Doppler or Inefficiency of uterine contractions for
‘2-DD’. Multi-sample volumes from multigate whatever reason is a high-risk factor for post-
along the scan line depict a bidimensional partum hemorrhage. Likewise, an increase in
dynamic representation of the blood flow, the areal flow of the arcuate arteries, i.e. higher
where the horizontal axis is the depth and the than one-fifth of the flow in the uterine arteries,
longitudinal axis is the velocity. Actually, this is is also a potential risk for postpartum hemor-
the best estimation in real time of the blood flow rhage. The differences of areal flow between
throughout the vessels, showing an areal flow the hypogastric, uterine and arcuate arteries,
(cm2/s) from depth (cm) × velocity (cm/s). Our in various physiologic conditions, including
experience shows that, during menses, areal postpartum, are shown in Figure 1.
59
81
30 August 2006 14:19:40
Figure 1 Multigate spectral Doppler analysis using GASP software for areal velocity in women at the
ovulation phase of the normal menstrual cycle, in the antepartum phase of labor and at 1 h postpartum. In
each column, the first image is the conventional bidimensional image of the area of interest during multigate
acquisition, the following from top to bottom are hypogastric (H) artery velocity profile (areal flow), uterine
(U) artery velocity profile (areal flow), and arcuate (A) artery velocity profile (areal flow). All images are
frozen in systolic peak
References 3. Jauniaux E, Jurkovic D, Campbell S, et al. Dopp-

ler ultrasonographic features of the developing
1. Ramsey EM, Donner MW. Placental Vasculature placental circulation; correlation with anatomic
and Circulation. Stuttgart: Georg Thieme, 1980 findings. Am J Obstet Gynecol 1992;166:585–7
2. Jurkovic D, Januniaux E, Kurjak A, et al. 4. Arduini D, Rizzo G, Romanini C. Doppler ultra-
Transvaginal color Doppler assessment of the sonography in early pregnancy does not predict
uteroplacental circulation in early pregnancy. adverse pregnancy outcome. Ultrasound Obstet
Obstet Gynecol 1991;77:365–9 Gynecol 1991;1:180–5
60
82
30 August 2006 14:19:44
Doppler evaluation of hemodynamic changes in uterine blood flow
5. Makikallio K, Tekay A, Jouppila P. Utero- 9. Tortoli P, Michelassi V, Bambi G, Guidi F,

placental hemodynamics during early human Righi D. Interaction between secondary veloci-
pregnancy: a longitudinal study. Gynecol Obstet ties, flow pulsation and vessel morphology in the
Invest 2004;58:49–54 common carotid artery. Ultrasound Med Biol
6. Fleischer A, Anyagebunam A, Schulman H, et al. 2003;29:407–15
Uterine and umbilical artery velocimetry during 10. Urban G, Paidas MJ, Bambi G, et al. Multigate
normal labor. Am J Obstet Gynecol 1987;157: spectral Doppler analysis, new application in
40–3 maternal-foetal science. Ultrasound Obstet
7. Bambi G, Morganti T, Ricci S, et al. A novel Gynecol 2004;24:217–18
ultrasound instrument for investigation of 11. Morganti T, Ricci S, Vittone F, Palombo C,
arterial mechanics. Ultrasonics 2004;42:731–7 Tortoli P. Clinical validation of common carotid
8. Tortoli P, Guidi G, Berti P, Guidi F, Righi D. artery wall distension assessment based on
An FFT-based flow profiler for high-resolution multigate Doppler processing. Ultrasound Med
in vivo investigations. Ultrasound Med Biol 1997; Biol 2005;31:937–45
23:899–910
61
83
30 August 2006 14:19:44
8
PATHOPHYSIOLOGY OF POSTPARTUM HEMORRHAGE
AND THIRD STAGE OF LABOR
R.-U. Khan and H. El-Refaey
INTRODUCTION uterine volume leads to a reduction in placental

site surface area. As the placenta is a relatively
The physiology of postpartum hemostasis
rigid and inelastic structure, the surface area of
depends primarily upon mechanical events
its attachment site decreases when it is tightly
mediated by hormones, which induce strong
compressed.
uterine muscular contraction. Virtually all
According to Brandt, compression of the
recent studies focus on the latter, but the
placenta forces placental blood back into the
phenomenon cannot be understood without
sinuses in the decidua basalis6. These sinuses
examining why uterine contraction stops
become blocked by the action of strong
bleeding. Broadly speaking, myometrium and
myometrial contraction, and thus the com-
decidua are arranged such that powerful muscu-
pressed placenta attempts to force blood back
lar contractions after delivery favor hemostasis
into a high-resistance system. Ultimately, the
(Figure 1)1–3. Spiral arteries ‘fan out’ to create a
sinuses become so congested that they rupture.
low-resistance vascular bed in the intervillous
The blood from the ruptured sinuses tears the
space, which facilitates placental blood flow.
fine septae of the spongy layer of the decidua
This flow has been shown to decrease with
basalis, and thus the placenta is sheared off7.
muscular activity4. Third-stage contractions
Dieckmann and colleagues implied that this
are powerful and prolonged: they act to stop
‘retroplacental hematoma’ has no functional
placental blood flow and to separate the
value, and a subsequent investigation suggested
placenta and membranes.
that it is the contraction and retraction of the
uterine wall itself that cause it to rend itself
PLACENTAL SEPARATION AND apart from the placenta8.
UTERINE ACTIVITY Ultrasonographic investigations recently
corroborated that the Dieckmann theory is
Mechanical events
correct. Herman and colleagues conducted
The biomechanical events which lead to real-time ultrasonographic imaging of the
delivery of the placenta and its membranes third stage of labor and identified a ‘detach-
begin to take place even before the start of the ment phase’, wherein the placenta completes
second stage of labor. Membrane detachment its separation9. This detachment is preceded by
starts during the first stage and slowly spreads a ‘contraction phase’, in which the placental-
upwards from the internal os5. site uterine wall undergoes thickening. How-
As the trunk of the baby is delivered, the ever, the ‘latent phase’ before this thickening
uterus muscle fibers undergo a very powerful occurred varied between patients and was
contraction. Muscle fibers shorten, and the thought to determine the overall length of the
uterus is reduced in size and volume, a process third stage. Of interest, neither the latent phase
characterized as retraction. These events are nor the contraction phase was associated
probably facilitated by the spiral arrangement of with ultrasound evidence of retroplacental
uterine muscle fibers, whereby the reduction in hematoma formation.
62
84
30 August 2006 14:19:44
Pathophysiology of postpartum hemorrhage
membranes which follow the placenta and may

form a retroplacental clot, whereas this blood
escapes immediately in the Matthew Duncan
method. Second, placental separation is slower
in the Matthew Duncan method, allowing more
time for bleeding10. As clinicians are able to nei-
ther predict nor alter the method of placental
separation, the distinction between the Schultze
and Matthew Duncan methods is most proba-
bly clinically irrelevant.
Control of postpartum bleeding occurs by
contraction and retraction of the interlacing
myometrial fibers surrounding maternal spiral
arteries of the placental bed. Myometrial con-
traction compresses the spiral arteries and veins,
thereby obliterating their lumina. It is for this
reason that the myometrial fibers involved are
often referred to as ‘living ligatures’10. In addi-
tion, it is thought that some hemostasis occurs
by means of direct pressure as the uterine walls
are forced to firmly oppose one another as a
result of myometrial contraction.
It is worth noting the physiological effect
of early cord clamping, a common intervention
which is part of the active management of the
third stage of labor, is to retain blood in the pla-
centa, which prevents it from being so tightly
Figure 1 (a) Circular uterine muscle at rest: two compressed by the uterus. This, in turn, reduces
sets of crossing spiral; (b) at term: stretching of the the amount of myometrial retraction and con-
spirals (Goerttler, 19311). The innermost part of the traction, leading to more, not less, bleeding.
muscular layer has been described as superficially
However, this blood is thought to form a retro-
‘circular’ musculature, which is in fact two sets of
crossing spirals2. An alternative description of
placental clot, which speeds up the shearing off
muscle fibers travelling in all directions has been of the placenta. Ultimately, the consequent
described3. Both descriptions suggest that blood speedy delivery of placenta should lead to
vessels are compressed during contraction of muscle quicker hemostasis, but the intervention of cord
cells clamping is a paradox in that it involves causing
increased initial bleeding to lessen ultimate total
bleeding.
The two classical methods of placental deliv- Unfortunately, apart from the recent ultra-
ery result in different bleeding patterns. In the sound studies mentioned above, there is a dis-
Schultze method, separation begins in the cen- tinct paucity of information about the physical
ter of the placenta (the fetal surface), and this changes which lead to hemostasis and placental
part descends first, with the remainder follow- separation.
ing. The Matthew Duncan separation method
involves detachment of the leading edge of the
Endocrine mechanisms leading to
placenta, and the entire organ slips down and
mechanical events
out of the uterus sideways. The latter method is
much less common (20% of the total), but is Like all muscular activity, uterine contractility
supposed to result in more bleeding for two pos- depends on both electrical and hormonal
sible reasons. First, in the Schultze method, any stimuli. ‘Intrinsic’ activity may be mediated by
extravasated blood is trapped within the stretch receptors, although it is unclear whether
63
85
30 August 2006 14:19:45
such mechanisms are neural or neurohormonal. both to prevent and to treat postpartum hemor-
Two classes of hormones have been implicated rhage. At the same time, however, therapeutic
in third-stage uterine contractility, namely oxytocic agents used to augment labor are
oxytocin and prostaglandins. sometimes associated with uterine atony in
the third stage. In this latter circumstance, the
non-pulsatile administration of these agents
Oxytocin
may be leading to down-regulation of oxytocin
Interest in the role of oxytocin in the third stage receptors, as has been demonstrated in in vitro
has been partly motivated by the long-standing studies15. Despite the acknowledged therapeu-
experience with therapeutic oxytocin to prevent tic role of oxytocic agents in the third stage of
postpartum hemorrhage. Broadly speaking, labor, the true physiological role of oxytocin in
oxytocin causes increased uterine contractions the third stage remains unclear. It appears to
by acting on myometrial oxytocin receptors. have an inconsistent or paradoxical relationship
However, research has failed to show a clear with the third stage.
and simple relationship between physiological
oxytocin action and third-stage events for a
Prostaglandins
number of reasons. Oxytocin assays are notori-
ously unreliable, because the decidua synthe- Prostaglandins are potent stimulators of
sizes its own oxytocin. As a result, plasma levels myometrial contractility, acting via cyclic AMP-
do not reflect oxytocin concentrations at the mediated calcium release. The therapeutic use-
myometrium. Moreover, plasma oxytocin levels fulness of prostaglandin agents in postpartum
take no account of the density of myometrial hemorrhage lends credence to the possibility
oxytocin receptors, which has been shown to of a physiological role for prostaglandins in
participate in a complex control mechanism the third stage of labor. The prostaglandins
with oxytocin itself and other factors. Finally, involved in uterine contraction are produced in
oxytocinase, a plasma enzyme, denatures decidual tissue, placental tissue and fetal
oxytocin before it reaches its site of action11. membranes16. The uterotonic action of prosta-
During labor, oxytocin is released in a glandins does not depend on gestation. There
pulsatile manner, and both the pulse frequency are many classes of prostaglandin; the two
and duration increase12. Exactly what triggers classes implicated in uterine contraction are
the pulsatile oxytocin release is presently PGE2 and PGF2α.
unclear. Ferguson speculated that uterine Several observers have noted that large
stretching of the cervix stimulates oxytocin amounts of prostaglandin are released in the
release, leading to uterine contractions13. This third stage of labor. In an elegant experiment,
phenomenon so far has not been demonstrated Noort and colleagues measured plasma levels of
in humans, but there may be significant pres- prostaglandin metabolites during and up to 48 h
sure changes on adjacent pelvic organs and the after labor17. PGF2α levels reached their maxi-
vagina which result in neurological stimulation. mum and started to decline within 10 min after
A pulse of oxytocin does not necessarily placental separation (Figure 2). The subsequent
correspond to a uterine contraction, and some rapid decline in these levels suggested that the
women do not experience a rise in plasma prostaglandins arise from either necrosis/cellu-
oxytocin after the delivery of the baby14. More- lar disruption at the placental site, or from the
over, it is not necessary to have an oxytocin fetal membranes. The latter are known to be a
pulse in order to deliver the placenta and major source of prostaglandins. In vitro experi-
achieve hemostasis. Additional methods of con- ments have shown that intrapartum amniotic
trol must be involved. Whereas it is known that fluid triggers prostaglandin synthesis in fetal
myometrial oxytocin receptor density increases membranes. The ‘active agent’ in the amniotic
during pregnancy and labor, the precise con- fluid remains unknown16; however, these obser-
trols of this up-regulation are unknown15. vations are thought to reflect the active role of
For many years, synthetic oxytocic agents prostaglandins in securing hemostasis by way of
have been successfully used in the third stage myometrial contraction in the third stage.
64
86
30 August 2006 14:19:45
1000
750
PGF2 (pg/ml)
500
250
0
I II III 1 2 3 4 5 6 12 24 36 48
Time
Figure 2 Plasma PGF2α levels (pg/ml; mean ± standard equivalent of the mean). (I) In early labor and at
full dilatation; (II) at delivery of the fetal head; (III) at placental separation and up to 48 h after placental
separation (Noort et al., 198917)
The interaction between prostaglandins and Before and after delivery, subtle changes take
endogenous or therapeutic oxytocin in the third place in both coagulation factors and fibrino-
stage is not well understood. Numerous animal lysis agents. Plasma concentrations of clotting
experiments have demonstrated interactions factors increase not only during pregnancy but
between prostaglandins and oxytocin at luteo- also after delivery, which suggests a hyper-
lysis, initiation and maintenance of pregnancy, coagulable state20. However, after placental
and possibly at onset of labor18. However, ther- separation, the fibrinolytic potential of the
apeutic oxytocic agents used in the third stage maternal blood also increases, and this tends to
do not appear to have a significant effect on reduce the potential of blood to clot21.
prostaglandin metabolite concentrations19. Fur- These conflicting changes are difficult to rec-
ther studies are required to better understand oncile and are further complicated by changes
from where the prostaglandins arise, and what in platelet activity before and after delivery.
controls their release. However, there are indications that an inflam-
In the next few years, it is likely that matory response arises at the placental bed after
misoprostol, a prostaglandin E1 analogue with placental delivery22. Such a response would pro-
uterotonic properties, will play an increasing mote local coagulation. This finding is impor-
role in the management of the third stage, as it tant in terms of evolutionary advantage, because
is both cheaper and more thermostable than it allows prevention of hemorrhage at the pla-
existing agents. cental site, while elsewhere (particularly in deep
pelvic and leg veins) thrombi are less likely to
persist, due to the increased fibrinolysis.
Coagulation
von Willebrand disease (factor VIII defi-
Many standard obstetric text books provide ciency) is an important example of a coagulo-
only the vaguest of suggestions that coagulation pathy which can result in increased risk of
at the placental site represents an important postpartum hemorrhage. This is especially true
hemostatic mechanism. Whilst this is certainly in the disease variant featuring factor VIIIc defi-
true, the exact pathway(s) involved are unclear. ciency. In many ways, von Willebrand disease
65
87
04 September 2006 13:40:59
mimics a platelet adhesion dysfunction, and time. Whilst uterine atony is responsible for
indeed the only aspect of hematological 75–90% of primary postpartum hemorrhage,
hemostasis after placental delivery which can be traumatic causes of primary postpartum hemor-
emphasized with any certainty is the formation rhage (including obstetric lacerations, uterine
of platelet plugs at arterioles. Postpartum inversion and uterine rupture) comprise about
hemorrhage rates in von Willebrand’s disease 20% of all primary postpartum hemorrhage (see
are in excess of 15%, and it has been suggested Chapter 9). Significant but less common causes
that this hemorrhage is largely preventable by of postpartum hemorrhage include congenital
minimizing maternal trauma at delivery and giv- and acquired clotting abnormalities, which
ing prophylactic treatment with desmopressin comprise around 3% of the total24. Uterine
(DDAVP)23. atony is responsible for the majority of primary
In summary, the hemostatic mechanisms postpartum hemorrhage originating from the
during and after placental separation probably placental bed. Although the most important
involve the contraction of muscle sheaths around risk factor is a previous history of atonic post-
the spiral arteries, leading to platelet plug forma- partum hemorrhage (relative risk 3.3)25, many
tion, retraction of the uterus causing mechanical other important risk factors often found in
occlusion of arterioles facilitating platelet plug combination.
formation, and the activation of both the clotting Failure of the uterus to contract may be
cascade and fibrinolysis. As of this writing, many associated with retained placenta or placental
of these events are vague assumptions rather than fragments, either as disrupted portions, or more
demonstrated fact, as third-stage physiology rarely a succenturiate lobe. The retained mate-
research has been grossly neglected. The fact rial acts as a physical block against strong uter-
that for decades effective treatments have been ine contraction, which is needed to constrict
available for postpartum hemorrhage in the placental bed vessels, but, in most cases, dys-
developed world has acted as a true disincentive functional postpartum contraction is the pri-
for novel work and ideas. It is tragic that the third mary reason for placental retention. It is more
stage of labor, the most dangerous moment of likely for the placenta to be retained in cases of
pregnancy, is so poorly understood. atonic postpartum hemorrhage, and so the con-
traction failure often becomes self-perpetuating.
The reasons for this contractile dysfunction are
PATHOPHYSIOLOGY OF
unknown. The exception is uterine fibroids,
where the source of distension cannot be
Although most of the physiological processes in removed by uterine contraction, and must
the third stage of labor remain unclear, they therefore cause the atony. However, the uterus
broadly help to explain the etiology of atonic does not even have to be distended during the
postpartum hemorrhage. In this section, the third stage for contractile dysfunction to occur.
etiology and accompanying pathophysiology Distension prior to delivery, which occurs with
will be discussed. multiple pregnancy and polyhydramnios, also
affects the ability of the uterus to contract
efficiently after delivery, and is thus another
Uterine atony
risk factor for atonic postpartum hemorrhage.
The most common cause of postpartum hemor- When postpartum hemorrhage occurs fol-
rhage is uterine atony, i.e. failure of the uterus lowing an antepartum hemorrhage, the scenario
to contract. Primary postpartum hemorrhage is particularly difficult since there have been two
due to uterine atony occurs when the relaxed episodes of blood loss. A rare but serious com-
myometrium fails to constrict these blood plication of abruption is extravasation of blood
vessels, thereby allowing hemorrhage. Since into the myometrium, known as a Couvelaire
up to one-fifth of maternal cardiac output, or uterus, which impairs the physiological uterine
1000 ml/min, enters the uteroplacental circula- contraction/retraction hemostatic process.
tion at term, postpartum hemorrhage is capable However, the relationship between the extra-
of exsanguinating the mother within a short vasation process and uterine dysfunction is
66
88
30 August 2006 14:19:50
not fully understood. Chorioamnionitis has a the task of postpartum contraction/retraction,

similar effect for unknown reasons. Both ante- and thus hemorrhage ensues26. As contraction/
partum hemorrhage and chorioamnionitis also retraction are considered essential prerequisites
impair uterine contraction during the first two for both placental detachment and postpartum
stages of labor, and prolonged labor in general is hemostasis, the inference is that physiological
a risk factor for postpartum hemorrhage. Con- hemostasis from a lower segment placental bed
ventional wisdom suggests that delay in the first is impossible. This is obviously not the case,
two stages leads to uterine atony, but it is more however, as clearly not all cases of Grade IV pla-
logical to suggest that uterine dysfunction centa previa necessitate hysterectomy. The only
before onset of labor results in delay in all three possible conclusion is that there are qualitative
stages, and thus causes postpartum hemor- and quantitative differences in the musculature
rhage. As far as we are aware, there is no of the lower segment in different patients. A
ongoing research into this ‘universal uterine recent literature search on this topic confirms
dysfunction’. that the nature and origin of these differences
have never been investigated.
Biswas and colleagues have compared pla-
The lower segment as an implantation site
cental bed biopsy changes in placenta previa
In both placenta previa and placenta accreta, and normally implanted placenta; they have
the placental bed (and thus the postpartum shown that previa is associated with significantly
bleeding site) is in the lower segment. The pres- higher trophoblastic giant cell infiltration and
ence of lower segment implantation makes physiological changes of the myometrial spiral
hemorrhage and placental retention much more arterioles27. This work is typical of modern
likely. Although existing evidence is scanty, obstetric research in that it concentrates on
there are indications that the etiology of patho- antenatal events while ignoring postpartum
logical bleeding is inextricably linked with the events. However, the findings are interesting
anatomical and physiological limitations of the because they suggest that the seeds of potential
lower segment. placenta accreta are sown in most cases of
placenta previa. Nonetheless, no knowledge
regarding the qualitative features of lower
Placenta previa
segment myometrium exists.
In placenta previa, the placental site is located in
an abnormally low position. Atonic postpartum
Placenta accreta
hemorrhage is a recognized complication and,
even if Cesarean section is performed, severe Placenta accreta is morbid adherence of
intraoperative bleeding is a significant risk26. placenta such that it invades the myometrium.
The usual pharmacological methods used to It is rare; in 1990, the quoted incidence was
stem hemorrhage are often less effective. Surgi- around 1 in 2000 to 1 in 3500 pregnant women
cal methods, such as oversewing of bleeding in North America28. The condition is strongly
sinuses and the B-Lynch suture, are sometimes linked to lower segment implantation; it occurs
also ineffective so that hysterectomy proves nec- in up to 15% of women with placenta previa26.
essary. Hemorrhage is often not stopped unless The adherence is also associated with a
the entire lower segment is removed; a subtotal deficiency of decidua in the lower segment.
hysterectomy is often inadequate, and many The most common cause of this decidual
surgeons perform total abdominal hysterectomy deficiency is endometrial scarring, which may
as the operation of choice. Thus, the involve- be secondary to previous Cesarean section
ment of the lower segment makes it more likely or myomectomy, previous endometritis, past
not only that hemorrhage will occur, but also history of evacuation of retained products of
that standard treatment modalities will fail. conception or uterine abnormalities.
Authors in conventional texts often suggest Uterine surgery is a major risk factor for pla-
that, in lower segment implantation, the muscle centa previa and placenta accreta29. There is an
surrounding the placental bed is inadequate to increased tendency for placental implantation in
67
89
30 August 2006 14:19:50
the vicinity of the uterine scar with secondary because they tend to be relatively less
trophoblast invasion of the myometrium. Uter- amenable to medical treatment and sometimes
ine scarring is also known to be associated with necessitate radical surgical intervention, such as
an increased risk of scar dehiscence, febrile mor- hysterectomy.
bidity and other factors30. Thus the scar is clas- Whereas knowledge of the ultrastructure
sically considered to be a ‘weak area’. Scarring of placental bed musculature is lacking with
of muscle results in the normal tissue being regards to the upper segment, it is virtually non-
replaced by fibrous tissue. Intrauterine retrac- existent for the lower segment. New research
tion forces induced during labor tend to thin out into this area is urgently needed, because all
the lower segment, and these forces stretch the non-surgical therapeutic modalities for post-
scar to the point of rupture. Uterine rupture is partum hemorrhage involve enhancement of
not considered predictable31, but is more likely uterotonicity and, in the absence of sufficient
with each Cesarean section. Although poorly myometrium, they will simply not work. We
described in the literature, our personal clinical hypothesize that lower segment placentation/
experience suggests that, with each ensuing surgery leads to structural and thus functional
Cesarean section, the entire lower segment changes in the muscle histology. Thus, we
often seems to become thinner. Indeed, the envisage a new, clinically important class
lower segment may take on a translucent of postpartum hemorrhage, ‘lower segment
quality. This appearance is not limited to the postpartum hemorrhage’. This new subclass
scar itself. It is possible that the ‘weak scar’ will be best managed by new protocols which
in fact represents a generalized lower segment address the features specific to lower segment
weakness induced by previous surgery. involvement.
Clinical experience also suggests to us that it
is not enough to assume that postpartum hem-
orrhage is more common with lower segment References
implantation purely because lower segment 1. Goerttler K. Die Architektur der Muskelwand
muscle is inadequate to the task. In cases of des menschlichen Uterus ind ihre funktionelle
placenta previa and placenta accreta, the lower Bedeutung. [The architecture of the muscle
segment looks even thinner than normal. We bonds of the human uterus and their functional
hypothesize that the contractile nature of lower behavior.] Gegenbaurs morphologisches Jahrbuch
segment muscle, which is already less than that 1931;45–128
of the upper segment, is further lowered by the 2. Fuchs A, Fuchs F. Physiology of parturition. In
presence of the placenta. This would mean Gabbe S, Niebyl J, Simpson J, eds. Obstetrics:
Normal and Problem Pregnancies, 2nd edn. New
that implantation itself has an adverse effect
York: Churchill Livingstone, 1991:147–74
on lower segment myometrium. Furthermore,
3. Renn K. Untersuchungen ueber die raeumliche
there is a body of anecdotal evidence which Anordnung der Muskelbuendel im Corpus
implies that placental size and trophoblast bereich des menschilichen Uterus. Z Anat
invasion are greater in areas of limited decidual Entwicklungsgesch 1970;132:75–106
tissue, including implantation on scars and 4. Lees M, Hill J, Ochsner A, et al. Maternal pla-
in ectopic pregnancies. We hypothesize that cental and myometrial blood flow of the rhesus
trophoblast would invade readily into the poorly monkey during uterine contractions. Am J Obstet
decidualized lower uterine segment, increasing Gynecol 1971;110:68–81
the likelihood that placenta accreta will develop. 5. de Groot A. Safe motherhood – the role of oral
In terms of the previous discussion, it is (methyl)ergometrine in the prevention of
postpartum hemorrhage. MD Thesis, University
unfortunate that a dramatic and remorseless rise
of Nijmegen, 1995
in the Cesarean section rate is being observed
6. Brandt M. The mechanism and management of
throughout the developed world. This phenom- the third stage of labor. Am J Obstet Gynecol
enon will inevitably give rise to an increase 1933;25:662–7
in the complications associated with placenta 7. Dieckmann W, Odell L, Williger V, et al. The
previa, placenta accreta and scar rupture. placental stage and postpartum hemorrhage. Am
The complications are particularly important J Obstet Gynecol 1947;54:415–27
68
90
30 August 2006 14:19:51
8. Inch S. Management of the third stage of labour pregnancy and puerperium. Thromb Res 1989;
– another cascade of intervention? Midwifery 54:91–8
1985;1:114–22 22. Louden K, Broughton Pipkin F, Symonds F,
9. Herman A, Weinrauth Z, Bukovsky I, et al. et al. A randomised placebo-controlled study of
Dynamic ultrasonographic imaging of the third the effect of low dose aspirin on platelet reactivity
stage of labor. New perspectives into third stage and serum thromboxane B2 production in non-
mechanisms. Am J Obstet Gynecol 1993;168: pregnant women, in normal pregnancy, and in
1496–9 gestational hypertension. Br J Obstet Gynaecol
10. Sweet D, Kiran B. Mayes’ Midwifery. London: 1992;99:371–6
Balliere Tindall, 1997 23. Kadir R, Lee C, Sabin C, et al. Pregnancy in
11. Hirst J, Chibbar R, Mitchell B. Role of oxytocin women with von Willebrand’s disease or factor
in the regulation of uterine activity during XI deficiency. Br J Obstet Gynaecol 1998;105:
pregnancy and in the initiation of labour. Semin 314–21
Reprod Endocrinol 1993;11:219–33 24. Prendiville W, Elbourne D. Care during the
12. Fuchs A, Romero R, Keefe D, et al. Oxy- third stage of labour. In Chambers I, Enkin M,
tocin secretion and human parturition: pulse Keirse M, eds. Effective Care in Pregnancy and
frequency and duration increase during Childbirth, Vol 2. Oxford: Oxford University
spontaneous labour in women. Obstet Gynecol Press, 1989:1145–70
1991;165:1515–23 25. Stones R, Paterson C, Saunders N. Risk factors
13. Ferguson J. A study of the motility of the intact for major obstetric hemorrhage. Eur J Obstet
uterus of the rabbit at term. Surg Gynecol Obstet Gynecol Reprod Biol 1993;48:15–18
1941;73:359–66 26. Konje J, Whalley R. Bleeding in late pregnancy.
14. Thornton S, Davison J, Baylis P. Plasma In James D, Steer P, Weiner C, Gonik B,
oxytocin during third stage of labour: compari- eds. High-risk Pregnancy Management Options.
son of natural and active management. Br Med J London: Saunders, 1994:119–36
1988;297:167–9 27. Biswas R, Sawhney H, Dass R, Saran R, Vasishta
15. Phaneuf S, Asboth G, Carrasco M, et al. Desen- K. Histopathological study of placental bed
sitisation of oxytocin receptors in human myo- biopsy in placenta previa. Acta Obstet Gynecol
metrium. Hum Reprod Update 1998;4:625–33 Scand 1999;78:173–9
16. Brennand J, Leask R, Kelly R, et al. The influ- 28. Zahan C, Yeomans E. Postpartum hemorrhage:
ence of amniotic fluid on prostaglandin synthesis placenta accreta, uterine inversion and puerperal
and metabolism in human fetal membranes. Acta hematomas. Clin Obstet Gynecol 1990;33:422–31
Obstet Gynecol Scand 1998;77:142–50 29. Dickinson J. Previous Caesarean section. In
17. Noort W, van Buick B, Vereecken A, et al. James D, Steer P, Weiner C, Gonik B, eds.
Changes in prostaglandin levels of PGF2α and High-risk Pregnancy Management Options.
PGI2 metabolites at and after delivery at term. London: Saunders, 1994;207–16
Prostaglandins 1989;37:3–12 30. Enkin M, Wilkinson C. Manual removal of
18. Jenkin G. Oxytocin and prostaglandin inter- placenta at caesarean section (Cochrane review).
actions in pregnancy and at parturition. J Reprod In Keirse MJNC, Renfrew MJ, Neilson JP,
Fertil 1992;45(Suppl):97–111 Crowther C, eds. Pregnancy and Childbirth
19. Ilancheran A, Ratnam S. Effect of oxytocics on Module. In The Cochrane Pregnancy and Child-
prostaglandin levels in the third stage of labour. birth Database [database on disk and CDROM].
Gynecol Obstet Invest 1990;29:177–80 The Cochrane Collaboration; Issue 2, Oxford:
20. Wallenburg H. Changes in the coagulation Update Software, 1999. Available from BMJ
system and platelets in pregnancy-induced Publishing Group, London
hypertension and pre-eclampsia. In Sharp F, 31. Beasley J. Complications of the third stage of
Symonds E, eds. Hypertension in Pregnancy. labour. In Whitfield C, ed. Dewhurst’s Textbook
Ithaca: Perinatology Press, 1987:227–48 of Obstetrics and Gynaecology for Postgraduates,
21. Shimada H, Takshima E, Soma M, et al. Source 5th edn. Oxford: Blackwall Science 1995:
of increased plasminogen activators during 368–76
69
91
30 August 2006 14:19:51
9
OBSTETRIC TRAUMA
D. G. Evans and C. B-Lynch
ACUTE UTERINE INVERSION firmly contracted uterus is used as a piston to

push the placenta out, in the same manner that
Acute uterine inversion, defined as when the
a piston is used to push fluid out of the barrel of
uterus is turned inside out, is a rare but serious
a syringe. Pressure is applied with the palm of
complication of the third stage of labor. The
the hand in the axis of the pelvic inlet, in a
estimated incidence is approximately 1 in
downward and backward direction with the aim
20–25 000 deliveries1–3. As the estimate of a
of forcing the placenta out through the lower
later report was < 1 : 20004, the true incidence
genital tract. Unfortunately, application of
is unclear because some of the milder forms
Crede’s maneuver when the uterus is not con-
correct themselves spontaneously and are thus
tracted may well facilitate acute inversion. On
not recognized or reported.
the other hand, the Brandt Andrews maneuver,
also mentioned in standard textbooks of mid-
Classification wifery and obstetrics, a modification of Aris-
totle’s method of delivering the placenta by cord
Uterine inversion may be complete or incom- traction, recommends applying tension, but not
plete, depending on whether the fundus has traction, to the umbilical cord with one hand,
passed through the cervix5. When the uterine whilst the other hand is placed on the abdomen
inversion occurs within the first 24 h post- gently moving the uterus upwards and back-
delivery, it is classified as acute. Inversion wards. Today, controlled cord traction is
occurring after the first 24 h and up to 4 weeks standard practice for the third stage of labor.
postpartum is classified as sub-acute, and the Other etiological factors include forcibly
rare chronic inversion occurs after the 4th week attempting to expel the placenta by using fundal
postpartum. pressure when the uterus is atonic, and traction
on the umbilical cord in a fundally placed
placenta when the uterus is relaxed. It may also
Etiology
be brought about by a local atony, more particu-
The expulsion of the placenta was probably larly of the fundal placental site together with
intended by Nature to occur as a result of active contractions of the rest of the uterus.
gravitational forces, with the mother in the Other etiological factors include macrosomia,
same squatting position that is often adopted for polyhydramnios, multiple pregnancy, primiparity
defecation. When the third stage is conducted in and oxytocin administration5. In other instances,
the dorsal position, however, help may be nec- however, the inversion occurs spontaneously
essary for placental expulsion. Accordingly, the from sudden increased abdominal pressure as a
inappropriate management of the third stage of result of coughing, sneezing or straining.
labor is often implicated in the etiology of acute Chronic inversion may result from an acute
uterine inversion. Indeed, Crede’s method of inversion left unrecognized or from a sub-
placental delivery with uncontrolled cord trac- mucous fibroid which has prolapsed through
tion, referred to in most textbooks of midwifery the cervix. A placental polyp resulting from a
and older textbooks of obstetrics, may indeed retained cotyledon of the placenta may present
increase the risk of acute uterine inversion. The in the same fashion.
70
92
30 August 2006 14:19:51
Obstetric trauma
Diagnosis must be instituted prior to attempting replace-

ment, and the bladder should be catheterized.
Symptoms are acute and pronounced. Gener-
Antibiotic prophylaxis is advisable.
ally, the mother is aware of something coming
Any delay increases the difficulty in replacing
down and this is usually quickly followed by
the uterus, and the first health-care professional
unanticipated profound shock. The uterus may
present should make the initial attempt at
appear at the introitus outside the vagina and
replacement. This will be aided if regional anes-
the fundus is no longer palpable abdominally.
thetic is already in place8. The placenta should
In partial inversion, the fundus of the uterus
be left in situ and no attempt made to remove it.
may be indented and may or may not pass
The portion of the uterus that came down last
through the cervical os. In such instances, it is
should go back first, that is, the lower segment
neither palpable abdominally nor visible at the
initially and the fundus later. The hand is lubri-
vulva. Vaginal examination detects the inverted
cated with hibitane cream (or other suitable
body of the uterus, and, above and encircling it,
antiseptic if available) and placed inside the
the ring of the cervix. In all instances, pain may
vagina. With gentle maneuvers of the fingers
be severe due to stretching of the infundibulo-
around the cervical rim and simultaneous
pelvic ligaments and other viscera.
upward pressure with the palm of the hand, the
Shock is the outstanding sign, and may in
uterus is gradually replaced. The employment
part be neurogenic due to stretching of the
of force is dangerous, as the thinned-out lower
viscera and in part due to hemorrhage and
segment may be torn or otherwise traumatized.
hypovolemia. The degree of shock is propor-
The vaginal vault may already have been torn
tional to blood loss and hemorrhage is variable,
in some cases. The degree of shock does not
depending on whether any attempt has been
diminish until the uterus is replaced. In the
made to remove the placenta. Some bleeding
majority of instances, replacement of the uterus
will always be present unless the placenta is
is successful using this conservative method9.
completely adherent to the uterine wall. It is
If replacement is successful, the placenta
important to recognize that severe hemorrhage
should be manually removed with the aid of
will accompany any attempt at removing the
ergometrine or an oxytocic infusion. In under-
placenta before the uterus is replaced5,6. This
developed countries or in a home setting, boiled
eventuality is a special risk if the birth has been
water brought to a bearable temperature can be
attended by a traditional birth attendant (TBA)
used to soak clean towels or cloths to assist in
in parts of the underdeveloped world.
pushing and packing the vagina. This may facili-
tate replacement attempts and control further
blood loss. Bimanual massage of the fundus
Management
may improve contraction.
Acute uterine inversion is a true obstetric If replacement is unsuccessful, measures to
emergency6, and clearly one which may lead to relax the cervical retraction ring should be the
severe postpartum hemorrhage. If present and next line of therapy. Beta mimetics or amyl
available, a supportive team should be sum- nitrite inhalation can often relax the retraction
moned to the delivery suite for resuscitation ring sufficiently to allow uterine replacement9.
and protocol management (see Chapter 20). A similar effect is seen with the administration
Uterotomics, if started, are to be stopped of halothane anesthesia, but, unfortunately, use
and manual replacement attempted under ade- of this agent in sufficient doses can result in
quate and appropriate anesthesia followed by the unwanted and life-threatening complica-
delivery of the placenta assisted by restart of tions of uterine atony, hypotension and severe
oxytocin7. hemorrhage. Halothane is no longer used for
Elevation of the foot of the delivery table or these and other reasons. A 2 g intravenous
bed may relieve the tension on the viscera and bolus of magnesium sulfate can be used in the
reduce the pain and shock. Immediate resusci- hypotensive patient (0.25 mg of intravenous
tation with intravenous fluids is indicated via terbutaline in the stable patient) to relax
large-gauge venous access. Adequate analgesia the cervical contraction ring10. Intravenous
71
93
30 August 2006 14:19:51
nitroglycerine can be tried although it is not head-down (Trendelenberg) tilt. The patient is
commonly used. catheterized with an indwelling catheter and
Further attempts at replacement of the broad-spectrum antibiotics are administered.
uterus should take place under general anesthe- With the bowels packed upward and away from
sia in an operating theater equipped and ready the uterus, the obstetric surgeon places his
to perform a laparotomy. Before resorting to a hands in front and back of the lower segment
laparotomy, however, the tried and tested with the finger tips between and below the level
O’Sullivan hydrostatic technique11 should be of the inverted fundus. With progressive pres-
attempted. Here, the patient is first resuscitated sure on the fingertips of both hands which
to restore vital signs including adequate blood flip up simultaneously, the internal dimple
volume and pressure. The obstetric team and is replaced progressively by the rising uterine
anesthetist are summoned. fundus (Figure 1a–e)13. Uterine perfusion
Adequate analgesia is essential before: returns with re-establishment of uterine pulse
pressure.
(1) Attempt at repositioning without the use of
If this technique fails, then the mid-line
uterine relaxant;
abdominal incision can be extended upwards if
(2) If response is not imminent or sustained, an necessary. The inverted uterus resembles a fun-
anesthetist should provide uterine relax- nel; it is best to exteriorize the uterus. Instru-
ation to facilitate repositioning and the mental upward traction is applied to the round
administration of uterotonics; ligaments bilaterally using Allis or ring forceps,
while the assistant exerts upward pressure on
(3) General anesthesia is preferable, adminis-
the inverted parts from the vagina below. This
tered by an obstetric anesthetist. Digital
maneuver is the Huntington technique14,15.
repositioning should be maintained to
Failure at this stage warrants employing the
support and establish good uterine muscle
Haultain technique whereby an incision is made
tone;
vertically in the posterior cervix via the abdomi-
(4) 1–2 liters of saline at body temperature nal route, following the dimple as a guide to
should be infused into the vagina through relieve the constriction at this level. The assis-
rubber tubes placed in the posterior fornix, tant exerts upward pressure from the vagina to
whilst obliterating the introitus with the effect reduction and replacement16.
obstetrician’s hand. As the vaginal walls
distend, the fundus of the uterus rises and
the inversion is usually promptly corrected.
Once this is achieved, fluid is allowed to
slowly escape from the vagina whilst the
placement of the uterine fundus is achieved
and maintained.
When O’Sullivan first described this technique,
he used a douche-can and wide rubber tubing to
deliver the solution. More recently, a silastic
vacuum cup has been used to instil the sterile
solution into the vagina12. Until replacement
is effected, however, towels soaked in warm
hypertonic saline solution and draped over the
inverted uterus may reduce the edema which
will inevitably occur and which further impedes
replacement of the uterus. In extremely difficult
cases, replacement may require mid-line laparo-
tomy, with the patient cleansed and draped in
the Lloyd Davis (frog-legged) position with a Figure 1a Acute uterine inversion
72
94
30 August 2006 14:19:53
Obstetric trauma
Figure 1b Acute uterine inversion. Finger tips Figure 1d Acute uterine inversion. Return of
placed below fundus of uterus to facilitate reduction vascularity
Figure 1c Acute uterine inversion. Progressive Figure 1e Acute uterine inversion. Complete
reduction with some ischemia reduction and revascularization with normal clinical
features. (B-Lynch technique of non-instrumental
On return of the uterus to its normal posi- reduction of acute uterine inversion at laparotomy.
tion, the placenta should be removed manually ©Copyright ’05)
from the vagina, and uterine contraction main-
tained abdominally by bi-manual stimulation.
Ergometrine, oxytocic intravenous infusion, or intensive care or the high-dependency unit for
mesoprostyl can be administered. The posterior 24 h.
uterine incision, if used, is then repaired in lay- A sub-acute inversion is managed in a similar
ers, and the abdomen closed in the usual fash- manner but may resolve spontaneously as the
ion. The patient should be monitored in the uterus involutes4.
73
95
30 August 2006 14:19:59
In chronic inversion, the uterus involutes in rupture of a scar confers a 10–20-fold increase
its inverted position and remains in the vagina in risk of a subsequent rupture18,19.
as a soft swelling, which bleeds readily to touch Rupture of the uterus is generally sudden,
and shows areas of superficial ulceration. Pro- accompanied by severe abdominal pain and
longed inversion may result in conversion of the followed by vascular collapse. In many cases,
columnar epithelium of the uterine wall into a however, asymptomatic dehiscence takes place
stratified squamous epithelium. Replacement of during a vaginal delivery after a previous Cesar-
a chronic inversion can prove extremely diffi- ean section, when the dehiscence is gradual and
cult, due partly to the inevitable edema present retraction of the uterus arrests hemorrhage from
and the friable nature of the tissues. The tech- the wound. Because of this possibility, it is
niques adopted for replacing the acutely always necessary to exclude silent dehiscence
inverted uterus are no longer helpful in this by manual exploration of the uterus after
chronic situation. Bed rest, elevation of the foot delivery of the fetus when a scar is present
of the bed, antibiotic prophylaxis, and vaginal on the uterus.
cleansing with hibitane packs may be helpful to A major factor in spontaneous uterine
reduce the edema and treat any infections, but rupture is obstructed labor, especially in the
it may eventually be necessary to perform a developing world when women routinely
hysterectomy. If the chronic inversion is due to delivery without the benefit of the presence
the presence of a fibroid or a placental polyp, of trained health-care providers. Rupture may
initial removal of the polyp by ligating and be due to maternal or fetal causes (generally
cutting the pedicle as near to the base as macrosomia). Examples of maternal causes are
possible may facilitate replacement of the cephalopelvic disproportion from pelvic con-
inverted uterus. traction due to developmental, constitutional or
nutritional causes, abnormal presentation such
as shoulder presentation, breech or brow, per-
RUPTURED UTERUS
sistent mentoposterior face presentation, trans-
Uterine rupture is a serious obstetric complica- verse lie, fetal abnormality, hydrocephalus, fetal
tion with high morbidity and mortality. In tumor, fetal ascites, conjoined twins, maternal
developed countries, the increasing number of tumors, intrinsic cervical lesions, extrinsic fib-
Cesarean sections performed for minor degrees roids or tumor, locked twins, and rarely uterine
of disproportion, fetal distress or pre-eclampsia misalignment such as incarcerated retroverted
in primiparae is of considerable importance in uterus, and pathological uterine anteversion.
calculating the long-term risks associated with Additionally, grand multiparity, the use of
Cesarean section, particularly in terms of the uterotonic drugs to induce or augment labor,
incidence and risk of uterine rupture. Both the placenta percreta, and intrauterine manipula-
short- and long-term risks are accentuated in tion have all been implicated as causes of
resource-poor countries. uterine rupture19,20.
Uterine rupture may be complete when the The most common predisposing cause of
tear extends into the peritoneal cavity, or rupture during pregnancy is a weak scar follow-
incomplete when the serosa remains intact. ing a previous Cesarean section20. Rarely, rup-
The rupture may be spontaneous, traumatic or ture can occur following unrecognized injury
the result of scar dehiscence and may occur to the uterus at a previous difficult delivery.
either during pregnancy, early in labor or It may present with sudden severe abdominal
following a prolonged labor17. pains and collapse, or the symptoms may pres-
In developed countries, the most common ent gradually, when rupture is based on scar
cause of uterine rupture is dehiscence of a previ- dehiscence. If the onset is gradual, diagnosis
ous lower segment transverse Cesarean section may be difficult as the abdominal pain may be
scar. Rupture of a classical scar is eight times slight and accompanied only by alterations in
more common than that of a previous lower the fetal heart tracing, maternal tachycardia and
segment incision, and is far more apt to occur minimal vaginal bleeding. This triad is then
before rather than during labor. Previous followed by patient collapse, cessation of fetal
74
96
30 August 2006 14:19:59
Obstetric trauma
movement and easy palpation of the fetal parts Rarely, the uterus may rupture during early
if the fetus has been expelled completely into to mid-pregnancy or during labor in patients
the peritoneal cavity. If the patient is in a hospi- who have had a previous cornual ectopic preg-
tal and the catastrophe recognized at its onset, nancy. Here also, the rupture is dramatic, is
the outcome should be the safe delivery of the located over the repair site of the ectopic and
baby and repair of the uterus. If the patient is is characterized as a fundal blow-out. Sudden
not in a hospital, on the other hand, the catas- severe abdominal pain is experienced over the
trophe is just that, a catastrophe of a dead child fundus of the uterus followed by collapse.
and its mother. Rupture of a previously unscarred uterus is
Uterine rupture during labor is also most usually a catastrophic event resulting in death of
commonly due to dehiscence of a previous the infant, extensive damage to the uterus and a
Cesarean scar with pain over the scar, followed very high risk of maternal death from blood loss.
by sudden severe abdominal pain and collapse. The damage to the uterus may be so extensive
In grand multiparae with a friable inelastic uter- that repair is impossible and a hysterectomy is
ine wall, rupture may occur in early labor even required. In developed countries, the incidence
where there has been no previous scar or diffi- of ruptured uterus in an unscarred uterus
cult delivery, although this eventuality is not is approximately 1 : 10 000 deliveries22; in
nearly as common as rupture in the previously the underdeveloped countries, the data are
scarred uterus. Here, however, diagnosis may unknown. The incidence of rupture of a uterus
be difficult initially as the presentation may be with a previous Cesarean section scar is 1%22,23.
confused with a small accidental hemorrhage A trial of labor following a previous Cesarean
and therefore missed. section increases the risk of perinatal death and
Rupture after a prolonged labor is commonly rupture of the uterus compared to elective
due to obstructed labor, with marked thinning repeat Cesarean section. In one large Canadian
of the lower segment and increased retraction of study, a trial of labor following a previous
the upper segment resulting in the formation of Cesarean section was associated with an
a retraction or Bandl’s ring. The tear begins in increased risk of rupture (by 0.56%) but fewer
the lower uterine segment, may extend up to the maternal deaths than in an elective section (1.6
fundus or down into the vagina, or proceed vs. 5.6 per 100 000)19.
laterally into the broad ligament. If the tear In less developed countries, the incidence
is posterior, it may go through the posterior of uterine rupture varies from 1.4% to 25%,
vaginal fornix into the Pouch of Douglas with 25% in Ethiopian women with obstructed
(colporrhexis)20. If the rupture is in the lower labor23. Uterine rupture accounted for 9.3% of
anterior segment, the bladder is stripped from maternal mortality in one study from India and
its attachment to the lower segment. The perito- 6.2% in a study from South Africa24.
neum remains intact and so the rupture is char- A laparotomy is indicated when rupture of
acterized as incomplete. A multiparous patient the uterus is suspected. The patient is anesthe-
in obstructed labor will continue to have tetanic tized, cleansed, draped and the bladder cathet-
contractions until the uterus ruptures, whilst erized with an indwelling catheter. A mid-line
a primiparous patient will usually go out of lower abdominal incision should be used as this
labor. Classical clinical signs of a rupture in a may be extended cephalad if necessary. The
multiparous patient can be dramatic; abdominal fetus should be delivered expeditiously and the
pain is constant, the contractions become virtu- uterus delivered from the abdominal incision to
ally continuous initially with only short intervals assist in controlling the bleeding and assessing
between them and later no interval between the situation while resuscitative measures are
contractile forces, with the formation of a undertaken. In the series of over 1300 world-
Bandl’s ring followed by rupture and collapse. wide reported successful applications of the
The contractions then usually stop20–22, the B-Lynch (Brace) suture, 25 cases were applied
fetus is expelled into the peritoneal cavity, for persistent uterine atony after repair of a
the fetal parts are easily palpable and the uterus uterine rupture. In these cases, successful bleed-
adopts an altered shape. ing control and hemostasis were achieved (CBL
75
97
30 August 2006 14:19:59
world-wide communication www.CBLynch. domestic physical abuse or intimate partner

com)25. physical abuse can be as high as 13.5%29.
Hysterectomy may be necessary and should Developed countries are not immune from
have been consented, if at all possible. It is not this problem, however, and a large review of the
necessary to remove the ovaries merely because prevalence of abuse during pregnancy in the
this is easier in a crisis. As with a Cesarean hys- United States documented that between 0.9%
terectomy performed in late labor, the cervix is and 20.1% of pregnant women were abused by
no longer a discrete and circumscribed solid their partners. This figure covers all forms of
structure, easily delineated and permitting accu- abuse, emotional, physical and sexual30.
rate placement of vaginal clamps. In the acute Direct abdominal trauma by punching or
situation, hemostasis and avoidance of further kicking the abdomen increases the risk of
dissection are of paramount importance, and adverse outcome of the pregnancy. Adverse out-
the removal of the distal cervix is not critical. comes are more common with direct physical
The most difficult surgical situation occurs assaults than with motor vehicle accidents29,30.
when the rupture is extraperitoneal into the Partner abuse also tends to be a repetitive event,
broad ligament, with a massive hematoma dis- increasing the risk to the fetus31. In some coun-
torting the anatomy and obscuring the bleeding tries, partner abuse and violence against women
points. Here, it may be necessary to pack the is accepted as a cultural norm, thus reducing the
space, the end of the pack being brought out numbers of reported cases. Even in the Chinese
through a gap in the uterine repair20. A balloon community in Hong Kong and despite western
catheter with light traction may be used for socialization, it is not uncommon for women to
enhanced tamponade with or without the submit to their husbands and endure humilia-
application of the B-Lynch (Brace) suture tion for the sake of keeping their family
application26. together. Providing help for these pregnant
Other conservative surgery may be appropri- women is challenging32.
ate on occasions, for example, when simple Motor vehicle accidents account for 60–75%
repair of the tear may be preferable to hysterec- of cases of blunt trauma. Most injuries are
tomy. With an anterior rupture, the bladder minor, but, in the United States, between 1300
may be involved; the appearance of hematuria and 3900 women each year suffer a fetal loss as
is almost pathognomonic. Repair is undertaken a result of a motor vehicle accident27,28. Despite
and the bladder catheterized for 2 weeks. A pos- the majority of the injuries being minor, the
terior fornix rupture (colporrhexis) is relatively fetus is always at risk and careful assessment
easy to repair. Incomplete rupture is not usually must be carried out in all cases of blunt abdomi-
apparent until delivery has been achieved. It nal trauma resulting from motor vehicle acci-
will commonly declare itself by intrapartum or dents. Assessments must be frequent and
postpartum hemorrhage. It should always be repeated with special attention to conditions
excluded by manual exploration after delivery of commonly seen after such trauma. These
the fetus. Both bladder tears and colporrhexis include abruptio placentae, preterm labor,
may be missed if not anticipated. If this is the uterine rupture, fetomaternal hemorrhage,
case, bleeding may continue, to the surgeon’s direct fetal injury and fetal demise33.
dismay. The pattern of injury following automobile
accidents depends on the type of seat belt
restraints. An unbelted driver or passenger is
BLUNT ABDOMINAL TRAUMA
usually ejected from the vehicle or sustains
The three main causes of serious blunt abdomi- injuries when they hit the interior of the car.
nal trauma in pregnancy are motor vehicle The injuries are mainly to the face, head, chest,
accidents, falls and domestic or intimate abdomen and pelvis. With shoulder and
partner physical abuse. In the developed world, abdominal restraints, rib, sternum and clavicu-
the most common cause of blunt abdominal lar fractures are common, whereas in the
trauma is motor vehicle accidents27,28. In the lap-only belted, lumbar spine and hollow viscus
less developed countries, the incidence of injuries are more frequent. Sharp objects in the
76
98
30 August 2006 14:19:59
Obstetric trauma
pockets of the clothing on the person can cause relationship to the degree of trauma; abruption
additional trauma; a fountain pen may perforate may occur with very little evidence of injury to
the lungs or heart. Even bulky outdoor over- the mother. It usually, but not always, follows
clothing represents a hazard. With thick cloth- soon after the trauma.
ing, there is a short distance between the body Vaginal bleeding, abdominal pain, increased
of the person and the restraint. On impact, the uterine tone, uterine tenderness, high frequency
weight of the body causes acceleration forwards. contractions, and abnormal fetal cardiotoco-
The speed of contact between the person graphy are the classical clinical signs of a placen-
and the restraint can compound the damage tal abruption. In a posteriorly inserted placenta,
sustained to the body. severe backache and vaginal bleeding may
During the first trimester, the uterus is be significant symptoms. The bleeding may be
well protected within the pelvis and sustains revealed or concealed within the uterus. If
very little damage from blunt trauma. With concealed, in severe cases, the uterus becomes
advancing pregnancy, however, the uterus woody hard as described by Couvelaire, blood
becomes an abdominal organ and therefore having been extravasated into the muscular wall
more susceptible to trauma. The blood supply of the uterus. Fetal parts are impossible to feel
to the pelvis is markedly increased the more and the patient’s condition rapidly deteriorates
advanced the pregnancy, giving rise to retro- due to hypovolemia and pain.
peritoneal hemorrhage which can be life- The management of abruptio placentae
threatening. Bowel injuries are less common, as depends on the severity of the abruption, the
the bowel occupies the upper abdominal space nature of the general injuries sustained, the con-
later in pregnancy, is a more movable entity and dition of the fetus and the duration of the preg-
is not in the direct line of the trauma. nancy. The trauma surgeon and the obstetrician
Assessing the extent of trauma can be diffi- should work together in managing the patient.
cult, as clinical signs initially may be sparse. Establishing wide-bore intravenous access is
Patients should be assessed frequently to detect essential. The hematologist should also be
deterioration in their condition. The presence of involved. A complete thrombophilia screen
bony injuries should raise suspicion of intra- should be requested and cross-matched blood
peritoneal hemorrhage: rib fractures are associ- organized, together with fresh frozen plasma.
ated with liver and spleen injuries and pelvic A preterm uncompromised fetus should be
fractures with retroperitoneal hemorrhage and observed by continuous cardiotocography for a
injury to the genitourinary system. minimum of 6–12 h or by a Pinard stethoscope
Difficulty is often encountered in detecting a in less developed communities and, if the gesta-
small amount of bleeding into the peritoneal tion is under 34 weeks, the mother should be
cavity. As blood may be non-irritant, ultrasound given corticosteroids to mimimize the adverse
examination may be equivocal, and CT scan- effect of prematurity on lung maturation. If the
ning exposes the fetus to a large radioactive fetus is previable and compromised, vaginal
dose. The decision to proceed to a laparotomy delivery is the safest for the mother.
may therefore be entirely based on clinical In a term pregnancy with abruptio and
judgement. an uncompromised fetus, vaginal delivery is an
The most common cause of fetal death option. However, Cesarean section is advised if
in non-fatal accidents is abruptio placentae. In the fetus is compromised. If the fetus, on the
minor injuries, the incidence is between 1 and other hand, has died, induction of labor and
5%, in contrast to major trauma where the inci- vaginal delivery are appropriate and safe for the
dence may be as high as 30%. At the time of mother.
impact, the intrauterine pressure may be as high Preterm labor following blunt abdominal
as ten times the pressure reached at the height of trauma may be precipitated by extravasation
a labor contraction. Blunt trauma causes the of blood into the myometrium stimulating uter-
uterus to compress and then expand and ine contraction. Prostaglandin release may stim-
the placenta shears away from the uterine ulate uterine activity. Preterm labor requiring
wall. The degree of separation may bear no tocolysis occurs in 10–30% of cases of blunt
77
99
30 August 2006 14:20:00
abdominal trauma, but less than 1% deliver given Anti-D immunoglobulin to prevent sensi-
before 34 weeks. Tocolytics should be used tization. Sensitization can occur as early as the
guardedly, lest they mask the sign of abruption. 5th week of pregnancy. A Kleihauer–Betke test
Contractions following blunt abdominal trauma is essential to assess the magnitude of the
abate without treatment in 90% of cases. All fetomaternal hemorrhage and adjust the dose of
tocolytics have side-effects which the obstetri- Anti-D immunoglobulin accordingly.
cian should be familiar with: beta mimetics In all cases of blunt abdominal injuries, fetal
induce tachycardia and may mask the early assessment is of paramount importance. Cardio-
signs of abruption; non-steroidal anti- tocography is the most sensitive method of
inflammatory agents affect platelet and renal immediate fetal surveillance. Ultrasonography
function; and calcium channel blockers cause is only accurate in predicting 40% of cases of
hypertension. The fetal heart rate and the abruption. Uterine activity is the most sensitive
uterine contractions should be continuously indicator for predicting abruption following
monitored34. blunt abdominal trauma. Frequent contractions
Uterine rupture is a rare (1%) occurrence in have an adverse effect on fetal outcome.
blunt abdominal trauma; when it does occur, it As a guideline, patients who have sustained
is usually in association with a fractured pelvis. blunt abdominal trauma, but have no abdomi-
The site of rupture is commonly the fundus of nal tenderness, no vaginal bleeding and no
the uterus or the site of a previous uterine scar. contractions should be monitored 2-hourly for
Fetal mortality in such cases is 100%, and 6–12 hours. Patients with abdominal tender-
maternal mortality 10%35–38. Diagnosis may be ness, vaginal bleeding and contractions should
difficult with vague abdominal pain, uterine be monitored continuously43,44.
tenderness, but with easily palpable fetal parts,
and a poor trace or absence of a fetal heart on
cardiotocography. Fetal demise and maternal References
shock are more dramatic presentations. 1. Spain AW. Acute inversion of the uterus. J Obstet
If suspected, exploratory laparotomy in the Gynaecol Br Empire 1946;53:219
presence of the trauma surgeon is indicated. 2. Das P. Inversion of the uterus. J Obstet Gynaecol
Uterine repair should be undertaken only if the Br Empire 1940;47:525–48
patient is hemodynamically stable. If not, hys- 3. Fahmy M. Acute inversion of the uterus. Int J
terectomy should be performed. However, the Surg 1977;62:100
risk of a rupture in a subsequent pregnancy is 4. Watson P, Besch N, Bowes WA. Management of
high, and the patient and her family should be acute and subacute puerperal inversion of the
uterus. Obstet Gynecol 1980;55:12
advised this at an appropriate time.
5. Brar HS, Greenspoon JS, Platt LD, Paul RH.
Fetal injury occurs very infrequently follow-
Acute puerperal uterine inversion. J Reprod Med
ing blunt abdominal trauma. Fracture of the 1989;34:173–7
long bones or the skull is the most common 6. Wendel PJ, Cox SM. Emergent obstetric man-
injury and occurs in approximately 1% of cases. agement of uterine inversion. Obstet Gynecol Clin
If the fetus is distressed, immediate delivery is N Am 1995;22:261–74
called for. In the preterm non-compromised 7. Abouleish E, Ali V, Joumaa B, et al. Anaesthetic
fetus, delivery may be delayed, but serial management of acute puerperal uterine inver-
monitoring is advised39,40. sion. Br J Anaesth 1995;75:486–7
Fetomaternal hemorrhage occurs in up to 8. Catanzarite VA, Moffitt KD, Baker ML, et al.
30% of cases of blunt abdominal trauma, espe- New approach to the management of acute
puerperal uterine inversion. Obstet Gynecol 1986;
cially if the placenta is situated anteriorly. Most
68(Suppl):7–10
fetuses will have a normal outcome, although
9. Clark SL. Use of ritodrine in uterine inversion.
anemia, supraventricular tachycardia and fetal Am J Obstet Gynecol 1984;151:705
demise can occur depending on the extent 10. Grossman RA. Magnesium sulphate for uterine
of the fetomaternal hemorrhage41,42. Victims of inversion. J Reprod Med 1981;26:261–2
blunt abdominal trauma should be screened for 11. O’Sullivan JV. Acute inversion of the uterus. Br
Rhesus factor, and all Rhesus-negative mothers Med J 1945;ii:282–3
78
100
30 August 2006 14:20:00
Obstetric trauma
12. Ogueh O, Ayida G. Acute uterine inversion: a 29. Valladares E, Pena R, Oersson LA, Hogberg U.
new technique of hydrostatic replacement. Br J Violence against pregnant women: prevalence and
Obstet Gynaecol 1997;104,951–2 characteristics. A population-based study in Nic-
13. B-Lynch C. Non instrumental atraumatic aragua. Br J Obstet Gynaecol 2005;112:1234–48
stepwise reduction of acute uterine inversion. 30. Gazmararian JA, Lazorick S, Spitz AM, Ballard
In press TJ, Saltzman LE, Marks JS. Prevalence of vio-
14. Huntington JL. Acute inversion of the uterus. lence against women: a review of the literature.
Boston Med Surg J 1921;184:376–80 JAMA 1996;275:1915–20
15. Huntington JL, Irving PC, Kellogg PS. Abdomi- 31. Godwin TM, Breen MT. Pregnancy outcome
nal reposition in acute inversion of the puerperal and feto-maternal hemorrhage after non cata-
uterus. Am J Obstet Gynecol 1928;15:34–40 strophic trauma. Am J Obstet Gynecol 1990;162:
16. Haultain FWN. The treatment of chronic uter- 665–71
ine inversion by abdominal hysterotomy with a 32. Tiwari A, Leung WC, Leung TW, Humphreys J,
successful case. Br Med J 1901;ii:974 Parker B, Ho PC. A randomised controlled trial
17. Schrinsky DC, Benson RC. Rupture of the of empowerment training for Chinese abused
pregnant uterus: a review. Obstet Gynaecol Surv pregnant women in Hong Kong. Br J Obstet
1978;33:217–32 Gynaecol 2005;112:1249–56
18. Ritchie EH. Pregnancy after rupture of the preg- 33. Connolly A, Katz VL, Bash KL, McMahon MJ,
nant uterus. J Obstet Gynaecol Br Commonwealth Hansen WF. Trauma and pregnancy. Am J
1971;78:642–8 Perinatol 1997;14:331–6
19. Aguero O, Kizer S. Obstetric prognosis of the 34. Elliott M. Vehicular accidents and pregnancy.
repair of uterine rupture. Surg Gynaecol Obstet Aust NZ J Obstet Gynaecol 1966;6:279–86
1968;127:528–30 35. Williams JK, McClain L, Rosemurgy AS, Colo-
20. Hudson CN. Obstructed labour and its sequelae. rado NM. Evaluation of blunt abdominal trauma
In Lawson JB, Harrison KA, Bergstrom S, eds. in the third trimester of pregnancy: maternal,
Maternity Care in Developing Countries. London: and fetal considerations. Obstet Gynecol 1990;75:
RCOG Press, 2001 33–7
21. Wen SW, Rusen ID, Walker M, et al. Compari- 36. American College of Obstetricians and Gynecol-
son of maternal mortality and morbidity between ogists. Trauma during pregnancy. ACOG Tech-
trial of labor and elective Caesarean among nical Bulletin No. 161, November 1991,
women with previous caesarean delivery. Am J Washington DC
Obstet Gynecol 2004;19:1263–9 37. Mighty H. Trauma in pregnancy. Crit Care Clin
22. Miller DA, Goodwin TM, Cherman RB, Oaul 1994;10:623–34
RH. Intrapartum rupture of the unscarred 38. Dahmus MA, Sibai BN. Blunt abdominal
uterus. Obstet Gynecol 1997;89:671–3 trauma. Are there any predictive factors for
23. Gaym A. Obstructed labour in a district hospital. abruptio placentae or maternal-fetal distress. Am
Ethiop Med J 2002;40:11 J Obstet Gynecol 1993;169:1054–9
24. Rajaram P, Agarwal A, Swain S. Determinants of 39. Lavin JP, PolSky SS. Abdominal trauma during
maternal mortality: a hospital based study from pregnancy. Clin Perinatol 1983;10:423–38
South India. Ind J Matern Child Health 40. Goodwin TM, Breen MT. Pregnancy outcome
1995;6:7–10 and feto-maternal hemorrhage after non-
25. B-Lynch C. Persistent uterine atony after catastrophic trauma. Am J Obstet Gynecol 1990;
successful repair of ruptured uterus treated by 162:665–71
Brace suture, world-wide reports and personal 41. Pearlman MD, Tintinalli JE, Lorenz RP. A
communication. www.cblynch.com prospective controlled study of outcome after
26. Danso D, Reginald P. Intrauterine balloon cath- trauma during pregnancy. Am J Obstet Gynecol
eter with B-Lynch suture. Br J Obstet Gynaecol 1990;162:1502–10
2002;109:963 42. Rose PG, Strohm PL, Zuspan FP. Fetomaternal
27. Esposito TJ, Gens DR, Smith IG, Scorpio R, hemorrhage following trauma. Am J Obstet
Buchman T. Trauma during pregnancy. A Gynecol 1985;153:844–7
review of 79 cases. Arch Surg 1991;126:1073–8 43. Pearlman MD, Phillips ME. Safety belt use dur-
28. Hoff WS, D’Amelio LF, Tinkoff GH, et al. ing pregnancy. Obstet Gynecol 1996;88:1026–9
Maternal predictors of fetal demise in trauma 44. Pearlman MD, Tintinalli JE, Lorenz RP. Blunt
during pregnancy. Surg Gynecol Obstet 1991;172: trauma during pregnancy. N Engl J Med
175–80 1991;323:1609–13
79
101
30 August 2006 14:20:00
10
PLACENTAL ABNORMALITIES
M. K. Mehasseb and J. C. Konje
INTRODUCTION (3) Grade III: edge covers the os and the

placenta is asymmetrical;
Obstetric hemorrhage is still considered to be
one of the leading causes of maternal mortality (4) Grade IV: placenta symmetrically covers
and morbidity1–7. Placental abnormalities are the os.
a major contributor to obstetric hemorrhage.
Although this classification/grading system is
The common abnormalities include placental
the most common, others reflect the ultrasound
abruption, placenta previa, morbidly adherent
definition of the placental site.
placentae (accreta, increta, percreta) and
An alternative anatomical grading is provided
retained placenta. These abnormalities, for
below:
example, accounted for 36% of pregnancy-
related deaths due to hemorrhage in one series8. (1) Total placenta previa: where the internal cer-
vical os is completely covered by placenta;
PLACENTA PREVIA (2) Partial placenta previa: where the internal os
is partially covered by placenta;
Placenta previa is defined as partial or complete
insertion of the placenta onto the lower uterine (3) Marginal previa: where the edge of the
segment after fetal viability (20 weeks in placenta is at the margin of the internal os
developed countries and 24–28 weeks in devel- but does not cover it;
oping countries). Four grades of placenta previa
(4) Low-lying placenta: where the placental edge
are recognized (Figure 1):
does not reach but is in close proximity to
(1) Grade I: placenta is in the lower segment the internal os.
but its edge does not reach the internal os;
This other classification depends on the state of
(2) Grade II: lower placental edges reach the os the cervix at the time of examination; for exam-
but do not cover it; ple, a low-lying placenta at 2 cm dilatation may
I II III IV
Figure 1 Grades of placenta previa: I, Encroaching on the lower segment; II, reaching the internal os;
III, asymmetrically covering the internal os; IV, symmetrically covering the internal os
80
102
30 August 2006 14:20:02
Placental abnormalities
become a partial placenta at 8 cm dilatation. Clinical

Since a digital examination is not recommended
The most characteristic feature is painless
in cases of placenta previa, this alternative
vaginal bleeding. This is usually recurrent and
classification has limited clinical application.
unprovoked and does not commonly appear until
the end of the second trimester. The first
episode is usually self-limiting and is rarely so
Incidence and risk factors profuse as to prove fatal. However, the earlier in
The overall incidence is variable in different pregnancy the first presentation of bleeding, the
series, but is on average 1 in 300 deliveries9,10. more likely is the later need for early inter-
Risk factors for placenta previa include: vention. ‘Fetal distress’ is unusual unless the
hemorrhage is severe enough to cause maternal
(1) It is thought to be more common with shock.
advanced maternal age. This may, however, Abdominal palpation is not diagnostic but,
be a reflection of increased parity rather where the presenting part is free in late preg-
than age. However, a rise in incidence nancy or abnormal, placenta previa should be
from 0.3% to 0.7% over a 10-year period suspected.
has been attributed to a shift to an older Sometimes, especially with minor degrees of
obstetric population11. placenta previa, bleeding might not appear until
(2) Women of higher parity have a higher the onset of labor. This may clinically mimic
incidence12. abruption (see below).
The possibility of placenta previa should
(3) Multifetal gestation: secondary to an increase always be considered in women who present
in the surface area occupied by the placental with bleeding in the latter half of pregnancy.
mass13. The diagnosis can seldom be made solely on a
(4) The incidence increases with the number clinical basis.
of previous Cesarean section deliveries14,15. A There is no role for digital examination in the
single Cesarean section increases the risk by diagnosis unless in the operating theater as part
0.65%, two by 1.5%, three by 2.2% and of the double set-up with adequate preparation
four or more by 10%. A previous Cesarean for proceeding to Cesarean section. Although
section in association with placenta previa uncommon, where imaging (see below) is easily
increases the risk of Cesarean hysterectomy available and reliable, it remains useful in
almost four-fold11. cases where the diagnosis is in doubt and where
double set-up facilities are unavailable.
(5) Smoking doubles the risk of placenta
previa13–16. This may be attributed to
placental hypertrophy secondary to carbon Imaging
monoxide hypoxemia16.
The most commonly used method of placental
(6) Patients with placenta previa have 12 times localization in modern obstetrics is ultrasound
the usual risk of having a recurrent previa in scan. It is safe, accurate and non-invasive and is
subsequent pregnancies. the method of choice for making the diagnosis.
(7) For unclear reasons, fetal anomalies are The gestational age at which diagnosis is made
increased with placenta previa even after significantly influences accuracy. The earlier the
control for maternal age9. It is also scan is performed, the more likely the placenta
uncertain if there is an association with is to be found in the lower pole of the uterus.
intrauterine fetal growth restriction17,18. Consequently, routinely localization of the pla-
centa at the 20–22 weeks’ gestation anomaly
scan poses several questions. For example, is a
low-lying placenta at this stage predictive of pla-
Diagnosis
centa previa at the time of delivery, or does such
This can either be clinical or by imaging. screening reduce the adverse outcome for the
81
103
30 August 2006 14:20:02
pregnancy? More importantly, should the scan placenta at a later gestation is higher. Taipale
be repeated at 32–34 weeks’ gestation and does and colleagues24, for example, observed that, if
the asymptomatic patient have to be admitted a placenta overlapped the internal os by at least
and if so when? 25 mm at 18–23 weeks, the positive predictive
About 28% of placentas in women scanned value for previa at delivery was 40% with a sen-
transabdominally before 24 weeks are found to sitivity of 80%. Indeed, Becker and colleagues25
be ‘low’ but by 24 weeks this drops to 18% and found that, when the lower edge overlapped the
only 3% are low-lying by term19. Conversely, a os by at least 25 mm at 20–23 weeks, a vaginal
false-negative scan for a low placenta is found in delivery was not possible at all at term (i.e. it
as many as 7% of cases at 20 weeks20. Such had a 100% positive predictive value).
results are more common when the placenta is Although the routine practice of localizing
posterior, the bladder is over-filled, the fetal the placenta at the anomaly scan will no doubt
head obscures the margin of the placenta, or the continue, its limitations should be recognized
operator fails to scan the lateral uterine wall21. and, wherever possible, transvaginal ultrasound
A low-lying placenta is more common in early scans should be offered to improve the accuracy
pregnancy because the lower segment does not of localization and also measure the distance
exist. This apparent ‘placental migration’ is due from the os to the placental edge to help define
to enlargement of the upper segment and for- the degree of ‘low lying’.
mation of the lower segment, with many appar- Since there are no randomized, controlled
ently low placentas being found to be above the trials of the effect of routine localization versus
lower segment. Comeau and colleagues22 and no localization on the mother and fetus, current
Ruparelia and Chapman23 have shown that the practice will have to be governed by large-
more advanced the pregnancy is, the more accu- cohort case studies. It is, however, easy to
rate a scan diagnosis of placenta previa will be. assume that, where there is a low-lying placenta,
Transvaginal ultrasound is not only more education of patients and carers enhances the
accurate in diagnosing placenta previa but it chances of a better outcome for mother and
is more precise in defining the relationship of baby. Whether such patients should be rou-
the lower edge of the placenta to the internal tinely admitted at a later gestation is debatable.
os (Figure 2). Placenta previa is diagnosed on Most units do not, however, routinely admit but
transvaginal ultrasound scan when the placental repeat the scans at 32–24 weeks’ gestation.
edge is less than 3 cm from the internal os. Dashe and colleagues26 observed that persis-
Where the distance between the lower edge tence of placental previa diagnosed at 20–23
of the placenta and the internal cervical os weeks occurred in 34% of cases at delivery,
is measured, the persistence of a low-lying whereas 73% of those present at 32–35 weeks
persisted at delivery. A policy of routine scan-
ning will therefore reduce the false-positive rates
but will be at the expense of increasing workload
and patient anxiety. Unfortunately, none of the
studies reported on the proportion of patients
with low-lying placentas diagnosed at 32–34
weeks that later presented with bleeding. For
units not routinely scanning for placental site at
20 weeks, scanning for placental site is only
indicated with abnormal presentation, vaginal
bleeding or a chance finding when ultrasound
was undertaken in late pregnancy for other rea-
sons. For such cases, a transvaginal approach is
recommended as it is associated with a better
Figure 2 Transabdominal ultrasound scan with diagnostic accuracy, especially with posterior
superimposed color Doppler signal showing an placenta previa27,28. This approach has been
anterior placenta previa shown to be safe and is well tolerated.
82
104
30 August 2006 14:20:03
Transperineal sonography has been used by (3) Severe life-threatening, non-stopping

some investigators29. It allowed easy visualiza- (continuing) hemorrhage < or > 37 weeks;
tion of the internal os in all cases and carried a
(4) Hemorrhage associated with uterine
positive predictive value of 90% and a negative
contractions.
predictive value of 100% for placenta previa.
Magnetic resonance imaging (MRI) has The management of the third and fourth
been used to visualize placental abnormalities categories is immediate delivery by Cesarean
including placenta previa (Figure 3). It has the section. In the presence of non-life-threatening
advantages of being an objective, reproducible hemorrhage after 37 weeks’ gestation, a planned
test, minimizing the operator error. However, delivery is also advisable. This must, however,
due to cost and logistic limitations, it is unlikely be with the recognition that such a hemorrhage
that it will replace ultrasonography for routine could very rapidly become life-threatening.
evaluation30,31. For category 1, the best approach is expectant
management, although this must not be to the
detriment of maternal life.
Management
Management depends on whether the patient is
Expectant management
symptomatic or not. Asymptomatic patients
(where the diagnosis is made on ultrasound The perinatal mortality in placenta previa
scan) are managed expectantly, often as for is directly related to gestational age at
those with mild symptoms that are non- delivery32–35. Macafee33 and Johnson and col-
threatening to either the mother or fetus. leagues35 introduced expectant management of
Those with symptoms can be divided into placenta previa with the aim of achieving maxi-
four categories depending on the maternal con- mum fetal maturity possible while minimizing
dition, severity of hemorrhage, the gestational the risks to both mother and fetus, the overall
age and the neonatal facilities available in the objective being to reduce perinatal mortality,
unit. These categories are: and, at the same time, reducing maternal mor-
tality. This management plan was based on the
(1) Pregnancy < 37 weeks’ gestation without
assumption that most episodes of bleeding are
threat to the mother;
usually small, self-limited and are not fatal to
(2) Pregnancy > 37 weeks without threat to the the fetus or mother in the absence of provoking
mother; trauma (e.g. intercourse, vaginal examination)
or labor, and that a proportion of cases, particu-
larly those presenting early with lesser degrees
of previa, may resolve to permit vaginal delivery.
More recently, an improvement in perinatal
mortality attributed mainly to prolongation of
pregnancy with expectant management has also
been reported36,37.
Although Macafee33, in his regimen, advo-
cated that the patient remained as an inpatient
in a fully equipped and fully staffed maternity
hospital from the time of initial diagnosis to
delivery, a policy of permitting a selection of
women to return home has also been advo-
cated38 as part of expectant management, but
remains controversial. Cotton and colleagues32
reported no difference in their perinatal and
maternal mortality rates in those sent home and
Figure 3 MRI of a grade IV placenta previa those managed in hospitals, whereas D’Angelo
(completely covering the internal os) and Irwin39 suggested keeping the mother in
83
105
30 August 2006 14:20:05
hospital until delivery was justified, on the contractions, placental abruption, which is
grounds that neonatal mortality and morbidity widely regarded as a contraindication to toco-
and cost of treatment were reduced. Kaunitz lysis, cannot be excluded. In addition, placental
and colleagues40, in a review of 355 maternities abruption is said to coexist with placenta previa
managed at home, however, reported one in 10% of cases, and tocolytics cause maternal
intrapartum death from placenta previa. tachycardia and palpitations, features that could
This controversy is more evident in cases be confused with hypovolemia. Sampson and
of asymptomatic transvaginally diagnosed colleagues47 advocate the use of tocolytics in
placenta previa. For this group, it is becoming cases of placenta previa and uterine contrac-
increasingly acceptable to manage them at tions after 21 weeks and cite a reduction in
home41,42. In a review of 15 930 deliveries in perinatal mortality from 126 to 41 per 1000.
Edinburgh, Love and Wallace43 concluded that, Mild blood loss in placenta previa is not
while clinical outcomes were highly variable associated with a significantly high perinatal
and cannot be predicted from antenatal events, mortality. In contrast, significant blood loss is
the majority of cases with or without bleeding, associated with a high perinatal loss. Liberal use
irrespective of the degree of previa, could be of blood transfusion has been reported to nullify
managed on outpatient bases. There are no this effect32. Although there is no theoretical
randomized, controlled trials on the different limit to the number of blood transfusions a
approaches to this aspect of placental previa patient can have, most blood banks do not have
and such evidence is urgently needed to enable endless supplies. To optimize oxygen supply to
rationale decision-making in clinical practice. the fetus and protect the mother against antici-
Although most experts will advocate immedi- pated future blood loss, the ideal aim of trans-
ate delivery where there is severe hemorrhage fusion should be to maintain a hemoglobin level
(heavy vaginal bleeding producing maternal of at least 10 g/dl or a hematocrit of 30%.
hypovolemia), it is, however, not considered a Despite expectant management, 20% of
contraindication to expectant management43. women with placenta previa are delivered earlier
An aggressive approach involving admission than 32 weeks. These cases account for 73% of
and repeated blood transfusions improves peri- perinatal deaths32. They remain a major prob-
natal morbidity and mortality, especially where lem and, although the use of cervical cerclage
the bleeding occurs very early in pregnancy. In has been advocated, this is generally not
one study, where approximately 20% of the used. The neonatal mortality and morbidity are
women lost over 500 ml of blood, half of them reduced in this group by maternal corticosteroid
were managed expectantly with a mean gain administration.
in gestation of 16.8 days44. Crenshaw and Continuous hospitalization is costly and has
colleagues34, on the other hand, managed only an associated psychological effect of separation
43–46% of patients successfully with an aggres- on families. In developing countries, this may be
sive expectant approach, whereas Cotton and unaffordable to many families. However, the
colleagues32, with an aggressive approach, advantages include easy access to resuscitation
successfully managed 66% of women and prompt delivery and ensuring bed rest
expectantly. (which anecdotally has been thought to
During expectant management, preterm decrease the occurrence of hemorrhage) as well
labor remains a problem. Brenner and col- as limitation of activities. With improvement
leagues45 found that 40% of women with in transportation facilities and ambulance ser-
placenta previa had prelabor rupture of vices in developed countries, highly motivated
membranes, and went into spontaneous labor women who clearly understand the necessity of
or other developed problems that resulted in restriction of activity and are within, for exam-
delivery before 37 weeks’ gestation. Inhibiting ple, 15–30 min of the hospital perhaps may be
contractions in those with preterm labor would monitored at home. This will only apply to cases
seem logical, but some regard antepartum of grades I–III placenta previa or asymptomatic
hemorrhage as a contraindication to the use of grade IV. In all cases of expectant manage-
tocolytics46. With vaginal bleeding and uterine ment, cross-matched blood (two units) must be
84
106
30 August 2006 14:20:05
available at all times. However, in many hospi- or spinal anesthesia should not be regarded
tals this requirement is the sine qua non of a as contraindicated provided an experienced
limitation on therapeutic options. anesthetist is available.
The uterine incision should be a transverse
lower segment incision (if possible), provided
Method of delivery
there is a lower segment. Where the lower seg-
A diagnosis of placenta previa means delivery ment is non-existent or is very vascular, some
by Cesarean section, but this is not inevitable, obstetricians advocate a classical or a De Lee’s
especially where the previa is to a minor degree. incision. Scott51, however, believes that such
For the minor degree of placenta previa (grade I incisions are rarely justified because of their
or II anterior) and an engaged fetal head, consequences and long-term disadvantages.
pregnancy may be allowed to continue beyond When difficulties are encountered with trans-
37–38 weeks and vaginal delivery anticipated. verse lower segment incisions, these may be
In such patients, amniotomy followed by converted to inverted T-, J- or U-shaped
syntocinon can be considered. incisions.
In patients with a major grade of placenta Where the placenta is anterior, two
previa (grade II posterior, grades III–IV), deliv- approaches are available for incising the uterus,
ery should be by elective or emergency Cesarean going through the placenta or defining its edge
section. The former is ideal since emergency and going through the membranes above or
delivery has a negative effect on perinatal mor- below the placenta. The former approach
tality and morbidity, independent of gestational requires speed and may result in significant
age. Cotton and colleagues32 found that 27.7% fetal blood loss52. The latter, however, may be
of babies born as emergencies had anemia associated with undue delay in the delivery of
compared to 2.9% delivered electively. the fetus, more troublesome bleeding from a
Cesarean section for placenta previa poses partially separated placenta and therefore fetal
several problems. It should, therefore, never be blood loss and anoxia. Myerscough52 advises
left to an inexperienced obstetrician. The Royal against cutting or tearing through the placenta
College of Obstetricians and Gynaecologists in because of the inevitable fetal blood loss that
the UK recommends that such Cesarean sec- occurs as fetal vessels are torn. Because the
tions are performed by consultants. Although lower segment is less muscular, contraction and
general anesthesia was preferred to regional in retraction, which result in the occlusion of the
the past, there is an increasing tendency to using sinuses of the placental bed, are inadequate,
the latter especially as Frederiksen and col- and intraoperative hemorrhage is therefore not
leagues48 demonstrated not only its safety but a uncommon53. Where hemostasis is difficult
reduction in intrapartum blood loss compared to achieve, bleeding sinuses could be oversewn
to that with general anesthesia. with atraumatic sutures51. If this is unsuccess-
ful, packing the uterus is possible, but the major
disadvantage is that, by leaving the pack in situ
Procedure
during closure of the uterus, the bleeding may
Epidural analgesia is increasingly being continue but remain concealed for some time as
advocated for Cesarean section (in developed the pack is soaking through. The use of balloons
countries) although Moir49 considers placenta with a tamponading effect on the bleeding
previa to be an absolute contraindication to an placenta bed or intramyometrial injection of
epidural. This is because epidurals, by lowering prostaglandin F2α has been shown to be useful
the blood pressure, may critically reduce uterine in such cases52. More recently, where the facili-
and placental perfusion. Crawford50, however, ties are available, uterine artery embolization
believes that, in experienced hands, an epidural has been used with excellent results. The diffi-
is safe. Indeed, an increasing number of culty with this is planning to ensure that the
anesthetists offer regional anesthesia to these facilities and the interventional radiologist are
patients30,31. Where the patient’s condition is available on the labor ward during the delivery.
stable and there is no active bleeding, epidural When the bleeding remains uncontrollable,
85
107
30 August 2006 14:20:05
ligation of the internal iliac artery or even (1) Parity: more common in women of higher
hysterectomy may be necessary as the last resort parity;
(see Chapters 32 and 34).
(2) Age: more common in older women but
this may again be a reflection of parity
PLACENTAL ABRUPTION rather than age;
The Latin term abruptio placentae means ‘rending (3) Previous placental abruption: this varies
asunder of the placenta’, implying and denoting from 6 to 16.7% after one episode and
a sudden accident, which is a valid clinical 25% after two episodes. Up to 7% of those
characteristic of most cases. It represents bleed- with abruption severe enough to result in
ing due to premature separation of a normally fetal death have the same outcome in a
sited placenta after fetal viability. The initial subsequent pregnancy and 30% of all
event in abruption is bleeding into the decidua future pregnancies in women who have
basalis. a placental abruption do not result in a
living child56–59.
(4) Premature rupture of fetal membranes: a
Incidence and risk factors meta-analysis of 54 studies demonstrated
It occurs in about 1 in 200 pregnancies10, a three-fold increase in the risk of
although higher incidences have been abruption60 and this risk was much higher
reported54. When placentas are examined rou- with rupture between 20 and 36 weeks’
tinely, the incidence is much higher at 4.5%55 gestation and if rupture was for longer
suggesting that small episodes are more com- than 24 hours61,62;
mon than those diagnosed clinically. Placental (5) Cigarette smoking: the incidence in smokers
abruption can be revealed or concealed (Figure is almost double that in non-smokers;
4), the former occurring in 65–80% of cases. smokers who quit have an associated
The concealed type is clinically more dangerous reduction in risk;
as it is often associated with more severe
complications. (6) Cocaine use: significantly increased risk
Risk factors for placental abruption include: compared to non-users63;
Figure 4 Concealed and revealed placental abruption
86
108
30 August 2006 14:20:08
(7) Abdominal trauma: placental abruption be difficult to distinguish from the abdominal
complicates 1–6% of minor injuries and pain of abruption which is often unremitting.
up to 50% of major injuries64; Where this distinction is possible, the contrac-
tions are characteristically very frequent with a
(8) Sudden decompression of the uterus after
rate often of over five in 10 min71.
membrane rupture, e.g. in twin pregnancies,
The absence of abdominal pain does not
external cephalic version or pregnancies
exclude placental abruption, especially where
with polyhydramnios;
the placenta is posteriorly sited. This is evi-
(9) Unexplained raised α-fetoprotein; denced by the so-called ‘unsuspected or silent
abruption’ referred to by Notelovitz and col-
(10) Hyperhomocysteinemia and thrombo-
leagues72 and the higher pathological incidence
philias, especially factor V Leiden65,66;
of placental abruption found by Fox55. The
(11) Hypertensive disorders of pregnancy. presence of pain is probably indicative of
extravasation of blood into the myometrium. In
severe cases (grade 3), the pain is sharp, severe
Diagnosis
and sudden in onset. Some patients may, in
Unlike placenta previa where ultrasound is the addition, present with nausea, anxiety, thirst,
mainstay of diagnosis, the diagnosis of placental restlessness and a feeling of faintness, whereas
abruption is usually made on clinical grounds others may complain of absent or reduced fetal
(Table 1). Ultrasonography may, however, be movements.
helpful in certain instances, for example, where Some patients present with signs of shock
there is a large retroplacental hematoma. This, where the blood loss is significant (tachycardia
however, is an uncommon finding even in predominates; blood pressure has a poor rela-
severe cases. The symptoms and signs are diag- tion with blood volume in this condition). The
nostic in moderate to severe cases. In the mild presence of hypertension may, however, mask
forms, the diagnosis may not be obvious until true hypovolemia but an increasing abdominal
after delivery when a retroplacental clot is girth or a rising fundal height must raise the
identified. suspicion of significant concealed hemorrhage.
Placental abruption classically presents with Typically, the uterus is ‘woody hard’ in severe
vaginal bleeding, abdominal pain, uterine cases where the fetus is difficult to palpate and a
contractions and tenderness. Vaginal bleeding, continuous fetal heart rate monitor or real-time
however, is a symptom in no more than 70–80% ultrasonography is essential to identify the fetal
of cases28. The bleeding which occurs after the heart beat. The fetus may be ‘distressed’ with
36th week of gestation in about 50% of cases28 is fetal heart rate abnormalities or it may be dead.
characteristically dark and non-clotting. Because The former occurs in grades 1–2, but, in grade
labor is the commonest factor precipitating pla- 3, the latter is an invariable occurrence by
cental separation70, nearly 50% of patients with definition73. In severe cases complicated by
placental abruption are in established labor. The disseminated intravascular coagulation, there
presence of uterine contractions may, however, may be absence of clotting in the vaginal blood
loss, which is dark-colored. The incidence of
coagulopathy varies from 35 to 38%57,74 and
Table 1 Clinical picture of placental abruption71 this occurs mainly in the severe forms.
Symptom/sign Frequency (%) A vaginal examination reveals blood clots
in the vagina, which is typically non-clotting.
Vaginal bleeding 78 Serous fluid from a retroplacental clot may
Uterine tenderness or back pain 66 be confused with liquor. The cervix may be
Fetal distress 60 dilating since 50% of cases are in labor. If the
Preterm labor 22
membranes are ruptured, blood-stained liquor
High-frequency contractions 17
is usually present.
Hypertonus 17
Dead fetus 15 Ultrasound scan is not a sensitive method of
diagnosing placental abruption but is useful in
87
109
06 September 2006 16:49:22
excluding coincident placenta previa, which maternal resuscitation, as fetal death occurs
is present in 10% of cases. Where the retro- commonly in the severe variety of placental
placental clot is large, ultrasonography identifies abruption, often with coagulopathy. Once
it as hyperechogenic or isoechogenic when resuscitation has been initiated, the fetal mem-
compared to the placenta. Such echogenicity branes should be ruptured to hasten the onset of
may therefore be misinterpreted as a thick pla- labor. This is effective in most cases but, in a
centa75. A resolving retroplacental clot appears few, augmentation with syntocinon may be
hyperechogenic within 1 week and sonolucent needed. This must be administered cautiously
within 2 weeks. as uterine rupture could occur from an
Though ultrasound scan is not an accurate overstimulated uterus.
diagnostic tool, it is useful in monitoring cases Where the fetus is alive, the decision on how
managed. The size of the hematoma, location best to achieve delivery is not always easy. This
and change in size over time and fetal growth is compounded by the fact that the outlook for
are all parameters monitored by ultrasound the fetus is poor, not only in terms of immediate
scan. A Kliehauer–Betke test may be useful in survival but also because studies have shown
making the diagnosis when a patient presents that as many as 15.4% of liveborn infants do
with abdominal pain but without vaginal bleed- not survive77. However, delivering by Cesarean
ing or even in cases of ‘unsuspected or silent section when the fetus is alive has been shown
abruption’. in non-randomized, controlled trials to have a
better outcome than vaginal delivery (52% vs.
16%78; 20% vs. 15%73). Indecision and unnec-
Management
essary delays in performing Cesarean sections70
The severity of the abruption, the state of the are responsible for most poor results from
fetus and the gestational age of the pregnancy all Cesarean section in the last quarter of preg-
impinge on management which can be divided nancy. Cesarean section must therefore be con-
into general and specific measures. Sher and sidered in all cases where the fetus is alive,
Statland76 divided placental abruption into particularly if there is evidence of fetal distress.
three degrees of severity upon which manage- However, the presence of coagulopathy adds
ment can be based. These are shown in Table 2. considerable risk to the mother, and morbidity
General management is similar to that for and mortality could be increased by surgery.
any patient presenting with bleeding (see above Once the decision is to deliver and the fetus
under placenta previa). The specific measures is alive, the degree of abruption and the state
include immediate delivery, expectant manage- of the fetus must be taken into consideration
ment and management of complications. before delivering. When the abruption is severe,
Cesarean section must be performed once
resuscitation has commenced. Such delivery
Immediate delivery
should be performed promptly, especially as
This depends on the severity of abruption and most post-admission fetal deaths occur in
whether the fetus is alive or dead. If the fetus is fetuses delivered more than 2 hours after
dead, vaginal delivery should be the goal after admission.
Table 2 Grading of placental abruption (Sher and Statland)76
Grade Description
0 Asymptomatic abruption with a small retroplacental clot (< 150 ml)
1 Vaginal bleeding (150–500 ml); uterine tetany and tenderness may be present; no signs of maternal
shock or fetal distress
2 Vaginal bleeding; no signs of maternal shock; signs of fetal distress
3 Vaginal bleeding; marked uterine tetany yielding a board-like consistency on palpation; persistent
abdominal pain, with maternal shock and fetal demise; coagulopathy may be evident in 30% of cases
88
110
30 August 2006 14:20:08
If the abruption is mild to moderate, the may be rejected because it is expensive or causes
mode of delivery should be determined by the significant family disruptions.
condition of the baby, its presentation and During expectant management, fetal growth
the state of the cervix. In the presence of abnor- should be monitored by regular ultrasound scan
mal fetal heart rate patterns, immediate delivery as fetal growth restriction is a common finding
by Cesarean section is the option of choice. in association with placental abruption. The
However, if the decision is to deliver vaginally, timing of delivery depends on further vaginal
continuous fetal monitoring should be available bleeding, the fetal condition, gestational age
to enable early identification of abnormal fetal and available neonatal care facilities. If the
heart rate patterns. Golditch and Boyce79, bleeding episodes are recurrent, induction at
Lunan80 and Okonufua and Olatubosun78 have 37–38 weeks is advisable, provided there is no
all shown that the perinatal mortality is higher fetal compromise. Where the initial episode is
with vaginal delivery in the absence of electronic small and self-limiting and there are no acute
fetal monitoring. There is a place for the use of features of fetal compromise (e.g. abnormal
prostaglandins in the ripening of the cervix cardiotocography or a biophysical profile score
of women with mild abruption, but then the < 6) or chronic fetal compromise (growth
danger of inducing tetanic contractions must restriction, oligohydramnios or abnormal
always be borne in mind. Where amniotomy is umbilical artery Doppler recording), no evi-
feasible, this often hastens delivery but, where dence supports induction of labor. Despite this,
it is not possible, syntocinon can be used, it is nevertheless common for induction of labor
though once again maintaining vigilance for at term to be advocated in such patients, using
hyperstimulation. the speculative argument that some undetected
damage might have occurred to the integrity
and function of the placenta and, in the face of
Expectant management
such uncertainty, delivery at term confers more
This is recommended where neither the fetus advantages.
nor the mother are at risk. Unfortunately, the In a small proportion of cases, mild
lack of signs of fetal compromise on monitoring abruption may co-exist with labor. Whether
does not guarantee absence of deterioration in abruption provoked labor, or vice-versa, is
the fetal condition. With expectant manage- difficult to establish in these cases. The use of
ment, pregnancy is prolonged in the hope of tocolytics in such patients is controversial, as
improving fetal maturity and therefore survival. their use in the presence of placental abruption
It is ideal for pregnancies less than 37 com- is regarded by many as contraindicated since
pleted weeks of gestation; however, since neo- they may worsen the process of abruption46.
natal survival is virtually guaranteed > 34–35 Sholl81, however, stated that a trial of tocolytics
weeks’ gestation, there is no place in persisting in the presence of mild placental abruption
with such an approach for pregnancies > 34 and labor may successfully prolong pregnancy
weeks where fetal monitoring cannot be without jeopardizing the mother and fetus.
maintained. Expectant management is recom- There have as yet been no large trials to confirm
mended for patients in whom vaginal bleeding Sholl’s statement.
is slight, abdominal pain is mild and usually
localized and they are cardiovascularly stable.
Once a decision has been made on conservative Management of the complications of placental
management, the fetal condition must be abruption
monitored closely as it may change very quickly.
Complications of placental abruption include:
Expectant management can be in the com-
munity or in the hospital; admission is not asso- (1) Maternal shock: this may be disproportion-
ciated with a better outcome. However, where ate to the revealed blood loss. The type
patient education and access to hospital are of resuscitation should therefore be deter-
poor, admission may provide a safer option. It is mined by the clinical state of the patient.
perhaps in such communities that admission In most cases of shock, features of
89
111
30 August 2006 14:20:08
disseminated intravascular coagulation preservation of the intravascular volume are the

must be excluded, as their presence will cornerstones of treatment. Heparin has no role
require additional measures to replace in the modern management of consumptive
coagulation factors. coagulopathy. The presence of coagulopathy per
se is not an indication for Cesarean delivery but
(2) Disseminated intravascular coagulation
rather a strong contraindication. Also, the pres-
(DIC): treatment will require correction of
ence of an unfavorable cervix is not an indica-
the coagulation factor deficits, in consulta-
tion for Cesarean delivery, unless the condition
tion with a hematologist. Monitoring of
of the mother necessitates prompt delivery.
renal function is essential as acute tubular
The abdominal and uterine incisions can bleed
necrosis is a recognized sequela.
excessively when coagulation defects persist (see
(3) Ischemic necrosis of the distal organs (e.g. Chapter 25).
kidneys and brain): this requires adequate
fluid replacement.
Delivery
(4) Postpartum hemorrhage (secondary to DIC
Unless there is an obstetric contraindication to
or Couvelaire uterus): treatment is with
vaginal delivery or hemorrhage is so brisk that it
uterotonic drugs and other methods of
cannot be safely managed with vigorous blood
managing postpartum hemorrhage.
transfusion, every attempt should be made
(5) Isoimmunization: the administration of to deliver these patients vaginally (without
anti-D needs to be within 72 h but the jeopardizing maternal health).
quantity administered should be deter- Amniotomy (artificial rupture of membranes)
mined by Kleihauers–Betke test. and syntocinon infusion should be started. The
rigidity of the uterus or the presence of a high
intrauterine pressure should not deter the use of
Management of cases with intrauterine
syntocinon. If no rhythmic uterine contractions
fetal death
are superimposed on the background uterine
Where there is fetal death (in 20% of cases), hypertonus, then syntocinon should be started
placental detachment is usually greater than in standard doses. The benefits of achieving
50%, and approximately 30% of patients show a vaginal delivery override the risks of using
evidence of coagulopathy. Such cases should syntocinon. There is no evidence that its use is
therefore be classified as severe. The manage- associated with enhanced passage of thrombo-
ment should consist of the following: plastin into the maternal circulation and thereby
initiating or enhancing maternal consumptive
coagulopathy82. With intrauterine fetal demise,
Evaluation and replacement of blood loss
no time limit for delivery is necessary. The
Blood loss > 2500 ml is common. At least 4 maternal outcome is mainly dependent on the
units of blood should be cross-matched and diligence of fluid and blood replacement rather
transfusion commenced with packed red blood than on the interval to delivery83. Where the
cells, regardless of the initial vital signs as the cervix is unfavorable and maternal health is not
initial hematocrit or hemoglobin levels may in danger, prostaglandins may be used to induce
be normal due to hemoconcentration. Once delivery.
resuscitation has been established, subsequent
hypotension and tachycardia may then appear.
PLACENTA ACCRETA, INCRETA AND
PERCRETA
Management of coagulopathy (30% of cases)
This is a group of morbidly adherent placenta of
Without evidence of excessive vaginal bleeding, varying severity. Such morbid adherence occurs
no therapy is warranted even in the presence when the implantation site is lacking a sufficient
of abnormal laboratory results. Appropriate amount of decidua. Consequently, the physio-
replacement of blood components and logical cleavage plane through the decidual
90
112
30 August 2006 14:20:09
spongy layer is missing. This leads to one serosal bladder interface and the retroplacental
or more cotyledons being firmly anchored vessels, and, second, the presence of large
to the decidua basalis and even to the intraplacental lakes89.
myometrium. Preliminary work suggests that the applica-
The term ‘placenta accreta’ is used to des- tion of three-dimensional color power Doppler
cribe any placental implantation that is firmly ultrasound can be complementary to other tech-
adherent to the uterine wall. Placental villi are niques for antenatal imaging. It has been shown
anchored to the myometrium due to defective to be superior to magnetic resonance imaging in
decidualization. If villi invade the myometrium, this context90.
the condition is called placenta increta. If the
invasion goes as deep as reaching the serosal
Management
surface, this is called placenta percreta (see
Chapter 8). In most cases, problems arise after delivery of
Although uncommon, they are associated the baby. Most of the complications of morbidly
with a significantly high maternal morbidity adherent placentas are related to the problems
and sometimes mortality primarily due to of delivery or failure to deliver. Management
hemorrhage, uterine perforation, infection must therefore aim to minimize these complica-
and the associated surgical difficulties and tions (see Chapter 24).
complications84. Hemorrhage is the most common and this is
associated with attempts to detach the placenta
from the uterus. In most of these cases, unfortu-
Incidence
nately, the ultimate treatment is usually hyster-
They occur in about 1 in 2500 deliveries. There ectomy. Alternative approaches to management
has been a marked increase in the last 50 years, include uterine/hypogastric artery ligation or
probably secondary to the increase in Cesarean angiographic embolization. Sometimes, the
section delivery rates85. percreta type might even invade the bladder
Risk factors include implantation over the base, further complicating the surgical proce-
lower uterine segment overlying a previous dure required and making the control of
surgical scar or excessive uterine curettage hemorrhage very difficult.
resulting in Asherman’s syndrome. Placenta In cases of extensive placenta accreta
previa is identified in one-third of cases, and (involving most of the placental surface), bleed-
25% of women have had a previous Cesarean ing might be very limited until attempts at
delivery. Nearly one-quarter have previously manual removal are made. At times, traction on
undergone curettage and another quarter are the cord may lead to uterine inversion. Manual
grand multigravida (five or more)86. removal is usually not successful as the plane of
cleavage between the uterus and the placenta
cannot be developed. The safest treatment
Diagnosis
is usually hysterectomy. Attempts at uterine
Diagnosis is often not made until after delivery. conservation include piece-meal removal of
Some patients may present with vague features as much placental tissue as possible followed
which include a raised maternal serum by packing of the uterine cavity, but this
α-fetoprotein87 and bleeding before delivery, approach is reported to carry an unacceptably
although this is usually a consequence of high mortality rate of 25%86. Another option
placenta previa. Uterine rupture may occur to conserve the uterus is to leave the entire
antenatally due to myometrial invasion by placenta in situ if there is no bleeding. Kayem
chorionic villi at the site of a Cesarean section describes a case where spontaneous resorption
scar88. of the placenta occurred over 6 months follow-
The use of ultrasound Doppler color flow ing uterine artery embolization91. Other groups
mapping improves the diagnostic sensitivity. describe a similar approach, but using metho-
The two most sensitive criteria are, first, a trexate. The placenta spontaneously delivered
distance less than 1 mm between the uterine after 4 weeks92,93.
91
113
30 August 2006 14:20:09
RARE TYPES OF PLACENTAL 3. Hazelgrove JF, Price C, Papapachan VJ, et al.

ABNORMALITIES: SHAPE Multicenter study of obstetric admission to 14
intensive care units in southern England. Crit
Some anatomical variations in the shape of Care Med 2001;29:770
the placenta can give rise to serious postpartum 4. Zeeman GG, Wendel GDJ, Cunningham FG,
hemorrhage. These include bipartite placentas, et al. A blueprint for obstetric critical care. Am J
succenturiate lobes and placenta membranacea. Obstet Gynecol 2003;188:532
5. Jegosathy R. Sudden maternal deaths in Malay-
sia: a case report. J Obstet Gynaecol Res 2002;28:
Bipartite placenta 186
6. Rahman MH, Akhter HH, Khan Chowdury
Bipartite placenta occurs when the placenta is
ME, et al. Obstetric deaths in Bengladesh. Int J
occasionally separated into two lobes, and the Gynaecol Obstet 2002;77:161
division is incomplete with vessels of fetal origin 7. Nagaya K, Fetters MD, Ishikawa M, et al.
extending from one lobe to the other before Causes of maternal mortality in Japan. JAMA
ending in the umbilical cord. Its incidence is 2000;283:2661
about 1 in 350 deliveries94. 8. Chichakli LO, Atrash HK, Mackay AP, et al.
Pregnancy-related mortality in the United States
due to hemorrhage: 1979–1992. Obstet Gynecol
Succenturiate lobes 1999;94:721
In this abnormal form, one or more small 9. Crane JMG, Van Den Hof MC, Dodds L, et al.
accessory lobes develops in the membranes Neonatal outcomes in placenta previa. Obstet
Gynecol 1999;93:541
at a distance from the main placenta. The
10. Martin JA, Hamilton BE, Ventura SJ, et al.
succenturiate lobes usually connect to the latter Births: Final data for 2001. National Vital
with vascular connections of fetal origin. It Statistics report. Hyattsville: National Center for
can be considered to be like a small version of Health Statistics, 2002
the lobate placenta. The accessory lobe may be 11. Frederiksen MC, Glassenberg R, Stika CS, et al.
retained in the uterus after delivery, causing Placenta previa: A 22-year analysis. Am J Obstet
serious hemorrhage. Its incidence has been Gynecol 1999;180:1432
reported to be as high as 5%95. 12. Babinszki A, Kerenyi T, Torok O, Grazi V,
Lapinski RH, Berkowitz RL. Perinatal outcome
in grand and great-grand multiparity: effects of
Placenta membranacea parity on obstetric risk factors. Am J Obstet
Gynecol 1999;181:669–74
This type of placenta develops as a thin mem-
13. Ananth CV, Smulian JC, Vintzileos AM. The
brane-like structure with the whole of the fetal effect of placenta previa on neonatal mortality:
membranes covering the functioning villi. The a population-based study in the United States,
diagnosis can be made with ultrasound scan. It 1989 through 1997. Am J Obstet Gynecol 2003;
can give rise to serious hemorrhage as an associ- 188:1299–304
ation with placenta previa or accreta. One varia- 14. Gesteland K, Oshiro B, Henry E, et al. Rates
tion is the ‘ring-shaped’ or ‘horse-shoe’ placenta of placenta previa and placental abruption in
where the process does not involve the whole women delivered only vaginally or only by cesar-
placenta, but only a central part. This might ean section. J Soc Gynecol Invest 2004;11:208A
occur in about 1 in 6000 deliveries95. 15. Gilliam M, Rosenberg D, Davis F. The likeli-
hood of placenta previa with greater number of
cesarean deliveries and higher parity. Obstet
References Gynecol 2002;93:973
16. Williams MA, Mittendorf R, Lieberman E,
1. Bonnar J. Massive obstetric haemorrhage. Monson RR, Schoenbaum SC, Genest DR.
Baillieres Best Pract Res Clin Obstet Gynaecol Cigarette smoking during pregnancy in relation
2000;14:1–18 to placenta previa. Am J Obstet Gynecol 1991;
2. Gilbert TT, Smulian JC, Martin AA. Obstetric 165:28–32
admission to the intensive care unit: outcomes 17. Brar HS, Platt LD, DeVore GR, Horenstein J.
and severity of illness. Obstet Gynecol 2003;102: Fetal umbilical velocimetry for the surveillance
897
92
114
04 September 2006 13:44:57
of pregnancies complicated by placenta previa. 32. Cotton DB, Read JA, Paul RH, Quilligan EJ.
J Reprod Med 1988;33:741–4 The conservative aggressive management of
18. Ananth CV, Demissie K, Smulian JC, Vintzileos placenta previa. Am J Obstet Gynecol 1980;137:
AM. Relationship among placenta previa, fetal 687–95
growth restriction, and preterm delivery: a 33. Macafee CH, Millar WG, Harley G. Maternal
population-based study. Obstet Gynecol 2001;98: and foetal mortality in placenta praevia. J Obstet
299–306 Gynaecol Br Emp 1962;69:203–12
19. Chapman MG, Furness ET, Jones WR, Sheat 34. Crenshaw C, Jr, Jones DE, Parker RT. Placenta
JH. Significance of the location of placenta site in previa: a survey of twenty years experience with
early pregnancy. Br J Obstet Gynaecol 1989;86: improved perinatal survival by expectant therapy
846–8 and cesarean delivery. Obstet Gynecol Surv 1973;
20. McLure N, Dornan JC. Early identification of 28:461–70
placenta previa. Br J Obstet Gynaecol 1990;97: 35. Johnson HW, Williamson JC, Greeley AV. The
959–61 conservative management of some varieties of
21. Laing FC. Placenta previa: avoiding false- placenta praevia. Am J Obstet Gynecol 1945;49:
negative diagnoses. J Clin Ultrasound 1981;9: 398–406
109–13 36. Besinger RE, Moniak CW, Paskiewicz LS,
22. Comeau J, Shaw L, Marcell CC, Lavery JP. Fisher SG, Tomich PG. The effect of tocolytic
Early placenta previa and delivery outcome. use in the management of symptomatic placenta
Obstet Gynecol 1983;61:577–80 previa. Am J Obstet Gynecol 1995;172:1770–5;
23. Ruparelia BA, Chapman MG. Early low-lying discussion 1775–8
placentae – ultrasonic assessment, progress and 37. Towers CV, Pircon RA, Heppard M. Is tocolysis
outcome. Eur J Obstet Gynecol Reprod Biol 1985; safe in the management of third-trimester
20:209–13 bleeding? Am J Obstet Gynecol 1999;180:1572–8
24. Taipale P, Hiilesmaa V, Ylostalo P. Transvaginal 38. Silver R, Depp R, Sabbagha RE, Dooley SL,
ultrasonography at 18–23 weeks in predicting Socol ML, Tamura RK. Placenta previa: aggres-
placenta previa at delivery. Ultrasound Obstet sive expectant management. Am J Obstet Gynecol
Gynecol 1998;12:422–5 1984;150:15–22
25. Becker RH, Vonk R, Mende BC, Ragosch V, 39. D’Angelo LJ, Irwin LF. Conservative manage-
Entezami M. The relevance of placental location ment of placenta previa: a cost-benefit analysis.
at 20–23 gestational weeks for prediction of pla- Am J Obstet Gynecol 1984;149:320–6
centa previa at delivery: evaluation of 8650 cases. 40. Kaunitz AM, Spence C, Danielson TS, et al.
Ultrasound Obstet Gynecol 2001;17:496–501 Perinatal and maternal mortality in a religious
26. Dashe JS, McIntire DD, Ramus RM, Santos- group avoiding obstetric care. Am J Obstet
Ramos R, Twickler DM. Persistence of placenta Gynecol 1984;150:826–31
previa according to gestational age at ultrasound 41. Rosen DM, Peek MJ. Do women with placenta
detection. Obstet Gynecol 2002;99:692–7 praevia without antepartum haemorrhage
27. Tan NH, Abu M, Woo JL, Tahir HM. The role require hospitalization? Aust N Z J Obstet
of transvaginal sonography in the diagnosis of Gynaecol 1994;34:130–4
placenta praevia. Aust N Z J Obstet Gynaecol 42. Anon MG. Editorial comment. Aust N Z J Obstet
1995;35:42–5 Gynaecol 1994;34:130–1
28. Knuppel AR, Drukker JE. Bleeding in Late 43. Love CD, Wallace EM. Pregnancies complicated
Pregnancy: Antepartum Bleeding. Philadelphia: by placenta praevia: what is appropriate manage-
Saunders, 1986 ment? Br J Obstet Gynaecol 1996;103:864–7
29. Hertzberg BS, Bowie JD, Carroll BA, Kliewer 44. Ananth CV, Smulian JC, Vintzileos AM. The
MA, Weber TM. Diagnosis of placenta previa association of placenta previa with history of
during the third trimester: role of transperineal cesarean delivery and abortion: a metaanalysis.
sonography. Am J Roentgenol 1992;159:83–7 Am J Obstet Gynecol 1997;177:1071–8
30. Powell MC, Buckley J, Price H, Worthington 45. Brenner WE, Edelman DA, Hendricks CH.
BS, Symonds EM. Magnetic resonance imaging Characteristics of patients with placenta previa
and placenta previa. Am J Obstet Gynecol 1986; and results of ‘expectant management’. Am J
154:565–9 Obstet Gynecol 1978;132:180–91
31. Fraser R, Watson R. Bleeding During the Latter 46. Besinger RE, Niebyl JR. The safety and efficacy
Half of Pregnancy. London: Oxford University of tocolytic agents for the treatment of preterm
Press, 1989 labour. Obstet Gynecol Surv 1990;45:415–40
93
115
30 August 2006 14:20:09
47. Sampson MB, Lastres O, Tomasi AM, an updated meta-analysis. Reprod Toxicol 2001;
Thomason JL, Work BA, Jr. Tocolysis with 15:341–69
terbutaline sulfate in patients with placenta 64. Schiff MA, Holt VL. The injury severity score in
previa complicated by premature labor. J Reprod pregnant trauma patients: predicting placental
Med 1984;29:248–50 abruption and fetal death. J Trauma 2002;53:
48. Frederiksen MC, Glassenberg R, Stika CS, et al. 946–9
Placenta previa: a 22-year analysis. Am J Obstet 65. Kupferminc MJ. Thrombophilia and pregnancy.
Gynecol 1999;180:1432 Curr Pharm Des 2005;11:735–48
49. Moir DD. Obstetric Anaesthesia and Analgesia, 66. Gherman RB, Goodwin TM. Obstetric
2nd edn. London: Bailliere Tindall, 1980 implications of activated protein C resistance
50. Crawford JS. Priciples and Practice of Obstetrics and factor V Leiden mutation. Obstet Gynecol
Anaesthesia, 15th edn. Oxford: Blackwell, 1985 Surv 2000;55:117–22
51. Scott JS. Antepartum haemorrhage. In 67. Ananth CV, Oyelese Y, Yeo L, Pradhan A,
Whitefield CR, ed. Dewhurst’s Textbook of Vintzileos AM. Placental abruption in the
Obstetrics and Gynaecology for Postgraduates, 4th United States, 1979 through 2001: temporal
edn. Oxford: Blackwell, 1986 trends and potential determinants. Am J Obstet
52. Myerscough PR. Munro Kerr’s Operative Obstet- Gynecol 2005;192:191–8
rics, 10th edn. London: Bailliere Tindall, 1982 68. Ananth CV, Berkowitz GS, Savitz DA, Lapinski
53. Williamson HC, Greeley AV. Management of RH. Placental abruption and adverse perinatal
placenta praevia: 12 year study. Am J Obstet outcomes. JAMA 1999;282:1646–51
Gynecol 1945;50:987–91 69. Eskes TK. Clotting disorders and placental
54. Rasmussen S, Irgens LM, Bergsjo P, Dalaker K. abruption: homocysteine – a new risk factor. Eur
The occurrence of placental abruption in J Obstet Gynecol Reprod Biol 2001;95:206–12
Norway 1967–1991. Acta Obstet Gynecol Scand 70. Hibbard BM. Bleeding in late pregnancy. In
1996;75:222–8 Hibbard BM, ed. Principles of Obstetrics. London:
55. Fox H. Pathology of the Placenta. London: Butterworths, 1988
Saunders, 1978 71. Hurd WW, Miodovnik M, Hertzberg V, et al.
56. McShane PM, Heyl PS, Epstein MF. Maternal Selective management of abruptio placentae:
and perinatal morbidity resulting from placenta a prospective study. Obstet Gynecol 1983;61:
previa. Obstet Gynecol 1985;65:176–82 467
57. Pritchard JA, Brekken AL. Clinical and labora- 72. Notelovitz M, Bottoms SF, Dase DF, Leichter
tory studies on severe abruptio placentae. Am J PJ. Painless abruptio placentae. Obstet Gynecol
Obstet Gynecol 1967;97:681–700 1979;53:270–2
58. Paterson MEL. The aetiology and outcome of 73. Page EW, King EB, Merrill JA. Abruptio placen-
abruption placentae. Acta Obstet Gynecol Scand tae; dangers of delay in delivery. Obstet Gynecol
1979;58:31–5 1954;3:385–93
59. Rasmussen S, Irgens LM, Dalaker K. The effect 74. Green-Thompson RW. Antepartum haemor-
on the likelihood of further pregnancy of placen- rhage. Clin Obstet Gynaecol 1982;9:479–515
tal abruption and the rate of its recurrence. Br J 75. Nyberg DA, Cyr DR, Mack LA, Wilson
Obstet Gynaecol 1997;104:1292–5 DA, Shuman WP. Sonographic spectrum of
60. Ananth CV, Savitz DA, Williams MA. Placental placental abruption. Am J Roentgenol 1987;148:
abruption and its association with hypertension 161–4
and prolonged rupture of membranes: a 76. Sher G, Statland BE. Abruptio placentae with
methodologic review and meta-analysis. Obstet coagulopathy: a rational basis for management.
Gynecol 1996;88:309–18 Clin Obstet Gynecol 1985;28:15–23
61. Kramer MS, Usher RH, Pollack R, Boyd M, 77. Abdella TN, Sibai BM, Hays JM Jr, Anderson
Usher S. Etiologic determinants of abruptio GD. Relationship of hypertensive disease to
placentae. Obstet Gynecol 1997;89:221–6 abruptio placentae. Obstet Gynecol 1984;63:
62. Major CA, de Veciana M, Lewis DF, et al. 365–70
Preterm premature rupture of membranes 78. Okonofua FE, Olatunbosun OA. Caesarean
and abruptio placentae: is there an association versus vaginal delivery in abruptio placentae
between these pregnancy complications? Am J associated with live fetuses. Int J Gynaecol Obstet
Obstet Gynecol 1995;172:672 1985;23:471–4
63. Addis A, Moretti ME, Ahmed Syed F, 79. Golditch IA, Boyce NE. Management of
Einarson TR, Koren G. Fetal effects of cocaine: abruptio placentae. JAMA 1970;212:288–93
94
116
30 August 2006 14:20:09
80. Lunan CB. The management of abruptio 89. Twickler DM, Lucas MJ, Balis AB, et al. Color
placentae. J Obstet Gynaecol Br Commonw 1973; flow mapping for myometrial invasion in women
80:120–4 with a prior cesarean delivery. J Matern Fetal
81. Sholl JS. Abruptio placentae: clinical manage- Med 2000;9:330–5
ment in nonacute cases. Am J Obstet Gynecol 90. Lam G, Kuller J, McMahon M. Use of magnetic
1987;156:40 resonance imaging and ultrasound in the antena-
82. Clark S, Cotton DB, Gonik B, et al. Central tal diagnosis of placenta accreta. J Soc Gynecol
hemodynamic alterations in amniotic fluid Investig 2002;9:37–40
embolism. Am J Obstet Gynecol 1995;158:1124 91. Kayem G, Davy C, Goffinet F, Thomas C,
83. Brame RG, Harbert GM Jr, McGaughey HS Jr, Clement D, Cabrol D. Conservative versus
Thornton WN Jr. Maternal risk in abruption. extirpative management in cases of placenta
Obstet Gynecol 1968;31:224–7 accreta. Obstet Gynecol 2004;104:531–6
84. Zelop CM, Harlow BL, Frigoletto FD Jr, Safon 92. Henrich W, Fuchs I, Ehrenstein T, Kjos S,
LE, Saltzman DH. Emergency peripartum Schmider A, Dudenhausen JW. Antenatal diag-
hysterectomy. Am J Obstet Gynecol 1993;168: nosis of placenta percreta with planned in situ
1443–8 retention and methotrexate therapy in a woman
85. Comstock CH. Antenatal diagnosis of placenta infected with HIV. Ultrasound Obstet Gynecol
accreta: a review. Ultrasound Obstet Gynecol 2002;20:90–3
2005;26:89–96 93. Nijman RG, Mantingh A, Aarnoudse JG. Persis-
86. Fox H. Placenta accreta, 1945–1969. Obstet tent retained placenta percreta: methotrexate
Gynecol Surv 1972;27:475 treatment and Doppler flow characteristics. Br J
87. Hung TH, Shau WY, Hsieh CC, Chiu TH, Hsu Obstet Gynaecol 2002;109:587–8
JJ, Hsieh TT. Risk factors for placenta accreta. 94. Fox H. Pathology of the placenta. Clin Obstet
Obstet Gynecol 1999;93:545–50 Gynaecol 1986;13:501–19
88. Liang HS, Jeng CJ, Sheen TC, Lee FK, Yang 95. Benirschke K, Kaufman P. Pathology of the
YC, Tzeng CR. First-trimester uterine rupture Human Placenta, 4th edn. New York:
from a placenta percreta. A case report. J Reprod Springer-Verlag, 2000
Med 2003;48:474–8
95
117
30 August 2006 14:20:09
118
30 August 2006 14:20:09
Section III
General preventive measures
119
30 August 2006 14:20:09
11
ACTIVE MANAGEMENT OF THE THIRD STAGE OF LABOR
W. Prendiville and M. O’Connell
THE EVIDENCE varied widely, with most practitioners clamping

and cutting immediately. However, this proce-
Traditionally, the third stage of labor is defined
dure was not performed in most units in
as that time between the delivery of the baby
Austria, Denmark, Finland, Hungary and
and delivery of the placenta. Separation of the
Norway until the cord stopped pulsating7.
placenta from the uterine wall results from a
[Editor’s note: to add to this confusion, there is
combination of capillary hemorrhage and uter-
some concern that early clamping may deprive
ine muscle contraction. The length of the third
the neonate of an important amount of blood
stage of labor, and its subsequent complica-
and its associated hemoglobin, a factor of great
tions, depends on a combination of the length of
importance in many countries of the world.
time it takes for placental separation and the
The components of AMTSL, as outlined in the
ability of the uterine muscle to contract.
November 2003 Joint Statement of the Inter-
Preventive clinical management of the third
national Confederation of Midwives (ICM) and
stage of labor varies from the purely expectant
the International Federation of Gynecology
to an active approach, or some variation thereof.
and Obstetrics (FIGO), are administration
The expectant (‘pure’ physiological) approach
of a uterotonic agent (oxytocin is the drug of
involves waiting for clinical signs of placental
choice), controlled cord traction, and uterine
separation (alteration of the form and size of the
massage, after delivery of the placenta. See
uterus, descent and lengthening of the umbilical
further discussion below.]
cord and blood loss) and allowing the placenta
Given these circumstances, we reiterate this
to deliver either unaided using gravity or with
definition as the combined approach using three
the aid of nipple stimulation, as described in
component interventions: (1) a prophylactic
most maternity books1,2. In contrast, the full
uterotonic agent; (2) early clamping and divi-
active approach involves administration of an
sion of the umbilical cord; and (3) controlled
oxytocic agent, early umbilical cord clamping
cord traction.
and division and controlled cord traction for
delivery of the umbilical cord3–6.
In daily practice, the term ‘active manage- UTEROTONIC AGENTS
ment’ does not mean the same thing to all
health-care professionals. Marked variation in The commonly used uterotonic agents are
practice is seen. A recent survey of management divided into three groups: oxytocin and oxyto-
of the third stage of labor in 14 European coun- cin agonists, ergot alkaloids and prostaglandins.
tries confirmed this variation7. Whereas all units
professed to practice active management of the Oxytocin
third stage of labor, prophylactic uterotonics
were infrequently employed in units in Austria Oxytocin (Syntocinon) is a cyclic nonapeptide
and Denmark. Controlled cord traction was that is obtained by chemical synthesis. This syn-
almost universally used in Ireland and the UK, thetic form is identical to the natural hormone
but in less than 50% of units in the other 12 that is stored in the posterior pituitary and
countries surveyed. Policies with respect to released into the systemic circulation in
clamping and cutting the umbilical cord also response to suckling and labor. Oxytocin
98
120
30 August 2006 14:20:10
Active management of the third stage of labor
stimulates the smooth muscle of the uterus, Oxytocin vs. no uterotonics

more powerfully towards the end of pregnancy,
Seven trials including more than 3000 women
during labor, and immediately postpartum.
were described in this comparison. Variation
At these times, the oxytocin receptors in the
was noted not only in sample size and dose of
myometrium are increased8,9. The oxytocin
oxytocin used, but also in mode of administra-
receptor is coupled via G9q proteins to
tion, with the intramuscular route being used in
phospholipase C. The resultant activation trig-
three trials15–17 and the intravenous route used
gers release of calcium from intracellular stores
in the other four trials18–21. Those who received
and thus leads to myometrial contraction10.
prophylactic oxytocin had clear benefit in terms
Low-dose intravenous infusion of oxytocin
of postpartum hemorrhage (Figures 1 and 2).
elicits rhythmic uterine contractions similar in
Although debate surrounds the precise defini-
frequency, force and duration to those observed
tion of postpartum hemorrhage, this benefit was
during labor. Higher infusions can cause sus-
seen whether the cut-off was taken as > 500 ml
tained uterine contractions. A transient relax-
(relative risk (RR) 0.5, 95% confidence interval
ation of smooth muscle, with an associated brief
(CI) 0.43–0.59) or > 1000 ml (RR 0.61, 95%
episode of hypotension, flushing and reflex
CI 0.44–0.87). A trend towards a decreased
tachycardia, has been observed with rapidly
need for therapeutic oxytocin was also demon-
administered intravenous bolus injections11.
strated (RR 0.50, CI 0.39–0.64) in those
Oxytocin acts rapidly, with a latency period
who received prophylactic oxytocin. A non-
of less than 1 min after intravenous injection
statistically significant trend was also seen in
and 2–4 min after intramuscular injection.
the need for manual removal of the placenta
When oxytocin is administered by a continuous
in the prophylactic oxytocin group (RR 1.17,
intravenous infusion, the uterine response
95% CI 0.79–1.73) as well as an insignificant
begins gradually and reaches a steady state
increase in blood transfusion (RR 1.30, 95%
within 20–40 min. Removal of oxytocin from
CI 0.50–3.39).
plasma is accomplished mainly by the liver
and kidneys, with less than 1% excreted un-
changed in urine. The metabolic clearance rate
Oxytocin vs. ergot alkaloids
amounts to 20 ml/kg/min in the pregnant
woman12,13. Six trials including over 2800 women were
The prophylactic use of oxytocin in the third described in this comparison. Variation was
stage of labor has been described in a Cochrane noted not only in sample size, dose of oxytocin
review, where oxytocin alone was compared used, and preparation of ergot alkaloid used,
to no uterotonic and also compared to ergot but also in the mode of administration, with the
alkaloids14. intramuscular route being used in one trial15,
Study Oxytocin Control Relative risk (fixed) Weight Relative risk (fixed)
n/N n/N 95% Cl (%) 95% Cl
De Groot 1996 25/78 55/143 10.2 0.83 [ 0.57, 1.22 ]

Howard 1964 15/470 25/470 6.6 0.06 [ 0.32, 1.12 ]
Ilancheran 1990 0/5 0/5 0.0 Not estimated
Nordstrom 1997 104/513 175/487 47.1 0.56 [ 0.46, 0.70 ]
Pierre 1992 37/488 126/482 33.3 0.29 [0.21, 0.41 ]
Poeschmann 1991 7/28 10/24 2.8 0.60 [0.27, 1.33 ]
Total (95% Cl) 1582 1611 100.0 0.50 [ 0.43, 0.59 ]

Total events 188 (Oxytocin), 391 (Control) 2
Test for heterogeneity chi-square=18.10 df=4 p=0.001 l =77.9%
Test for overall effect z=8.76 p<0.00001
0.1 0.2 0.5 1 2 5 10
Favors oxytocin Favors control
Figure 1 Comparison of oxytocin vs. no uterotonics (all trials), with outcome of postpartum hemorrhage
(clinically estimated blood loss ≥ 500 ml). Cochrane review14
99
121
30 August 2006 14:20:11
the intravenous route in four trials18,19,22,23 and tetanic contractions (in approximately 2 min),
both intravenous and intramuscular routes in as does intravenous administration, but with a
one trial24. longer duration of activity26. Oxytocin agonists
Little differential effects were demonstrated for the prevention of postpartum hemorrhage
between these two oxytocics (Figures 3 and 4). are currently the subject of an additional
Ergometrine was associated with more manual Cochrane review27.
removal of the placenta (RR 0.57, 95% CI
0.41–0.79) and a statistically insignificant
Syntometrine
tendency towards hypertension (RR 0.53, 95%
CI 0.19–1.58). Syntometrine is a mixture of 5 IU oxytocin
(Syntocinon) and 500 µg ergometrine maleate.
Ergometrine is a naturally occurring ergot
Oxytocin agonists
alkaloid which stimulates contractions of the
Carbetocin appears to be the most promising of uterine and vascular smooth muscle. Following
these agents in preventing postpartum hemor- administration, it increases the amplitude and
rhage25. Carbetocin is a long-acting synthetic frequency of uterine contractions and tone and
octapepetide analogue of oxytocin, with agonist thus impedes uterine blood flow. Intense con-
properties and similar clinical and pharmaco- tractions are produced and are usually followed
logical properties to naturally occurring oxy- by periods of relaxation. Hemostatis is caused
tocin. It binds to oxytocin receptors and causes by contractions of the uterine wall around
rhythmic contractions of smooth muscle of the bleeding vessels at the placental site.
uterus, increases the frequency of contractions The vasoconstriction caused by ergometrine
and increases uterine tone. Intramuscular injec- involves mainly capacitance vessels, leading to
tions of carbetocin provide similar responses to an increase in central venous pressure and blood
Study Oxytocin Control Relative risk (fixed) Weight Relative risk (fixed)
n/N n/N 95% Cl (%) 95% Cl
De Groot 1996 7/78 16/143 14.2 0.80 [ 0.34, 1.87 ]

Nordstrom 1997 32/513 43/487 55.3 0.71 [ 0.45, 1.10 ]
Pierre 1992 7/488 21/482 26.5 0.33 [ 0.14, 0.77 ]
Poeschmann 1991 2/28 3/24 4.0 0.57 [ 0.10, 3.14 ]
Total (95% Cl) 1107 1136 100.0 0.61 [ 0.44, 0.87 ]

Total events 48 (Oxytocin), 83 (Control) 2
0.1 0.2 0.5 1 2 5 10
Favors oxytocin Favors control
Figure 2 Comparison of oxytocin vs. no uterotonics (all trials), with outcome of severe postpartum
hemorrhage (clinically estimated blood loss ≥ 1000 ml). Cochrane review14
Study Oxytocin Ergot alkaloids Relative risk (fixed) Weight Relative risk (fixed)
n/N n/N 95% Cl (%) 95% Cl
De Groot 1996 1/78 2/146 1.7 0.94 [ 0.09, 10.16 ]

Fugo 1958 55/324 36/149 60.5 0.70 [ 0.48, 1.02 ]
Sorbe 1978 10/508 32/543 37.8 0.34 [ 0.17, 0.68 ]
Total (95% Cl) 908 838 100.0 0.57 [ 0.41, 0.79 ]

Total events: 66 (Oxytocin), 70 (Erot alkaloids) 2
0.001 0.01 0.1 1 10 100 1000

Favors oxytocin Favors ergots
Figure 3 Comparison of oxytocin vs. ergot alkaloids (all trials), with outcome of manual removal of the
placenta. Cochrane review14
100
122
30 August 2006 14:20:15
Study Oxytocin Ergot alkaloids Relative risk (fixed) Weight Relative risk (fixed)
n/N n/N 95% Cl (%) 95% Cl
McGinty 1958 4/50 15/100 100.0 0.53 [ 0.19, 1.52 ]
Total (95% Cl) 50 100 100.0 0.53 [ 0.19, 1.52 ]

Total events: 4 (Oxytocin), 15 (Ergot alkaloids )
Test for heterogeneity: not applicable
0.1 0.2 0.5 1 2 5 10
Favors oxytocin Favors ergot
Figure 4 Comparison of oxytocin vs. ergot alkaloids (all trials), with outcome of diastolic blood pressure
> 100 mmHg between delivery of the baby and discharge from the labor ward. Cochrane review14
pressure. Ergometrine produces arterial vaso- diastolic blood pressure both with ergometrine–
constriction by stimulation of the α-adrenergic oxytocin and oxytocin. However, the use of
and serotonin receptors and inhibition of ergometrine–oxytocin was associated with a
endothelial-derived relaxation factor release. greater rise in blood pressure than when using
Uterine contractions are initiated within 1 min oxytocin alone (OR 2.40, 95% CI 1.58–3.64).
of intravenous injection and last for up to The incidence of nausea and/or vomiting was
45 min, whilst, with the intramuscular injec- addressed in four trials32–35. In these trials,
tion, contractions are initiated within 2–3 min a greater incidence of these side-effects was
and last for 3 h or longer28–30. noted with ergometrine–oxytocin use compared
The prophylactic use of ergometrine– to oxytocin alone (vomiting: OR 4.92, 95%
oxytocin in the third stage of labor has CI 4.03–6.00; nausea: OR 4.07, 95% CI
also been the subject of a Cochrane review, 3.43–4.84; vomiting and nausea: OR 5.71, 95%
where ergometrine-oxytocin was compared to CI 4.97–6.57). The same trials studied the inci-
oxytocin31. dence of need for blood transfusion and found
no difference (OR 1.37, 95% CI 0.89–2.10).
In the two trials that addressed the issue of
Ergometrine–oxytocin vs. oxytocin
manual removal of the placenta, no significant
Six trials including 9332 women were described difference was shown (OR 1.03, 95% CI
in this comparison. Variation was noted not 0.80–1.33)33,36.
only in sample size but also in outcomes mea-
sured. Maternal outcomes in terms of nausea
Prophylactic use of ergot alkaloids in the
and vomiting, the need for blood transfusion
third stage of labor
and blood pressure measurements were consid-
ered in four trials32–35. Manual removal of the Ergot alkaloids are amide derivatives of the
placenta was considered in two trials33,36. All six tetracyclic compound lysergic acid. There are
trials addressed the issue of postpartum hemor- three categories: (1) the ergotamine group:
rhage, but much variation was seen in the ergotamine, ergosine and isomers; (2) the rego-
quantification of the amount of blood lost32–37. toxine group: ergocornine, ergocristine, ergo-
In terms of postpartum hemorrhage, all six kryptine and isomers; and (3) the ergotamine
trials32–37 demonstrated a significantly lower and isomers.
rate of postpartum hemorrhage with The ergot alkaloids act as partial agonists or
ergometrine–oxytocin regardless of the dose of antagonists at adrenergic, dopaminergic and
oxytocin used (odds ratio (OR) 0.82, 95% CI tryptaminergic receptors. All the ergot alkaloids
0.71–0.95). Four trials examined the effects significantly increase the motor activity of the
of uterotonics on diastolic blood pressure32–35. uterus. They produce persistent contractions in
Whilst there was a marked difference in the the inner zone of myometrium through calcium
criteria used to ascertain the changes in diastolic channel mechanism and actin–myosin interac-
blood pressure, a consistent picture, neverthe- tion, leading to the shearing effect on placental
less, emerges demonstrating an elevation of separation. The gravid uterus is very sensitive to
101
123
30 August 2006 14:20:16
ergot alkaloids, and small doses can be adminis- Moreover, the area under the misoprostol con-
tered immediately postpartum to obtain a marked centration vs. time curve is increased, implying
uterine response. The different preparations and greater exposure time45.
routes of administration have been the subject of The prophylactic use of prostaglandins in
a number of studies, both for therapeutic and the management of the third stage of labor has
prophylactic use15,38–41. All ergot alkaloids have been the subject of a Cochrane review wherein
qualitatively the same effect on the uterus; ergo- misoprostol was compared46 to (1) either
metrine is the most active and is also less toxic placebo or no uterotonic; (2) conventional
than ergotamine. For this reason, ergometrine injectable uterotonic; or (3) injectable prosta-
and its semi-synthetic derivative methylergo- glandin vs. injectable uterotonic.
metrine have replaced other ergot preparations
as uterine-stimulating agents in obstetrics. The
Misoprostol vs. placebo/no uterotonic
injectable forms of both preparations are
unstable when stored unrefrigerated and at high Six trials were included in this comparison.
temperatures. The oral forms similarly deterio- Misoprostol 400 µg was the dose used in three
rate within weeks when stored in increased of the trials47–49. A dose of 600 µg was used
temperatures. Methylergometrine differs little in an additional three trials50–52. One trial
from ergometrine in its pharmacokinetics. compared doses of 600 µg, and 400 µg with
Clinical trials have been conducted on placebo/no uterotonic53.
the use of ergot alkaloids in the third stage At both doses, misoprostol was either equal
of labor for prevention of postpartum hemor- or less effective than placebo/no treatment for
rhage15,23,38. The use of ergot alkaloids in the blood loss of 1000 ml or more and appeared to
third stage of labor compared with no utero- have a protective effect on the use of additional
tonic drugs and with different routes of admin- uterotonics, although this did not reach statisti-
istration is again the subject of a Cochrane cal significance. Misoprostol was, however,
review42. associated with more vomiting, shivering, and
pyrexia than placebo, and this was dose-related
and occurred across the trials.
Prostaglandins
Rectal misoprostol was compared to
Prostaglandins ripen the cervix by altering the placebo in one trial49. No statistically significant
extracellular ground substance, by increasing reduction in blood loss of at least 1000 ml (RR
the activity of collagenase, and by increasing the 0.69, 95% CI 0.35–1.37) or need to use addi-
elastase, glycoaminoglycans, dermatan sulfate, tional uterotonic agents (RR 0.70, 95% CI
and hyaluronic acid levels in the cervix43,44. 0.31–1.62) was observed.
They allow for cervical smooth muscle relax-
ation and increase intracellular calcium, thus
Misoprostol vs. conventional injectable uterotonics
facilitating contraction of the myometrium.
Misoprostol is a synthetic analogue of natu- Fourteen trials were included in this compari-
rally occurring prostaglandin E1. It is rapidly son51,54–69. The trials are heterogeneous in
absorbed following oral administration and its terms of dose of misoprostol used, route of
bioavailability exceeds 80%. Peak plasma levels administration and injectable uterotonic used.
are reached in 30–60 min, and it is converted to Overall, the risk of postpartum hemorrhage of at
its active misoprostol acid, which has a half-life least 1000 ml was higher for the misoprostol
of 30–60 min. It is metabolized in the liver, and group (RR 1.34, 95% CI 1.16–1.55) compared
less than 1% of the active metabolite is excreted to either intravenous or intramuscular injections
in the urine. In pregnancy, it is absorbed across of oxytocin70.
the vaginal mucosa. After oral administration,
the plasma concentration increases rapidly to
Injectable prostaglandins vs. injectable uterotonics
reach a peak in 30 min and rapidly declines,
whereas with vaginal administration the peak is Seven trials compared injectable prosta-
reached in 1.5 h before steadily declining. glandins with conventional injectable
102
124
30 August 2006 14:20:16
uterotonics17,41,71–75. The trials were heteroge- risk of respiratory distress syndrome in pre-term
neous, and reliable estimates of outcomes were infants81. At present, evidence is insufficient to
not possible. The injectable prostaglandins were recommend early or late cord clamping, and the
associated with less blood loss, a shorter dura- issue is the subject of a Cochrane review82.
tion of the third stage of labor, more vomiting,
diarrhea and abdominal pain than conventional
COMPARISON OF ACTIVE VERSUS
uterotonics. [Editor’s note: Interested readers
EXPECTANT MANAGEMENT
should see also Chapter 12 and Section IV, with
the tables in Chapter 19.] As noted above, the active management of the
third stage of labor consists of three interlocking
interventions: a prophylactic uterotonic agent,
EARLY CORD CLAMPING AND
early clamping and division of the umbilical
DIVISION
cord, and controlled cord traction.
The timing of umbilical cord clamping is vari- This management package has been com-
able76. In the active management of the third pared to expectant management of the third
stage of labor, early cord clamping is generally stage of labor in a Cochrane review83. Five trials
carried out in the first 30 s after birth, regardless were included in the analysis84–88. Active man-
of the presence or absence of cord pulsations77. agement was routinely practiced in the first four
Late cord clamping constitutes expectant man- of these trials, and both active and expectant
agement, whereby clamping is deferred until management were practiced in the fifth trial.
cord pulsations have ceased. A precise defini- The oxytocics used included oxytocin alone,
tion of early or late cord clamping is not ergometrine alone and a combination of
currently available78. oxytocin and ergometrine.
Delayed clamping of the cord facilitates The incidence of postpartum hemorrhage
placental transfusion. This results in an increase both at the 500 ml (RR 0.38, 95% CI
in infant blood volume by 30% and an increase 0.32–0.46) and 1000 ml (RR 0.33, 95% CI
in hematocrit and hemoglobin levels, with a 0.21–0.51) levels was significantly decreased in
resultant increase in iron stores and less anemia the actively managed group compared to the
in infancy78–80. However, the benefits associated expectantly managed group (Figures 5 and 6).
with this increase in infant blood volume are More importantly, the need for blood transfu-
short-lived, lasting no longer than 3 months79. sion was also significantly less in the actively
In Rhesus-negative women, early clamping of managed group (RR 0.34, 95% CI 0.22–0.53),
the cord may increase the likelihood of feto- and the duration of the third stage of labor
maternal transfusion and so exacerbate the risk was not unexpectedly of shorter duration in
of isoimmunization78. Early clamping of the the actively managed group (RR 0.15, 95% CI
cord has also been associated with a higher 0.12–0.19). A tendency toward an increase in
Study Treatment Control Relative risk (fixed) Weight Relative risk (fixed)
n/N n/N 95% Cl (%) 95% Cl
Abu Dhabi 1997 48/827 90/821 21.2 0.53 [ 0.38, 0.74 ]

Bristol 1988 50/846 152/849 35.6 0.33 [ 0.24, 0.45 ]
Dublin 1990 14/705 60/724 13.9 0.24 [ 0.14, 0.42 ]
Hinchingbrooke 1998 51/748 126/764 29.3 0.41 [ 0.30, 0.56 ]
Total (95% Cl) 3126 3158 100.0 0.38 [ 0.32, 0.46 ]

Total events: 163 (Treatment), 428 (Control) 2
0.1 0.2 0.5 1 2 5 10
Figure 5 Comparison of active vs. expectant management (all women), with outcome of postpartum
hemorrhage (clinically estimated blood loss ≥ 500 ml)83
103
125
30 August 2006 14:20:17
Study Treatment Control Relative risk (fixed) Weight Relative risk (fixed)
n/N n/N 95% Cl (%) 95% Cl
Abu Dhabi 1997 6/827 26/821 31.6 0.23 [ 0.09, 0.55 ]

Bristol 1988 7/846 26/849 31.4 0.27 [ 0.12, 0.62 ]
Dublin 1990 1/705 11/724 13.1 0.09 [ 0.01, 0.72 ]
Hinchingbrooke 1998 13/748 20/764 23.9 0.66 [ 0.33, 1.32 ]
Total (95% Cl) 3126 3158 100.0 0.33 [ 0.21, 0.51 ]

Total events: 27 (Treatment), 83 (Control) 2
0.1 0.2 0.5 1 2 5 10
Figure 6 Comparison of active vs. expectant management (all women), with outcome of severe
postpartum hemorrhage (clinically estimated blood loss ≥1000 ml)83
the need for manual removal of the placenta was research and audit. It is reproduced in full as an
noted in the actively managed group (RR 1.21, Appendix to this chapter.
95% CI 0.82–1.78), but this did not reach sta-
tistical significance. The incidences of nausea
and vomiting were increased in the actively References
managed group (RR 1.83, 95% CI 1.51–2.23
and RR 2.19, 95% CI 1.68–2.86, respectively). 1. Sweet D. Mayes Midwifery, 12th edn. London:
However, this was only noted where WB Saunders Co, 1997
2. Stables D. Physiology in Childbearing with Anat-
ergometrine was used as the oxytocic.
omy and Related Biosciences. London: Balliere
Based on the data presented above, the
Tindall, 1999
authors conclude that active management is 3. Prendiville WJ, Harding JE, Elbourne DR,
superior to expectant management in terms of Stirrat GM. The Bristol third stage trial: active vs
blood loss and other serious complications of physiological management of the third stage of
the third stage of labor, and that active manage- labour. Br Med J 1988;297:1295–1300
ment should be routine for women expecting a 4. Den Hertog CE, DeGroot AN, VanDongen PW.
vaginal delivery in a maternity hospital. History and use of oxytocics. Eur J Obstet
[Editor’s note: At the International Confer- Gynaecol Reprod Med 2001;94:8–12
ence on the Prevention of Post Partum Hemor- 5. McCormick ML, Sanghvi HC, Kinzie B,
rhage held in Goa on July 12–15, 2006, there McIntosh N. Preventing postpartum haemor-
rhage in low-resource settings. Int J Gynaecol
was considerable discussion on the appropriate-
Obstet 2002;77:267–75
ness of this intervention to be performed in
6. World Health Organisation. Pregnancy, Child-
the hands of skilled birth attendants who were birth, Postpartum and Newborn Care: a Guide
working in a domiciliary delivery, although it for Essential Practice. Geneva: World Health
was recognized that all such individuals would Organisation, 2003
not have access to an injectable uterotonic for 7. Winter C, Macfarlane A, Deneux C, et al.
logistic reasons.] Policies for management of the third stage
The European 5th Framework has funded an of labour and the immediate management of
expert group from 14 European Union (EU) postpartum haemorrhage in Europe: what is the
countries to address postpartum hemorrhage role of evidence? 2006 In press
in the EU. The group reviewed the literature, 8. Alexandrova M, Soloff MA. Oxytocin receptors
and parturition. I. Control of oxytocin receptor
surveyed participants with respect to current
concentration in the rat myometrium at term.
protocols and devised a consensus document89.
Endocrinology 1980;106:730–5
This group also clarified the definition of active 9. Fuchs AR, Fuchs F, Hurstein P, Soloff MS,
management of the third stage of labor. The Fernstrom MJ. Oxytocin receptors and human
consensus document has received wide support parturition: a dual role for oxytocin in the
from a large number of international authorities initiation of labor. Science (New York) 1982;215:
and forms the basis for future comparative 1396–8
104
126
30 August 2006 14:20:19
10. Sanborn BM, Dodge K, Monga M, Qian A, 23. Sorbe B. Active pharmacological management
Wang W, Yue C. Molecular mechanisms regu- of the third stage of labor. A comparison of
lating the effects of oxytocin on myometrial oxytocin and ergometrine. Obstet Gynecol 1978;
intercellular calcium. Adv Exp Med Biol 1998; 52:694–7
449:277–86 24. McGinty LB. A study of the vasopressor effects
11. Parker SL, Schimmer BP. Pituitary hormones of oxytocics when used intravenously in the third
and their hypothalamic releasing hormones. In stage of labour. Western J Surg 1956;64:22–8
Goodman and Gilman, eds. The Pharmacological 25. Chong YS, Su LL, Arulkumaran S. Current
Basis of Therapeutics, 11th edn. New York: strategies for the prevention of postpartum
McGraw Hill, 2006:1489–510 haemorrhage in the third stage of labour. Curr
12. Amico JA, Seitchik J, Robinson AG. Studies of Opin Obstet Gynecol 2004;16:143–50
oxytocin in plasma of women during hypocon- 26. Hunter DJ, Schulz P, Wassenaar W. Effects of
tractile labor. J Clin Endocrinol Metab 1984;58: carbetocin, a long acting oxytocin analog on the
274–9 postpartum uterus. Clin Pharm Therapeu 1992;
13. De Groot AN, Vree TB, Hekster YA, et al. Bio- 52:60–7
availability and pharmacokinetics of sublingual 27. Su LL, Chong YS, Chan ESY, Samuel M.
oxytocin in male volunteers. J Pharm Pharmacol Oxytocin agonists for preventing postpartum
1995;47:571–5 haemorrhage (Protocol). Cochrane Database of
14. Elbourne DR, Prendiville WJ, Carroli G, Wood Systematic Reviews 2005, Issue 3. Art No.:
J, McDonald S. Prophylactic use of oxytocin in CD005457. DOI:10.1002/14651858. CD005457
the third stage of labour. Cochrane Database 28. Rall TW. Oxytocin, prostaglandins, ergot alka-
of Systematic Reviews 2001, Issue 4. Art. loids, and other drugs; tocolytic agents. In Good-
No.: CD001808. DOI: 10.1002/14651858. man, Gilman A, Rall TW, Nies AS, Taylor P,
CD001808 eds. Goodman and Gilman’s The Pharamacological
15. De Groot ANJA, Van Roosmalen J, Van Dongen Basis of Therapeutics. Toronto: Pergamon Press,
PWJ, Borm GF. A placebo-controlled trial of 1990:933–53
oral ergometrine to reduce postpartum haemor- 29. Berde E, Stürmer E. Introduction to the
rhage. Acta Obstet Gynecol Scand 1996;75:464–8 pharmacology of ergot alkaloids and related
16. Newton M, Mosey LM, Egli GE, Gifford WB, compounds as a basis to their therapeutic
Hull CT. Blood loss during and immediately application. In Berde B, Schild HO, eds. Ergot
after delivery. Obstet Gynecol 1961;17:9–18 Alkaloids and Related Compounds. New York:
17. Poeschmann RP, Doesburg WH, Eskes TKAB. Springer Verlag, 1978:1–28
A randomised comparison of oxytocin, sulpros- 30. Müller-Schweinitzer E, Weidmann H. Basic
tone and placebo in the management of the pharmacological propertues. In Berde B, Schild
third stage of labour. Br J Obstet Gynaecol HO, eds. Ergot Alkaloids and Related Compounds.
1991;98:528–30 New York: Springer Verlag, 1978:87–232
18. Howard WF, McFadden PR, Keetek WC. 31. McDonald S, Abbott JM, Higgins SP. Prophy-
Oxytocic drugs in the fourth stage of labor. lactic ergometrine–oxytocin versus oxytocin for
JAMA 1964;189:411–13 the third stage of labour. Cochrane Database
19. Ilancheran A, Ratnam SS. Effect of oxytocin on of Systematic Reviews 2004, Issue1. Art.
prostaglandin levels in the third stage of labour. No.: CD000201. DOI: 10.1002/14651858.
Gynecol Obstet Invest 1990;29:177–80 CD000201.pub2
20. Nordstrom L, Fogelstam K, Friedman G, 32. Choy CMY, Lau WC, Tam WH, Yuen PM. A
Larsson A, Rydhstroem H. Routine oxytocin in randomised controlled trial of intramuscular
the third stage of labour: a placebo controlled syntometrine and intravenous oxytocin in the
randomised trial. Br J Obstet Gynaecol 1997;104: management of the third stage of labour. Br J
781–6 Obstet Gynaecol 2002;109:173–7
21. Pierre F, Mesnard L, Body G. For a systematic 33. Khan GQ, John LS, Chan T, Wani S, Hughes
policy of iv oxytocin where a fairly active man- AO, Stirrat GM. Abu Dhabi third stage trial:
agement of third stage of labour is yet applied: Oxytocin versus syntometrine in the active man-
results of a controlled trial. Eur J Obstet Gynaecol agement of the third stage of labour. Eur J Obstet
Reprod Med 1992;43:131–5 Gynaecol Reprod Med 1995;58:147–51
22. Fugo NW, Dieckmann WJ. A comparison of 34. McDonald SJ, Prendiville W, Blair E. Random-
oxytocic drugs in the management of the placen- ised controlled trial of oxytocin alone versus
tal stage. Am J Obstet Gynecol 1958;76:141–6 oxytocin and ergometrine in the active
105
127
30 August 2006 14:20:19
management of the third stage of labour. Br Med oral misoprostol in the third stage of labour. Br J
J 1993;307:1167–71 Obstet Gynaecol 1998;105:971–5
35. Yuen PM, Chan NST, Yim SF, Chang AMZ. 48. Hofmeyr GJ, Nikodem VC, deJager M, Drakely
A randomised double blind comparison of A, Gelbart B. Oral misoprostol for labour third
syntometrine and syntocinon in the management stage management: randomised assessment of
of the third stage of labour. Br J Obstet Gynaecol side effects (part 2). Proceedings of the 17th
1995;102:377–80 Conference on Priorities in Perinatal care; 1998,
36. Nieminen U, Jarvinen PA. A comparative study South Africa, 1998:53–4
of different medical treatments of the third stage 49. Bamigboye AA, Hofmeyr GJ, Merrell DA.
of labour. Ann Chirurig Gynaecol Fenniae 1963; Rectal misoprostol in the prevention of post-
53:424–9 partum hemorrhage: a placebo controlled trial.
37. Mitchell GG, Elbourne DR. The Salford third Am J Obstet Gynecol 1998;179:1043–6
stage trial: oxytocin plus ergometrine versus 50. Surbek DV, Fehr P, Hoesli I, Holzgreve W.
oxytocin alone in the active management of the Oral misoprostol for third stage of labor: a
third stage of labour. Online Journal of Current randomized placebo-controlled trial. Obstet
Clinical Trials 1993;2:Doc 83 Gynecol 1999;94:255–8
38. Andersen B, Andersen LL, Sorensen T. 51. Benchimol M, Gondry J, Mention J, Gagneur O,
Methylergometrine during the early puerperium; Boulanger J. Role of misoprostol in controlled
a prospective randomized double blind study. delivery [Place du misoprostol dans la direction
Acta Obstet Gynecol Scand 1998;77:54–7 de la deliverance]. J Gynaecol Obstet Biol Reprod
39. Borri P, Gerli P, Antignani FL, et al. Methyler- 2001;30:576–83
gonovine maleate: a proposal for its more specific 52. Hofmeyr GJ, Nikodem VC, deJager M, Drakely
use. Biol Res Preg Perinatol 1986;7:128–30 A. Side effects of oral misoprostol in the third
40. Moir DD, Amoa AB. Ergometrine or oxytocin? stage of labour: a randomised placebo controlled
Blood loss and side effects at spontaneous vertex trial. South Afr Med J 2001;91:432–5
delivery. Br J Anaes 1979;51:113–17 53. Hofmeyr GJ, Nikodem VC, de Jager M, Gelbart
41. Van Selm M, Kanhai HH, Keirse MJ. Preventing BR. A randomized placebo-controlled trial of
the recurrence of atonic postpartum hemorrhage: oral misoprostol in the third stage of labour. Br J
a double blind trial. Acta Obstet Gynecol Scand Obstet Gynaecol 1998;105:971–5
1995;74:270–4 54. Caliskan E, Dilbaz B, Meydanli M, Ozturk N,
42. Liabsuetrakul T, Choobun T, Islam M, Narin MA, Haberal P. Oral misoprostol for the
Peeyananjarassri K. Prophylactic use of ergot third stage of labor: a randomized controlled
alkaloids in the third stage of labour (Protocol). trial. Obstet Gynecol 2003;101:921–8
Cochrane Database of Systematic Reviews 2005, 55. Cook C, Spurrett B, Murray H. A randomized
Issue 3. Art. No.: CD005456.DOI: 19.1002/ clinical trial comparing oral misoprostol with
14651858.CD005456 synthetic oxytocin or syntometrine in the third
43. Uldbjerg N, Ekman G, Malmstrom A, Sporrong stage of labour. Aust NZ J Obstet Gynaecol 1999;
B, Ulmstein U, Wingerup L. Biochemical 39:414–19
and morphological changes of human cervix 56. Amant F, Spitz B, Timmerman D, Corremans
after local application of prostaglandin E2 in A, Van Assche FA. Misoprostol compared with
pregnancy. Lancet 1981;1:267–8 methylergometrine for the prevention of postpar-
44. Uldbjerg N, Ekman G, Malmstrom A, Olsson K, tum haemorrhage: a double-blind randomised
Ulmstein U. Ripening of the human uterine trial. Br J Obstet Gynaecol 1999;106:1066–70
cervix related to changes in collagen, glyco- 57. Lumbiganon P, Hofmeyr J, Gulmezoglu AM,
saminoglycans, and collagenolytic activity. Am J Villar J. Misoprostol dose related shivering and
Obstet Gynecol 1983;147:662–6 pyrexia in the third stage of labour. Br J Obstet
45. More B. Misoprostol: an old drug, new Gynaecol 1999;106:304–8
indications. J Postgrad Med 2002;48:336–9 58. Whalley RL, Wilson JB, Crane JM, Matthews K,
46. Gulmezoglu AM, Forna F, Villar J, Hofmeyr GJ. Sawyer E, Hutchens D. A double-blind placebo
Prostaglandins for prevention of postpartum controlled randomised trial of misoprostol and
haemorrhage. Cochrane Database of Systematic oxytocin in the management of the third stage of
Reviews2004, Issue 1. Art No.: CD000494. labour. Br J Obstet Gynaecol 2000;107:1111–15
DOI: 10.1002/14651858.CD000494.pub2 59. El-Refaey H, Nooh R, O’Brien P, Abdalla M,
47. Hofmeyr GJ, Nikodem VC, deJager M, Gelbart Geary M, Walder J, Rodeck C. The misoprostol
BR. A randomised placebo controlled trial of third stage of labour study: a randomised
106
128
30 August 2006 14:20:19
controlled comparison between orally adminis- 72. Bhattacharya P, Devi PK, Jain S, Kanthamani
tered misoprostol and standard treatment. Br J CR, Raghavan KS. Prophylactic use of 15(S) 15
Obstet Gynaecol 2000;107:1104–10 methyl PGF2 alpha by intramuscular route for
60. Ng PS, Chan ASM, Sin WK, Tang LCH, control of postpartum bleeding – a comparative
Cheung KB, Yuen PM. A multicentre random- trial with methylergometrine. Acta Obstet Gynecol
ized trial of oral misoprostol and i.m synto- Scand 1998;Suppl 145:13–15
metrine in the management of the third stage of 73. Chua S, Chew SL, Yeoh CL, et al. A randomized
labour. Hum Reprod 2001;16:31–5 controlled study of prostaglandin 15-methyl F2
61. Bugalho A, Daniel A, Faundes A, Cunha M. alpha compared with syntometrine for prophy-
Misoprostol for prevention of postpartum lactic use in the third stage of labour. Aust NZ J
haemorrhage. Int J Gynaecol Obstet 2001;73:1–6 Obstet Gynaecol 1995;35:413–16
62. Lokugamage A, Paine M, Bassaw-Balroop K, 74. Catanzarite VA. Prophylactic intramyometrial
et al. Active management of the third stage at carboprost tromethamine does not substantially
Cesarean section: a randomized controlled trial reduce blood loss relative to intramyometrial
of misoprostol versus syntocinon. Aust N Z oxytocin at routine caesarean section. Am J
Obstet Gynaecol 2001;41:411–14 Perinatol 1990;7:39–42
63. Gerstenfeld TS, WingDA. Rectal misoprostol 75. Chou MM, MacKenzie IZ. A prospective,
versus intravenous oxytocin for the prevention of double blind, randomized comparison of pro-
postpartum hemorrhage after vaginal delivery. phylactic intramyometrial 15-methyl prostaglan-
Am J Obstet Gynecol 2001;185:878–82 din F2 alpha, 125 micrograms, and intravenous
64. Gulmezoglu AM, Villar J, Ngoc NT, et al. The oxytocin, 20 units, for the control of blood loss at
WHO multicentre double-blind randomized trial elective caesarean section. Am J Obstet Gynecol
to evaluate the use of misoprostol in the manage- 1994;171:1356–60
ment of the third stage of labour. Lancet 2001; 76. Inch S. Management of the third stage of labour:
358:689–95 another cascade of intervention? Midwifery 1991;
65. Kundodyiwa TW, Majoko F, Rusakaniko S. 7:64–70
Misoprostol versus oxytocin in the third stage of 77. McDonald SJ. Management in the Third Stage
labor. Int J Obstet Gynaecol 2001;75:235–41 of Labour. Western Australia: University of
66. Karkanis SG, Caloia D, Salenieks ME, et al. Western Australia, 1996
Randomized controlled trial of rectal misoprostol 78. Prendiville WJ, Elbourne D. Care during the
versus oxytocin in third stage management. J third stage of labour. In Chalmers I, Enkin M,
Obstet Gynecol Can 2002;24:149–54 Keirse MJNC, eds. Effective Care in Pregnancy
67. Penaranda W, Arrieta O, Yances B. Active man- and Childbirth. Oxford: Oxford University Press,
agement of the childbirth with sublingual miso- 1989:1145–69
prostol: a clinical controlled trial in the Hospital 79. World Health Organisation. Care of the umbilical
de Maternidad Rafeal Calvo. Revista Colombiana cord: a review of the evidence. Geneva: World
de Obstetricia y Ginecologia 2002;53:87–92 Health Organisation, 1998
68. Caliskan E, Meydanli M, Dilbaz B, Aykan B, 80. Mercer JS. Current best evidence: a review
Sonmezer M, Haberal A. Is rectal misoprostol of the literature on umbilical cord clamping.
really effective in the treatment of third stage of J Midwifery Women’s Health 2001;46:402–14
labor? A randomized controlled trial. Am J Obstet 81. Rabe H, Reynolds G, Diaz-Rossello J. Early
Gynecol 2002;187:1038–45 versus delayed umbilical cord clamping in
69. Caliskan E, Dilbaz B, Meydanli M, Ozturk N, preterm infants. Cochrane Database of Systematic
Narin M, Haberal A. Oral misoprostol for the Reviews 2004, Issue 3. Art. No.: Cd003248.
third stage of labor: a randomized controlled DOI:10.1002/14651858/CD003248.pub2
trial. Obstet Gynecol 2003;101:921–8 82. McDonald SJ, Abbott JM. Effect of timing of
70. Gulmezoglu AM, Villar J, Ngov NT, et al. WHO umbilical cord clamping of term infants on
multicentre randomized controlled trial of miso- maternal and neonatal outcomes (Protocol).
prostol in the management of the third stage of Cochrane Database of Systematic Reviews 2003,
labour. Lancet 2001;358:689–95 Issue 1.Art. No.: CD004074.DOI:10.1002/
71. Abdel-Aleem H, Abol-Oyoun EM, Moustafa SAM, 14651858.CD004074
Kamel HS, Abdel-Wahab HA. Carboprost tro- 83. Prendiville WJ, Elbourne D, McDonald S.
metamol in the management of the third stage of Active versus expectant management in the third
labor. Int J Obstet Gynaecol 1993;42: 247–50 stage of labour. Cochrane Database of Systematic
107
129
30 August 2006 14:20:20
Reviews 2000, Issue 3. Art. No.:CD000007. physiological management of third stage of

DOI: 10.1002/14651858. CD000007 labour. BMJ 1988;297:1295–300
84. Khan GQ, John LS, Wani S, Doherty T, Sibai 87. Begley CM. A comparison of active and physio-
BM. Controlled cord traction versus minimal logical management of the third stage of labour.
intervention techniques in delivery of the Midwifery 1990;6:3–17
placenta: a randomized controlled trial. Am J 88. Rogers J, Wood J, McCandlish R, Ayers S,
Obstet Gynecol 1997;177:770–4 Truesdale A, Elbourne D. Active vs expectant
85. Thilaganathan B, Cutner A, Latimer J, Beard R. management of the third stage of labour: the
Management of the third stage of labour in Hitchingbrooke randomised controlled trial.
women at low risk of postpartum haemor- Lancet 1998;351:693–9
rhage. Eur J Obstet Gynaecol Reprod Biol 1993; 89. Euphrates group. European consensus on
48:19–22 prevention and management of postpartum
86. Prendiville WJ, Harding JE, Elbourne D, Stirrat haemorrhage. 2006, in press
GM. The Bristol Third Stage Trial: active vs.
108
130
30 August 2006 14:20:20
APPENDIX: EUROPEAN CONSENSUS ON PREVENTION AND

MANAGEMENT OF POSTPARTUM HEMORRHAGE
The EUPHRATES group(EUropean Project on obstetric Haemorrhage Reduction:
Attitudes, Trial, and Early warning System), European Union 5th Framework
INTRODUCTION Action is often taken following maternal

signs (e.g. hypotension, malaise) rather than on
The EUPHRATES study comprises five parts,
estimated blood loss.
the second of these being ‘the development
Blood loss at Cesarean section is generally
of a minimal European core consensus on
greater than at vaginal delivery.
prevention and management of post partum
Despite these three caveats, our group
hemorrhage’. This consensus is not a
endorses the following classical definitions:
protocol or guideline. It represents a European
consensus on what could be agreed on by all. ● ≥ 500 ml = postpartum hemorrhage
Each maternity unit should have its own written
● > 1000 ml = severe postpartum hemorrhage
protocol concerning prevention and treatment
of postpartum hemorrhage (PPH). ● ≤ 24 h = primary, or early, postpartum hem-
orrhage
● > 24 h = secondary, or late, postpartum
Method hemorrhage
This consensus is based on three pillars: (a) In regions and in groups where anemia of
review of literature, (b) survey of present proto- pregnancy is prevalent, the recognition of lesser
cols and practice, (c) consensus by experts gath- amounts is clinically important.
ered in a special board (see list of members at
the end of this Appendix).
The following principle was followed. Where 1(b) Communication
solid evidence was available (level of evidence Substandard care is often related to lack of
= 1), a consensus process was not necessary. communication within the team and between
Consensus was necessary in two circumstances: the team and other professionals. Managing
disagreement as to the clinical relevance of an difficult cases as a team may make the difference
outcome measure clearly shown to be affected between life and death. Identified communica-
by an intervention (e.g. active management of tion problems include the following:
third stage) and situations where action has to
be taken but no high-level evidence is available ● Failure by the first-line care providers to call
(e.g. medications in presence of continuing senior colleagues in time
postpartum hemorrhage). ● Reluctance of senior colleagues to come,
when informed of problem
● Failure by the obstetrical team to inform on
STATEMENTS time other specialists, e.g. intensive care,
anesthesiology, hematology.
1. General considerations
● In theater, failure of anesthesists and
1(a) Definition of postpartum hemorrhage in terms
obstetricians to keep each other informed
on milliliters lost
of relevant events, such as rapid blood
Evaluation of blood loss is unreliable. loss, tachycardia, blood pressure support
109
131
30 August 2006 14:20:20
interventions (fluid replacement and/or ● Caregivers should be trained to be proficient

vasopressor use), etc. in active third-stage management, and to
offer it to all women.
● Failure to obtain blood, because of lack of
perception by the laboratory/blood transfu- ● It is acknowledged, however, that, provided
sion staff of the severity of the case the woman and caregiver are fully informed,
a decision not to use active management in
some individual cases and/or settings should
1(c) Implementing local policies to ensure rapid not be considered substandard care.
availability of blood products at all times
It is mandatory that appropriate blood products
be available easily and rapidly in units where 2(b) Type, dosage, route, speed and timing of
women deliver. Different European countries administration of prophylactic uterotonic drugs
achieve this through different systems and There is a lack of randomized trials addressing
there is no evidence that one system should the questions of dosage, route and timing of
prevail. prophylactic uterotonic drug administration,
There should be a written document, detail- because most trials have compared the full
ing how this is to be implemented and including package made up of three interventions to no
practical information such as transfusion intervention.
department phone number, etc. This document
should be widely disseminated.
(i) Type of drug
1(d) Audits and enquiries ● Oxytocin is the most frequently used drug for
active management in Europe.
The impact of existing guidelines/consensus
statements on severe maternal hemorrhage ● In the United Kingdom and Ireland, Synto-
should be monitored by audit and/or confiden- metrine is widely used. This is a combination
tial enquiries. of oxytocin and ergometrine. Syntometrine is
more effective but is associated with more
side-effects than oxytocin3. Syntometrine is not
2. Prevention of postpartum hemorrhage suitable for all women, e.g. in hypertension.
at vaginal birth
● Ergometrine has been reported in the
2(a) Active management of the third stage of labor European survey as additional prophylaxis
(following the administration of oxytocin),
● Active management of the third stage of
after the placenta has been delivered in
labor is usually defined as a three-component
women with risk factors such as multiple
intervention: (1) prophylactic uterotonic, (2)
pregnancy or grand multiparae. This has
early (or less early) clamping of cord, and (3)
never been assessed in a randomized trial.
controlled cord traction. Active management
in the third stage of labor has been proven ● Misoprostol is less effective than injectable
to be effective in reducing blood loss in all uterotonics in reducing postpartum blood
women1. The evidence that active routine loss; however, its superiority over placebo as
management reduces severe maternal part of the active management of the third
adverse effects (morbidity) resulting from stage of labor remains uncertain4.
postpartum hemorrhage is less convincing.
Our group concludes:
The full package of active management is
● Oxytocin is the first drug of choice for all
certainly a valid (and validated) option.
women in the third stage of labor.
● Isolated uterotonics may also be a useful
● Syntometrine may be preferred by some
option2.
clinicians but is contraindicated in hyper-
Our group concludes: tension and pre-eclampsia.
110
132
30 August 2006 14:20:20
● Additional ergometrine (following the administer prophylactic oxytocic therapy ‘on

administration of oxytocin) in selected cases (= just after) delivery of the anterior shoulder’,
is considered acceptable practice. and that is also the timing in use in many
randomized trials. In practice, it is reported
● Misoprostol, although less effective, may be
in our survey that it is usually administered
considered in situations where injectable
after delivery of the baby. Two randomized,
uterotonics are not available.
controlled trials5,6 compared oxytocin given
before and after the placenta had delivered, and
(ii) Dosage found no benefit in providing the uterotonic as
● Oxytocin: most trials have used intramuscu- early as possible. Further research is needed.
lar (IM) or intravenous (IV) administration Our group concludes:
of 5 or 10 IU of oxytocin. The European ● The best time to administer prophylactic
survey shows this dosage to be widely prac- oxytocic therapy is just after birth.
ticed. Particular dosages have been reported
in various settings, e.g. 20 IU in 500 ml IV ● Whether it is administered before or after
bolus5 or lower doses such as 1 IU in 10 min cord pulsation has ceased seems relatively
(‘turning up the drip’). unimportant.
● For Syntometrine, there is only one dosage:

ergometrine 500 µg with oxytocin 5 units (Syn- 2(c) Manual removal of the placenta
tometrine® 1 ml contained in one ampoule). ● Should be performed without delay in
● Misoprostol: most trials have used presence of hemorrhage.
400–600 µg when administered orally, and ● No European consensus could be obtained
400 µg per rectum. as to when this should be performed in the
absence of bleeding. Some would act after
(iii) Route of administration 20 min while others would wait for more
than 1 hour. Evidence is lacking and further
● Oxytocin: If an IV line is in situ, the intrave-
research is needed.
nous route is the route of choice. ‘Turning up
the drip’ delivers low quantities, e.g. 1–2 IU
(1000–2000 mU) in 10 min. If no IV line, 2(d) Other
IM administration is preferable.
Nipple stimulation or early breastfeeding have
● Syntometrine/Ergometrine: Intramuscular been advocated for prevention of postpartum
administration. hemorrhage, as simple and physiological, in
● Misoprostol can be administered orally or particular in low-resource settings. The avail-
intrarectally. able evidence from two randomized controlled
trials7,8 is insufficient to reach a conclusion.
(iv) Speed of administration
3. Prevention of postpartum hemorrhage
A case of maternal death in the 1997–1999 UK
at Cesarean section
Confidential Enquiry was attributed to severe
hypotension following rapid administration of ● For women undergoing delivery by Cesarean
10 IU oxytocin IV. A key recommendation was section, there is an increased risk that blood
made that the administration should be ‘slow’. transfusion may be necessary.
However, no definition of ‘slow’ is available.
● It is reasonable to advise routine administra-
tion of an uterotonic drug immediately after
(v) Timing of administration the baby has been born by Cesarean section.
A recommendation often made, among others ● Accurate blood loss assessment at Cesarean
in the Bristish National Formulary, is to section is difficult. Measuring both vaginal as
111
133
30 August 2006 14:20:20
well as abdominal blood loss may increase Whether an anesthetist is available immediately
accuracy. and whether the woman has got an effective
epidural will determine the order in which the
● For Cesarean sections that are considered to
above and the following occur.
be at greater risk of hemorrhage (e.g. pla-
centa previa, especially in the presence of ● Keep communication open with the anesthe-
uterine scar), it is recommended that a senior tist and the rest of the team.
obstetrician be present.
(iv) If bleeding still not controlled
4. Management of postpartum ● Circulatory support as necessary with
hemorrhage colloids and/or blood products, and
4(a) Postpartum hemorrhage after vaginal delivery vasopressors if needed
We divided the event into three stages: ● Ensure appropriate oxygenation

(i) concern about possible excessive bleeding, ● Monitor for coagulation abnormalities
(ii) early management of hemorrhage, and
(iii) continuing hemorrhage. ● Uterine packing or intrauterine balloon
● Uterine artery embolization
(i) Concern about possible excessive bleeding
4(b) Hemorrhage at Cesarean section
● If relevant, remove placenta
(i) Immediate management
● Empty bladder, massage uterus until it is well
contracted, give additional uterotonics ● Ensure bladder is empty.
● Look for any obvious bleeding in episiotomy ● Explore the uterine cavity and remove the
or tear, and act on findings. placenta and/or clots
● Massage uterus until well contracted, give
(ii) Immediate management in case of hemorrhage additional uterotonics
● Call for help ● Look for and repair trauma, consider
exteriorization of uterus
● Measure blood loss, blood pressure, and pulse
rate, insert large gauge intravenous infusion if ● Measure blood loss
not yet in place and take blood samples
● Check the placenta for completeness (ii) Hemorrhage not controlled
● Continue circulatory support as necessary
(iii) If bleeding continues with colloids and/or blood products and
vasopressors if needed
● Circulatory support as necessary with crystal-
loids, colloids and/or blood products ● Ensure appropriate oxygenation and con-
sider mechanical ventilation when needed
● Ensure appropriate care with sufficient staff
or appropriate referral ● Ensure appropriate care with sufficient staff
● Administer additional uterotonic drugs ● Additional uterotonic drugs (injectable

(injectable prostaglandins) prostaglandins)
● Perform bimanual compression (time aware- ● Appropriate surgery

ness)
4(c) Factor VII
● Explore under anesthesia the genital tract
for retained placenta or part thereof, or Recombinant activated factor VII (Novo-
traumatic damage and act on findings. Seven®) may be a future option in catastrophic
112
134
30 August 2006 14:20:20
hemorrhage, permitting sometimes to avoid Alfirevic (Obstetrician), Peter Brocklehurst

hysterectomy. At present, NovoSeven is very (Obstetrician, Epidemiologist), Alison
expensive and its safety has not yet been ade- MacFarlane (Epidemiologist), Jane Rogers
quately evaluated. Therefore, the use of this (Midwife), Clare Winter (Midwife).
drug should be limited to units with adequate
expertise and resources, and participating in
ongoing registers of use. References
1. Prendiville WJ, Elbourne D, MacDonald S.
Consensus Special Board Active versus expectant management in the third
stage of labour (Cochrane Review). Cochrane
The Special Board was made up of experts from Library, Issue 2, 2004. Chichester, UK: John
14 European countries: Wiley & Sons, Ltd
Austria: Mathias Klein (Obstetrician), Heinz 2. Elbourne DR, Prendiville WJ, Carroli G, Wood J,
MacDonald S. Prophylactic use of oxytocin in the
Leipold (Obstetrician); Belgium: Sophie Alex-
third stage of labour (Cochrane Review).
ander (Obstetrician, Epidemiologist), Paul
Cochrane Library, Issue 2, 2004. Chichester, UK:
Defoort (Obstetrician), Corinne Hubinont John Wiley & Sons, Ltd
(Obstetrician), Wei hong Zhang (Epidemiolo- 3. MacDonald S, Abbott JM, Higgins SP. Prophy-
gist); Denmark: Jens Langhoff-Roos (Obstetri- lactic ergometrine-oxytocin versus oxytocin for
cian), Desiree Rosenborg (Anesthetist); the third stage of labour (Cochrane Review).
Finland: Risto Erkkola (Obstetrician), Vedran Cochrane Library, Issue 2, 2004. Chichester, UK:
Stefanovic (Obstetrician), Jukka Uotila (Obste- John Wiley & Sons, Ltd
trician); France: Marie-Hélène Bouvier-Colle 4. Villar J, Gülmezoglu AM, Hofmeyr J, Forna F.
(Epidemiologist), Gérard Breart (Epidemiolo- Systematic review of randomized controlled trials
gist), Catherine Deneux (Epidemiologist), of misoprostol to prevent postpartum hemor-
rhage. Obstet Gynecol 2002;100:1301–12
Thierry Harvey (Obstetrician), Frédéric
5. Jackson KW Jr, Allbert JR, Schemmer GK,
Mercier (anesthetist); Hungary: Istvan Berbik
Elliot M, Humphrey A, Taylor J. A randomized
(Obstetrician), Jeno Egyed (Obstetrician), Janos controlled trial comparing oxytocin administra-
Herczeg (Obstetrician); Ireland: Mikael tion before and after placental delivery in the pre-
O’Connell (Obstetrician), Walter Prendiville vention of postpartum hemorrhage. Am J Obstet
(Obstetrician); Italy: Anna Maria Marconi Gynecol 2001;185:873–7
(Obstetrician), Graziella Sacchetti (Obstetri- 6. Huh WK, Chelmow D, Malone FD. A double
cian); Nederlands: Kathy Herschderfer blinded, randomized controlled trial of oxytocin
(Midwife), Jos Van Roosmalen (Obstetrician); at the beginning versus the end of the third stage
Norway: Bente Ronnes (Midwife), Babill of labor for prevention of postpartum hemor-
Stray-Pedersen (Obstetrician); Portugal: rhage. Gynecol Obstet Invest 2004;58:72–6
7. Bullough C, Msuku R, Karonde L. Early sucking
Diogo Ayres-de-Campos (Obstetrician), Nuno
and post partum haemorrhage: controlled trial in
Clode (Obstetrician), Teresa Rodrigues (Obste-
deliveries by traditional birth attendants. Lancet
trician); Spain: Enrique Barrau (Obstetrician), 1989;334:522–5
Vicenç Cararach (Obstetrician), Dolores 8. Irons D, Sriskandabalan, Bullough C. A simple
Gomez (Obstetrician); Switzerland: Olivier alternative to parenteral oxytocic for the third
Irion (Obstetrician), Carolyn Troeger stage of labour. Int J Gynaecol Obstet 2004;46:
(Obstetrician); United Kingdom: Zarko 15–18
113
135
30 August 2006 14:20:21
12
MISOPROSTOL: THEORY AND PRACTICE
M. B. Bellad and S. Goudar
INTRODUCTION (13E)-11,16-dihydroxy-16-methyl-9-oxo-prost
-13-enoate), as shown in Figure 13.
Prostaglandins have revolutionalized obstetric
Misoprostol is manufactured as oral tablets
practice. In particular, the advent of miso-
of 200 µg scored and 100 µg unscored. It has
prostol has precipitated an enormous amount of
several advantages – stability in ambient tem-
innovative research as well as controversy. At
perature, long shelf-life and low cost – that have
present, misoprostol is being investigated for its
made it a central focus of research in obstetrics
role in the management of postpartum hemor-
and gynecology for 25 years4. Misoprostol
rhage, induction of labor, cervical ripening and
is rapidly absorbed via the oral route and,
termination of pregnancy. Initially, this drug
although not formulated for parenteral use, can
was approved by the US Food and Drug
also be administered sublingually (buccally),
Administration (FDA) in 1988 for oral adminis-
rectally and vaginally5–7.
tration for the prevention and treatment of
peptic ulcers associated with the use of non-
steroidal anti-inflammatory drugs. Since the Pharmacokinetics, physiology and
early 1990s, however, misoprostol has been teratogenicity profile
viewed with increasing interest by obstetricians
Misoprostol is extensively absorbed and under-
and gynecologists because of its uterotonic and
goes rapid de-esterification to misoprostol acid;
cervical ripening activity. The multiple off-label
this compound is responsible for its clinical
uses for misoprostol underlie its description
activity and, unlike the parent compound, it is
as ‘one of the most important medications in
detectable in plasma. After oral administration,
obstetrical practice’1. Even in 2005, misoprostol
the peak level of misoprostol acid is reached
was not approved by the FDA for use in preg-
within 12 ± 3 min, with a terminal half-life of
nant women, a stand strangely and strongly
20–40 min. Plasma levels of misoprostol acid
supported by its manufacturer2.
vary considerably between and within studies,
but mean values after single doses show a linear
relationship with the dose over the range of
200–400 µg. No accumulation of misoprostol
MISOPROSTOL acid was noted in multiple dose studies and a
Misoprostol is a synthetic PGE1 analog. Natu-
rally occurring PGE1 is not orally sustainable,
as it is unstable in acid media and is also not
suitable for parenteral use because of its rapid CH2 (CH2)5 COOCH3
degradation in the blood. Misoprostol, the syn-
thetic PGE1 analog, is produced by bringing
about an alteration in the chemical structure of H CH3
the naturally occurring compound, thereby C=C-CH2-C (CH2)3 CH3
making it orally stable and clinically useful.
Misoprostol is otherwise called alprostadil and
its chemical formula is C22H38O5 ((±)–methyl Figure 1 Chemical structure of misoprostol3
114
136
06 September 2006 16:37:31
Misoprostol: theory and practice
plasma steady state was achieved within 2 days. women after delivery11. The effects of
The bioavailability of misoprostol is decreased misoprostol on the reproductive tract are
when administered with food or antacids8. increased and gastrointestinal adverse effects
Misoprostol is primarily metabolized in the are decreased when it is administered vagi-
liver and less than 1% of its active metabolite is nally10,12,13. When misoprostol tablets are
excreted in the urine9. Patients with hepatic displaced in the posterior fornix of the vagina,
ease should receive decreased doses, whereas plasma concentrations of misoprostol acid peak
dose adjustment is not necessary for patients in 1–2 h and then decline slowly (Figure 2)5.
with renal disease who do not require dialysis. Vaginal application of misoprostol results in
Misoprostol has no known drug interactions slower increases and lower peak plasma concen-
and does not induce the hepatic enzyme trations of misoprostol acid than does oral
systems9. administration, but overall exposure to the drug
Pharmacokinetic studies in pregnant women is increased (indicated by the increased area
show that sublingual and oral misoprostol used under the curve in Figure 2)5. The peak plasma
for first-trimester termination of pregnancy levels of misoprostol are sustained for up to 4 h
produce earlier and higher peak plasma con- after vaginal administration5 (Figure 2).
centrations than vaginal or rectal misoprostol, Among women who were 9–11 weeks preg-
resulting in earlier, more pronounced uterine nant and given misoprostol before a surgical
tonus (oral misoprostol 7.8 ± 3.0 min vs. termination of pregnancy, intrauterine pressure
vaginal misoprostol 20.9 ± 5.3 min)6,7,10. These began to increase an average of 8 min after
findings have very recently been validated in oral administration and 21 min after vaginal
350
300
Plasma misoprostol acid (pg/ml)
250
200
150
Vaginal
100
50
Oral
0
0 60 120 180 240
Minutes
Figure 2 Mean (standard deviation) plasma concentrations of misoprostol acid after oral and vaginal
administration of misoprostol in 20 women. Reprinted from Zieman M, et al.5
115
137
30 August 2006 14:20:21
administration; it was maximal 25 min after side-effects include headache, abdominal

oral administration and 46 min after vaginal cramps, nausea and flatulence, chills, shivering,
administration, respectively. Uterine contractil- and fever, all of which are dose-dependent. It is
ity initially increased and then reached a plateau interesting to note that, before its use in preg-
1 h after oral administration, whereas uterine nant women, chills, shivering and fever were not
contractility increased continuously for 4 h after commonly reported side-effects, suggesting that
vaginal administration. Maximal uterine con- these are dose-dependent.
tractility was significantly higher after vaginal Package warnings are very clear that miso-
administration10. The maximum serum concen- prostol is not to be taken by pregnant women,
tration was achieved 23 min later in rectal and non-pregnant women should use contra-
administration and the peak levels were lower ceptives while taking misoprostol and should be
compared to oral administration of misoprostol7 warned about the effects of misoprostol if taken
(Figure 2). by pregnant women. Misoprostol should also be
In the pharmacokinetic study by Tang and avoided in nursing mothers because of concern
colleagues6, the peak plasma level of misopro- over causing diarrhea in the baby8,11.
stol acid was highest and earliest after adminis- Congenital anomalies sometimes associated
tration of misoprostol by the sublingual route. with fetal death have been reported subsequent
Misoprostol tablets dissolved in water and to the unsuccessful use of misoprostol for termi-
taken orally have also been shown to produce a nation of pregnancy, but the drug’s teratogenic
faster onset and stronger uterotonic effect than mechanism has not been elucidated17,18. Several
either oral tablet or rectal administration14,15. reports associate the use of misoprostol during
However, there was no significant difference the first trimester of pregnancy with skull
when misoprostol was used in the form of defects, cranial nerve palsies, facial malforma-
moistened tablets compared to dry tablets for tions (Mobius syndrome), and limb defects19.
first-trimester termination of pregnancy16. Misoprostol is listed as a pregnancy category X
drug.
Toxic doses of misoprostol have not been
Adverse effects
determined; however, pregnant women have
Common side-effects of misoprostol include tolerated cumulative doses up to 2200 µg
diarrhea and abdominal pain. Less common administered over a period of 12 h without
Serum misoprostol acid concentration (pg/ml)
400
300
oral
rectal
200
100
0
0 60 120 180 240
Time (mins)
Figure 3 Mean serum concentration of misoprostol acid over time with oral and rectal administration.
Error bars represent one standard deviation7
116
138
30 August 2006 14:20:22
any serious adverse effects20. A dose of 6000 µg 88–100% of women provided this regimen;
of misoprostol, taken orally to induce termina- 53–60% of women aborted within 24 h after
tion of pregnancy (with trifluoperazine), resulted one dose of misoprostol was administered33–39.
in abortion, hyperthermia, rhabdomyolysis, If complete abortion did not occur within
hypoxemia and a complex acid–base disorder21. that interval, repeating the misoprostol dose
resulted in complete termination of pregnancy
in 19–32% of women within 24 h after the
MISOPROSTOL IN THE FIRST
second dose33,34. The remaining 10–30% of
TRIMESTER
women who aborted successfully had a delayed
For first-trimester medical termination of response, with the abortion completed over an
pregnancy, misoprostol is used most extensively average period of 24–28 days33,34.
in conjunction with either mifepristone or Misoprostol has also been used alone for
methotrexate. Both regimens are effective. In medical termination of pregnancy, with variable
the initial studies of mifepristone and miso- efficacy. The earliest studies of misoprostol-
prostol for medical termination of pregnancy, induced termination of pregnancy in the first
both drugs were given orally. Only regimens of trimester reported complete abortion rates of
mifepristone in combination with oral miso- 5–11% among women given a total dose of
prostol have been licensed for abortion in 400 µg of oral misoprostol40,41. Up to three
any country. Administration of 600 mg of oral 800-µg doses of vaginal misoprostol given every
mifeprostone followed 48 h later by 400 µg of 48 h resulted in complete termination of preg-
oral misoprostol resulted in 91–97% complete nancy in up to 96% of women who were no
abortion in women who were no more than 49 more than 63 days pregnant42. However, in a
days pregnant, compared to 83–95% of women randomized trial comparing methotrexate plus
who were no more than 56 days pregnant22–25. vaginal misoprostol with vaginal misoprostol
Lowering the dose of mifepristone to 200 mg alone, only 47% of the women given miso-
and increasing the dose of oral misoprostol to prostol alone had complete termination of
600 µg increases the efficacy, with abortion pregnancy, as compared with 90% of the
rates of 96–97% among women no more than women given methotrexate plus misoprostol
49 days pregnant and 89–93% among women (p < 0.001)43.
50–63 days pregnant26,27. A combined regimen With regard to the use of misoprostol as a
of mifepristone and misoprostol can result in cervical-priming agent before vacuum aspira-
complete abortion in 94–95% for medical abortion of the uterus, numerous randomized, con-
tion to women of 9–13 weeks pregnancy but trolled studies have shown that misoprostol is
is associated with high incidence of heavy more effective than placebo and vaginal PGE2
bleeding28,29. The timing of administration of in terms of the degree of cervical dilatation
misoprostol after mifeprostone for medical ter- achieved44,45. As cervical priming facilitates sur-
mination of pregnancy ranges from 6 to 48 h. gical vacuum aspiration, the risks of dilatation
The complete abortion rates improve with one and evacuation of the uterus are therefore mini-
or two additional doses of misoprostol. mized. These results were replicated by numer-
Vaginal administration of misoprostol was ous other randomized, controlled trials involving
more effective and better tolerated than oral a large number of participants. The best regi-
administration for the induction of first- men for cervical ripening in the first trimester
trimester abortion30,31. However, some studies is 400 µg of vaginal misoprostol given 3–4 h
concluded that both oral and vaginal miso- before suction curettage44,46,47. In one study,
prostol were of similar efficacy. Sublingual misoprostol, when administered with mife-
administration of misoprostol had a success rate prostone for termination of early pregnancy in
of 92%32. scarred uteri, was safe and effective, but further
A single dose of intramuscular or oral metho- randomized trials are essential to confirm this48.
trexate (50 mg per square meter of body-surface Sublingual misoprostol was effective in facili-
area) followed 5–7 days later by 800 µg of vagi- tating cervical dilatation before surgical abor-
nal misoprostol resulted in complete abortion in tion, and its use significantly decreased the time
117
139
30 August 2006 14:20:22
of surgical evacuation and minimized blood loss 1000

during the procedure49,50.
800
Single dose (µg)

Danger
600
MISOPROSTOL IN EARLY Safe
PREGNANCY FAILURE 400
Single or repeated doses of misoprostol result in 200 Ineffective
complete expulsions with minimal side-effects
and complications in evacuation of first- 0
10 20 30 40
trimester missed abortions51,52. Vaginal miso-
prostol is more effective then oral admini- Gestation (weeks)
stration53. Misoprostol is also effective in
incomplete termination of pregnancy, and it is Figure 4 Safe single doses of misoprostol for
producing uterine contractions at various
safer than the surgical method54,55.
gestations63
MISOPROSTOL IN THE SECOND

abortion in the first trimester, whereas doses in
TRIMESTER
the range of 25–50 µg induce labor in the third
Indications for termination of pregnancy in trimester. The optimal dose of vaginal miso-
the second trimester include chromosomal and prostol for induction of labor in the second
structural fetal abnormalities as well as social trimester probably lies somewhere between
reasons. Surgical evacuation of the uterus, still 50 and 800 µg. Within this range, higher doses
being practiced in some centers, is associated may be needed to cause termination of preg-
with greater morbidity, mortality and complica- nancy early in the second trimester, whereas
tions. Intra-amniotic hypertonic saline/urea lower doses may be sufficient later in the second
instillation, intra-amniotic PGF2 infusion, trimester. Higher and more frequent doses are
extra-amniotic ethacridine lactate, oxytocin associated with shorter drug administration-to-
infusion and vaginal PGE2 were practiced abortion interval compared to lower and less
before the introduction of misoprostol. frequent doses1,63 (Figure 4).
Intravaginal misoprostol in the dose of
400 µg is effective and associated with fewer
side-effects56. Vaginal misoprostol was found to MISOPROSTOL IN THE THIRD
be as effective as or more effective than PGE2. TRIMESTER
Misoprostol was equally effective as extra-
Induction of labor
amniotic prostaglandins57–60. Misoprostol in the
dose of 400 µg every 3 h was more effective in This is one of the common obstetric inter-
terms of a significantly shorter drug administra- ventions primarily performed with the aim of
tion-to-abortion interval and a higher percent- reducing maternal and perinatal morbidity and
age of successful abortion within 48 h compared mortality. The success of induction of labor
to misoprostol 400 µg every 6 h, and the inci- not only lies with replication of physiological
dence of side-effects was similar in both groups mechanisms, but also depends upon the cervical
except for that of fever. However, the fever status. An unfavorable cervix presents the great-
returned to normal within 24 h after the last est challenge to successful induction. The devel-
dose of misoprostol61. opment of effective, safe (to both mother and
Vaginal misoprostol was significantly more fetus) and less expensive pharmacological
effective as judged by drug administration-to- agents to accomplish this task has been the
abortion interval and the need to augment the focus of much clinical research.
therapy with oxytocin infusion when compared The results of the first study (1993) sug-
to oral misoprostol62. gested that misoprostol is a cost-effective and
It is somewhat paradoxical that a greater dose safe alternative for induction of labor at term.
(800 µg) of vaginal misoprostol is essential for Many later studies, including randomized trials,
118
140
30 August 2006 14:20:23
not only confirmed this finding, but also have vigilance86,87. The use of prostaglandins includ-
shown that misoprostol is more effective than a ing misoprostol increases the uteroplacental
placebo or other prostaglandins; moreover, it is resistance but does not affect the umbilical
associated with a higher rate of vaginal delivery blood flow in a Doppler velocimetry study of
within 24 h, a shorter induction-to-delivery umbilical, uterine and arcuate arteries immedi-
interval and significantly lower Cesarean ately before and 2–3 h after the administration
section rates than pooled figures for the control of vaginal misoprostol or cervical PGE2, thus
groups64–67. suggesting misoprostol is as safe as PGE2 gel88.
Studies of different routes of misoprostol Available data suggest that vaginal misoprostol
administration were conducted for induction of in a dose of 25 µg every 6 h is as safe as PGE2
labor, including oral, vaginal, intracervical and in patients with a live fetus for induction of
sublingual68–73. Although all routes were suc- labor.
cessful, vaginal misoprostol is associated with
a shorter induction-to-delivery interval, lower
Induction of labor after fetal death
number of doses and diminished oxytocin
use68,70. Misoprostol gel is associated with fewer Misoprostol is ideally suited agent for induction
uterine contraction abnormalities and longer of labor after fetal death as there is no concern
induction-to-labor and delivery interval when about the adverse effects of uterine hyper-
compared to misoprostol tablets71. stimulation on the fetus. For fetal death at term,
The safety of misoprostol is crucial, as some a dose as low as 50 µg every 12 h may be ade-
studies have shown a high frequency of uterine quate for induction of labor, whereas higher
tachysystole and hyperstimulation, including doses are necessary in patients with fetal death
some reports of uterine rupture during the in the second trimester and early in the third
induction of the labor with misoprostol74–76. A trimester89,90.
vaginal dose of 25 µg is often recommended as
the more prudent dose, as it is associated with
THIRD STAGE OF LABOR
lower incidence of uterine hyperstimulation and
it is comparable to the 50 µg dose in achieving Postpartum hemorrhage is a major cause of
delivery within 24 h64,77–81. Doses higher than maternal morbidity and mortality. It is sudden,
50 µg have been associated with increased risk dramatic and unpredictable. In developing
of complications. The interval of administration countries, over 125 000 or approximately 28%
of misoprostol ranges from every 3 to 6 h. It is of total maternal deaths are caused by post-
better to use 6-h dosing intervals to avoid the partum hemorrhage each year. According to
possible risk of tachysystole82. Misoprostol is one estimate, the risk is approximately 1 in 1000
also effective as a cervical ripening agent for deliveries91.
prelabor rupture of the membranes83. Oral Based on misoprostol’s uterotonic effects,
misoprostol not only induced labor but also this drug has been evaluated for both prevention
resulted in delivery within 24 h without increas- and treatment of postpartum hemorrhage (see
ing maternal or neonatal complications66,84. Chapters 4 and 16–19). The WHO misoprostol
Misoprostol is not recommended for induc- multicenter trial concluded that use of an oral
tion in cases of previous Cesarean section, as it tablet of 600 µg was associated with a higher
is associated with higher frequency of disruption risk of severe postpartum hemorrhage, the need
of prior uterine incision compared with use of for additional uterotonic agents, shivering and
PGE2 or oxytocin. Misoprostol use is associated pyrexia compared with intramuscular or
with a 5.6% rupture of uterine scars compared intravenous oxytocin92. However, the dose
to 0.2% in patients attempting vaginal birth of misoprostol used in these trials varied from
after Cesarean delivery without stimulation, as 400 to 600 µg (orally and rectally). Moreover,
shown by meta-analysis85. Misoprostol use in the frequency of postpartum hemorrhage
grand multipara is not associated with adverse (blood loss > 1000 ml) was not lower in the
maternal or neonatal outcome. However, misoprostol group than in the control group in
its use in such patients warrants strict any of the trials. None the less, there was higher
119
141
04 September 2006 13:47:17
use of oxytocin in the control groups. In many Global Network for Women’s and Children’s
reports, misoprostol 600 µg oral or 400 µg Health Research randomized, placebo-
rectal is significantly less effective than inject- controlled trial of misoprostol in home delivery
able uterotonics in preventing postpartum settings in rural India will perhaps answer the
hemorrhage92–103. Misoprostol at the dose of question regarding the benefit of oral miso-
400–600 µg is associated with risk of shivering, prostol in the prevention of postpartum
and doses more than 400 µg also significantly hemorrhage108 (see also Chapter 8).
increase the risk of pyrexia. At present, oral or Oral misoprostol in the dose 600 µg was
rectal misoprostol is not as effective as conven- associated with lower incidence of measured
tional injectable uterotonic agents, and the high blood loss ≥ 500 ml and lower incidence of
rates of shivering and fever associated with reduction in postpartum hemoglobin (reduction
its use make it undesirable for routine use to of hemoglobin ≥ 2 g/dl was 16.4% with miso-
prevent postpartum hemorrhage, especially for prostol and 21.2% with ergometrine), but the
low-risk women. Thus, there is insufficient difference were not statistically significant.
evidence to date to support the routine use Shivering was significantly more common with
of misoprostol when oxytocin or methylergo- misoprostol, whereas vomiting was more
metirne is available. There is some evidence of common with ergometrine in a randomized,
increased uterotonic effect with the administra- controlled trial with misoprostol 600 µg and
tion of misoprostol, either by the sublingual ergometrine (0.5 mg, four tablets) in the home
route or as an oral solution6,14,15. Use of buccal delivery settings of rural Gambia109. Rectal
misoprostol in a placebo-controlled trial to pre- misoprostol has also been reported to control
vent hemorrhage at Cesarean delivery was not postpartum hemorrhage that is unresponsive to
associated with a significant difference between oxytocin and methylergometrine in the dose of
the two groups, both in the incidence of post- 1000 µg110.
partum hemorrhage and a difference in pre- Rectal misoprostol in the dose of 800 µg
and postoperative hemoglobin level. However, could be a useful first-line drug for the treat-
misoprostol reduced the need for additional ment of primary postpartum hemorrhage111. In
uterotonic agents during Cesarean delivery104. a randomized, double-blind, placebo-controlled
In all of these studies, it is important to note that trial with sublingual misoprostol at a primary
misoprostol was compared to conventional health center in Bissau, Guinea-Bissau, West
uterotonics. It is tragic but true, however, that Africa, the incidence of postpartum hemorrhage
these latter drugs are not available in many parts was not significantly different between the two
of the world where women deliver with no groups. However, significantly fewer women in
medical assistance whatsoever. the misoprostol group experienced a blood
Despite the lesser efficacy of misoprostol loss of ≥ 1000 ml or ≥ 1500 ml. The decrease in
compared with conventional injectable oxyto- hemoglobin concentration tended to be less
cics and the potential to cause side-effects, in the misoprostol group, the mean difference
several factors – ease of use, stability in field between the two groups being 0.16 mmol/l
conditions, longer shelf-life, and less expense (−0.01 to 0.32 mmol/l). From this study, it was
– underlie its continued evaluation as a concluded that sublingual misoprostol reduces
uterotonic agent. It remains of great interest, the frequency of severe postpartum hemor-
especially for use in home deliveries by tradi- rhage112. Further randomized, controlled trials
tional birth attendants and minimally qualified are necessary to identify the best drug dose and
nurse midwives in less developed areas where route of administration for the treatment of
administration of injectable uterotonics may not postpartum hemorrhage113, particularly when
be feasible or may not be available. Here, it use of conventional agents is not possible. Such
offers a plausible preventive strategy in such studies should differentiate the site of use for
areas for reducing maternal mortality related to misoprostol. When there is no syringe to inject
postpartum hemorrhage105–107. The results of methergin or oxytocin or no refrigeration facility
an ongoing National Institute of Child Health for these drugs, misoprostol may be the only
and Human Development (NICHD) sponsored viable option and should be compared to the
120
142
30 August 2006 14:20:24
local standard practice (where there is no use of manufacturer of misoprostol, the FDA has
uterotonic agents). approved a new label for the use of misoprostol
during pregnancy. The new labeling revises the
contraindications and the precaution that
OTHER USES
misoprostol should not be used in pregnant
Misoprostol is used prior to procedures such women by stating that the contraindication is
as intrauterine insemination and hystero- only for pregnant women who are using the
scopy114–116. Its use in cervical pregnancy is medication to reduce the risk of non-steroidal
documented with one case report; however, anti-inflammatory drugs. Misoprostol is now a
extreme caution is recommended with this legitimate part of the FDA-approved regimen
approach and methotrexate is favored by many for use with mifeprostone to induce abortion
authorities117. in early abortion and is also recognized for
induction of labor118.
CONCLUSION
Misoprostol is one of the most important medi- References
cations in obstetric practice. As of the time of 1. Goldberg AB, Greenberg MB, Darney PD.
writing, its use in pregnant women remains Misoprostol and pregnancy. N Engl J Med
unapproved by the US FDA, except in conjunc- 2001;344:38–47
tion with mifeprostone (or, in some cases, 2. Friedman MA. Manufacturer’s warning regard-
methotrexate) for first-trimester medical termi- ing unapproved uses of misoprostol. N Engl J
nation of pregnancy. Despite this, the interna- Med 2001;344:61
3. Barik S, Datta S, Gupta K. Misoprostol:
tional literature is replete with favorable reports
pharmacology. In Barik S, Datta S, Gupta K,
of off-label uses. For example, there is strong eds. Misoprostol in Obstetrics and Gynecology.
and consistent evidence to support the use of New Delhi: Jaypee Brothers, 2003:8–15
misoprostol for cervical ripening before surgical 4. Yap-Seng Chong, Lin Lin Su, Arulkumaran S.
abortion in the first trimester and for induction Misoprostol: a quarter century of use, abuse,
of labor in the second and third trimesters. and creative misuse. Obstet Gynecol Survey
Whereas lower dose and strict vigilance are 2004;59:128–40
required for use of misoprostol for induction of 5. Zieman M, Fong SK, Benowitz NL, et al.
labor with a live fetus, it is ideal for induction of Absorption kinetics of misoprostol with oral or
labor in patients with intrauterine fetal death. vaginal administration. Obstet Gynecol 1997;90:
Misoprostol may also prevent postpartum 88–92
6. Tang OS, Schweer H, Seyberth HW, et al.
hemorrhage when injectable uterotonic agents
Pharmacokinetics of different routes of admin-
are either impractical or unavailable. On the
istration of misoprostol. Hum Reprod 2002;17:
other hand, misoprostol may not be the pre- 332–6
ferred uterotonic for prevention of postpartum 7. Khan RU, El-Refaey H. Pharmacokinetics and
hemorrhage where injectable oxytocics are adverse-effect profile of rectally administered
readily available. Its use in the treatment of misoprostol in the third stage of labor. Obstet
postpartum hemorrhage in regions of the world Gynecol 2003;101:968–74
where the standard of care is delivery without 8. Searle: Cytotec (misoprostol) (information
uterotonic agents (i.e. delivery with no utero- package). Chicago: GD Searle & Co, 1995
tonic medication) needs further evaluation. The 9. Foote EF, Lee DR, Karim A, et al. Disposition
oral route is associated with faster effect and of misoprostol and its active metabolite in
patients with normal and impaired renal
with more side-effects. The other routes, such
function. J Clin Pharmacol 1995;35:384–9
as vaginal and rectal, have sustained and longer
10. Danielsson KG, Marions L, Rodriguez A, et al.
effects with less side-effects. The sublingual and Comparison between oral and vaginal adminis-
buccal routes and dose need further evaluation. tration of misoprostol on uterine contractility.
Finally, after considerable dialogues between Obstet Gynecol 1999;93:275–80
the American College of Obstetricians and 11. Abdel-Aleem H, Villar J, Gulmezoglu
Gynecologists (ACOG) and Searle, the AM, et al. The pharmacokinetics of the
121
143
30 August 2006 14:20:24
prostaglandin E1 analogue misoprostol in 23. Peyron R, Aubeny E, Targosz V, et al. Early

plasma and colostrum after postpartum oral termination of pregnancy with mifepristone
administration. Eur J Obstet Gynecol Reprod Biol (RU 486) and the orally active prostaglandin
2003;108:25–8 misoprostol. N Engl J Med 1993;328:1509–13
12. Creinin MD, Darney PD. Methotrexate and 24. Spitz IM, Bardin CW, Benton L, Robbins A.
misoprostol for early abortion. Contraception Early pregnancy termination with mifepristone
1993;48:339–48 [Erratum, Contraception 1994; and misoprostol in the United States. N Engl J
49:99] Med 1998;338:1241–7
13. Toppozada MK, Anwar MY, Hassan HA, 25. Winikoff B, Sivin I, Coyaji KJ, et al. Safety,
el-Gazaerly WS. Oral or vaginal misoprostol for efficacy, and acceptability of medical abortion
induction of labor. Int J Gynaecol Obstet 1997; in China, Cuba, and India: a comparative trial
56:135–9 of mifepristone–misoprostol versus surgical
14. Chong YS, Chua S, Arulkumaran S. Sublingual abortion. Am J Obstet Gynecol 1997;176:431–7
misoprostol for first trimester termination of 26. McKinley C, Thong KJ, Baird DT. The effect
pregnancy: safety concerns. Hum Reprod 2002; of dose of mifepristone and gestation on the
17:2777–8 efficacy of medical abortion with mifepristone
15. Chong YS, Chua S, Shen L, et al. Does the and misoprostol. Hum Reprod 1993;8:1502–5
route of administration of misoprostol make 27. Baird DT, Sukcharoen N, Thong KJ. Random-
a difference? The uterotonic effect and side ized trial of misoprostol and cervagem in com-
effects of misoprostol given by different bination with a reduced dose of mifepristone for
routes after vaginal delivery. Eur J Obstet induction of abortion. Hum Reprod 1995;10:
Gynecol Reprod Biol 2004;113:191–8 1521–7
16. Creinin MD, Carbonell JL, Schwartz JL, 28. Ashok PW, Flett GM, Templeton A. Termina-
Varela L, Tanda R. A randomized trial of tion of pregnancy at 9–13 weeks’ amenorrhoea
the effect of moistening misoprostol before with mifepristone and misoprostol. Lancet
vaginal administration when used with metho- 1998;352:542–3
trexate for abortion. Contraception 1999;59: 29. Gouk EV, Lincoln K, Khair A, Haslock J,
11–16 Knight J, Cruickshank DJ. Medical termination
17. Pastuszak AL, Schuler L, Speck-Martins CE, of pregnancy at 63 to 83 days gestation. Br J
et al. Use of misoprostol during pregnancy and Obstet Gynaecol 1999;106:535–9
Mobius’ syndrome in infants. N Engl J Med 30. el-Refaey H, Rajasekar O, Abdalla M, et al.
1998;338:1881–5 Induction of abortion with mifepristone
18. Gonzalez CH, Marques-Dias MJ, Kim CA, (RU486) and oral or vaginal misoprostol. N
et al. Congenital abnormalities in Brazilian Engl J Med 1995;332:983–7
children associated with misoprostol misuse in 31. Carbonell JL, Velazco A, Rodriguez Y, et al.
first trimester of pregnancy. Lancet 1998;351: Oral versus vaginal misoprostol for cervical
1624–7 priming in first-trimester abortion: a random-
19. Orioli IM, Castilla EE. Epidemiological assess- ized trial. Eur J Contracept Reprod Health Care
ment of misoprostol teratogenicity. Br J Obstet 2001;6:134–40
Gynaecol 2000;107:519–23 32. Tang OS, Ho PC. Pilot study on the use of
20. el-Refaey H, Templeton A. Induction of abor- sublingual misoprostol for medical abortion.
tion in the second trimester by a combination Contraception 2001;64:315–17
of misoprostol and mifepristone: a randomized 33. Creinin MD, Vittinghoff E, Galbraith S, Klaisle
comparison between two misoprostol regimens. C. A randomized trial comparing misoprostol
Hum Reprod 1995;10:475–8 three and seven days after methotrexate for
21. Bond GR, Van Zee A. Overdosage of miso- early abortion. Am J Obstet Gynecol 1995;173:
prostol in pregnancy. Am J Obstet Gynecol 1994; 1578–84
171:561–2 34. Creinin MD, Vittinghoff E, Keder L, Darney
22. Wu YM, Gomex-Alzugaray M, Haukkamaa M, PD, Tiller G. Methotrexate and misoprostol for
et al. Task force on Post ovulatory Methods early abortion: a multicenter trial. I. Safety and
of Fertility Regulation (WHO). Comparison of efficacy. Contraception 1996;53:321–7
two doses of mifepristone in combination with 35. Creinin MD, Vittinghoff E, Schaff E, Klaisle C,
misoprostol for early medical abortion: a Darney PD, Dean C. Medical abortion with
randomised trial. Br J Obstet Gynaecol 2000; oral methotrexate and vaginal misoprostol.
107:524–30 Obstet Gynecol 1997;90:611–16
122
144
30 August 2006 14:20:24
36. Creinin MD. Oral methotrexate and vaginal 50. Vimala N, Mittal S, Kumar S. Sublingual miso-
misoprostol for early abortion. Contraception prostol for preabortion cervical ripening in first-
1996;54:15–18 trimester pregnancy termination. Contraception
37. Carbonell Esteve JL, Varela L, Velazco A, 2003;67:295–7
Tanda R, Sanchez C. 25 mg or 50 mg of oral 51. Herabutya Y, O-Prasertsawat P. Misoprostol
methotrexate followed by vaginal misoprostol 7 in the management of missed abortion. Int J
days after for early abortion: a randomized trial. Gynaecol Obstet 1997;56:263–6
Gynecol Obstet Invest 1999;47:182–7 52. Wakabayashi M, Tretiak M, Kosasa T, et al.
38. Hausknecht RU. Methotrexate and misoprostol Intravaginal misoprostol for medical evacuation
to terminate early pregnancy. N Engl J Med of first trimester missed abortion. Prim Care
1995;333:537 1998;5:176
39. Carbonell JL, Varela L, Velazco A, Cabezas E, 53. Creinin MD, Moyer R, Guido R. Misoprostol
Fernandez C, Sanchez C. Oral methotrexate for medical evacuation of early pregnancy
and vaginal misoprostol for early abortion. failure. Obstet Gyneco1 1997;89:768–72
Contraception 1998;57:83–8 54. Henshaw RC, Cooper K, el-Refaey H, et al.
40. Lewis JH. Summary of the 29th meeting of the Medical management of miscarriage: non-
Gastrointestinal Drugs Advisory Committee, surgical uterine evacuation of incomplete and
Food and Drug Administration, June 10, 1985. inevitable spontaneous abortion. Br Med J
Am J Gastroenterol 1985;80:743–5 1993;306:894–5
41. Norman JE, Thong KJ, Baird DT. Uterine 55. Chung TK, Lee DT, Cheung LP, et al. Sponta-
contractility and induction of abortion in early neous abortion: a randomized, controlled trial
pregnancy by misoprostol and mifepristone. comparing surgical evacuation with conserva-
Lancet 1991;338:1233–6 tive management using misoprostol. Fertil Steril
42. Carbonell JL, Varela L, Velazco A, Fernandez 1999;71:1054–9
C. The use of misoprostol for termination of 56. Bugalho A, Bique C, Almeida L, et al. The
early pregnancy. Contraception 1997;55:165–8 effectiveness of intravaginal misoprostol (Cyto-
43. Creinin MD, Vittinghoff E. Methotrexate and tec) in inducing abortion after eleven weeks of
misoprostol vs misoprostol alone for early pregnancy. Stud Fam Plann 1993;24:319–23
abortion: a randomized controlled trial. JAMA 57. Nuutila M, Toivonen J, Ylikorkala O, et al. A
1994;272:1190–5 comparison between two doses of intravaginal
44. Bugalho A, Bique C, Almeida L, et al. Applica- misoprostol and gemeprost for induction of sec-
tion of vaginal misoprostol before cervical dila- ond-trimester abortion. Obstet Gynecol 1997;
tation to facilitate first-trimester pregnancy 90:896–900
interruption. Obstet Gynecol 1994;83:729–31 58. Dickinson JE, Godfrey M, Evans SF. Efficacy
45. Ngai SW, Yeung KC, Lao T, et al. Oral miso- of intravaginal misoprostol in second-trimester
prostol versus vaginal gemeprost for cervical pregnancy termination: a randomized con-
dilatation prior to vacuum aspiration in women trolled trial. J Matern Fetal Med 1998;7:115–19
in the sixth to twelfth week of gestation. Contra- 59. Wong KS, Ngai CS, Wong AY, et al. Vaginal
ception 1995;51:347–50 misoprostol compared with vaginal gemeprost
46. Singh K, Fong YF, Prasad RN, Dong F. in termination of second trimester pregnancy:
Randomized trial to determine optimal dose a randomized trial. Contraception 1998;58:
of vaginal misoprostol for preabortion cervical 207–10
priming. Obstet Gynecol 1998;92:795–8 60. Munthali J, Moodley J. The use of misoprostol
47. Singh K, Fong YF, Prasad RN, Dong F. for mid-trimester therapeutic termination of
Evacuation interval after vaginal misoprostol pregnancy. Trop Doct 2001;31:157–61
for preabortion cervical priming: a randomized 61. Wong KS, Ngai CS, Yeo EL, et al. A compari-
trial. Obstet Gynecol 1999;94:431–4 son of two regimens of intravaginal misoprostol
48. Xu J, Chen H, Ma T, et al. Termination of early for termination of second trimester pregnancy:
pregnancy in the scarred uterus with mife- a randomized comparative trial. Hum Reprod
pristone and misoprostol. Int J Gynaecol Obstet 2000;15:709–12
2001;72:245–51 62. Gilbert A, Reid R. A randomised trial of oral
49. Saxena P, Salhan S, Sarda N. Role of sub- versus vaginal administration of misoprostol for
lingual misoprostol for cervical ripening prior to the purpose of mid-trimester termination of
vacuum aspiration in first trimester interruption pregnancy. Aust N Z J Obstet Gynaecol 2001;41:
of pregnancy. Contraception 2003;67:213–17 407–10
123
145
30 August 2006 14:20:24
63. Fiala C, Weeks A. Misoprostol in obstetrics prostaglandin E1 analog misoprostol for labor
and gynaecology, summary of evidence, induction. Am J Obstet Gynecol 1997;177:
www.misoprostol.org 364–9; discussion 369–71
64. Hofmeyr GJ. Vaginal misoprostol for cervical 78. Srisomboon J, Tongsong T, Tosiri V. Pre-
ripening and labour induction in late pregnancy induction cervical ripening with intravaginal
(Cochrane review). Cochrane Library, Issue 4, prostaglandin E1 methyl analogue misoprostol:
Oxford: Update Software, 1999 a randomized controlled trial. J Obstet Gynaecol
65. Hofmeyr GJ, Gulmezoglu AM, Alfirevic Z. Miso- Res 1996;22:119–24
prostol for induction of labour: a systematic 79. Wing DA, Paul RH. A comparison of differing
review. Br J Obstet Gynaecol 1999;106:798–803 dosing regimens of vaginally administered
66. Sanchez-Ramos L, Chen AH, Kaunitz AM, misoprostol for preinduction cervical ripening
et al. Labor induction with intravaginal miso- and labor induction. Am J Obstet Gynecol 1996;
prostol in term premature rupture of mem- 175:158–64
branes: a randomized study. Obstet Gynecol 80. Diro M, Adra A, Gilles JM, et al. A double-
1997;89:909–12 blind randomized trial of two dose regimens
67. Sanchez-Ramos L, Kaunitz AM, Wears RL, of misoprostol for cervical ripening and labor
et al. Misoprostol for cervical ripening and labor induction. J Matern Fetal Med 1999;8:114–18
induction: a meta-analysis. Obstet Gynecol 81. Meydanli MM, Caliskan E, Burak F, et al.
1997:89:633–42 Labor induction post-term with 25 micrograms
68. Toppozada MK, Anwar MY, Hassan HA, et al. vs. 50 micrograms of intravaginal misoprostol.
Oral or vaginal misoprostol for induction of Int J Gynaecol Obstet 2003;81:249–55
labour. Int J Gynaecol Obstet 1997;56:135–9 82. Wing DA, Paul RH. A comparison of differing
69. Adair CD, Weeks JW, Barrilleaux S, et al. dosing regimens of vaginally administered
Oral or vaginal misoprostol administration for misoprostol for preinduction cervical ripening
induction of labor: a randomized, double-blind and labor induction [Erratum). Am J Obstet
trial. Obstet Gynecol 1998;92:810–13 Gynecol 1997;176:1423
70. Nopdonrattakoon L. A comparison between 83. Ngai SW, To WK, Lao T, et al. Cervical prim-
intravaginal and oral misoprostol for labor ing with oral misoprostol in pre-labor rupture of
induction: a randomized controlled trial. J membranes at term. Obstet Gynecol 1996;87:
Obstet Gynaecol Res 2003;29:87–91 923–6
71. Liu HS, Chu TV, Chang YK, et al. Intra- 84. Shetty A, Stewart K, Stewart G, et al. Active
cervical misoprostol as an effective method of management of term prelabour rupture of
labor induction at term. Int J Gynaecol Obstet membranes with oral misoprostol. Br J Obstet
1999;64:49–53 Gynaecol 2002;109:1354–8
72. Shetty A, Mackie L, Danielian P, et al. Sub- 85. Plaut MM, Schwartz ML, Lubarsky SL. Uter-
lingual compared with oral misoprostol in term ine rupture associated with the use of miso-
labour induction: a randomized controlled trial. prostol in the gravid patient with a previous
Br J Obstet Gynaecol 2002;109:645–50 cesarean section. Am J Obstet Gynecol 1999;
73. Shetty A, Daliellan P, Templeton A. Sublingual 180:1535–42
misoprostol for the induction of labour at term. 86. Bique C, Bugalho A, Bergstrom S. Labor
Am J Obstet Gynecol 2002;186:72–6 induction by vaginal misoprostol in grand
74. Wing DA, Tran S, Paul RH. Factors affecting multiparous women. Acta Obstet Gynecol Scand
the likelihood of successful induction after 1999;78:198–201
intravaginal misoprostol application for cervical 87. Induction of labor. ACOG Practice Bulletin 10.
ripening and labor induction. Am J Obstet Washington, DC: American College of Obste-
Gynecol 2002;186:1237–40 tricians and Gynecologists, 1999
75. Bennett BB. Uterine rupture during induction 88. Urban R, Lemancewicz A, Urban J, et al.
of labor at term with intravaginal misoprostol. Misoprostol and dinoprostone therapy for labor
Obstet Gynecol 1997;89:832–3 induction: a Doppler comparison of uterine and
76. Wing DA, Lovett K, Paul RH. Disruption of fetal hemodynamic effects. Eur J Obstet Gynecol
prior uterine incision following misoprostol for Reprod Biol 2003;106:20–4
labor induction in women with previous cesar- 89. Bugalho A, Bique C, Machungo F, Faundes A.
ean delivery. Obstet Gynecol 1998;91:828–30 Induction of labor with intravaginal miso-
77. Farah LA, Sanchez-Ramos L, Rosa C, et al. prostol in intrauterine fetal death. Am J Obstet
Randomized trial of two doses of the Gynecol 1994;171:538–41
124
146
30 August 2006 14:20:25
90. Bugalho A, Bique C, Machungo F, Bergstrom the third stage of labour. Hum Reprod 2001;16:
S. Vaginal misoprostol as an alternative to oxy- 31–5
tocin for induction of labor in women with late 102. Kundodyiwa TW, Majoko F, Rusakaniko S.
fetal death. Acta Obstet Gynecol Scand 1995;74: Misoprostol versus oxytocin in the third stage of
194–8 labor. Int J Gynaecol Obstet 2001;75:235–41
91. Drife J. Management of postpartum hemor- 103. Caliskan E, Oilbaz B, Meydanli MM, et al. Oral
rhage. Br J Obstet Gynaecol 1997;104:275–7 misoprostol for the third stage of labor: a
92. Gulmezoglu AM, Villar J, Ngoc NT, et al. randomized controlled trial. Obstet Gynecol
WHO multicentre randomised trial of miso- 2003;101:921–8
prostol in the management of the third stage of 104. Hamm J, Russel Z, Botha T, et al. Buccal
labour. Lancet 2001;358:689–95 misoprostol to prevent hemorrhage at cesarean
93. Hofmeyr GJ, Nikodem VC, de Jager M, et al. A delivery: a randomized study. Am J Obstet
randomized placebo controlled trial of oral Gynecol 2005;192: 1404–6
misoprostol in the third stage of labour. Br J 105. Joy SD, Sanchez-Ramos L, Kaunitz AM. Miso-
Obstet Gynaecol 1998;105:971–5 prostol use during the third stage of labor. Int J
94. Hofmeyr GJ, Nikodem C, de Jager M, et al. Gynaecol Obstet 2003;82:143–52
Oral misoprostol for labour third stage manage- 106. Chong YS, Chua S, Arulkumaran S. Severe
ment: randomised assessment of side effects hyperthermia following oral misoprostol in the
(part 2). Proceedings of the 17th Conference immediate postpartum period. Obstet Gynecol
on Priorities in Perinatal Care in South Africa, 1997;90:703–4
1998:53–4 107. Chong YS, Chua S, El-Refaey H, et al. Post-
95. Surbek DV, Fehr PM, Hosli I, et al. Oral miso- partum intrauterine pressure studies of the
prostol for third stage of labor: a randomized uterotonic effect of oral misoprostol and intra-
placebo-controlled trial. Obstet Gynecol 1999; muscular syntometrine. Br J Obstet Gynaecol
94:255–8 2001;108:41–7
96. Hofmeyr GJ, Nikodem VC, de Jager M, et al. 108. Kodkany BS, Derman RJ, Goudar SS, et al.
Side-effects of oral misoprostol in the third Initiating a novel therapy in preventing post-
stage of labour – a randomised placebo- partum hemorrhage in rural India: a joint
controlled trial. S Afr Med J 2001;91:432–5 collaboration between the United States and
97. Lumbiganon P, Hofmeyr J, Gulmezoglu AM, India. Int J Fertil 2004;49:91–6
et al. Misoprostol dose-related shivering and 109. Walraven G, Blum J, Dampha Y, et al. Miso-
pyrexia in the third stage of labour. WHO Col- prostol in the management of the third stage
laborative Trial of Misoprostol in the Manage- of labor in the home delivery setting in rural
ment of the Third Stage of Labour. Br J Obstet Gambia: a randomized controlled trial. Br J
Gynaecol 1999;106:304–8 Obstet Gynaecol 2005;112:1277–83
98. Cook CM, Spurrett B, Murray H. A random- 110. O’Brien P, El-Refaey H, Gordon A, et al.
ized clinical trial comparing oral misoprostol Rectally administered misoprostol for the treat-
with synthetic oxytocin or syntometrine in the ment of postpartum hemorrhage unresponsive
third stage of labour. Aust N Z J Obstet Gynaecol to oxytocin and ergometrine: a descriptive
1999;39:414–19 study. Obstet Gynecol 1998;92:212–14
99. Amant F, Spitz B, Timmerman D, et al. Miso- 111. Mousa HA, Alfirevic Z. Treatment for primary
prostol compared with methylergometrine for postpartum hemorrhage. Cochrane Database of
the prevention of postpartum haemorrhage: systematic reviews 2003. 1. A meta analysis
a double-blind randomised trial. Br J Obstet of randomized trials on the use of misoprostol
Gynaecol 1999;106:1066–70 for the treatment of postpartum hemorrhage.
100. el-Refaey H, Nooh R, O’Brien P, et al. The Cochrane Database Syst Rev 2003;1:CD003249
misoprostol third stage of labour study: a 112. Høj L, Cardoso P, Nielsen BB, Hvidman L,
randomised controlled comparison between Nielsen J, Aaby P. Effect of sublingual miso-
orally administered misoprostol and standard prostol on severe postpartum haemorrhage in
management. Br J Obstet Gynaecol 2000;107: a primary health centre in Guinea–Bissau:
1104–10 randomised double blind clinical trial. Br Med J
101. Ng PS, Chan AS, Sin WK, et al. A multicentre 2005:331:723–8
randomized controlled trial of oral misoprostol 113. Oboro VO, Tabowei TO, Bosah JO. Intra-
and 1M syntometrine in the management of uterine misoprostol for refractory postpartum
125
147
30 August 2006 14:20:25
hemorrhage. Int J Gynaecol Obstet 2003;80: hysteroscopy: a randomized controlled trial (1).
67–8 Obstet Gynecol 2000;96:890–4
114. Ngai S, Chan YM, Liu KL, Ho PC. Oral miso- 117. Mendilcioglu I, Zorlu, CG, Simsek M. Suc-
prostol for cervical priming in non-pregnant cessful termination of cervical pregnancy with
women. Hum Reprod 1997;12:2373–5 misoprostol. Eur J Obstet Gynecol Reprod Biol
115. Preutthipan S, Herabutya Y. A randomized 2003;106:96
controlled trial of vaginal misoprostol for 118. New US Food and Drug Administration
cervical priming before hysteroscopy. Obstet Labeling on Cytotec (Misoprostol) Use and
Gynecol 1999;94:427–30 Pregnancy. ACOG Committee Opinion 283.
116. Preutthipan S, Herabutya Y. Vaginal miso- Washington, DC: American College of Obste-
prostol for cervical priming before operative tricians and Gynecologists, 2003
126
148
30 August 2006 14:20:25
13
LABOR WARD DRILLS
M. Tipples and S. Paterson Brown
INTRODUCTION GENERAL PRINCIPLES OF ADULT

EDUCATION
As the leading cause of maternal mortality
world-wide and a major contributor to maternal Adult learning
morbidity, massive obstetric hemorrhage
Before embarking, it is worth reflecting on how
deserves center stage in the training of mid-
adults learn and appreciating that they are not
wifery and obstetric staff. That this need for
satisfied with facts alone, but also like to under-
training is global is highlighted by the recent
stand and be able to apply the knowledge they
increase in deaths due to obstetric hemorrhage
acquire. Three different processes are involved
in the UK (CEMACH1). Although much
in learning, all of which can be complementary
knowledge is gained at the bedside, practical
and are featured in practical teaching sessions.
teaching with a structured approach to this
unique life-threatening emergency provides a
sense of security and preparedness that cannot Visual
be obtained by reading or attending lectures.
Several well-established courses focus on This includes the learning we do through
practical emergency teaching, and further reading, but also includes what is assimilated
information is available through their through watching a person or people doing
websites: ALSO at www.also.org.uk, MOET something practical. Being able to picture the
at www.alsg.org.uk, and MOSES at www. scene and actions that were taken enables one
bartsandthelondon.org.uk/simulationcentre/co to easily recall them when a similar situation
urses.asp. These courses present a structured presents itself.
approach to resuscitation with skills, drills and
scenarios taught and applied to the seriously ill Auditory
patient. As good as such courses may be, they
cannot begin to train everyone in all things This includes learning through listening,
and there remains a need for strong local but also includes dialogue, questions and
supplementation. discussion.
All functioning obstetric units should possess
a multidisciplinary massive hemorrhage proto-
Kinesthetic
col, which should be updated and rehearsed
regularly. Running these sessions as a local drill This involves learning through doing and
helps to test such systems in place to deal with includes both hands-on practice and role play.
obstetric hemorrhage and makes them particu- Hands-on practice is especially useful for practi-
larly useful to local staff. Clinical scenario and cal skills, whereas role play encourages the
skills training add detail and depth to this train- learner to work logically through a sequence of
ing, but efficiency in the system is an essential events in a clinical scenario.
prerequisite to effective care. This chapter des- All three forms of learning are variably suited
cribes how various practical training techniques to different things. For example, learning to tie
(drills, skills and scenarios) work and how such a knot can be visualized and explained, but one
programs are set up locally. needs to do it to finally realize the skill. Of
127
149
30 August 2006 14:20:25
importance, different individuals tend to gain elective surgery to facilitate training at a given
more from one approach compared to another: time, remembering, of course, that labor ward
some prefer watching what is going on, others workloads are unpredictable and a back-up
benefit most from open discussion and feed- teaching location needs to be available (for
back, and still others relish the challenge of example, a seminar room or antenatal
being the doers in the practical teaching demon- classroom).
stration. Appreciating these differences and
staying sensitive to the particular needs of those
Setting the tone
being taught helps keep practical teaching fun
and effective while, at the same time, avoiding The instructor should give a general explanation
what can be extremely stressful for some at the beginning of the teaching session. This
individuals. will establish the mood and motivate the learn-
ers by outlining the usefulness of the content.
For example ‘Obstetric hemorrhage is the lead-
Practical teaching
ing cause of maternal death globally, and today
The same preparations should be made whether we are going to run through a simulated case
teaching skills, drills or scenarios. of placental abruption. The aim is for you to
consolidate and apply your knowledge of this,
which should assist you when you face a similar
Knowledge
situation in a real emergency’. At this stage, it
A sound knowledge base is required before may be useful to place the clinical problem in
practical teaching can be undertaken success- the context of recent local events.
fully. An initial lecture/workshop/discussion The specific objectives of the session should
should be organized if staff are unfamiliar with then be explained together with what is
practical teaching or if new material is to be expected of everyone in terms of who is going
taught, as this allows staff to prepare them- to do what, and whether questions can be
selves. It also helps reinforce the idea that asked throughout, or be kept till the end. It is
practical teaching is an opportunity to put extremely useful to allow questioning through-
what one knows into practice. out, as many people will forget if asked to wait
till the end, but it can spoil the momentum of
the scenario and role play. This must be judged
Environment
anew in each session.
A suitable location should be found that is con-
ducive to the teaching that has been planned.
Dialogue
The layout of the room should allow those
involved in the scenario to access the patient The actual ‘doing’ in practical teaching and role
and those watching to see clearly. Heating and play works through the simulation starting
ventilation should be considered, but acoustics from very specific instructions. Progress can
are vital and can sometimes conflict (e.g. noise vary according to what the learner does, and the
from an open window). When teaching about instructor needs to stay alert and flexible in
obstetric hemorrhage, a delivery room or an order to remain in control, to cover all intended
operating theater makes for a very realistic envi- teaching points and to guide the session to an
ronment, but it occasionally conflicts with clini- appropriate conclusion.
cal needs. To try to avoid this, one can plan
impromptu teaching when the delivery suite is
Feedback
quiet. Such ‘unannounced’ teaching is good for
testing how the systems are working (i.e. drills), This is sometimes known as critique and is an
but, as it does not allow planning in terms of essential part of the learning process as it pro-
who or how many people can be taught, it may motes retention of important points. One form
be less useful when running clinical scenarios. of systematic feedback, described by Pendleton
Another alternative is to consider reducing and known as Pendleton’s rules, comprises four
128
150
30 August 2006 14:20:26
Labor ward drills
stages: the learner says what she/he did well, knowledge, the features and differences of
then what she/he could improve upon; this is which are summarized in Table 1, together with
followed by the trainer saying what the learner examples of suitable teaching material.
did well and then what can be improved upon.
Allowing the learner to comment first gives the
Drills
instructor an opportunity to assess the candi-
date’s insight into her or his own ability and These are practice or ‘dummy’ runs and are
behavior. The instructor then has the opportu- comparable to fire practices in testing the local
nity to highlight both good practice and areas systems. Running a drill not only allows local
for improvement not already covered by the scrutiny (i.e. what actually happens when the
learner in order to stress and reinforce learning alarm is put out), but also can be a very effective
points to all present. test of local arrangements and services and of
staff knowledge of them.
Closure
Preparations for a drill
Bearing in mind that adults need to understand
before they change behavior, it is crucial that When running drills, the staff that are going to
questions and discussion be encouraged. A be involved should be faced with the drill in a
summary of the key learning points from the normal clinical area, unprepared, in order to
session should then be given, so that everyone receive a realistic idea of what would happen in
leaves with a clear message of the most a true situation. Clearly, the drill should not
important issues. conflict with patient care and therefore the tim-
ing will depend to some extent on existing work-
load. The lead clinician for the teaching session
DRILLS, SKILLS AND SCENARIOS
should, however, have informed the lead mid-
These three styles of teaching differ in their wife and, in the case of an obstetric hemor-
aims. All require and test different skills and rhage, the transfusion hematologist and other
Table 1 Key features and differences in skills, drills and scenario teaching
Skill Drill Scenario

Definition Acquisition of a skill A chain of events in Improvised clinical role
response to a problem play
Aim of the teaching Ensure correct technique Test the local emergency Apply and practice clinical
system care in a improvised set-up
Teaching environment Seminar room Throughout hospital in Seminar room, operating
day-to-day environment theater or delivery room
Examples of things Brace suture Response to the APH
suitable for teaching Rusch balloon emergency massive – abruption
and testing in relation Aortocaval compression obstetric hemorrhage call – placental previa
to obstetric hemorrhage CPR PPH
– atony
– trauma
– RPOC
Skill mix Doctors and midwives All delivery suite staff Multidisciplinary:
and laboratory staff, obstetricians, midwives,
hematologists and porters anesthetists, pediatricians
CPR, cardiopulmonary resuscitation; APH, antepartum hemorrhage; PPH, postpartum hemorrhage;

RPOC, retained products of conception
129
151
30 August 2006 14:20:26
necessary individuals, such as transportation ● When does the anesthetist/consultant/

staff. This is not only as a matter of courtesy but hematologist arrive?
also to plan timings in order to avoid clashes
Such analyses can help to illustrate system fail-
of interests. The transfusion hematologist may
ures and modify local policy. The identification
prepare spare serum for grouping and make
of problems stimulates and informs guideline
empty blood bags available for the ‘dummy
development. Clarifying the roles of different
run’.
staff and streamlining activity can also improve
future responses and improve care. Such devel-
Running the drill opments can be monitored at future drills and
improvements in the system should be fed back
Table 2 illustrates an example of an assessment
to staff.
sheet for a massive obstetric hemorrhage drill,
Having run drills for obstetric hemorrhage
suggesting things that can usefully be moni-
at Queen Charlotte’s and Chelsea Hospital
tored; these include the following:
for many years, the following are examples of
● Who responds to the initial emergency problems identified and system changes made
buzzer? in response.
● Is the appropriate emergency call put out? Communication As identified in numerous
confidential enquiries, problems in communica-
● How effective is the emergency bleeping
tion often hamper emergency responses. We
system?
found that we struggled with instructions
● Is transportation alerted and does she/he between clinicians and blood transfusion staff
respond? regarding what was needed when: could we wait
for group-compatible blood or even cross-
● Do transfusion staff receive any communica-
matched blood? How long could we wait to
tion?
have blood at the bedside? What clotting
● How quickly does blood arrive at the bed- products did we need when? These are some
side? examples of questions often not clarified over
the ‘phone’. It became obvious that this job was
● How quickly is the patient transferred to the
being delegated to someone very junior on the
operating theater?
delivery suite and misunderstandings were com-
mon. As solutions we, first, installed a red
phone in the operating theater that linked
Table 2 Example of an assessment sheet for mas-
exclusively with a red phone in the transfusion
sive obstetric hemorrhage drill. This assessment
sheet can be expanded to include the response times
laboratory. This enabled blood requirements to
for individual doctors, and their reactions and be discussed by the anesthetist directly with
actions transfusion staff without them having to leave
the patient to go to a phone outside the theater.
● Time emergency buzzer pulled Second, we identified time limits for transfusing
● Staff responding to the initial buzzer blood at the bedside (for example, ‘we need 4
● Time switchboard received emergency call
units of blood within 30 minutes’), rather than
● Staff responding to the emergency bleep
discussing whether we could wait for blood to
● Initial treatment of ABC (airway, breathing and
circulation) resuscitation instituted quickly and be cross-matched or not. This left the labora-
effectively tory in no doubt of the clinical needs and has
● Time transportation person arrives in blood minimized delay in blood arriving at the bedside
transfusion when needed.
● Time blood samples received in the laboratory
The role of transportation The transportation
● Time appropriate blood arrives at patient’s
bedside person initially arrived in the delivery suite when
● Time patient transferred to the operating theater a hemorrhage call was put out in order to take
blood samples to the laboratory for grouping/
130
152
30 August 2006 14:20:26
Labor ward drills
cross-matching, but this was deemed inefficient with the tendency for the Brace suture to slip off
and delayed blood being brought to the bedside the ‘shoulders’ while pulling it tight).
in the most urgent cases. As solutions, we, first,
changed the process so that the transportation
Running the skills teaching
person now goes straight to the laboratory in
readiness for the most urgent of needs in col- This teaching process is best done in four steps:
lecting O-negative blood. Second, the installa-
Step 1 The instructor demonstrates the skill in
tion of a chute for samples to be sent to the
silence. The skill is performed at normal speed
laboratory has also helped in this context. If the
so that the candidates appreciate their ultimate
clinical condition of the patient can wait for
aim.
group-compatible blood, the transportation
person stays in the transfusion laboratory until Step 2 The instructor then demonstrates the
the sample has arrived by chute and has been skill slowly with a commentary. Providing the
grouped, ultimately bringing the appropriate commentary and breaking the technique down
blood to the delivery suite. add understanding to the process and can
highlight points of caution and safety as well
as adding helpful hints.
Skills
Step 3 The learner provides the commentary,
The teaching of practical skills can be useful
which the instructor follows while demonstrat-
in obstetric hemorrhage teaching sessions. The
ing the skill for the third time. The instructor
need for specific teaching often becomes appar-
must be careful not to assume knowledge dur-
ent during the discussion and questioning when
ing this process and stop in mid-flow if errors
running a scenario. Things may have been men-
are made. This step is crucial in terms of surgi-
tioned which are not fully understood, and such
cal safety, as the instructor can tell what the
circumstances illustrate how important it is for
learner understands. Any errors or omissions
scenario teaching to be constructive. Staff must
can be addressed immediately (this step may
feel able to question about what something is or
need to be repeated).
how it is done.
In obstetric hemorrhage, the following skills Step 4 Once step 3 is completed satisfactorily,
may be needed: the learner is allowed to perform the skill
while providing a commentary under direct
● Medical skills
supervision.
– bimanual uterine compression
– aortic compression
– cardiopulmonary resuscitation Scenario teaching
● Surgical skills These practical teaching sessions describe a
– insertion of an inflatable uterine balloon clinical picture and facilitate role play to manage
– insertion of a Brace suture the problem. The aim of such teaching is
– intravenous cut-down for venous access to demonstrate appropriate clinical behavior.
This includes clinical knowledge and how it is
applied, but also how individuals work together
Preparation for skills teaching
as a team and communicate. Such interactions
When teaching any practical skill that may be can be complex and are worth detailing
required in an emergency, it should be done further before illustrating massive hemorrhage
slowly and calmly, giving ample time for scenarios.
reflection, questions and practice. The use of
manikins and surgical aids works well, but one
Teamwork
must remember to point out the differences to
be expected when working in vivo (such as the The ability to work together as a team is funda-
need to keep an inflatable uterine balloon well mental to good clinical care. Individuals possess
into the cavity while inflating it, or how to deal differing expertise and the group’s dynamics
131
153
30 August 2006 14:20:26
and ability to carry out specific tasks depend The patient Depending on the subject to be
upon the interpersonal skills of all team mem- taught, either a manikin or a live person can
bers. Watching a group working together can be used. Manikins tend to be good for collapse
highlight problems and help focus remedial and cardiopulmonary resuscitation, whereas live
action in terms of teamwork and occasionally models are better when responses are needed
individual behavior. (for example, eclampsia). Either can suit mas-
Every team needs a leader, and deciding who sive obstetric hemorrhage. The advantage of
the leader is to be can sometimes be difficult. It a live model is that everyone usually learns
is important to recognize that the team leader a great deal with regard to how all levels of
need not be the most senior person and, as the staff communicate with a patient in such
scenario develops, sometimes the leader will emergencies
need to change. In any event, the leader should
have appropriate knowledge and skills, be a The equipment It helps to be more realistic if
good communicator and motivator, be able to there is some equipment available when running
maintain situation awareness (see the whole pic- clinical scenarios. This should be kept simple,
ture) and distribute the workload. At the same but using it helps to illustrate what important
time, watching staff adapt to each other can be features have been dealt with (e.g. lateral tilt
hugely instructive, and discussing these issues and oxygen) and what omissions have occurred
afterwards can help them understand each (e.g. intravenous access or urinary catheter).
other, as well as individual needs and stresses. Table 3 is a suggested minimum equipment list
for a massive hemorrhage scenario.
Communication
Running the scenario
The process of asking for and providing infor-
mation, and of listening to what other people Who should be involved? Deciding who should
are trying to say, should be simple. It clearly is be involved in the role play and who is better left
not and is repeatedly raised as a problem area in to watch quietly can be difficult. If the members
confidential mortality reports. In the Confiden- of staff are new to scenario teaching, it is best,
tial Enquiry Report of 1997–19992, the greatest initially, to ask for volunteers. Lack of volun-
cause of substandard care in maternal deaths teers may be due to simple factors like being
was failure of communication and team working shy, but it may result from fear of ignorance
between professionals. When running practical
teaching sessions, communication within the
team can be witnessed and discussed after- Table 3 Basic equipment list for practical obstetric
wards. Generally speaking, when dealing with hemorrhage training
any emergency, single precise commands
Airway and breathing
should be addressed to specific individuals.
● Guedel airway
Voices should not be raised and an air of calm ● Oxygen mask with bag and tubing
control should be apparent. Unfortunately, ● Stethoscope
some individuals tend to become overexcited, Circulatory

and noise levels can build up, which can affect ● Wedge (to provide lateral tilt for the pelvis)
everyone’s behavior, as well as making it diffi- ● Tape
cult to hear what is being said without resorting ● Two large-bore intravenous cannulae (14FG)
to shouting. Pointing out such behavioral fea- ● 20-ml syringe
● Blood bottles for full blood count (FBC),

tures under stress during mock emergencies can
only help to raise awareness. cross-match (XM), clotting studies
● 2-liter bags of crystalloid run through giving sets
● Catheter
Preparing for scenario teaching Specific equipment for massive obstetric

hemorrhage
When preparing for role-play, it is important to ● Intrauterine inflatable balloon and bladder syringe
try to make things as realistic as possible.
132
154
30 August 2006 14:20:26
Labor ward drills
being exposed or raising issues of competency. the scenario can progress and more complex
It is for this reason that didactic teaching is features can be added.
needed prior to running scenario training, so Prompting can be difficult if it is to be done
that the theoretical material has already been sensitively without demoralizing or embarrass-
covered. Those previously unsure of the theory ing the learner, and is the real skill in making
behind the problem can build on their newly this form of teaching constructive. The exam-
acquired knowledge in a practical way. Indeed, ples that follow may be useful in the massive
once members of staff become used to this hemorrhage situation:
method of teaching, more will come forward.
● Lateral tilt can be forgotten in the pregnant
Occasionally, someone may need to be invited
woman and a prompt asking whether there
to join in, but this should be done sensitively
is ‘anything else that could improve the circu-
and with support.
lation?’ may jog a response;
Give people defined roles People need to be ● If there has been no apparent registering of,
given a defined role and told what they can or or response to, observations, information
cannot expect in terms of back-up. For exam- about tachycardia or hypotension can be
ple, ‘You are the senior house officer who has repeated;
just answered the emergency buzzer to this
multiparous patient. She has just bled briskly
● Comment that uncross-matched blood is
following spontaneous vaginal delivery. The now available if staff have lost their train of
midwife is here, but all other staff are busy with thought and had already mentioned they
an emergency in theater and you should not would request blood but then forgotten
expect help for at least 10 minutes. Please carry about it;
on as you would in real life. I will give you any ● Providing the patient’s physiological res-
observations you request’. ponses can intervene to slow down/speed up
the action as required. For example, once
Keeping the scenario going The patient can be intravenous fluids have been commenced,
primed to give certain responses, and monitors inform the candidate that the blood pressure
can be prepared with readings (cardiotocograph is improving but that the bleeding is still brisk
paper sticking out of a machine/blood pressure vaginally. This will encourage the candidate
recordings on a monitor, etc.), but it is the to move on to assess the cause for this;
instructor’s role to keep the scenario flowing
and give as much or as little information as is ● If the candidate moves away from the intra-
requested. The scenario needs to progress, how- venous access without taking any bloods for
ever, and gentle encouragement and occasional laboratory investigation, the instructor may
subtle prompts can assist the learner in achiev- slow things down by asking if she/he would
ing the key treatment points. The aim of run- do anything else before moving on to assess
ning scenarios is not to display ignorance, but to the cause of the bleeding. The candidate
empower individuals to apply their knowledge could also be prompted with an empty
in a logical and timely manner. Depending on syringe and blood bottles, if necessary, to
the performance and ability of the candidate, make a teaching point.
the scenario can be resolved early or become
more complex. This needs to have been antici- Drawing things to a logical conclusion When the
pated by the instructor well in advance. If the scenario has run its course, all people who have
candidate is becoming stressed, but has done been involved in the role play should be con-
all the basic key treatment points, then the gratulated and thanked for their participation,
scenario can resolve and the candidate can be and then encouraged to engage in the feedback
congratulated. If the key treatment points have process as described above. Questions and dis-
not been achieved, then help can be at hand in cussion should then be encouraged before clo-
the form of a registrar or consultant arriving to sure, with particular emphasis given to the key
help. If the learner is doing a fantastic job, then treatment points.
133
155
30 August 2006 14:20:26
Examples of possible massive obstetric hem- emergency situation. It is best kept simple
orrhage scenarios are given, together with their and, because it can be stressful to those
key treatment points in Tables 4 and 5. involved in role play, it must be introduced
sensitively and conducted within an
encouraging atmosphere. Staff need to know
SUMMARY
what style of teaching will be used, and what
Setting up practical teaching locally improves its aims are. Advertising the planned content
local processes, builds on teamwork, aids of the session in advance will encourage staff
with communication, and improves clinical to prepare and capitalize on enthusiasm and
knowledge and its application in the learning.
Table 4 Sample scenario for postpartum hemorrhage due to atonic uterus
Candidate information
A 34-year-old grand multipara delivered a healthy baby boy weighing 4.00 kg 40 minutes ago. She had
physiological management of her third stage and the placenta was delivered 10 minutes ago. The midwife
has noticed some fresh brisk vaginal bleeding and accosted you as you were walking past the delivery room.
Initial observations
Talking but very pale. Pulse 110/min, blood pressure 120/80 mmHg; large volume of blood on floor and
bed.
Please proceed as you would in real life together with the midwife who called you. I will give you any
observations you request. (The candidate can be obstetric or midwifery as either should be able to manage
this emergency initially. If further progress to theater is needed, more senior help can arrive as requested.)
Instructor’s notes/Key treatment points Achieved
● Call for help and initiate the massive obstetric hemorrhage drill
● Recognize that this is a circulatory problem: progress rapidly through airway and breathing and
attach face mask for oxygen
● Establish intravenous access
● Send blood for full blood count, cross-match, coagulation, and U&Es
● Commence intravenous fluids
● Do clinical examination and diagnose uterine atony
● Give uterine massage
● Administer a uterotonic agent
● Do a vaginal examination and evacuate clots
● Check no obvious vaginal lacerations
● Do bimanual uterine compression
● Go through drugs cascade logically and give intravenous fluids and blood appropriately
● Consider examination under anesthetic if patient fails to respond and consider other causes of
postpartum hemorrhage
● Knowledge of surgical techniques to control hemorrhage, i.e. Rüsch balloon, Brace suture, etc.
134
156
30 August 2006 14:20:27
Labor ward drills
Table 5 Sample scenario for postpartum hemorrhage not due to atony
Candidate information
A 24-year-old primipara is induced at 42 weeks’ gestation. She is having intermittent abdominal pain when
the prostaglandin is inserted. Within 1 hour, she is transferred to the delivery suite where she delivers a baby
boy weighing 3.8 kg followed by the placenta.
Initial observations
Talking; pulse is 100/min, blood pressure 115/70 mmHg; steady trickle of blood vaginally.
Please proceed as you would in real life and I will give you any observations you request.
(This scenario is more complex – a precipitate labor with the possibility of a concealed abruption. The focus
will be on distinguishing between genital tract trauma and disseminated intravascular coagulation (DIC).
How this scenario will unfold will depend on the learner’s experience and ability.)
Instructor’s notes/Key treatment points Achieved
● Call for help and institute massive hemorrhage call
● Recognize circulatory problem. Move swiftly through airway and breathing. Administer face
mask for oxygen
● Insert intravenous access
● Send blood for full blood count, group and save, and coagulation screen
● Commence intravenous fluids
● Abdominal examination to confirm uterus well contracted
● Vaginal examination to check for vaginal lacerations
● Transfer to theater for analgesia and examination
● Catheterize
● Full EUA: check vagina, cervix and uterine cavity
EUA, examination under anesthetic
References Firth-Cozens J. Teams, culture and managing risk.

In Vincent C, ed. Clinical Risk Management:
1. Lewis G, ed. and Confidential Enquiry into Enhancing Patient Safety. London: BMJ Books,
Maternal and Child Health (CEMACH). Why 2001: 355–68
Mothers Die 2000–2002 – The Sixth Report of Confi- Gaba DM, Fish KJ, Howard SK. Crisis Management
dential Enquiries into Maternal Deaths in the United in Anesthesiology. New York: Churchill Living-
Kingdom. London: RCOG Press stone, 1994
2. Why Mothers Die 1997–1999: The Confidential Glavin R, Maran N. Simulation and non-technical
Enquiry into Maternal Deaths in the UK. London: skills. In Greaves JD, Dodds C, Kumar C, Mets
RCOG Press, 2001 B, eds. Clinical Teaching: a Guide to Teaching Prac-
tical Anaesthesia. Lisse: Swets & Seitlinger, 2003:
219–29
Further reading
Helmreich RL, Schaefer HG. Team performance in
Mackway-Jones K, Walker M. Pocket Guide to Teach- the operating room. In Bogner MS, ed. Human
ing for Medical Instructors. London: BMJ Books, Error in Medicine. Hillsdale, New Jersey: Lawrence
1999 Erlbaum Associates, 1994:225–53
135
157
30 August 2006 14:20:27
14
NON-PNEUMATIC ANTI-SHOCK GARMENT
S. Miller and P. Hensleigh
INTRODUCTION facilities or home to the CEOC facilities for

definitive therapies.
The FIGO/ICM recommendations for active
The NASG is a lightweight, relatively
management of third-stage labor, including
inexpensive (US$160.00 per garment), wash-
uterotonic prophylaxis with additional utero-
able and reusable (at least 50 times), neoprene
tonic treatment when necessary, reduce the
garment that resembles the lower part of a wet
incidence of severe postpartum hemorrhage due
suit. The NASG, manufactured by the Zoex
to atony. However, at least 1% of all women still
Company, received a United States Food and
suffer intractable postpartum hemorrhage from
Drug Administration (FDA) 510(k) medical
uterine atony or other obstetric causes, such as
device regulations number (FDA device #
genital lacerations, ruptured uterus, ruptured
K904267/A, Regulatory Class: II, January 17,
ectopic pregnancies, or placenta previa, accreta,
1991) (Section 510(k) Medical Device Amend-
and abruption. Multiple blood transfusions are
ment, FDA, Office of Device Evaluation, 1991)
often needed to resuscitate and stabilize these
and can be exported to countries outside the
women and hemostasis may require surgical
United States. The NASG is designed in hori-
interventions. Until the time when quality com-
zontal segments, with three segments on each
prehensive emergency obstetric care (CEOC),
leg, a segment over the pelvis, and a segment
including surgery and/or blood transfusions, is
over the abdomen that contains a small, foam
readily available for all women, strategies and
compression ball (see schematic in Figure 1).
technologies for hemorrhage treatment and
Unlike the pneumatic anti-shock trousers
hypovolemic shock resuscitation are needed
(PAST), or medical anti-shock trousers
such that they can be readily provided and easily
(MAST), which preceded the NASG, there are
applied, even by persons with no medical train-
no pumps, tubing, or gauges to add complexity
ing. Promising technologies to reduce maternal
and risk of malfunction. Using the three-way
mortality include the non-pneumatic anti-shock
elasticity of neoprene and the tight grip of
garment (NASG) as a technology for reducing
the Velcro fasteners, the garment can apply
the mortality and morbidity associated with
30–40 mmHg of circumferential counter pres-
obstetric hemorrhage1,2.
sure to the lower body from the ankles up to the
level of the diaphragm (see Figure 2). This
amount of pressure is effective in reversing
hypovolemic shock by shunting blood from the
THE NON-PNEUMATIC ANTI-SHOCK
capacitance veins of the abdomen and lower
GARMENT
extremities to the vital core organs, the heart,
The NASG is a simple device, proposed as the lungs and brain. The effect of this autotrans-
immediate first-aid treatment for reversing fusion, estimated to be 500–1500 ml, is almost
hypovolemic shock and decreasing blood immediate upon application. The moderate
loss secondary to obstetric hemorrhage, by NASG pressures are generated by manually
application of lower body counter pressure. The stretching the neoprene material. Excess pres-
NASG also has the potential to keep women sures and the resultant tissue ischemia, known
alive during long transports from lower level complications with the pneumatic PAST, are
136
158
30 August 2006 14:20:27
Non-pneumatic anti-shock garment
Figure 1 Schematic diagram of the non-pneumatic anti-shock garment
Figure 2 Patient wearing the non-pneumatic anti-shock garment in hospital
137
159
30 August 2006 14:20:30
not an issue. Application of the garment intractable hemorrhage in the 1998 guidelines
requires about 2 min and, within 2–5 min after of the American College of Obstetricians and
its application, most patients with severe shock Gynecologists8.
regain consciousness and vital signs begin The PASG requires inflation and careful
to recover. With the bleeding slowed and management of pressure levels, both to main-
the blood the pressure restored, panic levels tain adequate pressure and to prevent over-
decrease, and there is time to deliberately assess inflation resulting in compartment syndromes,
the situation. Patients can remain in stable ischemia, and necrosis. The valves and manom-
condition for hours while blood transfusions eters that maintain inflation are subject to
are initiated and arrangements made for leaks and malfunctions. In addition, specialized
surgery or other required therapies. The patient training for safe and effective use of the MAST
and her family can be given emotional support is necessary, which makes widespread use in
and prepared for transport to a referral-level developing countries difficult.
facility or any required surgical procedures.
If the hemorrhage is due to uterine atony, the
EFFICACY OF THE PASG FOR
continued infusion of uterotonics and passage
TRAUMA
of time may be all that is required for
management. McSwain reviewed the physical principles
responsible for the physiologic effects of lower-
body counter pressure with application of the
DEVELOPMENT OF THE NASG
PASG3. Numerous studies in hypovolemic
The modern NASG is a non-pneumatic refine- animals and humans demonstrate that the
ment of the pneumatic anti-shock garment PASG increases blood pressure by decreasing
(PASG). The PASG was adapted from a device the vascular volume and increasing vascular
developed by George Crile, in the early 1900s resistance within the compressed region of the
before the advent of technologies to allow blood body. When blood flow or arterial vascular size
transfusions. Crile, a surgeon who wrote text- is measured above the device, the flow is greater
books on blood pressure and shock, designed and vessels are larger. In the compressed region,
the first inflatable pressure suit to maintain the radius of blood vessels is decreased, thus
blood pressure during surgery. This pneumatic slowing down blood flow. In the hypovolemic
suit underwent multiple modifications and was model, the PASG increases venous return and
further refined for use as an anti-gravity suit the increase in preload is associated with
(G-suit) for the Army Air Corps in 1942. Dur- increased cardiac output9.
ing the Vietnam War, the G-suit was modified These changes in vessels and blood flow are
for resuscitating and stabilizing soldiers with also associated with the translocation of blood
traumatic injuries before and during transport. from the lower body enclosed in the device to
The G-suit was then modified to a half-suit, the upper body (heart, lungs and brain), a vol-
which became known as military anti-shock ume which is estimated at 750–1000 ml10. In
trousers (MAST). summary, the physiologic bases for restoration
The MAST/PASG has been used since the of blood pressure are: the reduced vessel size
1970s by emergency rescue squads in the beneath the device, which produces increased
United States to stabilize patients with a variety systemic vascular resistance, the decreased con-
of disorders: pelvic and lower limb fractures, tainer size, the translocation of blood to the
hypovolemic shock, septic shock, and to control upper body, and increased preload resulting in
intra-abdominal, pelvic, and thigh hemorrhage, increased cardiac output.
as well as for gynecological and obstetric hemor- Although animal studies show improved
rhage3–6. Andrea and colleagues used the survival rates except for thoracic injuries11,12, it
MAST to stabilize women with intractable is uncertain if the rapid, positive changes in vital
uterine bleeding while preparations were made signs and blood loss affect survival in humans.
for transcatheter emobization7. The PASG is Despite widespread acceptance of the PASG for
included as a recommended treatment for military and civilian trauma injury and reports
138
160
30 August 2006 14:20:31
of small series demonstrating successful treat- What is unknown about the PASG is what
ment of various bleeding conditions in adults the difference in outcomes might be if the
and children, there are no definitive prospective PASG (or the non-pneumatic adaptation, the
randomized treatment trials to show improve- NASG) were to be used in low-resource settings
ment in survival. The three United States- where there are longer transport times, slower
based, prospective, alternate-day, randomized responses at the hospital level, and longer
treatment trials of civilian trauma cases did not delays in obtaining definitive therapy by blood
find consistent results and are confounded by transfusions and/or surgery.
inclusion of injuries to the upper body, a contra-
indication to use of the PASG13–15. In some
POTENTIAL BENEFITS OF THE NASG
studies, the pH on hospital admission was lower
IN LOW-RESOURCE SETTINGS
with PASG use, intensive care unit and hospital
stays were longer, and survival worse. These The NASG was adapted from the PASG by the
studies were all conducted in a metropolitan National Aeronautics and Space Administration
setting where high-level hospital care was avail- (NASA) in 1971, when the NASA/Ames
able within minutes, and even the brief delay in Research Center developed a prototype pres-
applying the PASG may have been a detriment sure suit designed to protect hemophiliac chil-
to the benefits of early hospital care. dren from bleeding into elbow and knee joints
Despite the lack of randomized, controlled by straightening and compressing the joint until
trials supporting its use, the Committee on medical attention was available21. Both PASG
Trauma of the American College of Surgeons and NASG provide circumferential counter
included the PASG as essential equipment for pressure in the lower body, but the NASG
ambulances16 and the PASG remains on the is simpler in design, more quickly and easily
curriculum and in the textbooks for emergency applied, less expensive and avoids the risk of
medical technicians in the United States17,18. A over-inflation and excessive pressure22.
position paper on the PASG by the National The NASG is particularly suited to use in
Association of EMS Physicians19 cited the lack low-resource settings. Lighter and more flexible
of controlled trials, but, on the basis of other than the PASG, it is more comfortable for a
reports, deemed the PASG as ‘Class I usually woman to be inside the suit for longer periods of
indicated and effective for hypotension due to time, something necessary in the long transport
ruptured aortic aneurysm, but of uncertain effi- times and delayed treatment conditions of low-
cacy for other emergency situations’. Its use for resource settings. As with the PASG, within
uncontrolled gynecologic hemorrhage, urologic minutes of being placed in the NASG, a
hemorrhage, and ruptured ectopic pregnancy patient’s vital signs are restored and, if confused
was a ‘Class IIb acceptable, but uncertain effi- or unconscious, their sensorium generally
cacy, may be helpful, probably not harmful’. clears1. Women can remain in the NASG for as
PASG proponents particularly recommend its long as is required to restore their circulatory
use for bleeding in the abdomen, retro- volume with crystalloids and to replace blood.
peritoneum, pelvis, or thighs9. The current In prior reports of cases where blood transfu-
recommendations in the US are that, while its sions were not readily available, this has often
effects on survival are unknown, it is probably required 18–24 h, and, in one case, a woman
indicated in patients with bleeding in the very remained safely in the NASG for 57 h23.
areas (abdomen, retroperitoneum and pelvis) Compared to the PASG, with pressures of
that are the bleeding sites for women with 100 mmHg or more, the NASG only applies
obstetric hemorrhage. Likewise, in France, 30–40 mmHg. Higher pressures appear to be
use of the ‘pantalon antichoc’ is questioned responsible for skin and muscle ischemia and
for widespread use, but its use for post- adverse effects on pH as well as the occasional
partum hemorrhage, disseminated intravascular anterior compartment syndrome.
coagulopathies associated with pregnancy and A second benefit of the NASG for obstetric
labor, and other obstetric and gynecological indications is that the design of the garment per-
bleeding is endorsed20. mits complete perineal access so that genital
139
161
30 August 2006 14:20:31
lacerations can be repaired, speculum or skilled birth and critical care attendance are
bimanual examinations can be performed, and limited or absent.
manual removal of placenta or emptying of the The improvement in maternal morbidity
uterus with manual vacuum aspiration or curet- and mortality can presently only be discussed as
tage can all be accomplished with the NASG potential benefits, because there have been no
in place. Thus, the source of most obstetric definitive trials of the NASG, and there is only
hemorrhages can be located and attended to very limited experience with its use for obstetric
while the garment maintains vital signs. hemorrhage in low-resource settings. The case
A third benefit of the NASG is that it sig- series and pilot studies discussed below indicate
nificantly reduces further blood loss. When the how the NASG functions to decrease blood
NASG is applied, the external circumferential loss, reverse shock, and stabilize women for
counter pressure is distributed evenly through- many hours while awaiting blood transfusions.
out the abdominal cavity and to the outside of As such, use of the NASG might contribute to
the circulatory vessels – tamponading venous decreased maternal mortality and morbidity.
bleeding. In the event of an arterial injury, Experience with transport from lower-level
continued bleeding results from the tension in facilities to referral centers, while theoretically
the wall of the artery keeping the defect open. beneficial, is only anecdotal at this point.
However, the NASG compresses all the intra-
abdominal vessels including the internal iliac
SUGGESTED PROTOCOL FOR USING
and uterine arteries. This compression reduces
THE NASG
the radius of the arteries and reduces the
transmural pressure (the difference between The NASG is recommended for cases of obstet-
the pressure inside the artery and the pressure ric hemorrhage meeting the American College
outside the artery) which, in turn, reduces the of Surgeons’ criteria for Class II hypovolemic
tension in the arterial wall, closing the defect shock: > 750 ml blood loss, pulse > 100 and
and reducing blood loss. Although the mean blood pressure normal or slightly decreased25.
pressure applied by the NASG is only in the The NASG is not recommended for use with a
range of 30–40 mmHg, this low pressure, which viable fetus, for patients with mitral stenosis,
is below arterial pressures, can still stop arterial congestive heart failure, pulmonary hyper-
bleeding when applied externally to the abdo- tension, or in clinical conditions where there
men and the lower extremities24. Because the could be bleeding sites above the level of the
applied pressures could interfere with uterine diaphragm. Availability of the NASG does not
blood flow, the NASG is not recommended for negate the importance of preventive measures
obstetric bleeding when the fetus is still viable, such as the active management of the third stage
such as might be the case with placenta previa or of labor or administration of uterotonics to treat
abruption. Post-delivery, however, or when the uterine atony. The authors recommend cardio-
fetus is not viable or is dead, the NASG can be vascular resuscitation using limited crystalloid
used for any obstetric hemorrhage. infusion with the goal of ‘permissive hypovol-
Another potential benefit for the use of emia’26–30. This means infusing 1000–1500 ml
the NASG for obstetric hemorrhage in low- of saline rapidly followed by a slower rate of
resource settings is that persons with no medical infusion, 150 ml/h, to achieve a mean arterial
background can learn to apply the garment pressure of about 60 mmHg (blood pressure
safely with minimal training. Such training, 80–50 mmHg) and urine output of 30 ml/h.
which includes hands-on practice in application Supplemental oxygen should be given until the
and removal of the NASG, takes approximately patient is resuscitated, the hemorrhage arrested,
1 h. Once the garment has been properly and the circulation fully normalized.
applied, patients can safely be transported
and/or await definitive treatment in a more
Application of the NASG
stable physical condition. This final point is
critical, as the majority of maternal deaths due The technique for application is for one person
to obstetric hemorrhage occur in areas where to stretch the neoprene panels with all their
140
162
30 August 2006 14:20:31
strength and fasten them with the Velcro as documented in two published reports based on
tightly as possible. The lowest (ankle) segment a series of 150 obstetric cases in one hospital in
is applied first and the abdominal segment Silkot, Pakistan1,23. There was no blood bank in
last. If the woman experiences difficulty breath- the hospital; if blood transfusions were neces-
ing, the abdominal panel should be loosened sary, they could only be obtained through direct
slightly, but not removed. However, if dyspnea donor transfusion. The researchers documented
continues, the NASG should be removed and that patients placed in the NASG experienced
the cause of the respiratory problem evaluated. rapid resuscitation from hypovolemic shock, as
A woman with normal cardiorespiratory func- well as an extended period of stabilization while
tion should experience no problems with venti- awaiting definitive treatment. In a combined
lation. If there is no prompt response in terms of analysis of the two reports24, there appeared to
vital signs with placement of the NASG, the be no adverse effects of prolonging this stabiliza-
application should be checked for adequate tion period, even though the average interval
tightness, and additional saline infusion given from diagnosis of hemorrhage to blood trans-
promptly. As soon as the patient is stable, there fusion was 5 h 30 min, and the mean time in the
must be a diligent evaluation for the specific NASG was more than 30 h.
source and cause of the blood loss.
If pelvic examination or vaginal procedures
Pilot studies of the NASG
are needed, the NASG should be left in
place; if laparotomy is necessary, open the Based on the successes of the case series in
abdominal segment. Often there will be a drop Pakistan, the authors and their in-country col-
in blood pressure when this panel is removed; leagues are conducting NASG pilot studies in
this should respond to additional saline comprehensive emergency obstetric care facili-
infusion. ties in Nigeria (Dr Dosu Ojengbede, University
of Ibadan), teaching facilities in Egypt (John
Snow International), and in primary and sec-
Removal of the NASG ondary health facilities in Mexico (Population
The NASG is left in place as long as needed to Council and IMSS-Opportunidades). These
achieve hemostasis and replace red blood cell studies compare use of a standardized protocol
volume with transfusion of donor blood. The of shock and hemorrhage care in the pre-
NASG can be removed when the hemoglobin intervention period with the same standardized
level is > 7 or the hematocrit 20%, the pulse protocol plus the NASG in the post-
< 100, and the systolic pressure > 100 mmHg. intervention phase. The primary outcome was
Removal of the NASG begins with the lowest volume of measured blood loss after initiation of
segment (#1) and proceeds upwards, allowing treatment with or without the NASG. To obtain
15 min between removing each segment for a relatively objective measure of blood loss,
redistribution of blood. If the blood pressure maternal bleeding after admission to the study
falls by 20 mmHg or the pulse increases by is measured using a specially designed, closed-
20 beats/min after a segment is removed, end, calibrated plastic blood collection drape.
replace the NASG and consider the need for Prior studies of this drape indicate that it
more saline or blood transfusions. If there is is more accurate than visual assessment in
recurrent bleeding, replace the NASG and measuring postpartum blood loss31. Secondary
determine the source of bleeding. outcomes include severe acute maternal
morbidities (SAMMs: acute respiratory distress
syndrome, cardiac deficiency, central nervous
EXPERIENCES WITH THE NASG system damage, and renal failure) and need
for emergency hysterectomy. The standard
Case series in Pakistan
protocol included: active management of third-
Recently, examples of the potential benefits of stage labor, immediate use of uterotonics for
using the NASG in cases of obstetric hemor- suspected postpartum uterine atony, training
rhage in a resource-challenged setting were in limited intravenous crystalloid fluid
141
163
30 August 2006 14:20:31
replacement32,33, and, in the post-intervention, women had a statistically significant greater loss
prompt application of the NASG. of blood on study entry and more severe signs of
Inclusion criteria were obstetric hemorrhage shock than did the no-NASG group. Estimated
with shock (minimum requirements: estimated blood loss at entry to the study was 1000 ml
blood loss > 750 ml, systolic blood pressure for NASG cases and 750 ml for no-NASGs
< 100 mmHg and/or pulse > 100). There were (p < 0.001), mean systolic blood pressure for
a variety of obstetric hemorrhage etiologies, NASG cases was 88.3 mmHg vs. 97.2 mmHg
including postpartum hemorrhage due to atony for no-NASGs (p < 0.001), and mean diastolic
or lacerations, ectopic pregnancies, ruptured blood pressure was 56.7 mmHg for NASGs vs.
uterus, complications of abortion, and a variety 60.8 mmHg for no-NASGs (p = 0.005). A
of placental pathologies. Overall, postpartum probable reason for the NASG group’s worse
uterine atony and/or retained placenta or condition on study entry was that the clinical
placental fragments accounted for 44% of all researchers waited until women were in deeper
participants. Pre-intervention data were col- shock before putting on the garment, a new
lected from May to September 2004. Post- technology that they were not accustomed to
intervention data collection has been completed applying.
in Egypt and is underway in Nigeria and As shown in Table 1, NASG patients
Mexico. had 46% less mean measured blood loss (50%
less median measured blood loss) than did
those patients not treated with the NASG
NASG pilot study in Egypt
(p < 0.001). NASG and no-NASG patients had
The study sites in Egypt comprised high- similar blood loss during surgery: NASG
volume referral CEOC teaching facilities (El patients lost only 32 ml more (p = 0.748), and
Galaa, Alexandria, Assiut, and Al Minya). All NASG patients had a lower amount of esti-
are staffed by senior obstetricians and obstetric mated ‘other’ blood loss (spilled on floor, on
residents with immediate access to banked gauze or towels) compared to no-NASG cases
donor blood and surgery. Pre-intervention data, (p = 0.209). Patients treated with the NASG
including measured blood loss, were collected had a statistically significant lower amount
for 3 months, after which all providers were of post-study entry blood loss (drape +
trained in the use of the NASG. The only intraoperative + ‘other’) compared to no-
change to the pre-intervention clinical manage- NASGs (p < 0.001). The NASG group
ment was the use of the NASG. Post- received 193.3 ml more blood than those
intervention data were collected for another not receiving the NASG (p = 0.034), and the
3 months. NASG group also received 225.6 ml more of
The primary outcome was mean measured intravenous fluids (p = 0.06). There was a non-
amount of blood lost after a woman entered the statistically significant 84% lower incidence of
study. A sample size of 150 pre-intervention severe acute maternal morbidities (SAMMs)
obstetric hemorrhage patients (no-NASGs) and and mortalities, which were combined as
150 post-intervention obstetric hemorrhage ‘extreme adverse outcomes’. One patient
patients (NASGs) was needed to detect a 50% (0.5%) had an extreme adverse outcome among
difference in blood loss between the two groups the NASG patients (renal failure), whereas five
with power (β) of 80% and a significance level patients (3.2%) died or suffered SAMMs
(α) of 5%. The final study sample comprised among the no-NASG patients (odds ratio
156 no-NASG and 204 NASG patients with (OR) 0.15, 95% confidence interval (CI)
obstetric hemorrhage who met the entry crite- 0.02–1.31)34.
ria. Diagnoses of obstetric hemorrhage covered A greater percentage of patients in the NASG
a range of primary diagnoses with no statistically group had surgeries compared to no-NASG
significant differences between no-NASGs and patients (49.0% vs. 37.8%, p < 0.03), perhaps
NASG patients. The three most common diag- due to their worse condition on study entry.
noses were uterine atony, genital lacerations, The most common surgeries were Cesarean
and complications of abortion. The NASG section and salpingectomy. Both groups spent
142
164
30 August 2006 14:20:31
Table 1 Results from NASG Pilot study in Egypt
No-NASG (n = 156) NASG (n = 204)

Mean ± SD Mean ± SD p Value
Mean blood loss measured with drape (ml) 561.6 ± 447.4 303.5 ± 219.5 < 0.001a
(median 500) (median 250)
Mean intraoperative blood loss (ml)c 394.1 ± 537.5 426.0 ± 641.2 0.748b
Estimated ‘other’ blood loss (ml)c 196.3 ± 333 140.2 ± 221.3 0.209 b
Blood loss post-entry to study 902.7 ± 696.5 591.3 ± 532.0 < 0.001a
(drape + intraoperative + other) (ml) (median 700) (median 450)
Volume blood transfusion (ml)d 963.2 ± 668.4 1156.5 ± 707.8 0.034b
Volume intravenous fluids (ml) 2035.5 ± 1190.6 2261.1 ± 1033.3 0.057b
Extreme adverse outcomes 5 women (3.2%) 1 woman (0.5%) OR 0.15, 95%
CI 0.02–1.31
Selected variables: NASG

aMann-Whitney U test used for comparison of groups with unequal variances; bfrom Student’s T test; cfor
the 159 women who had operations, 59 in the no-NASG group and 100 in the NASG group; dfor the 248
women who had blood transfusions, 95 in the no-NASG group and 153 in the NASG group
almost equal time in the hospital following primary outcome of measured mean blood loss
admission, NASG patients a mean of 2.0 ± 1.5 was statistically significantly less. The non-
days, and the no-NASGs a mean of 1.9 ± 1.6 statistically significant difference in extreme
days (p = 0.50). Both groups received compara- adverse outcomes is promising, but will require
ble dose of oxytocin; for those with uterine a larger sample over a longer period of time in
atony, the mean total was 30.7 units for order to demonstrate if there is a true difference
no-NASG and 32.0 for NASG patients. in this most important outcome.
A sub-analysis of women who entered
the study with severe hemorrhage (> 1500 ml)
NASG pilot studies in Nigeria and Mexico
revealed no difference in frequency by study
group, no-NASGs 26.9%, NASG 31.4%, Pilot studies are currently underway at four
p = 0.36. However, the mean volume of blood urban acute obstetric hospitals in Nigeria and
lost in the drape for the NASG group was 66% in rural primary health-care facilities (Unidades
less than for the no-NASGs (278.1 ± 221.5 ml Medicales Rurales or UMRs) with long trans-
NASG vs. 818.1 ± 641.3 ml no-NASG, port times to rural CEOC facilities in Mexico.
p < 0.001). This was a greater mean blood loss The design and methodology of these studies
difference by study group than among all are similar to those in the Egypt study. One dif-
patients (with and without severe hemorrhage ference is that, in Mexico, pre-weighed adult
at study entry). A non-statistically significant diapers are used to collect blood during long
decrease was found for extreme adverse out- transports between the UMRs and the CEOC
comes; among those who did not receive the facilities where women are brought for definitive
NASG, such outcomes occurred in four treatment. (The blood collection drapes used in
patients (9.5%) and in only one NASG patient the CEOC tend to slip off during the often mul-
(4.6%) (OR 0.5, 95% CI 0.1–1.9). tiple vehicle changes involved in transport from
The results among this sample of women in the UMRs.) The used diapers are weighed at
university teaching hospitals appear very prom- the CEOC facilities and blood volume is calcu-
ising. While the women with the NASG had lost lated from the weight of the diaper; at the
more blood and had greater signs of shock, the CEOC facilities, the diapers are then replaced
143
165
30 August 2006 14:20:32
by the closed-end plastic blood collection drape. liter greater in the no-NASG group, 1385 ml vs.
In Mexico, the number of no-NASG cases was 257 ml.
ten and the number of NASG cases 27. An interim analysis from the Nigeria study
The following case history illustrates the concerns the recovery of 33 women with
application of the NASG in a rural Mexican severe postpartum hemorrhage35. This subset of
community. A 25-year-old woman, gravida 2 women met three of the four American College
para 1, delivered at home attended by a tradi- of Surgeons’ Class III or IV shock criteria:
tional birth attendant (TBA) who had been sen- > 1500 ml blood loss, packed cell volume
sitized to the need for transport in the NASG by < 20%, pulse > 120, blood pressure
community outreach. When the woman began < 80/50 mmHg, and altered mental status.
to hemorrhage from a retained placenta, the During resuscitation, the vital signs were
TBA arranged transport to a study UMR. The recorded every 15 min. Although not always
trip took over 2 h; on arrival at the UMR, it was accomplished, the protocol advised giving
estimated that the patient had lost more than 1500 ml saline rapidly, followed by additional
2000 ml blood and had a pulse of 160. The saline to achieve minimum blood pressure of
NASG was applied and the woman transported 80/50 mmHg, corresponding to a mean arterial
by ambulance another 2 h to a CEOC facility. pressure of 60 mmHg.
Blood loss during the trip was measured at Thirty-two women survived without morbid-
300 ml. At the hospital, intravenous fluids were ity; one woman expired. Among the 32 surviv-
started, a manual removal of the retained pla- ing women, the estimated mean blood loss upon
centa was performed, and two units of blood study entry was 1471 ml and mean packed cell
were transfused. The patient was discharged volume 18.9%. Mean measured and estimated
with a hemoglobin level of 8. blood loss with the NASG in place was 220 ml.
In contrast to the community setting in Mean volume of blood transfused was 1442 ml
Mexico, the Nigerian study is set in urban and the discharge packed cell volume 23.8%.
CEOC hospitals. Before introducing the NASG After placing the NASG, 28 of the 32 women
protocol, data were collected for 3 months, as in had improvement in their vital signs by the next
the other pilots. While data are still being col- 15-min recording. The resuscitation interval
lected in this study, an interim analysis includes (the time from placement of the NASG until
27 no-NASG patients and 63 with NASG. achieving a mean arterial pressure ≥ 60 mmHg)
The most common diagnoses are antepartum was 52 min. The patients with more rapid intra-
hemorrhage and postpartum hemorrhage, i.e. venous saline infusion responded more rapidly;
12 women (44%) in the non-NASG group those receiving an infusion at > 1000 ml/min
versus 39 (65%) in the NASG group. The recovered in 24 min while those receiving
estimated blood losses on study entry were simi- < 1000 ml/min took 83 min.
lar for both groups (1258 ml vs. 1487 ml), but
the NASG women had more severe symptoms
Preliminary analysis: Nigeria and Egypt severe
of shock, with a mean blood pressure of
hemorrhage
53/23 mmHg compared to 91/55 mmHg. The
percentage with Class 3 or Class 4 shock in the A recent analysis was conducted combining the
NASG group was also greater, 85% vs. 22%. Egypt data34 and data from six tertiary-care hos-
According to the protocol, measurements of pitals in Nigeria on all women (n = 263) with
blood loss were to be made using the same severe hemorrhage (estimated blood loss on
closed-end blood collection drapes used in the admission ≥ 1000 ml, pulse > 120). There were
Egypt study. However, in some cases, this could 104 in the pre-intervention, standard-care
not be accomplished and visual estimates were group, and 159 treated with standard care and
made of blood loss. In all cases, the amounts of the NASG. There were no differences in the
blood lost and replaced were reconciled with the pre-intervention and NASG groups by country
admission and subsequent hematocrit values. or age. There were a variety of diagnoses; the
The difference in measured plus estimated leading cause of hemorrhage was uterine atony,
blood loss after study entry was more than a approximately 33%. The patients’ conditions
144
166
30 August 2006 14:20:32
on study entry were statistically similar in the lightweight, reusable, relatively inexpensive,
two groups. The difference in median blood loss and can be used at the lowest level of the
in the drape after treatment was statistically sig- health-care system; it has the potential to make
nificantly different; those in the NASG group a great contribution to reducing maternal mor-
lost 64% less, 250 ml vs. 700 ml (median differ- tality and morbidity from obstetric hemorrhage
ence 490 ml, 95% CI: 350–600 ml). Neither and hypovolemic shock if it proves efficacious
mortality alone, n = 7 (6.7%) for pre-interven- in clinical trials or by strong evidence from
tion vs. 3 (1.9%) for the NASG groups, nor multiple quasi-experimental trials.
severe morbidity, n = 5 (5.3%) vs. 2 (1.3%),
were statistically significantly different. How-
ever, a combined severe adverse outcome vari-
able, a combination of mortality and SAMMs, References
was statistically significantly different, with
1. Hensleigh PA. Anti-shock garment provides
those in the NASG group less likely to suffer a resuscitation and haemostasis for obstetric
severe adverse outcome, RR = 0.28, 95% CI haemorrhage. Br J Obstet Gynaecol 2002;109:
0.10–0.7736. 1377–84
2. Tsu VD, Langer A, Aldrich T. Postpartum hem-
orrhage in developing countries: is the public
CONCLUSIONS health community using the right tools? Int J
The data from these pilot trials are promising. Gynaecol Obstet 2004;85(Suppl 1):S42–51
The Pakistan data, while only descriptive, 3. McSwain NE Jr. Pneumatic anti-shock garment:
indicate that women can remain stabilized for state of the art 1988. Ann Emerg Med 1988;17:
506–25
long periods while awaiting blood transfusion, a
4. McSwain NE. PASG in the 90’s. Emergency
situation typical in low-resource settings. The 1994;26:28–33
Egypt study was adequately powered to demon- 5. Pearse CS, Magrina JF, Finley BE. Use of
strate a statistically significant difference in MAST suit in obstetrics and gynecology. Obstet
measured blood loss for women suffering Gynecol Surv 1984;39:416–22
obstetric hemorrhage, with symptoms of 6. Pelligra R, Sandberg EC. Control of intractable
hypovolemic shock treated with the NASG and abdominal bleeding by external counterpressure.
standardized hemorrhage and shock protocol JAMA 1979;241:708–13
compared with women with similar diagnoses 7. Andrae B, Eriksson LG, Skoog G. Anti-shock
and clinical symptoms treated only with stan- trousers (MAST) and transcatheter embolization
dardized hemorrhage protocols. The data from in the management of massive obstetrics hemor-
rhage. A report of two cases. Acta Obstet Gynecol
Mexico are extremely preliminary, but may indi-
Scand 1999;78:740–1
cate the utility of the garment for stabilization
8. American College of Obstetricians and Gynecol-
and transport of hemorrhaging women who ogists. Educational Bulletin #243 1998
deliver outside of facilities and who are not 9. McSwain MJ, McSwain NE. Pneumatic
attended by skilled providers. The preliminary antishock garment: state of the art at the turn of
Nigerian data also indicate promising use in the century. Trauma 2000;2:63–75
facilities to which women with severe shock are 10. Civetta JM, Nussenfeld ST, Nagel EL, Kaplan
brought as a last resort. BH, Rowe TR, Pettijohn F. Reversal of pretreat-
At this time, with only these pilot studies and ment hypotension and control of hemorrhage in
descriptive case series, definitive evidence for trauma patients by a simple device. Am Surg
the use of the NASG for management of 1977;43:20–9
11. Ali J, Duke K. Pneumatic antishock garment
obstetric hemorrhage and hypovolemic shock
decreases hemorrhage and mortality from splenic
is not available. In order to demonstrate that
injury. Can J Surg 1991;34:496–501
the NASG not only decreases blood loss and 12. Gardner WJ. Hemostasis by pneumatic compres-
facilitates resuscitation from shock, but also sion. Am Surg 1969;35:635–7
decreases maternal mortality and morbidity, a 13. Bickell WH, Pepe PE, Bailey ML, Wyatt CH,
much larger trial with a strong experimental Mattox KL. Randomized trial of pneumatic anti-
design is needed. Given that, the NASG is shock garments in the prehospital management
145
167
30 August 2006 14:20:32
of penetrating abdominal injuries. Ann Emerg DV, Mattox KL, eds. Trauma, 5th edn. New
Med 1987;16:653–8 York: McGraw-Hill Med Publ Div, 2004:220–5
14. Mattox KL, Bickell W, Pepe PE, Burch J, 27. Pepe PE, Eckstein M. Reappraising the pre-
Feliciano D. Prospective MAST study in 911 hospital care of the patient with major trauma.
patients. J Trauma 1989;29:1104–11; discussion Emerg Med Clin N Am 1998;16:1–15
1111–12 28. Jacobs LM. Timing of fluid resuscitation in
15. Chang FC, Harrison PB, Beech RR, Helmer SD. trauma. N Engl J Med 1994;331:1153–4
PASG: does it help in the management of 29. Capone AC, Safar P, Stezoski W, Tisherman S,
traumatic shock? J Trauma 1995;39:453–6 Peitzman AB. Improved outcome with fluid
16. Schwab CW, Gore D. MAST: medical antishock restriction in treatment of uncontrolled hemor-
trousers. Surg Annu 1983;15:41–59 rhagic shock. J Am Coll Surg 1995;180:49–56
17. Johnson M, Bruce, D., Gilbertson, J., 30. Soucy DM, Rude M, Hsia WC, Hagedorn FN,
Murkowski, F. Alaska EMS Goals: A Guide for Illner H, Shires GT. The effects of varying
Developing Alaska’s Emergency Medical Services fluid volume and rate of resuscitation during
System. Juneau: Alaska Department of Health uncontrolled hemorrhage. J Trauma 1999;46:
and Social Services – Section of Community 209–15
Health & Emergency Medical Services, 2003 31. Patel A, Goudar SS, Geller SE, et al. Drape
18. Stoy W. Mosby’s Emt-Basic Textbook. CV Mosby, estimation vs. visual assessment for estimating
1995 postpartum hemorrhage. Int J Gynecol Obstet
19. Domeier RM, O’Connor RE, Delbridge TR, 2006;93:220–4
Hunt RC. Use of the pneumatic anti-shock 32. Bickell WH, Wall MJ Jr, Pepe PE, et al.
garment (PASG). National Association of EMS Immediate versus delayed fluid resuscitation
Physicians. Prehosp Emerg Care 1997;1:32–5 for hypotensive patients with penetrating torso
20. Quinot J, Cantais E, Kaiser E. Le pantalon injuries. N Engl J Med 1994;331:1105–9
antichoc: A-ti-il reelement une place dans le 33. Pope A, French G, Longnecker D. Fluid
traitement du choc? Medecine d’urgence 2001: resuscitation. State of the science for treating combat
119–126 casualties and civilian injuries. Washington, DC:
21. Haggerty J. Anti shock garment. In National National Academy Press, 1999
Aeronautical Space Administration, Office of 34. Miller S, Hamza S, Bray E, et al. First aid for
Space Access and Technology, Commercial obstetrical hemorrhage: the pilot study of the
Development and Technology Transfer non-pneumatic anti-shock garment (NASG) in
Division, 1996 Egypt. Br J Obstet Gynaecol 2005;113:424–9
22. Pelligra R. Non-pneumatic antishock garment 35. Hensleigh P, Miller S, Ojengbede O, et al.
use. Emergency 1994;26:53–6 Non-pneumatic anti shock garment resuscitation
23. Brees C, Hensleigh PA, Miller S, Pelligra R. A with post partum hemorrhage. Presented at
non-inflatable anti-shock garment for obstetric Society of Gynaecology and Obstetrics of Nigeria
hemorrhage. Int J Gynaecol Obstet (SOGON) Annual Conference, November 22–25,
2004;87:119–24 2005, Ibadan, Nigeria
24. Miller S, Lester F, Hensleigh P. Prevention and 36. Miller S, Turan JM, Ojengbede, A, et al. The
treatment of postpartum hemorrhage: new pilot study of the non-pneumatic anti-shock gar-
advances for low-resource settings. J Midwifery ment (NASG) in women with severe obstetric
Womens Health 2004;49:283–92 hemorrhage: combined results from Egypt and
25. American College of Surgeons. Advanced Nigeria. Proceedings of the International Congress
Trauma Life Support Manual. Chicago: American on Evidence Based Interventions to Prevent Post
College of Surgeons, 1992 Partum Hemorrhage: Translating Research into
26. Harbrecht BG, Alarcon LH, Peitzman AB. Practice. July 12–15, 2006, Goa, India
Management of shock. In Moore EE, Feliciano
146
168
30 August 2006 14:20:32
15
SPECIAL CIRCUMSTANCES: JEHOVAH’S WITNESSES, THOSE
WHO REFUSE BLOOD TRANSFUSION AND/OR CONSENT
J. C. W. Marsh, M. O. Elebute and D. H. Bevan
THE JEHOVAH’S WITNESSES and the Royal College of Surgeons Code of

SOCIETY Practice (2002)7.
Jehovah’s Witnesses (JW) belong to the
religious organization, the Watch Tower
BLOOD ‘PRODUCTS’ ACCEPTABLE
Bible and Tract Society. They number an
FOR JW
estimated 6 million world-wide, of whom
145 000 live in the UK1. In Australia, a 2001 The JW Society demands that the ‘primary com-
population survey showed that the 81 000 ponents’ of blood must be refused. These are red
JW represented 0.4% of the population2. and white blood cells, platelets and plasma (fresh
JW refuse blood transfusion with the ‘primary frozen plasma). Autologous blood collection and
components’ of blood (see below for definition) storage for later re-infusion (pre-deposit) is not
and are prepared to die rather than be trans- acceptable to most JW as the blood is separated
fused. Until 2000, the church would have from the individual for a period of time by
expelled any member who had been transfused storage. In contrast, ‘fractions’ of plasma or
with any prohibited component of blood. Such cellular components such as albumin, immuno-
an individual would have been ostracized and globulins, non-recombinant clotting factors and
shunned by the members of the church and hemoglobin-based oxygen carriers are left to
their family, leading to social isolation. In 2000, individual choice (‘matters of conscience’). This
rejection by the church was abandoned and it circumstance highlights the importance of full
was left to the individual to revoke his own discussion with each individual JW patient to
membership from the Society. Although this elucidate what may be and what may not be
change in policy was seen as a relaxation of the acceptable to them. Regardless of their personal
JW policy on blood transfusion, the JW Society choice, all JW will accept crystalloids, synthetic
felt that no JW would wish to dissociate them- colloids (e.g. dextrans), hydroxyethylstarch and
selves3. In practical terms, this change may gelatins (for example Gelofusine).
mean that some JW may, in absolute medical During the mid-1980s, there was enormous
confidentiality, accept transfusion under certain interest in the development of red cell substi-
circumstances. Regardless, under British law, tutes, as summarized in a more recent review8.
any competent adult has an absolute right to These were either based on perfluorocarbons or
refuse transfusion. Readers are directed to hemoglobin-based oxygen carriers which used
previously published works on this area of hemoglobin from one of three sources: bovine
ethics (discussed in reference 4), whereas blood, outdated units of human red cells
issues of consent in adult JW are discussed or recombinant technology. Anecdotal reports
later in this review. Transfusion in children of describe their use in JW patients on a compas-
JW is not discussed in this review. Instead, the sionate basis9,10. Not all JW will accept them,
reader is referred to recent summaries4,5 based and their use in JW would be a matter of con-
on guidelines from the Association of Anaes- science for the individual. Disappointingly, their
thetists of Great Britain and Ireland (2005)6 undesirable safety profile and lack of efficacy
147
169
30 August 2006 14:20:32
have meant that today only a few products to rFVIIa. However, rFVIIa is currently so
are still being tested clinically and none are expensive that cost issues may act as ethical
licensed for human use in the USA, Europe issues, given that a single dose currently costs
or Canada11,12. Similarly, none of the platelet about £3600 in the UK. Case reports of its
substitutes under consideration are available successful use in JW to control life-threatening
for clinical use (reviewed in reference 4). bleeding in idiopathic thrombocytopenic
These used modified platelets, infusible platelet purpura17, and to correct the coagulopathy and
membranes, fibrinogen-coated albumin micro- bleeding in liver cirrhosis18 both exist.
spheres or semi-artificial platelet substitutes In addition to fractions of blood components
such as autologous erythrocytes or liposomes. outlined above, the following blood maneuvers
Their derivation from blood components would may or may not be acceptable to JW patients as
provide the same restrictions of acceptability to a matter of individual choice (‘matters of con-
JW as other blood components. science’): acute normovolemic hemodilution,
Recombinant products are accepted by most intraoperative and postoperative blood salvage
JW, such as granulocyte-colony stimulating techniques, hemodialysis and heart bypass
factor (G-CSF) used to treat neutropenia and to surgery where non-blood fluids must be used
mobilize peripheral blood stem cells for auto- to prime the pumps19,20.
logous and allogeneic transplantation. Erythro- Acute normovolemic hemodilution involves
poietin (rHuEPO) has been given to many JW, removing blood from the patient at the
although some JW refuse the Epoetin-beta beginning of an operation, replacing it with
preparation (NeoRecormon) because it con- crystalloid or colloid, and replacing the blood
tains a trace of albumin. In contrast, Epoetin- (which has been kept in direct contact with the
alfa (Eprex) does not contain any albumin as patient) near the end of surgery. Intraoperative
an expedient. Darbepoietin-alfa is a novel cell salvage is often acceptable, although post-
erythropoiesis-stimulating factor with two addi- operative collection of blood from surgical
tional carbohydrate side-chains and extra sialic drains is not acceptable. Contamination with
acid residues compared to rHuEPO, resulting in malignant cells, bacteria and amniotic fluid may
a longer half-life and increased activity in vivo13. be relative contraindications. Intraoperative cell
The use of rHuEPO in JW is discussed in salvage has been reported in a JW with placenta
greater detail below. The availability of recom- previa21. There were no reports of amniotic
binant factors VIII and IX permits the treatment fluid embolism in 174 women during Cesarean
of JW with hemophilia A and B, respectively, section, but with an incidence of only 1 : 8000
and desmopressin (DDAVP) for mild von to 1 : 80 000 of all deliveries, this study was too
Willebrand’s disease. small to assess the risk adequately. Never-
Recombinant coagulation factor VIIa theless, this maneuver, if available, may be
(rFVIIa; Eptacog-alfa; Novoseven: Novo- considered for JW women with a high risk of
Nordisk) is licensed for treatment of hemophilia life-threatening hemorrhage.
patients with inhibitors and for congenital disor- Organ transplantation may be acceptable to
ders of platelet function14,15. The drug is also JW. Although autologous blood donation is
being used (off-license) to treat life-threatening not acceptable, there are reports of autologous
hemorrhage in non-hemophilic patients16. In and allogeneic stem cell transplantation in JW4.
the setting of severe thrombocytopenia, if the Liver and pancreatic transplantations have also
thrombocytopenia is due to decreased produc- been performed in JW22,23.
tion of platelets (as in leukemia and other bone
marrow failures), it is anticipated that rFVIIa
would be ineffective because rFVIIa needs THE IMPACT OF WITHHOLDING
access to platelet surfaces to prevent bleeding. BLOOD IN JW PATIENTS
In autoimmune thrombocytopenia, however, a
Non-obstetric patients
low platelet count is hemostatically more effec-
tive as the platelets are younger and have more A series of 300 JW patients accrued from multi-
membrane surface, making it more responsive ple centers in the USA between 1981 and 1994
148
170
30 August 2006 14:20:32
Jehovah’s Witnesses, those who refuse blood transfusion and/or consent
and undergoing surgery with a postoperative chemotherapy regimens have been used to try
hemoglobin (Hb) < 8 g/dl showed an inverse to reduce the risk of death from anemia or
correlation between mortality and Hb level, bleeding, but at the expense of either failure to
with a sharp rise in mortality when the Hb fell achieve complete remission or early relapse
below 5–6 g/dl24. There were no deaths in 99 of the leukemia. An exception is acute pro-
patients with a nadir Hb between 7.1 and myelocytic leukemia where the use of all-trans
8.0 g/dl, but mortality rose from 8.9% with Hb retinoic acid (ATRA) and arsenic trioxide can
levels between 6.1 and 7.0 g/dl to 100% when the induce remissions as single agents without caus-
nadir Hb was 1.1–2.0 g/dl. The odds of death ing major myelosuppression27,28. The treatment
increased 2.5 times for each gram decrease in of acute lymphoid leukemia is somewhat more
postoperative Hb. This study, although not in successful in JW patients as the drugs are not
patients with postpartum hemorrhage, empha- so myelosuppressive as those used in acute
sizes the fact that mortality (and morbidity) are myeloid leukemia.
extremely high with very low Hb levels.
In the setting of intensive care (ICU), a trend
Obstetric patients
to higher mortality was observed among 21 JW
patients compared with non-JW patients treated The outcome of 332 JW women who had 391
between 1999 and 2003, although APACHE II deliveries during the period 1988–1999 was
scores, APACHE II risks of death and ICU reported from Mount Sinai Hospital in New
lengths of stay were similar between the two York29. All women were offered rHuEPO after
groups2. Of interest, three of the 21 patients had 28 weeks if the hematocrit level was < 36%, but
postpartum hemorrhage; of these, one died. Just most refused because the drug contained albu-
over half the patients did not receive rHuEPO. min. Obstetric hemorrhage occurred in 24 (6%)
The authors comment that this was not only of patients, two of whom died. The mortality
because of patient refusal but also because of rate was 521/100 000 live births, which repre-
the lack of a formal protocol for managing JW sented a 44-fold increase compared with mater-
patients which left the use of rHuEPO to the nal deaths from all causes in the general
discretion of individual physicians. As in the obstetric population at that institution during
previous study, no JW patient with an Hb level the same time interval. In addition to the two
of < 2 g/dl survived. fatal cases of postpartum hemorrhage, a third
A single-center study of 85 JW patients JW patient with massive postpartum hemor-
undergoing neurosurgery reported that JW rhage survived; this patient chose, before the
patients had longer hospital stay and longer delivery, to accept blood products.
operation times, but less blood loss peri- The off-label use of rFVIIa has been found
operatively25. The authors suggest that longer effective in several reported cases of peripartum
operation times may have been due to more hemorrhage unresponsive to all conventional
careful and slower surgical technique. Cell measures. These reports included women with
salvage was used in 47% of JW compared with disseminated intravascular coagulation (DIC),
4% of non-JW. There was no report of the use eclampsia or HELPP syndrome, previously
of rHuEPO in their patients. A similar outcome considered relative contraindications to rFVIIa
compared with non-JW patients was noted use. The use of rFVIIa in peripartum bleeding
despite performing potentially hemorrhagic spi- in the JW context has not been specifically
nal and intracranial cerebrovascular procedures. reported, but, as a recombinant non-blood-
The treatment of hematological disorders derived hemostatic agent, it is acceptable to JW
such as leukemia in JW patients poses extreme and should therefore be strongly considered
problems because red cell and platelet trans- in life- or function-threatening bleeding. The
fusions are critical to support the cytopenias, recommended dose is 90–100 µg/kg, rounded
which are due not only to the bone marrow up to the nearest number of vials, given as
failure associated with the leukemia but an intravenous bolus. The use of rFVIIa in
also due to the chemotherapy4,26. In most obstetric hemorrhage is discussed in more detail
instances of acute myeloid leukemia, modified elsewhere in this book.
149
171
30 August 2006 14:20:33
Published case reports of JW obstetric intraoperative cell salvage, the preoperative use
patients merit further discussion. Kalu and of rHuEPO and use of antifibrinolytic agents
colleagues reported a triplet pregnancy in a JW such as aprotonin and tranexamic acid to pre-
woman30. The prophylactic antenatal use of vent blood loss, as well as meticulous surgical
rHuEPO at 28 weeks (600 IU/kg intravenously hemostasis. A major issue to address preopera-
on three occasions over 2 weeks), along with tively in JW is their management of unexpected
oral iron supplementation, produced a rise in life-threatening blood loss. Before any elective
the Hb level from 11.1 to 13.2 g/dl. The patient surgery, it is vital that there is a formal planning
had an uncomplicated elective Cesarean section meeting with the patient, involving the surgeon
with 500 ml estimated blood loss and Hb level and anesthetist who will be performing the sur-
of 12.2 g/dl on day 2. The authors highlighted gery, along with input from the hematological
the need for further studies to establish the team. The latter member could be either a con-
safety and optimal dose of rHuEPO in preg- sultant hematologist or a transfusion nurse spe-
nancy, the target Hb level and the optimal route cialist who is an official member of the hospital
for iron administration (see below). De Souza transfusion team (see above). The main aims of
and colleagues report the antenatal use of this meeting are to (1) ensure the patient is fully
rHuEPO at a dose of 50 IU/kg twice weekly informed of the risks of bleeding and (2) estab-
from 29 weeks in a JW woman, resulting in an lish and document the extent of the patient’s
increase in the Hb level from 10.9 to 13.3 consent in terms of exactly what blood prod-
6 weeks later21. At 34 weeks, antenatal ucts/maneuvers are acceptable and what are
hemorrhage necessitated Cesarean section not acceptable for that individual patient. Below
with intraoperative cell salvage. Maternal post- are described the current guidelines used at our
operative Hb level was 12.0 g/dl and recovery hospital (protocol of St George’s Hospital, Lon-
uneventful, despite a theoretical risk of amniotic don, 2006) and updated from our previously
fluid embolism (as discussed earlier). The use of published guidelines4.
rHuEPO in pregnancy is not contraindicated, as
there is no evidence of harm in pregnancy. An
additional advantage is that it is secreted in Guidelines for consent to elective surgery
breast milk and stimulates erythropoiesis in in JW patients
premature infants when breast-fed31. A case in Early warning system
the USA reported a pregnant JW patient with
placental abruption and intrauterine death at 31 The fact that a JW patient requires an elective
weeks who developed DIC after vaginal deliv- operation must be ascertained and communi-
ery32. Her Hb level fell from 11.8 to 2.9 g/dl. cated to the ‘JW ad hoc team’ (see below)
The patient agreed to have rHuEPO and also a at least 2 weeks (10 working days) before
polymerized human Hb solution (PolyHeme) the operation date. Notification of less than
which resulted in a rise of her Hb level to 2 weeks before planned surgery may lead to
4.5 g/dl and clinical stabilization. The problems cancellation.
associated with the clinical development of red
cell substitutes are discussed earlier.
The JW ad hoc team
This consists of the consultant surgeon and
MANAGEMENT OF ELECTIVE consultant anesthetist (key participants) and the
SURGERY consultant hematologist or transfusion nurse
specialist (facilitator). The consultant anesthe-
Consent and planning
tist is self-identified as prepared to accede to the
The management of JW patients undergoing JW patient’s beliefs and wishes (a list is held
elective surgery has helped further the develop- by the Anesthetic Office). All three consultants
ment of ‘bloodless surgery’ for the general must have sufficient time and, ideally, concur-
population19,20. Important practical maneuvers rent time, for the preoperative meeting. This
include acute normovolemic hemodilution, meeting is then arranged with the JW patient. If
150
172
30 August 2006 14:20:33
this meeting cannot be accomplished by three have any further questions or concerns. The
working days before surgery, or if key partici- clinical team then agree (or disagree) with the
pants (including the JW patient) cannot attend patient to go forward with the operation on
the meeting, then surgery should be cancelled. the terms agreed upon, and this commitment
is documented. If the patient has made an
Advance Directive, it should be read and a copy
Format of the preoperative meeting
placed in the notes.
This is flexible, but should contain the follow- On occasion, these guidelines are difficult to
ing. The patient is assured that the meeting is to achieve, as it may not always be possible for
formulate a plan for surgery that complies with all interested parties to meet simultaneously.
her wishes and beliefs, and that no attempt will In this situation, the transfusion nurse specialist/
be made to frighten her or place her under hematologist and the patient can meet with
duress. She should be asked if she has consulted the surgeon and anesthetist separately, using a
the written advice of the JW Transfusion Com- single checklist form to document discussions
mittee on permissible products for infusion, and (and results of investigations) for both
if she differs from the view of the JW Transfu- consultations.
sion Committee in any respect. The duration of In most instances, JW patients will insist
the meeting should be set at a maximum of on being accompanied by either a relative or
45 min except in unusual circumstances. All associate who is also a JW. Under such circum-
comments, questions and answers must be stances, it may then be difficult to know for
documented. certain whether peer pressure has prevented the
The surgeon outlines the proposed opera- patient from making her own decisions.
tion, describes possible complications that may Although the Association of Anesthetists of
result in bleeding, and reminds the patient of Great Britain and Ireland (AAGBI) issued
the ever-present risk of bleeding with any sur- guidelines in 20056 and state that ‘it is very
gery. This description and its understanding by important to take the opportunity to see the
the patient are also documented. The anesthe- patient without relatives or members of the
tist outlines techniques used to avoid transfu- local community’, the patient may insist on
sion of blood. The patient’s informed consent their presence at this meeting. The patient
to these matters is obtained and documented. should be offered private consultation, but, if
The anesthetist asks what actions are and are it is declined, one has to accept her desire, as
not sanctioned by the patient if she is uncon- well as her written or verbal consent, as a true
scious or otherwise unable to communicate and indication of her will.
dying of unexpected blood loss, and this too is
documented.
The hematologist (or transfusion nurse spe-
Preoperative assessment
cialist) asks the JW patient which therapeutic
agents are acceptable to infuse to support blood The patient should be seen by a hematologist
volume and/or hemostatic function in the event to review her medical history for bleeding
of bleeding. The written or spoken advice of the episodes, hypertension, previous anemia and
JW Transfusion Committee to the patient may any drug history for medications that may exac-
be helpful at this stage (see below). The answer erbate bleeding, such as aspirin, non-steroidal
to this question is documented. If clinically anti-inflammatory drugs and warfarin. The
appropriate and timely, the hematologist presence of infection, inflammation or malig-
explains the technique of preoperative Hb nancy predicts a poor response to rHuEPO.
enhancement using rHuEPO. If the patient Any evidence of anemia should be thoroughly
accepts this therapy, clinical assessment and investigated and treated preoperatively. The
rHuEPO therapy (if appropriate) is arranged on following investigations are recommended: full
the Hematology Day Unit. blood count, coagulation screen, serum B12,
At the end of the discussion, the JW patient folate and ferritin, serum urea and creatinine,
and her supporter(s) should be asked if they electrolytes and liver function.
151
173
30 August 2006 14:20:33
Other practical issues who is at immediate risk of dying from blood

loss, whether in ICU or the Accident and Emer-
Rationalizing the frequency and volume of blood
gency department, and the relatives inform the
sampling is important to reduce blood loss post-
medical team that the patient is a JW and pro-
operatively. The use of pediatric blood sample
duce an Advance Directive signed by the patient
tubes is recommended. Postoperative folic acid
confirming his/her wish not to be transfused,
should be considered when reduced oral intake
even if their death is imminent from massive
is anticipated, and folinic acid used when oral
bleeding. A Canadian case found a doctor liable
intake is not possible. Iron supplementation
for assault when a blood product was transfused
should be used if there is postoperative bleed-
to an unconscious JW. In contrast, a British case
ing. If the patient is unable to take oral iron or if
was upheld because the patient’s refusal was
rHuEPO is being used, then intravenous iron
open to doubt, as it was potentially subject
infusions, such as iron sucrose (Venofer), can be
to influence by co-religionists (cases discussed
given safely. There is some evidence that intra-
in reference 4). There must be the following
venous iron may be more efficacious than oral
certainties: (1) that the patient is a committed
iron generally when rHuEPO is used preopera-
JW, (2) they have independently and freely
tively33 and when used in dialysis patients34, but
decided to refuse transfusion, and (3) they had
some centers use oral iron supplementation and
considered this to the point of death at the
only reserve intravenous iron if there is a poor
time of making their Advance Directive. The
response on the basis of iron studies.
Association of Anaesthetists of Great Britain
and Ireland advise in their guidelines that,
Erythropoietin (rHuEPO) administration ‘In the management of an unconscious patient
whose status as a JW may be unknown, the
In critically ill patients, one randomized study doctor caring for the patient will be expected
showed that rHuEPO significantly increased to perform to the best of his ability, and this
Hb levels and reduced blood transfusion may include the administration of blood trans-
requirements, although it had no impact on fusion’6. However, this guideline only applies
clinical outcome or mortality35. rHuEPO is when the JW status is unclear and/or the
effective in boosting the Hb level in individuals relatives/associates cannot produce an Advance
undergoing autologous blood donation Directive. The reader is also referred to The
(although this maneuver is unacceptable to JW Royal College of Surgeons Code of Practice,
patients). For some JW, it can be used pre- and 2002, for further discussion on this issue7.
postoperatively, or in the ante- or postnatal Described below are the current guidelines
period. High doses are needed compared with used at St George’s Hospital, which were drawn
chronic kidney disease patients. The current up with the help of our Trust solicitors4.
UK licensed dose for Eprex (Epoetin-alfa) is
600 U/kg weekly for 3 weeks and on the day
of surgery. An alternative regimen is 300 U/kg Guidelines for life-threatening bleeding in
given for 15 days, starting 10 days before an unconscious adult JW
surgery. From a practical point of view, having (1) Any documentary evidence, for example, an
started on a program of rHuEPO preopera- Advocate Directive (living will), stating that
tively, it is important to ensure the date of the patient will not accept blood transfusion
surgery. In the event of cancellation of surgery, even in the event of life-threatening bleed-
rHuEPO would need to be continued to avoid a ing, should be requested from relatives or
fall in Hb level when it is discontinued. associates of the patient and examined, if
time permits.
Management of life-threatening bleeding (2) A copy should be put in the case notes and
in an unconscious adult JW patient its contents respected.
The most difficult aspect in the management of (3) The doctor (who should be of Consultant
JW is when faced with an unconscious patient status), if time permits, should discuss with
152
174
30 August 2006 14:20:33
the patient’s relatives the implications of 12. Stowell CP. Whatever happened to blood
withholding blood. substitutes? Transfusion 2004;44:1403–4
13. Smith, R. Applications of darbepoietin-a, a novel
(4) The doctor should act in the best interests erythropoiesis-stimulating protein, in oncology.
of the patient and will be expected to per- Curr Opin Hematol 2002;9:228–33
form to the best of his/her ability, which 14. Hedner U. Treatment of patients with factor
may involve giving blood if steps 1, 2 and 3 VIII and factor IX inhibitors with special focus
are impossible. on the use of recombinant factor VIIa. Thromb
Haemost 1999;82:531–4
(5) A clear and signed entry of the steps taken 15. D’Oiron R, Menart C, Trzeciak MC, et al. Use
should be written in the patient’s case of recombinant factor VIIa in 3 patients with
notes. inherited type 1 Glanzmann’s thrombaesthenia
undergoing invasive procedures. Thromb
Haemost 2000;83:644–7
16. Thompson AR. When all else fails to stop mas-
References
sive bleeding from trauma. J Thromb Haemost
1. Anonymous. 2003 report of Jehovah’s Witnesses 2005;3:638–9
worldwide: Watch Tower Bible and Tract 17. Waddington DP, McAuley FT, Hanley JP,
Society of Pennsylvania, 2004 Summerfield GP. The use of recombinant
2. Maclaren G, Anderson M. Bloodless intensive Factor VIIA in a Jehovah’s Witness with
care: a case series and review of Jehovah’s autoimmune thrombocytopenia and post-
Witnesses in intensive care unit. Anaesth Intens splenectomy haemorrhage. Br J Haematol
Care 2004;32:798–803 2002;119:286–8
3. Muramoto O. Bioethical aspects of the recent 18. Papatheodoridis GV, Chung S, Keshav S, Pasi J,
changes in the policy of refusal of blood by Burroughs AK. Correction of both prothombin
Jehovah’s Witnesses. Br Med J 2001;322:37–9 time and primary haemostasis by recombinant
4. Marsh JCW, Bevan DH. Haematological care of factor VII during therapeutic alcohol injection of
the Jehovah’s Witness patient. Br J Haematol hepatocellular cancer in liver cirrhosis. J Hepatol
2002;119:25–37 1999;31:747–50
5. Woolley S. Children of Jehovah’s Witnesses and 19. Nash JM., Cohen H. Management of Jehovah’s
adolescent Jehovah’s Witnesses: what are their Witness patients with haematological problems.
rights? Arch Dis Child 2005;90:715–19 Blood Rev 2004;18:211–17
6. Association of Anaesthetists of Great Britain 20. Goodnough LT, Shander A, Spence R. Blood-
and Ireland, London. Management of less medicine: clinical care without allogeneic
Anaesthesia for Jehovah’s Witnesses, 2nd edn, blood transfusion. Transfusion 2003;43:668–76
2005. Website: www.aagbi.org.uk. 21. de Souza A, Permezel M, Anderson M, Ross A,
7. Royal College of Surgeons of England. Code McMillan J, Walker S. Antenatal erythropoietin
of practice for the surgical management of and intra-operative cell salvage in a Jehovah’s
Jehovah’s Witnesses, 2002. Website: www. Witness with placenta praevia. Br J Obstet
rcseng.ac.uk. Gynaecol 2003;110:524–6
8. Klein HG. The prospects for red-cell substitutes. 22. Gagandeep JN, Mateo S, Sher L, et al. Live
N Engl J Med 2000;342:1666–8 donor liver transplantation without blood
9. Agrawal YP, Freedman M, Szczepiorkowski products: strategies developed for Jehovah’s
ZM. Long-term transfusion of polymerized Witnesses offer broad application. Ann Surg
bovine hemoglobin in a Jehovah’s Witness 2004;240:350–7
following chemotherapy for myeloid leukemia: 23. Detry O, Deroover A, Delwade J, et al. Avoiding
a case report. Transfusion 2005;45:1735–8 blood products during liver transplantation.
10. Anton N, Hitzler JK, Kavanagh BP. Treatment Transplant Proc 2005;37:2869–70
of life-threatening post-haemorrhagic anaemia 24. Carson J, Noveck H, Berlin J, Gould S. Mortality
with cell-free haemoglobin solution in an and morbidity in patients with very low post-
adolescent Jehovah’s Witness. Br J Haematol operative haemoglobin levels who decline blood
2002;118:1183–6 transfusion. Transfusion 2002;42:812–18
11. Buehler PW, Alayash AI. Toxicities of hemoglo- 25. Suess S, Suess O, Brock M. Neurosurgical
bin solutions: in search of in-vitro and in-vivo procedures in Jehovah’s Witnesses: an increased
model systems. Transfusion 2004;44:1516–30 risk? Neurosurgery 2001;49:266–72
153
175
30 August 2006 14:20:33
26. Cullis JO, Duncombe AS, Dudley JM, Lumley 31. Kraft A, Breymann C, Huttner C, Huch R,
HS, Apperley JF, Smith AG. Acute leukaemia in Huch A. Erythropoietic quality of maternal milk.
Jehovah’s Witnesses. Br J Haematol 1998;100: Lancet 1999;354:778
664–8 32. Cothren CC, Moore EE, Long JS, Haenel JB,
27. Lin C-P, Liu H-J, Tsai C-H. Successful Johnson JL, Ciesla DJ. Large volume poly-
treatment of acute promyelocyte leukaemia in merized haemoglobin solution in a Jehovah’s
a pregnant Jehovah’s Witness with all-trans Witness following abruptio placentae. Transfu-
retinoic acid, rhG-CSF and erythropoietin. Am J sion Med 2004;14:241–6
Hematol 1996;51:251–2 33. Rohling R, Zimmerman A, Breymann C. Intra-
28. Menedez A, Svarch E, Martinez G, Hernandez venous versus oral iron supplementation from
P. Successful treatment of acute promyelocyte preoperative stimulation of hemoglobin synthesis
leukaemia using all-trans retinoic acid and eryth- using recombinant human erythropoietin. J
ropoietin in a Jehovah’s Witness. Ann Haematol Haematother Stem Cell Res 2000;9:497–500
1998;76:43–4 34. Macdougall IC, Tucker B, Thompson J,
29. Singla AK, Lapinski RH, Berkowitz RL, Saphier Tomson CR, Baker LR, Raine AE. A random-
CJ. Are women who are Jehovah’s Witnesses ised controlled study of iron supplementation in
at risk of maternal death? Am J Obstet Gynecol patients treated with erythropoietin. Kidney Int
2001;185:893–5 1996;50:1694–9
30. Kalu E, Wayne C, Croucher C, Findley I, 35. Corwin HL, Gettinger A, Pearl RG, et al.
Manyonda I. Triplet pregnancy in a Jehovah’s Efficacy of recombinant human erythropoietin in
Witness: recombinant human erythropoietin and critically ill patients. JAMA 2002;288:2827–35
iron supplementation for minimising the risks of
excessive blood loss. Br J Obstet Gynaecol 2002;
109:723–5
154
176
30 August 2006 14:20:34
Section IV
Special preventive measures:
misoprostol in action
177
30 August 2006 14:20:34
‘With the emerging evidence on the use of various routes of administration of misoprostol,
particularly in the non-hospital setting, it is becoming clear that this drug should be available at
the community level in the hands of trained personnel, especially where oxytocin, Uniject and other
uterotonics are not present or practical for use.’
The Working Group of the Goa International Conference on the Prevention of
Post Partum Hemorrhage, July 15, 2006, Goa, India
16
MISOPROSTOL IN PRACTICE
M. Potts
Prior to the availability of misoprostol, it was Although these measures may seem revolu-
impossible to carry any significant element of tionary at first glance, they should be viewed as
emergency obstetric care into homes where an essential step towards a long-term strategy
women deliver without a skilled birth attendant. where all women can be delivered by a certified
As a low-cost, easy-to-administer, powerful midwife or physician practicing active manage-
uterotonic with an excellent safety profile and ment of the third stage of labor. Over the past
long shelf-life, misoprostol has a revolutionary half-century, countries such as Sri Lanka and
potential to reduce death and morbidity from Thailand have brought maternal mortality to
postpartum hemorrhage in precisely those situa- low levels by ensuring over 90% of deliveries are
tions where it is most common – delivery at attended by a skilled person able to use an
home without a skilled birth attendant. oxytocic, and ultimately all countries should
In a placebo-controlled, community-based follow such a path.
trial in India, administration of 600 µg miso- Unfortunately, rapid population growth,
prostol orally immediately after delivery sig- economic collapse and the spread of HIV/
nificantly reduced postpartum hemorrhage (see AIDS in some African countries and the endless
Addendum). Research in Indonesia, Nepal and recruitment of skilled health professions from
elsewhere is showing that community volun- developing to developed countries will make the
teers with minimal training can teach illiterate road to providing comprehensive obstetric care
women to self-administer misoprostol effectively long and slow. During this interval, widespread
and responsibly1 (see Chapter 19). A 1000 µg access to misoprostol and the education to use it
rectal dose of misoprostol can be used to treat safely during home births have the potential to
postpartum hemorrhage, in situations where an make a significant contribution – perhaps even
appropriate technology exists to diagnose blood the single most important contribution – to
loss (such as blood-soaked sarong or ‘kanga’), reducing the global burden of deaths from
and where births are attended by traditional postpartum hemorrhage. The only other practi-
birth attendants (TBAs). In Tanzania, illiterate cal intervention with the potential to reduce
TBAs, with a brief training, used misoprostol to postpartum hemorrhage in low-resource set-
bring about a highly significant reduction in the tings is realistic access to family planning, as all
number of women who needed to be referred to women who wish to limit childbearing are at risk
hospital or receive intravenous treatment2. of postpartum hemorrhage, and the older,
156
178
06 September 2006 14:27:54
Misoprostol in practice
higher-parity women, who have the greatest program effectiveness (SAFE) demonstration
unmet need for family planning, are at even project of community-based distribution of miso-
higher risk. prostol for prevention of PPH in rural Indonesia.
Proceedings of Preventing Postpartum Hemorrhage:
From Research to Practice, Bangkok, Thailand,
References January 20–24, 2004:31–7
2. Prata N, Mbaruka G, Campbell M, Potts M,
1. Wiknjosastro G, Sanghvi H. Preventing PPH
Vahidnia F. Controlling postpartum hemorrhage
among women living in areas where a high
after home births in Tanzania. Int J Gynaecol
proportion of births are not attended by skilled
Obstet 2005;90:51–5
providers: Safety, acceptability, feasibility and
Editors’ Addendum
The Editors wish to bring the reader’s atten- R. J. Derman1, B. S. Kodkany2, S. S.
tion to the paper referred to by Professor Goudar2, S. E. Geller3, V. A. Naik2, M. B.
Potts on page 156. This paper has been pub- Bellad2, S. S. Patted2, A. Patel4, S. A.
lished in the October 7, 2006 issue of The Edlavitch1, T. Hartwell5, H. Chakraborty5,
Lancet. To the Editors’ knowledge, this is the N. Moss6. Oral misoprostol in preventing post-
largest placebo-controlled study of miso- partum hemorrhage in a community setting.
prostol for the prevention of postpartum 1University of Missouri-Kansas City School
hemorrhage, and the results showed that of Medicine; 2Jawaharlal Nehru Medical
misoprostol significantly reduced the rate of College, Belgaum, Karnataka, India; 3Univer-
postpartum hemorrhage in the patients who sity of Illinois, Chicago College of Medicine,
were administered this agent in comparison to USA; 4John H. Stroger Jr. Hospital of Cook
the patients who received the placebo control. County, USA; 5Statistics and Epidemiology,
The full title of the paper and all authors RTI International; 6National Institute of
are: Child Health and Human Development.
157
179
04 September 2006 13:53:27
17
MANAGEMENT OF POSTPARTUM HEMORRHAGE AT THE
COMMUNITY LEVEL
N. Prata
The ability to manage postpartum hemorrhage all of them produced encouraging results.
at the community level is an essential element in During one intervention trial in rural Kigoma,
any program to decrease maternal mortality Tanzania, Prata and colleagues demonstrated
from postpartum hemorrhage1, as most of that traditional birth attendants (TBAs), who
the deliveries in developing countries occur at assist in most home deliveries, were able to diag-
home without the presence of a skilled birth nose postpartum hemorrhage and effectively
attendant2. and safely administer 1000 µg of rectal miso-
The efficacy, safety, and importance of misoprostol to control postpartum hemorrhage4.
prostol use for postpartum hemorrhage man- Blood loss measurement was standardized by
agement are well established for hospital-based employing the traditional blood collection tool
settings3. However, misoprostol’s most signifi- used by women in the region – the local garment
cant impact will probably be at the household that is colloquially referred to as a ‘kanga’5. This
level, where most deliveries occur. Some studies study also showed that the ability to manage
have tested such technology in home births, and postpartum hemorrhage in home births resulted
Table 1 Controlling postpartum hemorrhage (PPH) with a 1000 µg of misoprostol (intervention) in home
births, Kigoma, Tanzania
Intervention Non-intervention
Odds ratio
Main outcomes of the study n % n % (95% CI)
PPH (blood loss ≥ 500 ml) 111 24.5 73 18.5 1.3 (1.0–1.7)
Referrals 8 1.8 75 19.0 0.1 (0.0–0.2)
Additional interventions among PPH cases n = 111 n = 73
Type of additional interventionsa b 1b 0.9 c69c
94.5
Intravenous fluids 1 0.9 25 34.3
Blood transfusion 1 0.9 16 21.9
Manual removal of placenta 0 0.0 17 23.3
Repair of tears 0 0.0 4 5.5
Hysterectomy 0 0.0 1 1.4
Other medical interventionsd 0 0.0 7 9.6
aNumber of cases do not add up to total referred, some women had more than one intervention; bhospital
records not available for one patient; three patients did not need additional interventions; another three were
referred for other reasons than PPH; chospital records not available for four patients; four patients did not
need additional interventions; two cases referred for other reasons than PPH; dmedical interventions
included: Amoxyl tablets, methergin, and misoprostol.
Source: Prata N, et al. Controlling postpartum hemorrhage after home births in Tanzania. Int J Gynaecol
Obstet 2005;90:51–5
158
180
30 August 2006 14:20:34
Management at the community level
in important reductions in the number of refer- including home births, and particularly in those
rals and the need for additional interventions, where it must be self-administered.
key factors in resource-poor settings (Table
1).This is particularly helpful in rural areas.
References
In settings where culturally appropriate
methods of measuring blood loss after delivery 1. Khan KS, Wojdyla D, Say L, Gulmezoglu AM,
are difficult to devise, all women could be Van Look PF. WHO analysis of causes of mater-
administered a prophylactic dose of 600 µg nal death: a systematic review. Lancet 2006;367:
misoprostol after delivery of the baby. Its safety 1066–74
2. AbouZahr C, Wardlaw T. Maternal mortality at
and efficacy in the hands of TBAs were shown
the end of a decade: signs of progress? Bull WHO
in a randomized, controlled trial in the Gam-
2001;79:561–8
bia6. In addition, in places where, for cultural or 3. Villar J, Gulmezoglu AM, Hofmeyr GJ, Forna F.
other reasons, women deliver at home alone or Systematic review of randomized controlled trials
in the presence of a family member, self-admin- of misoprostol to prevent postpartum hemor-
istration of a prophylactic dose of misoprostol, rhage. Obstet Gynecol 2002;100:1301–12
distributed during pregnancy by trained com- 4. Prata N, Mbaruku G, Campbell M, Potts M,
munity health-care workers, are both viable Vahidnia F. Controlling postpartum hemorrhage
options that can produce promising results, as after home births in Tanzania. Int J Gynaecol
was shown in other studies in Indonesia, Nepal, Obstet 2005;90:51–5
and Afghanistan. 5. Prata N, Mbaruku G, Campbell M. Using the
Kanga to measure postpartum blood loss. Int J
It will be many decades before all women in
Gynaecol Obstet 2005;89:49–50
low-resource settings can receive skilled atten-
6. Walraven G, Blum J, Dampha Y, et al. Miso-
tion at delivery in their homes. In the meantime, prostol in the management of the third stage of
misoprostol has the potential to make a signifi- labour in the home delivery setting in rural Gam-
cant difference in reducing maternal mortality. bia: a randomised controlled trial. Br J Obstet
It should be made available for use in all settings Gynaecol 2005;112:1277–83
159
181
30 August 2006 14:20:34
18
ORAL MISOPROSTOL FOR PREVENTION OF POSTPARTUM
HEMORRHAGE BY PARAMEDICAL WORKERS IN INDIA
(AN ICMR TASK FORCE STUDY)
N. Chandhiok
Paramedical workers conduct deliveries in the hemorrhage. Six hundred women each received
rural areas of India where active management of either active management of the third stage
the third stage of labor is not routinely practised of labor with 600 µg of oral misoprostol (inter-
and uterotonic agents are only provided for the vention) or the current government guidelines
management of postpartum bleeding. A multi- for prevention of postpartum hemorrhage
site, cluster-randomized, feasibility study was (controls). The primary outcome was blood loss
carried out to determine if paramedical workers after delivery and this was measured using a
from rural Peripheral Health Centers (PHCs) calibrated blood collection drape.
could actively manage the third stage of labor Baseline characteristics were comparable in
using oral misoprostol to prevent postpartum both groups and over 70% of women had
Table 1 Outcome of intervention. Figures in parentheses are percentages
Intervention Comparison
Tablet Injection Tablet

misoprostol methergine methergine None† Total
(n = 600) (n = 531) (n = 58) (n = 11) (n = 600)
Duration of third 7.9 ± 4.2 11.1 ± 4.1*** 9.6 ± 5.0*** 5.9 ± 2.4 10.9 ± 4.3***
stage of labor (min)
(mean ± SD)
Blood loss (ml)
Mean ± SD 139.7 ± 100.4 211.8 ± 80.6*** 211.6 ± 83.0*** 171.8 ± 178.3 211.0 ± 83.4***
Median 100 200*** 200*** 100 200***
Q1–Q3 90–150 150–250 150–280 100–160 150–250
Range 25–1300 30–750 25–415 100–700 25–750
Postpartum hemorrhage 4 (0.7) 4 (0.8)NS – 1 (9.1) 5 (0.8)NS
Additional measures
Uterotonics 4 2*** – 1 3***
Intravenous fluids 4 2*** – 1 3***
Blood transfusion 1 – – – –
Referred to higher level 2 (0.3) 1 (0.2) 1 (0.2) 2 (0.3)
of health facility for
PPH
***p < 0.001; **p < 0.01 when compared with the intervention; †group was not compared with intervention
due to small sample
NS, not significant; PPH, postpartum hemorrhage
160
182
30 August 2006 14:20:34
Prevention by paramedical workers in India
moderate anemia. The paramedical workers to address the reduction in postpartum hemor-
were able to provide the intervention according rhage at such low incidence (Table 1).
to the guidelines in almost all deliveries (99%). As most deliveries in rural areas take place at
There was a significant reduction in the dura- home, there is a need to extend this study for all
tion of the third stage of labor (7.9 ± 4.2 vs. domiciliary deliveries.
10.9 ± 4.3 min, p < 0.001), and the measured
blood loss after delivery in the intervention
group (139.7 ± 100.4 ml vs. 211.0 ± 83.4 ml,
ACKNOWLEDGEMENT
p < 0.001). This magnitude of reduction is
significant for a country such as India where This communication is based on the following
80% of the women are anemic at the time of previously published article at the Editor’s
delivery and any reduction in blood loss is request: Chandhiok N, Dhillon BS, Datey S,
considered highly beneficial. The overall Mathur A, Saxena NC. Oral misoprostol for
incidence of postpartum hemorrhage observed prevention of postpartum hemorrhage by
in the study was extremely low (< 1% in both paramedical workers in India. Int J Gynaecol
groups), and the study size was not adequate Obstet 2006;92:170–5
161
183
30 August 2006 14:20:34
19
OVERVIEW OF MISOPROSTOL STUDIES IN
A. Hemmerling
INTRODUCTION non-randomized trials, and they represent a

diversity of situations. Table 1 provides an
These tables of peer-reviewed misoprostol
overview of 32 studies in the prevention of
studies were compiled to provide the reader
postpartum hemorrhage (including dosage
with comprehensive references to the use of
and route of administration). Table 2 gives an
misoprostol in practice since 1997, for both
overview of seven studies in the treatment of
prevention and treatment of postpartum hemor-
postpartum hemorrhage (including dosage and
rhage. The tables include both randomized and
route of administration).
162
184
30 August 2006 15:04:27
Table 1 Misoprostol for prevention
Participants Participants
in misoprostol Dosage of Route of in control
30 August 2006 14:20:36

Authors Institutions Study title Journal n group misoprostol administration group(s) Control agent(s)
Prata N, Bixby Program in Misoprostol and active Int J Gynaecol 8 2532 1189 600 µg oral 1343 current AMTSL
Hamza S, Population, Family management of the third Obstet 2006 practices
Gypson R, et al. Planning and Maternal stage of labor Jul 6 [epub
Health, School of Public ahead of
Health, University of print]
California, Berkeley, USA

Nellore V, Dept. of Obstetrics and Rectal misoprostol vs. Int J Gynaecol 8 120 60 400 µg rectal 60 125 µg 15-methyl
Mittal S, Gynecology, All India 15-methyl prostaglandin Obstet 2006; prostaglandin F2α
Dadhwal V Institute of Medical F2α for the prevention of PPH 94:45–6 i.m.
Sciences, Ansari Nagar,
New Delhi, India
Chandhiok N, Division of Reproductive Oral misoprostol for Int J Gynaecol 8 1200 600 600 µg oral 600 current government
Dhillon BS, Health and Nutrition, prevention of PPH by Obstet 2006; guidelines for PPH
Datey S, et al. Indian Council of Medical paramedical workers in India 92:170–5 prevention
Research, New Delhi, India
163
185
Zachariah ES, Dept. of Obstetrics and Oral misoprostol in the third Int J Gynaecol 8 2023 730 400 µg oral [1] 617 [1] 10 IU oxytocin
Naidu M, Gynecology, Christian stage of labor Obstet 2006; [2] 676 i.m.
Seshadri L Medical College Hospital 92:23–6 [2] 2 mg
Vellore, India ergometrine i.v.
Garg P, Maulana Azad Medical Oral misoprostol vs. injectable Int J Gynaecol 8 200 100 600 µg oral 100 0.2 mg
Batra S, College and Lok Nayak methylergometrine in Obstet 2005; methylergometrine
Gandhi G Hospital, Delhi, India management of the third 91:160–1 i.v.
stage of labor
Ozkaya O, Dept. of Obstetrics and Placebo-controlled randomized J Obstet 8 150 [1] 50 400 µg [1] rectal 50 placebo

Sezik M, Gynecology, School of comparison of vaginal with Gynaecol Res [2] 50 [2] oral
Kaya H, et al. Medicine, Suleyman rectal misoprostol in the 2005;31:
Demirel University, Turkey prevention of PPH 389–93
Hoj L, Dept. of Obstetrics and Effect of sublingual BMJ 2005; 8 661 330 600 µg sublingual 331 placebo
Cardoso P, Gynecology, Aarhus misoprostol on severe PPH 331:723
Nielsen BB, University Hospital, in a primary health center in
et al. Denmark Guinea-Bissau: randomized
double-blind clinical trial
Overview of misoprostol studies in postpartum hemorrhage
Continued
Table 1 Continued
30 August 2006 14:20:36

Walraven G, Farafenni Field Station, Misoprostol in the BJOG 2005; 8 1229 630 600 µg oral 599 2 mg ergometrine
Blum J, Medical Research management of the third stage 112:1277–83 oral
Dampha Y, Council Laboratories, of labor in the home delivery

et al. Farafenni, Gambia setting in rural Gambia: a
randomized controlled trial
Vimala N, Dept. of Obstetrics and Sublingual misoprostol versus Int J Gynaecol 8 120 60 400 µg sublingual 60 0.2 mg
Mittal S, Gynecology, All India methylergometrine for active Obstet 2004; methylergometrine
Kumar S, Institute of Medical Sciences, management of the third 87:1–5 i.v.
et al. Ansari Nagar, New Delhi, stage of labor
India
Lam H, Tang Dept. of Obstetrics and A pilot-randomized Acta Obstet 8 60 30 600 µg sublingual 30 1 ml syntometrine
OS, Lee CP, Gynecology, Queen Mary comparison of sublingual Gynecol Scand i.v. (5 IU
et al. Hospital, Hong Kong SAR, misoprostol with syntometrine 2004;83: syntocinone and
164
186
China on the blood loss in third stage 647–50 0.5 mg ergometrine
of labor maleate)
Caliskan E, SSK Maternity and Women’s Oral misoprostol for the third Obstet Gynecol 8 1574 388 600 µg oral [1] 404 [1] 600 µg
Dilbaz B, Health Teaching Hospital, stage of labor: a randomized 2003;101: [2] 384 misoprostol plus
Meydanli MM, Ankara, Turkey controlled trial 921–8 [3] 398 10 IU oxytocin i.v.
et al. [2] 10 IU oxytocin
i.v.
[3] 10 IU oxytocin
i.v. plus 0.2 mg
methylergonovine

maleate
Oboro VO, Maternity Unit, Zonal A randomized controlled trial Obstet Gynecol 8 496 247 600 µg oral 249 10 IU oxytocin i.m.
Tabowei TO General Hospital, Kwale, of misoprostol vs. oxytocin in 2003;23:
Delta State, Nigeria the active management of the 13–16
third stage of labor
Lumbiganon P, Dept. of Obstetrics and Side-effects of oral BJOG 8 1686 843 600 µg oral 843 10 IU oxytocin i.m.
Villar J, Piaggio Gynecology, Faculty of misoprostol during the first 2002;109: or i.v.
G, et al. Medicine, Khon Kaen 24 h after administration in 1222–6
University, Thailand the third stage of labor
Quiroga Diaz Hospital de Ginecologia Vaginal misoprostol in the Ginecol Obstet 8 400 208 800 µg vaginal 192 current AMTSL
R, Esparaza y Obstetricia de Monterrey, prevention of PPH Mex 2002;70: practices
Arechiga M, N. L. Mexico 572–5
Batiza Resendiz
30 August 2006 14:20:36

V, et al.
Caliskan E, Social Security Council: Is rectal misoprostol really Am J Obstet 8 1606 396 600 µg rectal [1] 401 [1] 10 IU oxytocin
Meydanli MM, Maternity and Women’s effective in the treatment of Gynecol 2002; [2] 407 i.v. plus 600 µg
Dilbaz B, et al. Health Teaching Hospital, third stage of labor? A 187:1038–45 [3] 402 misoprostol rectal
Kucukesat, Ankara, Turkey randomized controlled trial [2] 10 IU oxytocin
i.v.
[3] 10 IU oxytocin
i.v. plus 1 ml
methylergometrine
i.m.
Karkanis SG, University of Toronto, Randomized controlled trial J Obstet 8 214 110 400 µg rectal 113 5 IU oxytocin i.v.
Caloia D, Toronto, Canada of rectal misoprostol vs. Gynaecol Can or 10 IU oxytocin
Salenieks ME, oxytocin in third stage 2002;24: i.m.
et al. management 149–54
Kundodyiwa Dept. of Obstetrics and Misoprostol vs. oxytocin in Int J Gynaecol 8 499 243 400 µg oral 256 10 IU oxytocin i.m.
TW, Majoko F, Gynecology, University the third stage of labor Obstet 2001;
165
Rusakaniko S of Zimbabwe, Harare, 75:235–41
187
Zimbabwe
Benchimol M, Centre de Gynecologie Role of misoprostol in the J Gynecol 8 600 200 600 µg oral [1] 200 [1] 2.5 IU oxytocin
Gondry J, Obstetrique, Amiens, delivery outcome Obstet Biol [2] 200 i.v.
Mention JE, France Reprod (Paris) [2] placebo
et al. 2001;30:
576–83
Gerstenfeld TS, Women’s and Children’s Rectal misoprostol vs. Am J Obstet 8 325 159 400 µg rectal 166 20 IU oxytocin i.v.
Wing DA Hospital, Dept. of Obstetrics intravenous oxytocin for Gynecol 2001;
and Gynecology, University the prevention of PPH after 185:878–82

of Southern California Keck vaginal delivery
School of Medicine, Los
Angeles, USA
Gulmezoglu WHO Collaborative Group WHO multicenter randomized Lancet 2001; 18 530 9264 600 µg oral 9266 10 IU oxytocin i.m.
AM, Villar J, to Evaluate Misoprostol in trial of misoprostol in the 358:689–95 or i.v.
Ngoc NT, et al. the Management of the management of the third stage
Third Stage of Labour of labor
Continued
30 August 2006 14:20:37
Table 1 Continued
Hofmeyr GJ, Dept. of Obstetrics and Side-effects of oral S Afr Med J 8 600 300 600 µg oral 300 placebo
Nikodem VC, Gynecology, Coronation misoprostol in the third stage 2001;91:
de Jager M, Hospital and Effective Care of labor – a randomized 432–5

et al. Research Unit, University placebo-controlled trial
of the Witwatersrand,
Johannesburg, South Africa
Bugalho A, Maternity of the Hospital Misoprostol for prevention of Int J Gynaecol 8 663 324 400 µg rectal 339 10 IU oxytocin i.m.
Daniel A, Central de Maputo, Maputo, PPH Obstet 2001;
Faundes A, Mozambique 73:1–6
et al.
Ng PS, Chan Dept. of Obstetrics and A multicenter randomized Hum Reprod 8 2058 1026 600 µg oral 1032 1 ml syntometrine
AS, Sin WK, Gynecology, The Chinese controlled trial of oral 2001;16: i.v. (5 IU
166
188
et al. University of Hong Kong, misoprostol and i.m 31–5 syntocinone and
Prince of Wales Hospital, syntometrine in the 0.5 mg ergometrine
Shatin, New Territories management of the third maleate 0.5 mg)
stage of labor
Walley RL, Dept. of Obstetrics and A double-blind placebo BJOG 2000; 8 401 203 400 µg oral 198 10 IU oxytocin i.m.
Wilson JB, Gynaecology, St John’s, controlled randomized trial 107:1111–15
Crane JM, et al. Memorial University of of misoprostol and oxytocin in
Newfoundland, Canada the management of the third
stage of labor
El-Refaey H, Dept. of Obstetrics and The misoprostol third stage BJOG 2000; 8 1000 501 500 µg oral 499 standard oxytocic

Nooh R, Gynaecology, University of labor study: a randomized 107:1104–10 regimens (10 IU
O’Brien P, et al. College Hospital, London, controlled comparison between oxytocin or 0.5 mg
UK orally administered misoprostol ergometrine or
and standard management 1 ml syntometrine)
Cook CM, Dept. of Obstetrics and A randomized clinical trial Aust N Z J 8 863 424 400 µg oral 439 standard oxytocic
Spurrett B, Gynaecology, University comparing oral misoprostol Obstet regimens (10 IU
Murray H of Sydney at Nepean with synthetic oxytocin or Gynaecol oxytocin i.m. or
Hospital Penrith, New syntometrine in the third 1999;39: 1 ml syntometrine
South Wales, Australia stage of labor 414–19 i.m.)
Amant F, Spitz Dept. of Obstetrics and Misoprostol compared with Br J Obstet 8 200 100 600 µg oral 100 0.2 mg
B, Timmerman Gynaecology, University methylergometrine for the Gynaecol methylergometrine
D, et al. Hospitals Leuven, Belgium prevention of PPH: a 1999;106: i.v.
double-blind randomized trial 1066–70
30 August 2006 14:20:37

Surbek DV, Dept. of Obstetrics and Oral misoprostol for the third Obstet Gynecol 8 65 31 600 µg oral 34 placebo
Fehr PM, Gynecology, University stage of labor: a randomized 1999;94:
Hosli I, et al. of Basel, Switzerland placebo-controlled trial 255–8
Bamigboye AA, Dept. of Obstetrics and Rectal misoprostol in the Am J Obstet 8 546 271 400 µg rectal 275 placebo
Hofmeyr GJ, Gynecology, Coronation prevention of PPH: a Gynecol 1998;
Merrell DA Hospital, and University of placebo-controlled trial 179:1043–6
the Witwatersrand,
Hofmeyr GJ, Dept. of Obstetrics and A randomized placebo Br J Obstet 8 500 250 400 µg oral 250 placebo
Nikodem VC, Gynaecology, Coronation controlled trial of oral Gynaecol
de Jager M, Hospital and University of misoprostol in the third 1998;105:
et al. the Witwatersrand, stage of labor 971–5
Bamigboye AA, Dept. of Obstetrics and Randomized comparison of Acta Obstet 8 491 241 400 µg rectal 250 1 ml syntometrine
Merrell DA, Gynecology, Natalspruit rectal misoprostol with Gynecol Scand i.m. (5 IU
167
Hofmeyr GJ, Hospital and the University syntometrine for management 1998;77: syntocinone and
189
et al. of the Witwatersrand, of third stage of labor 178–81 0.5 mg ergometrine
Johannesburg, South Africa maleate 0.5 mg)
El-Refaey H, Dept. of Obstetrics and Use of oral misoprostol in the BJOG 1997; 8 237 237 600 µg oral 0 –
O’Brien P, Gynaecology, University prevention of PPH 104:336–9
Morafa W, et al. College Hospital, London,
UK
PPH, postpartum hemorrhage

Table 2 Misoprostol for treatment
30 August 2006 14:20:37

Prata N, Bixby Population Program, Controlling PPH after home Int J Gynaecol 849 454 1000 µg rectal 395 current practices
Mbaruku G, School of Public Health, births in Tanzania Obstet 2005;
Campbell M, University of California, 90:51–5
et al. Berkeley, USA
Walraven G, Reproductive Health Misoprostol in the treatment BJOG 2004; 160 79 600 µg 200 µg oral 81 placebo
Dampha Y, Programme, Medical of PPH in addition to routine 111:1014–17 and 400 µg

Bittave B, et al. Research Council management: a placebo sublingual
Laboratories, Farafenni, randomized controlled trial
The Gambia, South Africa
Hofmeyr GJ, Effective Care Research Misoprostol for treating PPH: BMC 238 117 1000 µg 200 µg oral 121 placebo
Ferreira S, Unit, University of a randomized controlled trial Pregnancy and 400 µg
Nikodem VC, Witwatersrand and Fort [ISRCTN72263357] Childbirth sublingual
et al. Hare, and East London 2004;4:16 and 400 µg
Hospital Complex, East rectal
London, South Africa
168
190
Shojai R, Service de gynecologie- [Rectal misoprostol for PPH] Gynecol Obstet 41 41 1000 µg rectal 0 –
Desbriere R, obstetrique, CHU Nord, Fertil 2004;
Dhifallah S, et al. Marseille, France 32:703–7
Lokugamage Dept. of Obstetrics & A randomized study Acta Obstet 64 32 800 µg rectal 32 1 ml syntometrine
AU, Sullivan Gynaecology, Royal Free comparing rectally Gynecol Scand i.m. (5 IU
KR, Niculescu and University College administered misoprostol 2001;80: syntocinone and
I, et al. London School, London, versus Syntometrine combined 835–9 0.5 mg ergometrine
UK with an oxytocin infusion for maleate) plus
the cessation of primary PPH 10 IU oxytocin i.v.
Abdel-Aleem H, Dept. of Obstetrics & Management of severe PPH Int J Gynaecol 18 18 600 µg or rectal 0 –

El-Nashar I, Gynecology, Faculty of with misoprostol Obstet 2001; 1000 µg
Abdel-Aleem A Medicine, Assiut University, 72:75–6
Assiut, Egypt
O’Brien P, Dept. of Obstetrics and Rectally administered Obstet Gynecol 14 14 1000 µg rectal 0 –
El-Refaey H, Gynaecology, University misoprostol for the treatment 1998;92:
Gordon A, et al. College Hospital, of PPH unresponsive to 212–14
London, UK oxytocin and ergometrine: a
descriptive study
PPH, postpartum hemorrhage

Section V
Hospital preparation
191
30 August 2006 14:20:37
20
RESUSCITATION
M. J. Cowen
INTRODUCTION in the blood bank, all of whom must be alerted

at an early stage, as any undue delay or failures
The key feature of all resuscitation efforts is
of communication at the initial stage will invari-
the early recognition and management of
ably result in a poor outcome. In particular, the
hypovolemic shock with an overall objective
prompt involvement of experienced senior anes-
of restoring the circulating blood volume
thetists is mandatory along with intensive care
to maintain normal tissue perfusion and
back-up facilities (see Chapters 13 and 22).
oxygenation.
The most important first step is to secure good
The reason for the high mortality associated
venous access whilst veins are still available and
with obstetric hemorrhage is simple, i.e. the
before shut-down occurs, preferably by two
delayed recognition of hypovolemia and failure
large-bore cannulae, i.e. 14 gauge (flow rate
to provide adequate volume resuscitation.
315 ml/min) or 16 gauge (210 ml/min). The
Common problems include the failure to recog-
importance of cannula size and flow rate cannot
nize risk factors, frequent under-estimation of
be overstated, and all too often very small
the degree of blood loss, and failure to involve
cannulae are inserted with poor flow rates (e.g.
key personnel early enough. These problems
20 gauge cannula flow rate = 65 ml/min) with
can be operational even in developed nations,
disastrous results. In those circumstances where
such as the UK and the US1–3.
veins are collapsed, however, a small cannula is
better than nothing.
GENERAL CONSIDERATIONS The use of a cannula and fluid infusion are the
mainstay of treatment. Most of the other activities
All resuscitation strategies include two principal
required, including monitoring and laboratory
objectives:
sampling, whilst important are not actually
(1) Achievement of hemostasis by arresting treatment. It is easy in the busyness of the emer-
the source of bleeding by whatever means gency scene to be distracted by these peripheral
necessary, including surgical intervention activities at the expense of treatment, i.e. intra-
with or without anesthesia; venous fluid therapy, and it is important for the
team leader to keep a sense of perspective if
(2) Restoration of an adequate circulating
there is to be a good outcome.
blood volume by maintaining a normal
The second most important step is to send
blood pressure and urine output (> 30 ml/h
an urgent blood sample to the laboratory for
in adults) (0.5 ml/kg/h).
baseline readings of full blood count, including
The early recognition of problems followed by hemoglobin, hematocrit and platelet count, plus
an immediate response is of paramount impor- clotting tests, including prothrombin time (PT),
tance, as mobilization of the key personnel and activated partial thromboplastin time (APTT)
equipment takes time. It is not possible for one and fibrinogen levels as a baseline. A biochemi-
individual to do all the necessary work and there cal profile and blood gas analysis are also useful,
is no place for acting solo in isolation. Key play- especially to measure the level of acidosis and
ers must include senior obstetricians, midwives, base deficit. Because of the often rapid changes
anesthetists, hematologists and laboratory staff occurring, it is essential throughout the episode
170
192
30 August 2006 14:20:38
Resuscitation
to keep close liaison with hematologists and the postpartum hemorrhage is that there may be lit-
blood transfusion laboratory. tle apparent fall in blood pressure until a very
Observations, monitoring and readings must late stage. This is because most patients are
be kept continuously, preferably by a team often young and previously fit and their cardio-
member or midwife solely dedicated to this vascular system will compensate until a very late
activity. The most useful parameter of all is the stage when it suddenly crashes and
trend recorded on the basic TPR observation decompensates.
chart, with particular attention to pulse rate, Table 1 shows a useful staging of schemes to
blood pressure and urine output, because these assess the degree of obstetric hemorrhage.
will provide the overall picture at a glance. In cases of severe shock, blood transfusion
In severe cases, a central line, preferably will normally be required if 30–40% of blood
multi-lumen, inserted in the neck, subclavian volume is lost, whilst, if over 40% blood
or femoral zones is invaluable, not only for volume is lost, the situation is immediately
monitoring of central venous pressure as a guide life-threatening4.
to hypovolemia and as an assist in determining
infusion requirements, but also as a good
TRANSFUSION CONSIDERATIONS
venous access for rapid infusion especially when
peripheral veins are not available or collapse. Blood transfusions are not without risk and it is
In most instances, this is not a first priority, essential that they are restricted to those in real
especially considering that routine labor ward need. Because of the known risks, which include
personnel may not have the skill to perform this incompatibility reactions, and infectious disease
safely. transmission, an initial attempt to correct hypo-
Insertion of an arterial line is useful to volemia should be made by using crystalloid
monitor blood gases, acidosis, and base deficit, solutions (e.g. normal saline, Ringers lactate) or
and also to serve as a direct dynamic blood colloids (gelatins, starches).
pressure monitor in the intensive care situation. No absolute parameters can help one to
This too is not, however, a priority in the decide when to start transfusing blood but
immediate management, although it will be clinical commonsense indicates that blood
useful later, especially in severe cases. Not should be strongly considered after 2 liters of
all patients require blood and, in deciding crystalloid and 1 liter colloid have been given if
which fluid to use, the hazards and risks of the cardiovascular system remains unstable with
blood transfusion warrant an initial attempt to continued bleeding.
correct hypovolemia using crystalloid or colloid Blood will invariably be required if > 40%
solutions. blood volume is lost, which correlates in a 70 kg
Having said this, many patients will require adult to approximately 2.5 liters measured
rapid infusion. Proper equipment must be blood loss.
readily available and include a rapid infuser More precise arbiters are the measurement of
pump and/or pressure infusion bags, and prefer- hemoglobin and hematocrit decrease, results
ably an individual team member dedicated to of which can be available almost immediately
this activity. It is also important for all fluids using rapid I-stat analyzers. The long-standing
to be warmed since the rapid infusion of cold practice tradition that suggested that ideally the
fluids will create hypothermia and increase the hemoglobin should be maintained at a level of
chances of disseminated intravascular coagula- 10 g/dl has changed, with rising awareness of
tion and clotting failure. infection risks from donated blood products,
and this has led to a re-evaluation of require-
ments5–7. Recent work shows that healthy
THE RECOGNITION OF
patients can tolerate a hemoglobin concentra-
HYPOVOLEMIA
tion as low as 7 g/dl8. Oxygen delivery will still
The loss of circulating blood volume results be maintained as long as the patient is
in the signs of tachycardia, hypotension and normovolemic. More recently, the American
oliguria. The most dangerous feature of Society of Anesthesiologists Taskforce on Blood
171
193
30 August 2006 14:20:38
Table 1 Staging scheme for assessment of obstetric hemorrhage
Severity of shock Findings % blood loss
None None < 15–20%

Mild Tachycardia (< 100 bpm) < 20–25%
Mild hypotension
Peripheral vasoconstriction
Moderate Tachycardia (100–120 bpm) < 25–35%
Hypotension (systolic blood pressure 80–100 mmHg)
Restlessness
Oliguria
Severe Tachycardia (> 120 bpm) > 35%
Hypotension (systolic blood pressure < 60 mmHg)
Altered consciousness
Anuria
From Gonik B. Intensive care monitoring of the critically ill patient. In Creasy RK, Resnik R, eds. Maternal-
Fetal Medicine, 3rd edn. Philadelphia: WB Saunders, 1994:865–90
Component Therapy concluded that transfu- blood, or in appropriate cases for 20–30 min for
sion is rarely indicated when the hemoglobin fully cross-matched blood. In general, one unit
level is > 10 g/dl, but almost always indicated of red cells increases the hemoglobin by approx-
when it is < 6 g/dl9. It therefore seems reason- imately 1.5 g/dl and the hematocrit by 5% in a
able to conclude that transfusion should be 70 kg woman. In the dynamic bleeding situa-
commenced in most obstetric patients after a tion, however, it is appropriate to calculate
postpartum hemorrhage if the hemoglobin level transfusion requirements so as to restore the
falls below 7 g/dl10, but that may be difficult to hemoglobin to approximately 10 g/dl, albeit
assess in real time under emergency situations. using blood conservatively.
In these circumstances, it is useful to have
an agreed Action Plan posted in the labor and
Platelets
delivery ward, as published guidelines on trans-
fusion practice are not easily referred to in the Platelets should be transfused only to prevent or
emergency situation11–13. As an example, Table correct bleeding associated with a decrease in
2 is an example of current practice which is platelet count or abnormality of platelet func-
in use at the author’s hospital (taken from a tion as they are difficult to store and in short
national guideline14 with local modifications15). supply16.
ADDITIONAL CONSIDERATIONS FOR Fresh frozen plasma

BLOOD PRODUCTS
Fresh frozen plasma (FFP) maintains the viabil-
Additional discussion is provided in Chapter ity of all clotting factors. The only clear indica-
25. tion for FFP is the replacement of coagulation
factors in clotting disorders17. FFP should be
given if the PT and APTT exceed 1.5 times the
Red cells
control level in the presence of continuous
If blood is required instantly, it may be neces- bleeding. In situations with massive bleeding,
sary to use the emergency supply of Group O however, it may be necessary to give FFP even
Rh(D)-negative uncross-matched stock, but before clotting results are available. The dose
preferable to wait 5–10 min for type-specific required is 12–15 ml/kg or normally 4 units for
172
194
30 August 2006 14:20:38
Resuscitation
Table 2 Acute massive blood loss: a template guideline
Immediate actions Key points Other considerations
● Arrest bleeding ● Early surgical or obstetric intervention

● Upper G/I tract procedures
● Interventional radiology
● Contact key personnel ● Most appropriate surgical team
● Duty anesthetist
● Blood bank
● Restore circulating volume ● Insert wide-bore peripheral cannulae ● Blood loss is often
N.B. In patients with major ● Give adequate volumes of underestimated
vessel or cardiac injury, crystalloid/blood ● Refer to local guidelines for the
it may be appropriate to ● Aim to maintain normal blood resuscitation of trauma patients
restrict volume replacement pressure and urine output > 30 ml/h and for red cell transfusion
after discussion with surgical in adults (or 0.5 ml/kg/h) ● Monitor CVP if
team hemodynamically unstable
● Request laboratory ● FBC, PT, APTT, fibrinogen; blood ● Take samples at earliest
investigations bank sample, biochemical profile, opportunity as results may be
blood gases affected by colloid infusion
● Repeat FBC, PT, APTT, fibrinogen ● Misidentification is most
every 4 h, or after one-third blood common transfusion risk
volume replacement, or after infusion ● May need to give FFP &
of FFP platelets before the FBC and
coagulation results available
● Request suitable red cells ● Blood needed immediately – use ● Contact blood transfusion
N.B. All red cells are now ‘Emergency stock’ group O Rh laboratory or oncall BMS and
leukocyte-depleted. The (D)-negative provide relevant details
volume is provided on each ● Blood needed in 5–10 min – type-specific ● Collect sample for group and
pack, and is in the range of will be made available to maintain cross-match before using
190–360 ml O Rh (D)-negative stocks emergency stock
● Blood needed in 30 min or longer – fully ● Blood warmer indicated if large
cross-matched blood will be provided volumes are transfused rapidly
● Consider the use of ● Anticipate platelet count < 50 × 109/l ● Target platelet count:-
platelets after > 2 liters blood loss with continued > 100 × 109/l for multiple/CNS
bleeding trauma
● Dose: 10 ml/kg body weight for a > 50 × 109/l for other situations
neonate or small child, otherwise one ● Consider early use of platelets
‘adult therapeutic dose’ (one pack) if clinical situation indicates
continued excessive blood loss
despite the count
● Consider the use of FFP ● Anticipate coagulation factor deficiency ● PT/APTT > 1.5 × mean
after > 2 liters blood loss with continued control correlates with
bleeding increased surgical bleeding
● Aim for PT & APTT < 1.5 × mean ● May need to use FFP before
control laboratory results available:
● Allow for 20-min thawing time take sample for PT, APTT,
● Dose: 12–15 ml/kg body wt = 1 liter fibrinogen before FFP
or 4 units for an adult transfused
continued
173
195
30 August 2006 14:20:38
Table 2 Continued
Immediate actions Key points Other considerations
● Consider the use of ● To replace fibrinogen & FVIII ● Fibrinogen < 0.5 strongly
cryoprecipitate ● Aim for fibrinogen > 1.0 g/l associated with microvascular
● Allow for 20-min thawing time bleeding
● Dose: 10 packs or 1 pack/10 kg in
children
● Suspect DIC ● Treat underlying cause if possible ● Shock, hypothermia, acidosis,
risk of DIC
● Mortality if DIC is high
For abbreviations, see text
an adult, and the objective should be to aim for coagulopathy wherein a vicious cycle consumes
a PT and APTT less than 1.5 control level. FFP clotting factors and platelets rapidly. DIC can
requires a thawing time of 20 min, and hence develop dramatically in obstetric patients, espe-
early anticipation of a potential requirement is cially in association with placental abruption
helpful. and amniotic fluid embolism. It also occurs
suddenly after massive bleeding with shock,
acidosis and hypothermia. This latter risk
Cryoprecipitate
emphasizes the importance of warming all
It is appropriate to administer cryoprecipitate infused fluids whenever possible. DIC carries
which contains fibrinogen and factor VIII a high mortality and, once established, can
when there is evidence of a consumptive be difficult to reverse. Patients with prolonged
coagulopathy with a fibrinogen level less hypovolemia are particularly at risk. The diag-
than 0.5 g/l. The normal dose is 10 units. As nosis can be made by frequent estimation of
with FFP, cryoprecipitate needs thawing time. platelets, fibrinogen, PT and APTT. Treatment
The aim is to restore the fibrinogen level to consists of administering platelets, FFP and
> 1.0 g/l. cryoprecipitate sooner rather than later.
Coagulopathy Complications of blood transfusion

Coagulopathy can develop rapidly in an obstet- Increasing awareness of the risks of transfusion
ric patient. Confirmatory laboratory tests are has led to diminished use of blood and blood
required for precise diagnosis, but in the clinical products in recent years. Complications can
setting of postpartum hemorrhage the presence occur because of incompatibility, storage prob-
of microvascular bleeding is a good clinical indi- lems, and transmission of infection.
cator18,19. Absence of clotting with continued The most common cause of a transfusion-
bleeding strongly suggest a coagulopathy. related death is incompatibility leading to a
Hemostasis is normally adequate when clotting hemolytic reaction22. Most of such deaths are
factors are greater than 30% of normal18–21. If due to misidentification and are entirely pre-
bleeding continues in the presence of clotting ventable, emphasizing the importance of safe
factors > 30% normal and a PT and APTT less systems for cross-checking all blood products.
than 1.5 times control level, it is unlikely that Storage problems include hyperkalemia, as
low coagulation levels are responsible18,19. potassium levels rise in stored blood which, if
given rapidly and repeatedly, can give rise to
hyperkalemia, especially in an acidotic, hypo-
Disseminated intravascular coagulopathy
thermic patient. Similarly, hypothermia can
Disseminated intravascular coagulopathy increase if large volumes of cold stored blood
(DIC) represents the most deadly form of are given rapidly without a blood warmer.
174
196
30 August 2006 14:20:38
Resuscitation
The transmission of infection is arguably the be considered for Jehovah Witness patients (see
most feared complication especially in terms of Chapter 15 for full discussion of perioperative
HIV, hepatitis B and C and cytomegalovirus salvage).
(CMV). Estimated HIV transmission risks vary In the technique of hemodilution,
widely from 1 in 200 000 to 1 in 2 000 000 500–1000 ml blood may be collected and
transfusions23. But the most common trans- reinfused later; however, overall experience in
mission is of viral hepatitis, although this is massive postpartum hemorrhage is limited29,30.
decreasing with improved screening. Currently,
the incidence is 1 per 103 000 units of blood
ANESTHETIC CONSIDERATIONS
transfused23. CMV is carried in asymptomatic
donors in the neutrophil. CMV infection can be Postpartum hemorrhage is the most frequent
prevented by using CMV-negative blood or by reason for emergency surgery and anesthesia in
eliminating neutrophils from donor blood24. the postpartum period. The principal causes
include uterine atony, trauma, retained placenta
and uterine inversion, all of which are discussed
Alternatives to transfusion
in detail in other parts of this book. A large pro-
Three alternative methods of autologous portion of these will require anesthesia as part of
transfusion are presently available: preoperative the therapy to arrest the hemorrhage.
donation antepartum, perioperative cell salvage, The choice of anesthetic will be dictated by
and hemodilution. Rarely, if ever, are these circumstances, the degree of blood loss and the
feasible in the unexpected massive postpartum urgency of the situation. A general anaesthetic
hemorrhage, but they nevertheless merit consid- is preferable in most instances of significant
eration especially when treating patients who postpartum hemorrhage with hypovolemia. The
are adherent to the Jehovan Witness belief. problem in using a regional block is that unrec-
Antepartum donation may be considered for ognized hypovolemia in combination tends to
high-risk patients and for those with rare blood aggravate hypotension and increase maternal
types, but it is recommended that, before dona- morbidity and mortality. However, if a patient
tion, the hemoglobin should not be less than is already receiving a regional block (spinal or
11 g/l and the hematocrit 33%25–27. However, epidural), bleeding is controlled and the cardio-
many obstetric patients may not be able to vascular system stable, it may be appropriate to
donate more than one unit of blood, whereas continue with a regional technique. If instability
most patients requiring blood after postpartum occurs in such circumstances, early conversion
hemorrhage require considerably more than one to a general anesthetic is indicated.
unit and thus would need homologous blood. Crucial items for the safe conduct of an
Furthermore, such patients are difficult to pre- anaesthetic include the involvement of experi-
dict. Accordingly, preoperative donation may enced senior/consultant anesthetists and
not be beneficial or even cost-effective taking additional helpers, pre-sited two wide-bore
into account the low frequency of blood transfu- cannulae, knowledge of hemoglobin/hematocrit
sion even in high-risk patients and the difficulty levels, rapid infusion devices and fluid warmers,
of predicting these in advance27. immediate availability of crystalloid and colloid
Perioperative blood salvage is a technique infusions and, as soon as possible, blood and
of scavenging blood lost during an operation, blood products especially FFP, and, finally,
washing it and then transfusing the scavenged available equipment for central venous access
red cells28. Of concern is that washing may not and direct arterial line monitoring.
adequately remove amniotic fluid and fetal A suitable general anesthetic technique
debris which, when re-transfused, may precipi- includes pre-oxygenation and rapid sequence
tate the anaphylactoid amniotic fluid embolism induction with cricoid pressure using either
response. Blood salvage may nevertheless be thiopentone in reduced dose (e.g. 4 mg/kg) or
appropriate in cases of massive obstetric hemor- ketamine (1 mg/kg) or etomidate (0.2 mg/kg),
rhage when blood bank resources are limited. followed by intubation after suxamethonium.
Where the technique is available, it should also Maintenance agents will include further muscle
175
197
30 August 2006 14:20:38
relaxants (e.g. rocuronium 0.6 mg/kg) with

nitrous oxide, oxygen and either a very low con-
centration of volatile anesthetic (e.g. isoflorane)
to combat awareness, or possibly opiates such as
fentanyl, alfentanil or remifentanil.
In some circumstances, e.g. uterine inversion
where intensive relaxation is required, an addi-
tional volatile agent may be helpful. Equipotent
doses of all volatile halogenated agents produce
similar degrees of uterine relaxation31,32. Other
alternatives include use of nitroglycerine given
intravenously33,34.
CARDIOPULMONARY
RESUSCITATION
The prognosis is poor in the event of cardiac
arrest in a patient with severe hypovolemia
after a postpartum hemorrhage because of
hypoxemia and rapidly accelerating acidosis.
Nevertheless, most patients are young and pre-
viously fit, as no attempts should be spared to
resuscitate.
Cardiac arrest will present with sudden loss
of consciousness, absent major pulses and
absent respiration. Response needs to be imme- Figure 1 Adult basic life support (Resuscitation
diate to have any chance of success and should Council, UK)
follow the agreed Cardiac Arrest Procedure
along conventional lines in three phases, e.g.
UK Resuscitation Guidelines as in Figures 1
and 2. Three items are of crucial importance:
(1) Basic life support – the ABC system. This (1) External cardiac massage must be com-
includes Airway control, Breathing support menced without delay if there are no palpa-
and Circulatory support. ble major pulses;
(2) Advanced life support. This includes (2) Adrenaline 1 mg given every 3 min will fre-
intubation and ventilation, continued quently be required;
circulatory support often with epinephrine (3) Given that the root cause of the arrest is
(adrenaline), defibrillation and ECG moni- hypovolemia, vigorous attempts to restore a
toring, drugs and fluids, and management circulatory blood volume must be contin-
of complex arrhythmias. ued throughout the cardiopulmonary resus-
(3) Prolonged life support, including all citation process if there is to be any chance
intensive care systems. of success.
176
198
30 August 2006 14:20:46
Resuscitation
Figure 2 Advanced life support algorithm for the management of cardiac arrest in adults (Resuscitation
Council UK). BLS, basic life support; VF, ventricular fibrillation; VT, ventricular tachycardia; CPR,
cardiopulmonary resuscitation; ETT, endotracheal tube
References Kingdom. London: Department of Health,

HMSO, 2004
1. Rochat RW, Koonin LM, Atrash HK, et al. 4. American College of Surgeons. Advanced
Maternal mortality in the United States: report Trauma Life Support Course Manual. Chicago:
from the maternal mortality collaborative. Obstet American College of Surgeons, 1997:103–12
Gynecol 1988;72:91 5. Combs CA, Murphy EL, Laros RK. Cost-
2. Li XF, Fortney JA, Kotelchuck M, Glover LH. benefit analysis of autologous blood donation in
The postpartum period: the key to maternal obstetrics. Obstet Gynecol 1992;80:621–5
mortality. Int J Gynaecol Obstet 1996;54:1–10 6. Camann WR, Datta S. Red cell use during
3. Why Mothers Die 2000–2002. Confidential cesarean delivery. Transfusion 1991;31:12–15
Enquiries into Maternal Deaths in the United
177
199
30 August 2006 14:20:50
7. Consensus Conference. The impact of routine 22. Honig CL, Bove JR. Transfusion associated
HLTV-III antibody testing of blood and plasma fatalities: Review of Bureau of Biologics report.
donors on public health. JAMA 1986;256: Transfusion 1980;20:653–6
1178–80 23. Goodnough LT, Brecher ME, Kanter MH,
8. Consensus Conference. Perioperative red blood AuBuchon JP. Transfusion medicine. 1. Blood
cell transfusion. JAMA 1988;260:2700–3 transfusion. N Engl J Med 1999;350:438–47
9. American Society of Anaesthesiologists Task 24. Pamphilon DH, Rider JH, Barbara JA, William-
Force. Practice Guidelines for Blood son LM. Prevention of transfusion-transmitted
Component Therapy. Anesthesiology 1996;84: cytomegalovirus infection. Transfus Med 1999;9:
732–47 115–23
10. Chestnut DH, ed. Antepartum and postpartum 25. Droste S, Sorensen T, Price T, et al. Maternal
hemorrhage. In Obstetric Anesthesia: Principles and fetal hemodynamic effects of autologous
and Practice. Amsterdam: Elsevier Mosby, 2004: blood donation during pregnancy. Am J Obstet
676–7 Gynecol 1992;167:89–93
11. British Committee for Standards in Haematol- 26. Kruskall MS, Leonard S, Klapholz H.
ogy. Guidelines for transfusion for massive blood Autologous blood donation during pregnancy:
loss. Clin Lab Haematol 1988;10:265–73 analysis of safety and blood use. Obstet Gynecol
12. British Committee for Standards in Haematol- 1987;70:938–40
ogy. Guidelines for the use of fresh frozen 27. Andres RL, Piacquadio KM, Resnick R. A reap-
plasma. Transfus Med 1992;2:57–63 praisal of the need for autologous blood donation
13. British Committee for Standards in Haematol- in the obstetric patient. Am J Obstet Gynecol
ogy. Guidelines for platelet transfusions. Transfus 1990;163:1551–3
Med 1992;2:311–18 28. Williamson KR, Taswell HF. Intraoperative
14. Stainsby D, MacLennan S, Hamilton PJ. blood salvage. A review. Transfusion 1991;31:
Management of massive blood loss: a template 662–75
guideline. Br J Anaesth 2000;85:487–91 29. Estella NM, Berry DL, Baker BW, et al. Normo-
15. Milton Keynes General NHS Trust. Acute mas- volemic hemodilution before cesarean hyster-
sive blood loss – a template guideline. 2002:1–10 ectomy for placenta percreta. Obstet Gynecol
16. Consensus Conference. Platelet transfusion 1997;90:669–70
therapy. JAMA 1987;257:1777–80 30. Grange CS, Douglas MJ, Adams TJ, Wadsworth
17. Transfusion alert: Indications for the use of red LD. The use of acute hemodilution in
blood cells, platelets, and fresh frozen plasma. parturients undergoing cesarean section. Am J
US Department of Health and Human Services, Obstet Gynecol 1998;178:156–60
Public Health Service, National Institutes of 31. Munson ES, Embro WJ. Enflurane, isoflurane,
Health, 1989 and halothane and isolated human uterine
18. Ciaverella D, Reed RL, Counts RB, et al. muscle. Anesthesiology 1977;46:11–14
Clotting factor levels and the risk of diffuse 32. Turner RJ, Lambros M, Keyway L, Gatt SP.
microvascular bleeding in the massively trans- The in-vitro effects of sevoflurane and desflurane
fused patient. Br J Haematol 1987;67:365–8 on the contractility of pregnant human uterine
19. Murray DJ, Olson J, Strauss R, et al. Coagulation muscle. Int J Obstet Anesth 2002;11:246–51
changes during packed red cell replacement of 33. Altabef KM, Spencer JT, Zinberg S. Intravenous
major blood loss. Anesthesiology 1988;69:839–45 nitroglycerin for uterine relaxation of an inverted
20. Consensus Conference. Fresh-frozen plasma: uterus. Am J Obstet Gynecol 1992;166:1237–8
indications and risks. JAMA 1985;253;551–3 34. Bayhi DA, Sherwood CDA, Campbell CE.
21. Aggeler PM. Physiological basis for transfusion Intravenous nitroglycerin for uterine inversion.
therapy in hemorrhagic disorders: a critical J Clin Anesth 1992;4:487–8
review. Transfusion 1961;1:71–85
178
200
30 August 2006 14:20:50
21
EQUIPMENT TRAY FOR POSTPARTUM HEMORRHAGE
T. F. Baskett
Primary postpartum hemorrhage is most often and suture of cervical and/or vaginal lacera-
due to uterine atony which usually responds to tions11. The most common predisposing causes
the appropriate application of oxytocic drugs. In of its use were placenta previa, with or without
a minority of cases, however, the atonic uterus partial accreta, and uterine atony refractory to
will not contract with any uterotonic agents, oxytocic agents11.
particularly in cases of prolonged and aug- The contents of an obstetric hemorrhage tray
mented labor with an exhausted and infected are shown in Table 1. As individual obstetric
uterus. In these instances, a variety of surgical units undoubtedly have a varying availability
techniques may be necessary, including uterine of supplies, local conditions may modify these
tamponade with packing1 or balloon devices2–4, contents. Three vaginal retractors are necessary
uterine compression sutures5–8, major vessel for access to and exposure of high vaginal and or
ligation9,10, and hysterectomy, all of which are cervical lacerations. Heaney or Breisky–Navratil
discussed in detail in other chapters of this
book. In addition to uterine atony unresponsive
Table 1 Contents of obstetric hemorrhage equip-
to oxytocic agents, numerous other causes of ment tray
postpartum hemorrhage may require surgical
intervention with more equipment than is avail- Access/exposure
able in the standard vaginal delivery or Cesar- ● Three vaginal retractors (Heaney,
ean section packs. These include high vaginal or Breisky–Navratil)

● Four sponge forceps
cervical lacerations with poor exposure, placenta
previa and/or placenta accreta at the time of Eyed needles
Cesarean section, and uterine rupture. In most ● straight 10 cm
obstetric units, and for the individual obstetri- ● curved 70–80 mm, blunt point
cian and nursing personnel who work there, the

Sutures
additional equipment and instruments for these
● No. 1 polyglactin (vicryl)
surgical techniques are rarely used. Thus, when ● O and No. 2 chromic catgut with curved needle
they are needed they may not be readily avail- ● Ethiguard curved, blunt point monocryl
able and valuable time will be lost searching for

them. For these reasons, every obstetric unit Uterine/vaginal tamponade
should have a readily available, sterile ‘obstetric ● Vaginal packs
● Kerlix gauze roll
hemorrhage equipment tray’ upon which is
● Uterine balloon (depending on local availability):
placed all the necessary material for surgical
Sengstaken–Blakemore, Rüsch urological balloon,
management of postpartum hemorrhage. Bakri balloon, surgical glove and catheter,
Experience with one such equipment tray in condom and catheter
a large Canadian unit has shown it is used
in about 1 in 250 Cesarean deliveries and 1 in Diagrams (Figures 1–4)
1000 vaginal deliveries11. The most common Pages with diagrams and instructions:
● Uterine and ovarian artery ligation
surgical techniques that called for use of the
● Uterine compression suture techniques: B-Lynch,
tray were uterine compression sutures, uterine
square and vertical
tamponade, uterine and ovarian artery ligation,
179
201
04 September 2006 14:04:56
vaginal retractors are suitable for this purpose. standard vaginal packing tied together can be
Four sponge forceps are useful to identify and adequate, but the ideal is a kerlix gauze roll
compress cervical lacerations, to provide com- which has a thicker six-ply gauze than the
pression to the edges of extensive vaginal four-ply of the usual vaginal pack. In recent
lacerations or to uterine edges at the time of years, balloon tamponade has also been used
laparotomy for uterine rupture. Standard pack- for uterine atony unresponsive to oxytocic
aged suture material often contains needles that drugs following vaginal delivery. A variety of
are too small for the placement of uterine com- balloon devices have been used, including the
pression sutures. Thus, a pair of eyed needles, Sengstaken-Blakemore tube2, the Rüsch uro-
preferably blunt point, one straight Keith 10 cm logical balloon4 and the Bakri balloon3 – the
and one 70–80 mm curved, are advisable. A latter is commercially available (see Chapters 28
number of standard sutures should also be and 29). Others have improvised, for example
included: No. 1 polyglactin (vicryl) has a small using a surgical glove tied at the wrist around a
needle but the vicryl can be cut off and inserted plain urethral catheter which, when filled with
into the eyed needles. For the full B-Lynch water or saline, will mould to the contour of the
compression suture, two of the standard suture uterus11. A condom has also been adapted for
lengths of vicryl may need to be tied together. this purpose12. Depending on local availability,
If available, Ethiguard monocryl on a curved one or more of these balloon tamponade kits
blunt point needle is ideal for the B-Lynch com- should be provided on the tray.
pression suture. The standard O and No. 2 Because uterine compression sutures will
chromic needles are suitable for uterine and rarely be used by an individual obstetrician and
ovarian artery ligation. For the vertical uterine the technique may be forgotten, it is useful to
compression sutures and square uterine com- have diagrams, which can be easily sterilized
pression sutures, the straight 10-cm needle and included in the tray or placed on a wall
threaded with No. 1 vicryl is appropriate. chart under glass (Figures 1–4)11.
Material and equipment for uterine and For postpartum hemorrhage due to uterine
vaginal tamponade should be provided. For atony refractory to oxytocic agents, or second-
vaginal tamponade, which may be necessary ary to trauma of the genital tract, the rapid
to prevent hematoma formation following the application of surgical techniques for hemo-
suture of extensive vaginal lacerations, standard stasis is essential to reduce the need for blood
vaginal packing should suffice, although it may transfusion, with its inherent potential morbid-
be necessary to tie more than one of these ity. Often hysterectomy is the final definitive
packs together. For packing the uterine cavity, treatment and may be necessary as a life-saving
Ovarian
artery
• Use curved needle with No. 0/1 or No. 2 suture

• Include a ‘cushion’ of myometrium
Uterine
artery
Figure 1 Uterine and ovarian artery ligation
180
202
30 August 2006 14:20:52
Equipment tray for postpartum hemorrhage
• For use with lower segment Cesarean incision

• Use large curved needle with No. 1 or No. 2 suture
• Can use large 3/8 circle curved cutting needle for same
technique without Cesarean incision
• Or use Ethiguard curved blunt point monocryl
• Check that compression sutures have worked by
observing blood loss p.v. before closing the abdomen
Figure 2 Uterine compression sutures: B-Lynch technique
• Suture through and through with straight 10-cm Keith

• needle
• Multiple square sutures may be used to cover the whole
• body of the uterus; may be useful for placenta previa
• (make sure to leave a drainage portal)
• Sub-endomyometrial injections of 1–2 ml of dilute
• vasopressin (5 units in 20 ml saline) may reduce local
• bleeding in the lower uterine segment
• observing blood loss p.v. before closing the abdomen
Figure 3 Uterine compression sutures: square
• Alternative to the B-Lynch technique if no lower segment

Cesarean incision
• May be placed without opening the uterus using straight
10-cm Keith needle
• Ensure downward bladder retraction
• Two to four vertical sutures may be placed
observing blood loss p.v. before closing the abdomen
Figure 4 Uterine compression sutures: vertical
181
203
04 September 2006 14:06:22
maneuver. However, hysterectomy was avoided control of massive postpartum haemorrhage: an

in all instances in one hospital using an obstetric alternative to hysterectomy? Five cases reported.
hemorrhage tray on nine occasions in 1 year11. Br J Obstet Gynaecol 1997;104:372–5
Thus, if the instruments and equipment are 6. Hayman RC, Arulkumaran S, Steer PJ. Uterine
brace sutures – a simple modification of the
readily available for the rapid application of
B-Lynch surgical procedure for the management
alternative surgical methods, then one is less
of postpartum hemorrhage. Obstet Gynecol 2002;
likely to have resort to hysterectomy with 99:502–6
its attendant morbidity and fertility-ending 7. Smith KL, Baskett TF. Uterine compression
implications. sutures as an alternative to hysterectomy
for severe postpartum haemorrhage. J Obstet
Gynaecol Can 2003;25:197–200
References 8. Cho JH, Jun HS, Lee CN. Hemostatic
1. Maier RC. Control of postpartum hemorrhage suturing technique for uterine bleeding during
with uterine packing. Am J Obstet Gynecol 1993; Cesarean delivery. Obstet Gynecol 2000;96:
169:17–23 129–31
2. Chan C, Razyi K, Tham KA, Arulkumaran S. 9. Fahmy K. Uterine artery ligation to control post-
The use of the Sengstaken–Blakemore tube to partum haemorrhage. Int J Gynaecol Obstet 1987:
control postpartum haemorrhage. Int J Gynaecol 25:363–7
Obstet 1997;58:251–2 10. Evans S, McShane P. The efficacy of internal
3. Bakri YN, Amri A, Jabbar FA. Tamponade iliac ligation. Surg Gynecol Obstet 1985;162:
balloon for obstetrical bleeding. Int J Gynaecol 250–3
Obstet 2001;74:139–42 11. Baskett TF. Surgical management of severe
4. Johanson R, Kumar M, Obhari M, Young P. obstetric haemorrhage: experience with an
Management of massive postpartum haemor- obstetric haemorrhage equipment tray. J Obstet
rhage: use of hydrostatic balloon catheter to Gynaecol Can 2004;26:805–8
avoid laparotomy. Br J Obstet Gynaecol 2001; 12. Akhter S, Begum MR, Kebir Z, Rashid M, Laila
108:420–2 TR, Zabean F. Use of a condom to control
5. B-Lynch C, Cocker A, Lowell AH, Abu J, massive postpartum hemorrhage. Medscape Gen
Cowan MJ. The B-Lynch surgical technique for Med 2003;5:3
182
204
30 August 2006 14:20:56
22
BUILDING HOSPITAL SYSTEMS FOR MANAGING MAJOR
OBSTETRIC HEMORRHAGE
D. W. Skupski, G. S. Eglinton, I. P. Lowenwirt and F. I. Weinbaum
INTRODUCTION TACKLING THE PROBLEM OF MAJOR

OBSTETRIC HEMORRHAGE
Maternal death from major obstetric hemor-
rhage is a leading killer of women world-wide, Recently developed programs to improve out-
as most of the chapters in this book amply comes for women with major obstetric hemor-
demonstrate. Attention to this topic is not rhage have focused on at least two important
glamorous, unfortunately, but few topics can factors: first, the initial response to the hemor-
be more important in improving the health rhage, and, second, the prevention of hemor-
of reproductive-aged women throughout the rhage in those patients who can be identified as
world. This chapter demonstrates a proven, being at high risk for it. This latter effort is in
in-hospital approach to decreasing morbidity recognition of the fact that two of the three
and mortality of women with major obstetric most common causes of hemorrhage cannot be
hemorrhage1. The program hinges on building, identified in advance. These are uterine atony
developing and improving existing hospital and/or placenta previa and placenta accreta4. In
systems that are necessary for the care of such contrast, only placenta previa is reliably able to
women. be diagnosed in advance.
Any program aimed at improving outcomes
from major obstetric hemorrhage must also
consider the interface of individuals and depart-
BACKGROUND ments that may not traditionally be thought of
as important in the process of caring for women
In the United States, the need for Cesarean
with obstetric hemorrhage. The remainder of
hysterectomy as well as the incidence of major
this chapter describes the details of these hospi-
obstetric hemorrhage have both increased in
tal systems and, in particular, how they have
recent years2–4, most likely due to the known
recently been revised with good effect in a major
increase in Cesarean and repeat Cesarean deliv-
New York teaching hospital.
ery with their respective increases in placenta
previa and accreta, especially in patients under-
going repeat Cesarean delivery2–4. In the setting
IMPORTANCE OF COMMUNICATION
of intractable obstetric hemorrhage, emergency
AND EDUCATION
peripartum hysterectomy is used as a life-saving
procedure. According to one recent article, the Two extremely important and overarching pro-
incidence of emergency peripartum hyster- cesses must be initially addressed in order for
ectomy is approximately 2.5/1000 deliveries3 any program aimed at improving outcomes to
and hemorrhage associated with uterine atony is be successful: communication and education. It
the most frequent indication, followed by cannot be over-emphasized that clear channels
placenta accreta5. Apart from whether or not of communication must be developed between
hysterectomy need be performed, maternal all the people and departments that are involved
death is a known complication of major obstet- in caring for women with major obstetric
ric hemorrhage6. hemorrhage. This includes the immediate and
183
205
30 August 2006 14:20:56
coordinated communications that are inevitably year 2000 and one in the year 2001. This
necessary for any rapid response team to work at circumstance prompted the creation of a patient
maximum capacity. This communication must safety team that worked to improve the hospital
be far more comprehensive than just the mem- systems at NYHQ for caring for women at risk
bers of the obstetric department and may need for, or suffering from, major obstetric hemor-
to include members of the emergency depart- rhage. This patient safety team chose as its
ment, anesthesiology, the labor and delivery mission and was successful in the creation
suite, nursing administration, the operating of an improved management scheme (clinical
rooms, and the blood bank, to name just a few. pathway) for the identification and management
Basic education is equally important, and it is of major obstetric hemorrhage, with the express
imprudent to believe that attending or house intent of reducing maternal deaths due to this
staff will know (a priori) all the component parts cause.
of the program based on their past experience
and training. All care providers who evaluate
these patients and institute therapy must Patient safety teams
possess the requisite knowledge of the patho-
Beginning in 2001, a multidisciplinary patient
physiology of hemorrhagic shock in order to
safety team was established that included indi-
identify the presence and assess the severity of
viduals from the medical divisions of Obstetric
this problem, and to begin the process of initial
Anesthesiology, Maternal Fetal Medicine, Neo-
treatment. It cannot be over-emphasized to all
natology and the Blood Bank, as well as the hos-
levels of staff that the diagnosis is not always as
pital departments of Nursing, Communication
easy as training manuals might suggest. The
and Administration. Over the course of 6–12
involvement of departmental leaders who are
months, meeting usually every week for 1–2 h,
experienced with the management of obstetric
the newly created patient safety team evaluated
hemorrhage and available on a 24/7/365 basis is
the totality of the medical center’s care of the
key. When they become primary stakeholders in
two women who died from major obstetric
the educational process, training for less experi-
hemorrhage, considered both the proximate
enced care providers should be developed and
and systems-related causes of these unfortunate
be repeated on a regular basis. Training such as
outcomes, discussed possible recommended
this should be thought of as a continuous pro-
changes in the management, and decided on
cess – something that has to be repeated to every
how best to change the systems at NYHQ that
new rotation of house staff and attending
were then present for the care of women who
consultants.
might find themselves in similar circumstances.
EVENTS AT NYHQ
Objective of our study
The New York Hospital Medical Center of
Queens (NYHQ) is an acute care 480-bed hos- In order to assess the impact of the proposed
pital in Flushing, New York, affiliated with the changes in hospital systems on the outcomes
Weill Medical College of Cornell University, of our patients, we began to carefully record
and the New York Presbyterian Hospital. The a variety of pertinent outcomes prospectively
hospital serves an urban community of great from that point forward, and looked back retro-
ethnic diversity who are insured by both com- spectively to record the same outcomes for the
mercial and governmental payers; the hospital is 2 years in which the deaths had occurred. The
designated for the highest level (Level III) of committee was of the opinion that the accurate
Neonatal Intensive and Maternal Care, and also recording of outcomes was essential to demon-
has the highest designation for a Trauma Center strate any effect of changes in management
(Level I). Separate critical care units are dedi- over time. Specifically, we hypothesized that
cated to Surgical, Medical and Cardiac services. the changes we implemented in our hospital
Two maternal deaths due to major obstetric systems would lead to improved outcomes for
hemorrhage occurred in recent years, one in the women with major obstetric hemorrhage.
184
206
30 August 2006 14:20:56
Hospital systems for managing major obstetric hemorrhage
Methods changes (i.e. to improve patient safety) and

the importance of early diagnosis of major
A multifaceted approach included the following:
hemorrhage.
(1) We formed an obstetric rapid response (7) We established the role of the existing
team (Team Blue), modeled it after the Trauma Team (with the full agreement of
cardiac arrest team, and included quarterly the Director of the Trauma Division) to
mock drills on all shifts for various emer- specifically respond and assist in cases of
gency clinical scenarios. severe obstetric hemorrhage, because the
(2) We developed clinical pathways – guide- Trauma Team was the most experienced in
lines and protocols – specifically designed to resuscitation of patients with hemorrhagic
provide for early diagnosis of patients at risk shock within our institution. The Trauma
for major obstetric hemorrhage and for Team includes surgical house officers work-
streamlined care in emergency situations. ing under the direction of the surgical
trauma attending physician. These team
(3) In response to a marked increase in the vol- members are expert in the placement of
ume of gynecologic emergency cases and large-bore intravenous lines (by venous
births at NYHQ, we separated the in-house cut-down if necessary), are knowledgeable
obstetric and gynecologic responsibilities to about the physiology of volume resuscita-
allow the in-house obstetrician to focus tion, assist in obtaining adequate amounts
on obstetric emergencies without fear of of blood products for massive blood
neglecting gynecological emergencies. replacement, and also are most experienced
(4) We revised the duties of the 24-h in-house in inserting intraluminal lines directly into
staff (consultant) obstetrician to include the major vessels for monitoring and
continuous and frequent monitoring of all obtaining requisite samples.
patients on the Labor and Delivery unit.
This monitoring included those patients
who had private obstetricians who might The creation of new protocols and
not be present on a continuous basis. guidelines
(5) We empowered all obstetric care providers The following protocols and guidelines
(including physician assistants, nurses, resi- were created to enhance the reception and
dent physicians and the in-house attending perpetuation of the new activities.
physician) to immediately involve senior ● We prepared for major hemorrhage in
members of the Department whenever patients with known placenta previa (Figure
there was disagreement with or concern 1). This preparation included antenatal
about the management scheme (particu- consultation with Maternal Fetal Medicine,
larly when there was a possible delay in rec- Obstetric Anesthesiology and senior
ognition of the severity of hemorrhage). A gynecologic surgeons; liberal use of ultra-
senior member of the Department was then sound to identify placenta accreta in patients
required to discuss the issue immediately with prior uterine surgery and/or placenta
with the attending physician to avoid delay. previa. When such patients were identified,
they received twice-weekly type and screen
(6) Through weekly didactic sessions, we
to allow for more rapid availability of blood
educated all of our staff to recognize
products if major hemorrhage occurred.
the severity of hemorrhage described in the
Amniocentesis was performed for fetal lung
Advanced Trauma Life Support Manual of
maturity at 36 weeks of gestation followed by
the American College of Surgeons7, and
planned Cesarean delivery if the fetal lungs
disseminated information regarding the
were shown to be mature.
new protocols for patient care. The attend-
ing, nursing and ancillary staffs were all ● We prepared for major hemorrhage in
informed regarding the intent of the patients in whom we suspected placenta
185
207
30 August 2006 14:20:57
E ar l y p reg n an c y p r e v ia o r l o w - l y i n g pl a ce n t a
U l tr as o u nd at 3 0 w e e k s
P re v i a n o t s e e n P r e v i a s ee n
St a n d a rd ma n ag e m en t Look for accreta*
Accreta suspected Accreta not suspected
Cou ns eli ng Counseling
P l a n n ed a m n i o c e n t e s i s f o r P l an ne d amn io c en t es i s f or
fetal lung maturity at 36 fetal lung maturity at 36 weeks
w eeks followed by p lann ed f o l l o w ed b y Ce s a r e a n de l i v er y
Cesarean hysterectomy
Figure 1 Proposed management scheme for patients at risk for major obstetric hemorrhage. CD,
Cesarean delivery. *Suspicion for accreta is markedly increased with prior CD and anterior placenta;
†includes bed rest, pelvic rest, preparation for CD, serial CBC, consider erythropoeitin, iron and vitamin
supplements and serial autologous blood donation; ‡includes the counseling above and a recommendation
for Cesarean hysterectomy. Low parity may decrease the strength of the recommendation if future
child-bearing is desired
accreta (Figure 1). This included autologous intensive care unit as needed. In addition,
blood donation as often as every week for we used the Cell Saver, but only after
a period of 4–5 weeks before the planned delivery of the fetus and after copious
Cesarean delivery; erythropoietin, iron and peritoneal irrigation had been performed4.
vitamin therapy in an effort to boost red Weekly autologous blood donation not
blood cell production; consultation with only was used to prevent introduction of
interventional radiology in which we would blood-borne infection with transfusion but
consider placement of ports preoperatively, also contributed to resolving any potential
so that embolization of major pelvic blood shortage of blood in our area.
vessels could occur rapidly in the event of
● We obtained consultation with the Trauma
substantial hemorrhage during the operation;
Team as necessary.
judicious placement of additional intra-
venous lines and a 7.5 French internal ● For patients with suspected placenta accreta,
jugular cordis for invasive monitoring and we discussed the likely decreased maternal
volume replacement; intraoperative monitor- mortality of planned Cesarean hysterec-
ing with an arterial line and central venous tomy8. Planned Cesarean hysterectomy was
pressure; and transfer to the surgical then performed for those who agreed.
186
208
30 August 2006 14:20:57
● For patients with suspected placenta accreta, retrospectively during the time period of
Cesarean delivery and Cesarean hysterec- January 2000 to May 2001 and prospectively
tomy were scheduled in the main operating beginning in June 2001.
room under the direction of senior The data collection program also involves
gynecologic surgeons (Figure 1), because monitoring by senior Departmental leaders
staff and facilities of the main operating room who receive reports on a daily basis from care
are better equipped to perform hysterectomy providers regarding all cases of major obstetric
than is the case with the Labor and Delivery hemorrhage. These cases were highlighted and
suite. This procedural change also avoided included in the database as they occurred.
the problem of consuming staff and resources Outcomes analyzed included maternal deaths,
on Labor and Delivery that were considered lowest documented maternal pH, lowest
necessary for the care of other patients. documented maternal temperature, and the
occurrence of coagulopathy.
Table 1 shows the hospital systems involved, Our definition of major obstetric hemorrhage
along with an assessment of the impact on included one or more of the following:
improving outcomes in women with major estimated blood loss = 1500 ml, need for blood
obstetric hemorrhage and the relative amount of transfusion, need for uterine packing, perfor-
work involved in the change. mance of uterine artery ligation, and perfor-
In addition to the changes in systems detailed mance of Cesarean hysterectomy. Admittedly,
above, data on obstetric volume, mode of deliv- this definition is different from that of post-
ery, occurrence of major obstetric hemorrhage partum hemorrhage that has been detailed in
and outcomes important in identifying improve- other chapters of this volume. Accordingly, the
ments were collected from 2000 to 2005. Cases rate of major obstetric hemorrhage by our
were identified prospectively for the entire definition was expected to be lower than the
patient cohort (2000–2005). Demographic and known incidence of postpartum hemorrhage.
outcome data on each patient were recorded Data were compared between the 2 years before
Table 1 Impact of hospital system changes on the outcomes of women with major obstetric hemorrhage
Specific change Impact Amount of work
Administrative
Patient safety team critical extensive
Trauma Team involvement minor moderate
Departmental
Obstetric rapid response team critical extensive
Development of clinical pathways or guidelines major moderate
Dissemination of clinical pathways or guidelines major moderate
Separation of in-house obstetrician and gynecologist minor moderate
Culture change to proactive attending physician major moderate
Care provider empowerment major moderate
Didactic teaching about physiology and treatment of hemorrhagic shock major moderate
Clinical pathways or guidelines
Antenatal management of known placenta previa major moderate
Preparation for hemorrhage in suspected placenta accreta minor moderate
Counseling about planned Cesarean hysterectomy minor minimal
Scheduled Cesarean delivery for previa and accreta in the main operating room minor minimal
Nursing
Culture change to team participation major extensive
Empowerment of nurses major moderate
187
209
30 August 2006 14:20:57
Table 2 Major obstetric hemorrhage in the period 2000–2005
Total Repeat Cases of

Cesarean Cesarean major obstetric Cesarean
Year Births births* births† hemorrhage‡ hysterectomy§ Mortality
2000 2705 516 217 3 1 1

2001 3106 801 287 8 5 1
2002 3323 903 332 8 5 0
2003 3395 932 326 14 4 0
2004 3648 1053 374 18 5 0
2005 (8 months) 2546 759 275 12 4 0
Total 18 723 4964 1811 63 24 2
*2000–2001 compared to 2002–2005, p < 0.0001; †2000–2001 compared to 2002–2005, p = 0.002;

‡2000–2001 compared to 2002–2005, p = 0.02; §rate of Cesarean hysterectomy as a function of the total
number of major obstetric hemorrhage cases 2000–2001 compared to 2002–2005, p = 0.37
and the 3 years after the systemic changes were lowest pH (p = 0.004) and lowest temperature
implemented, 2000–2001 vs. 2002–2005. (p < 0.0001). There also was a trend toward
less coagulopathy (p = 0.09). These diverse
findings were very important, because it is
Results
known that a triad of physiologic derangements
During each successive year of the study, the occurs in hemorrhagic shock that can lead to
following important changes occurred simulta- death. This triad comprises acidemia, hypother-
neously: increasing obstetric volume, increasing mia and coagulopathy. Its presence helps to
rate of Cesarean delivery, an increasing rate of confirm that our major finding of reduced
repeat Cesarean delivery, and an increasing maternal death is not a statistical chance event,
number of cases of major obstetric hemorrhage and also argues that our response to the event of
(Table 2). The increases in Cesarean delivery, a major obstetric hemorrhage became better as
repeat Cesarean delivery, and cases of major time passed and as care providers became more
obstetric hemorrhage all were significant experienced and knowledgeable.
between the time periods of 2000–2001 vs. The two time periods were also analyzed
2002–2005, but no difference was shown in the according to other characteristics, such as need
rate of Cesarean hysterectomy (Table 2). for Cesarean hysterectomy, volume of trans-
Clinical characteristics, measures of severity fusion, operative time, need for intubation for
of hemorrhage and outcomes are shown in greater than 24 h, and number of hours
Table 3. The patient groups from the two time intubated (Table 3). No significant differences
periods (2000–2001 vs. 2002–2005) were simi- were present in these measures in the periods
lar in demographics as measured by age, parity 2000–2001 vs. 2002–2005. The incidence
and incidence of prior Cesarean delivery. The of peripartum hysterectomy was 1.3/1000
severities of obstetric hemorrhage also appeared (24/18 723) during the entire study period
to be similar between the time periods. The (2000–2005). Placenta accreta with prior
severity measures were APACHE II scores9, Cesarean delivery accounted for 14/24 (58.3%)
occurrence of placenta accreta and amount of cases of Cesarean hysterectomy, and we sus-
estimated blood loss (Table 3). pected accreta in seven cases and confirmed it in
The major result of the combined effort was four cases at delivery. The operative characteris-
that maternal deaths were significantly reduced tics, morbidity and mortality of patients under-
in the time period following the systemic going peripartum hysterectomy are shown in
changes (p = 0.036). This was supported by the Table 4. The numbers here are different from
additional findings of significant differences in Table 3, because Table 3 shows all patients
188
210
30 August 2006 14:20:57
Table 3 Major obstetric hemorrhage: comparison of demographics, measures of severity and outcomes
2000–2001 2002–2005
(n = 12) (n = 49) p Value
Demographics
Age, mean (SD) 36.5 (6.0) .– 34.2 (5.9) .– < 0.23 *
Parity, median (range) 36.1 (0–3) .. 36.1 (0–5) .. < 0.70 *
Prior Cesarean delivery, n (%) 36.6 (50.0) .– .532 (65.3) .– < 0.33 *
Severity measures
Occurrence of placenta accreta, n (%) 36.4 (33.3) .– .511 (22.4) .– < 0.46 *
APACHE score, median (range) 11.5 (7–31) .. .510 (6–18) .. < 0.07 *
Estimated blood loss, mean (SD) 2725 (1289) 2429 (1214) < 0.46 *
Outcomes
Maternal death, n (%) 36.2 (16.7) .– 36.0 (0.0) .– < 0.036*
Lowest pH, median (range) 7.23 (6.8–7.39) 7.34 (7.08–7.44) < 0.004*
Lowest temperature (°C), median (range) 35.2 (30.2–35.8) 36.1 (35.2–37.8) < 0.0001*
Coagulopathy, n (%) 36.7 (58.3) .– .515 (30.6) .– < 0.09 *
Cesarean hysterectomy, n (%) 36.6 (50.0) .– .518 (36.7) .– < 0.51 *
Volume of transfusion, mean (SD) 1313 (1029) 1194 (1547) < 0.80 *
Operative time, mean (SD) .185 (91) ..– .184 (79) ..– < 0.99 *
Intubation > 24 h, n (%) 36.7 (58.3) .– .516 (32.7) .– < 0.18 *
*Significant difference
during the entire study period and the data hemorrhagic shock is not recognized or when
in Table 4 is confined to those patients who a patient presents in an advanced state of
underwent Cesarean hysterectomy. Interest- hemorrhagic shock, a need to improve hospital
ingly, a significant difference was also present in systems to provide a safety net for patients is
the lowest pH in patients undergoing Cesarean as important as is the education of a specific
hysterectomy between the time periods of physician or group of physicians after an adverse
2000–2001 vs. 2002–2005. We think this outcome.
underscores that our response to women with Our findings indicate that there were
hemorrhagic shock from blood loss improved significant improvements in outcomes after we
over the course of time. introduced systemic changes at our institution,
including improvements in maternal deaths,
lowest pH and lowest temperature. There were
Deciphering the data
no difference in measures of severity of obstetric
The response to major obstetric hemorrhage hemorrhage and significant increases in the
must be multifaceted and rapid in order to number of cases of major obstetric hemorrhage
be successful. A quality assurance committee between the study time periods, leading us to
would be the traditional departmental or insti- the conclusion that this improvement in out-
tutional response to a poor outcome such as a comes is a true finding. When comparing the
maternal death from hemorrhage, and, after this time periods before and after the systemic
peer review, specific physician education would changes, the significant differences in lowest
occur regarding the components of early identi- temperature and in lowest pH (Table 3) suggest
fication and ‘best’ treatment, as determined by that the team’s response to massive hemorrhage
departmental leaders. However, this traditional improved after system-wide interventions. The
response ignores the lessons learned from the reduction in maternal mortality, however, can-
Institute of Medicine report regarding errors not be considered a robust observation, because
that lead to morbidity and mortality during hos- this observation is hospital-based and may
pital stays10. When clinical judgment fails and not be replicated in a population-based sample.
189
211
30 August 2006 14:20:58
Table 4 Peripartum hysterectomy in the period 2000–2005. Incidence: 24/18 723 (1.3/1000). All data are
number of cases unless otherwise designated
2000–2001† 2002–2005‡ Total§
Etiology
Placenta accreta 4 10 14
Placenta accreta with prior CD 4 10 14
Uterine atony 2 6 8
Morbidity
Cystotomy 1 1 2
Pulmonary embolus 1 0 1
Coagulopathy 5 8 13
Acute tubular necrosis 0 0 0
ARDS 0 0 0
Myocardial infarction 0 0 0
Pneumonia 0 0 0
Mortality
Placenta percreta 1 0 1
Other characteristics
Operative time (min), mean (SD) 259 (52.3) 250 (66.6) 252 (62.4)
EBL (ml), median (range) 3500 (2500–5200) 3000 (1000–7000) 3250 (1000–7000)
Transfusion total volume (ml), mean (SD) 2125 (847.8) 2292 (2076.4) 2250 (1829.9)
FFP/platelets given (n) 5 10 15
Lowest pH, mean (SD) 7.15* (0.17) 7.27* (0.07) 7.24 (0.12)
Intubated 5 12 17
Intubated > 24 h 3 3 6
Days to discharge, median (range) 6 (4–7) 4 (3–11) 5 (3–11)
Anesthetic management
Regional anesthesia only 1 3 4
Conversion to general 2 12 14
General anesthesia only 3 3 6
†2000–2001 hysterectomy n = 6, total births n = 5811; ‡2002–2005 hysterectomy n = 18, total births
n = 12 912; §2000–2005 (total) hysterectomy n = 24, total births n = 18 723; *significant difference
p = 0.02
CD, Cesarean delivery; ARDS, adult respiratory distress syndrome; SD, standard deviation; EBL, estimated
blood loss; FFP, fresh frozen plasma
This caveat in no way diminishes the value of interest that the entire staff accepted these addi-
our findings in terms of their broad applicability tional time expenditures as part of their ongoing
in other hospitals throughout this and other self-education and were proud of the outcome
countries. and the results (Table 1).
The process of implementing the systemic This study design does not allow a determi-
changes required considerable effort by many nation of which of several interventions may
individuals and was very time-intensive. The have accounted for improvements in outcome.
patient safety team met numerous times and We strongly believe that the data presented in
deliberated on the specifics of our response. this chapter support the conclusion that a
These efforts included repeated education of well-reasoned, carefully constructed and multi-
care providers on the diagnosis and manage- faceted program focusing on patient safety
ment of hypovolemic shock. It is of considerable can improve outcomes, although we cannot
190
212
30 August 2006 14:20:58
attribute any specific improvement to any spe- References

cific change that we instituted. We also strongly
1. Skupski DW, Lowenwirt IP, Weinbaum FI,
believe that our experience demonstrates that Brodsky D, Danek MM, Eglinton GS. Improv-
focusing on the problem of obstetric hemor- ing hospital systems for the care of women with
rhage by the medical and administrative depart- major obstetric hemorrhage. Obstet Gynecol
ments in a given hospital can and does lead 2006:107:97–83
to improved outcomes. The effort involved is 2. Kastner ES, Figueroa R, Garry D, Maulik D.
substantial, but rewarding. Emergency peripartum hysterectomy: experience
at a community teaching hospital. Obstet Gynecol
2002;99:971–5
FINAL COMMENTS 3. Miller DA, Chollet JA, Goodwin TM. Clinical
risk factors for placenta previa-placenta accreta.
The risk of placenta previa with or without
Am J Obstet Gynecol 1997;177:210–14
accreta in patients with multiple Cesarean
4. Placenta accreta. ACOG Committee Opinion
deliveries is difficult to quantitate11. However, No. 266. American College of Obstetricians and
recently published prospective data12,13 corrob- Gynecologists. Obstet Gynecol 2002;99:169–70
orate previously published retrospective data on 5. Forna F, Miles AM, Jamieson DJ. Emergency
the substantial risk of accreta associated with peripartum hysterectomy: a comparison of cesar-
previa and prior Cesarean14. Placenta previa is a ean and postpartum hysterectomy. Am J Obstet
detectable condition, allowing for a preventive Gynecol 2004;190:1440–4
clinical pathway such as that developed in 6. Frieden TR, Novello AC, King J. Health Alert:
Figure 1 to be implemented. We believe that prevention of maternal deaths through improved
the preparation that takes place after the early management of hemorrhage. Letter from State of
New York Department of Health and The New
identification of patients at risk is an important
York City Department of Health and Mental
component in the ability to improve outcomes
Hygiene, August 9, 2004
for our program. 7. American College of Surgeons Committee on
When confronted with adverse outcomes, Trauma. Advanced Trauma Life Support for Doc-
principles of quality improvement require that tors, Chapter 3. Shock. Chicago: American Col-
‘systems’ thinking take place. It is tempting to lege of Surgeons, 1997
attempt to correct the proximate cause (e.g. an 8. Sheiner E, Levy A, Katz M, Mazor M. Identify-
individual physician’s lack of attention to detail ing risk factors for peripartum cesarean hysterec-
or suboptimal clinical judgment on an individ- tomy. A population-based study. J Reprod Med
ual case) without addressing the ‘systems’. We 2003;48:622–6
believe these data support the clear need for a 9. Knaus WA, Draper EA, Wagner DP,
Zimmerman JE. APACHE II: a severity of dis-
systemic response and hope they are useful to
ease classification system. Crit Care Med
others faced with the task of improving safety in
1985;13:818–29
obstetric suites. The specific series of changes in 10. Kohn LT, Corrigan JM, Donaldson M. To err is
systems at our institution was uniquely adapted human: building a safer health system. Washing-
to the circumstances we encountered. It is pos- ton, DC: Institute of Medicine, 1999
sible that these changes may not be as important 11. Greene MF. Vaginal birth after Cesarean revis-
nor as easily achievable in other areas of the ited. N Engl J Med 2004;351:2647–9
world. However, in any institution’s response to 12. Silver RM for the MFMU Network of the
major obstetric hemorrhage, it is important to NICHD. The MFMU cesarean section registry:
keep in mind the numerous and potentially maternal morbidity associated with multiple
changing nature of obstacles to system changes repeat cesarean delivery. Am J Obstet Gynecol
2004:191:S17 Abstr
and the need to put together a multidisciplinary
13. Rashid M, Rashid RS. Higher order repeat cae-
response to overcome these obstacles. Though
sarean sections: how safe are five or more? Br J
this is a challenging task, the result of improve- Obstet Gynaecol 2004;111:1090–4
ments in outcomes for women with obstetric 14. Clark SL, Koonings PP, Phelan JP. Placenta
hemorrhage remains rewarding and, most previa/accreta and prior cesarean section. Obstet
importantly, achievable. Gynecol 1985;66:89–92
191
213
30 August 2006 14:20:58
214
30 August 2006 14:20:58
Section VI
Therapy for non-atonic conditions
215
30 August 2006 14:20:58
23
BLEEDING FROM THE LOWER GENITAL TRACT
A. Duncan and C. von Widekind
INTRODUCTION (2) Vaginal tears (above and below the levator

ani muscle, see Figure 1);
In the first comprehensive English Language
textbook on the subject, William Smellie, in his (3) Vulva and perineal tears;
1752 Treatise on the Theory and Practise of Mid-
(4) Episiotomies.
wifery1, correctly identifies the atonic uterus as a
major cause of postpartum hemorrhage with his With the exception of cervical tears without
statement ‘This dangerous efflux is occasioned by vaginal extension, all of the above can lead to
every thing that hinders the emptied uterus from paravaginal hematomas, which in turn can be
contracting’. Although he refers to vaginal pack- divided into those above and below the levator
ing with Tow or linen rags (dipped in astringents ani muscle (Figure 1). Infralevator hematomas
such as oxycrate, red tart wine, alum or include those of the vulva, perineum, para-
Sacchar-saturni), he does not specifically refer vaginal space and ischiorectal fossa. Supra-
to bleeding from the lower genital tract. Because levator bleeding is more dangerous, as it is more
this omission was repeated in subsequent years difficult to identify and control the source of
by many standard textbooks and reviews of bleeding, and blood loss into the retroperitoneal
postpartum hemorrhage, it is not surprising that space can be massive.
the present evidence base is poor, and a 2005
MESH search in PubMed of the National
Library USA combining the terms ‘Postpartum INCIDENCE
hemorrhage’ AND ‘Lacerations’ OR ‘Rupture’ In the UK, postpartum hemorrhage of more
NOT ‘Uterine rupture’ came up with only 28 than 500 ml occurs in between 5 and 17% of all
publications. deliveries and postpartum hemorrhage of more
Maternal deaths specifically from lower geni- than 1000 ml in 1.3% of deliveries.
tal tract bleeding as the cause of postpartum
hemorrhage are rare in the developed world.
The 2000–2002 United Kingdom Confidential Cervical tears
Enquiries2 reported only one death from this
Minor cervical tears are common and are likely
cause. World-wide, no accurate figures exist,
to remain undetected. However, bleeding which
but it is likely that the numbers are significant,
occurs despite a well-contracted uterus and which
particularly where there is significant co-
does not appear to be arising from the vagina
morbidity and a poorly resourced maternity
or perineum is an indication for examining the
infrastructure3.
cervix. Numerous cases have been described of
women dying from hemorrhage due to a cervical
tear, following operative vaginal delivery.
CLASSIFICATION
Possible sources of bleeding from the lower Postpartum hematoma
genital tract include:
Because there is no agreed definition, there
(1) Cervical tears; is no consensus as to the incidence. After
194
216
30 August 2006 14:20:58
Bleeding from the lower genital tract
8
3
/$
3
E
/$
Figure 1 Paravaginal hematomas. (a) The hematoma lies beneath the levator ani muscle; (b) the
hematoma lies above the levator ani and is spreading upwards into the broad ligament. H, hematoma;
LA, levator ani, U, uterus; P, pelvic peritoneal reflection
195
217
30 August 2006 14:20:59
spontaneous delivery, up to 50% of parturients with a relative risk of 5, carried the same weight
develop a minor self-limiting infralevator/vulva as a cause of postpartum hemorrhage as did
hematoma5. In contrast, the formation of a sig- multiple pregnancy and retained placenta.
nificant postpartum hematoma is an uncommon Rotational forceps are a particular risk factor for
but serious complication after delivery, with the spiral vaginal tears9.
reported incidence of around 1 in 500–700 Coagulation disorders, if present, are likely to
deliveries6. Major pelvic (supralevator) hema- significantly increase the risk of lower genital
tomas are rare, with widely varying reported tract hemorrhage and hematoma and therefore
incidence of between 1 in 500 and 1 in 20 0007. should always be corrected where possible. If
vaginal lacerations require repair in this situa-
tion, the threshold for the use of a vaginal pack
Episiotomy
should be low.
An episiotomy can bleed heavily, and, although
there are no data on the incidence of hemor-
PREVENTION
rhage from this cause alone, observational stud-
ies suggest that the relative risk of postpartum The three main areas in which risk can be
hemorrhage is increased four to five times if an reduced all require a proactive approach:
episiotomy is performed8.
(1) Antenatal co-morbidities such as anemia
and diabetes should be treated so that
RISK FACTORS women entering labor are as healthy as
possible.
The major causes of postpartum hemorrhage
are uterine atony, retained placental fragments, (2) A consistent proactive approach is required
morbid adherence of the placenta and lower in both the first and second stages of labor.
genital tract lacerations. Data from the North Active monitoring (partogram) and early
West Thames District of the UK (Table 1) intervention are essential where progress is
reviewed the obstetric factors associated with a inadequate or cephalic-pelvic disproportion
blood loss of more than 1000 ml and appor- is diagnosed. Coagulation defects (includ-
tioned a relative risk to each factor4. Of these, ing iatrogenic defects due to anticoagulat-
assisted delivery (forceps or vacuum extrac- ion) should be corrected where possible
tion), prolonged labor, maternal obesity (and (see Chapter 25).
associated large baby) and episiotomy were
(3) Postpartum, the early identification of
most relevant to the risks of lower genital tract
excessive blood loss and a proactive
hemorrhage. It is worth noting that episiotomy,
approach to resuscitation/fluid replacement
as well as identification of the source of
bleeding and stopping it, are vital.
Table 1 Risk factors for postpartum hemorrhage
and approximate increase in risk4 Because operative delivery and episiotomy are
both significant risk factors for postpartum
Relative Relative
hemorrhage from the lower genital tract, efforts
Antenatal risk Intrapartum risk
to reduce the incidence of both are likely to
Placenta 13 Emergency Cesarean 9 reduce the risk of hemorrhage. Where operative
previa section vaginal delivery is required, however, then
Obesity 2 Assisted delivery 2 a proper technique as described in standard
Prolonged labor (> 12 h) 2 textbooks10 will reduce the risk of vaginal and
Placental abruption 13 cervical tears.
Multiple pregnancy 5
Retained placenta 5
Elective Cesarean section 4 DIAGNOSIS
Mediolateral episiotomy 5
Pyrexia in labor 2 Careful and well-documented observation after
delivery is imperative as the seriousness of
196
218
30 August 2006 14:20:59
concealed or persistent low-grade blood loss can polyglactin/polyglycolic acid suture on a taper-
be underestimated. cut needle, obliteration of dead spaces and
Bleeding, especially after instrumental taking care that sutures are not inserted too
vaginal delivery, that occurs despite a well- tightly. If dead spaces cannot be closed securely,
contracted uterus and that does not appear to then a vaginal pack should be inserted.
be arising from the lower vagina or perineum
is an indication for examination of the upper
Vaginal tear repair
vagina and cervix. The characteristic feature of
bleeding from upper vaginal and cervical tears is The technique for repair of superficial vaginal
a steady loss of fresh red blood. tears is similar to that of perineal repair, as
Exclusion of upper vaginal and cervical tears described above. Use an absorbable, continuous
requires examination in the lithotomy position interlocking stitch, which must start and finish
with good relaxation, good light and proper beyond the apices of the laceration, and should
assistance7. A tagged vaginal tampon to absorb where possible reach the full depth of the tear
blood loss from the uterine cavity and the use in order to reduce the risk of subsequent
of flat-bladed vaginal retractors will assist in hematoma formation.
visualizing the vaginal walls. For deeper tears, an attempt should be made
The cervix should always be examined where to identify the bleeding vessel and ligate it.
there is continuing bleeding despite a well- If there is any significant dead space or if the
contracted uterus and also after use of all vagina is too friable to accept suturing, then
rotational forceps, which are associated with a packing is indicated (see below), because access
significant increase in the risk of upper vaginal to deeper tears is usually difficult in an inade-
and cervical tears11. The method for doing this quately anesthetized patient. Thus, repair of
is to grasp the anterior lip with one ring forceps such lacerations should be done in theater with
and to place a second ring forceps at the adequate anesthesia.
2-o’clock position, followed by progressively Lacerations high in the vaginal vault and
‘leap-frogging’ the forceps ahead of one another those extending up from the cervix may involve
until the entire circumference has been the uterus or be the cause of broad ligament or
inspected. retroperitoneal hematomas. The proximity of
the ureters to the lateral vaginal fornices, and
the base of the bladder to the anterior fornix,
TREATMENT
must be kept in mind when any extensive repair
Hemorrhage from the lower genital tract should is undertaken in these areas. Poorly placed
always be suspected when there is ongoing stitches can lead to genitourinary fistulas.
bleeding despite a well-contracted uterus. Vaginal packing for at least 24 h is always wise
Generally, high vaginal or cervical tears require under these conditions.
repair under regional anesthesia in theater. Vaginal packing using gauze is the most
The Scottish Obstetrics Guidelines and common method to achieve vaginal tamponade.
Audit Project (SOGAP) group provides detailed As with uterine packing, the technique of
guidelines on the management of postpartum vaginal packing involves ribbon gauze inserted
hemorrhage12. A summary of the ORDER uniformly side-to-side, front-to-back and top-
protocol as described by Bonnar13 is shown to-bottom. Vaginal packing using thrombin-
in Table 2, with additional boxes relating to soaked packs, as described for uterine packing,
hemorrhage from the lower genital tract. can also be considered15, especially where
closure of all lacerations has not been possible.
Because of the risk that the raw vaginal sur-
Perineal tear repair
face will bleed on removal of the pack, povidone
The technique has been well described else- iodine-soaked double lengths of 4.5 × 48 inch
where14. The principles include ensuring that packs can be inserted inside sterile plastic
the first suture is inserted above the apex of the drapes (this has been well described for the
tear or episiotomy incision, use of a continuous management of uterine hemorrhage, but the
197
219
30 August 2006 14:21:00
Table 2 Management of major postpartum hemorrhage (blood loss > 1000 ml or clinical shock) (see
reference 13)
Failure to control bleeding

First-line management Second-line management
Bleeding despite a well-
contracted uterus is likely to be
due to genital tract trauma
O: ORGANIZATION/Call for help 1. Inform

Consultant Obstetrician
Consultant anesthetist
2. Examine under anesthesia:
R: RESTORATION OF BLOOD VOLUME
Remove retained products
Blood and crystalloid transfusion
Repair any tear
Bimanual compression
D: DEFECTIVE BLOOD COAGULATION Prostaglandin F2 intramyometrial and
Correct as dictated by clotting studies (250 µg maximum eight injections i.m.)
Continue bimanual compression
Continue resuscitation and monitoring
E: EVALUATION OF RESPONSE Failure to control bleeding
R: REMEDY THE CAUSE Under anesthetic (general or regional)

1. Improve the tone 1. Repair cervix
Bimanual compression Circumferential examination with ring
Oxytocin 10 units by slow i.v. forceps
injection Repair with interrupted figure-of-eight
Ergometrine 0.5 mg by slow i.v. dissolvable suture
injection 2. Repair vaginal tear if possible
Oxytocin infusion 40 units in 500 Epithelial repair with continuous
ml at 125 ml/h dissolvable suture
Prostaglandin F2 intramuscular Individual figure-of-eight ligation of
(Carboprost 250 µg i.m.) bleeding vessels
2. If no better, consider lower genital Vaginal pack & catheter 24 h (+
tract bleeding and move to second-line antibiotic cover)
management
Failure to control bleeding
If continuing bleeding from vaginal tear despite vaginal pack consider:

1. Alternative form of vaginal tamponade
Blood pressure cuff in glove inflated to just above systolic pressure26*
Rüsch catheter or Sengstaken–Blakemore tube (aspiration channel for drainage of lochia)
2. If the cervical tear extends into the uterus, laparotomy and hysterectomy may be required
3. Angiographic embolization of bleeding vessels
4. Bilateral internal iliac artery ligation
198
220
30 August 2006 14:21:01
principle is the same for vaginal packing) to (a)

allow for easy removal16. Generally, packs are
left in place for 24–36 h before removal17. A
urinary Foley catheter and broad-spectrum
antibiotic cover should be given where packs
are used. Balloon tamponade using Rüsch
catheters18 or Blakemore-Sengstaken19 tubes,
as described for treatment of uterine bleeding
(see Chapters 28 and 29), can also be used.
Pinborg and colleagues20 described the
successful use of the blood pressure cuff in
two patients to control intractable vaginal
bleeding following evacuation of vaginal
hematoma that developed after spontaneous
vaginal delivery. A blood pressure cuff was
inserted into a sterile glove, which in turn was
inserted into the vagina and the pressure then
(b)
gradually increased to 120 mmHg, 10 mmHg
above the systolic pressure, to stop the bleeding.
Eight hours later, the pressure of the cuff
was reduced by 10 mmHg/h and the cuff then
taken out after 32 h. Both patients made an
uneventful recovery.
Cervical tear
Any cervical tear extending above the internal
os warrants laparotomy. Small, non-bleeding
lacerations of the cervix do not need to be
sutured. Any bleeding cervical tear, and
certainly any tear longer than 2 cm, however,
should be sutured by using an absorbable suture
on a tapered (rather than a cutting) needle.
A suitable method for suturing is shown in (c)
Figure 2.
Both edges of the most caudal part of the
laceration are grasped with a ring forceps and
then sutured with an interrupted or figure-of-
eight stitch. This is then held with a hemostat to
bring down into view the next part of the tear,
which is sutured in the same way, and so on
until the apex is secured. The laceration should
be observed for a few minutes after suturing, to
ensure adequate hemostasis. The ring forceps
can be replaced and left on for some time if
oozing persists.
Cervical and vaginal vault lacerations that
continue to ooze despite treatment as detailed
above or those that are associated with hema-
tomas may be amenable to selective arterial
embolization (see below). Figure 2 (a)–(c) Suturing cervical tear
199
221
04 September 2006 17:20:15
Hematoma management possibility that a supralevator/broad liga-

ment hematoma is due to a ruptured uterus
The literature on the management of para-
or where a cervical tear appears to have
genital hematomas is limited and no random-
extended up into the uterus. At laparotomy,
ized studies of the efficacy of various treatments
if there is continuing bleeding from the
exist21.
upper vagina, then the anterior division of
the internal iliac artery should be ligated in
Infralevator hematomas continuity, which will reduce the pulse
pressure to the distal internal iliac artery
As always, initial management consists of resus-
branches (that supply the uterus and
citation measures and analgesia followed by a
vagina) by 85% and the blood flow by
period of observation. For hematomas that are
about 50%24 (see Chapters 32 and 34). A
less than 5 cm and not expanding, conservative
further vaginal pack should be inserted.
treatment with ice packs, pressure dressing and
analgesia is recommended22. The visible skin (2) Selective arterial embolization Where there
margin of the hematomas should be marked is continuing expansion of a supralevator
to help establish whether it is expanding. For hematoma without extension into the cervix
hematomas that are expanding or more than or uterus, selective arterial embolization is
5 cm in size, surgical intervention is recom- seen as the treatment of choice25 over inter-
mended. Where possible, the surgical incision nal iliac artery ligation, which in itself has an
should be made via the vagina to minimize uncertain chance of success26 and involves
visible scarring. Distinct bleeding points should imposing a laparotomy on an already unsta-
be under-run with figure-of-eight dissolvable ble patient. The blood supply to the upper
sutures. The presence of any residual bleeding vagina is from a rich anastomotic network
or a hematoma cavity is an indication for of vessels, arising mainly from branches of
insertion of a drain, a vaginal pack and a Foley the anterior trunk of the internal iliac artery
catheter, all of which should be left in place (vaginal, uterine, middle rectal arteries) and
for at least 24 h. Usually, however, no distinct the internal pudendal artery, which is the
bleeding point can be seen, in which case a most inferior branch of the posterior trunk
drain and pack should be inserted10. of the internal iliac artery. The technique of
selective arterial embolization investigates
these vessels by preliminary transfemoral
Supralevator hematomas
arteriography, followed by embolization
Approximately 50% of broad ligament hema- using Gelfoam (gelatin) pledglets. Pelage
tomas present early with symptoms of lower and colleagues25 reported a series of 35
abdominal pain, hemorrhage and in severe cases, patients who underwent this procedure
shock. The other 50% present after 24 h. Broad for unanticipated postpartum hemorrhage.
ligament and retroperitoneal hematomas are ini- Bleeding was controlled in all but one,
tially managed expectantly if the patient is stable who required hysterectomy 5 days later for
and the lesions are not expanding23. Ultrasound, re-bleeding. All women who had successful
CT scanning and MRI may all be used to assess embolization resumed menstruation. The
the size and progress of these hematomas. Close procedure, however, is not without risk and
observation, intravenous fluid resuscitation, deaths have been reported due to sepsis and
blood transfusion, vaginal packing or balloon/ multiple organ failure27 (see Chapter 44).
blood pressure cuff tamponade and antibiotics
are commenced as appropriate, but, if it is not
possible to maintain a stable hemodynamic state, SUMMARY
then active intervention is indicated, with options
In summary, bleeding from the lower genital
including the following:
tract should always be considered as a possible
(1) Laparotomy ± total abdominal hysterectomy cause of primary postpartum hemorrhage
This is indicated where there is any where there is continuing bleeding despite a
200
222
30 August 2006 14:21:06
well-contracted uterus. Primary repair of vagi- other methods compared. Br J Obstet Gynaecol
nal or cervical tears with full-thickness sutures 1982;89:501–6
using a dissolving suture on a taper-cut needle, 12. Management of Postpartum haemorrhage – A
followed by insertion of a vaginal pack and Clinical Practice Guideline for Professionals
involved in Maternity Care in Scotland. Aberdeen:
catheter for at least 24 h will stem most bleed-
Scottish Programme for Clinical Effectiveness in
ing. Urgent resort to laparotomy is necessary if
Reproductive Health, 1998
there is a cervical tear extending beyond the 13. Bonnar J. Massive obstetric hemorrhage.
internal cervical os up into the uterus, or if Baillieres Best Pract Res Clin Obstet Gynaecol
bleeding fails to settle despite an attempt at 2000;14:1–18
vaginal tamponade. Internal iliac artery ligation 14. Johanson R. Continuous vs. interrupted sutures
or selective arterial embolization should be for perineal repair. In Keirse M, Renfrew M,
considered where there is continuing expansion Neilson J, Crowther C, eds. Pregnancy and
of a supralevator hematoma or upper vaginal Childbirth Module. The Cochrane Pregnancy and
bleeding despite the above measures. As always, Childbirth Database. London: BMJ Publishing
regular assessments, clear documentation, a Group, 1994
15. Bobrowski R, Jones T. A thrombogenic uterine
proactive approach and early intervention are
pack for postpartum hemorrhage. Obstet Gynecol
vital to obtain a good outcome.
1995;85:836–7
16. Wax J, Channell J, Vandersloot J. Packing of the
lower uterine segment: new approach to an old
References
technique? Int J Gynaecol Obstet 1993;43:197–8
1. Smellie W. A Treatise on the Theory and Practice of 17. Maier R. Control of postpartum haemorrhage
Midwifery, 1792 with uterine packing. Am J Obstet Gynecol 1993;
2. Millward-Sadler H. Why Mothers Die 2000–2002. 169:317
The Confidential Enquiries into Maternal Deaths in 18. Johanson R, Kumar M, Obhrai M, et al. Man-
the United Kingdom. London: Royal College of agement of massive postpartum haemorrhage:
Obstetricians and Gynaecologists, 2004:227 use of a hydrostatic balloon catheter to avoid
3. Etuk S, Asuqo E. Effects of community and laparotomy. Br J Obstet Gynaecol 2001;108:
health facility interventions on postpartum 420–2
haemorrhage. Int J Gynaecol Obstet 2000;70: 19. Katesmark M, Brown R, Raju K. Successful
381–3 use of a Sengstaken-Blakemore tube to control
4. Stones R, Paxton C, Saunders N. Risk factors massive postpartum haemorrhage. Br J Obstet
for major obstetric haemorrhage. Eur J Obstet Gynaecol 1994;101:259–60
Gynecol Reprod Biol 1993;48:15–18 20. Pinborg A, Bodker B, Hogdall C. Postpartum
5. Drife J. Management of primary postpartum haematoma and vaginal packing with a blood
haemorrhage. Br J Obstet Gynaecol 1997;104: pressure cuff. Acta Obstet Gynecol Scand 2000;
275–7 79:887–9
6. Hankins G, Zahn C. Puerperal haematomas and 21. Ridgway LE. Puerperal emergency. Vaginal and
lower genital tract lacerations. In Hankins G, vulvar haematomas. Obstet Gynecol Clin North
et al., eds. Operative Obstetrics. Connecticut: Am 1995;22:275–83
Appleton & Lange, 1995:57–72 22. Zahn C, Yeomans E. Postpartum haemorrhage:
7. Cheung TH, Chang A. Puerperal haematomas. placenta accrete, uterine inversion and puerperal
Asia-Oceania J Obstet Gynaecol 1991;17:119–23 haematomas. Clin Obstet Gynaecol 1990;33:422
8. Combs C, Murphy E, Laros R. Factors associ- 23. Lingam K, Hood V, Carty M. Angiographic
ated with postpartum hemorrhage with vaginal embolisation in the management of pelvic haem-
birth. Obstet Gynecol 1991;77:69–76 orrhage. Br J Obstet Gynaecol 2000;107:1176–8
9. Stones R, Paterson C, Saunders N. Risk factors 24. Burchell R. Physiology of internal iliac artery
for major obstetric haemorrhage. Eur J Obstet ligation. J Obstet Gynaecol Br Commonwealth
Gynecol Reprod Biol 1993;48:15–18 1968;75:642–51
10. James D, Steer P, Weiner C, et al. High-risk 25. Pelage J, Le Dref O, Jacob D, et al. Selective
Pregnancy Management Options, 2nd edn. arterial embolisation of the uterine arteries in
London: WB Saunders, 1999:1187–204 the management of intractable postpartum
11. Healy D, Quinn M, Pepperell R. Rotational haemorrhage. Acta Obstet Gynecol Scand 1999;
delivery of the fetus: Kielland’s forceps and two 78:698–703
201
223
30 August 2006 14:21:06
26. Evans S, McShane P. The efficacy of internal 27. Ledee N, Ville Y, Musset D, et al. Management
iliac artery ligation in obstetric haemorrhage. in intractable obstetric haemorrhage: an audit
Surg Gynecol Obstet 1985;160:250–3 study on 61 cases. Eur J Obstet Gynecol Reprod
Biol 2001;94:189–96
202
224
30 August 2006 14:21:06
24
ADHERENT PLACENTA: NEW MANAGEMENT OPTIONS
G. Kayem, T. Schmitz, V. Tsatsaris, F. Goffinet and D. Cabrol
INTRODUCTION MANAGEMENT OF ADHERENT

PLACENTA
Placenta accreta occurs when a defect of the
decidua basalis results in abnormally invasive The classical approach most often recommended
placental implantation1. It is often diagnosed in cases of placenta accreta is extirpative4. If risk
only after delivery when manual removal of the factors and prenatal imaging both strongly sug-
placenta has failed. Attempting forcible manual gest this diagnosis, a Cesarean hysterectomy is
removal of a placenta accreta can easily lead to generally planned, especially for patients who
dramatic hemorrhage that may result in hyster- do not wish continued fertility. If the placenta
ectomy. Thus, placenta accreta and especially accreta is discovered after delivery, the placenta
placenta percreta reportedly result in a maternal is removed as soon as possible to empty the
mortality rate of 7%, and cause intra- and post- uterine cavity. In most cases, however, this
operative morbidity associated with massive forced placental delivery induces massive
blood transfusions, infection, ureteral damage, hemorrhage and leads to hysterectomy.
and fistula formation2. Its incidence, along with When the diagnosis of adherent placenta is
the Cesarean section rate, has increased 10-fold not suspected before labor and a postpartum
over the past 50 years3. With a frequency hemorrhage is obviously related to attempting
of approximately 1 per 1000 deliveries, this forcible removal of a placenta accreta, several
disorder has become more common in today’s options are possible, dependent on the patient’s
medical practice4. wishes and the cervical situation.
If there is no wish for continued fertility
or if the hemodynamic status is unstable, a
DIAGNOSIS OF PLACENTA ACCRETA
hysterectomy must be performed. Otherwise, an
In practice, placenta accreta is diagnosed accord- attempt can be made to preserve the uterus
ing to clinical or histological criteria as follows5. using surgical (ligating hypogastric arteries) or
If suspected before labor, prenatal diagnosis of radiological (embolization of the uterine arter-
placenta accreta is confirmed by the failure of its ies) techniques (see Chapters 30 and 32). Other
gentle attempted removal during the third stage methods have been published in case reports
of labor. If not suspected before delivery, pla- describing uterine packing, oversewing the pla-
centa accreta can be diagnosed if manual cental bed, prostaglandin administration, direct
removal of the placenta is partially or totally aortic compression and argon beam coagulation
impossible and no cleavage plane exists between in order to decrease blood loss6. More recently,
part or the entire placenta and the uterus; a a simple method using parallel sagittal ligatures
heavy bleeding occurs from the implantation of the lower segment has been described; it is
site after forced placental removal. particularly useful if the hemorrhage is located
After a hysterectomy performed because of to the lower segment7. Other similar methods,
postpartum hemorrhage, placenta accreta is more complex to perform, have also been des-
shown by histologic confirmation of accreta on cribed, but seem to be associated with serious
the hysterectomy specimen. side-effects (uteropyosis, synechia)8–10.
203
225
30 August 2006 14:21:06
We believe these methods can be used only DESCRIPTION OF CONSERVATIVE

when the diagnosis of adherent placenta has MANAGEMENT
been made after attempting forcible removal
Depending on how the placenta accreta is
and in case of severe hemorrhage.
discovered, two different types of conservative
An alternative therapeutic approach to the
treatment can be used.
placenta is conservative rather than extirpative.
Some cases of successful conservative manage- (1) When discovered during the third stage of
ment of placenta accreta have previously been labor, removal of the placenta is not forced;
reported11–15. the conservative treatment leaves the pla-
Conservative strategy was initiated in our centa, in part or entirely, in the uterus when
center in 1997 and followed the successful the patient’s hemodynamic status is stable
conservative management of one case of and no septic risk is present.
placenta accreta, by leaving the placenta in
(2) When the placenta accreta is strongly sus-
place16. Since this date, our protocol is to
pected before delivery (based on history and
manage most cases of placenta accreta con-
ultrasound and/or magnetic resonance
servatively, leaving in situ each placenta that
imaging suggestive of the diagnosis), the
adheres either partially or totally to the
case is discussed at the daily obstetric
myometrium. We evaluated this management
staff meeting and conservative treatment
by a historical consecutive study to compare the
is proposed to the patient. In this case,
impact of conservative and extirpative strategies
management includes the following steps
for placenta accreta on maternal morbidity and
(Figure 1). The precise position of the
mortality17.
placenta is determined by ultrasound. A
Two consecutive periods, A and B, were
Cesarean section is planned, with the
compared. During period A (January 1993 to
abdominal incision at the infraumbilical
June 1997), our written protocol called for the
midline, enlarged above the umbilicus if
systematic manual removal of the placenta, to
necessary, and a vertical uterine incision at
leave the uterine cavity empty. In period B (July
a distance from the placental insertion.
1997 to December 2002), we changed our
After extraction of the infant, delivery of the
policy by leaving the placenta in situ. The
placenta is attempted prudently, with an
following outcomes over the two periods were
intravenous injection of 5 IU oxytocin and
compared: need for blood transfusion, hysterec-
moderate cord traction. If this fails, the pla-
tomy, intensive care unit admission, duration
centa is considered to be ‘accreta’. The cord
of stay in intensive care unit, and postpartum
is cut at the placental insertion and the pla-
endometritis. Thirty-three cases of placenta
centa left in the uterine cavity; the uterine
accreta were observed among 31 921 deliveries
incision is closed. Prophylactic antibiotic
(1.03/1000). During period B, there was a
therapy (amoxicillin and clavulanic acid) is
reduction in the hysterectomy rate (from
administered for 10 days.
11 (84.6%) to 3 (15%); p < 0.001), the
mean number of red blood cells transfused
(3230 ± 2170 ml vs. 1560 ± 1646 ml; p < 0.01),
FOLLOW-UP AFTER CONSERVATIVE
and disseminated intravascular coagulation
MANAGEMENT
(5 (38.5%) vs. 1 (5.0%); p = 0.02), compared
with period A. There were three cases of sepsis During the postpartum period, all patients are
in period B and none in period A (p = 0.26). seen weekly until complete resorption of the
One hysterectomy was required at day 26, placenta. Ultrasonography and clinical exami-
because of sepsis and hemorrhage, after a nation are performed to detect hemorrhage,
conservative management of an entire pain or clinical signs of infection. To improve
placenta accreta. Two women with conservative clinical follow-up and to help choose antibiotic
management have subsequently had successful therapy in cases of endometritis with or without
pregnancies. sepsis, C-reactive protein and blood counts are
204
226
30 August 2006 14:21:06
Adherent placenta
Prenatal suspicion of placenta accreta

(placenta previa + previous Cesarean
section)
Discussion with the patient

Medical staff meeting
Patient does not Cesarean

Patient wishes for continued wish for continued section +
fertility fertility hysterectomy
Cesarean section with:

Ultrasound location of the placenta
Vertical hysterotomy at a distance from the placenta
Fetal delivery
Delivery of the placenta is attempted prudently, with oxytocin 5 IU Success: placenta

injection and moderate cord traction normally inserted
Failure: Confirm the diagnosis of placenta accreta
Section of the umbilical cord

Closure of the uterine incision
Sulprostone (8.3 ml/min for 1 h)

Systematic
Radiological embolization except if there is no bleeding after
surgical treatment
Follow-up once a week

- Clinical examination (bleeding, fever, pelvic pain)
- Hemoglobin level, leukocyte numeration, C-reactive protein,
vaginal sample for bacteriological examination
- Ultrasound examinations (size of the retained placenta)
Figure 1 Conservative management of placenta accreta that is strongly suspected before delivery
205
227
30 August 2006 14:21:07
assayed and vaginal samples are taken for accreta is strongly suspected before labor,
bacteriological study. should be preferable to confirm the diagnosis.
COMPLICATIONS OF CONSERVATIVE
OPTIMAL ADJUVANT THERAPY IN
MANAGEMENT
CONSERVATIVE MANAGEMENT
Conservative management is a strategy that
Methotrexate, uterine artery embolization and
must be applied with discretion. Complications
sulprostone are three adjuvant treatments des-
are possible and include sepsis and hemorrhage
cribed in several case reports involving conser-
with failure of conservative management21,27. In
vative treatment14,18–21. The outcome when
case of secondary hemorrhage and/or sepsis
the placenta is left in place after methotrexate
following a conservative management, hysterec-
administration varies widely; it ranges from
tomy may become necessary. At present, the
expulsion at 7 days to progressive resorption
number of patients managed with this strategy
in roughly 6 months14,18–20. We do not use
is too low for an adequate evaluation of the
methotrexate at all. Similarly, only a few reports
risk of rare severe maternal morbidity or
describe the outcome after embolization and
mortality. Accordingly, this type of manage-
leaving the placenta in situ22. In our practice, we
ment is presently appropriate only when rigor-
perform, almost systematically, embolization of
ous monitoring can follow, in centers with
uterine arteries to diminish or prevent a post-
adequate equipment and resources26.
partum hemorrhage. Sulprostone is a well-
Ideally, these complications should be
known uterotonic agent utilized in case of
discussed prenatally with the patient to give her
postpartum hemorrhage. It can be used to pre-
complete information about the different thera-
vent or treat immediate abnormal postpartum
peutic strategies (extirpative or conservative).
bleeding. Data do not currently prove the bene-
Given the difficulties mentioned above for pre-
fit of adding this therapy to conservative treat-
natal diagnosis, this discussion is rarely possible.
ment; however, its utilization may contribute to
Accordingly, one possible option is to preserve
the prevention of major postpartum bleeding in
maternal fertility and to diminish the risk of
the 2 or 3 days after delivery.
hemorrhage when placenta accreta is discovered
during delivery.
PRENATAL IDENTIFICATION OF
PLACENTA ACCRETA FOR FERTILITY AFTER CONSERVATIVE
CONSERVATIVE MANAGEMENT MANAGEMENT AND RISK OF
RECURRENCE
Prenatal identification of placenta accreta
would facilitate the choices about management In our experience, seven patients managed
of delivery and allow the appropriate precau- conservatively were contacted from 1–5 years
tions (reinforcement of obstetric, anesthetic and afterwards, whereas ten were lost to long-term
radiology teams, blood transfusion readiness). follow-up. Of these seven, one had another
However, the sensitivity and specificity of successful pregnancy 2 years later and another
transvaginal or transabdominal ultrasound and had two consecutive successful pregnancies,
magnetic resonance imaging vary from 33% to both complicated by placenta accreta, located
95% in different studies; they depend greatly on at the same place, and treated conservatively
placenta location23–26. For these reasons, imag- again. The others chose, for various personal
ing should be considered only when placenta reasons, not to become pregnant again. None
accreta is suspected for clinical reasons (mainly sought subsequent treatment for sterility.
placenta previa associated with previous Cesar- The possibility of recurrence should thus be
ean section). Moreover, systematic attempts at a discussed with the woman when deciding on
careful and gentle intraoperative delivery of the the initial conservative management. Moreover,
placenta (intravenous injection of 5 IU oxytocin in any subsequent pregnancies following a
and moderate contraction), even when placenta conservative management, the risk of placenta
206
228
30 August 2006 14:21:07
Adherent placenta
accreta should be monitored carefully by appro- 11. Legro RS, Price FV, Hill LM, Caritis SN.
priate investigations, particularly if the placenta Nonsurgical management of placenta percreta: a
is located in the same site as before. case report. Obstet Gynecol 1994;83:847–9
12. Hollander DI, Pupkin MJ, Crenshaw MC,
Nagey DA. Conservative management of
CONCLUSIONS placenta accreta. A case report. J Reprod Med
1988;33:74–8
Conservative management of placenta accreta 13. Komulainen MH, Vayrynen MA, Kauko ML,
appears to be a safe alternative to extirpative Saarikoski S. Two cases of placenta accreta man-
management. However, it must be applied cau- aged conservatively. Eur J Obstet Gynecol Reprod
tiously and should be proposed only in centers Biol 1995;62:135–7
with adequate resources, and the capability of 14. Mussalli GM, Shah J, Berck DJ, Elimian A,
securing a strict follow-up in order to detect and Tejani N, Manning FA. Placenta accreta
treat subsequent complications. and methotrexate therapy: three case reports.
J Perinatol 2000;20:331–4
15. Clement D, Kayem G, Cabrol D. Conservative
References treatment of placenta percreta: a safe alternative.
Eur J Obstet Gynecol Reprod Biol 2004;114:
1. Khong TY, Robertson WB. Placenta creta 108–9
and placenta praevia creta. Placenta 1987;8: 16. Kayem G, Pannier E, Goffinet F, Grange G,
399–409 Cabrol D. Fertility after conservative treatment
2. O’Brien JM, Barton JR, Donaldson ES. The of placenta accreta. Fertil Steril 2002;78:637–8
management of placenta percreta: conservative 17. Kayem G, Davy C, Goffinet F, Thomas C,
and operative strategies. Am J Obstet Gynecol Clement D, Cabrol D. Conservative versus
1996;175:1632–8 extirpative management in cases of placenta
3. Miller DA, Chollet JA, Goodwin TM. Clinical accreta. Obstet Gynecol 2004;104:531–6
risk factors for placenta previa-placenta accreta. 18. Arulkumaran S, Ng CS, Ingemarsson I, Ratnam
Am J Obstet Gynecol 1997;177:210–14 SS. Medical treatment of placenta accreta with
4. ACOG Committee Opinion. Placenta accreta. methotrexate. Acta Obstet Gynecol Scand 1986;
Number 266, January 2002. American College 65:285–6
of Obstetricians and Gynecologists. Int J 19. Buckshee K, Dadhwal V. Medical management
Gynaecol Obstet 2002;77:77–8 of placenta accreta. Int J Gynaecol Obstet 1997;
5. Gielchinsky Y, Rojansky N, Fasouliotis SJ, 59:47–8
Ezra Y. Placenta accreta – summary of 20. Gupta D, Sinha R. Management of placenta
10 years: a survey of 310 cases. Placenta 2002; accreta with oral methotrexate. Int J Gynaecol
23:210–14 Obstet 1998;60:171–3
6. Scarantino SE, Reilly JG, Moretti ML, Pillari 21. Jaffe R, DuBeshter B, Sherer DM, Thompson
VT. Argon beam coagulation in the management EA, Woods JR. Failure of methotrexate treat-
of placenta accreta. Obstet Gynecol 1999;94: ment for term placenta percreta. Am J Obstet
825–7 Gynecol 1994;171:558–9
7. Hwu YM, Chen CP, Chen HS, Su TH. Parallel 22. Lemercier E, Genevois A, Descargue G, Clavier
vertical compression sutures: a technique to con- E, Benozio M. [MRI evaluation of placenta
trol bleeding from placenta praevia or accreta accreta treated by embolization. Apropos of a
during caesarean section. Br J Obstet Gynaecol case. Review of the literature]. J Radiol 1999;80:
2005;112:1420–3 383–7
8. Wu HH, Yeh GP. Uterine cavity synechiae after 23. Lam G, Kuller J, McMahon M. Use of magnetic
hemostatic square suturing technique. Obstet resonance imaging and ultrasound in the antena-
Gynecol 2005;105:1176–8 tal diagnosis of placenta accreta. J Soc Gynecol
9. Ochoa M, Allaire AD, Stitely ML. Pyometria Investig 2002;9:37–40
after hemostatic square suture technique. Obstet 24. Finberg HJ, Williams JW. Placenta accreta:
Gynecol 2002;99:506–9 prospective sonographic diagnosis in patients
10. Cho JY, Kim SJ, Cha KY, Kay CW, Kim MI, with placenta previa and prior cesarean section.
Cha KS. Interrupted circular suture: bleeding J Ultrasound Med 1992;11:333–43
control during cesarean delivery in placenta 25. Levine D, Hulka CA, Ludmir J, Li W, Edelman
previa accreta. Obstet Gynecol 1991;78:876–9 RR. Placenta accreta: evaluation with color
207
229
30 August 2006 14:21:07
Doppler US, power Doppler US, and MR 27. Butt K, Gagnon A, Delisle MF. Failure of
imaging. Radiology 1997;205:773–6 methotrexate and internal iliac balloon
26. ACOG educational bulletin. Postpartum hemor- catheterization to manage placenta percreta.
rhage. Number 243, January 1998 (replaces No. Obstet Gynecol 2002;99:981–2
143, July 1990). American College of Obstetri-
cians and Gynecologists. Int J Gynaecol Obstet
1998;61:79–86
208
230
30 August 2006 14:21:07
25
ACQUIRED AND CONGENITAL HEMOSTATIC DISORDERS IN
PREGNANCY AND THE PUERPERIUM
R. V. Ganchev and C. A. Ludlam
During normal pregnancy, a series of progres- trophoblasts may contribute to the pro-
sive changes in hemostasis occur that are overall coagulant effect1. Although concentrations of
procoagulant and help prevent excessive bleed- soluble tissue factor (TF) remain constant
ing at the time of delivery. The concentrations during normal pregnancy2, monocyte TF
of coagulation factors V, VII, VIII, IX, X, XII activity and expression are lower when
and von Willebrand factor (vWF) rise signifi- compared with those in non-pregnant women,
cantly (Table 1) and are accompanied by a pro- possibly acting to counterbalance the pro-
nounced increase in fibrinogen levels (up to coagulant changes3,4. Local hemostasis at the
two-fold from non-pregnant levels). Factor XIII placental trophoblast level is characterized by
levels tend to decrease in late pregnancy after an increased TF expression and low expression
initial increase in the beginning of pregnancy. of tissue factor pathway inhibitor (TFPI)1.
Markers of coagulation activation such as Approximately 4 weeks’ post-delivery, the
prothrombin fragments (PF1+2), thrombin– hemostatic system returns to that of the
antithrombin complexes (TAT) and D-dimer non-pregnant state5.
are increased, while a decrease in physiological Although the overall balance shifts towards
anticoagulants is manifested by a significant hypercoagulability, occasionally medical condi-
reduction in protein S activity and acquired tions coincident with pregnancy and complica-
activated protein C (APC) resistance. Fibrinoly- tions of pregnancy itself put excessive demands
sis is inhibited not only by the rise in endothe- on maternal physiology and may result in a
lium-derived plasminogen activator inhibitor-1 bleeding tendency. This chapter describes
(PAI-1) but also by placenta-derived PAI-2. acquired and congenital hemostatic disorders
Microparticles derived from maternal endo- that may lead to hemorrhagic complications in
thelial cells and platelets, and from placental the obstetric patient.
Table 1 Coagulation system changes in normal pregnancy
Increased Decreased No change

Systemic changes
Procoagulant factors I, V, VII, VIII, IX, X, XII XIII PC, AT
Anticoagulant factors soluble TM PS
Adhesive proteins vWF
Fibrinolytic proteins PAI-1, PAI-2 t-PA soluble TF
Tissue factor (TF) monocyte TF
Microparticles (MP) MP
Local placental changes TF TFPI
TM, thrombomodulin; PS, protein S; PC, protein C; AT, antithrombin; vWF, von Willebrand factor; PAI,
plasminogen activator inhibitor; t-PA, tissue plasminogen activator; TFPI, tissue factor pathway inhibitor.
Adapted from Brenner B. Haemostatic changes in pregnancy. Thromb Res 2004;114:409–14
209
231
30 August 2006 14:21:07
ACQUIRED DISORDERS OF thrombocytopenia in pregnancy, affecting 5%

HEMOSTASIS of all pregnant women and accounting for more
than 75% of cases of pregnancy-associated
Thrombocytopenia
thrombocytopenia7,8. It presents as a mild to
Thrombocytopenia is the most common hemo- moderate thrombocytopenia (100–150 × 109/l),
static abnormality and may complicate up to which is detected incidentally often for the first
10% of all pregnancies. The normal platelet time during the third trimester of pregnancy.
count ranges from 150 to 400 × 109/l, and The platelet count returns to normal within 7
thrombocytopenia is defined as a count of days of delivery. GT is the physiologic thrombo-
less than 150 × 109/l. The platelet count may cytopenia that accompanies normal pregnancy
decline by approximately 10% during normal and is thought to be due to hemodilution and/or
pregnancy6. Spontaneous bleeding is unusual accelerated platelet clearance7,8. It is an entirely
unless the count has fallen to below 30 × 109/l, benign condition, which is not associated with
but surgical bleeding or postpartum hemor- maternal hemorrhage or fetal or neonatal
rhage may occur as a consequence of platelets thrombocytopenia. It is, however, necessary to
less than 50 × 109/l. Thrombocytopenia in preg- monitor the platelet count during pregnancy
nancy may result from variety of causes (Table and, if it falls below 100 × 109/l, the diagnosis
2). The timing of onset of these disorders must be reviewed. Rare cases, subsequently
during pregnancy and their clinical manifesta- confirmed as GT, have had counts as low as
tions often overlap, making the identification 50 × 109/l9. Epidural anesthesia is considered
of individual causes of thrombocytopenia safe if the maternal platelet count is greater than
sometimes problematic. 80 × 109/l. Delivery should proceed according
It is important to consider spurious thrombo- to obstetric indications and the cord platelet
cytopenia as a possible cause of decreased count should be checked. GT is difficult to
platelet count before embarking on extensive distinguish from idiopathic thrombocytopenic
investigations or treatment. This is a laboratory purpura, when thrombocytopenia is identified
artefact due to EDTA-induced platelet aggrega- for the first time during pregnancy and no
tion in vitro and can be diagnosed by visual previous counts have been documented.
inspection of the blood film, when platelet
changes are readily visible.
Idiopathic thrombocytopenic purpura
Idiopathic thrombocytopenic purpura (ITP)
Gestational thrombocytopenia
accounts for one to five cases of thrombocyto-
Gestational, or incidental, thrombocytopenia penia per 10 000 pregnancies10 and 5% of cases
(GT) is the most common cause of of pregnancy-associated thrombocytopenia7; it
Table 2 Causes of pregnancy-associated thrombocytopenia
Pregnancy-specific Not pregnancy-specific

Gestational (incidental) thrombocytopenia Idiopathic thrombocytopenic purpura (ITP)
Pre-eclampsia Thrombotic thrombocytopenic purpura (TTP)
HELLP syndrome (hemolysis, elevated liver enzymes Hemolytic uremic syndrome (HUS)
and low platelets) Systemic lupus erythematosus
Acute fatty liver of pregnancy (AFLP) Viral infection (HIV, CMV, EBV)
Antiphospholipid antibodies
Consumptive coagulopathy
Drug-induced thrombocytopenia
Type 2B von Willebrand disease
Congenital
From McCrae KR. Thrombocytopenia in pregnancy: differential diagnosis, pathogenesis and management.
Blood Rev 2003;17:7–14
210
232
30 August 2006 14:21:08
Acquired and congenital hemostatic disorders
is the most common cause of significant regarded as safe for normal vaginal delivery, and
thrombocytopenia in the first trimester. ITP is those > 80 × 109/l are safe for Cesarean section,
characterized by premature clearance of plate- spinal or epidural anesthesia14.
lets by antiplatelet antibodies and consequent The major treatment options for maternal
increased production of platelets by the bone ITP are corticosteroids or intravenous immuno-
marrow. The most common presentation is the globulin (IVIg). There is no evidence, however,
finding of an asymptomatic thrombocytopenia that either of these treatment modalities admin-
on a routine blood count, when the distinction istered to the mother affects the platelet count
from GT may be difficult. Patients occasionally in the fetus or neonate. If the duration of treat-
present for the first time with severe thrombo- ment is likely to be short, i.e. starting in the
cytopenia in pregnancy, and women with third trimester, corticosteroids are an effective
previously diagnosed ITP often experience an option. An initial dose of 1 mg/kg prednisolone
exacerbation in pregnancy11. Symptomatic (based on pregnancy weight) is recom-
patients present with minor bruises or mended11,14, which can be subsequently
petechiae, bleeding from mucosal surfaces, or tapered. In addition to their toxicities in non-
rarely fatal intracranial bleeding. pregnant individuals, such as osteoporosis and
As in the non-pregnant patient, ITP is a diag- weight gain, corticosteroids increase the
nosis of exclusion with thrombocytopenia and incidence of pregnancy-induced hypertension
normal or increased megakaryocytes in the bone and gestational diabetes, and may promote
marrow in the absence of other causes. There is premature rupture of the fetal membranes.
no confirmatory laboratory test, and documen- Concerns about potential adverse maternal
tation of a low platelet count outside pregnancy effects of steroids have led some to use IVIg
is invaluable. Practically, however, in the as a first-line therapy in pregnancy15,16. Others
absence of a platelet count prior to pregnancy, reserve this treatment for patients in whom
significant thrombocytopenia (< 100 × 109/l) in steroid therapy is likely to be prolonged or in
the first trimester, with a declining platelet whom an unacceptably high maintenance dose
count as gestation progresses, is most consistent is required (> 7.5 mg prednisolone daily). The
with ITP. In contrast, mild thrombocytopenia conventional dose of IVIg is 0.4 g/kg/day for 5
developing in the second or the third trimester days, although 1 g/kg/day for 2 days has been
and not associated with hypertension or used successfully and may be more conve-
proteinuria most likely represents GT12. Bone nient11. A persistent and predictable response is
marrow examination is unnecessary unless obtained in 80% of the cases. The response to
there is suspicion of leukemia, lymphoma or therapy usually occurs within 24 h (more rapid
malignant infiltration. than with steroids) and is maintained for 2–3
The decision to treat a pregnant woman with weeks. After an initial response, repeat single
ITP is based on assessment of the risk of signifi- infusions can be used to prevent hemorrhagic
cant maternal hemorrhage. The count usually symptoms and ensure an adequate platelet
falls as pregnancy progresses, with a nadir in the count for delivery.
third trimester11, and active treatment may have Therapeutic options for those women with
to be instituted to ensure a safe platelet count severely symptomatic ITP refractory to oral
at the time of delivery. The incidence of ante- steroids or IVIg include high-dose intravenous
partum hemorrhage is not increased in maternal methylprednisolone (1.0 g), perhaps combined
ITP, but there is a small increased risk of post- with IVIg, or azathioprine14, but these should
partum hemorrhagic complications, not from only be considered after careful assessment of
the placental bed but from surgical incisions the potential risks. Splenectomy is now rarely
such as episiotomies and from soft-tissue performed in pregnancy. It remains an option if
lacerations13. all other attempts to increase the platelet count
Asymptomatic patients with platelet counts fail and is best performed in the second
> 20 × 109/l do not require treatment until trimester.
delivery is imminent but should be carefully The offspring of mothers with ITP may also
monitored. Platelet counts of > 50 × 109/l are develop thrombocytopenia, as a result of the
211
233
30 August 2006 14:21:08
transplacental passage of maternal antiplatelet Secondary autoimmune

IgG7,12. The incidence of severe neonatal thrombocytopenia
thrombocytopenia (< 50 × 109/l) has been
Antiphospholipid syndrome
reported between 9 and 15%, with intracranial
hemorrhage occurring in 0–1.5% of infants17. The diagnosis of primary antiphospholipid
Due to the inability of maternal clinical charac- syndrome requires the coexistence of clinical
teristics to predict neonatal thrombocytopenia, manifestations (either vascular thrombosis or
antenatal (cordocentesis) and perinatal (fetal pregnancy morbidity) with laboratory evidence
scalp blood sampling) procedures for determi- of reproducible antiphospholipid antibodies
nation of fetal platelet count have been consid- (either lupus anticoagulant or anticardiolipin
ered in the past. Cordocentesis carries a antibody)19. Primary antiphospholipid syn-
mortality of 1–2%, however, whereas scalp drome is associated with autoimmune
blood sampling is associated with artefactually thrombocytopenia in 20–40% of cases20.
low results and risk of significant hemorrhage. Thrombocytopenia is rarely severe and usually
For these reasons, both procedures are now does not require treatment. If treatment is nec-
largely abandoned in the management of essary, management options during pregnancy
ITP in pregnancy. The most reliable predictor are similar to those for primary ITP. However,
of fetal thrombocytopenia is a history of primary antiphospholipid syndrome is associ-
thrombocytopenia at delivery in a prior ated with recurrent spontaneous abortions
sibling18. before 10 weeks of gestation, and women with
In view of the very low risk of serious the condition are at risk of intrauterine fetal
neonatal hemorrhage, it is now agreed that the growth restriction or death, pre-eclampsia and
mode of delivery in ITP should be determined maternal thrombosis19,21.
by purely obstetric indications11,14. If the A combination of low-dose aspirin and
maternal platelet count remains low at the low-dose subcutaneous heparin is helpful in
time of delivery, despite optimal antenatal preventing recurrent spontaneous abortions
management, platelet transfusion may be in antiphospholipid syndrome22. Antenatal and
required to treat maternal bleeding. Mothers postnatal thrombosis prophylaxis is indicated in
with thrombocytopenia are unlikely to bleed women with antiphospholipid syndrome and a
from the uterine cavity after the third stage history of thrombosis23. Moderate thrombo-
of labor, provided that there are no retained cytopenia should not alter decisions about
products of conception. However, bleeding may antiplatelet or antithrombotic therapy in
occur from surgical wounds, episiotomies or antiphospholipid syndrome20.
perineal tears. Non-steroidal anti-inflammatory
drugs should be avoided for postpartum analge-
Systemic lupus erythematosus
sia. ITP should not exclude women from
consideration for peripartum thrombosis pro- Immune platelet destruction may occur in sys-
phylaxis. Prophylactic doses of low-molecular temic lupus erythematosus (SLE) because of
weight heparin are generally safe if the platelet antiplatelet antibodies or immune complexes,
count is greater than 50 × 109/l. Following but thrombocytopenia is seldom severe; less
delivery, a cord blood platelet count should than 5% of cases have platelet count
be determined in all cases. Since the neonatal < 30 × 109/l during the course of the disease13.
platelet count may decline for 4–5 days after Thrombocytopenia is often the first presenting
delivery11, daily monitoring is indicated. Infants feature and may precede any other manifesta-
should be closely observed and treatment is tions of the condition by months or years. It is
rarely required. In those with clinical hemor- difficult to document any special effect of preg-
rhage or platelet count < 20 × 109/l, treatment nancy on SLE; the general consensus is that
with IVIg produces a rapid response. Life- pregnancy does not affect the long-term prog-
threatening hemorrhage should be managed nosis of SLE, but that pregnancy itself may be
with platelet transfusion combined with associated with more flare-ups, particularly in
IVIg11. the puerperium24. The management of isolated
212
234
30 August 2006 14:21:08
thrombocytopenia associated with SLE in preg- Table 3 Drugs causing thrombocytopenia

nancy is governed by the principles outlined
A. Immune mediated
for ITP. Women with SLE are also at risk for
Acetaminophen
pre-eclampsia which may be complicated by Aminosalicylic acid
thrombocytopenia. Amiodarone
Amphotericin B
Cimetidine
HIV-associated thrombocytopenia
Diclofenac
HIV-related thrombocytopenia can be caused Gold/gold salts
by increased platelet destruction by antiplatelet Levamisole
antibodies or immune complexes, commonly Methyldopa
during early-onset HIV. In advanced disease, Quinine and quinidine
Ranitidine
drugs and infection may lead to marrow dys-
Sulfasalazine
function that results in thrombocytopenia. Vancomycin
In one series of HIV-positive women, approxi-
mately 3% were thrombocytopenic and, in B. Unique antibody-mediated process
most cases, thrombocytopenia was believed to Heparin
be directly related to HIV infection25. Slightly C. Suppression of platelet production
fewer than half of the thrombocytopenic women Anagrelide
had a platelet count < 50 × 109/l, and 20% had Valproic acid
hemorrhagic complications25.
D. Suppression of all hematopoietic cells
Treatment with antiretroviral therapy tends
Chemotherapeutic agents
to improve the defective thrombopoiesis and
increase the platelet count in HIV-positive Adapted from George JN, Raskob GE, Shah SR,
patients, but some antiretroviral drugs may also et al. Drug-induced thrombocytopenia: a systematic
cause thrombocytopenia. When immune des- review of published case reports. Ann Intern Med
truction is believed to be a significant component 1998;129:886–90
of thrombocytopenia, IVIg may be required to
treat hemorrhagic symptoms or to increase the
platelet count before delivery in thrombocyto- (HIT). It occurs in 1–5% of patients receiving
penic HIV-positive women25. Corticosteroids unfractionated heparin but is considerably less
are also effective but may be associated with common in patients treated with low-molecular
increased risk of further immunosuppression weight heparins. HIT is caused by an antibody
and infection. Thrombotic thrombocytopenic directed against the heparin–platelet factor 4
purpura is found more frequently in HIV- complex, which can induce platelet activation
infected patients and should be treated and aggregation in vivo. Unlike other thrombo-
accordingly. Cesarean delivery reduces the risk cytopenias, HIT is complicated by arterial
of transmission of HIV from mother to fetus. and/or venous thrombosis which may be life-
threatening. Laboratory tests are available to
confirm the diagnosis. HIT has been reported in
Drug-induced thrombocytopenia
pregnancy26,27, although it may be less common
Drug-induced thrombocytopenia may be in pregnant than in non-pregnant individuals28.
caused by immune- or non-immune-mediated Fetal thrombocytopenia does not occur because
platelet destruction or suppression of platelet heparin does not cross the placenta. Heparin
production. Both are uncommon in pregnancy, should be withdrawn immediately on clinical
but drug-induced causes should be considered suspicion of HIT. If ongoing anticoagulation is
and excluded. Drugs which are commonly asso- urgently required, the heparinoid danaparoid
ciated with thrombocytopenia are shown in may be used in most patients. Danaparoid has
Table 3. been used successfully to treat HIT in preg-
A unique form of drug-induced thrombo- nancy27. Hirudin is an alternative in non-
cytopenia is heparin-induced thrombocytopenia pregnant patients, but experience is limited in
213
235
30 August 2006 14:21:08
pregnancy and its use is not recommended after 20 weeks of gestation6. Although the clini-
unless there is no suitable alternative29. Platelet cal manifestations of pre-eclampsia generally do
transfusion should be avoided in patients not become evident until the third trimester, the
with HIT. Because HIT is potentially life- lesions underlying this disorder occur early in
threatening, all women must have a platelet pregnancy and involve deficient remodelling of
count before treatment with heparin begins. the maternal uterine vasculature by placental
The count must be repeated on day 4 of first trophoblast cells30,31. Thrombocytopenia devel-
exposure to heparin or day 1 of repeat exposure ops in approximately 50% of patients, with the
and then at least weekly for the first 3 weeks. severity usually proportional to the severity of
the pre-eclampsia. Occasionally, the onset of
thrombocytopenia precedes other manifesta-
Thrombocytopenia with microangiopathy tions of pre-eclampsia7. Current understanding
Several syndromes are associated with thrombo- of the pathogenesis of thrombocytopenia in
cytopenia as a result of platelet activation, red pre-eclampsia is that it is due to excessive
cell fragmentation, and a variable degree of platelet activation, adhesion of platelets to
hemolysis (microangiopathic hemolytic anemia, damaged or activated endothelium, and/or
MAHA). Some syndromes are unique to clearance of IgG-coated platelets by the
obstetric practice. The differential diagnosis reticuloendothelial system7.
is particularly pertinent for obstetricians and is Activation of the coagulation cascade occurs
important because management options differ. in most patients with pre-eclampsia; however,
The differential diagnosis is summarized in screening coagulation tests such as activated
Table 4. partial thromboplastin time (APTT), pro-
thrombin time (PT) and fibrinogen are usually
normal. Regardless, more sensitive markers of
Pre-eclampsia and HELLP syndrome hemostatic activity such as D-dimer and TAT
Pre-eclampsia affects approximately 6% of all complexes are often elevated. In severe
pregnancies, most often those of primigravidas pre-eclampsia, the activation of coagulation
less than 20 or greater than 30 years of age7. results in consumption of clotting factors and
The criteria for the condition include hyperten- therefore prolongation of the clotting test times
sion and proteinuria > 300 mg/24 h developing and a fall in plasma fibrinogen.
Table 4 Differentiation of pregnancy-associated microangiopathies
Diagnosis TTP HUS HELLP Pre-eclampsia AFLP

Time of onset 2nd trimester postpartum 3rd trimester 3rd trimester 3rd trimester
Hemolysis +++ ++ ++ + +
Thrombocytopenia +++ ++ ++ ++ +/±
Coagulopathy - - ± ± +++
Liver disease ± ± +++ ± +++
Renal disease ± +++ + + ±
Hypertension rare ± ± +++ ±
CNS disease +++ ± ± ± +
Effect of delivery none none recovery recovery recovery
on disease
Management early plasma supportive supportive supportive supportive
exchange ± plasma consider plasma plasma exchange
exchange exchange if persists rarely required
TTP, thrombotic thrombocytopenic purpura; HUS, hemolytic uremic syndrome; HELLP, hemolysis,
elevated liver enzymes, and low platelets; AFLP, acute fatty liver of pregnancy.
Adapted from Horn EH. Thrombocytopenia and bleeding disorders. In James DK, Steer PJ, Weiner CP,
Gonik B, eds. High-Risk Pregnancy: Management Options, 3rd edn, Elsevier, 2006:901–24
214
236
30 August 2006 14:21:08
The HELLP (hemolysis, elevated liver disease occurs exclusively during pregnancy, the
enzymes and low platelets) syndrome is often incidence of both is increased in this setting,
considered to be a variant of pre-eclampsia and and up to 10% of all cases of TTP occur in
is the most common cause of severe liver disease pregnant patients6.
in pregnant women32. Criteria for the HELLP TTP is defined by a pentad of symptoms that
syndrome include microangiopathic hemolytic include MAHA, thrombocytopenia, neurologi-
anemia, aspartate aminotransferase (AST) cal abnormalities, fever, and renal dysfunction,
> 70 U/l and thrombocytopenia, with a platelet although the complete pentad is present at the
count < 100 × 109/l33. Patients may present time of diagnosis in less than 40% of patients34.
with severe epigastric and right upper quadrant The clinical manifestations of HUS are similar.
pain, which need not be accompanied by Neurological abnormalities are particularly a
hypertension and proteinuria. Exacerbation of feature of patients with TTP; renal dysfunction
HELLP syndrome may occur postpartum and is more severe in patients with HUS. Congenital
there is a recurrence risk of approximately 3% in or acquired deficiency of a specific von Wille-
subsequent pregnancies. The syndrome occa- brand factor-cleaving protease, ADAMTS 13,
sionally presents postpartum, usually within and the consequent increased level of high-
48 h, but rarely as late as 6 days after delivery. molecular weight multimers of vWF play a
Despite their similarities, HELLP is associated central role in the pathogenesis of TTP. Inter-
with significantly greater maternal and fetal estingly, levels of ADAMTS 13 decrease during
morbidity and mortality than pre-eclampsia7. normal pregnancy, perhaps accounting, at least
Management of pre-eclampsia/HELLP syn- in part, for the predisposition to development of
drome is supportive and should be focused on thrombotic microangiopathy in this setting36.
stabilizing the patient medically prior to early TTP and HUS may be difficult to discern
delivery of the fetus. Platelet transfusions may from one another, as well as from other
be needed if bleeding occurs or if thrombo- pregnancy-associated microangiopathies such
cytopenia is severe and Cesarean delivery is as pre-eclampsia or the HELLP syndrome. The
planned, though the survival time of transfused extent of microangiopathic hemolysis is gener-
platelets in patients with pre-eclampsia is dimin- ally more severe in TTP or HUS than in
ished6. If required, the consumptive coagulo- pre-eclampsia or HELLP, and the former disor-
pathy resulting from pre-eclampsia should be ders are not associated with hypertension. The
treated with fresh frozen plasma (FFP). Con- time of onset of these disorders is also helpful
sumptive coagulopathy severe enough to result in differentiating between them. TTP usually
in depletion of fibrinogen is uncommon in these presents in the second trimester, HUS in the
disorders, but, if severe hypofibrinogenemia is postpartum period and pre-eclampsia and
present, plasma fibrinogen levels can be raised the HELLP syndrome almost exclusively in the
with cryoprecipitate. In most cases, the clinical third trimester7,34,37. Plasma antithrombin lev-
manifestations of pre-eclampsia resolve within els are normal in TTP and HUS and reduced in
several days after delivery, although the platelet pre-eclampsia and HELLP34. Another feature
count may decline for additional 24–48 h34. If distinguishing these disorders is their response
severe thrombocytopenia, hemolysis or organ to delivery. Whereas pre-eclampsia and the
dysfunction persists after delivery, plasma HELLP syndrome usually improve following
exchange may be considered35, but the delivery, the courses of TTP and HUS do not.
diagnosis should also be reviewed. Hence, pregnancy termination should not be
considered therapeutic in patients with TTP or
HUS38. However, TTP responds equally well to
Thrombotic thrombocytopenic purpura and
plasma exchange in pregnant and non-pregnant
hemolytic uremic syndrome
patients with > 75% of patients achieving remis-
Thrombotic thrombocytopenia (TTP) and sion6. Plasma exchange should be instituted as
hemolytic uremic syndrome (HUS) share the soon as possible after the diagnosis of TTP.
central features of microangiopathic hemolytic Daily plasma exchange should continue until at
anemia and thrombocytopenia. Though neither least 48 h after complete remission is obtained.
215
237
30 August 2006 14:21:08
Repeated plasma exchange cycles are usually clinicopathologic syndrome, characterized by

maintained until delivery. Management of HUS activation of the coagulation system, and result-
is supportive and includes renal dialysis and red ing in widespread intravascular deposition of
cell transfusion. Plasma exchange has no proven fibrin-rich thrombi. Consumption of clotting
benefit in the treatment of HUS. factors usually leads to a bleeding diathesis,
The placental ischemia and increased inci- although a small percentage of affected individ-
dence of premature delivery that complicate uals may go on to develop widespread thrombo-
pregnancies in patients with TTP and HUS sis with peripheral organ ischemia. Some degree
may lead to poor fetal outcomes, but these are of consumptive coagulopathy accompanies
markedly improved by good management of most forms of obstetric hemorrhage; however,
these conditions. the greater risk of coagulopathy usually arises
from consumption of clotting factors and plate-
lets as a result of massive obstetric hemorrhage.
Acute fatty liver of pregnancy
The combination of massive hemorrhage and
Acute fatty liver of pregnancy affects one of coagulation failure is recognized as one of the
every 5000–10 000 pregnancies and is most most serious complications in pregnancy.
common in primagravidas during the third Obstetric consumptive coagulopathy is usu-
trimester39. The cause of the condition is ally acute in onset (except as an uncommon late
unknown in the majority of instances, but some complication of retained dead fetus) and can be
patients may have a long-chain 3-hydroxy-acyl caused by a variety of disease processes. It
CoA dehydrogenase (LCHAD) deficiency40. is triggered by several mechanisms including
Patients present with overt signs of hepatic release of TF into the circulation, endothelial
damage and may have hemorrhagic manifesta- damage to small vessels and production of
tions, perhaps the result of decreased synthesis procoagulant phospholipids in response to
of clotting factors and consumptive coagulo- intravascular hemolysis42 (Table 5). Blood loss
pathy. Evidence for consumptive coagulopathy
is provided by thrombocytopenia, prolonged
APTT and PT and by decrease in fibrinogen Table 5 Mechanism of consumptive coagulopathy
and antithrombin levels. in pregnancy
AFLP is most aptly viewed as part of the
pregnancy-associated microangiopathies; up to A. Injury to vascular endothelium
50% of patients with AFLP may also meet Pre-eclampsia
Hypovolemic shock
criteria for pre-eclampsia. The extent of micro-
Septicemic shock
angiopathic hemolysis and thrombocytopenia is
generally mild compared to that observed in B. Release of tissue factor (TF)
HELLP, TTP, or HUS41. Placental abruption
Delivery is the most important aspect of Amniotic fluid embolism
management, as it starts the reversal of the Retained dead fetus
pathological process. Coagulation defects are Placenta accreta
Acute fatty liver
managed supportively with fresh frozen plasma,
cryoprecipitate and platelet concentrates. In C. Production of procoagulant
these patients, normalization of hemostatic Fetomaternal hemorrhage
abnormalities may not occur for up to 10 days Phospholipids
after delivery. Fetal mortality in this disorder Incompatible blood transfusion
approaches 15%, though maternal mortality Septicemia
occurs in less than 5% of cases39. Intravascular hemolysis
From Anthony J. Major obstetric hemorrhage and

CONSUMPTIVE COAGULOPATHY disseminated intravascular coagulation. In James
DK, Steer PJ, Weiner CP, Gonik B, eds. High-Risk
Consumptive coagulopathy (disseminated Pregnancy: Management Options, 3rd edn. Elsevier,
intravascular coagulation) is an acquired 2006:1606–23
216
238
30 August 2006 14:21:09
itself with transfusion and volume replacement of severe pulmonary hypertension following
may also trigger consumptive coagulopathy. embolization of the pulmonary vessels by fetal
With obstetric complications associated with squames. If the mother survives this acute
coagulation failure, there may be interaction of event, there may be an anaphylactoid reaction
several mechanisms. to the presence of the fetal tissues in the mater-
These triggers lead to the generation of nal circulation associated with cardiovascular
thrombin, cause defects in inhibitors of coagula- collapse, pulmonary edema and the develop-
tion and suppress fibrinolysis. Thrombin pro- ment of consumptive coagulopathy. Sepsis
motes platelet activation and aggregates form, causes consumptive coagulopathy via the
which occlude the microvasculature and result release of proinflammatory cytokines such as
in thrombocytopenia. Thrombin becomes tumor necrosis factor α (TNF-α), interleukin 1
bound to antithrombin (AT) and thrombo- (IL-1) and IL-6, which may trigger TF expres-
modulin, and these proteins are soon con- sion by monocytes and endothelial cells43.
sumed. Following binding to thrombomodulin, Severe pre-eclampsia with intense vasospasm
thrombin activates the anticoagulant protein C, and resulting ischemia causes endothelial injury
which also becomes depleted, predisposing and expression of TF.
to microvascular thrombosis. In consumptive Acute consumptive coagulopathy in preg-
coagulopathy secondary to sepsis, increased lev- nancy presents almost invariably with bleeding –
els of C4b-binding protein result in the binding either as a genital tract bleeding from the
of more free protein S, and therefore render it placental site or bleeding from the wound
unavailable to be a cofactor of the anticoagulant after Cesarean section. There may be excessive
protein C. PAI-1 is increased out of proportion bleeding from venepuncture sites.
to the level of tissue plasminogen activator Laboratory investigations are essential to
(tPA), resulting in depressed fibrinolysis. Fibrin establish the diagnosis of consumptive coagulo-
is formed, but its removal is impaired, leading pathy. The characteristic changes are a low or
to thrombosis of small and middle-size vessels. falling platelet count and a prolongation of the
The passage of erythrocytes through partially APTT and PT. Fibrinogen level falls with the
occluded vessels leads to red cell fragmentation progression of the coagulopathy; the normal
and microangiopathic hemolytic anemia. range in late pregnancy is 4–6 g/l which is sig-
Placental abruption is the most common nificantly higher than the non-pregnant range,
cause of obstetric consumptive coagulopathy 2–4 g/l; coagulation fails at levels < 1 g/l. FDPs
(60% of cases; 5% of all abruptions), but the are increased, reflecting the excessive deposi-
syndrome is uncommon unless the abruption tion of fibrin and enhanced fibrinolysis. The
is severe enough to cause fetal death. Initially, D-dimer is the most commonly used parameter
increased intrauterine pressure forces TF-rich to assess FDP levels, as it is specific for fibrin
decidual fragments into the maternal circula- breakdown. Normal D-dimer levels are under
tion. However, in severe abruption, hypo- 200 ng/ml, but often exceed 2000 ng/ml in
volemic shock, large volume transfusion and cases of consumptive coagulopathy. The blood
high levels of fibrin degradation products film may show evidence of microangiopathic
(FDPs) that act as anticoagulants themselves hemolysis with fragmentation of red cells.
exacerbate the situation. Retained dead fetus The basic principles in treatment of con-
may cause chronic consumptive coagulopathy sumptive coagulopathy are removal of the
by release of TF from the dead fetus into the precipitating cause if possible, correction of
maternal circulation, but generally only if the aggravating factors, and replacement of missing
fetus is at least 20 weeks’ size and the period coagulation factors and platelets. Correction of
of death is more than 4 weeks. Amniotic fluid aggravating factors such as shock and hypoxia is
embolism occurs during labor, Cesarean section important. This includes red cell transfusion if
or within a short time of delivery. Amniotic fluid necessary and oxygen administration. Intra-
is rich in TF and may enter uterine veins when venous antibiotics should be given if sepsis is
there has been a tear in the uterine wall. The suspected. Replacement of clotting factors is
condition may lead to maternal death as a result most effectively done with fresh frozen plasma.
217
239
30 August 2006 14:21:09
If there is severe hypofibrinogenemia, cryo- Recombinant activated protein C (raPC) has

precipitate may be required. Platelets should been successfully used in sepsis-related obstetric
be maintained > 50 × 109/l in the presence of consumptive coagulopathy at a dose of 24 µg/
active bleeding by the administration of blood kg/h in a 96-h infusion47,48. Caution is needed
group-compatible platelets. Any etiological in patients with severe thrombocytopenia
condition should be promptly treated; it often (< 30 × 109/l) because of the increased inci-
requires delivery of the fetus. Heparin use often dence of intracerebral hemorrhage associated
leads to excessive bleeding and therefore does with its use; monitoring of the platelet count
not usually have a role in obstetric consumptive and transfusion of platelets as necessary are
coagulopathy except in the cases of a retained important considerations. In addition to acting
dead fetus. Similarly, antifibrinolytic drugs as an anticoagulant, raPC has direct anti-
(tranexamic acid, aprotinin) are not helpful and inflammatory and anti-apoptotic properties49.
are usually contraindicated because they inhibit This may explain in part why the other endo-
the removal of deposited fibrin by fibrinolysis. genous anticoagulants (antithrombin and tissue
The usual regimen, when there is coagulation factor pathway inhibitor) used in severe sepsis
failure in obstetric practice, includes adminis- have not shown such good efficacy.
tration of FFP, platelets and cryoprecipitate. The treatment of such underlying conditions
FFP contains fibrinogen and all coagulation such as abruptio placentae, uterine rupture
factors. Each unit is approximately 250 ml and and fetal death require immediate obstetric
the usual requirement is 4–6 units. Platelet attention. Usually, there has been extensive
concentrates are used to increment platelet hemorrhage and red cell transfusion is needed
count. A unit of platelets is approximately 60 ml in addition to correction of the coagulation
in volume; it should raise the platelet count by failure.
5000 × 109/l and the usual dose is five packs.
Cryoprecipitate is enriched in fibrinogen, factor
FACTOR VIII INHIBITORS
VIII and vWF and is particularly useful for
the treatment of hypofibrinogenemia. Ten bags Acquired hemophilia is due to the development
(each 30 ml) of cryoprecipitate should increase of an autoantibody to factor VIII (FVIII). The
the fibrinogen level by 1 g/l. One 250 ml unit of estimated incidence is approximately 1 per
FFP contains a similar amount of fibrinogen 1 000 000 per annum. Most cases occur in
(500 mg) as one 30 ml bag of cryoprecipitate healthy individuals without discernible risk
(435 mg). factors, but the condition is associated with
The D-dimer, platelet count and fibrinogen autoimmune conditions such as rheumatoid
level are clinically useful tests in monitoring arthritis and SLE, inflammatory bowel disease,
replacement therapy if the patient is bleeding. multiple sclerosis and malignancies. In up to
The aim should be to achieve a platelet count 11% of cases, the associated factor is a recent or
> 50 × 109/l, a fibrinogen level > 1.0 g/l and ongoing pregnancy50.
significant shortening of the APTT and PT to Acquired hemophilia may occur in relation to
approach their normal values. any pregnancy, but the risk appears to be great-
Although recombinant activated factor VII est after the first delivery. Onset is usually at
(rFVIIa) is not licensed for use in pregnancy, term or within 3 months postpartum, but may
it has been used in obstetric patients with con- only become evident 12 months post-delivery51.
sumptive coagulopathy and severe bleeding not Clinical manifestations do not necessarily corre-
responsive to other treatment options44,45 (see late with inhibitor levels and can range from
Chapter 26). Consumptive coagulopathy is not spontaneous bruising to life-threatening hemor-
a contraindication to the use of rFVIIa if mas- rhage. FVIII inhibitors may cross the placenta
sive bleeding is occurring. However, caution and persist in the neonate for up to 3 months,
should be used in patients with major consump- but neonatal complications are rare51. Sponta-
tive coagulopathy because there are occasional neous resolution occurs in almost 100% of
reports of thrombosis and consumptive women first diagnosed in the postpartum period
coagulopathy after the use of rFVIIa46. after 30 months50.
218
240
30 August 2006 14:21:09
Basic coagulation studies in acquired hemo- (2) Prevention and treatment of systemic
philia demonstrate a prolonged APPT with a embolism in patients with mechanical heart
normal PT and thrombin time (TT). If plasma valve prostheses;
from the patient is mixed with normal plasma,
(3) Prevention of pregnancy complications in
the APPT remains prolonged due to the inhibi-
women with antiphospholipid syndrome
tor antibody neutralizing the FVIII in the
(APS) or other thrombophilia and prior
normal plasma. FVIII inhibitors must be differ-
pregnancy complications.
entiated from a lupus inhibitor by specific tests
because the clinical implications are profoundly The anticoagulants currently available for
different. Quantification of FVIII inhibitor is by the prevention and treatment of VTE and
the Bethesda assay, and checking this level may arterial thromboembolism include heparin
help in determining the choice of therapy and and heparin-like compounds (unfractionated
monitoring the progress of the patient. heparin (UFH), low-molecular weight heparin
Treatment is aimed at control of bleeding (LMWH), and heparinoids) and coumarin
and accelerating the elimination of inhibitors. derivatives, e.g. warfarin. The ‘direct’ thrombin
Hematological measures to minimize blood loss inhibitors, such as hirudin, cross the placenta
aim to compensate for the loss of FVIII. Choice and have therefore not yet been evaluated
of product to attempt to normalize hemostasis during pregnancy53.
depends on various considerations, including Heparins are the anticoagulant of choice
the severity of bleeding, availability of clotting during pregnancy for situations in which their
factor concentrates, inhibitor level and cross- efficacy is established. Neither UFH, LMWH
reactivity of inhibitor to porcine FVIII. Human nor heparinoids cross the placenta54. Heparins
FVIII may be effective if the titer of inhibitor is are not associated with any known teratogenic
low, i.e. less than 10 Bethesda units. At higher risk, and the fetus is not anticoagulated as a
levels, use of porcine FVIII which may not result of maternal heparin use. LMWHs have
cross-react with the inhibitor, and recombinant potential advantages over UFH during preg-
FVIIa or prothrombin complex concentrate nancy because they have a longer plasma
(PCC) becomes necessary52. half-life and a more predictable dose-response
Inhibiting the production of the inhibitor is than UFH, with the potential for once-daily
the second management aim. Prednisolone at administration. In addition, LMWHs are asso-
dose of 1 mg/kg is associated with a loss of ciated with a lower risk of HIT and osteoporosis
inhibitor in 50% of patients with acquired than UFH.
hemophilia52. Other immunosuppressives Coumarin derivatives such as warfarin cross
should be considered if there is no response the placenta and have the potential to cause
to steroids. Addition of cyclophosphamide teratogenicity as well as anticoagulate the fetus
(2.0–3.0 mg/kg) should be considered at 3 predisposing to bleeding in utero. It is probable
weeks if there is no decline in the inhibitor titer, that oral anticoagulants are safe during the first
or earlier if there is continued bleeding. Other 6 weeks of gestation, but there is an approxi-
methods to reduce inhibitor levels include mately 5% risk of developmental abnormalities
azathioprine, plasma exchange or infusion of of fetal cartilage and bone if they are taken
IVIg. between 6 and 12 weeks’ gestation55. The risk of
warfarin embryopathy is dose-dependent, with
an increased risk when the daily warfarin dose
ANTICOAGULANT THERAPY DURING exceeds 5 mg56. Fetal intracranial bleeds in utero
PREGNANCY AND THE PERIPARTUM are a well-established complication after expo-
PERIOD sure to these drugs during any trimester. In
general, coumarins should not be used for the
Anticoagulant therapy is indicated during
prevention or treatment of VTE in pregnancy,
pregnancy in the following cases:
but they remain the anticoagulants of choice
(1) Prevention and treatment of venous for the management of pregnant women with
thromboembolism (VTE); mechanical heart valve prostheses. Because of
219
241
30 August 2006 14:21:09
the hemorrhagic risk to both mother and fetus, data regarding the efficacy and safety of
warfarin should be avoided beyond 36 weeks antithrombotic therapy during pregnancy.
gestation. However, it appears reasonable to adopt one of
LMWHs are currently widely used for the the following three approaches:
prevention and treatment of gestational VTE.
(1) Oral anticoagulants throughout pregnancy;
In our institution, women on prophylactic doses
of LMWH are advised to have the dose of the (2) Replacing oral anticoagulants with UFH
LMWH tailed off at the end of pregnancy and from weeks 6 to 12;
omit their dose if labor is suspected. Women on
(3) UFH throughout pregnancy.
a therapeutic dose of LMWH are admitted in
advance of planned induction to be converted to In the first two regimens, heparin is usually
the therapeutic dose of intravenous UFH. They substituted for the oral anticoagulant close to
should omit LMWH on the day of admission term. The use of LMWH for anticoagulation
and should be started on UFH, aiming for an in patients with artificial heart valves is still
APTT ratio of 1.5–2.0. UFH should be reduced debatable.
to 500 IU/h when contractions start, aiming for As Walker57 has so succinctly stated, deci-
an APTT ratio < 1.5 and should be stopped at sions about the most appropriate anticoagulant
the second stage of labor or earlier if it appears regimen during pregnancy for women with
that a Cesarean section may be required. In the mechanical heart valve prostheses must be
latter case, protamine sulfate may be needed made on an individual patient basis after careful
for reversal of UFH if the APTT ratio remains counseling, and should be based as far as possi-
> 1.5. Postpartum, the heparin infusion can be ble on the relative risks of the various thrombo-
restarted 4 h post-delivery at 500 IU/h, provid- prophylaxis regimens and on whether the
ing there is no bleeding. Patients are restarted patient is perceived to be at higher or lower
on a therapeutic dose of LMWH 2–3 days after thromboembolic risk.
delivery. Warfarin can be started 4–5 days post- Women with the older type of mechanical
partum, and LMWH should be continued until prostheses (e.g. Starr-Edwards or Bjork-Shiley),
an international normalized ratio (INR) of 2.0 women with a prosthesis in the mitral position,
or greater is reached on two consecutive days. women with multiple prosthetic valves and
Breastfeeding is safe on UFH, LMWH and women with atrial fibrillation may be regarded
warfarin. as being at high thromboembolic risk. Women
Epidural anesthesia is generally safe in with newer and less thrombogenic valves (e.g.
women following discontinuation of UFH, pro- St Jude’s or Duromedics), particularly if they
viding their coagulation screen is normal and are in the aortic position and providing they are
their platelet count is > 80 × 109/l. It remains in normal sinus rhythm, may be regarded as
unclear what period of time should elapse being at lower thromboembolic risk.
between the last dose of LMWH and insertion With the information currently available, it
or removal of an epidural or spinal catheter, or would be prudent to advise women in the
how long the time interval should be until the high-thromboembolic-risk category to use an
next dose. In practice, it is reasonable to allow at oral anticoagulant with an INR target of
least 12 h to elapse after a prophylactic dose of 3.5 throughout pregnancy, although some may
LMWH before inserting an epidural or spinal choose to substitute adjusted doses of heparin
catheter, but a delay up to 24 h may be neces- between 6 and 12 weeks’ gestation. Warfarin
sary in patients on therapeutic doses of LMWH. should be avoided close to term and UFH or
At least 2 h should elapse after insertion of the LMWH substituted. However, if labor com-
catheter before LMWH is given again. If there mences in a woman on warfarin, intravenous
have been difficulties with the procedure, then vitamin K or fresh frozen plasma can be used to
it is prudent to delay prior to giving further reverse its effect.
prophylaxis. On the basis of one report that the risk of
Pregnant women with prosthetic heart valves fetal complications with warfarin appears to be
pose a problem because of the lack of reliable dose-related56, women with mechanical heart
220
242
30 August 2006 14:21:10
valves in the lower thromboembolic risk cate- minor trauma; menorrhagia is a common
gory may feel reassured about the relatively low presentation. Laboratory findings include
risk to their fetus if they use warfarin throughout thrombocytopenia, large platelets, prolonged
pregnancy, or with substitution of UFH or bleeding time and poor platelet aggregation in
LMWH from weeks 6 to 12 if their daily warfa- vitro to ristocetin.
rin requirement does not exceed 5 mg. Women Eleven cases of Bernard–Soulier syndrome in
in this category requiring higher daily doses pregnant women have been described to date59.
of warfarin may wish to minimize the risk of Most have been diagnosed prior to pregnancy,
fetal complication, and be prepared to rely on and postpartum hemorrhage has been more
adjusted doses of UFH and LMWH, but they common than antepartum bleeding60. Manage-
must be made aware that there is less good ment of bleeding in Bernard–Soulier syndrome
evidence to support the use of these latter in pregnancy is debatable; single-donor platelet
regimens. In general, women with bioprosthetic transfusions (preferably HLA-matched),
valves do not require anticoagulation, but desmopressin (DDAVP) and antifibrinolytic
anticoagulation may be necessary for other agents have been successfully used60.
indications. Glanzmann’s thrombasthenia is due to a
Clear recommendations for heparin use dur- spectrum of mutations in platelet membrane
ing labor and delivery in women with artificial GP IIb/IIIa, resulting in failure to bind
heart valves are not available. Intravenous UFH fibrinogen. It is characterized by excessive
at therapeutic doses may be administered until menstrual blood loss, bleeding from mucous
6 h before delivery. If the UFH is to be reversed, membranes, and major hemorrhage following
it is usually sufficient to stop the infusion (as the trauma or surgery. The platelet count is normal,
half-life of the UFH is approximately 1 h). If but clot retraction is greatly impaired and agents
more rapid reversal is necessary, protamine sul- such as adenosine diphosphate (ADP), epi-
fate is used. One mg of protamine sulfate neu- nephrine and collagen fail to induce platelet
tralizes 100 IU of heparin if the latter has been aggregation. Patients with this condition are at
given within the previous 30 min. Protamine increased risk of primary postpartum hemor-
sulfate should be given slowly at 5 mg/min, with rhage. Single-donor platelets (again, HLA-
a maximum single dose of 50 mg. Protamine matched if possible) and recombinant activated
sulfate is much less effective in reversal of FVII have been used to control bleeding during
LMWH. the peripartum and the postpartum period61.
Warfarin is initiated in the postpartum period The May-Hegglin anomaly is a rare auto-
in patients with mechanical valves. Antico- somal dominant condition with thrombo-
agulation with intravenous UFH while awaiting cytopenia and giant platelets. Platelet count
therapeutic levels of warfarin is probably not varies between 40 and 80 × 109/l, but platelet
warranted. The risk of bleeding, particularly function appears normal. Excess hemorrhage
after Cesarean section, exceeds the risk of is uncommon, but patients may need a
thrombotic complications58. Subcutaneous platelet transfusion to achieve hemostasis at
UFH in prophylactic doses (5000–7500 units delivery62.
twice daily) may be given.
von Willebrand disease
CONGENITAL DISORDERS OF von Willebrand disease (vWD) is the most com-
HEMOSTASIS mon of the inherited bleeding disorders, found
in approximately 1% of the general population
Congenital platelet disorders
without ethnic variations. It is caused by a
Bernard–Soulier syndrome is a rare autosomal reduced plasma concentration of structurally
recessive platelet disorder due to a variety of normal von Willebrand factor (vWF) or the
mutations in membrane glycoproteins Ib, IX presence of a structurally abnormal molecule
and V. Patients usually present early in life with with reduced activity. vWF is the carrier protein
spontaneous bruising, epistaxis or bleeding after in plasma for FVIII, and it also acts as a bridge
221
243
30 August 2006 14:21:10
between platelets and subendothelial collagen easy bruising or bleeding from mucosal sur-
fibers. faces. The most frequent problem found in the
vWF is synthesized in endothelial cells as a non-pregnant female is menorrhagia, which
polypeptide of 2813 amino acids, which under- may be quite severe. Patients with mild abnor-
goes initial dimerization and then multimerizat- malities may be asymptomatic, with the diag-
ion up to a multimer with a molecular weight nosis made only after significant hemostatic
of 20 000 kDa. High-molecular weight (HMW) challenges such as operations and trauma.
multimers are functionally more effective in Laboratory tests in patients with vWD
promoting platelet adhesion and aggregation. show prolonged bleeding time and may show
The vWF protein is released into the plasma, a prolonged APTT. More definitive diagnostic
and is also stored in Weibel–Palade bodies in tests depend on the finding of reduced vWF
the endothelial cells. vWF is also synthesized in activity measured by ristocetin cofactor activity
megakaryocytes, stored in the platelet α-gran-
ules and, on activation, secreted by the platelet
release reaction. This allows accumulation of Table 6 Classification of von Willebrand disease
vWF at the site of vascular injury where it can (VWD)
promote further platelet adhesion and thus
Type 1 Partial quantitative deficiency of
hemostasis. The mature vWF protein possesses
apparently normal vWF
a number of specific binding sites, which repre- Type 2 Qualitative deficiency of vWF
sent its different activities (Figure 1). Circulat- Type 2A Qualitative variants with decreased
ing HMW multimers are cleaved by a protease, HMW multimers
known as ADAMTS 13, which is lacking in Type 2B Qualitative variants with increased
patients with the rare congenital thrombotic affinity for platelet GP Ib
thrombocytopenic purpura. Type 2M Qualitative variants with normal HMW
vWD is subclassified into six categories multimers appearance
(Table 6), which correspond to distinct Type 2N Qualitative variants with markedly
pathophysiological mechanisms and are impor- decreased affinity for factor VIII
Type 3 Virtually complete deficiency of vWF
tant in determining therapy. Of all the catego-
ries, about approximately 70–80% of patients VWF, von Willebrand factor; HMW multimers,
have type 1 disease. high-molecular weight multimers
The condition commonly presents as a mild Adapted from Sadler JE. Thromb Haemost
to moderate bleeding disorder, typically with 1994;71:520–5
Figure 1 The von Willebrand factor. The protein consists of a series of domains with different binding
sites for factor VIII, heparin, collagen and platelet glycoprotein (Gp) Ib and IIb/IIIa. The sites of gene
mutations giving rise to different subtypes of VWD are marked. From Green D, Ludlam CA. VWD in
bleeding disorders. Health Press 2004, pp. 63–69
222
244
30 August 2006 14:21:13
(vWF:RCo) and collagen-binding assay The loading dose is 40–60 IU/kg, and this
(vWF:CB), accompanied by variable reductions can be followed by repeat doses every 12–24 h
in vWF antigen (vWF:Ag) and FVIII. Several to maintain vWF activity (vWF:RCoF) > 50%.
further tests that aid in classification include All currently available concentrates are derived
analysis of ristocetin-induced platelet aggrega- from plasma. As at least one viral inactivation
tion (RIPA), vWF multimer and assay of FVIII step is included in their manufacture, they are
binding to vWF63. The diagnosis may not be unlikely to transmit hepatitis or HIV, but there
straightforward, as one or more of the activities is still a risk of parvovirus infection.
of FVIII and vWF may be borderline and even
normal. It is often necessary to repeat the
von Willebrand disease and pregnancy
estimations on at least three occasions. Stress,
physical exercise, recent surgery and pregnancy von Willebrand disease is the most common
all increase plasma vWF levels and FVIII levels, congenital hemostatic disorder in pregnancy. In
and diagnosis may be difficult in these circum- a normal pregnancy, both FVIII and vWF levels
stances64. When investigating patients with bor- progressively increase (Figure 2)67. vWF starts
derline results, it should be taken into account to rise as early as the 6th week and by the third
that FVIII and vWF levels are 15–20% lower in trimester may have increased three- to fourfold.
individuals with blood group O compared to FVIII and vWF levels also increase in most
individuals with blood group A64. women with vWD, which may explain the fre-
The aim of therapy for vWD is to correct quent improvement in minor bleeding manifes-
the impaired primary hemostasis and impaired tations during pregnancy. The hemostatic
coagulation. Treatment choice depends on the response to pregnancy depends on both the type
severity and the type of disease, and on the and severity of disease. Most women with type 1
clinical setting. Treatment options usually vWD have an increase in FVIII and vWF levels
include DDAVP and vWF-containing blood into the normal non-pregnant range, which may
products65. mask the diagnosis during pregnancy. However,
DDAVP, a synthetic vasopressin analogue, levels may remain low in severe cases. FVIII and
releases vWF from endothelial stores; there is vWF antigen levels often increase in pregnant
also an increase in the plasma FVIII level. It is women with type 2 vWD with minimal or
usually given by slow intravenous infusion of no increase in vWF activity levels. In type 2B
0.3 µg/kg over 20 min, which can be repeated vWD, the increase in the abnormal vWF can
every 4–6 h on two or three occasions. The drug cause progressive and severe thrombocytopenia,
can also be given subcutaneously or as a nasal but intervention is not usually required. Most
spray. Side-effects include hypotension, facial women with type 3 vWD have no improvement
flushing, fluid retention for up to 24 h and con- in FVIII or vWF levels during pregnancy68.
sequent hyponatremia. DDAVP can safely be After delivery, FVIII and vWF in normal
used during pregnancy66 and after delivery. It is women fall slowly to baseline levels over a
effective in securing in many situations in type 1 period of 4–6 weeks. However, the postpartum
vWD with a 3–5-fold increase in the plasma decline of these factors may be rapid and signifi-
vWF and FVIII levels. It is of no therapeutic cant in women with vWD68. As the individual
benefit in type 3 vWD because of the very low hemostatic response to pregnancy is variable,
basal levels of vWF and FVIII. The response in vWF and FVIII levels should be monitored
types 2 is less predictable. DDAVP is contrain- during pregnancy and 3–4 weeks after delivery.
dicated in patients with type 2B because it may Antepartum hemorrhage is uncommon in
exacerbate the coexisting thrombocytopenia. women with vWD, but may occur after sponta-
Patients should have a test of DDAVP (if neous miscarriage or elective termination,
possible when not pregnant) to see if it is occasionally as the initial presentation of vWD.
effective in their individual case. Women with vWD are at substantial risk for
Plasma-derived vWF concentrates are neces- secondary postpartum hemorrhage, especially
sary in patients who do not respond adequately 3–5 days after delivery. vWD may also exacer-
to DDAVP or in whom it is contraindicated. bate bleeding due to other obstetric causes, such
223
245
30 August 2006 14:21:14
450
400
350
Factor level iu/dl
300
250 FVIII
200 vWF:Ag
150
100
50
0
13 18 23 28 33 38 Post Basal
Weeks gestation
Figure 2 Levels of factor VIII and vWF in normal pregnancy. From Giangrande PL. Management of
pregnancy in carriers of haemophilia. Haemophilia 1998;4:779–84
as uterine atony or a trauma to the birth canal. FVIII and vWF:RCo should be maintained in
Other pregnancy-associated reasons for bleed- the normal range for at least 3–7 days after
ing in women with vWD include extensive Cesarean section. It is difficult and unnecessary
bruising and hematomas at intramuscular to diagnose vWD in the neonate, except when
injection, episiotomy and surgical wound sites. type 3 vWD is suspected. Generally, diagnosis
For patients whose vWD profile has normal- can be postponed until later in childhood.
ized in pregnancy, no specific hemostatic sup-
port is required. Regional analgesia may
HEMOPHILIAS
proceed in these patients after discussion with
an obstetric anesthetist. Although neonatal Hemophilias A and B are the most common
bleeding is rare, ventouse delivery and high- severe congenital bleeding disorders associated
cavity forceps should be avoided. Careful and with reduced or absent coagulation FVIII and
prompt repair of episiotomy wounds or perineal FIX, respectively. The incidence of hemophilia
tears is advisable. A is around 1 in 10 000 live male births. Hemo-
For patients whose vWF activity (vWF:RCo) philia B is about five times less common than
has not normalized, decisions about regional hemophilia A. The genes for both conditions
analgesia should be individualized69. Hemo- are located on the X-chromosome; they are
static supportive therapy with DDAVP or vWF therefore sex-linked disorders that almost exclu-
concentrate should be given to cover delivery or sively affect males. Clinically, the hemophilias
Cesarean section if the FVIII level is less than have an identical presentation and can only be
50% or if vWF:RCo has not normalized66. distinguished by measuring plasma levels of the
Because of the high incidence of secondary specific clotting factors. The clinical severity is
postpartum hemorrhage in patients with vWD, directly related to plasma concentrations of
efforts should be made to ensure that placenta is FVIII/FIX. Individuals with levels of below 1%
complete upon expulsion or removal. of normal have severe hemophilia and the most
After delivery, all patients should be closely frequent bleeds. Females in families with a his-
observed for postpartum hemorrhage and tory of hemophilia may be obligate, potential or
uncorrected hemostatic defects treated. In sporadic carriers, depending on the details of
responsive patients, DDAVP is the treatment of the pedigree71. An obligate carrier is a woman
choice to prevent and treat mild to moderate whose father has hemophilia, or a woman who
postpartum bleeding70. FVIII and vWF:RCo has family history of hemophilia and who has
should be checked a few days postpartum given birth to a hemophiliac son, or a woman
because they may fall rapidly after delivery. who has more than one child with hemophilia.
224
246
30 August 2006 14:21:14
A potential carrier of hemophilia is a woman Coagulation studies should also be carried

who has a maternal relative with the disorder. A out to identify carriers with low FVIII/FIX lev-
woman with one affected child and no family els. Phenotypic data may be helpful in assessing
history may be a sporadic carrier71. Female car- the statistical risk of carrriership if molecular
riers of hemophilia may have reduced FVIII/IX diagnosis is not possible. However, normal lev-
levels because of random inactivation of the els of FVIII/FIX do not exclude carriership72.
X-chromosome (lyonization). If the FVIII/IX Women who have low levels of FVIII may have
level is less than 50%, abnormal bleeding may a useful hemostatic response to DDAVP. To
occur after trauma or surgery. establish whether this response is occurring, a
There are two main risks for a female carrier trial of intravenous DDAVP can be attempted,
of hemophilia in pregnancy. First, women with with measurement of the response in FVIII
a low FVIII/IX level may be at risk of bleeding levels over the next 24 h.
after delivery or during invasive procedures Once carriership has been established,
in the first trimester. Second, there is a 50% women should be offered prepregnancy coun-
chance of each son inheriting hemophilia and seling to provide them with the information nec-
50% of her daughters being carriers. essary to make informed reproductive choices.
As discussed earlier, the levels of FVIII and A new technique of preimplantation diagnosis
vWF rise during normal pregnancy (Figure 2). is potentially useful for carriers of hemophilia
The increase is particularly marked during the who, after counseling, do not wish to contem-
third trimester, when levels of FVIII may rise to plate bringing up a hemophilic child, but would
double that of the normal baseline value. Simi- not consider termination. Following in vitro
larly, the vast majority of carriers of hemophilia fertilization (IVF) treatment, it is possible to
A will have increased their FVIII production to remove a single embryonic cell at the 8–16-cell
within the normal range by late gestation; factor stage and carry out genetic diagnosis. Female or
replacement therapy is thus only rarely required unaffected male embryos can then be trans-
during pregnancy in carriers of hemophilia A. ferred into the uterus. In the UK, each such test
By contrast, the level of FIX does not increase requires a license from the Human Fertilization
significantly during pregnancy, and thus a and Embryology Authority.
woman with a low initial baseline FIX is more If prenatal diagnosis is requested, testing
likely to require replacement to control bleeding is usually carried out by chorionic villus sam-
complications during delivery. pling (CVS) at 11–12 weeks’ gestation; DNA
All women who are obligate or potential extracted from fetal cells is analyzed. The prin-
carriers of hemophilia should be offered genetic cipal advantage of this procedure is that it may
testing and counseling. In particular, they be applied during the first trimester, so that, if
should have their carrier status determined termination of the pregnancy is required, this
to allow for the optimal management of is easier to carry out. The main adverse event
their pregnancies. Genetic testing should be related to CVS is miscarriage, which is esti-
offered when the individual is able to under- mated at about 1–2%. Fetal cells are karyotyped
stand the issues concerned (usually at age of so that the fetal sex is established. If the fetus is
13–15 years) and after having given informed female, no further tests are done. If the fetus is
consent72. In many individuals in the UK with male, additional tests are conducted to establish
hemophilia A and B, the causative mutation whether the affected gene has been inherited.
has been identified. If the mutation within the Cells for karyotyping and as a source of DNA
family is known, it is straightforward to screen can also be obtained from amniotic fluid
the potential carrier. If, on the other hand, the (amniocentesis) after 15 weeks’ gestation; here,
mutation is not known, then linkage analysis the miscarriage rate is about 0.5–1%. Fetoscopy
using informative genetic polymorphisms may to allow for fetal blood sampling is rarely per-
be possible. If neither of these approaches formed; it can only be performed after about 16
is suitable, then direct mutation detection weeks’ gestation and has a substantial risk of
may be possible by sequencing the FVIII/FIX fetal death (1–6%). The use of prenatal diag-
gene. nosis is decreasing in developed countries. As
225
247
30 August 2006 14:21:14
hemophilia care improves, more couples are RARE COAGULATION DISORDERS

willing to contemplate bringing up a child with
Fibrinogen deficiency
hemophilia67. When prenatal diagnosis has not
been carried out but there is a risk that the child The hypo- and dysfibrinogenemias comprise
may have hemophilia, fetal sex should be diag- a collection of disorders that are usually
nosed by ultrasonography67. This information is dominantly inherited and associated with both
necessary for the obstetrician even if the parents bleeding and venous thrombotic manifestations.
do not wish to know the sex of the infant. Women are at risk of recurrent miscarriage, and
Factor VIII/IX levels in female carriers of both antenatal and postnatal hemorrhage. In
hemophilia should be monitored regularly in hypofibrinogenemia, both antigenic and func-
pregnancy. It is particularly important to mea- tional fibrinogen levels are reduced. The
sure coagulation factor levels toward the end of diagnosis of dysfibrinogenemia is made by
the third trimester (34–36 weeks) to plan man- demonstrating a prolonged TT with a normal
agement of delivery67. If maternal FVIII/FIX antigenic fibrinogen level.
levels remain low at 34–36 weeks in hemophilia Prophylaxis with fibrinogen concentrates
carriers, treatment is necessary for delivery67. A improves pregnancy outcome and prevents
FVIII/FIX plasma level of 40% is safe for vagi- antepartum and postpartum hemorrhage in
nal delivery, and a level of 50% or greater is safe women with hypo- and dysfibrinogenemia.
for Cesarean section. Epidural anesthesia may Cryoprecipitate is a good source of fibrinogen
be used if coagulation defects have been cor- but should not usually be used, as it is not virally
rected67. Recombinant FVIII/FIX or DDAVP inactivated. Its use may be considered in an
(for carriers of hemophilia A only) should emergency situation if no other alternatives are
be used. Plasma-derived factor concentrate available. The half-life of infused fibrinogen is
products, including those subjected to dual- 3–5 days, and treatment is unlikely to be needed
inactivation processes, have the potential to more often than on alternate days. Levels above
transmit non-lipid coated viruses, e.g. parvo- 1.5 g/l are required toward the end of pregnancy
virus, and should not be used. Infection of the and at the time of delivery75.
fetus with parvovirus may result in hydrops
fetalis and fetal death.
If the fetus is a known hemophiliac, is male
Factor VII deficiency
and of unknown hemophilia status, or is of
unknown sex, care should be taken to avoid Congenital FVII deficiency is the most common
traumatic vaginal delivery. Routine Cesarean of the rare inherited coagulation disorders with
delivery is unnecessary67, but should be carried an estimated prevalence of 1 in 500 000. It is
out if obstetric complications are anticipated. inherited in an autosomal recessive manner and
Most bleeding problems in carriers of hemo- its frequency is significantly increased in coun-
philia occur postpartum. Replacement therapy tries where there are consanguineous marriages.
should be given immediately after delivery to FVII levels are usually less than 10% in homo-
mothers with uncorrected hemostatic defect. zygotes and around 50% in heterozygotes.
Treatment options at this stage are the same Although there is a poor correlation between
as those during labor and delivery. Supportive FVII levels and bleeding risk, hemorrhages
therapy to maintain hemostasis should be occur in patients with factor VII levels below
continued for 3–4 days after vaginal delivery 10–15%76. Individuals with a moderate FVII
and for 5–10 days after Cesarean section73. deficiency often bleed from the mucous
In the infant, intramuscular injections should membranes, and epistaxis, bleeding gums and
be avoided until hemophilia has been excluded. menorrhagia are common. In severe FVII defi-
Cord blood should be obtained for FVIII/FIX ciency (FVII level < 2%), bleeding into the cen-
assays74. Routine administration of coagulation tral nervous system very early in life leads to a
factor concentrates to neonates with hemophilia high morbidity and mortality. Congenital FVII
is unnecessary if delivery has been atraumatic deficiency is usually suspected when an isolated
and there are no clinical signs of hemorrhage74. prolongation of the PT is found in a patient
226
248
30 August 2006 14:21:14
without liver disease, and a normal APTT and defined by FX levels of 6–10%; these individu-
fibrinogen level. als are often diagnosed incidentally but may
The FVII level may increase up to four-fold experience easy bruising or menorrhagia. The
during normal pregnancy76. However, it is diagnosis of FX deficiency is suspected follow-
unknown whether FVII levels increase to the ing the finding of a prolonged APTT and PT
same degree in pregnant women with congenital and is confirmed by measuring plasma FX levels.
FVII deficiency as they do in normal preg- Thirteen pregnancies in eight patients with
nancy77. FVII deficiency during pregnancy is a isolated FX deficiency have been reported in
risk factor for postpartum hemorrhage. Bleed- the literature81. The complications described
ing may occur from the placental implantation include spontaneous abortions, placental
site, episiotomies, lacerations to the birth canal, abruptions, premature births and postpartum
or surgical trauma occurring with Cesarean hemorrhage. FX levels increase during preg-
delivery78. nancy and antenatal replacement therapy is not
Recombinant activated FVII (rFVIIa) has usually needed. However, women with severe
been approved in the European Union for use in FX deficiency and a history of adverse outcome
congenital FVII deficiency79. In places where in pregnancy may benefit from aggressive
this product is not available, fresh frozen replacement therapy75. As the half-life of FX
plasma, prothrombin complex concentrates is 24–40 h, a single daily infusion is usually
(PCCs) or plasma-derived FVII concentrate adequate. FX levels of 10–20% are generally
may be used. Because the patient may poten- sufficient for hemostasis75 and are required at
tially need a Cesarean delivery and because peri- the time of delivery.
neal trauma cannot be anticipated, prophylaxis FX is present in intermediate-purity FIX con-
is usually recommended at the time of deliv- centrates (prothrombin complex concentrates,
ery78. Recombinant FVIIa has been given as an PCCs). FX levels should be monitored as cau-
initial bolus injection of 20–50 µg/kg, followed tion is required because of the prothrombotic
by further boluses of 10–35 µg/kg every 4–6 properties of these concentrates. Fresh frozen
hours to cover vaginal delivery or Cesarean plasma may be an alternative when prothrombin
section in patients with congenital FVII defi- complex concentrates are not available.
ciency78,80. It has also been used as an initial
bolus injection of 13 µg/kg with subsequent
Combined deficiencies of the vitamin
continuous infusion at 1.7–3.3 µg/kg/h for 4
K-dependent factors II, VII, IX and X
days76 (see Chapter 26).
Congenital combined deficiency of factors II,
VII, IX and X is an autosomal recessive bleed-
Factor X deficiency
ing disorder. It is caused by deficiency of
Congenital FX deficiency is an autosomal enzymes associated with vitamin K metabolism
recessive disorder. The prevalence of the severe (e.g. γ-glutamyl carboxylase) as a result of
(homozygous) form is 1 : 1 000 000 in the gen- homozygous genetic mutations. Muco-
eral population and is much higher in countries cutaneous and postoperative related bleeding
where consanguineous marriages are more have been reported. Severe cases may present
common. The prevalence of heterozygous FX with intracranial hemorrhage or umbilical cord
deficiency is about 1 : 500, but individuals bleeding in infancy. Some individuals have
are usually clinically asymptomatic. Severe FX associated skeletal abnormalities (probably
deficiency (FX level < 1%) is associated with a related to abnormalities in bone vitamin
significant risk of intracranial hemorrhage in the K-dependent proteins such as osteocalcin).
first weeks of life and umbilical stump bleeding. Severe bleeding is usually associated with
The most frequent symptom is epistaxis, which activities of the vitamin K-dependent factors of
is seen with all severities of deficiency. < 5%. Affected individuals show prolongation
Menorrhagia occurs in half of the women. of the APTT and PT associated with variable
Severe arthropathy may occur as a result of reductions in the specific activities of factors II,
recurrent joint bleeds. Mild deficiency is VII, IX and X.
227
249
30 August 2006 14:21:14
The clinical picture and response to vitamin challenge, treatment is not usually required
K is variable, some responding to low-dose oral during vaginal delivery. In women with partial
vitamin K but others are non-responsive even to deficiency and significant bleeding history or no
high-dose intravenous replacement. In those previous hemostatic challenges, tranexamic acid
individuals who are non-responsive to vitamin is often used for 3 days, with the first dose being
K, prothrombin complex concentrates are the administered during labor. Tranexamic acid is
product of choice. also used to manage prolonged mild intermit-
There is a single report of a pregnancy pro- tent secondary postpartum hemorrhage which
gressing to term in an individual with severe is a common presentation of FXI-deficient
congenital vitamin K-dependent clotting factor patients84. FXI concentrate is needed for
deficiency managed with oral vitamin K 15 mg severely deficient women to cover vaginal
daily throughout pregnancy. Bleeding from an delivery and also for Cesarean section. The aim
episiotomy wound in this case required fresh is to maintain the FXI level > 50% during labor
frozen plasma82. and for 3–4 days after vaginal delivery and 7
days after Cesarean section. FXI concentrate is
potentially thrombogenic; the single dose
Factor XI deficiency
should not exceed 30 IU/kg with the aim of
FXI deficiency is an autosomally inherited con- raising FXI level to no greater than 70%84.
dition, which is particularly common in Ashke- Concurrent use of tranexamic acid or other
nazi Jews in whom heterozygote frequency is antifibrinolytic drugs with FXI concentrate
8%. Overall, the prevalence of severe deficiency should be avoided. Fresh frozen plasma can be
is approximately 1 : 1 000 000 but partial defi- used, but, in patients with severe deficiency, it is
ciency is much more common. FXI deficiency is difficult to produce a sufficient rise (to more
unlike most of the other rare coagulation disor- than 30%) without the risk of fluid overload75.
ders in that heterozygotes may have a significant Recombinant FVIIa has been used successfully
bleeding tendency that is poorly predicted to manage adult patients with FXI deficiency
by the FXI level. Spontaneous bleeding is undergoing surgery, although it is not licensed
extremely rare, even in those with undetectable for this indication75.
FXI levels. Bleeding is provoked by injury or
surgery, particularly in areas of high fibrinolytic
Factor XIII deficiency
activity (e.g. genitourinary tract). Menorrhagia
is common, and women with FXI deficiency Congenital FXIII (fibrin stabilizing factor)
may be diagnosed as a consequence of this. FXI deficiency is an autosomal recessive disorder.
deficiency rarely results in bleeding during It is characterized by features of delayed
pregnancy, but women with severe or partial and impaired wound healing with bleeding
deficiency may suffer postpartum bleeding75. occurring 24–36 h after surgery or trauma.
The APTT is usually prolonged and diag- Umbilical bleeding in the first few weeks of life
nosis is confirmed by finding a low FXI level. is very suggestive of the disorder. Soft tissue
The deficiency is classified as severe if the FXI bleeds are more common than hemarthroses,
level is less than 15% and partial at 15–70%; the which usually only occur after trauma. Sponta-
lower limit of the normal range is 70%. There is neous intracranial bleeds are a characteristic
controversy about changes in FXI levels during feature. Spontaneous abortions occur in early
normal pregnancy, some studies demonstrating pregnancy because FXIII is required for
an increase and others a decrease83. Changes successful implantation. Women with FXIII
in FXI levels in women with FXI deficiency deficiency are also at increased risk of postnatal
have been inconsistent during pregnancy84. It is bleeding75. The severity of the bleeding state
therefore recommended that FXI levels should varies markedly between individuals with appar-
be checked at the initial visit, and during the ently similar FXIII plasma levels. The routine
third trimester in FXI-deficient women. tests (APTT and PT) are normal and the FXIII
In women with partial FXI deficiency and level has to be specifically requested of the
no bleeding history but previous hemostatic laboratory.
228
250
30 August 2006 14:21:15
FXIII has a half-life of 7–10 days and there- adults, children and in pregnancy. Br J Haematol
fore only needs to be given at 4–6-weekly inter- 2003;120:574–96
vals to maintain a level > 3% which is necessary 10. Kessler I, Lancet M, Borenstein R, et al. The
to prevent spontaneous intracranial bleeds. Up obstetrical management of patients with immu-
nologic thrombocytopenic purpura. Int J
to 50% of severely (FXIII level < 1%) affected
women may miscarry without appropriate FXIII
11. Burrows RF, Kelton JG. Thrombocytopenia
treatment75. All severely affected individuals during pregnancy. In Greer IA, Turpie AG,
should be started on monthly infusions of plasma-- Forbes CD, eds. Haemostasis and Thrombosis in
derived FXIII concentrate from the time of diag- Obstetrics and Gynaecology. London: Chapman &
nosis to prevent intracranial bleeds and these Hall, 1992
should be continued during pregnancy75. FXIII 12. Gill KK, Kelton JG. Management of idiopathic
levels fall throughout pregnancy and should be thrombocytopenic purpura in pregnancy. Sem
monitored, aiming to keep the trough level > 3%. Hematol 2000;37:275–83
FXIII deficiency may also cause life- 13. Letsky EA. In de Swiet, ed. Coagulation Defects in
threatening hemorrhage in the neonate with Medical Disorders in Obstetric Practice, 4th edn.
Oxford: Blackwell Science, 2002:61–96
levels < 3%. The disorder can be diagnosed
14. Letsky EA, Greaves M. Guidelines on the
from cord or peripheral blood samples. Treat-
investigation and management of thrombocyto-
ment of an acute bleeding episode is with FXIII penia in pregnancy and neonatal alloimmune
concentrate at a dose of 20 IU/kg75. thrombocytopenia. Maternal and Neonatal
Haemostasis Working Party of the Haemostasis
and Thrombosis Task Force of the British Soci-
References
ety for Haematology. Br J Haematol 1996;95:
1. Brenner B. Haemostatic changes in pregnancy. 21–6
Thromb Res 2004;114:409–14 15. Crowther MA, Burrows RF, Ginsberg J,
2. Bellart J, Gilabert R, Miralles RM, et al. Kelton JG. Thrombocytopenia in pregnancy:
Endothelial cell markers and fibrinopeptide A to diagnosis, pathogenesis and management. Blood
D-dimer ratio as a measure of coagulation and Rev 1996;10:8–16
fibrinolysis balance in normal pregnancy. 16. Gill KK, Kelton JG. Management of idiopathic
Gynecol Obstet Invest 1998;46:17–21 thrombocytopenic purpura in pregnancy. Sem
3. Øian P, Omsjø I, Maltau JM, Østerud B. Hematol 2000;37:275–83
Reduced thromboplastin activity in blood mono- 17. Bussel JB, Druzin ML, Cines DB, Samuels P.
cytes and reduced sensitivity to stimuli in vitro of Thrombocytopenia in pregnancy. Lancet 1991;
blood monocytes from pregnant women. Br J 337:251
Haematol 1985;59:133–7 18. Godelieve C, Christiaens ML, Nieuwenhuis
4. Holmes VA, Wallace JMW, Gilmore WS, et al. HK, Bussel JB. Comparison of platelet counts in
Tissue factor expression on monocyte sub- first and second newborns of mothers with
populations during normal pregnancy. Thromb immune thrombocytopenic purpura. Obstet
Haemost 2002;87:953–8 Gynecol 1997;90:546–52
5. Holmes VA, Wallace JM. Haemostasis in normal 19. Miyakis S, Lockshin MD, Atsumi T, et al. Inter-
pregnancy: a balancing act? Biochem Soc Trans national consensus statement on an update of the
2005;33:428–32. classification criteria for definite antiphospho-
6. McCrae KR. Thrombocytopenia in pregnancy: lipid syndrome (APS). J Thromb Haemost 2006;
differential diagnosis, pathogenesis and manage- 4:295–306
ment. Blood Rev 2003;17:7–14 20. Galli M, Finazzi G, Barbui T. Thrombocyto-
7. McCrae KR, Samuels P, Schreiber AD. penia in the antiphospholipid syndrome: patho-
Pregnancy-associated thrombocytopenia: pathophysiology, clinical relevance and treatment. Ann
genesis and management. Blood 1992;80: Med Intern 1996;147:24–7
2697–714 21. Harris EN. A reassessment of the antiphospho-
8. Shehata N, Burrows RF, Kelton JG. Gestational lipid syndrome. J Rheumatol 1990;17:733–5
thrombocytopenia. Clin Obstet Gynecol 1999;42: 22. Rai R, Cohen H, Dave M, Regan L. Randomised
327–34 controlled trial of aspirin and aspirin plus hepa-
9. Guidelines for the investigation and manage- rin in pregnant women with recurrent miscar-
ment of idiopathic thrombocytopenic purpura in riage associated with phospholipid antibodies (or
229
251
30 August 2006 14:21:15
antiphospholipid antibodies). BMJ 1997;314: 35. Martin JN, Files JC, Blake PG, et al. Plasma
253–7 exchange for preeclampsia: Postpartum use for
23. Royal College of Obstetricians and Gynaecolo- persistently severe preeclampsia-eclampsia with
gists: Guidelines. Thromboprophylaxis during HELLP syndrome. Am J Obstet Gynecol 1990;
pregnancy, labour and after vaginal delivery. 162:126–37
Guidelines No. 37. London, RCOG Press, 36. Mannucci PM, Canciani T, Forza I, et al.
2004 Changes in health and disease of the metallo-
24. de Swiet M. Antiphosholipid syndrome, systemic protease that cleaves von Willebrand Factor.
lupus erythematosus and other connective tissue Blood 2001;98:2730–5
diseases. In de Swiet, ed. Medical Disorders in 37. Lain KY, Roberts JM. Contemporary concepts
Obstetric Practice, 4th edn. Oxford: Blackwell of the pathogenesis and management of
Science, 2002:267–81 preeclampsia. JAMA 2002;287:3183–6
25. Mandelbrot L, Schlienger I, Bongain A, et al. 38. Esplin MS, Branch DW. Diagnosis and manage-
Thrombocytopenia in pregnant women infected ment of thrombotic microangiopathies during
with human immunodeficiency virus: maternal pregnancy. Clin Obstet Gynecol 1999;42:360–8
and neonatal outcome. Am J Obstet Gynecol 39. Bacq Y. Acute fatty liver of pregnancy. Sem
1994;171:252–7 Perinatol 1998;22:134–40
26. van Besien K, Hoffman R, Golichowski A. 40. Tyni T, Ekholm E, Pihko H. Pregnancy
Pregnancy associated with lupus anticoagulant complications are frequent in long-chain
and heparin induced thrombocytopenia: man- 3-hydroxyacyl-coenzyme A dehydrogenase defi-
agement with a low molecular weight heparinoid. ciency. Am J Obstet Gynecol 1998;178:603–8
Thromb Res 1991;62:23–9 41. Vigil-De Gracia P. Acute fatty liver and HELLP
27. Greinacher A, Eckhardt T, Mussmann J, syndrome: two distinct pregnancy disorders. Int
Mueller-Eckhardt C. Pregnancy complicated by J Gynaecol Obstet 2001;73:215–21
heparin associated thrombocytopenia: manage- 42. Anthony J. Major obstetric hemorrhage and
ment by a prospectively in vitro selected disseminated intravascular coagulation. In James
heparinoid (Org 10172). Thromb Res 1993;71: DK, Steer PJ, Weiner CP, Gonik B, eds. High
123–6 Risk Pregnancy: Management Options, 3rd edn.
28. Fausett MB, Vogtlander M, Lee RM, et al. Amsterdam: Elsevier, 2006:1606–23
Heparin-induced thrombocytopenia is rare in 43. Levi M. Current understanding of disseminated
pregnancy. Am J Obstet Gynecol 2001;185: intravascular coagulation. Br J Haematol 2004;
148–52 124:567–76
29. Huhle G, Geberth M, Hoffmann U, et al. 44. Moscardo F, Perez F, de la Rubia J, et al.
Management of heparin-associated thrombo- Successful treatment of severe intra-abdominal
cytopenia in pregnancy with subcutaneous bleeding associated with disseminated intra-
r-hirudin. Gynecol Obstet Invest 2000;49:67–9 vascular coagulation using recombinant acti-
30. Khong TY, De Wolf F, Robertson WB, Brosens vated factor VII. Br J Haematol 2001;114:174–6
I. Inadequate maternal vascular response to 45. Zupancic Salek S, Sokolic V, Viskovic T, et al.
placentation in pregnancies complicated by Successful use of recombinant factor VIIa for
preeclampsia and by small for gestational age massive bleeding after caesarean section due
infants. Am J Obstet Gynecol 1987;157:360–3 to HELLP syndrome. Acta Haematol 2002;108:
31. Goldman-Wohl D, Yagel S. Regulation of 162–3
trophoblast invasion: from normal implantation 46. Ludlam CA. The evidence behind inhibitor
to preeclampsia. Mol Cell Endocrinol 2002;187: treatment with recombinant factor VIIa. Pato-
233–8 physiol Haemost Thromb 2002;32(Suppl 1):13–18
32. Tank PD, Nadanwar YS, Mayadeo NM. Out- 47. Maclean A, Almeida Z, Lopez P. Complications
come of pregnancy with severe liver disease. Int J of acute fatty liver of pregnancy treated with acti-
Gynaecol Obstet 2002;76:27–31 vated protein C. Arch Gynecol Obstet 2005;273:
33. Sibai BM. The HELLP syndrome (hemolysis, 119–21
elevated liver enzymes, and low platelets): much 48. Mikaszewska-Sokolewicz M, Mayzner-
ado about nothing? Am J Obstet Gynecol 1990; Zawadzka E. Use of recombinant human
162:311–16 activated protein C in treatment of severe sepsis
34. McCrae KR, Cines DB. Thrombotic micro- in a pregnant patient with fully symptomatic
angiopathy during pregnancy. Sem Hematol ovarian hyperstimulation syndrome. Med Sci
1997;34:148–58 Monit 2005;11:27–32
230
252
30 August 2006 14:21:15
49. Toh CH, Dennis M. Disseminated intravascular 64. Laffan M, Brown SA, Collins PW, et al. The
coagulation: old disease, new hope. BMJ 2003; diagnosis of von Willebrand disease: a guideline
327:974–7 from the UKHCDO. Haemophilia 2004;10:
50. Kashyap R, Choudhry VP, Mahapatra M, et al. 199–217
Postpartum acquired haemophilia: clinical rec- 65. Pasi KJ, Collins PW, Keeling DM, et al.
ognition and management. Haemophilia 2001;7: Management of von Willebrand disease: a guide-
327–30 line from the UKHCDO. Haemophilia 2004;10:
51. Porteous AO, Appleton DS, Hoveyda F, Lees 218–31
CC. Acquired haemophilia and postpartum 66. Mannucci PM. How I treat patients with von
haemorrhage treated with internal pudendal Willebrand disease. Blood 2001;97:1915–19
embolisation. Br J Obstet Gynaecol 2005;112: 67. Giangrande PL. Management of pregnancy in
678–9 carriers of haemophilia. Haemophilia 1998;4:
52. Boggio LN, Green D. Acquired hemophilia. Rev 779–84
Clin Exp Hematol 2001;5:389–404 68. Kujovich JL. Von Willebrand disease and
53. Bates SM, Ginsberg JS. How we manage venous pregnancy. J Thromb Haemost 2005;3:246–53
thromboembolism during pregnancy. Blood 69. Stedeford JC, Pittman JA. Von Willebrand’s
2002;100:3470–8 disease and neuroaxial anaesthesia. Anaesthesia
54. Bates SM, Greer IA, Hirsh J, Ginsberg JS. Use 2000;55:1228–9
of antithrombotic agents during pregnancy. 70. Horn EH. Thrombocytopenia and bleeding
Presented at the Seventh ACCP Conference disorders. In James DK, Steer PJ, Weiner CP,
on Antithrombotic and Thrombolytic Therapy. Gonik B, eds. High-Risk Pregnancy: Management
Chest 2004;126:627–44 Options, 3rd edn. Amsterdam: Elsevier, 2006:
55. Ginsberg JS, Hirsh J, Turner C, et al. Risks to the 901–24
fetus of anticoagulant therapy during pregnancy. 71. Miller R. Counselling about diagnosis and inher-
Thromb Haemost 1989;61:197–203 itance of genetic bleeding disorders: haemophilia
56. Vitale N, De Feo M, De Santo LS, et al. Dose- A and B. Haemophilia 1999;5:77–83
dependent fetal complications of warfarin in 72. Ludlam CA, Pasi KJ, Bolton-Maggs P, et al.
pregnant women with mechanical heart valves. A framework for genetic service provision for
J Am Coll Cardiol 1999;33:1637–41 haemophilia and other inherited bleeding
57. Walker ID. In O’Shaughnessy D, Makris M and disorders. Haemophilia 2005;11:145–63
Lillicrap D, eds. Obstetrics in Practical Hemostasis 73. Walker ID, Walker JJ, Colvin BT, et al. Investi-
and Thrombosis, 1st edn. Oxford: Blackwell gation and management of haemorrhagic disor-
Publishing, 2005:139–48 ders in pregnancy. J Clin Pathol 1994;47:100–8
58. Kearon C, Hirsh J. Management of anticoagu- 74. Kulkarni R, Lusher JM, Henry RC, Kallen DJ.
lation before and after elective surgery. N Engl J Current practices regarding newborn intracranial
Med 1997;336:1506–11 haemorrhage and obstetrical care and mode of
59. Rahimi G, Rellecke S, Mallmann P, Nawroth F. delivery of pregnant haemophilia carriers: a
Course of pregnancy and birth in a patient with survey of obstetricians, neonatologists and
Bernard–Soulier syndrome – a case report. Pato- haematologists in the United States, on behalf of
physiol Haemost Thromb 2002;32(Suppl 1):13–18 the National Hemophilia Foundation’s Medical
60. Kriplani A, Singh BM, Sowbernika R, and Scientific Advisory Council. Haemophilia
Choudhury VP. Successful pregnancy outcome 1999;5:410–15
in Bernard–Soulier syndrome. J Obstet Gynaecol 75. Bolton-Maggs PH, Perry DJ, Chalmers EA, et al.
Res 2005;31:52–6 The rare coagulation disorders – review with
61. Kale A, Bayhan G, Yalinkaya A, Yayla M. guidelines for management from the UKHCDO.
The use of recombinant factor VIIa in a Haemophilia 2004;10:593–628
primigravida with Glanzmann’s thrombasthenia 76. Jimenez-Yuste V, Villar A, Morado M, et al.
during delivery. J Perinat Med 2004;32:456–8 Continuous infusion of recombinant activated
62. Pajor A, Nemes L, Demeter J. May Hegglin factor VII during caesarean section delivery in
anomaly and pregnancy. Eur J Obstet Gynecol a patient with congenital factor VII deficiency.
Reprod Biol 1999;85:229–31 Haemophilia 2000;6:588–90
63. Favaloro EJ. Laboratory assessment as a critical 77. Fadel HE, Krauss JS. Factor VII deficiency and
component of the appropriate diagnosis and pregnancy. Obstet Gynecol 1989;73:453–4
sub-classification of von Willebrand’s disease. 78. Eskandari N, Feldman N, Greenspoon JS.
Blood Rev 1999;13:185–204 Factor VII deficiency in pregnancy treated with
231
253
30 August 2006 14:21:15
recombinant factor VIIa. Obstet Gynecol 2002;99: 82. McMahon MJ, James AH. Combined deficiency
935–7 of factors II, VII, IX, and X (Borgschulte–
79. Mariani G, Konkle BA, Ingerslev J. Congenital Grigsby deficiency) in pregnancy. Obstet Gynecol
factor VII deficiency: therapy with recombinant 2001;97:808–9
activated factor VII – a critical appraisal. Haemo- 83. David AL, Paterson-Brown S, Letsky EA. Factor
philia 2006;12:19–27 XI deficiency presenting in pregnancy: diagnosis
80. Muleo G, Santoro R, Iannaccaro PG, et al. and management. Br J Obstet Gynaecol 2002;
The use of recombinant activated factor VII in 109:840–3
congenital and acquired factor VII deficiencies. 84. Kadir RA, Economides DL, Lee CA. Factor XI
Blood Coagul Fibrinolysis 1998;9:389–90 deficiency in women. Am J Hematol 1999;60:
81. Romagnolo C, Burati S, Ciaffoni S, et al. Severe 48–54
factor X deficiency in pregnancy: case report and
review of the literature. Haemophilia 2004;10:
665–8
232
254
30 August 2006 14:21:15
26
THE USE OF RECOMBINANT FACTOR VIIa
S. Sobieszczyk and G. H. Brfborowicz
INTRODUCTION acute, uncontrolled, or otherwise profound

bleeding, and in other clinical circumstances
As described in detail in other chapters of this
associated with excessive bleeding in patients
volume, conditions with excessive bleeding, as
without pre-existent coagulation defects5,6.
are seen with uterine rupture, placenta accreta,
Indeed, the early descriptions of the benefits of
abruption and uterine atony, often require
rFVIIa in trauma patients7–9 were bolstered by a
intensive resuscitation with blood components
compassionate use study, which suggested that
and coagulation factors. In such circumstances,
rFVIIa administration could reverse massive
blood transfusion may be life-saving, but on
bleeding, and thus significantly decrease trans-
occasion involves exposing the patient to addi-
fusion requirements observed in critically ill,
tional risks. Over the years, numerous efforts
multi-transfused trauma patients10,11. Recently,
have been put forward to reduce these risks.
rFVIIa was approved for the treatment of
One of the most spectacular is discussed in this
hemorrhage associated with congenital
chapter.
factor VII deficiency12,13 and Glanzmann’s
Recombinant activated factor VII (rFVIIa)
thrombasthenia14,15.
(NovoSeven®; Novo Nordisk A/S, Bagsvaerd,
Denmark) was developed for the treatment of
spontaneous and/or surgical bleeding episodes
PECULIARITIES OF OBSTETRIC
in patients with hemophilia A or B with forma-
HEMORRHAGE
tion of allo-antibodies to FVIII or FIX after
replacement therapy1–3. rFVIIa is currently Patients who develop massive, life-threatening
licensed for this indication in most countries postpartum hemorrhage often have a combina-
world-wide. The US Food and Drug Adminis- tion of ‘coagulopathic’ diffuse bleeding in addi-
tration (FDA) licensed rFVIIa on March 25, tion to ‘surgical bleeding’. Whereas bleeding
1999 for bleeding episodes in patients with from larger vessels may be controlled by
hemophilia A or B and inhibitors to FVIII or surgeons using a variety of operations (see
FIX. The FDA approved use of rFVIIa in 2005 Chapters 30–32), the ability to control diffuse
for additional indications such as surgical proce- bleeding is limited and, in many cases, not feasi-
dures in patients with hemophilia A or B and ble. Thus administration of hemostatic drugs
inhibitors, and treatment of bleeding episodes that can control the coagulopathic component
in patients with factor VII deficiency4. In of blood loss may reduce mortality and morbid-
Europe, it is also approved for use in bleeding ity in such patients. Clinical experience pres-
episodes in patients with acquired hemophilia ently suggests that rFVIIa is a safe and effective
due to auto-antibodies against endogenous hemostatic measure in severe obstetric hemor-
FVIII or FIX, surgical procedures in this group rhage, both as a adjunctive treatment to surgical
of patients, and Glanzmann’s thrombasthenia. hemostasis as well as a ‘salvage’ or ‘rescue’ ther-
Beyond its currently recognized indications, apy where postpartum hemorrhage is refractory
rFVIIa has been effectively used ‘off label’ on an to current pharmaceutical and ‘uterus sparing’
empirical basis as a general hemostatic agent in surgical techniques. The ‘evidence’ behind the
a wide range of conditions associated with preceding statement comes from three sources:
233
255
30 August 2006 14:21:16
(1) Studies on its mechanism of action; a two-chain molecule consisting of a light chain
of 152 amino acid residues and a heavy chain
(2) Accumulating reports in the literature; and
of 254 amino acid residues held together by a
(3) Data from clinical studies. single disulfide bond19,20 (Figures 1 and 2).
All suggest that rFVIIa has the potential to func-
tion as a ‘universal hemostatic agent’16 across a Production of rFVIIa using recombinant
range of indications characterized by impaired DNA technique
thrombin generation in non-hemophilic patients,
The development of rFVIIa was undertaken to
many of whom are critically ill and refractory to
alleviate the problems associated with the use of
other hemostatic treatment options.
plasma-derived factor VIIa, such as limited sup-
The usual manner for treating postpartum
ply and possible viral contamination. Multiple
hemorrhage includes, first, non-invasive/non-
steps were involved in the development of this
surgical methods, including administration of
recombinant protein. First, the human gene for
crystalloid solutions and/or red blood cells,
factor VII, located on chromosome 13, com-
uterine massage, uterotonic medications
prising eight exons (coding regions), was iso-
(oxytocin, ergotamine, prostaglandins), and,
lated from the liver gene library. After standard
second, invasive/surgical methods, e.g. ligation
amplification procedures used to generate mul-
of uterine vessels, ligation of iliac arteries, angio-
tiple copies of the hFVII gene, it was transfected
graphic embolism of uterine/iliac arteries, or the
into a baby hamster kidney cell line. A master
B-Lynch method. Unfortunately, the overall
cell bank of the transfected cell line that secretes
effectiveness of such procedures to arrest hem-
factor VII in a single-chain form into the culture
orrhage and prevent the need for emergency
medium was then established. During the last
hysterectomy is estimated to be only about
steps, proteolytic conversion by autocatalysis to
50%17,18. Moreover, comparatively few centers
the active two-chain form (rFVIIa) takes place
world-wide have access to the physical equip-
in a chromatographic purification process,
ment or surgical manpower resources necessary
which was shown to remove exogenous viruses.
to conduct all the aforementioned procedures
No human serum or other proteins are used in
the production of rFVIIa (see Chapter 15). The
COAGULATION FACTOR VII: protein backbone is identical with human puri-
THE HUMAN PROTEIN AND fied factor VIIa. The final product (rFVIIa),
RECOMBINANT PRODUCT despite minor differences in carbohydrate
composition, is structurally similar to plasma-
Structure of the human FVII (hFVII)
derived factor VIIa. The activity of rFVIIa is
Human factor VII (eptacog alpha) is a serine similar to that of natural factor VIIa present in
protease (molecular weight 50 kDa) composed the body21,22 (see Table 1).
of 406 amino acid residues, belonging to the Human activated factor VII (hFVIIa) or
group of vitamin K-dependent coagulation recombinant activated factor VII (rFVIIa) is a
glycoproteins. The primary site of FVII naturally occurring initiator of hemostasis that
synthesis in humans is the liver. Factor VII is vital to the coagulation process, as it combines
is composed of four discrete domains: with tissue factor (TF) at the site of blood vessel
a γ-carboxyglutamic acid (Gla)-containing damage in a natural way, stimulates thrombin
domain, two epidermal growth factor (EGF)- generation, permits stable fibrin clot formation,
like domains, and a serine protease domain. All and thereby the cessation of bleeding.
appear to be involved, to different extents, in an
optimal interaction with tissue factor (TF). The
PHARMACOKINETIC STUDIES OF
Gla domain of factor VII is also essential for
rFVIIa IN HUMANS
activation of factor X and other macromolecular
substrates. The activation of factor VII to factor The pharmacokinetics of single-bolus doses
VIIa involves the hydrolysis of a single peptide of rFVIIa have been studied in various adult
bond between Arg152 and Ile153. The result is populations: patients with hemophilia, patients
234
256
06 September 2006 16:35:56
The use of recombinant factor VIIa
Figure 1 Three-dimensional molecular structure of factor VII. Reproduced with permission from Novo
Nordisk
with cirrhosis, and healthy volunteers. The average percentage of the preparation found in
pharmacokinetic parameter values of rFVIIa plasma was significantly lower after administra-
after bolus administration were similar. The tion of rFVIIa in a dose of 70 µg/kg (42.7%)
elimination half-life (t1/2) ranged from 2.45 to compared to doses of 17.5 µg/kg (50.1%)
2.72 h and clearance (CL) ranged from 32.8 or 35 µg/kg (49.0%) (p = 0.0067). Additional
to 34.9 ml/h.kg23. Lindley and colleagues doses for specific patient populations are war-
investigated the single-dose pharmacokinetics ranted however23,24. An increased elimination
of rFVIIa, evaluated in three dose levels (17.5, rate and lower recovery of rFVIIa during bleed-
35.0, 70 µg/kg) in hemophilic A/B patients with ing may be related to consumption through
inhibitors. The results of these investigations complex formation with TF exposed at the site
demonstrate that the mean t1/2 of recombinant of vessel damage and on the phospholipids
factor VIIa is independent of dose level24. exposed on the activated platelet surface. The
Pharmacokinetic evaluations suggest the volume of distribution at steady state (Vss), is
elimination of rFVIIa follows linear kinetics two to three times that of plasma and similar to
with a faster clearance rate and shorter t1/2 when the half-life of recombinant factor VIIa24.
rFVIIa is administered for bleeding episodes
(medians: 2.70 and 2.41 h, respectively) com-
MECHANISM OF HEMOSTATIC
pared to non-bleeding indications (medians:
ACTION OF rFVIIa (see Figure 3)
3.44 and 2.89 h, respectively). Therefore, the
duration of action may by shorter when rFVIIa Recombinant factor VIIa induces hemostasis at
is used to control bleeding episodes. The the site of injury. The mechanism of action
235
257
06 September 2006 16:35:17
Table 1 Recombinant vs. plasma-derived FVIIa21
Amino acid sequence identical

Amino acid composition identical
Gamma-carboxylation identical
Peptide map identical
Biological activity identical
Carbohydrate composition similar
following injury. Tissue injury disrupts the

endothelial cell barrier that normally separates
TF-bearing cells from the circulating blood.
Once exposed to the blood, TF serves as a
high-affinity receptor for FVIIa. FVIIa is found
in the circulation, comprising about 1% of the
total circulating FVII protein mass in the
plasma. It is endowed with very weak enzymatic
activity, which only becomes fully realized upon
binding to its cofactor, TF, at a site of vascular
injury25,26. Factor VIIa alone shows very little
proteolytic activity, only attaining its full
enzymatic potential when complexed to TF.
In studies using TF incorporated into lipid
vesicles, van’t Veer and colleagues demon-
strated that zymogen FVII acts as an inhibitor
of FVIIa:TF-initiated thrombin generation.
The addition of FVIIa at a concentration of
10 nmol/l in hemophilic conditions overcomes
this inhibition and results in a thrombin genera-
tion equivalent to normal. These data suggest
that the therapeutic effect of rFVIIa is due in
part to its ability to overcome the inhibitory
effect of physiologic FVII on FVIIa:TF-initiated
thrombin generation27.
However, if TF is no longer available or
exposed to the clotting factors in the blood-
stream, e.g. when a platelet plug covers the TF-
Figure 2 The active two-chain enzyme factor
VIIa, is generated by specific cleavage AT Arg 152.
containing subendothelial space, or when TF
Reproduced with permission from Novo Nordisk activity is inhibited by TFPI (tissue factor path-
way inhibitor), then rFVIIa-mediated large-
scale thrombin generation could take place on
includes the binding of factor VIIa to the the activated platelet surface independently of
exposed tissue factor-dependent pathway and, TF28.
independently of tissue factor, activation of The initial formation of a TF/FVIIa or TF/
factor X directly on the surface of activated rFVIIa complex allows activation of FIX and
platelets localized to the site of injury25,26. FX, and is crucial in generating the initial con-
The formation of the TF/FVIIa or TF/ version of small amounts of prothrombin into
rFVIIa complex at the site of injury is necessary thrombin (on the TF-bearing cells), which is
to initiate hemostasis. TF is a membrane-bound essential to the amplification and propagation
glycoprotein, which normally is expressed on phase of coagulation. FXa cannot move to the
cells in the subendothelium and is only exposed platelet surface because of the presence of
236
258
30 August 2006 14:21:23
Figure 3 Mode of action of Eptacog alfa (activated) (with permission Novo Nordisk). (1) Tissue factor
(TF)/FVIIa, or TF/rFVIIa interaction, is necessary to initiatiate hemostasis. (2) At pharmacological
concentrations, rFVIIa directly activates FX on the surface of locally activated platelets.This activation will
initiate the ‘thrombin burst’ independently of FVIII and FIX. This step is independent of TF. (3) The
thrombin burst leads to the formation of a stable clot
normal plasma inhibitors, but instead remains platelets provide for further thrombin genera-
on the TF-bearing cell and activates a small tion. Platelet-surface FXa is relatively protected
amount of thrombin. Thrombin leads to the from normal plasma inhibitors and can complex
activation of platelets and FV and FVIII at the with platelet-surface FVa, where it activates
site of injury. thrombin in quantities sufficient to provide for
This small amount of thrombin is not suffi- fibrinogen cleavage.
cient for fibrinogen cleavage, but is critical for FIXa, FVIIIa and FVa bind efficiently to the
hemostasis, as it can activate platelets, activate surface of the activated platelet and further acti-
and release FVIII from von Willebrand factor vation of FX into FXa occurs via the complex
(vWF) or activate platelet and plasma FV, and between FIXa and FVIIIa. During amplifica-
FXI. FIXa moves to the platelet surface, where tion, FXa complexes with FVa to generate
it forms a complex with FVIIIa and activates thrombin and subsequently activate FV, FVIII
FX on the platelet surface. The activated and platelets.
237
259
30 August 2006 14:21:28
At pharmacological concentrations (supra- time (PT) and the activated partial thrombo-
physiological doses), rFVIIa also directly plastin time (APTT). The PT generally short-
activates FX on the surface of locally activated ens to around 7–8 s except in FV- or
platelets, helping to generate thrombin and FX-deficient plasma, suggesting that patients
fibrin (platelet-dependent TF-independent completely deficient in FV and/or FX will not
pathways). rFVIIa does not bind to resting benefit from therapy with this product33. PT
platelets. Instead, the effect of high-dose rFVIIa may not adequately reflect coagulation func-
(which only activates FX on activated platelets) tion. The APTT shortening is due to the direct
is localized to the sites of vessel injury where TF activation of FX by circulating FVIIa on the
is exposed and platelets are activated29,30. This phospholipids used in the partial thrombo-
results in the conversion of prothrombin into plastin time test. Data indicate that clinical
large amounts of thrombin. The full thrombin improvement during rFVIIa treatment is associ-
burst mediated by FXa in complex with FVa is ated with a shortening of APTT of 15–20 s33.
necessary for the formation of a fully stabilized Post-rFVIIa coagulation parameters normalize
and solid fibrin hemostatic plug. as early as 20 min after infusion. Thus, the
rFVIIa works by producing a stable fibrin shortening of these two screening tests of
clot directly at the site of vascular injury, both coagulation does not necessarily reflect clinical
dependently and independently of TF. This effectiveness, which is judged subjectively.
reaction provides an extremely strong activation Coagulopathy is usually easy to recognize
of thrombin at the site of tissue damage, leading by the clinical assessment of ongoing bleeding,
to the formation of a stable fibrin network. physical examination and observation of oozing
Administration of rFVIIa might result in forma- from cut surfaces, intravascular catheter sites
tion of a more stable hemostatic plug by a or mucus membranes. The initial evaluation
variety of mechanisms, including enhancement during hemorrhage includes the PT, APTT,
of activation of thrombin activatable fibrinolysis thrombin time (TT) and fibrinogen concentra-
inhibitor31, improvement of the physical prop- tion, antithrombin and platelet count. In the
erties of the fibrin clot, enhancement of platelet interpretation of these tests, it is important to
activation32, and possibly enhancement of know the normal range and to be aware of the
FXIII activation. sensitivity of the screening tests for each coagu-
Lisman and colleagues observed that the lation factor, as these vary from laboratory
enhanced thrombin generation from FVIIa not to laboratory. In addition, assays of clotting
only accelerates clot formation, but also inhibits parameters may provide different results with
fibrinolysis by activation of thrombin activatable different reagents, although these parameters do
fibrinolytic inhibitor (TAFI) in factor VIII- not show a direct correlation to the level of
deficient plasma28. rFVIIa binding to thrombin- hemostasis achieved. Finally, it is important to
activated platelets provides extra thrombin and remember that laboratory coagulation para-
thus ensures both full activation of TAFI and meters may be used as an adjunct to the
FXIII, and the formation of a dense fibrin struc- clinical evaluation of hemostasis for monitoring
ture. The full thrombin burst generated con- the effectiveness and treatment schedule of
verts fibrinogen into a firm plug that is resistant rFVIIa34.
to premature lysis, thereby facilitating full Clotting parameters obtained prior to rFVIIa
hemostasis. administration are often outside the normal
range, perhaps indicating the development
of dilutional or consumption coagulopathy in
these patients. Post rFVIIa, clotting parameters
MONITORING THE CLINICAL EFFECT
improve, but do not normalize, and thus cannot
OF rFVIIa
be used as predictors of rFVIIa efficacy.
Currently, there is no good and/or satisfactory Laboratory monitoring of the efficacy of
laboratory method for monitoring the clinical rFVIIa treatment is helpful. The effect on PT
effectiveness of rFVIIa. Administration of is particularly marked, but this does not
rFVIIa results in shortening of the prothrombin always translate to clinically improved blood
238
260
06 September 2006 16:34:11
coagulation. Similarly, measurement of the level of thromboembolic complications following

of FVII in plasma does not correlate with clini- administration of the drug both in animal mod-
cal efficacy. Study of the effects of rFVIIa on els and in clinical use, all suggest that rFVIIa is a
monitoring plasma FVIIa levels demonstrates a useful adjunctive therapy for control of severe
linear relationship between the concentration of postpartum hemorrhage. Recombinant FVIIa is
FVIIa and FVII:C (functional clotting ability), a manufactured product, does not contain any
but the therapeutic concentration range for human plasma components, and therefore is
FVIIa has not yet been established. The use of free from viral contamination. Neither albumin
plasma VIIa levels is controversial, and is not an nor any other human protein is used in its man-
assay that is widely available. ufacturing process. This means that there is no
Levels of functional fibrinogen and risk of transmission of human viruses or prions.
antithrombin do not change during repeated Strict quality control standards are applied to
injections of rFVIIa for the treatment of hemor- the fermentation process as well as the subse-
rhage. The minimal changes that occur post- quent extensive purification measures. Genetic
operatively are not greater than those seen with recombination eliminates the dependency on
patients who do not have coagulation disorders. donors and allows for the production of
Nonetheless, it is still advisable to monitor unlimited amounts of the medication20.
patients at risk of systemic activation. Safety analyses demonstrate that rFVIIa
Telgt and colleagues showed that low is associated with very few treatment-related
concentration of rFVIIa, in the absence of TF, adverse events and is very well tolerated. Thus,
can activate FX as assayed by the PT33,35. experience with recombinant factor VIIa in
Higher concentration of rFVIIa had no addi- several thousand patients has shown that the
tional effect on the PT. At rFVIIa doses well incidence of non-serious adverse events is 13%
below the clinically therapeutic dose, a maxi- and serious adverse events are less than 1%37.
mum shortening of the PT occurs. Thus, at Aledort calculated that the risk of rFVIIa-
doses in the clinically therapeutic range, no fur- related thrombosis is 25 per 105 infusions38.
ther effect on the PT is observed. This suggests Despite the mechanism of action, use of rFVIIa
that, at concentrations typical for clinical use, in DIC and sepsis remains controversial. Sev-
tests based on the PT are not useful for eral reports suggest that rFVIIa may be used
monitoring the effect of rFVIIa. Telgt and safely in such situations, without induction of
colleagues, in an experimental study, observed thrombotic complications or when conventional
that rFVIIa effectively reduced PT and APTT replacement therapy with fresh frozen plasma
in normal and deficient (FVIII, FIX, FXI, and red blood cell concentrates fails to provide a
FXII) plasma. This reduction of both para- hemostatic response. Non-serious side-effects
meters (PT and APTT) has been attributed to are rarely seen during treatment with recombi-
the ability of rFVIIa to directly activate FX, nant factor VIIa; the most common being pain
even in the absence of TF34,35. at the infusion site, fever, headache, vomiting,
The best available indicator of rFVIIa changes in the blood pressure and skin-related
efficacy is the arrest of hemorrhage judged hypersensitivity reactions. Adverse events have
by visual evidence, hemodynamic stabilization not been related to dose.
and reduced demand for blood components36.
There is currently no satisfactory laboratory test
OUR EXPERIENCE
to monitor the clinical effectiveness of rFVIIa.
Between 2000 and 2006 in the Department
of Gynecology and Obstetrics, University of
SAFETY OF rFVIIa
Medical Sciences, Poznan we used rFVIIa in
The complex coagulopathy and high complica- almost 45 cases of postpartum hemorrhage39–46.
tion rates seen in patients with intractable According to data gathered from other areas of
postpartum hemorrhage, together with the Poland, we estimate that it has been used
understanding of the localized mechanism in approximately 100 cases of postpartum
of action of rFVIIa, and the low risk hemorrhage.
239
261
06 September 2006 16:33:17
The data presented below concern our first Recombinant FVIIa was administered intra-
18 patients in whom rFVIIa was used. Detailed venously at doses of 16.6–48 µg/kg. In most
information is presented in Tables 2–5. Our cases, single administration of rFVIIa was suffi-
patient data were obtained when we were using cient. However, in severe coagulopathy coexist-
a study protocol and were prepared to use the ing with postpartum hemorrhage or prolonged
drug. This was not always the case in other periods of treatment (transfusions, complica-
centers (see Table 6). tions of shock) and recurrent bleeding, a second
dose similar to the initial dose was necessary to
Table 2 Clinical details of patients with severe, control the bleeding.
recurring and uncontrollable bleeding post-delivery
Number Conclusions
of patients The analysis of our data clearly shows that
Number of postpartum hemorrhages 18
rFVIIa was an effective hemostatic drug, which
significantly decreased bleeding and led to the
Cause of bleeding/complications rapid stabilization of our patients’ conditions.
Uterine atony 8 Clearly, the early use of this agent decreases the
Genital tract trauma 1 amount of transfused preparations. An impor-
Disseminated intravascular coagulation 8
tant secondary observation was the contraction
Shock 18
of the uterus after the drug application in
Reoperations before rFVIIa administration 7 patients who had qualified for hysterectomy
Obstetric hysterectomy* 2 shortly before the drug was administered. We
suggest that rFVIIa should be administered in
*In six cases, hysterectomy was not performed.
every case in which embolization of uterine
rFVIIa was administered after the decision to oper-
arteries is being considered. Coagulation
ate was made due to uncontrolled, life-threatening
bleeding. After its administration, the bleeding parameters showed typical shortening of PT
stopped and the operation was not necessary. In and APTT; however, the clinical effect – control
two women, hysterectomy was performed in another of bleeding – was the most important overall
hospital, before the patients were transported to our effect of the drug. There were no complications
department of rFVIIa administration. The dose, timing of
administration after the diagnosis of postpartum
Table 3 Blood loss before and after rFVIIa hemorrhage, and the apparent ability to
administration enhance uterine contractility will need further
study in the future.
Blood loss Median (range) (ml)
Before rFVIIa 3000 (1800–6800) WORLD-WIDE EXPERIENCE

After rFVIIa 0.00 (0–350)
Tables 6–8 present the world-wide experience
with rFVIIa in obstetric hemorrhage. The
Table 4 Transfusion needed before and after results reported in the literature support the
rFVIIa administration benefit of rFVIIa therapy in obstetric cases with
Before rFVIIa After rFVIIa major/life-threatening hemorrhage, even in
the presence of disseminated intravascular
Median Median coagulopathy (DIC)-like ‘coagulopathy’. They
(range) U/P (range) U/P demonstrate that rFVIIa is highly effective and
safe in allowing quick arrest of life-threatening
Red blood 6 (3–13) 6 4 (0–9) 3
postpartum hemorrhage unresponsive to con-
cell (IU)
Fresh frozen 4 (1–8) 4 2 (0–9) 2
ventional treatments. Treatment with rFVIIa
plasma (IU) led to a reduction in the use of blood products
in this relatively large group of patients, decreas-
U/P, units per patient ing blood product exposure for patients and
240
262
30 August 2006 14:21:29
Table 5 Selected laboratory tests before and after rFVIIa administration. Data are given as median (range)
Normal Before 2 hours 4 hours 12 hours

Parameter range rFVIIa after rFVIIa after rFVIIa after rFVIIa
PT (s) 11.5–13.5 17.35 11.10 11.25 12.65

(11.9–26.7) (9.1–18.3) (9.1–17.6) (11.2–17.1)
APTT (s) 25–37 55.00 35.00 36.80 39.10
(26–81) (26–76) (22–69) (24–60)
PLT (Gpt/l) 140–440 76.50 70.00 69.50 70.50
(21–223) (20–197) (19–186) (37–165)
PT, prothrombin time; APTT, activated partial thromboplastin time; PLT, platelets
sparing an expensive and limited resource. other clinical measures had failed. There was
Administration of rFVIIa should be also con- no clear correlation between the severity of
sidered before hysterectomy and as an adjunct bleeding and the dose of rFVII administered.
to invasive/surgical procedures, before they are Possibly the ‘timing’ determined the level of the
undertaken. This is particularly true in patients dosing.
who wish to preserve fertility
Efficacy
Conclusions
Bleeding either stopped, markedly decreased
Randomized controlled studies are required to or decreased following rFVIIa administration in
determine the optimal dose and dose schedule 54 of the cases. In one patient, there was no
of rFVIIa for intractable postpartum hemorrhage response to therapy with rFVIIa. Also only in
and to investigate whether the need for hyster- one patient after an early significant reduction
ectomy/surgical procedures and overall morbid- of bleeding, recurrence was observed. In gen-
ity rates can be reduced by earlier treatment eral, however, the rapid onset of action means
with higher doses of rFVIIa. In the meanwhile, that rFVIIa can be used in the perioperative
clinicians caring for acutely bleeding obstetric period. There was no clear correlation between
patients should be aware of the potential of the speed of response and either the type of pro-
rFVIIa to arrest life-threatening postpartum cedure performed, the severity of the bleeding
hemorrhage. Although an expensive product, condition, or the dose of rFVIIa given.
a trial of one to four doses of rFVIIa can be Most patients continued to require some
justified in cases of uncontrolled bleeding which form of blood product replacement therapy
persists despite maximal medical and surgical during the 24 h following rFVIIa administra-
treatment to achieve hemostasis. tion, but the need was greatly reduced
Although the limitations of anecdotal case compared with the 24 h prior to rFVIIa admin-
data are recognized, in the absence of efficacy istration. No correlation existed between
and safety data from randomized trials, volun- baseline and post-rFVIIa administration in
tary registry submissions are being used to pro- laboratory measurements and the predictability
vide a preliminary insight into the scope of the of response to rFVIIa (data obtained from
low incidence of clinical problems, as well as the references but not presented in tables). Further-
usefulness and adverse effects of this medication more, of great importance, the results observed
when it is used ‘off-label’. in these tables of cases of postpartum hemor-
rhage suggest that rFVIIa may be administered
even in the presence of DIC-like ‘coagulo-
rFVIIa dose
pathy’. In the patients shown in Tables 6–8,
When a rationale for using rFVIIa was stated, it major conditions reported to be associated
was most commonly ‘last-resort’ therapy, after with postpartum hemorrhage included some
241
263
30 August 2006 14:21:29
30 August 2006 14:21:30
Table 6 Clinical characteristics of patients with risk of severe, recurring and uncontrollable blood loss during delivery and postpartum: literature review
Blood products Blood loss Dose of rFVIIa

given pre-rFVIIa before (µg/kg) Overall bleeding
Provocation of Type of (units) (hemostatic rFVIIa Timing (when (number of response to
Year Ref. n bleed delivery Surgical treatment agents) (ml) rFVIIa given) doses) rFVIIa (min) Comments
2001 47 1 DIC, liver CS HYS NA > 3000 Post hysterectomy; 90 (9) Response after
dysfunction, renal last resort 3-h intervals 2 single doses;

failure; severe significantly
intra-abdominal reduced
bleeding after CS
2002 48 1 Congenital FVII VD No No No evidence Prophylactic first 50; 35 No evidence The first case of a
deficiency (1% of bleeding dose at complete 4-h intervals of bleeding pregnant woman
before application dilatation of the with FVII deficiency
of rFVIIa) cervix receiving rFVIIa
intrapartum
2002 49 1 Acquired VD HYS RBC (65); FFP NA 11 days 160 Bleeding
242
264
hemophilia (FVIII (60); CRYO (60); (massive) post-delivery; last stopped
0.5%) vWF (3 × 500); resort (rapidly)
FVIII (30 × 1068);
FIX (26 × 600);
18 g sandoglobulin
2002 50 1 2-h post CS CS No RBC (12); FFP NA Last resort 90 Bleeding Normalization of
massive vaginal (10); PPTs (8); stopped coagulation tests
bleeding; shock; CRYO (950)
DIC, HELLP
2003 51 1 Bleeding from the CS Under-running RBC (1.5); FFP > 3000 Last resort 90 Bleeding The balloon was

placenta bed in sutures in the (500 ml) stopped (15) removed on the first
lower uterine placenta bed; postoperative day
segment and application of hot
cervical canal packs; direct manual
tamponade with
surgical gauze;
insertion of
intra-cervical
Foley’s catheter
balloon
2003 52 2 (case 1) uterine (case 1) (case 1) subtotal (case 1) RBC (10); NA (case 1) NA Bleeding (case 2)
rupture, shock VD HYS FFP (4) intraoperative stopped (few Hysterectomy was
(case 2) uterine (case 2) (case 2) RBC (5) (case 2) before minutes) avoided
30 August 2006 14:21:30

atony eCS planned
hysterectomy
2003 53 1 Uterine atony; IVD laparotomy: Before 1st NA Post laparotomy 60; 120 Bleeding Cardiac arrest,
shock bilateral artery administration RBC 2-h interval, stopped resuscitation;
ligation; subtotal (42); FFP (31); (2nd for high-pressure
HYS; packing of PPTs (4); consolidation) ventilation,
pelvis (desmopressin) pulmonary edema,
before 2nd pneumothorax,

administration FFP ARDS
(3); PPTs (2)
2003 54 1 Uterine atony, CS HYS RBC (3); FFP (2); NA Intraoperative 12 Bleeding During general
pre-eclampsia CRYO (6) (CS) before significantly anesthesia
hysterectomy reduced induction, failed
intubation was
followed by cardiac
arrest;
postoperatively
243
DIC; ARDS; transit
265
encephalopathy,
and brachial
venouse thrombosis
(Folckmann
syndrome)
2003 55 2 (case 1) congenital (case 1) No No No evidence (case 1) (case 1) 60; No evidence (case 2) No evidence
deficiency of FVII VD of bleeding Prophylactic first 30 (5) every of bleeding of FVII deficiency
(2% before (case 2) dose at complete 2 h.
application rFVIIa) CS dilatation of the (case 2) 90
(case 2) liver cervix

dysfunction (case 2) prophylactic
before CS
2003 56 1 AFE, DIC CS HYS; pelvic RBC (12); FFP (8); NA Last resort 60 Bleeding MOF, died
packing (aprotinin) significantly
reduced
continued
Table 6 Continued
30 August 2006 14:21:30

2004 57 1 Uterine rupture; IVD 3 laparotomy: Before 1st 4000 to 2nd Before, intra- and 120 (19), start Bleeding Cardiac arrest,
shock; DIC 1st: HYS; 2nd: administration: laparotomy; postoperative before 2nd significantly resuscitation before
packing of pelvis; RBC (26); FFP before 3rd period; last resort laparotomy, reduced or 2nd laparotomy
3rd: small arteries (11); PPTs (10); laparotomy repeated stopped; (hyperkalemia
ligated in the PCC (1200). sudden following recurrent 8.5 mmol/l,
broad ligaments Total: RBC (27); increase of next 2 days. bleeding was hypothermia 32°C);
FFP (27); PPTs bleeding First two observed MOF
(10); 22 1350 l in 1 h doses (1st Recurrent bleeding
plateletpheresis; laparotomy) at was observed

(tranexamic acid) 1-h intervals, because patient
next doses developed severe
during the 2nd hypothermia,
day 3 doses; acidosis, hypoxia,
next day two dilution
doses at 1-h coagulopathy, all
intervals these reduced the
244
266
followed further efficacy of rFVIIa in
doses every 3 h vivo.
2004 58 2 Uterine atony, (case 1) (case 1) ligation of (case 1) RBC (19); (case 1) Last resort (case 1) 60 Bleeding (case 1) 4 weeks
shock; severe CS hypogastric arteries FFP (3350 ml); 200 ml/h (case 2) 60 stopped later developed
coagulopathy (case 2) (case 2) laparotomy; PPTs (900 ml); (case 2) (rapidly) thrombosis of both
CS ligation of fibrinogen (3 g); 2000 ml, ovarian veins
hypogastric arteries [aprotynin] hemo-
(case 2) RBC (22); peritoneum
FFP (3400 ml);
PPTs (3400 ml);

PPTs (300 ml);
fibrinogen (2 g);
[aprotynin]
2004 59 1 Placenta previa; CS No RBC (11); FFP (4); 1000 (in the 12 mg Bleeding
accreta; DIC CRYO (6) drain) 5 h stopped (few
after CS hours)
2004 60 1 Glanzmann’s VD No PPTs (4) No evidence Prophylactic 36 (2) 800 ml rFVIIa may offer
thrombasthenia of bleeding 1st during (intra- and an alternative option
vaginal postpartum in patients with
delivery, 2nd blood loss) Glanzmann’s
2 h after thrombasthenia
delivery during delivery
2004 61 1 AFE; DIC CS No RBC (6); FFP (1); 3000 90 Hemostasis
(developed 2 min PPTs (2) was secured
after delivery) within 30 min
30 August 2006 14:21:30

2004 62 3 (case 1) Eclampsia; (case 1) No (case 1) RBC (16); (case 1) 2500 last resort (case 1) Bleeding (case 1) patient
HELLP; CS FFP (14); PPTs (case 2) 3000 90 (2) controlled developed anuric
consumptive (case 2) (18); CRYO (10); (case 3) 1300 (case 2) 120 renal failure; cardiac
coagulopathy; CS (case 2) RBC (8); (3); 90 (2) 2-h arrest; patient died;
subcapsular liver (case 3) FFP (4); PPTs (6) intervals no evidence of
hematoma with CS (case 3) RBC (2); (case 3) 90 (2) systemic thrombosis
capsule rupture FFP (4); PPTs (6); 2-h interval identified
(case 2) placenta CRYO (10) (case 2) no future

percreta, transfusion
pre-eclampsia; requirement;
HELLP coagulation profile
(case 3) stabilized
pre-eclampsia;
HELLP; placenta
accreta;
consumptive
coagulopathy;
severe vaginal
245
267
bleeding and uterine
cramping
2004 63 1 Pre-eclampsia; eCS Laparotomy 12 h RBC (22); FFP 3500 in Post laparotomy 90 Bleeding
HELLP; DIC; after CS, because (18); PPTs (30); abdominal reduced (30),
shock intra-abdominal CRYO (20); cavity and 600 Bleeding
hemorrhage (aprotynin) postoperatively stopped (180)
from drains
2005 64 3 (case 1) Uterine (case 1) (case 1) (case 1) RBC (7); NA (case 1) before (case 1) (case 1) Improve coagulation
atony, shock CS relaparotomy with FFP (9); relaparotomy 120 (2) Bleeding parameters
(case 2) placenta (case 2) intracavitary (case 2) RBC (10); (cases 2, 3) 1-h interval stopped

previa, uterine VD oxytocin injected FFP (13); PPTs (2) last-resort (case 2) 60 (case 2)
atony (case 3) (case 3) into the uterus; (case 3) RBC (13); (2) within 3 h Bleeding
laceration of IVD ligature of both FFP (16); PPTs (2) (case 3) stopped
vagina, atony, uterine arteries; 120 (2) (case 3)
consumptive placement of Bleeding
coagulopathy B-Lynch sutures stopped
continued
Table 6 Continued
30 August 2006 14:21:30


2005 65 1 Uterine atony; CS HYS; packing of NA NA Before relaparotomy 2.4 mg Bleeding Resolution of the
shock; DIC the pelvis and ligation controlled coagulopathy
hypogastric artery (rapid
response)
2005 66 4 Uterine atony VD Uterus and vagina NA (case 1) before developed (case 1) 82 (case 1) Lower than
tamponade 1600 severe (case 2) 73 Bleeding standard doses may
(case 2) coagulopathy, (case 3) 61 stopped (15) be effective when

2400 surgical procedures; (case 4) 72 (case 2) respect good timing,
(case 3) avoided massive Bleeding before complication
1100 transfusion stopped (25) develops
(case 4) (case 3)
2500 Bleeding
stopped (35)
246
(case 4)
268
Bleeding
stopped (40)
2005 67 3 (case 1) dehiscence (case 1) (case 1) 3 (case 1) WB (12); (case 1) (case 1) 90 (case 1)
of uterine scar eCS laparotomy; FFP (17); PPTs (2); 225 ml/h (case 2) 90 Bleeding
(case 2i) placenta (case 2) bilateral internal (case 2) WB (11); (case 2) 600 (case 3) 80 controlled (16)
percreta, adherent; VD iliac ligation FFP (7) within 40 min (case 2)
dehiscence of (case (case 2) subtotal (case 3) 500 Bleeding
uterine scar 3)ieCS HYS hematoma stopped (14)
(previous CS) (case 3)

(case 3) NA Bleeding
stopped
RBC, red blood cell concentrates; FFP, fresh frozen plasma; PPTs, platelets; CRYO, cryoprecipitates; WB, whole blood; PCC, prothrombin complex
concentrate; vWF, von Willebrand factor; CS, Cesarean section (e, emergency); VD, vaginal delivery; IVD, instrumental vaginal delivery; DIC, dissemi-
nated intravascular coagulation; MOF, multiple organ failure; NA, not available; HYS, hysterectomy; laceration – uterine or vaginal; AFE, amniotic fluid
embolism; HELLP, hemolysis, elevated liver enzymes, low platelets; ‘last resort’, therapy, after other clinical measures had failed; n, number of cases
Table 7 Patients with severe postpartum hemorrhage, presented by Ahonen and colleagues (2005)68. The authors concluded that treatment with rFVIIa
may be of benefit in life-threatening postpartum hemorrhage of up to 20 l of blood in 5–8 h. For comments on this article, see reference 69
30 August 2006 14:21:31

Type of Additional surgeries Blood products given Blood loss before Timing (when rFVIIa Dose of rFVIIa (µg/kg) Overall bleeding response
Case Provocation of bleed delivery (number of surgery) pre-rFVIIa (units) rFVIIa (l) administered) (number of doses) to rFVIIa (min)
1 Placenta accreta VD HYS RBC (42); FFP (25); 25.0 After HYS 44 Partial
PPTs (40)
2 Adherent placenta CS HYS RBC (35); FFP (14); 20.0 After HYS 95 Good
PPTs (24)
3 Uterine atony, LAC VD Surgery (3) RBC (19); FFP (8); 11.0 Before HYS 78 Good
PPTs (8)
4 Laceration VD Surgery (2), RBC (25); FFP (16); 14.0 NA 103 Partial
embolization PPTs (24)
5 Laceration CS HYS (3 laparotomy) RBC (32); FFP (20); 19.0 After HYS 90 Good
PPTs (40)
6 Uterine atony CS Surgery, embolization RBC (10); FFP (8); 5.5 NA 116 Partial
PPTs (16)
247
269
7 Placenta accreta VD HYS RBC (14); FFP (6); 7.5 After HYS 42 Partial
PPTs (4)
8 Laceration CS Surgery (2), right RBC (11); FFP (4); 5.3 NA 120 None
uterine artery ligation PPTs (8)
9 Placenta percreta CS HYS (2) RBC (25); FFP (14); 14.0 After HYS 77 Good
PPTs (16)
10 Laceration IVD Surgery, embolization RBC (12); FFP (10); 8.8 NA 74 Partial
PPTs (32)
11 Laceration VD Surgery RBC (11); FFP (6); 5.5 NA 86 Good

PPTs (6)
12 Laceration VD Surgery, embolization RBC (10); FFP (8); 5.8 NA 96 Partial
PPTs (16)
RBC, red blood cell concentrates; FFP, fresh frozen plasma; PPTs, platelets; CS, Cesarean section (e, emergency); VD, vaginal delivery; IVD, instrumen-
tal vaginal delivery; DIC, disseminated intravascular coagulation; MOF, multiple organ failure; NA, not available; HYS, hysterectomy; laceration – uterine
or vaginal; AFE, amniotic fluid embolism; HELLP, hemolysis, elevated liver enzymes, low platelets; ‘last resort’, therapy, after other clinical measures had
failed
30 August 2006 14:21:31
Table 8 Patients with severe postpartum hemorrhage presented by Segal and colleagues70,71
Blood loss Timing Dose of rFVIIa Overall bleeding
Type of Blood products given before (when rFVIIa (µg/kg) response to
Case Provocation of bleed delivery Additional surgeries pre-rFVIIa (units) rFVIIa (l) administered) (number of doses) rFVIIa (min)
1 Placenta accreta CS HYS; ligation of internal iliac RBC (44); FFP (24); NA NA 90 (2) Bleeding stopped
arteries; packing PPTs (60); CRYO (54)

2 Uterine rupture VD HYS; ligation of internal iliac RBC (20); FFP (16); NA NA 100 (2) Bleeding stopped
arteries PPTs (60); CRYO (60)
3 Uterine atony CS CS, packing of uterus; laparotomy; RBC (19); FFP (8); NA NA 90 (2) Bleeding reduced
packing of tears on liver PPTs (8)
4 Uterine atony NA Subtotal HYS; ligation of internal RBC (14); FFP (12); NA NA 90 (2) Bleeding stopped
iliac arteries PPTs (10); CRYO (10)
5 Uterine rupture NA Ligation of internal iliac arteries; RBC (26); FFP (16); NA NA 90 (2) Bleeding stopped
subtotal HYS PPTs (30); CRYO (60)
248
270
6 Placenta accreta NA Arterial embolization; HYS; iliac RBC (100); FFP (50); NA NA 90 (2) Bleeding
ligation; 4 laparotomies; packing PPTs (50); CRYO (50) controlled
7 Uterine rupture NA HYS; ligation of internal iliac RBC (10); FFP (6); NA NA 90 (2) Bleeding reduced
arteries; packing CRYO (4)
8 Uterus myomatosus, NA HYS RBC (6); FFP (9) NA NA 60 (2) No bleeding
menorrhagia
9 Uterine rupture NA HYS RBC (15); FFP (6); PPTs NA NA 90 (2) Bleeding stopped
(15); CRYO (30)

10 Placenta accreta NA HYS; ligation of internal iliac RBC (27); FFP (30); NA NA 90 (2) Bleeding stopped
arteries; aortic clamp PPTs (10); CRYO (30)
RBC, red blood cell concentrates; FFP, fresh frozen plasma; PPTs, platelets; CRYO, cryoprecipitates; CS, Cesarean section; VD, vaginal delivery; NA, not
available; HYS, hysterectomy
individuals with HELLP syndrome and others PROPOSAL OF RECOMMENDATION

with both laboratory and clinical signs of DIC FOR THE USE OF rFVIIa IN SEVERE
before rFVIIa was administered. However, POSTPARTUM HEMORRHAGE
none of these patients developed an objectively
Based on our own experience and data from
confirmed, clinically manifest thrombo-
the literature36,72–80, we have prepared guide-
embolism (deep vein thrombosis, pulmonary
lines for treatment of postpartum hemorrhage
embolism, myocardial infarction, cerebro-
that include administration of rFVIIa.
vascular embolism) after rFVIIa therapy, even if
some patients had pre-existing signs of DIC
(often severe). Definitions of severe hemorrhage
Patients with HELLP syndrome who develop
DIC are recognized as being at particular risk (1) Loss of entire blood volume within 24 h;
for life-threatening complications. The HELLP (2) Loss of 50% of blood volume within 3 h;
syndrome is a form of severe pre-eclampsia, and
may be confused with the development of DIC. (3) Blood loss at a rate of 150 ml/min (for
Data presented in Tables 6–8 suggest a high 20 min > 50% blood volume);
efficacy and safety profile of rFVIIa in the treat- (4) Blood loss at a rate of 1.5 ml/kg/min
ment of HELLP syndrome and/or DIC with for ≥ 20 min;
massive bleeding. These findings are supported
by clinical experiences about the therapeutic (5) Sudden blood loss > 1500–2000 ml (uter-
effectivity of rFVIIa in three patients with ine atony; 25–35% blood volume).
massive obstetric hemorrhage due to placenta
previa, accreta, rupture of the uterus and
Definition of insufficient standard
pre-eclampsia with HELLP recently published
management
by an Israeli group71. As mentioned by Segal
and colleagues, these results raise the possibility The hemorrhage continues despite:
that rFVIIa may be administered in obstetric
(1) All standard pharmacological and surgical
cases with life-threatening bleeding episodes,
treatment methods have been used;
even in the presence of DIC-like coagulopathy.
Injection of rFVIIa should be also considered (2) Replacement therapy was performed;
before hysterectomy in a young patient with
(3) Coagulopathy was confirmed by laboratory
severe bleeding, or after internal iliac artery
testing
ligation, if bleeding continues.
The series of patients reported here provides (a) PT or APTT > 1.5 × times the control
data on the safety and efficacy of rFVIIa in value
intractable early postpartum hemorrhage. How-
(b) Thrombocytopenia < 50 × 109/l
ever, as with any case series, there are difficulties
in data analysis because data were collected ret- (c) Fibrinogen < 0.6–0.8 g/l.
rospectively after the bleeding episode had
occurred.
Preconditions for rFVIIa administration
(1) Hematological parameters
Safety Hemoglobin levels > 70 g/l (4.3 mmol/l)
International normalized ratio (INR) < 1.5
Adverse thromboembolic events were reported
Fibrinogen levels ≥ 1 g/l
in one case that was considered to be directly
Platelets levels ≥ 50 × 109/l
related to the use of rFVIIa54. In general, rFVIIa
administration was associated with an excellent (2) pH correction (≥ 7.2) (suggest using
safety profile. NaHCO3)
249
271
30 August 2006 14:21:31
(3) Body temperature should be restored if pos- especially in situations of coexisting coagulo-
sible to physiological values: rFVIIa retains pathy, we therefore recommend administration
its activity in the presence of hypothermia of rFVIIa as soon as possible under the follow-
ing circumstances:
Correction of the pH to ≥ 7.2 is recommended
before rFVIIa administration (efficacy of rFVIIa (1) When no blood is available;
decreases at a pH ≤ 7.1). We also suggest using
(2) Before metabolic complications develop;
bicarbonate to elevate the serum pH. It should
be noted that NaHCO3 has not been shown to (3) In women refusing transfusions (e.g.
provide benefits to patients in hemorrhagic shock. Jehovah Witnesses) (see Chapter 15);
(4) In acquired hemophilia (see Chapter 25);
Recommended replacement therapy
(5) Before the symptoms of severe thrombo-
(1) Fresh frozen plasma: 5–10 ml/kg (4–5 units); cytopathies, hypoxia and organ injury
appear;
(2) Cryoprecipitates: 1–1.5 units/10 kg (8–10
units); (6) If correction of INR (PT) is urgently
needed;
(3) Platelets: 1 units/10 kg (5–8 units);
(7) Before packing of the uterus or pelvis;
(4) Correction of acidosis (defined as pH ≥ 7.2);
(8) Before surgical procedures such as
(5) Warming of hypothermic patients (recom-
hysterectomy, laparotomy;
mended, but not mandatory for administra-
tion of rFVIIa). (9) Before medical procedures such as
embolization, ligation of the uterine and
internal iliac arteries (see Chapter 32).
Dosing administration protocol proposal
Information obtained from the literature
(1) The recommended initial dose of rFVIIa
allows us to summarize the advantages and
for treatment of severe postpartum
disadvantages of rFVIIa as follows.
hemorrhage is ~40–60 µg/kg administered
intravenously.
(2) If bleeding still continuous beyond Advantages
15–30 min, following the first dose of (1) Recombinant product;
rFVIIa, an additional dose of ~40–60 µg/kg
should be considered. Repeat 3–4 times (2) Not subject to blood shortage;
at 15–30-min intervals if clinical signs of (3) No viral transmission;
bleeding are still present (based on visual
evidence). (4) No human protein;
(3) If the response remains inadequate follow- (5) Localized hemostasis;

ing a total dose of > 200 µg/kg, the precon- (6) Low risk of anaphylaxis;
ditions for rFVIIa administration should
be re-checked, and corrected as necessary (7) No anamnestic responses;
before another dose is considered. (8) Low thrombogenicity;
(4) Only after these corrective measures have (9) Effective during and after surgery.
been applied should the next dose of rFVIIa
~100 µg/kg be administered.
Disadvantages
Recommended timing of administration (1) Short t1/2 requires frequent, repetitive
dosing;
Because our experience suggests that rFVIIa
permits effective control of obstetric bleeding, (2) Not 100% effective;
250
272
06 September 2006 16:31:46
(3) No measurable lab parameter for efficacy; 11. Dutton RP, Hess JR, Scalea TM. Recombinant
factor VIIa for control of haemorrhage: early
(4) Limited vial sizes; experience in critically ill trauma patients. J Clin
(5) Venous access; Anesth 2003;15:184–8
12. Mariani G, Testa MG, Di Paolantonio T,
(6) Cost. Molskov Bech R, Hedner U. Use of recombi-
nant, activated factor VII in the treatment of
congenital factor VII deficiencies. Vox Sang
References 1999;77:131-6
1. Luster JM, Roberts HR, Davignon G, et al. 13. Hunault M, Bauer KA. Recombinant factor
A randomized, double-blind comparison of two VIIa for the treatment of congenital factor VII
doses of recombinant factor VIIa in the treat- deficiency. Semin Thromb Hemost 2000;26:401–5
ment of joint, muscle and mucocutaneous haem- 14. d’Oiron R, Menart C, Trzeciak MC, et al. Use
orrhages in persons with haemophilia A and B, of recombinant factor VIIa in 3 patients with
with and without inhibitors. rFVIIa Study inherited type I Glanzmann’s thrombasthenia
Group. Haemophilia 1998;4:790–8 undergoing invasive procedures. Thromb
2. Shapiro AD, Gilchrist GS, Hoots WK, Cooper Haemost 2000;83:644–7
HA, Gastineau DA. Prospective, randomized 15. van Buuren HR, Wielenga JJ. Successful surgery
trial of two doses of rFVIIa (NovoSeven®) using recombinant factor VIIa for recurrent idio-
in haemophilia patients undergoing surgery. pathic nonulcer duodenal bleeding in a patient
Thromb Haemost 1998;80:773–8 with Glanzmann’s thrombasthenia. Dig Dis Sci
3. Abshire T, Kenet G. Recombinant factor VIIa: 2002;47:2134–6
review of efficacy, dosing regimens and safety 16. Hedner U. Recombinant factor VIIa
in patients with congenital and acquired factor (NovoSeven) as a haemostatic agent. Bloodline
VIII or IX inhibitors. J Thromb Haemost 2004;2: Rev 2001;1:3–4
899–909 17. Dildy GA 3rd. Postpartum hemorrhage: new
4. O’Connell KA, Wood JJ, Wise RP, Lozier JN, management options. Clin Obstet Gynecol
Braun MM. Thromboembolic adverse events 2002;45:330–44
after use of recombinant human coagulation 18. Yamamoto H, Sagae S, Nishikawa S, Kudo R.
factor VIIa. JAMA 2006;295:293–8 Emergency postpartum hysterectomy in
5. Michalska-Krzanowska G, Sajdak R, Stasiak- obstetric practice. J Obstet Gynaecol Res 2000;
Pikula E. Effects of recombinant factor VIIa 26:341–5
in haemorrhagic complications of urological 19. Sakai T, Lund-Hansen T, Thim L, Kisiel W.
operations. Acta Haematol 2003;109:158–60 The gamma-carboxyglutamic acid domain of
6. Naik VN, Mazer DC, Latter DA, Teitel JM, human factor VIIa is essential for its interaction
Hare GMT. Successful treatment using recom- with cell surface tissue factor. J Biol Chem
binant factor VIIa for severe bleeding post 1990;265:1890–94
cardiopulmonary bypass. Can J Anesth 2003;50: 20. NovoSeven® summary of product characteris-
599–602 tics. Available at: http://www.novoseven.com/
7. Danilos J, Goral A, Paluszkiewicz P, Przesmycki content/product_information/summary_of_
K, Kotarski J. Successful treatment with recom- product_characteristics/product_information_
binant factor VIIa for intractable bleeding at spc.asp
pelvic surgery. Obstet Gynecol 2003;101:1172–3 21. Lund-Hansen T, Petersen LC. Comparison of
8. Martinowitz U, Kenet G, Lubetski A, Luboshitz enzymatic properties of human plasma FVIIa
J, Segal E. Possible role of recombinant activated and human recombinant FVIIa. Thromb Haemost
factor VII (rFVIIa) in the control of haemorrhage 1987;58:270
associated with massive trauma. Can J Anaesth 22. Jurlander B, Thim L, Klausen NK, et al. Recom-
2002;49:S15–20 binant factor VII (rFVIIa): characterization,
9. Martinowitz U, Kenet G, Segal E, et al. manufacturing, and clinical development. Semin
Recombinant activated factor VII for adjunctive Thromb Hemost 2001;27:373–84
haemorrhage control in trauma. J Trauma 2001; 23. Erhardtsen E. Pharmacokinetics of recombinant
51:431–9 activated factor VII (rFVIIa). Semin Thromb
10. Kenet G, Walde R, Eldad A, Martinowitz U. Hemost 2000;26:385–91
Treatment of traumatic bleeding with recombi- 24. Lindley CM, Sawyer WT, Macik BG, et al.
nant factor VIIa. Lancet 1999;354:1879 Pharmacokinetics and pharmacodynamics of
251
273
30 August 2006 14:21:31
recombinant factor VIIa. Clin Pharmacol Ther VIIa versus factor VIII inhibitor bypass activity.
1994;55:638–48 J Thromb Haemost 2004;2:1700–8
25. Monroe DM, Hoffman M, Oliver JA, et al. 39. Breborowicz GH, Sobieszczyk S [Usefulness
Platelet activity of high-dose Factor VIIa is of recombinant active factor VIIa (rFVIIa,
independent of tissue factor. Br J Haematol NovoSeven®) in obstetric practice – own
1997;99:542–7 experiences]. Przydatnosd rekombinowanego
26. Hoffman M, Monroe DM, Roberts HR. Acti- aktywnego czynnika VIIa (rFVIIa, NovoSeven®)
vated Factor VII activates Factor IX and X on w praktyce polozniczej – doswiadczenia wlasne.
surface of activated platelets: thoughts on the Klin Perinat Ginek 2001;34:7–12
mechanism of action of high-dose activated 40. Breborowicz GH, Sobieszczyk S, Szymankiewicz
Factor VII. Blood Coagul Fibrinolysis 1998;9: M. Efficacy of recombinant activated factor VII
S61–5 (rFVIIa, NovoSeven®) in prenatal medicine.
27. van’t Veer C, Golden NJ, Mann KG. Inhibition Arch Perinat Med 2002;8:21–7
of thrombin generation by the zymogen factor 41. Breborowicz GH, Sobieszczyk S, Szymankiewicz
VII: implications for the treatment of hemophilia M. Efficacy of recombinant activated factor VII
A by factor VIIa. Blood 2000;95:1330–5 (rFVIIa, NovoSeven®) in prenatal medicine.
28. Lisman T, De Groot PG. Mechanism of action J Perinat Med 2003;31(Suppl 1):18
of recombinant factor VIIa. J Thromb Haemost 42. Breborowicz GH, Sobieszczyk S, Szymankiewicz
2003;1:1138–9 M. Recombinant factor VIIa in the management
29. Hoffman M, Monroe DM. A cell-based model of of major postpartum hemorrhage. Arch Perinat
haemostasis. Thromb Haemost 2001;85:958–65 Med 2004;10:17–19
30. Hedner U. Recombinant factor VIIa 43. Sobieszczyk S, Breborowicz GH. Management
(Novoseven) as a hemostatic agent. Semin recommendations for postpartum hemorrhage.
Hematol 2001;38:43–7 Arch Perinat Med 2004;10:53–6
31. Friederich PW, Levi M, Bauer KA, et al. Ability 44. Sobieszczyk S, Breborowicz GH, Kubiaczyk B,
of recombinant factor VIIa to generate thrombin Opala T. Efficacy of recombinant activated
during inhibition of tissue factor in human factor VII (r FVIIa; NovoSeven®) in obstetrical
subjects. Circulation 2001;103:2555–9 haemorrhagic shock. Crit Care 2003;7(Suppl 2):
32. Monroe DM, Hoffman M, Allen GA, Roberts S52
HR. The factor VII–platelet interplay: effective- 45. Sobieszczyk S, Skrzypczak J, Szymankiewicz M,
ness of recombinant factor VIIa in the treatment Kruszyñski Z, Kornacki J, Breborowicz GH.
of bleeding in severe thrombocytopathia. Semin Zastosowanie rekombinowanego aktywnego
Thromb Hemost 2000;26:373–7 czynnika VII (rFVIIa, NovoSeven®) podczas
33. Telgt DS, Macik BG, McCord DM, Monroe ciecia cesarskiego u ciezarnej z mechaniczna
DM, Roberts HR. Mechanism by which recom- zastawka serca. [Application of recombinant
binant factor VIIa shortens the aPTT: activation activated factor VII (rFVIIa, NovoSeven®) dur-
of factor X in the absence of tissue factor. Thromb ing cesarean section of a woman with artificial
Res 1989;56:603–9 heart valve]. Klin Perinat Ginek 2001;34:173–9
34. Kessler CM. Antidotes to haemorrhage: recom- 46. Sobieszczyk S, Breborowicz GH, Markwitz W,
binant Factor VIIa. Best Pract Res Clin Haematol Mallinger S, Adamski D, Kruszynski Z. Effect
2004;17:183–97 of recombinant activated factor VII (rFVIIa;
35. Gabriel DA, Carr M, Roberts HR. Monitoring NovoSeven®) in a patient in haemorrhagic shock
coagulation and the clinical effects of recombi- after obstetrical hysterectomy. Ginekol Pol 2002;
nant factor VIIa. Semin Hematol 2004;41:20–4 73:230–3
36. Martinowitz U, Michaelson M, on behalf of the 47. Moscardó F, Pérez F, de la Rubia J, et al.
Israeli Multidisciplinary rFVIIa Task Force. Successful treatment of severe, intra-abdominal
Guidelines for the use of recombinant activated bleeding associated with disseminated intra-
factor VII (rFVIIa) in uncontrolled bleeding: vascular coagulation using recombinant acti-
a report by the Israeli Multidisciplinary rFVIIa vated factor VII. Br J Haematol 2001;113:174–6
Task Force. J Thromb Haemost 2005;3:640–8 48. Eskandari N, Feldman N, Greenspoon JS.
37. Roberts HR, Monroe DM III, Hoffman M. Factor VII deficiency in pregnancy treated
Safety profile of recombinant factor VIIa. Semin with recombinant factor VIIa. Obstet Gynecol
Hematol 2004;41:101–8 2002;99:935–7
38. Aledort LM. Comparative thrombotic event 49. Ciadma A, Langie T, Zajac K, Fabian W.
incidences after infusion of recombinant factor Acquired haemophilia in parturient. Case report.
252
274
30 August 2006 14:21:32
Anaesthesiology Intensive Therapy 2002;34: 60. Kale A, Bayhan G, Yalinkaya A, Yayla M. The
269–70 use of recombinant factor VIIa in a primigravida
50. Zupancid Šalek S, Sokolid V, Viskovid T, Šanjug with Glanzmann’s thrombasthenia during deliv-
J, Šimid M, Kaštelan M. Successful use of ery. J Perinat Med 2004;32:456–8
recombinant factor VIIa for massive bleeding 61. Lim Y, Loo CC, Chia V, Fun W. Recombinant
after caesarean section due to HELLP syn- factor VIIa after amniotic fluid embolism and
drome. Acta Haematol 2002;108:162–3 disseminated intravascular coagulopathy. Int J
51. Loo CC, Kwek Bielanom, Tan HM, Yeo SH, Gynaecol Obstet 2004;87:178–9
Tien SL, Loh Bielanom. Successful treatment 62. Merchant-Shakil H, Prasad M, Vanderjagt TJ,
of postpartum hemorrhage with recombinant Howdieshell TR, Crookston P. Recombinant
activated coagulation factor VII (NovoSeven®). factor VIIa in management of spontaneous
7th Novo Nordisk Symposium on Haemostasis subcapsular liver hematoma associated with
Management. Clinical and Scientific Posters. pregnancy. Obstet Gynecol 2004;103:1055–8
2003; May:23 63. Price G, Kaplan J, Skowronski G. Use of
52. Bielanow T, Sidor M, Maciejewski M, Skrzypek recombinant factor VIIa to treat life-threatening
W. [Effectiveness of recombinant activated fac- non-surgical bleeding in a post-partum patient.
tor VIIA (NovoSeven) in case of severe obstetric Br J Anaesth 2004;93:298–300
complication with coagulopathy]. Ginekol Pol 64. Hollnberger H, Gruber E, Seelbach-Goebel B.
2003;74:1055–9 Major post-partum hemorrhage and treatment
53. Bouwmeester FW, Jonkhoff AR, Verheijen RH, with recombinant factor VIIa. Anesth Analg
van Geijn HP. Successful treatment of life- 2005;101:1886–7
threatening postpartum hemorrhage with 65. Holub Z, Feyereisi J, Kabelik L, Rittstein T.
recombinant activated factor VII. Obstet Gynecol Successful treatment of severe post-partum
2003;101:1174–6 bleeding after caesarean section using recombi-
54. Michalska- Krzanowska G, Stasiak-Pikula E, nant activated factor VII. Ceska Gynekol 2005;
Sajdak R. Recombinant activated factor VII: A 70:144–8
new treatment for obstetric haemorrhage? Anest 66. Tanchev S, Platikanov V, Karadimov D. Admin-
Intens Ther 2003;35:110–12 istration of recombinant factor VIIa for the
55. Pehlivanov B, Milchev N, Kroumov G. Factor management of massive bleeding due to uterine
VII deficiency and treatment in delivery with atonia in the post-placental period. Acta Obstet
recombinant factor VII. Eur J Obstet Gynecol Gynecol Scand 2005;84:402–3
Reprod Biol 2004;116:237–8 67. Shamsi TS, Hossain N, Soomro N, et al. Use of
56. Kretzschmar M, Zahm DM, Remmler K, recombinant Factor VIIa for massive haemor-
Pfeiffer L, Victor L, Schirrmeister W. Patho- rhage: Case series and review of literature. J Pak
physiological and therapeutic aspects of amniotic Med Assoc 2005;55:512–15
fluid embolism (anaphylactoid syndrome of 68. Ahonen J, Jokela R. Recombinant factor VIIa for
pregnancy): Case report with lethal outcome and life-threatening post-partum haemorrhage. Br J
overview. Anaesthesist 2003;52:419–26 Anaesth 2005;94:592–5
57. Boehlen F, Morales MA, Fontana P, Ricou B, 69. Butwick AJ, Riley ET, Ahonen J, Jokela R.
Irion O, de Moerloose P. Prolonged treatment Recombinant factor VIIa for life-threatening
of massive postpartum haemorrhage with recom- post-partum haemorrhage. Br J Anaesth 2005;
binant factor VIIa: case report and review of the 95:558
literature. Br J Obstet Gynaecol 2004;111:284–7 70. Segal S, Shemesh IY, Blumental R, et al. The use
58. Brice A, Hilbert U, Roger-Christoph S, of recombinant factor VIIa in severe postpartum
et al. [Recombinant activated factor VII as a hemorrhage. Acta Obstet Gynecol Scand 2004;83:
life-saving therapy for severe postpartum haem- 771–2
orrhage unresponsive to conservative traditional 71. Segal S, Shemesh IY, Blumenthal R, et al. Treat-
management]. Ann Fr Anesth Reanim 2004;23: ment of obstetric hemorrhage with recombinant
1084–8 activated factor VII (rFVIIa). Arch Gynecol Obstet
59. Gidiri M, Noble W, Rafique Z, Patil K, Lindow 2003;268:266–7
SW. Caesarean section for placenta praevia com- 72. Hedner U. Dosing with recombinant factor
plicated by postpartum haemorrhage managed VIIa based on current evidence. Semin Hematol
successfully with recombinant activated human 2004;41(Suppl 1):35–9
coagulation Factor VIIa. J Obstet Gynaecol 2004; 73. Monroe DM, Roberts HR. Mechanism of action
24:925–6 of high-dose factor VIIa: points of agreement and
253
275
30 August 2006 14:21:32
disagreement. Arterioscler Thromb Vasc Biol 2003; 78. Meng ZH, Wolberg AS, Monroe DM III,
23:8–9; discussion 10 Hoffman M. The effect of temperature and
74. Poon MC, D’Oiron R, Von Depka M, et al. pH on the activity of factor VIIa: implications
Prophylactic and therapeutic recombinant factor for the efficacy of high-dose factor VIIa in
VIIa administration to patients with Glanzmann’s hypothermic and acidotic patients. J Trauma
thrombasthenia: results of an international 2003;55:886–91
survey. J Thromb Haemost 2004;2:1096–103 79. DeLoughery TG. Management of bleeding
75. Mayer SA. Ultra-early hemostatic therapy for emergencies: when to use recombinant activated
intracerebral hemorrhage. Stroke 2003;34:224–9 Factor VII. Expert Opin Pharmacother 2006;7:
76. Dutton RPHJ, Scalea TM. Recombinant factor 25–34
VIIa for the control of hemorrhage: early experi- 80. Friederich P, Heny C, Messelink E, et al.
ence in critically ill trauma patients. Can J Effects of recombinant activated factor VII on
Anaesth 2003;5:184–8 perioperative blood loss in patients undergoing
77. Aldouri M. The use of recombinant factor retropubic prostatectomy: a double blind
VIIa in controlling surgical bleeding in non- placebo controlled trial. Lancet 2003;361:
haemophiliac patients. Pathophysiol Haemost 201–5
Thromb 2002;32(Suppl 1):41–6
254
276
30 August 2006 14:21:32
Section VII
Therapy for atony
277
30 August 2006 14:21:32
27
STANDARD MEDICAL THERAPY
F. Breathnach and M. Geary
INTRODUCTION successful until the early 20th century, when

Barber and Dale isolated ergotoxine in 19062.
Failure of the uterus to contract and retract
Initially thought to be a pure substance, this
following childbirth has for centuries been
agent was subsequently found to comprise four
recognized as the most striking cause of post-
alkaloids and in 1935 Moir and Dudley were
partum hemorrhage. Uterine atony is a condi-
credited for isolating ergometrine, the active
tion which, in spite of the presence of effective
aqueous extract ‘to which ergot rightly owes
medical interventions, still claims thousands of
its long-established reputation as the pulvis
maternal lives. In the developing world, lack of
parturiens’5,6. Moir reported on its clinical use
access to uterotonic therapies that have been
in 1936, stating6:
available for almost a century represents one of
the most glaring disparities in obstetric care ‘. . . the chief use of ergometrine is in the preven-
today. tion and treatment of postpartum haemorrhage.
Here the ergometrine effect is seen at its best. If
In the 19th century, uterine atony was
after the delivery of the placenta the uterus is
treated by intrauterine placement of various
unduly relaxed, the administration of ergo-
agents with the aim of achieving a tamponade metrine, 1 mg by mouth or 0.5 mg by injection,
effect. ‘A lemon imperfectly quartered’ or ‘a will quickly cause a firm contraction of the organ.
large bull’s bladder distended with water’ were If severe haemorrhage has already set in, it is
employed for this purpose, with apparent suc- highly recommended that the drug should be
cess. Douching with vinegar or iron perchloride given by the intravenous route. For this purpose
was also reported1,2. Historically, the first utero- one-third of the standard size ampoule may be
tonic drugs were ergot alkaloids, followed by injected or, for those who wish accurate dosage,
oxytocin and, finally, prostaglandins. a special ampoule containing 0.125 mg is manu-
Ergot, the alkaloid-containing product of the factured. An effect may be looked for in less than
one minute.’
fungus Claviceps purpurea that grows on rye, was
recognized for centuries as having uterotonic Oxytocin, the hypothalamic polypeptide hor-
properties and is the substance referred to by mone released by the posterior pituitary, was
John Stearns in 1808 as ‘pulvis parturiens’ (a discovered in 1909 by Sir Henry Dale7 and syn-
powder [for] childbirth), at which time it was thesized in 1954 by du Vigneaud8. The develop-
used as an agent to accelerate labor3. By the end ment of oxytocin constituted the first synthesis
of the 19th century, however, recognition of the of a polypeptide hormone and gained du
potential hazards associated with ergot use in Vigneaud a Nobel prize for his work.
labor, namely its ability to cause uterine hyper- The third group of uterotonics comprises
stimulation and stillbirth, had tempered enthu- the ever-expanding prostaglandin family. The
siasm for its use. Focus was diverted toward prostaglandins were discovered in 1935 by a
its role in preventing and treating postpartum group led by Swedish physiologist Ulf von
hemorrhage at a time when, according to an Euler9 who found that extracts of seminal vesi-
1870 report, maternal mortality in England cles or of human semen were capable of causing
approached one in 20 births4. Attempts to iso- contraction of uterine tissue and lowering blood
late the active alkaloids from ergot were not pressure. The term ‘prostaglandin’ evolved
256
278
30 August 2006 14:21:33
Standard medical therapy
from von Euler’s belief that the active material Table 1 Risk factors for uterine atony
came exclusively from the prostate gland. This
Factors associated with uterine overdistension
family of ‘eicosanoids’, 20-carbon fatty acids, Multiple pregnancy
was subsequently found to be produced in a Polyhydramnios
variety of tissues and capable of mediating a Fetal macrosomia
myriad of physiologic and pathologic processes.
Prostaglandins, by virtue of their ability to cause Labor-related factors
strong myometrial tetanic activity, are increas- Induction of labor
Prolonged labor
ingly being employed as adjunctive therapy
Precipitate labor
to standard oxytocin and ergometrine to treat Oxytocin augmentation
postpartum hemorrhage resulting from uterine Manual removal of placenta
atony (see Chapter 12).
This chapter is devoted to critical evaluation Use of uterine relaxants
of the standard pharmacological methods avail- Deep anesthesia (especially halogenated anesthetic
agents)
able to overcome uterine atony, with particular
Magnesium sulfate
focus on agent selection based on effectiveness,
safety profile, ease of administration, cost and Intrinsic factors
applicability in low-resource settings. Previous postpartum hemorrhage
Antepartum hemorrhage (abruptio or previa)
Obesity
UTERINE ATONY Age > 35 years
Powerful efficient contractions of the myo-
metrium are essential to arrest blood loss after
delivery. The resultant compression of the uter- hemorrhage confers a 2–4-fold increased risk of
ine vasculature serves to halt the 800 ml/min hemorrhage compared to women without such
blood flow in the placental bed. Recognition of a history12,13.
a soft, boggy uterus in the setting of a post- It is appropriate that women with these pre-
partum bleed alerts the attendant to uterine disposing risk factors should deliver in a hospital
atony. The contribution that uterine atony with adequate facilities to manage postpartum
makes toward postpartum hemorrhage is so hemorrhage. Prophylactic measures adopted
well-known that a universal reflex action when include appropriate hospital booking for women
faced with excessive postpartum bleeding is at risk, active management of the third stage of
to massage a uterine contraction. Prompt labor, intravenous access during labor and
recognition of this condition and institution of ensuring the availability of cross-matched
uterotonic therapy will effectively terminate the blood. However, it is noteworthy that uterine
majority of cases of hemorrhage. Once effective atony occurs unpredictably in women with
uterine contractility is assured, persistent bleed- no identifiable predisposing risk factors. This
ing should prompt the search for retained underpins the need for strict protocols for the
placental fragments, genital tract trauma or a management of postpartum hemorrhage to be
bleeding diathesis (see Chapters 9 and 25). in place in every unit that provides obstetric
Astute risk assessment is crucial in identify- care.
ing women at increased risk of uterine atony,
thereby allowing for preventive measures to
OXYTOCIN
be instituted and for delivery to take place
where transfusion and anesthetic facilities are With timely and appropriate use of uterotonic
available. The established risk factors associated therapy, the majority of women with uterine
with uterine atony are outlined in Table 1. It is atony can avoid surgical intervention. Stimula-
worth noting that multiparity, hitherto believed tion of uterine contraction is usually achieved in
to be a significant risk factor, has not emerged the first instance by bimanual uterine massage
as having an association with uterine atony and the injection of oxytocin (either intra-
in recent studies10-12. Previous postpartum muscularly or intravenously), with or without
257
279
30 August 2006 14:21:33
ergometrine. The mode of action of oxytocin of action of 2–5 min. Metabolism is via the
involves stimulation of the upper uterine seg- hepatic route and the mean plasma half-life
ment to contract in a rhythmical fashion. Owing is 30 min. Nonetheless, the clinical effect of
to its short plasma half-life (mean 3 min), a ergometrine persists for approximately 3 h. The
continuous intravenous infusion is required in co-administration of ergometrine and oxytocin
order to maintain the uterus in a contracted therefore results in a complementary effect, with
state14. The usual dose is 20 IU in 500 ml oxytocin achieving an immediate response and
of crystalloid solution, with the dosage rate ergometrine a more sustained action.
adjusted according to response (typical infusion Common side-effects include nausea, vomit-
rate 250 ml/h). When administered intra- ing and dizziness and these are more striking
venously, the onset of action is almost instanta- when given via the intravenous route. As a result
neous and plateau concentration is achieved of its vasoconstrictive effect via stimulation
after 30 min. By contrast, intramuscular admin- of α-adrenergic receptors, hypertension can
istration results in a slower onset of action occur. Contraindications to use of ergometrine
(3–7 min) but a longer lasting clinical effect (up therefore include hypertension (including
to 60 min). pre-eclampsia), heart disease and peripheral
Metabolism of oxytocin is via the renal and vascular disease. If given intravenously, where
hepatic routes. Its antidiuretic effect, which its effect is seen as being almost immediate, it
amounts to 5% of the antidiuretic effect of should be given over 60 s with careful monitor-
vasopressin, can result in water toxicity if given ing of pulse and blood pressure. Relevant to the
in large volumes of electrolyte-free solutions. developing world in particular is its heat lability.
This degree of water overload can manifest itself It is both heat- and light-sensitive and should
with headache, vomiting, drowsiness and con- be stored at temperatures below 8°C and away
vulsions. Furthermore, rapid intravenous bolus from light.
administration of undiluted oxytocin results in The product Syntometrine® (5 units oxy-
relaxation of vascular smooth muscle, which can tocin and 0.5 mg ergometrine) combines the
lead to hypotension. It is therefore best given rapid onset of oxytocin with the prolonged
intramuscularly or by dilute intravenous infu- effect of ergometrine. The mild vasodilatory
sion. Oxytocin is stable at temperatures up property of oxytocin may counterbalance the
to 25°C but refrigeration may prolong its vasopressor effect of ergometrine.
shelf-life. First-line treatment of uterine atony, there-
A disadvantage of oxytocin is its short half- fore, involves administration of oxytocin or
life. The long-acting oxytocin analog carbetocin ergometrine as an intramuscular or diluted
has been studied in this context as its more intravenous bolus, followed by repeat dosage
sustained action, similar to that of ergometrine if no effect is observed after 5 min and comple-
but without its associated side-effects, may offer mented by continuous intravenous oxytocin
advantages over standard oxytocic therapy15. infusion. Atony that is refractory to these
Comparative studies of carbetocin for the first-line oxytocics will warrant prostaglandin
prevention of postpartum hemorrhage have therapy.
identified enhanced effectiveness of this analog
when compared with an oxytocin infusion16,17.
CARBOPROST
Carboprost (15-methyl PGF2α) acts as a
ERGOMETRINE
smooth muscle stimulant and is a recognized
In contrast to oxytocin, the administration of second-line agent for use in the management
ergometrine results in a sustained tonic uterine of postpartum uterine atony unresponsive to
contraction via stimulation of myometrial oxytocin/ergometrine. It is an analog of PGF2α
α-adrenergic receptors. Both upper and lower (dinoprost) with a longer duration of action
uterine segments are thus stimulated to contract than its parent compound, attributed to its
in a tetanic manner14. Intramuscular injection resistance to inactivation by oxidation at the
of the standard 0.25 mg dose results in an onset 15-position. Available in single-dose vials of
258
280
30 August 2006 14:21:33
0.25 mg, it may be administered by deep intra- or sublingual route. The results of an inter-
muscular injection or, alternatively, by direct national multicenter, randomized trial of oral
intramyometrial injection. The latter route of misoprostol as a prophylactic agent for the third
administration is achieved either under direct stage of labor showed it to be less effective
vision at Cesarean section or transabdominally at preventing postpartum hemorrhage than
or transvaginally following vaginal delivery and parenteral oxytocin21. Fifteen percent of women
has the advantage of a significantly quicker in the misoprostol arm required additional
onset of action18,19. Peripheral intramuscular uterotonics compared with 11% in the oxytocin
injection yields peak plasma concentrations at group. This may be due to its longer onset of
15 min in contrast to less than 5 min for the action (20–30 min to achieve peak serum levels
intramyometrial route. Using a 20-gauge spinal compared to 3 min for oxytocin). However,
needle, intravascular injection can be avoided owing to the fact that its more prolonged time
by pre-injection aspiration, and intramyometrial interval required to achieve peak serum levels
rather than intracavitary placement of the may make it a more suitable agent for pro-
needle can be confirmed by observing resistance tracted uterine bleeding, there is mounting
on injection, as described by Bigrigg and interest in its role as a therapeutic rather than a
colleagues20. The dose may be repeated every prophylactic agent.
15 min up to a maximum cumulative dose of The use of rectal misoprostol for the treat-
2 mg (eight doses), although, in reported case ment of postpartum hemorrhage unresponsive
series, the majority of patients require no more to oxytocin and ergometrine was first reported
than one dose. by O’Brien and colleagues22 in a descriptive
Reported efficacy is high. Successful arrest study of 14 patients. Sustained uterine contrac-
of atonic hemorrhage is reported in 13/14 tion was reported in almost all women within
patients by Bigrigg and colleagues20. The largest 3 min of its administration. However, there
case series to date19 involved a multicenter sur- was no control group included for comparison.
veillance study of 237 cases of postpartum hem- A single-blinded, randomized trial of miso-
orrhage refractory to standard oxytocics and prostol 800 µg rectally versus Syntometrine®
reported an efficacy of 88%. The majority of intramuscularly plus oxytocin by intravenous
women in this study required a single dose only. infusion found that misoprostol resulted in
Owing to its vasoconstrictive and broncho- cessation of bleeding within 20 min in 30/32
constrictive effects, carboprost can result in cases (93%) compared to 21/32 (66%) for
nausea, vomiting, diarrhea, pyrexia and the comparative agents23. A Cochrane review
bronchospasm. Contraindications therefore supports these findings, suggesting that rectal
include cardiac and pulmonary disease. The misoprostol in a dose of 800 µg could be a use-
cost of carboprost makes it unsuitable for ful ‘first-line’ drug for the treatment of primary
consideration in low-resource settings. Further- postpartum hemorrhage24.
more, it is both light- and heat-sensitive and A strong need exists for high-dose miso-
must be kept refrigerated at 4°C. prostol to be evaluated in randomized control
trials. As an alternative to the aforementioned
uterotonics, misoprostol has the significant
MISOPROSTOL
advantage of low cost, thermostability, light
Misoprostol is a synthetic analog of prostaglan- stability and lack of requirement for sterile
din E1 which selectively binds to myometrial needles and syringes for administration, making
EP-2/EP-3 prostanoid receptors, thereby pro- it an attractive option for use in the developing
moting uterine contractility. It is metabolized world. It has a shelf-life of several years.
via the hepatic route. It may be given orally, Side-effects of misoprostol are mainly gastro-
sublingually, vaginally, rectally or via direct intestinal and are dose-dependent. A frequently
intrauterine placement. The rectal route of reported side-effect of misoprostol is the occur-
administration is associated with a longer onset rence of shivering and pyrexia. Side-effects are
of action, lower peak levels and a more favorable less marked when the rectal route of administra-
side-effect profile when compared with the oral tion is used.
259
281
30 August 2006 14:21:33
OTHER PROSTAGLANDINS However, to date there is only one case report in

the literature of the use of this agent in the set-
Dinoprost (prostaglandin F2α) has been used
ting of postpartum hemorrhage; that particular
via intramyometrial injection at doses of
case involved a placenta accreta where the
0.5–1.0 mg with good effect25. Low-dose
source of the persistent bleeding was the lower
intrauterine infusion via a Foley catheter has
uterine segment and the uterine body was
also been described, consisting of 20 mg dino-
described as being well contracted31. The dose
prost in 500 ml saline at 3–4 ml/min for 10 min,
employed was 1 g given intravenously 4-hourly
then 1 ml/min. The bleeding was arrested in
to a cumulative dose of 3 g.
all but one of 18 patients and no adverse
The use of recombinant activated factor VII
outcome was reported. As mentioned earlier,
(rFVIIa) as a hemostatic agent for refractory
however, this agent has a shorter duration
postpartum hemorrhage has recently been
of activity than carboprost and indeed has
described in a number of case reports32,33. The
been unavailable in the US since the 1980s
mode of action of this agent involves enhance-
where its withdrawal was attributed to financial
ment of the rate of thrombin generation, leading
reasons.
to formation of a fully stabilized fibrin plug that
Prostaglandin E2 (dinoprostone), in spite
is resistant to premature lysis. Reported cases
of its vasodilatory properties, causes smooth
involve hemorrhage unresponsive to a myriad
muscle contraction in the pregnant uterus, thus
of conventional treatments including hysterec-
making it a potentially suitable uterotonic agent.
tomy and pelvic vessel ligation, where use of this
Its principal indication is in pre-induction
agent was remarkably successful at arresting
cervical priming, but intrauterine placement of
seemingly intractable bleeding within a matter
dinoprostone has been successfully employed
of minutes. Doses of 60–120 µg/kg intra-
as a treatment for uterine atony26. The vaso-
venously were used. A more complete discus-
dilatory effect of dinoprostone, however, ren-
sion of this agent is found in Chapter 26.
ders it unsuitable for use in the hypotensive or
hypovolemic patient. It may, however, be of
use in women with cardiorespiratory disease in CONCLUSIONS
whom carboprost is contraindicated.
The identification of ‘substandard care’ in 71%
Experience with gemeprost, a prostaglandin
of maternal deaths attributed to hemorrhage
E1 analog, in pessary formulation delivered
in the 2000–2002 Confidential Report (UK)34
directly into the uterine cavity or placed in
underscores the need for a standard of care to
the posterior vaginal fornix, is again largely
be established in every unit where childbirth
anecdotal27-29. Its mode of action resembles
takes place and for all relevant health-care work-
that of PGF2α. Rectal administration has
ers to be keenly familiar with that standard (see
also been reported. A retrospective series of
Chapter 22). Integral to any protocol on man-
14 cases in which rectal gemeprost 1 mg was
agement of postpartum hemorrhage will be a
used for postpartum hemorrhage unresponsive
stepwise approach to achieving effective uterine
to oxytocin and ergometrine reported prompt
contractility. The successful management of
cessation of bleeding in all cases, with no
uterine atony will depend on staff being familiar
apparent maternal adverse sequelae30.
with the pharmacologic agents available to them
with respect to dosage, route of administration
and safety profile (Table 2). Application of such
HEMOSTATICS: TRANEXAMIC ACID
protocols has been shown to achieve successful
AND RECOMBINANT ACTIVATED
reduction in the morbidity associated with
FACTOR VII
postpartum hemorrhage35.
The antifibrinolytic agent tranexamic acid, It is tempting to credit the second- or
which prevents binding of plasminogen and third-line agent with successfully controlling a
plasmin to fibrin, may well have a role in the postpartum hemorrhage; however, it is certainly
control of intractable postpartum hemorrhage, plausible that a synergistic effect is observed
particularly where coagulation is compromised. where a combination of uterotonics is used.
260
282
30 August 2006 14:21:33
Table 2 Medical uterotonic therapy
Agent Dose Cautions

®,
Oxytocin (Pitocin 10 IU i.m./i.v. followed by i.v. Hypotension if given by rapid i.v.
Syntocinon®) infusion of 20 IU in 500 ml bolus. Water intoxication with
crystalloid titrated versus response large volumes
(e.g. 250 ml/h)
Ergometrine (Ergonovine®) 0.25 mg i.m./i.v. Contraindicated in hypertensive
patients. Can cause nausea/
vomiting/dizziness
Carboprost (15-methyl PGF2α) 0.25 mg i.m./myometrial. Can be Bronchospasm (caution in patients
(Hemabate®) repeated every 15 min. Max. 2 mg with asthma, hypertension,
cardiorespiratory disease)
Dinoprost (PGF2α) 0.5–1 mg intramyometrial or 20 mg Bronchospasm, nausea, vomiting
(Prostin F2α®) in 500 ml N/saline infused via Foley and diarrhea can occur
catheter into uterine cavity
Dinoprostone (Prostin®/ 2 mg p.r. 2-hourly Hypotension
Prepidil®)
Gemeprost (Cervagem®) 1–2 mg intrauterine placement/ Gastrointestinal disturbance
1 mg p.r.
Misoprostol (Cytotec®) 600–1000 µg p.r./intracavitary Gastrointestinal disturbance,
shivering, pyrexia
Tranexamic acid 1 g 8-hourly i.v. Can increase risk of thrombosis
(Cyclokapron®)
rFVIIa (Novoseven®) 60–120 µg/kg i.v. Fever, hypertension
i.m., intramuscularly; i.v., intravenously; p.r., per rectum
The global quest for an ‘ideal’ uterotonic 2. De Costa C. St Anthony’s fire and living liga-
agent must take into account the fact that what tures: a short history of ergometrine. Lancet
is applicable in one setting may have no rele- 2002;359:1768–70
vance in another. This is particularly true of the 3. Thoms H. John Stearns and pulvis parturiens.
Am J Obstet Gynecol 1931;22:418–23
need to study the potential of a low-cost agent
4. Edgar JC. The Practice of Obstetrics. Philadelphia:
such as misoprostol for use in the developing
Blakiston, 1913:475-7
world. The cost and instability of standard 5. Dudley HW, Moir C. The substance responsible
oxytocic drugs are prohibitive in many for the traditional clinical effect of ergot. Br Med
low-resource settings. Safety and parallel effi- J 1935;1:520–3
cacy should therefore suffice as parameters 6. Moir C. Clinical experiences with the new
whereby an agent such as misoprostol is judged alkaloid, ergometrine. Br Med J 1936;ii:799–801
rather than demonstration of clinical superiority 7. Dale HH. The action of extracts of the pituitary
over established uterotonics. body. Biochem J 1909;4:427–47
8. duVigneaud V, Ressler C, Swan JM, et al. The
synthesis of an octapeptide amide with the
hormonal activity of oxytocin. J Am Chem Soc
1954;75:4879–80
References
9. von Euler H, Adler E, Hellstrom H, et al. On the
1. Davis DD. The Principles and Practice of Obstetric specific vasodilating and plain muscle stimulat-
Medicine. London: Rebman, 1896:602 ing substance from accessory genital glands in
261
283
30 August 2006 14:21:34
man and certain animals (prostaglandin and 23. Lokugamage AU, Sullivan KR, Niculescu I, et al.
vesiglandin). J Physiol (London) 1937;88:213–34 A randomized study comparing rectally adminis-
10. Stones RW, Paterson CM, Saunders NJ. Risk tered misoprostol versus syntometrine combined
factors for major obstetric haemorrhage. Eur J with an oxytocin infusion for the cessation of
Obstet Gynaecol Reprod Biol 1993;48:15–18 primary postpartum haemorrhage. Acta Obstet
11. Tsu VD. Postpartum haemorrhage in Zimba- Gynecol Scand 2001;80:835–9
bwe: a risk factor analysis. Br J Obstet Gynaecol 24. Mousa HA, Alfirevic Z. Treatment for primary
1993;100:327–33 postpartum haemorrhage. Cochrane Database of
12. Waterstone M, Bewley S, Wolfe C. Incidence Systematic Reviews 2003;1 CD 003249
and predictors of severe obstetric morbidity: 25. Kupferminc MJ, Gull I, Bar-Am A, et al.
case-control study. Br Med J 2001;322:1089–94 Intrauterine irrigation with prostaglandin F2α for
13. Hall MH, Halliwell R, Carr-Hill R. Concomi- management of severe postpartum haemorrhage.
tant and repeated happenings of complications of Acta Obstet Gynecol Scand 1998;77:548–50
the third stage of labour. Br J Obstet Gynaecol 26. Peyser MR, Kupferminc MJ. Management of
1985;92:732–8 severe postpartum hemorrhage by intrauterine
14. Dollery C, ed. Therapeutic Drugs, 2nd edn. irrigation with prostaglandin E2. Am J Obstet
Edinburgh: Churchill Livingstone, 1999 Gynecol 1990;162:694–6
15. Hunter DJ, Schulz P, Wassenaar W. Effect of 27. Barrington JW, Roberts A. The use of gemeprost
carbetocin, a long-acting oxytocin analog on the pessaries to arrest postpartum haemorrhage. Br J
postpartum uterus. Clin Pharmacol Ther 1992;52: Obstet Gynaecol 1993;100:691–2
60–7 28. El-Lakany N, Harlow RA. The use of gemeprost
16. Boucher M, Nimrod CA, Tawagi GF, et al. pessaries to arrest postpartum haemorrhage. Br J
Comparison of carbetocin and oxytocin for the Obstet Gynaecol 1994;101:277
prevention of postpartum hemorrhage following 29. Bates A, Johansen K. The use of gemeprost pes-
vaginal delivery: a double-blind randomized saries to arrest postpartum haemorrhage. Br J
trial. J Obstet Gynaecol Can 2004;26:481–8 Obstet Gynaecol 1994;101:277–8
17. Dansereau J, Joshi AK, Helewa ME, et al. 30. Craig S, Chau H, Cho H. Treatment of severe
Double-blind comparison of carbetocin versus postpartum haemorrhage by rectally adminis-
oxytocin in prevention of uterine atony after tered gemeprost pessaries. J Perinat Med 1999;
caesarean section. Am J Obstet Gynecol 1999; 27:231–5
180:670–6 31. Alok K, Hagen P, Webb JB. Tranexamic acid in
18. Jacobs M, Arias F. Intramyometrial PGF2α in the management of postpartum haemorrhage. Br
treatment of severe postpartum hemorrhage. J Obstet Gynaecol 1996;103:1250
Obstet Gynecol 1980;55:665–6 32. Segal S, Shemesh IY, Blumenthal R, et al. Treat-
19. Oleen MA, Mariano JP. Controlling refractory ment of obstetric hemorrhage with recombinant
postpartum hemorrhage with hemabate sterile activated factor VII (rFVIIa). Arch Gynecol Obstet
solution. Am J Obstet Gynecol 1990;162:205–8 2003;268:266–7
20. Bigrigg A, Chui D, Chissell S, et al. Use of 33. Bouwmeester FW, Jonkhoff AR, Verheijen RH,
intramyometrial 15-methyl prostaglandin F2α et al. Successful treatment of life-threatening
to control atonic postpartum haemorrhage postpartum hemorrhage with recombinant
following vaginal delivery and failure of conven- activated factor VII. Obstet Gynecol 2003;101:
tional therapy. Br J Obstet Gynaecol 1991;98: 1174–6
734–6 34. Why Mothers Die. Report on Confidential
21. Gulmezoglu AM, Villar J, Ngoc NT, et al. WHO Enquiry into Maternal Deaths in the United
multicentre randomised trial of misoprostol in Kingdom 2000–2002. London: RCOG Press,
the management of the third stage of labour. 2004
Lancet 2001;358:689–95 35. Rizvi F, Mackey R, Geary M, et al. Successful
22. O’Brien P, El-Refaey H, Geary M, et al. Rectally reduction of massive postpartum haemorrhage
administered misoprostol for the treatment by use of guidelines and staff education. Br J
of postpartum haemorrhage unresponsive to Obstet Gynaecol 2004;111;495–8
oxytocin and ergometrine: a descriptive study.
Obstet Gynecol 1998;92:212–14
262
284
30 August 2006 14:21:35
28
INTERNAL UTERINE TAMPONADE
D. Danso and P. W. Reginald
INTRODUCTION deciduas) from within the uterus, thereby

resulting in either a significant reduction or
The origin of the word tamponade appears to
stoppage of uterine bleeding.
have come from an old French word for tam-
pon, which carries the connotation of a plug, a
bung or a stopper inserted into an open wound BASIC GENERAL PRINCIPLES
or a body cavity to stop the flow of blood1.
Today, the common usage of this term includes After failure of medical intervention to stop
the collection of menstrual effusion by insertion or reduce postpartum hemorrhage, one
of a preformed sanitary pledget into the vagina. should consider performing internal uterine
In the context of postpartum hemorrhage, tamponade. This should be carried out in the
tamponade refers to plugging the uterus with operating theater with anesthetic and nursing
some type of device to stop the flow of blood. staff present as well as blood transfusion service
Normally, this is in the form of a gauze pack or back-up. The woman should be placed in the
a balloon catheter. Internal tamponade proce- Lloyd Davies or lithotomy position with an
dures have been used successfully alone2–5 or in indwelling urethral catheter. Examination
combination with the Brace suture6 to reduce or under anesthesia should be carried out to
arrest massive postpartum hemorrhage. exclude lacerations, retained placenta, and to
empty the uterus of clots. Only then should
tamponade procedures be attempted. Utero-
tonics and hemostatics are advised as adjunct
PRINCIPLES OF UTERINE therapy and may be given simultaneously. Any
TAMPONADE of the internal uterine tamponade methods
Uterine tamponade requires developing intra- described below can be embarked upon before
uterine pressure to stop bleeding. This can be resorting to surgical interventions.
accomplished in two ways: The following is a description of the
‘tamponade test’ and various other methods of
(1) By insertion of a balloon that distends in the tamponade with their potential advantages and
uterine cavity and occupies the entire space, disadvantages.
thereby creating an intrauterine pressure
that is greater than the systemic arterial
pressure. In the absence of lacerations, the THE TAMPONADE TEST
blood flow into the uterus should stop the
This test, first described in 2003 by Condous
moment the pressure in the tamponade bal-
and colleagues7, was proposed as a prognostic
loon is greater than that of the systemic
index as to whether laparotomy would be
arterial pressure.
needed in patients with major postpartum hem-
(2) By insertion of a uterine pack consisting of a orrhage unresponsive to medical therapy. In the
gauze roll that is tightly packed into the original description, a Sengstaken–Blakemore
uterus in such a manner that pressure is esophageal catheter was inserted into the uter-
applied directly on capillary/venous bleed- ine cavity via the cervix, using ultrasound guid-
ing vessels or surface oozing (of the ance when possible, and filled with warm saline
263
285
30 August 2006 14:21:35
until the distended balloon was palpable per and may not easily adapt to the shape of the
abdomen surrounded by the well-contracted uterine cavity. Moreover, it contains latex and
uterus, and visible at the lower portion of the may not be affordable in resource-poor settings.
cervical canal. The position of the Sengstaken–
Blakemore esophageal catheter was checked to
ensure it was firmly fixed in situ within the uter- RÜSCH HYDROSTATIC UROLOGICAL
ine cavity by the application of gentle traction. If BALLOON
no or only minimal bleeding was observed via
This is a two-way Foley catheter (simplastic 20
the cervix or there was only minimal bleeding
ch, 6.7 mm, 30 ml), which can also be used for
into the gastric lumen of the Sengstaken–
postpartum hemorrhage. It has a capacity
Blakemore esophageal catheter, the tamponade
greater than 500 ml (see Figure 2)3. The tech-
test result was considered to be positive. If this
nique of insertion is similar to the description
were the case, surgical intervention, with possi-
already given for the Sengstaken–Blakemore
ble hysterectomy, was avoided. On the other
esophageal catheter. A 60-ml bladder syringe
hand, if significant bleeding continued via
can be used for inflating the balloon with warm
the cervix or the gastric lumen of the tube, the
saline via the drainage port. It is a simple tech-
tamponade test was deemed a failure and
nique and therefore junior residents can easily
laparotomy was performed. In this study, 14
learn and become adept in its use, especially
out of 16 women (87%) with intractable hemor-
if practised after a manual removal of the
rhage responded positively. Of the women
placenta.
who did not respond, one continued to bleed
because of an overlooked cervical extension of
the lower transverse uterine incision at Cesarean
BAKRI BALLOON
delivery. The balloon was inadequately inflated
in the other. The Rüsch urological balloon has The SOS Bakri tamponade balloon catheter
also been used successfully for the tamponade (Cook Ob/Gyn) is marketed as 100% Silicon
test3. Chapter 29 describes in more detail a (no latex), purpose-designed two-way catheter,
longitudinal study still in progress to determine to provide temporary control or reduction of
the effectiveness of the Rüsch urological balloon postpartum uterine bleeding when conservative
for the tamponade test. management is warranted (see Figure 3)4.
Again, the insertion technique is simple. Insert
the balloon portion of the catheter in the uterus,
making sure that the entire balloon is inserted
SENGSTAKEN–BLAKEMORE TUBE
past the cervical canal and internal os, under
The Sengstaken–Blakemore esophageal cathe- ultrasound guidance if possible. At Cesarean
ter was originally designed for the treatment of delivery, the tamponade balloon can be passed
esophageal variceal bleeds and the introduction via the Cesarean incision into the uterine cavity
of contrast media. It is a three-way catheter with the inflation port passing into the vagina
tube with stomach and esophageal balloon via the cervix. An assistant pulls the shaft of
components (see Figure 1). It can be inflated to the balloon through the vaginal canal until the
volumes greater than 500 ml. Several reports on deflated balloon base comes into contact with
its successful use to arrest major postpartum the internal cervical os. The uterine incision is
hemorrhage are available2,7,8–11. Before inser- closed in the usual fashion, taking care to avoid
tion of the tube, the distal end of the tube puncturing the balloon while suturing. A gauze
beyond the stomach balloon is severed to pack soaked with iodine or antibiotics can then
minimize the risk of perforation. The main be inserted into the vaginal canal to ensure
advantage is its simplicity of use and, therefore, maintenance of correct placement of the bal-
junior residents can easily learn and perform the loon and maximize the tamponade effect. The
test while waiting for help. balloon is then inflated with sterile fluid to the
The main disadvantages are that it is not desired volume for tamponade effect. Gentle
purpose-designed for postpartum hemorrhage traction on the balloon shaft ensures proper
264
286
30 August 2006 14:21:35
Internal uterine tamponade
Figure 1 Sengstaken–Blakemore tube Figure 3 Bakri balloon
contact between the balloon and the tissue sur-

face and may enhance the tamponade effect.
Success can be judged by the declining loss of
blood seen through the drainage port and the
fluid connecting bag.
The main disadvantage of this method is
that it may not be affordable in resource-poor
countries because of the expense.
FOLEY CATHETER
The successful use of the Foley catheter balloon
for internal uterine tamponade is also des-
cribed12,13. A Foley catheter with a 30-ml
balloon capacity is easy to acquire and may
routinely be stocked on labor and delivery
suites. Using a No. 24F Foley catheter, the tip
is guided into the uterine cavity and inflated
with 60–80 ml of saline (anecdotally, a volume
of 150 ml can be reached before it bursts).
Additional Foley catheters can be inserted, if
necessary, until bleeding stops. As attractive,
Figure 2 Rüsch hydrostatic balloon catheter easy and cheap as this method is, some concerns
265
287
30 August 2006 14:58:45
have been raised regarding the use of the Foley (1) Experience is required to pack properly and
catheter for uterine tamponade. First, the tightly and therefore junior residents may
capacity of the immediate postpartum uterine not be able to perform proficiently, espe-
cavity, especially if term, is too large for effective cially if they have large hands. Speed is also
tamponade to be achieved with one inflated necessary because the intrauterine/vaginal
balloon, and the risk of one balloon falling out hand becomes numb rapidly;
of the uterus is increased14. Second, significant
(2) Delay in recognizing continual hemorrhage
bleeding may occur above the Foley bulb, as it
as blood needs to soak through yards of
may not fill the entire uterine cavity. Even the
gauze before it becomes evident;
use of multiple Foley catheters cannot ensure
a complete compression effect on the entire (3) Success of the procedure will not be
uterine surface. known immediately, as the blood must soak
through the pack to reveal itself;
(4) The tightness of the pack is difficult to
HYDROSTATIC CONDOM CATHETER
determine, especially if blood soaks
This innovative approach from Bangladesh uses through, leading to a loss of the tamponade
a sterile rubber catheter fitted with a condom as effect;
a tamponade balloon device14. The sterile cath-
(5) Potential risk of trauma and infection;
eter is inserted within the condom and tied near
the mouth of the condom with a silk thread, and (6) Removing the pack may often require a
the outer end of the catheter is connected to a separate surgical procedure to dilate
saline set. In its original description, after place- and extract the intrauterine material, thus
ment in the uterus, the condom is inflated with falling short of an ideal option.
250–500 ml normal saline according to need,
Notwithstanding, uterine packing remains an
and the outer end of the catheter was folded and
option, especially, if balloon catheters or
tied with thread after bleeding had stopped14.
balloons are not available. The risk of intra-
Vaginal bleeding is observed and further infla-
uterine infection can be minimized by prophy-
tion is stopped when bleeding has ceased. To
lactic antibiotics.
keep the balloon in situ, the vaginal cavity
is packed with roller gauze and sanitary pads.
Success is gauged by the amount of blood loss CARE AFTER SUCCESSFUL UTERINE
per vaginum. Hemorrhage was arrested within TAMPONADE
15 min in all 23 cases in the original series14.
All patients should be managed in a high-
Although the sample size was small, this method
dependency or intensive care unit with very
represents a cheap, simple and quick interven-
close monitoring of their vital signs, fluid
tion which may prove invaluable in, especially,
input/output, fundal height and vaginal blood
resource-poor countries.
loss. Continued oxytocin infusion may be
necessary to keep the uterus contracted over
12–24 h. Prophylactic broad-spectrum anti-
UTERINE PACKING
biotic cover should be administered. The mean
Uterine packing entails placing, carefully and time for leaving tamponade balloons or uterine
systematically, several yards of gauze inside the packs ranges from 8 to 48 h2,7,9–12. A graduated
uterine cavity to occlude the whole intrauterine deflation of the balloon is advised to reduce the
space and, thus, control major hemorrhage. potential risk of further bleeding.
The technique fell out of favor in the 1950s, as it In summary, tamponade procedures are sim-
was thought to conceal hemorrhage and cause ple, cheap, easy to use, and effective measures
infection. It re-emerged in the 1980s and 1990s that should be considered in women with
after these concerns were not verified15. The intractable postpartum hemorrhage, especially
main disadvantages of this technique are: when other options may be unavailable.
266
288
30 August 2006 14:21:47
Internal uterine tamponade
References 8. Condie RG, Buxton EJ, Paynes ES. Successful

use of Sengstaken–Blakemore tube to control
1. Collins English Dictionary, 5th edn. London: massive postpartum haemorrhage. Br J Obstet
Collins, 2000:1563 Gynaecol 1994;101:1023–4
2. Katesmark M, Brown R, Raju KS. Successful 9. Chan C, Razvi K, Tham KF, Arulkumaran S.
use of a Sengstaken–Blakemore tube to control The use of a Sengstaken–Blakemore tube to
massive postpartum haemorrhage. Br J Obstet control postpartum hemorrhage. Int J Gynaecol
Gynaecol 1994;101:259–60 Obstet 1997;58:251–2
3. Johanson R, Kumar M, Obrai M, Young P. 10. Japaraj RP, Raman S. Sengstaken–Blakemore
Management of massive postpartum haemor- tube to control massive postpartum haemor-
rhage: use of a hydrostatic balloon catheter to rhage. Med J Malaysia 2003;58:604–7
avoid laparotomy. Br J Obstet Gynaecol 2001; 11. Frenzel D, Condous GS, Papageorghiou AT,
108:420–2 McWhinney NA. The use of the ‘tamponade
4. Bakri YN, Amri A, Abdul Jabbar F. test’ to stop massive obstetric haemorrhage in
Tamponade-balloon for obstetrical bleeding. Int placenta accreta. Br J Obstet Gynaecol 2005;112:
J Gynaecol Obstet 2001;74:139–42 676–7
5. Ferrazzani S, Guariglia L, Caruso A. Therapy 12. De Loor JA, van Dam PA. Foley catheters
and prevention of obstetric haemorrhage by for uncontrollable obstetric or gynaecologic
tamponade using a balloon catheter. Minerva hemorrhage. Obstet Gynecol 1996;88:737
Ginecol 2004;56:481–4 13. Marcovici I, Scoccia B. Postpartum hemorrhage
6. Danso D, Reginald P. Combined B-Lynch and intrauterine balloon tamponade. A report of
suture with intrauterine balloon catheter tri- three cases. J Reprod Med 1999;44:122–6
umphs over massive postpartum haemorrhage. 14. Akhter S, Begum MR, Kabir Z, Rashid M, Laila
Br J Obstet Gynaecol 2002;109:963 TR, Zabeen F. Use of a condom to control
7. Condous GS, Arulkumaran S, Symonds I, massive postpartum hemorrhage. Med Gen Med
Chapman R, Sinha A, Razvi K. The 2003;115:38
‘Tamponade test’ in the management of massive 15. Maier RC. Control of postpartum hemorrhage
postpartum hemorrhage. Obstet Gynecol 2003; with uterine packing. Am J Obstet Gynecol 1993;
101:767–72 169:317–21
267
289
30 August 2006 14:21:47
29
THE BALLOON INTERNAL UTERINE TAMPONADE
AS A DIAGNOSTIC TEST
S. Ferrazzani, L. Guariglia and C. Dell’Aquila
INTRODUCTION easily carried out by doctors in training while

awaiting help from a senior colleague.
During the last several years, a number of new
and simpler techniques have been developed in
the attempt to avoid major surgical procedures
UTERINE PACKING
for treatment of postpartum hemorrhage1–8.
Although a variety of surgical options have Control of postpartum hemorrhage by uterine
been proposed to avoid hysterectomy including packing is not new15. For many years, uterine
uterine artery ligation, ovarian artery ligation, packing with sterile gauze has been used in
internal iliac artery ligation, and B-Lynch Brace the clinical management of severe postpartum
suture9, a suitable conservative technique is still hemorrhage and as the last resort before hyster-
lacking1 and all proposed options have risks as ectomy16. Because of the availability of better
well as advantages10. uterotonic medications, this practice lost its
In most cases, procedures are effective in appeal, but reports on its successful use con-
avoiding hysterectomy, but a delay carries a tinued to appear17–19. Recently, some authors
poorer prognosis. Moreover, each of these tech- raised concerns about concealed bleeding and
niques entails a laparotomy and skilled person- infection20; a newer technique, however, has
nel must perform the procedure. Rarely, major allayed some of these concerns21.
complications follow radical surgery for post- Uterovaginal packing may sometimes obviate
partum hemorrhage; these include loss of fertil- the need for surgery altogether. In cases of
ity, other morbidity and even maternal death11. deliveries complicated by postpartum hemor-
B-Lynch and colleagues12,13 used brace rhage, after excluding uterine rupture, genital
sutures to compress the uterus without com- tract lacerations, and retained placental tissue,
promising major vessels. The advantage of the efforts are directed toward contracting the
B-Lynch procedure is that identification of spe- uterus by bimanual compression and uterotonic
cific blood vessels is not necessary, a process agents. If these are not successful, one must
which is often difficult. Although helpful during resort to surgical techniques. At this stage, an
Cesarean section, the B-Lynch procedure alternative option to remember is uterovaginal
requires a laparotomy and therefore may not packing. Easy and quick to perform, it may be
be ideal as the first approach in cases of used to control bleeding by tamponade effect
postpartum hemorrhage following vaginal and stabilize the patient until a surgical
delivery14. procedure is arranged.
This chapter will focus on one of the recently Chapter 28 describes the technique of
reported conservative measures to control hem- uterine packing in more detail.
orrhage – internal uterine tamponade. Although
uterine atony is the main indication for internal
uterine tamponade, this methodology is also BALLOON TAMPONADE
useful for postpartum hemorrhage arising from
Tamponade with different types of balloon
placenta previa/accreta. The technique can be
catheters used prior to surgery is a conservative
268
290
30 August 2006 14:21:47
The balloon internal uterine tamponade as a diagnostic test
procedure that is available before invasive The first therapeutic approach to this situa-
surgical techniques are needed1,3. Chapter 28 tion is mechanical stimulation by massage of the
describes the various types of balloon catheters uterus and then the use of uterotonic drugs. In
that are available. this case, the efficacy of the tamponade balloon
Balloon tamponade of the uterus is a recog- may derive from the mechanical stimulation of
nized procedure in those with massive and myometrial contraction caused by the balloon’s
intractable hemorrhage17,22–29. elasticity pressing against the myometrial wall.
The use of the Sengstaken–Blakemore eso- The simultaneous and continuous stimulation
phageal or gastric catheter is described in the of myometrial contraction and the tamponade
literature for the control of massive postpartum effect on the open vessels, reached with the
hemorrhage due to an atonic uterus not contraction, explain its efficacy. However, the
responding to oxytocics including prosta- uterus must be empty for the tamponade to
glandins3,17,23,30,31 or due to placenta accreta32. be successful.
Multiple Foley catheters in the case of a vaginal In the presence of placenta accreta, the
delivery22,25 have also been used and even balloon must be used with great caution, as a
rubber catheters fitted with a condom have failure or delay to control hemorrhage in such
been used successfully to control postpartum patients could be catastrophic.
hemorrhage in undeveloped countries33. Uro- In the small series reported in the literature,
logical fluid-filled catheters (300–500 ml)26,28,29 in which the different types of balloon catheter
or of silicone balloons designed for tamponade were filled with various volumes, ranging from
function27 also seem to be very effective, with 30 to 500 ml, Seror and colleagues chose an
further possibilities in cases of hemorrhage after inflation volume of 250 ml, since this value
Cesarean section for placenta previa/accreta. corresponds to the approximate volume of the
uterine cavity after delivery31.
Theoretical principle of action
The theoretical principle of the balloon
BALLOON TAMPONADE AS A TEST
tamponade is that temporary and steady
mechanical compression of the bleeding surface To date, there is no diagnostic test to identify
of the placental site can be performed while those patients with intractable hemorrhage who
waiting for the natural hemostatic mechanisms will need surgery. Condous and colleagues3
of the blood to take effect. The balloon, inflated proposed the use of an inflated Sengstaken–
inside the uterine cavity in order to stretch the Blakemore balloon catheter as a test to create
myometrial wall, can exert an intrauterine pres- tamponade and identify patients who will or will
sure that overcomes the systemic arterial pres- not need surgery (‘tamponade test’). When its
sure, resulting in cessation of the intrauterine results are positive, the tamponade test not only
blood flow. Probably, a quite different mecha- halts the blood loss and preserves the uterus,
nism can be advocated for its efficacy in the case but also gives an opportunity to reverse and
of uterine atony. With separation of the pla- correct any consumptive coagulopathy. More
centa, the many uterine arteries and veins that than 87% of their patients (14/16) with intracta-
carry blood to and from the placenta are severed ble postpartum hemorrhage responded to the
abruptly. Elsewhere in the body, hemostasis in tamponade test3. More recently, Seror and col-
the absence of surgical ligation depends upon leagues reported that, in a series of 17 cases,
intrinsic vasospasm and formation of blood tamponade treatment prevented surgery in 88%
clots locally. At the placental implantation site, of patients31.
the most important factors for achieving According to these clinical experiences, an
hemostasis are contraction and retraction of the early use of the balloon catheter may reduce
myometrium in order to compress the vessels the total blood loss, and it is probable that
and obliterate their lumens. Uterine atony any type of inflatable balloon with high fluid-
from any origin can prevent this physiological filling capacity could be used for the same
mechanism, leading to massive hemorrhage. purpose.
269
291
30 August 2006 14:21:47
The experience at the Catholic University In the three patients delivering by the vaginal
of Rome, Italy route, an examination was performed under
regional or general anesthesia for retained tissue
A longitudinal study is currently running in the
and lacerations and, when necessary, retained
Obstetrics and Gynecology Department of the
tissue or placenta was removed and lacerations
Catholic University of Rome, Italy; it started in
were sutured.
January 2002 and the Institute review board
Coagulation studies were carried out simul-
approved the study.
taneously to exclude coagulopathy as the first or
the complimentary cause of the hemorrhage.
In those patients considered for the study
Patients and methods
who showed no response to these measures, a
In the period January 2002–August 2005, sterile hydrostatic (bladder distention) balloon
10 773 patients delivered in our maternity catheter size Ch. 16, 5.3 mm (Rüsch UK High
ward. During this period, there were 124 Wycombe, England) (Figure 1) was inserted
(1.15%) instances of postpartum hemorrhage. into the uterine cavity via the cervix. This was
Of these, 13 were considered critical and the achieved using minimal analgesia or regional
women underwent treatment by intrauterine anesthetic. The insertion was facilitated by
tamponade. grasping the anterior and lateral margins of the
An atonic uterus caused postpartum hemor- cervix with sponge forceps and placing the bal-
rhage in one case and placenta previa/accreta loon into the uterine cavity with another sponge
was noted in 12 cases, of which two were forceps. The balloon catheter was then filled
associated with uterine atony. with 120–300 ml of warm saline solution until
The mean age of the patients was 35 years a contracted uterus was palpable through the
(26–39 years). The mean gestational age at abdomen. Applying gentle traction at this stage
delivery was 36 weeks from the first day of confirmed that the filled balloon was firmly
the last menstrual period (26 weeks and 5
days to 40 weeks and 1 day). Nine patients
were multiparous (69.2%). The mean parity
was 2.1.
Labor was spontaneous in three cases, and
stimulated with dinoprostone intravaginal gel
or oxytocin infusion in two cases. The mean
duration of labor was 6 h and 42 min (5–9 h).
Three patients had a vaginal delivery, and ten
had a Cesarean section (of which seven were
planned).
Routinely, the patients who delivered by the
vaginal route had prophylactic intramuscular
oxytocin/ergometrine in the third stage of labor
and all the patients who underwent Cesarean
section had intramyometrial and intravenous
oxytocin during/after the placenta was
delivered.
In the 13 cases of postpartum hemorrhage
considered in this study, patients were treated
with appropriate oxytocic agents and prosta-
glandin analogues (intravenous infusions
of oxytocin (40–100 U), intramyometrial
oxytocin (20 U), intramuscular ergometrine
(0.25–0.5 mg), and/or intravenous infusion of
sulprostone (500 mg)). Figure 1 The Rüsch hydrostatic balloon catheter
270
292
30 August 2006 14:21:49
fixed in the uterine cavity. If no or minimal warm saline solution, using a 60 ml bladder
bleeding was observed through the cervix, syringe. Tamponade was achieved by pulling
laparotomy was avoided and a gauze packing the distal extremity of the catheter shaft out of
of the vagina was performed to avoid self- the vagina. The uterine contraction over the
expulsion of the balloon from the completely balloon was maintained, after the uterine clo-
dilated cervical os. If significant bleeding sure, by a slow oxytocin infusion (20–40 U) that
continued through the cervix, the ‘tamponade was given over the next 24 h. A single-layer
test’ had failed and laparotomy was performed. closure of the uterine incision was performed,
In all the patients delivering by urgent or taking care not to include the balloon in the
planned Cesarean section, the problem of suture line. The Cesarean section was con-
abnormal insertion or suspicion of morbid cluded following the classical technique. Only
adhesion of the placenta was detected by ultra- when the bleeding was adequately controlled
sound scan before surgery. The placenta was was the abdominal wall closed.
delivered by firmly controlled cord traction, or Those who responded to the balloon catheter
by manual removal if it was abnormally adherent therapy were stabilized in the labor and delivery
to the uterine wall. If severe bleeding persisted unit for ongoing management. In all cases,
despite a contracted uterus after local intramyo- intravenous broad-spectrum antibiotics were
metrial and endovenous infusion of oxytocin administered for at least the first 24 h. The bal-
and prostaglandin analogues, the hydrostatic loon catheter was left in situ until the next day.
balloon catheter previously described was During this time interval, blood transfusion and
inserted intrabdominally, through the uterine coagulopathy correction were possible. Once
incision, into the cervical opening and through the above parameters were within acceptable
the cervical canal by a sponge forceps, leaving limits, the balloon catheter was slowly deflated
the balloon in the uterine cavity (Figure 2). The and withdrawn and the patient observed for any
balloon was then filled with 180–300 ml of active bleeding.
Figure 2 Intra-abdominal insertion of the hydrostatic balloon catheter into the uterine cavity
271
293
30 August 2006 14:21:53
Table 1 Clinical details of patients with postpartum hemorrhage who underwent a balloon tamponade
Age Gestation Duration of

Case (years) Gravidity Parity (weeks) labor (h) Mode of delivery
1 36 2 1 35 – planned CS
2 29 5 1 35 – planned CS
3 38 4 2 37 – planned CS
4 30 1 0 37 – planned CS
5 36 1 0 40 6 spontaneous labor
6 34 3 0 34 – urgent CS
7 34 2 1 37 – planned CS
8 36 2 1 36 – planned CS
9 39 5 1 30 – urgent CS
10* 35 2 0 26 – urgent CS
11 39 3 2 40 9 induced/oxytocin
12 33 6 1 38 5 induced/oxytocin
13 26 4 1 35 – planned CS
CS, Cesarean section; *failed ‘tamponade test’
Results explored and the placenta accurately removed.

A 3-cm fragment of the placenta was lacking but
The ‘tamponade test’ was positive in 12 out of
even a very vigorous examination of the uterine
13 cases and the hydrostatic catheter immedi-
cavity was unsuccessful in removing that frag-
ately arrested hemorrhage. In one case, tampon-
ment. A catheter balloon inserted through the
ade failed after 3 h and bleeding re-occurred. In
vagina soon arrested the severe bleeding. The
our series of 13 cases, the primary cause for
patient was administered 2 units of blood and
postpartum hemorrhage was bleeding at the
the balloon was removed after 24 h. The patient
placental site alone (ten cases), uterine atony
was discharged from the hospital 7 days later
associated with bleeding at the placental site
and her progress was uneventful until 11 days,
(two cases), and uterine atony with cervical
when she was re-admitted to hospital for further
laceration and pre-eclampsia-associated dis-
hemorrhage due to the expulsion of the placen-
seminated intravascular coagulation (one case).
tal fragment. An examination of the uterine cav-
Tables 1 and 2 provide details of the 13 cases.
ity resolved the case with no other intervention
Because, according to Benirschke and
or further blood transfusion.
Kaufmann34, the diagnosis of accreta cannot
Another patient (case 12) had a normal
be made when the placenta is not removed with
vaginal delivery with no pre-delivery suspicion
the uterus, in such a condition the diagnosis of
of abnormal adherence of the placenta. After
placenta accreta was based on clinical criteria
delivery of the placenta, the lack of a 3-cm pla-
and consisted of the inability to remove it by
cental fragment was observed. The examination
controlled cord traction because of a severe
of the uterine cavity was unsuccessful but, in the
adherence to the underlying myometrium and
absence of further bleeding, the patient was kept
failure to develop a cleavage plane between the
under observation with no other intervention.
placenta and uterus.
During the subsequent 24 h, sub-acute vaginal
Among the three patients delivering by the
bleeding was associated with a progressive fall of
vaginal route, two deserve a more detailed
the hematocrit level. A further examination of
description. One patient (case 5) had a pre-
the uterine cavity was planned, and the removal
delivery ultrasound diagnosis of marginal pla-
of the retained placental fragment caused a
centa previa. The woman had a normal labor,
severe hemorrhage that was quickly stopped by
but soon after delivery of the placenta a profuse
introducing a balloon catheter through the
hemorrhage began. The uterine cavity was
vagina into the uterine cavity.
272
294
30 August 2006 14:21:53
Table 2 Cause of bleeding and medical treatment before tamponade procedure
Postpartum
Estimated Intrapartum hospital
blood loss RBC and Postpartum admission
Case Cause of bleeding (ml) FFP RBC Medical treatment (days)
1 total placenta previa 1000 0 0 intramyometrial oxytocin, 5

oxytocin infusion
2 total placenta previa 3000 RBC 2 U 0 intramyometrial oxytocin, 11
oxytocin infusion
oxytocin infusion
4 total placenta previa 1200 RBC 1 U RBC 2 U intramyometrial oxytocin, 6
oxytocin infusion
5 marginal placenta previa, 1200 0 RBC 2 U oxytocin infusion, i.m. 8
focally accreta ergometrine
oxytocin infusion
oxytocin infusion
8 focal placenta accreta 1000 0 0 intramyometrial oxytocin, 4
oxytocin infusion
9 focal placenta accreta, 3300 RBC 6 U, 0 intramyometrial oxytocin, 5
atony FFP 6 U oxytocin infusion,
sulprostone infusion
10* marginal placenta previa, 5000 RBC 9 U 0 intramyometrial oxytocin, 6
focally accreta, atony oxytocin, sulprostone
infusion
11 atony, cervico-isthmic tear 1100 0 RBC 2 U i.m. oxytocin, oxytocin 7
and DIC in pre-eclampsia infusion
12 focal placenta accreta 1500 RBC 2 U RBC 2 U oxytocin and sulprostone 5
infusion, i.m. ergometrine
13 total placenta 1100 0 0 intramyometrial oxytocin, 5
previa/accreta oxytocin infusion
RBC, red blood cell; FFP, fresh frozen plasma; DIC, disseminated intravascular coagulation; i.m., intra-
muscular; *failed ‘tamponade test’
Among the ten cases resulting in Cesarean O’Leary suture35 was positioned around the
section, one patient (case 13) showed at ultra- uterine arteries immediately after delivery of the
sound scan high suspicion of morbid adhesion of infant, with the placenta still in situ, using a
the placenta (accreta/increta) before the planned 2-monofilament absorbable suture on a high-
Cesarean section for total placenta previa. Dur- curve needle. Subsequently, a bilateral utero-
ing Cesarean section, in order to prevent severe ovarian vessel ligation was performed with a
bleeding at delivery of the placenta by reducing 1-monofilament absorbable suture, including the
the blood flow to the uterus, a prophylactic broad ligament close under the tubal insertion to
273
295
30 August 2006 14:21:53
the uterus and the utero-ovarian ligament. The therapy of postpartum hemorrhage has been
placenta was found to extend across the internal shown in two cases described by Johanson36 and
cervical os. The inability to remove it by firmly in four cases more recently reported28,29. How-
controlled cord traction because of a severe ever, its efficacy in severe postpartum hemor-
adherence to the underlying myometrium and to rhage needed to be evaluated in a larger series.
develop a cleavage plane between the placenta In the present provisional study, the insertion
and uterus became the clinical confirmation of of the Rüsch urological hydrostatic balloon in
placenta accreta34. Therefore, the placental tis- patients with massive postpartum hemorrhage
sue was manually removed in fragments and the was very successful and was associated with no
placental site inspected. No myometrial defects significant complications. The procedure failed
were found, as the adhesion was limited to the in only two cases. As opposed to the traditional
myometrial layer. In order to control a persistent, gauze uterine packing, the technique with the
although moderate, bleeding from the placental balloon catheter provides immediate knowledge
site which did not respond to pharmacological of its effectiveness in controlling the postpartum
uterotonic therapy, a hydrostatic balloon cathe- hemorrhage, so that subsequent surgery can be
ter was inserted through the uterine incision, expedited in failed cases.
leaving the balloon in the uterine cavity as previ- If bleeding continues despite the insertion of
ously described. The patient did not need blood a balloon, the Rüsch urological hydrostatic bal-
transfusion and a 6-month follow-up by Doppler loon gives less information than a Sengstaken–
ultrasound demonstrated regular reperfusion of Blakemore catheter, since bleeding is noted only
the uterus. through the cervix but not from the uterine
Conservative treatment with the balloon fundal cavity. However, the Rüsch urological
catheter was unsuccessful in two cases and hys- hydrostatic balloon is simpler and cheaper than
terectomy was performed (cases 9 and 10). In the other. At the same time, its overturned
case 9, the balloon catheter was inserted, after pear-shape better fits in the uterine cavity, with
Cesarean section was concluded, by the vaginal probably less risk of self-expulsion. The uterus
route because of a persistent vaginal bleeding. must be empty for successful tamponade. If the
The ‘tamponade test’ was successful and the uterine cavity is completely empty and uterine
patient was monitored for 3 h. However, the contraction sustained by adequate pharmaco-
patient then had a hemorrhage due to secondary logical assistance, there is probably no need for
uterine atony not responding to oxytocics and monitoring bleeding from the uterine fundal
sulprostone infusion. Even further filling of the cavity. A larger series of cases will be necessary
balloon was unsuccessful and, soon after the to support this last opinion.
removal of the balloon, a large amount of blood The Rüsch urological hydrostatic balloon
and clots were expelled from the cervical os, so takes a few minutes to insert, is unlikely to cause
that urgent hysterectomy was mandatory. In trauma and is easy to place with minimal or no
case 10, the ‘tamponade test’ failed and no anesthesia, whereas its removal is painless and
other surgical approach was attempted before simple. Whether the patient is going to bleed
hysterectomy. The reason for the failure of the after removal of the balloon is a general con-
‘tamponade test’ was uterine atony refractory to cern, but this series demonstrates that there
any pharmacological treatment. were no cases of rebleeding after the planned
The 13 patients had a total estimated blood removal of the Rüsch urological hydrostatic
loss of 23.4 liters. The lowest and highest balloon. In case of rebleeding, it is possible to
estimated blood losses experienced were 1 and replace the balloon while planning an oppor-
5 liters. A total of 28 U of blood and 6 U of tune uterine arterial embolization in a patient
fresh frozen plasma were transfused. who is now in a stable condition36–39.
There were two cases of failure; atony was
the cause of failure and subsequent hysterec-
Discussion
tomy in both. In these cases, an attempt to
The effectiveness of the Rüsch urological hydro- mechanically favor uterine contraction by
static balloon as a conservative procedure in the applying a B-Lynch Brace suture of the uterus
274
296
30 August 2006 14:21:53
combined with an additional insertion in the 9. Tamizian O, Arulkumaran S. The surgical man-
uterine cavity of a balloon catheter could possi- agement of post-partum haemorrhage. Best Pract
bly have resolved the problem, with the com- Res Clin Obstet Gynecol 2002;16:81–98
bined conservative approach already described 10. Drife J. Management of primary postpartum
haemorrhage. Br J Obstet Gynaecol 1997;104:
by Danso and Reginald40.
275–7
One of the difficulties in the management of
11. El-Hamamy E, B-Lynch C. A worldwide review
patients with intractable postpartum hemor- of the uses of the uterine compression suture
rhage, not responding to uterotonic agents, is techniques as alternative to hysterectomy in the
the decision to perform a laparotomy and, in management of severe post-partum haemor-
case of Cesarean section, the decision to per- rhage. J Obstet Gynecol 2005;25:143–9
form a hysterectomy. The delay can be cata- 12. Vangsgaard K. ‘B-Lynch-suture’ in uterine
strophic. In the present series, average blood atony. Ugeshr Laeger 2000;162:3468
loss was considerably less than that of other 13. B-Lynch C, Coker A, Lawal AH, Abu J, Cowen
series recently reported3,31. In all the cases but MJ. The B-Lynch surgical technique for the
two, the risk of postpartum hemorrhage was control of massive postpartum haemorrhage: an
alternative to hysterectomy? Five cases reported.
known in advance. When there is confidence
Br J Obstet Gynaecol 1997;104:372–5
that the management of postpartum hemor-
14. Allam MS, B-Lynch C. The B-Lynch and other
rhage can be conservative, easy and effective, as uterine compression suture techniques. Int J
in the case of application of a balloon catheter, Gynaecol Obstet 2005;89:236–41
there is no reason for a delay. 15. Drucker M, Wallach RC. Uterine packing: a
In conclusion, the safe, low-cost, and easy re-appraisal. Mt Sinai J Med 1979;46:191–4
procedure of utilizing a balloon catheter can 16. American College of Obstetrician and Gynecolo-
be applied in any situation of life-threatening gists. Diagnosis and management of postpartum
postpartum hemorrhage and avoids radical sur- hemorrhage. ACOG technical bulletin no. 143.
gery in patients so that reproductive capacity is Washington, DC: American College of
preserved. Obstetricians and Gynecologists, 1990
17. Katesmark M, Brown R, Raju KS. Successful
use of a Sengstaken-Blakemore tube to control
References massive postpartum haemorrhage. Br J Obstet
Gynaecol 1994;101:259–60
1. Tamizian O, Arulkumaran S. The surgical 18. Kauff ND, Chelmow D, Kawada CY. Intracta-
management of postpartum haemorrhage. Curr ble bleeding managed with Foley catheter
Opin Obstet Gynecol 2001;13:127–31 tamponade after dilatation and evacuation. Am J
2. Papp Z. Massive obstetric hemorrhage. J Perinat Obstet Gynecol 1995;173:957–8
Med 2003;31:408–14 19. Bagga R, Jain V, Kalra J, Chopra S, Gopalan S.
3. Condous GS, Arulkumaran S, Symonds I, Uterovaginal packing with rolled gauze in post-
Chapman R, Sinha A, Razvi K. The ‘tamponade partum hemorrhage. Med Gen Med 2004;13:50
test’ in the management of massive postpartum 20. Hsu S, Rodgers B, Lele A, Yeh J. Use of packing
hemorrhage. Obstet Gynecol 2003;101:767–72 in obstetric hemorrhage of uterine origin. J
4. El-Refaey H, Rodeck C. Post-partum haemor- Reprod Med 2003;48:69–71
rhage: definitions, medical and surgical manage- 21. Roman AS, Rebarber A. Seven ways to control
ment. A time for change. Br Med Bull 2003;67: postpartum hemorrhage. Contemp Obstet Gynecol
205–17 2003;48:34–53
5. Pahlavan P, Nezhat C, Nezhat C. Hemorrhage 22. De Loor JA, van Dam PA. Foley catheters
in obstetrics and gynecology. Curr Opin Obstet for uncontrollable obstetric or gynecologic
Gynecol 2001;13:419–29 hemorrahage. Obstet Gynecol 1996;88:737
6. Gielchiensky Y, Rojansky N, Fasoulitios SJ, 23. Chan C, Razvi K, Tham KF, Arulkumaran S.
Ezra Y. Placenta accrete – summary of 10 years: The use of a Sengstaken-Blakemore tube to
a survey of 310 cases. Placenta 2002;23:210–14 control post-partum hemorrhage. Int J Gynaecol
7. Shevell T, Malone FD. Management of obstetric Obstet 1997;58:251–2
hemorrhage. Semin Perinatol 2003;27:86–104 24. Bakri YN. Uterine tamponade-drain for hemor-
8. Mousa HA, Walkinshaw S. Major postpartum rhage secondary to placenta previa-accreta. Int J
haemorrhage. Curr Opin Obstet Gynecol 2001;13: Gynaecol Obstet 1992;37:302–3
595–603
275
297
30 August 2006 14:21:53
25. Marcovici I, Scoccia B. Postpartum hemorrhage 33. Akhter S, Begum MR, Kabir Z, Rashid M, Laila
and intrauterine balloon tamponade: a report of TR, Zabeen F. Use of a condom to control
three cases. J Reprod Med 1999;44:122–6 massive postpartum hemorrhage. Med Gen Med
26. Johanson R, Kumar M, Oberai M, Young P. 2003;5:38
Management of massive postpartum haemor- 34. Benirschke K, Kaufmann P, eds. Pathology of the
rhage: use of a hydrostatic balloon catheter to Human Placenta, 4th edn. New York: Springer,
avoid laparotomy. Br J Obstet Gynaecol 2001; 2000:554
108:420–2 35. O’Leary JA. Uterine artery ligation in the control
27. Bakri YN, Amri A, Abdul Jabbar F. of postcaesarean haemorrhage. J Reprod Med
Tamponade-balloon for obstetrical bleeding. Int 1995;40:189–93
J Gynaecol Obstet 2001;74:139–42 36. Mitty H, Sterling K, Alvarez M, Gendler R.
28. Ferrazzani S, Guariglia L, Caruso A. Therapy Obstetric haemorrhage: prophylactic and
and prevention of obstetric hemorrhage by emergency arterial catheterization and embolo-
tamponade using a balloon catheter. Minerva therapy. Radiology 1993;188:183–7
Ginecol 2004;56:481–4 37. Pelage JP, Le Dref O, Jacob D, Soyer P,
29. Ferrazzani S, Guariglia L, Triunfo S, Caforio L, Herbreteau D, Rymer R. Selective arterial
Caruso A. Successful treatment of post-Cesarean embolization of the uterine arteries in the man-
hemorrhage related to placenta praevia using agement of intractable post-partum hemorrhage.
an intrauterine balloon. Two case reports. Fetal Acta Obstet Gynecol Scand 1999;78:698–703
Diagn Ther 2006;21:277–80 38. Corr P. Arterial embolization for haemorrhage in
30. Condie RG, Buxton EJ, Payne ES. Successful the obstetric patient. Best Pract Res Clin Obstet
use of a Sengstaken-Blakemore tube to control Gynecol 2001;4:557–61
massive postpartum haemorrhage [letter]. Br J 39. Tourn G, Collet F, Seffert P, Veyret C. Place
Obstet Gynaecol 1994;101:1023–4 of embolization of the uterine arteries in the
31. Seror J, Allouche C, Elhaik S. Use of management of post-partum haemorrhage: a
Sengstaken-Blakemore tube in massive post- study of 12 cases. Eur J Obstet Gynecol Reprod
partum hemorrhage: a series of 17 cases. Acta Biol 2003;110:29–34
Obstet Gynecol Scand 2005:84:660–4 40. Danso D, Reginald P. Combined B-Lynch
32. Frenzel D, Condous GS, Papageorghiou AT, suture with intrauterine balloon catheter tri-
McWhinney NA. The use of the ‘tamponade umphs over massive postpartum haemorrhage.
test’ to stop massive obstetric haemorrhage in Br J Obstet Gynaecol 2002;109:963
placenta accreta. Br J Obstet Gynaecol 2005;112:
676–7
276
298
04 September 2006 14:09:34
30
EMBOLIZATION
K. Choji and T. Shimizu
INTRODUCTION percutaneous transcatheter arterial emboliza-

tion (hereafter referred to as embolization) may
The standard treatments of postpartum
be indicated. The main objective of embo-
hemorrhage are described throughout this
lization is to stop active bleeding from the
book. When they are unsuccessful, however,
uterus or the birth canal and to prevent
(a) (b)
Figure 1 Branch patterns of the arteries to the uterus and the birth canal. (a) The most frequent pattern
of branching. The internal iliac artery (IIA) is initially divided into the superior and inferior gluteal trunks
(SGT and IGT, respectively), i.e. the gluteal bifurcation (GB). The uterine, vaginal and inferior pudendal
arteries (UA, VA and IPA, respectively) are the branches of the IGT together with the obturator and cystic
arteries (OA and CA, respectively). (b) Example of less common patterns include the uterine artery (UA)
arising at the gluteal bifurcation, the obturator artery (OA) arising directly from the internal iliac artery (IIA)
proximal to the iliac bifurcation, the internal pudendal artery (IPA) arising from the superior gluteal trunk
(SGT). Ao, aorta; AB, aortic bifurcation; IB, iliac bifurcation; CIA, common iliac artery; EIA, external iliac
artery; MRA, middle rectal artery; SGA, superior gluteal artery; IGA, inferior gluteal artery
277
299
30 August 2006 14:59:25
recurrent hemorrhage. In case this is not poss- arteries. In addition, the internal pudendal
ible, the last resort is to occlude the internal iliac artery may arise from the posterior branch that
arteries on a temporary basis to aid subsequent supplies the superior gluteal artery. To avoid
surgical intervention. confusion due to anatomical variation, we
When embolization is successful, on the would like to refer to the anterior and posterior
other hand, the patient can rapidly recover branches as the inferior and superior gluteal
without undergoing additional surgery. trunks, respectively. This nomenclature be-
Embolization not only saves the life of the comes more appropriate when performing
patient, but also the uterus and adnexal organs, angiography.
thus preserving fertility. Significant radiation On angiographic images, the inferior gluteal
effect is unlikely, as described below. The artery is seen as descending laterally and
procedure is also useful in those patients who extending lower than bony pelvis. The impor-
cannot accept transfusion due to religious or tance of this artery gives off the sciatic branch
other reasons (see Chapter 15). In those hospi- which supplies the sciatic nerve. Therefore, the
tals where embolization is available, it should accidental embolization of the inferior gluteal
be the procedure of choice for postpartum artery could result in transient or long-term
hemorrhage prior to surgical intervention. injury to the sciatic nerve.
High success rates in achieving hemorrhage The intramural portion of the uterine artery
cessation are possible. In an extensive review has a distinctive tortuous configuration. How-
of the literature by Vedantham and colleagues ever, its origin lacks any characteristic appear-
in 19971, cessation of hemorrhage was reported ance and is often superimposed on other
in 100% of 49 cases after vaginal delivery branches in the frontal projection. Therefore,
and 89% in 18 cases after Cesarean sections. oblique views of the inferior gluteal trunk are
Other recent reports include 75%2, 83%3 and frequently required to clarify the branching
100%4. point of the uterine artery. The superior cystic
artery can be identified by superselective
catheterization and manual contrast injection
which demonstrates either the distal network of
VASCULAR ANATOMY ON IMAGING
the artery in the bladder wall or sometimes the
The internal iliac artery is the first major branch cystic artery on the opposite side. The internal
of the common iliac artery, which descends pudendal artery, which is usually a branch from
into the pelvis (see Chapter 32). There is only the inferior gluteal trunk, is harder to confirm,
minimal variation in the distance between the often requiring some guess work. Further
aortic and the iliac bifurcations, making the difficulties may arise from the presence of a
identification of the internal iliac artery easy. In hematoma which can alter the appearances and
contrast, a number of variations in the distribu- distribution of these arteries.
tion of the branches of the internal iliac artery The middle rectal and the inferior rectal
are possible5,6. The proximal bifurcation of the arteries originate from the inferior gluteal and
internal iliac produces two trunks that are the internal pudendal arteries, respectively.
commonly termed the anterior and posterior These supply the middle and lower portions of
branches. The posterior branch supplies the the rectum, anal canal and the perianal skin.
superior gluteal artery, whilst the anterior sup- Theoretically, superselective embolization of
plies the remainder of the pelvis. In the majority the middle rectal or the inferior rectal artery
of instances, the branches of this anterior trunk may result in necrosis of these areas. However,
include the uterine, vaginal, superior cystic, surprisingly such serious complications have not
middle rectal, obturator, internal pudendal and been reported so far.
inferior gluteal arteries (Figure 1a). In 30% of The vaginal artery may originate from the
patients, these arteries have more proximal uterine artery at the level of the cervix or from
origins at the level of the bifurcation of the ante- the inferior gluteal trunk. In addition, the vagina
rior and posterior branches (Figure 1b). This is is also supplied by branches of the internal
especially true with the obturator and uterine pudendal artery.
278
300
30 August 2006 14:22:04
Embolization
TECHNICAL ASPECTS premedication. If no interacting drugs have

been administered, the authors recommend the
Preparation
combination of opiate and sedative antihista-
Unless it is an absolute emergency, obtaining a mines, such as pethidine 50–100 mg i.m. (in
coagulation panel including the platelet count, two divided doses if more than 50 mg is given)
APTT and PT (INR) is worthwhile (see Chapter and promethazine hydrochloride 25–50 mg i.m.
25). Deranged coagulation does not necessarily
contraindicate arteriography or embolotherapy7;
Location for embolization and arterial
however, its correction may help in preparation
puncture
for post-procedural hemostasis and the preven-
tion of complications relating to this. Occult The best location for embolization is the
coagulopathy may also be revealed8. As emboli- interventional suite where vascular procedures
zation is an invasive procedure, informed routinely take place. However, interventional
consent from the patient is essential, with expla- radiologists may be requested to perform proce-
nation and discussion of the possible complica- dures in surgical theaters in some emergency
tions, future fertility and the effects of the situations.
radiation. In situations where the patient is The optimal method in embolization is
sedated or unable to consent, the appropriate to achieve superselective catheterization of the
consenting process should be considered. Ideally, arterial branches that are the sources of hemor-
the patient is kept nil by mouth for an appropri- rhage, such as the uterine arteries on both sides.
ate duration prior to procedure in order to avoid When this is not possible, temporary occlusion
complications from vomiting. Bladder catheter- of the internal iliac arteries using balloon
ization is not essential, although it is helpful catheters is an option to stabilize the patient’s
in preventing the bladder from filling with condition and facilitating subsequent surgical
contrast-containing urine during the procedure. procedures. Removal of a uterine compression
pack may be attempted under such transient
arterial occlusion. If the temporary occlusion
Cross-sectional imaging
has been performed outside the angiography
Localization and measurement of the size of suite (such as in the operating theater) in an
the hematoma prior to arteriography and emergency, the patient could be subsequently
embolization can be extremely useful, although transferred to the angiography suite for proper
not essential. Confirming whether the hema- embolization. In some cases, temporary bilat-
toma is within or outside the uterus and its eral occlusion of the iliac bifurcations may be
relationship to pelvic structures will dictate the performed using angioplasty balloon catheters
course of the embolization (Figures 2a and b). placed and inflated at the iliac bifurcation bilat-
Magnetic resonance imaging (MRI) is the best erally. Acute ischemia of the lower limbs will
test of the pelvis, requiring a small number of occur as a result. The risk of injury to the ner-
examinations with different radiofrequency vous and muscular systems of the lower limb is
signal maneuvers (sequences), demonstrating minimized by shorter occlusion time of external
the sagittal, coronal and axial cross-sections. iliac arteries. Occlusion times of less than 1–2 h
It is recommended to include both T1- and are safe; irreversible injury may occur if it is
T2-weighted sequences in two to three exami- more than 6 h.
nations, such as T1-weighted coronal and
T2-weighted sagittal scans. Should MRI be
The order of arteriogram and catheter
unavailable, either computed tomography (CT)
maneuvers
or ultrasound examination may be an option.
At the puncture site in the groin, an introducer
sheath is used to stabilize the arterial entrance.
Premedication
The standard diameter of the sheath is 5 French
The interventional radiologist needs to gauge; a 6 French gauge sheath is necessary for
decide the type and quantity of agents for balloon occlusion.
279
301
30 August 2006 14:22:04
(a) (c)
(b)
Figure 2a–g Case study: a 23-year-old woman, who had been diagnosed to have double uterus and
double vagina, with the right uterus having been removed several years before. Following vaginal delivery at
full term weeks, she became anemic, with the hemoglobin measuring approximately 6.0 g/dl. Intrapelvic
pain was reported, mainly on the left. Hemorrhage per vagina was only of a moderate degree. (a and b)
T2-weighted magnetic resonance images of the pelvis, in coronal (a) and axial (b) cross-sections. A
hematoma (H) is detected in the left pelvic floor. The right side of the pelvis is preserved. R, rectum and
adjacent tissue; B, bladder. It was anticipated that left-sided embolization would achieve hemostasis based
on these images. (c) Whole pelvic arteriography. The right common femoral artery was punctured and a 5
French gauge hook-shaped catheter was inserted to the distal aorta (Ao) where radiological contrast was
infused. The outline of the common, internal and external iliac arteries (CIA, IIA and EIA, respectively)
and their major branches are demonstrated. The intramural branches of the uterine artery (UA) distribute
both above and within the pelvis. The hematoma is shown as a relatively hypovascular zone (H). (d) Left
internal iliac arteriography in the left anterior oblique position (LAO). Identification of the uterine and
vaginal arteries (UA and VA, respectively) is achieved: the origin of the uterine artery (UAO) is shown. The
superior and inferior gluteal trunks are superimposed (*). This falls into the category of vascular anatomy
shown in Figure 1b. A 5 French cobra-shaped catheter is used. (e) Left uterine arteriography. Superselective
catheterization was achieved using a 3 French gauge catheter inserted through the 5 French cobra-shaped
catheter. The intramural branches with their characteristic tortuosity are shown. Although no extravasation
is demonstrated, unilateral and partial embolization using grated particles of gelatine sponge was performed
in view of increased hemorrhage per vagina and the anatomical communication between the uterine artery
and the arteries to the upper vagina. (f) Left vaginal arteriography. Extravasation is clearly revealed
(arrowheads) on hand injection of radiological contrast through the 3 French catheter (arrow). Embolization
was performed using grated particles of gelatine sponge until the extravasation was barely detectable.
(g) Left inferior gluteal arterial trunk post-embolization. The uterine artery (UA) and a smaller number of
its intramural branches are opacified, the vaginal artery and the branches to the hematoma are no longer
opacified. Following embolization, the hemorrhage per vagina reduced to within normal losses; hemoglobin
increased to 11 g/dl on the next day and 12 g/dl on the following day. The patient was discharged 2 days
post-embolization without undergoing any other intervention; outpatient follow-up confirmed satisfactory
recovery continued
280
302
30 August 2006 14:22:09
Embolization
(d) (f)
(e) (g)
Figure 2a–g Continued
The first arteriogram is an image of the pelvis A 4 or 5 French gauge Cobra tip is a suitable
from the aortic bifurcation to the groins, in standard catheter for superselective access to
order to obtain a global view of the pelvic the uterine artery and other smaller branches, if
arteries (Figure 2c). A range of hook-shaped the hook catheter is inadequate for super-
catheters are useful, as they are helpful in selective catheterization (Figure 3). It is prefera-
accessing the common and internal iliac arteries bly made of soft polyurethane. 5 French gauge
on either side (Figure 3). Subsequently, the catheters have a risk of causing spasm when
internal iliac artery is selectively catheterized inserted into the uterine artery and other
and its arteriogram should be obtained (Figure branches of the inferior gluteal trunk. This can
2d). Oblique views may aid demonstration of be prevented and treated by nitrate vasodilators,
the uterine artery origin and facilitate its such as isosorbide dinitrate 0.05–0.20 mg
catheterization. per branch. Where suitable 4 French gauge
281
303
30 August 2006 14:22:18
uterus, the source is likely to be a branch such as

the vaginal or internal pudendal artery. If
branches other than the uterine artery are the
source of hemorrhage, superselective catheteriz-
ation and arteriogram of each branch are
required to assess the extent of extravasation
(Figure 2f). The advent of smaller diameter
catheters and hydrophilic coated guidewires has
made such superselective catheterization less
challenging. Extravasation is unlikely to be
demonstrated on non-superselective angio-
grams such as the global pelvic arteriogram and
the internal iliac arteriogram.
In case extravasation is confirmed, embolic
material is infused to occlude the artery (Figures
2f and g). If extravasation is not proven,
embolization of each of the branches supplying
the region of hemorrhage is performed. Hemo-
stasis can be achieved with embolization of
the regional arteries, including the source of
hemorrhage, even without actual demonstration
Figure 3 Standard catheters of use in of the bleeding artery9,10. The most accurate
embolization. (a) A 5 French gauge hook-shaped demonstration of the flow distribution of
(Modified hook 2 catheter, Merit Medical, USA); transcatheterally infused material is obtained
(b) a 5 French gauge cobra-shaped (Terumo, Japan) with combined angiography C-arm and CT
and (c) a 3 French gauge microcatheter which goes
equipment. Unfortunately, such machines are
through 5 French gauge catheters (Terumo, Japan):
not universally available. Therefore, the inter-
this catheter is coupled with a hydrophilic
polymer-coated floppy guidewire with an angled ventional radiologist needs to judge the vascular
head (arrowhead) anatomy and the distribution of the embolic
material mainly on the basis of the simple
two-dimensional angiography radiographs in
frontal or oblique projections.
catheters are available, they would reduce the
risk of vasospasm. Guidewires with angled tips
and hydrophilic coatings are also extremely Embolic material
useful tools. For difficult branches with steep Practical embolic materials are summarized in
angulation and tortuosity, finer catheters (less Table 1. Gelatine particles are the most com-
than 3 French in diameter) with their own spe- monly used embolic material in embolization
cific fine and floppy wires are indicated (Figure for postpartum hemorrhage as they are expected
3), although they are costly in general. These to dissolve in several weeks’ time, leading to
are fed through the standard catheters and recanalization of the embolized artery. How-
preferably have an angled tip. ever, these are not free from embolic complica-
tions2,11. Other advantages of gelatine particles
include that they are economical and easily
Targets of embolization
available. Where the particle form of gelatine is
The prime target of embolization is the source unavailable, gelatine plate or sponge could be
artery of hemorrhage. Commonly, this is the cut into particles or grated. Despite the popular
uterine artery when the source of hemorrhage is usage of gelatine particles, there is no evidence
in the myometrium, cervix or endometrium to contraindicate the use of permanent embolic
(Figure 2e). If the hemorrhage is due to lacera- material, such as polyvinyl alcohol (PVA)
tion of the birth canal below the level of the particles (Figure 4).
282
304
30 August 2006 14:22:20
30 August 2006 14:22:21
Table 1 Embolic materials
Duration Approximate
Materials of effect size Mechanism of effect Advantages Disadvantages
Particles
Gelatin sponge (cut) temporary 1–5 mm blockage of blood vessel, economic, safe cutting is time-consuming,
inflammatory occlusion (partly) proximal embolization due
to particle size
Gelatin sponge (grated) temporary 0.3–5 mm.5 blockage of blood vessel, economic, safe, easy to make effect could be short-lived,
inflammatory occlusion (partly) proximal embolization could
occur
Polyvinyl alcohol (PVA) permanent 100–700 µm blockage of blood vessel, readily available could be expensive, proximal
inflammatory occlusion (partly) embolization could occur
Autologous blood clot temporary 1–5 mm blockage of blood vessel, available by adding thrombin in duration of effect is
inflammatory occlusion (partly) vitro unreliable
283
Autologous blood clot permanent < 1 mm to 5 mm blockage of blood vessel, available by heating or exposing to
305
(degenerate) inflammatory occlusion (partly) alcohol or its derivatives
Liquid
Alcohol permanent destruction of blood vessel by economic and ultra-potent painful, unwanted vessels
ablating the intima, producing could be affected
degenerate thrombi
Ethanolamine oleate permanent similar to alcohol but milder similar to alcohol but milder similar to alcohol but milder
effect effect; the effect could be effect
stabilized when combined with

gelatin sponge
Glues such as permanent blockage of blood vessel readily available, stable requires expertise in the use
cyanoacrylate of material; adherence of the
delivery catheter could occur
Embolization
(3) Sciatica This is described above.

(4) Infection Intra-pelvic abscess formation14,15,
post-embolic pyrexia and pain/tenderness
in the pelvis are frequently observed, all
of which can be managed with anti-
inflammatories and antibiotics.
(5) Coagulopathy Difficult hemostasis at the
groin may be a result of coagulopathy.
(6) Acute intra-arterial thrombosis of the lower limb
This may be due to limited arterial flow in
the lower limb following arterial puncture
and catheter maneuver; thrombosis and
Figure 4 Embolization materials. (a) Gelatin occlusion of the lower limb artery may
sponge; (b) grater for gelatine sponge; (c) grated occur2. The risk is increased when balloon
gelatine sponge; and (d) polyvinyl alcohol (PVA) occlusion is performed for a long period.
particles in a bottle syringe
(7) Ischemia of the lower limb This is described
above.
Embolic material should not be infused into
the inferior gluteal artery for the reason des- (8) Radiation The biological effect of radiation
cribed above. In spite of this, there are reports has been studied from the data of measured
where infusion of gelatine particles into the absorption doses of the skin and estimated
inferior gluteal trunk either did not result in doses to the ovaries in a series of 20 cases of
sciatic nerve symptoms3 or only in a minority uterine artery embolization16. In this study,
of instances2. It is assumed that the amount fluoroscopy was performed up to a maxi-
of embolic material infused is the key factor mum of 52.5 min with a mean of 21.9 min,
as to whether sciatica presents or not. Even if resulting in a maximum skin dose of
superselective catheterization is achieved, care 304 cGy (mean 162 cGy). The estimated
needs to be taken to minimize overflow of maximum ovarian dose was 65 cGy (mean
embolic material. As embolization of a branch 22.3 cGy). These figures were greater than
approaches completion, some overflow is usu- the doses of other image examinations of
ally unavoidable. Particular caution is necessary the pelvis such as hysterosalpingography
when liquid embolic material is used, such as (0.04–0.55 cGy), recanalization of the
cyanoacrylate, alcohol and its derivatives. Fallopian tube (0.2–2.75 cGy), computed
tomography of the body trunk (0.1–1.9 cGy);
on the other hand, they were smaller than
COMPLICATIONS the dose in radiotherapy for intrapelvic
The reported frequency of complications is Hodgkin’s lymphoma (263–3500 cGy). On
small. The causes of complications include: the basis of the known risks of pelvic
irradiation for Hodgkin disease, the dose
(1) Technical errors These include hematoma associated with uterine artery embolization
at the puncture site (groin)12 and vascular is unlikely to result in acute or long-term
injury13. Allergic reactions to iodine con- radiation injury to the patient or to a mea-
trast and nephrotoxicity are also possible. surable increase in the genetic risk to the
(2) Post-embolic ischemia Infarct and necrosis of patient’s future children. In embolization
the uterus requiring hysterectomy11, as well for postpartum hemorrhage, there may
as the cervix and upper vagina2 and blad- be cases where longer fluoroscopy time
der11 have been reported. A decision is required than uterine artery only
between surgical and conservative manage- embolization; however, it would be still in
ment needs to be made in each case. the similar region to that of uterine artery
284
306
30 August 2006 14:22:26
Embolization
embolization, and, therefore, the injury appointment with the interventional radiology
from irradiation in embolization is unlikely. team. Such a protocol should be established
with input from both the obstetricians and
(9) Fertility A 35-month follow-up survey on six
interventional radiologists. It would include
patients, who underwent uterine artery
a list of the resources required, including
embolization with polyvinyl alcohol (PVA)
the personnel involved, the equipment, the
particles for therapy of fibromyomata and
consumables and the setting. It should also
wished subsequent conception, confirmed
make consideration for out-of-hours emergency
eight pregnancies in five patients (83%),
work and the case load. Therefore, the protocol
seven births including five transvaginal and
will depend on the requirements and resources
two Cesarean deliveries, and an abortion
of each specific department.
due to the patient’s request for termina-
tion17. The authors of this study concluded
that uterine embolization with PVA parti- CONCLUSION
cles did not affect fertility. In embolization
for postpartum hemorrhage, although the Though embolization has had a relatively short
non-uterine artery branches could be life of practice, it is a highly feasible, safe and
embolized and the embolic material could beneficial procedure, as it may preclude an indi-
be others than PVA, the effect on fertility is cation for further laparotomy and hysterectomy.
unlikely. Therefore, embolization should be the choice
of treatment prior to surgical intervention,
anywhere in the world, when the first line of
LOGISTICS conservative treatment fails.
Postpartum hemorrhage is essentially an emer-

gency situation, which may arise at any time. ACKNOWLEDGEMENTS
The incidence of truly intractable hemorrhage is The authors are grateful to Dr Neel Patel,
small, and, in the majority of cases, there is time Oxford, Dr Thejavanthi Narayan, Milton
during which the obstetricians perform the first Keynes (MK), Mrs Deborah Lee-Smith, Super-
line of treatment, including transfusion, and intendent Radiographer (MK), Mrs Corinne
wait for preparation by the interventional Ward and her colleagues, Radiology Sister/
radiology team. However, urgent intervention is nurses (MK) and the PACS team (MK), for
requested in the minority of cases. This could their help in preparation of the manuscript.
cause a strain in the management of staff in the
interventional radiology department. It could
also be a reason why embolization has not been References
widely recognized or discussed among the 1. Vedantham S, Goodwin SC, McLucas B, Mohr
obstetricians and radiologists as the choice G. Uterine artery embolization: an underused
of treatment, despite a number of successful method of controlling pelvic hemorrhage. Am J
reports both in postpartum and post-Cesarean Obstet Gynecol 1997;176:938–48
cases8–10,12–15,18–27. Nevertheless, the safety, 2. Ojala K, Perala J, Kariniemi J, et al. Arterial
feasibility and low complication rate of embol- embolization and prophylactic catheterization
ization cannot be emphasized enough. The idea for the treatment for severe obstetric hemor-
of offering embolization is simply kinder to the rhage. Acta Obstet Gynecol Scand 2005;84:
patient compared to hysterectomy or other 1075–80
3. Hansch E, Chitkara U, McAlpine J, et al. Pelvic
surgical intervention. The ability to offer
arterial embolization for control of obstetric
embolization would require an obstetric depart-
hemorrhage: a five-year experience. Am J Obstet
ment which is well aware of the implications Gynecol 1999;180:1454–60
of embolization in postpartum uterine hemor- 4. Pelage J, Sorer P, Repiquet D, et al. Secondary
rhage. Such a change in thinking will invariably postpartum haemorrhage: treatment with selec-
necessitate a proactive protocol providing easy tive arterial embolization. Radiology 1999;212:
access for the obstetricians to an emergency 385–9
285
307
30 August 2006 14:22:26
5. Ito T. The pelvis. In Kaibougaku kougi (Lectures polyvinyl alcohol particles for patients with uter-
in Anatomy, in Japanese), 1st edn. Tokyo: ine fibroid or adenomyosis. Cardiovasc Intervent
Nanzando, 1983:475–84 Radiol 2005;28:611–15
6. Lippert H, Pabst R. Arterial Variations in Man: 18. Heaston DK, Nineau DE, Brown BJ, Miller FJ.
Classification and Frequency, 1st edn. Munich: Transcatheter arterial embolization for control
Bergmann, 1985 of persistent massive puerperal hemorrhage after
7. Porteous AO, Appleton DS, Hoveyda F, Lees bilateral surgical hypogastric artery ligation. Am
CC. Acquired haemophilia and postpartum J Roentgenol 1979;133:152–4
haemorrhage treated with internal pudental 19. Pais SO, Glickman M, Schwartz PE, Pingoud E,
embolisation. Br J Obstet Gynaecol 2005;112: Berkowitz R. Embolization of pelvic arteries
678–9 for control of postpartum hemorrhage. Obstet
8. Heffner LJ, Mennuti MT, Rudoff JC, Gynecol 1980;55:754–8
McLean GK. Primary management of post- 20. Minck RN, Palestrant A, Chemey WB. Success-
partum vulvovaginal hematomas by angiographic ful management of postpartum vaginal hemor-
embolization. Am J Perinatol 1985;2:204–7 rhage by angiographic embolization. Ariz Med
9. Yamashita Y, Harada M, Yamamoto H, et al. 1984;41:537–8
Transcatheter arterial embolization of obstetrica 21. Rosenthal DM, Colapinto R. Angiographic arte-
and gynaecological bleeding: efficacy and clinical rial embolization in the management of postop-
outcome. Br J Radiol 1994;67:530–4 erative vaginal hemorrhage. Am J Obstet Gynecol
10. Abbas FM, Currie JL, Mitchell S, Osterman F, 1985;151:227–31
Rosenshein NB, Horowitz IR. Selective vascular 22. Ito M, Matsui K, Mabe K, Katabuchi H,
embolization in benign gynaecologic conditions. Fujisaki S. Transcatheter embolization of pelvic
J Reprod Med 1994;39:492–6 arteries as the safest method for postpartum
11. Porcu G, Roger V, Jacquier A, et al. Uterus and hemorrhage. Int J Gynaecol Obstet 1986;24:
bladder necrosis after uterine artery embolisation 373–8
for postpartum haemorrhage. Br J Obstet 23. Shweni PM, Bishop BB, Hansen JN, Subvayen
Gynaecol 2005;112:122–3 KT. Severe secondary postpartum haemorrhage
12. Bakri YN, Linjawi T. Angiographic embolization after caesarean section. S Afr Med J 1987;72:
for control of pelvic genital tract hemorrhage. 617–19
Acta Obstet Gynecol Scand 1992;71:17–21 24. Finnegan MF, Tisnado J, Bezirdjian DR, Cho S.
13. Greenwood LD, Glickman MG, Schwartz PE, Transcatheter embolotherapy of massive
Morse SS, Denny DF. Obstetric and non- bleeding after surgery for benign gynecologic
malignant gynaecologic bleeding: treatment with disorders. J Can Assoc Radiol 1988;39:
angiographic embolization. Radiology 1987;164: 172–7
155–9 25. Yamashita Y, Takahashi M, Ito M, Okamura H.
14. Chin HG, Scott DR, Resnik R, et al. Angio- Transcatheter arterial embolization in the
graphic embolization of intractable puerperal management of postpartum hemorrhage due
hematomas. Am J Obstet Gynecol 1989;160: to genital tract injury. Obstet Gynecol 1991;77:
434–8 160–3
15. Gilbert WM, Moore TR, Resnik R, et al. 26. Mitty HA, Sterling KM, Alvarez M, Gendler R.
Angiographic embolization in the management Obstetric hemorrhage: prophylactic and emer-
of hemorrhagic complications of pregnancy. Am gency arterial catheterization and embolo-
J Obstet Gynecol 1992;166:493–7 therapy. Radiology 1993;188:183–7
16. Nikolic B, Spies JB, Lundsten MJ, Abbara S. 27. Joseph JF, Mernoff D, Donovan J, Metz SA.
Patient radiation dose associated with uterine Percutaneous angiographic arterial embolization
artery embolization. Radiology 2000;214:121–5 for gynaecologic and obstetric pelvic hemor-
17. Kim MD, Kim NK, Kim HJ, Lee MH. Preg- rhage: a report of three cases. J Reprod Med 1994;
nancy following uterine artery embolization with 39:915–20
286
308
30 August 2006 14:22:27
31
CONSERVATIVE SURGICAL MANAGEMENT
C. B-Lynch
INTRODUCTION in 19973. Along with later modifications by

Hayman and colleagues11 and Cho and col-
A key factor in the surgical management of
leagues12, this13 may prove more effective than
postpartum hemorrhage is the awareness of pre-
radical surgery for the control of life-threatening
disposing factors1–3 and the readiness of thera-
postpartum hemorrhage3,11,12. Although sub-
peutic teams consisting of obstetric, anesthetic
total and total abdominal hysterectomy are still
and hematology staff3,4.
available and indeed useful in their own right,
In the past, the surgical management of post-
they should be considered as a last resort.
partum hemorrhage included use of an intra-
Common causes of postpartum hemorrhage
uterine pack, with or without thromboxane5,
are listed in Table 1, which is not to mean that
thrombogenic uterine pack6, ligation of uterine
additional causes cannot or do not exist.
arteries7, ligation of internal iliac artery8,
Most, if not all, are considered in references
stepwise devascularization9 and, finally, sub-
to postpartum hemorrhage in modern standard
total or total abdominal hysterectomy10. Most
textbooks of obstetrics and further described
of these are discussed in detail in other chapters
in the other chapters of this volume. Three
of this text.
important points merit attention.
A more conservative procedure, now collo-
First, there is significant increase in cardiac
quially known as the Brace suture technique,
output in pregnancy in accordance with red cell
was first described by B-Lynch and colleagues
Table 1 Common causes of postpartum hemorrhage
Uterine overdistention, atony

and disseminated intravascular Disorders of placenta, uterine and
Pre-existing conditions coagulation (DIC) genital tract trauma
Thrombocytopenic purpura Polyhydramnios Acute uterine inversion

Hypertensive disease Multiple gestation Lower segment Cesarean section
Uterine myoma Macrosomia Operative vaginal delivery
Anticoagulation therapy Prolonged labor Precipitate delivery
Coagulation factor deficiency Chorionamnionitis Previous uterine surgery
Systemic disease of hemorrhagic Tocolytic agents Internal podalic version
nature Halogenated anesthetic agents Breech extraction
Consumptive coagulopathy High parity Mid-cavity forceps
Müllerian malfunction Abruptio placentae Obstructed labor
Anemia Courvelliar’s uterus Abnormal fetal presentation
Placenta previa Vacuum site extraction
Placenta accreta, increta, percreta Placental subinvolution
Retained products of conception
Ruptured uterus
287
309
30 August 2006 14:22:27
mass and plasma volume, which provides a according to established labor ward protocols,
compensative reserve for acute blood loss and which involve the anesthetists, hematologists,
hemostatic response following massive hemor- the obstetric team and intensive care support
rhage14. Second, the arrangement of the uterine (see Chapters 13 and 22).
muscle fibers, vis-à-vis the course of the uterine
arteries, facilitates the use of compression
NEW DEVELOPMENTS IN
techniques for effective control of postpartum
THERAPEUTIC OPTIONS
hemorrhage and, finally, conservative treatment
such as bimanual compression of the uterus The type of surgical intervention depends
may control blood loss (Figure 1), whilst upon several factors, paramount of which is the
intensive resuscitative measures are undertaken experience of the surgeon. Other factors include
parity and desire for future children, the extent
of the hemorrhage, the general condition of the
patient and place of confinement. Women at
high risk of postpartum hemorrhage should
not be delivered in isolated units or units
ill-equipped to manage sudden, life-threatening
emergencies. Immediate access to specialist
consultant care, blood products and intensive
care are essential.
The B-Lynch suture compression

technique
The procedure was first performed and
described by Mr Christopher B-Lynch, a
consultant obstetrician, gynecological surgeon,
Fellow of the Royal College of Obstetricians
and Gynaecologists of the UK and Fellow of the
Royal College of Surgeons of Edinburgh, based
at Milton Keynes General Hospital National
Health Service (NHS) Trust (Oxford Deanery,
UK), during the management of a patient with a
massive postpartum hemorrhage in November
1989. This patient refused consent to an emer-
gency hysterectomy3! Table 2 provides an audit
summary of five case histories of other patients
with severe life-threatening postpartum hemor-
rhage managed with this technique.
Figure 1 Bimanual compression of the uterus,
illustrating the first-line approach to mechanical
hemostasis. This in itself might control bleeding The principle
significantly by assisting the uterus to use its The suture aims to exert continuous vertical
anatomical and physiological properties such as the compression on the vascular system. In the
cross-over interlinked network of myometrial fibers
case of postpartum hemorrhage from placenta
for vascular compression and bleeding control. The
previa, a transverse lower segment compression
patient should be placed in stirrups or frog-legged
position in the labor ward or in theater whilst suture is effective.
intravenous fluid and/or appropriate blood product
runs freely. In some cases and commonly so, there The technique2–4
may be failure to achieve satisfactory and lasting
hemostasis by this method See Figures 2a (i and ii), 2b and 2c.
288
310
30 August 2006 15:00:04
Conservative surgical management
Surgeon’s position In outlining the steps and bimanual compression performed. To do

involved, we assume that the surgeon is right- this, the bladder peritoneum is reflected inferi-
handed and standing on the right-hand side of orly to a level below the cervix (if it has been
the patient. A laparotomy is always necessary to taken down for a prior LSCS, it is pushed down
exteriorize the uterus. A lower segment trans- again). The whole uterus is then compressed by
verse incision is made or the recent lower placing one hand posteriorly with the ends of
segment Cesarean section suture (LSCS) the fingers at the level of the cervix and the other
removed to check the cavity for retained hand anteriorly just below the bladder reflec-
placental fragments and to swab it out. tion. If the bleeding stops on applying such
compression, there is a good chance that
Test for the potential efficacy of the B-Lynch suture
application of the B-Lynch suture will work
before performing the procedure The patient is
and stop the bleeding.
placed in the Lloyd Davies or semi-lithotomy
Even in the presence of coagulopathy,
position (frog leg). An assistant stands between
bimanual compression will control diffuse
the patient’s legs and intermittently swabs the
bleeding points. If this test is successful, the
vagina to determine the presence and extent of
application of the suture will also succeed.
the bleeding. The uterus is then exteriorized
(a(i)) (b)
(a(ii))
(c)
Figure 2a–c Summary of the application of the B-Lynch procedure
289
311
30 August 2006 15:00:52
Table 2 Audit summary of five selected case histories of patients with severe life-threatening postpartum hemorrhage treated by ecbolics and the B-Lynch
brace suture application in the period 1989–1995 at Milton Keynes General Hospital, UK2
30 August 2006 14:22:47

Age Presenting Mode of Infant sex and Apgar score at Treatment and Intensive care
(years) Parity GA diagnosis delivery weight (g) 5 and 10 min Type of PPH volume transfused admission Outcome
28 PP 39/40 placental spontaneous male (2800) 4, 7 primary ecbolics, 48 h; full good; 3 years later
abruption, vertex 20 units fresh antibiotic spontaneous
PPH, DIC blood, 8 units cover vertex delivery;
FFP female (3890 g);

no problems
22 PP 43/40 prolonged emergency male (4190) 7, 10 primary ecbolics, 48 h; full good; normal CT
labor, CS 13 units blood, antibiotic pelvimetry 2 years
persisting 5 units packed cover later; elective CS
occipito cells, BSA at 39 weeks;
position?, female (3820 g);
cephalopelvic no problems
disproportion
290
312
23 PP 37/40 eclampsia emergency (1) female (1) 3, 8 primary ecbolics, 72 h; full good; no
(twin) in labor, CS (2735), (2) 5, 8 19 units blood, antibiotic complications
PPH, DIC (2) female 5 units FFP, cover
(2430) BSA
35 PP 38/40 major elective female (3370) 9, 10 secondary, ecbolics, 72 h; full good; no
(IVF) placenta CS 9th day 15 units blood, antibiotic complications
previa readmission 5 units FFP, cover
BSA

30 PP 40/40 uterine atony spontaneous female (3890) 9, 10 primary ecbolics, 48 h; full good; no
vertex 15 units blood, antibiotic complications
7 units packed cover
cells, BSA
PP, primiparous; GA, gestational age in weeks; PPH, postpartum hemorrhage; CS, Cesarean section; CT, computerized tomography; DIC, disseminated
intravascular coagulations; BSA, brace suture application; IVF, in vitro fertilization; FFP, fresh frozen plasma
However, application of the B-Lynch suture is it will prevent suture slipping and uterine
not a substitute for the medical treatment of trauma. The suture now lies horizontally on
coagulopathy, which should take place along the cavity side of the posterior uterine wall.
with the operative intervention (see Chapter
(5) The fundus As the needle pierces the uterine
25).
cavity side of the posterior wall, it is placed
Suture application Given that the test criteria over the posterior wall, bringing the suture
for the B-Lynch suture placement are met, the over the top of the fundus and onto the
uterus remains exteriorized until application anterior right side of the uterus. The needle
of the suture is complete. The senior assistant re-enters the cavity exactly in the same way
takes over in performing compression and as it did on the left side, that is 3 cm above
maintains it with two hands during the place- the upper incision and 4 cm from the lateral
ment of the suture by the principal surgeon. side of the uterus through the upper inci-
sion margin, into the uterine cavity and
(1) First stitch relative to the low transverse then out again through 3 cm below the
Cesarean section/hysterotomy wound. With the lower incision margin (Figure 2a(ii)).
bladder displaced inferiorly, the first stitch
is placed 3 cm below the Cesarean section/ (6) Later role of the assistant The assistant main-
hysterotomy incision on the patient’s left tains the compression as the suture material
side and threaded through the uterine cav- is milked through from its different portals
ity to emerge 3 cm above the upper incision to ensure uniform tension and no slipping.
margin approximately 4 cm from the lateral The two ends of the suture are put under
border of the uterus (Figure 2a(i)). tension and a double throw knot is placed
for security to maintain tension after the
(2) The fundus The suture is now carried over lower segment incision has been closed by
the top of the uterus and to the posterior either the one- or two-layer method.
side. Once situated over the fundus, the
suture should be more or less vertical and (7) Relation to the hysterotomy incision The ten-
lie about 4 cm from the cornu. It does not sion on the two ends of the suture material
tend to slip laterally toward the broad liga- can be maintained while the lower segment
ment because the uterus has been com- incision is closed, or the knot can be tied
pressed and the suture milked through, first, followed by closure of the lower seg-
ensuring that proper placement is achieved ment (Figure 2c). If the latter option is cho-
and maintained (Figure 2a). sen, it is essential that the corners of the
hysterotomy incision be identified and stay
(3) The posterior wall The location on the sutures placed before the knot is tied. This
posterior uterus where the suture is placed ensures that, when the lower segment is
through the uterine wall is actually easy closed, the angles of the incision do not
to surface mark posteriorly. It is on the escape it. Either procedure works equally
horizontal plane at the level of the uterine well. It is important to identify the corners
incision at the insertion of the uterosacral of the uterine incision to make sure no
ligament (Figure 2b). bleeding points remain unsecured, particu-
larly when most of these patients are hypo-
(4) Role of the assistant As the operation pro-
tensive with low pulse pressure at the time
ceeds, the assistant continues to compress
of the B-Lynch suture application.
the uterus as the suture is fed through the
posterior wall into the cavity. This will (8) Post-application and hysterotomy closure It is
enable progressive tension to be maintained probable that the maximum effect of suture
as the suture begins to surround the uterus. tension lasts for only about 24–48 h.
Assistant compression will also help to pull Because the uterus undergoes its primary
the suture material through to achieve max- involutionary process in the first week after
imum compression, without breaking it, vaginal or Cesarean section delivery, the
at the end of the procedure. Furthermore, suture may have lost some tensile strength,
291
313
30 August 2006 14:22:48
but hemostasis would have been achieved

by that time. There is no need for delay in
closing the abdomen after the application of
the suture. The assistant standing between
the patient’s legs swabs the vagina again
and can confirm that the bleeding has been
controlled.
Application after normal vaginal delivery If
laparotomy is required for the management of
atonic postpartum hemorrhage, hysterotomy is
necessary to apply the B-Lynch suture. Hystero-
tomy will also allow exploration of the uterine
cavity, exclude retained products of conception,
evacuate large blood clots and diagnose abnor-
mal placentation and decidual tears, damage
and bleeding. B-Lynch suture application or
any modification of it (see below) without
hysterotomy or re-opening of the Cesarean sec-
tion wound runs the potential risk of secondary
postpartum hemorrhage. Therefore, confirma-
tion that the uterine cavity is completely empty
is essential. Furthermore, hysterotomy ensures
that the correct application of the suture pro-
vides maximum and even distribution of the
compressive effect during and after application
of the B-Lynch suture (Figures 2 and 3). Also,
it avoids blind application of the suture and
the possibility of obliteration of the cervical
and/or uterine cavities that may lead to clot
retention, infected debris, pyometria, sepsis and Figure 3 The in vivo effect of correct application
morbidity3,11,12,15. of the B-Lynch surgical technique seen immediately
after successful suture application. No congestion,
Application for abnormal placentation The no ischemia and no ‘shouldering’ of the sutures at
B-Lynch suture may be beneficial in cases the fundus
of placenta accreta, percreta and increta. In a
patient with placenta previa, a figure-of-eight
or transverse compression suture to the lower
anterior or posterior compartment or both is 10 days after her B-Lynch suture for suspected
applied to control bleeding. If this is not intestinal obstruction (unpublished data,
completely successful, then, in addition, the B-Lynch). Magnetic resonance imaging and
longitudinal Brace suture component may be hysterosalpingography were performed on one
applied for further/complete hemostasis3. patient, showing no intraperitoneal or uterine
sequelae16 (Figure 4a–c). No complications
have been observed in the five patients of the
first published series2 (see Table 1). Moreover,
POSTOPERATIVE FOLLOW-UP
all have succeeded in further pregnancy and
Three patients from the original series had delivery17,18.
laparoscopy postoperatively for sterilization, Tables 3–5 lists the clinical points of the
suspected pelvic inflammatory disease or B-Lynch surgical technique, the Hayman uter-
appendicitis. One patient who had a history of ine compression suture (see Figure 5) and the
ileostomy for surgical reasons had laparotomy Cho multiple square sutures (see Figure 6).
292
314
30 August 2006 15:01:29
(a)
(b)
Figure 4a–c Normal MRI 6 months after massive postpartum hemorrhage treated by B-Lynch surgical
technique followed by uneventful spontaneous vertex vaginal delivery 22 months later. (a) Sagittal view
showing normal endometrial cavity and treated Cesarean incision site; (b) coronal view, with no uterine
cavity synechiae19; (c) view at level of incision for Cesarean section, showing well-healed features
293
315
30 August 2006 14:23:02
(c)
Figure 4a–c continued
Table 3 The B-Lynch surgical technique: clinical points
1. User-friendly suture material monocryl No.1 mounted on 90-cm curved ethigard blunt needle
(codeW3709) (Ethicon, Somerville, NJ). Other rapidly absorbable sutures can be used according to the
surgeon’s preference. A good length and needle are essential19
2. Basic surgical competence required
3. Uterine cavity checked, explored and evacuated
4. Suture bends maintain even and adequate tension without uterine trauma or ‘shouldering’
5. Allows free drainage of blood, debris and inflammatory material
6. Transverse compression suture applied to the lower segment for abnormal placentation effectively
controls bleeding
7. Simple, effective and cost-saving
8. Fertility preserved and proven3
9. Mortality avoided3
10. World-wide application and successful reports (> 1300) (B-Lynch, personal data base,
christopherbl@aol.com)
11. Potential for prophylactic application at Cesarean section when signs of imminent postpartum
hemorrhage develop, e.g. placenta accreta, or where blood transfusion is declined, e.g. placenta previa
surgery on a Jehovah’s Witness
WORLD-WIDE REPORTS
The Indian subcontinent has the largest number
The current level of application of the B-Lynch of reported successful applications, over 250,
suture world-wide includes over 1300 success- followed by Africa, South America, North
ful cases; of these, there are only 19 failures. America, Europe and other countries. The
294
316
30 August 2006 14:23:05
Table 4 The Hayman uterine compression suture: Obstetricians and Gynaecologists in the UK,
clinical points and the Cochrane Database of systematic
reviews. To date, no serious adverse outcomes
1. Lower uterine segment or uterine cavity not
have been associated with the B-Lynch surgical
opened
2. Uterine cavity not explored under direct vision technique3,17,20,22,24. Furthermore, the latest
3. Probably quicker to apply 2000–2002 Triennial Confidential Enquiry
4. No feed-back data on fertility outcome states that no deaths were reported in women
5. Morbidity feed-back data limited who had had interventional radiology or
6. Unequal tension leads to segmented ischemia B-Lynch suture in the management of
secondary to slippage of suture – ‘shouldering’ postpartum hemorrhage23.
with venous obstruction It is important to remember that, if a patient
is a known or appreciated risk for postpartum
hemorrhage, then the elective delivery should be
Table 5 The Cho multiple square sutures: clinical performed in the day time, with prearranged
points co-operation between the imaging department
and the obstetric team. Theater staff should be
1. Multiple full-thickness square sutures applied,
alerted in time so that conservative surgery can
probably time-consuming if many square
sutures required be carried out quickly if needed. Patients at
2. Uterine cavity drainage restriction – pyometra particular risk are those with obesity, cardio-
risk15 myopathy, coagulopathy, abnormal placenta-
3. No feed-back data on fertility outcome tion, polyhydramnios and specific religious
4. Morbidity feed-back data limited convictions contraindicating blood transfusion.
5. Rhythmic contraction not facilitated and
involution impeded
6. The production of multiple uterine senechiae PLACEMENT OF LIGATURES IN
(see Chapter 24) STEPWISE DEVASCULARIZATION
The essential requirements are not simple and
may not be available in every unit. First, there is
a need for a competent obstetrician who is con-
17 reported failures were because of delay in versant and competent at pelvic gynecological
application, poor technique, defibrination and procedures, and who has a working knowledge
inappropriate material. Various suture materials of the pelvic anatomy, including the vascular
have been used. However, the monocryl suture and neurological supply of the pelvic organs.
(code WC3709) is recommended because it is Second, there is a need for an obstetric anesthe-
user- and tissue-friendly with uniform tension tist, as well as a vascular and/or gynecological
distribution and is easy to handle20. Holtsema cancer surgeon on standby. Finally, provisions
and colleagues recently opined, in a review, that must be available for admission postoperatively
the B-Lynch technique for postpartum hemor- to the intensive care unit.
rhage should be an option for every gyne- This set of requirements takes full account
cologist21. Wohlmuth and colleagues published of the extraordinarily generous blood supply to
outcome of a large series with a 91% success the uterus and the pelvic organs (see Figure 7).
rate (world-wide cumulative success rate The surgical approach starts with ligature of the
98%)22. uterine artery and its distribution to the uterus,
preferably as it emerges from crossing over the
ureter or as it approaches the uterine wall to
CONCLUSION
penetrate and establish its division25. This could
Of the compression suturing techniques des- be carried out unilaterally or bilaterally about
cribed above, the B-Lynch procedure has been 2 cm from the uterine angle at Cesarean section
recommended by the 2000–2002 Triennial or where the lower segment is opened after con-
Confidential Enquiry into Maternal Deaths in servative surgery for postpartum hemorrhage
the United Kingdom23, The Royal College of has failed (Figure 8).
295
317
30 August 2006 14:23:05
It is absolutely essential to remember that achieved by independent ligation sutures

the internal iliac (hypogastric artery) gives applied bilaterally to the cervix and/or vagina.
off independent branches that descend to the The ovarian vascular supply to the uterus is
cervix and vagina (vaginal branch), respectively also ligated, either unilaterally or bilaterally.
(see Chapter 32). Devascularization can be Unilateral or bilateral ligature of the internal
Figure 5 The Hayman uterine compression suture Figure 6 The Cho multiple square sutures
without opening the uterine cavity11 compressing anterior to posterior uterine walls12
Figure 7 Placement of ligatures in the process of stepwise devascularization, including ligature of the
descending uterine and vaginal arteries
296
318
30 August 2006 15:02:04
Figure 8 The complex vascular distribution to the pelvic organs. In this procedure of stepwise
devascularization, the patient must be in the Lloyd Davis or modified lithotomy position, with one of the
assistants able to access and swab the vagina to assess bleeding control
iliac artery may become necessary as a further control of massive post partum hemorrhage:
step to control massive postpartum hemor- an alternative to hysterectomy? Five cases
rhage. A skilful surgeon should aim to ligate the reported. Br J Obstet Gynaecol 1997;104:
anterior division of the internal iliac artery in 372–5
3. B-Lynch C, Cowen M.J. A new non-radical
order to achieve further devascularization of
surgical treatment of massive post partum
the uterus without compromising blood supply
hemorrhage. Contemp Rev Obstet Gynaecol 1997;
to the posterior division. However, ligation of March:19–24
the internal iliac directly could be done unilater- 4. Chez RA, B-Lynch C. The B-Lynch suture
ally or bilaterally without devascularizing the for control of massive post partum hemorrhage.
pelvic organs8,26. This may save time, life and Contemp Obstet Gynaecol 1998;43:93–8
organ. 5. Day LA, Mussey RD, DeVoe RW. The intra-
uterine pack in the management of post partum
hemorrhage. Am J Obstet Gynecol 1948;55:
References 231–43
1. Drife J. Management of primary post partum 6. Bobrowski RA, Jones JB. A thrombogenic
haemorrhage. Br J Obstet Gynaecol 1997;104: uterine pack for post partum hemorrhage. Obstet
275–7 Gynecol 1995;85:836–7
2. B-Lynch C, Coker A, Lawal AH, Cowen MJ. 7. Waters EG. Surgical management of post
The B-Lynch surgical technique for the partum hemorrhage with particular reference to
297
319
30 August 2006 15:02:49
ligation of uterine arteries. Am J Obstet Gynecol hysterectomy in the management of severe

1952;64:1143–8 post-partum haemorrhage. J Obstet Gynaecol
8. Evans S, McShane P. the efficacy of internal 2005;25:143–9
iliac artery ligation in obstetric hemorrhage. Surg 18. Tsitpakidis C, Lalonde A, Danso D, B-Lynch C.
Gynaecol Obstet 1985;160:250–3 Long term anatomical and clinical observations
9. Abdrabbo SA. Step-wise uterine devasculariza- of the effects of the B-Lynch uterine compression
tion: a novel technique for management of suture for the management of post partum
uncontrollable post partum hemorrhage with the hemorrhage – ten years on. J Obstet Gynaecol
preservation of the uterus. Am J Obstet Gynecol 2006; in press
1999;171:694–700 19. Wu HH, Yeh GP. Uterine cavity synechiae after
10. Baskett TF. Surgical management of severe hemostatic square suturing technique. Obstet
obstetric hemorrhage: experience with an Gynecol 2005;105:1176–8
obstetric hemorrhage equipment tray. J Obstet 20. Price N, B-Lynch C. Technical description of
Gynaecol Can 2004;26:805–8 the B-Lynch suture for treatment of massive
11. Hayman RG, Arulkumaran S, Steer PJ. Uterine hemorrhage and review of published case. Int J
compression sutures: surgical management of Fertil Womens Med 2005;50:148–63
post partum hemorrhage. Obstet Gynecol 2002; 21. Holtsema H, Nijland R, Huisman A, Dony J,
99:502–6 van den Berg PP. The B-Lynch technique for
12. Cho JH, Jun HS, Lee CN. Hemostatic post partum haemorrhage: an option for every
suturing technique for uterine bleeding during gynaecologist. Eur J Obstet Gynaecol Reprod Biol
cesarean delivery. Obstet Gynecol 2000;96: 2004;115:39–42
129–31 22. Wohlmuth C, Gumbs J, Quebral-Ivie J. B-Lynch
13. Roman A, Rebarbar A. Seven ways to control suture, a case series. Int J Fertil Womens Med
post partum haemorrhage. Contemp Obstet 2005;50:164–73
Gynaecol 2003;48:34–53 23. Department of Health. Why Mothers Die:
14. WHO Report of Technical Working Group. Report on Confidential Enquiries into Maternal
The Prevention and Management of Post Partum Deaths in the United Kingdom 2000–2002
Haemorrhage. Geneva: World Health Organisa- Triennial Report. London: RCOG Press, 2004:
tion, 1999;WHO/MCH/90–7 94–103
15. Ochoa M, Allaire AD, Stitely ML. Pyometra 24. Allam MS, B-Lynch C. The B-Lynch and
after hemostatic square suture technique. Obstet other uterine compression suture techniques. Int
Gynecol 2002;99:506–9 J Gynaecol Obstet 2005;89:236–1
16. Ferguson JE, Bourgeois FJ, Underwood PB, 25. O’Leary JA. Uterine artery ligation in the control
B-Lynch C. Suture for post partum hemorrhage. of post-caesarean hemorrhage. J Reprod Med
Obstet Gynecol 2000;95:1020–2 1995;40:189–93
17. El-Hammamy E, B-Lynch C. A worldwide 26. Clarke SL, Koonings P, Phelan JP. Placenta
review of the uses of the uterine compres- accreta and prior cesarean section. Obstet Gynecol
sion suture techniques as alternative to 1985;66:89–92
298
320
30 August 2006 14:23:26
32
INTERNAL ILIAC (HYPOGASTRIC) ARTERY LIGATION
C. B-Lynch, L. G. Keith and W. B. Campbell
BACKGROUND HISTORY impairment of function of the pelvic viscera. If

the procedure is performed correctly, there is no
The historical background of ligature of the
morbidity, either short- or long-term7.
internal iliac artery for the control of hemor-
Historically, the practice of internal iliac
rhage is not clear1. Numerous publications have
ligation was within the competency of most
attributed the procedure to different surgeons
obstetricians and gynecologists. Today, how-
in diverse specialties world-wide2–4. In the
ever, subspecialization means that their training
United Kingdom and the United States,
and experience may be insufficient, so pelvic
the operation was reported before 1900 and,
floor specialists or vascular surgeons are often
since then, many surgeons have practiced it and
called upon when internal iliac artery ligation is
found it useful.
required.
Howard Kelly first pioneered ligation of the
internal iliac (hypogastric) artery in the treat-
ment of intraoperative bleeding from cervical INTRODUCTION
cancer prior to this technique being applicable
to postpartum hemorrhage5. Studies have In 1963, Lane and Aldemann reported that
shown that, in postpartum hemorrhage, the hemorrhage was one of the major causes of
reduction of pulse pressure may only be maternal mortality in the United States8. East-
achieved in 48% of cases. It is for this reason man correlated this in an extensive review of the
that other workers have advocated bilateral liga- literature9. Any obstetrician who attends and
tion of the internal iliac arteries to significantly experiences a case of severe postpartum hemor-
improve the chances of reducing pelvic pulse rhage clearly understands the risk of losing a
pressure and facilitate hemostasis5. Reported patient from hemorrhage. The memory will last
complications include nerve injury, inadvertent for ever. Modern methods offer the likelihood of
ligature of the common iliac artery, prolonged resuscitation and survival through competent
blood loss and prolonged operative time. It has management by medical means or conservative
also been reported that there is a high rate surgery before such patients reach the point of
of complication and low rate of success for exsanguination10. However, when it becomes
hemostasis if the procedure is not done obvious that conservative methods have failed,
correctly6. Therefore, this procedure should be unilateral or bilateral internal iliac artery
reserved for hemodynamically stable patients of ligation should be considered urgently5,11,12.
low parity in whom future child-bearing itself is
of paramount concern.
ANATOMICAL CONSIDERATIONS
Unilateral or bilateral hypogastric artery
ligation can be life-saving in patients with The pelvic vasculature is arranged in such a
massive postpartum hemorrhage6,7. Although manner that there is ample collateral circula-
surgeons may be reluctant to perform bilateral tion13 (Tables 1–3). The common iliac artery
hypogastric artery ligation for fear of injury to bifurcates into two main branches – the external
the pelvic viscera, there is no evidence that iliac artery (which becomes the femoral artery at
this is the case or that there is any significant the inguinal ligament) and the internal iliac
299
321
30 August 2006 14:23:26
(hypogastric) artery which descends into the anterior superior iliac spines. The bifurcation of
true pelvis. The latter divides into anterior and the common iliac artery is found at the level
posterior branches. It is essential to identify this of both of these landmarks in the majority of
division because the uterine artery branches off patients. Reich and co-workers in 196414 used
from the anterior division. dissection of fresh cadavers to show that numer-
ous variations occur in the anatomy of these ves-
sels. It is not always true, for example, that one
Clinical anatomy
or both of the internal iliac (hypogastric)
The level of bifurcation of the common iliac branches of the common iliac artery are of
artery is quite constant, and there are two similar diameter along the entire length. There-
easily identifiable guides. These are the sacral fore, visual observation alone can be misleading.
promontory and a line drawn between both Likewise, there may be some difference in the
length and diameter of the right and left internal
iliac arteries. Surgeons should therefore be
Table 1 Branches of the internal iliac artery aware of the fact that subdivision of the main
internal iliac trunk may be into branches that
Posterior division Anterior division
are not significantly narrower than the main
Parietal Visceral trunk.
Iliolumbar Umbilical The important anatomical relations of
Lateral sacral Superior vesical the internal iliac (hypogastric) artery can be
Superior gluteal Middle hemorrhoidal summarized as follows:
Obturator Uterine
Internal pudendal Vaginal (1) Anterior medial – covered by peritoneum
Inferior gluteal (the internal iliac artery is entirely retro-
peritoneal);
Table 2 Anastomoses of internal iliac arteries
Branch of internal iliac Anastomotic vessels
1. The uterine arteries 1. Right and left ovarian arteries (direct branches of the aorta)
2. Inferior and middle hemorrhoidal 2. Superior hemorrhoidal artery (branch of inferior mesenteric)
3. Pubic branches of obturator 3. Inferior epigastrics (branch of external iliac)
4. Inferior gluteal 4. Circumflex and perforating branches of the deep femoral artery
5. Superior gluteal 5. Lateral sacral (posterior branches)
6. Iliolumbar 6. Lumbar artery (from aorta)
7. Lateral sacral 7. Middle sacral
8. Vesical arteries 8. Branches of the uterine and vaginal arteries
Table 3 Major pelvic anastomoses
Vertical
1. Ovarian artery (branch of aorta) with the uterine artery
2. Superior hemorrhoidal artery (branch of inferior mesenteric) with middle hemorrhoidal artery
3. Middle hemorrhoidal artery with inferior hemorrhoidal (branch of internal pudendal from hypogastric)
4. Obturator artery with inferior epigastric artery (branch of external iliac)
5. Inferior gluteal artery with circumflex and perforating branches of deep femoral artery
6. Superior gluteal artery with lateral sacral artery (posterior branches)
7. Lumbar arteries with iliolumbar artery
Horizontal
1. Branches of vesical arteries from each side
2. Pubic branches of obturator from each side
300
322
30 August 2006 14:23:27
Internal iliac artery ligation
(2) Anterior – the ureter (retroperitoneal and prior to total or subtotal hysterectomy when all
attached to the peritoneum); conservative measures have failed.
Patients also considered to be at high risk for
(3) Posterolateral – the external iliac vein and
recurrent postpartum hemorrhage, those with
the obturator nerve;
recurrent major placenta previa, or Jehovah’s
(4) Posteromedial – the internal iliac vein; Witnesses with important risk factors may be
candidates for prophylactic internal iliac liga-
(5) Lateral – the psoas major and minor
tion. Good clinical judgement is essential and, if
muscles.
prophylactic ligation is thought to be the best
course, then it should not be delayed.
Physiology of internal iliac artery ligation
Because of the excellent collateral circulation Therapeutic
in the pelvis, vascular compromise does not
Therapeutic ligation may become necessary:
occur when one or both internal iliac arteries
are ligated. At one time, ligation of the hypo- (1) Before or after hysterectomy for post-
gastric system was regarded as equivalent to partum hemorrhage;
shutting off all the blood to the area. Fortu-
(2) Where bleeding continues from the base
nately, this is not true. If it were, it is likely
of the broad ligament;
that the procedure would not be harmless. In
reality, the hypogastric artery distal to the point (3) Where there is profuse bleeding from the
of ligation is never emptied of blood because pelvic side-wall;
the rich anastomotic network starts to function
(4) Where there is profuse bleeding from the
immediately after ligation15. What does occur
angle of the vagina;
is the virtual abolition of the arterial pulse
pressure. This is associated with reduced mean (5) Where areas of diffuse bleeding are pres-
blood pressure and rate of blood flow in the ent without a clearly identifiable vascular
collateral system. As a result, the trip-hammer bed;
effect of arterial pulsations is abolished. The
(6) In the case of ruptured uterus in which the
surgeon must be aware that bilateral ligature of
uterine artery may be torn at the site of its
the internal iliac artery is more effective than the
origin from the internal iliac artery;
unilateral procedure in that the patient has less
chance of returning to theater for secondary (7) When there are additional indications
surgery to control hemorrhage. The reduced including atony of the uterus where
pressure and lack of pulsation do, however, conventional methods have failed;
mean that thrombosis in the vessels may remain
(8) Where extensive lacerations of the
in situ.
cervix have occurred following difficult
instrumental delivery;
INDICATIONS FOR LIGATION OF THE
(9) Where there is significant bleeding from
INTERNAL ILIAC ARTERY
the lower part of the broad ligament;
Prophylactic
(10) When there are gunshot wounds to the
Differentiation between prophylactic and thera- lower abdomen;
peutic use of internal iliac artery ligation is by no
(11) In the case of fracture of the pelvis and
means absolute. Conditions that may indicate
intraperitoneal hemorrhage.
ligation as a prophylactic measure include post-
abortion bleeding, postpartum hemorrhage, In such circumstances, hysterectomy alone
atonic uterus prior to hysterectomy, abruptio may not be sufficient to control hemorrhage.
placenta with uterine atony, abdominal preg- Internal iliac artery ligation, unilateral or bilat-
nancy with pelvic implantation of the placenta, eral, may become necessary and should not be
placenta accreta with intractable bleeding, and delayed in such life-threatening situations.
301
323
06 September 2006 16:30:41
SURGICAL TECHNIQUES peritoneum. If the uterus is present, this inci-

sion can be started in the peritoneum on the
General considerations
posterior surface of the round ligament, at the
All obstetric surgeons should be fully aware of junction of the middle and medial thirds.
the indications, timing and technical aspects of The incision is extended proximally for about
unilateral or bilateral hypogastric artery ligation. 10 cm. If the uterus is absent, the incision can
Experimental evidence by Burchell has be started over the external iliac artery and car-
shown that it is the abolition of the ‘trip- ried proximally to the level of the bifurcation.
hammer effect’ of arterial pulsations that allows Another method is to incise into the peritoneum
effective clotting to take place, so that small directly over the bifurcation. The incision is
vessels stop bleeding2,3. This may explain why then extended distally a few centimeters. All
bilateral ligation works better than unilateral these incisions have one feature in common:
ligation. they result in the formation of a medial and
Either a mid-line or a transverse abdominal lateral peritoneal flap. The ureter is always
incision may be used. The surgeon should not beneath the medial flap and may be visualized,
use an unfamiliar incision. A transverse incision reflected, and protected with ease. The ureter
may take more time, especially in obese normally crosses the common iliac artery from
patients. Visualization is considerably better lateral to medial at a point just proximal to the
from the opposite side of the pelvis. To work on bifurcation.
the contralateral side, the surgeon may elect to Once the peritoneum is opened, loose areolar
change sides during the operation. tissue is separated from it by blunt dissection in
In most situations, bilateral ligation is prefer- the direction of the vessels, not across them to
able to unilateral ligation. Not only is hemo- avoid unnecessary trauma. Small pieces of den-
stasis more secure, but, in addition, it allows tal cotton (‘pledgets’ on long, curved forceps)
greater confidence in making a decision not to are effective. The fingers also may be used.
re-explore the patient. Although it is possible to When the areolar tissue has been separated, the
perform the operation by the extraperitoneal bifurcation comes into view. If the arteries are
approach, the intra-abdominal approach is difficult to find, feel for a pulse (but remember
preferable except in cases of extreme obesity. that pulses may be difficult to palpate if a
Some surgeons advocate complete tran- patient is hypotensive). The bifurcation feels
section of the internal iliac artery between two like an inverted Y. The branch coming off at
ligatures. This has no practical or physiologic right angles is the hypogastric (internal iliac)
advantage. On the contrary, its practice may artery. It courses medially and inferiorly to the
lead to injury of the underlying veins. palpating finger. The continuing branch is the
The choice of material for ligating the artery external iliac artery. It courses laterally and
depends on the preference of the surgeon. For superiorly out over the psoas muscles to the leg,
example, 1–0 Vicryl and umbilical artery tape where it becomes the femoral artery.
have been used. Two ties should be placed The surgeon must accurately identify these
firmly but gently in continuity approximately two branches because inadvertent ligation of
0.5 cm apart and 0.5–1 cm below the the external iliac artery will produce an acutely
bifurcation (Figure 1). ischemic leg, and limb loss is then a risk. If the
external iliac artery is ligated, the ligature can be
cut but it must then be checked for adequate
Transabdominal approach
flow, because the inner layer of the wall may
The abdomen is opened and the viscera packed have been disrupted. If the artery has been
away in the usual manner. Identification of the transected, then it needs to be formally repaired
bifurcation of the common iliac artery is made and a graft may be required. The attendance of
by the two bony landmarks: the sacral promon- a vascular surgeon becomes essential.
tory and an imaginary line drawn through The common and internal iliac arteries are
both anterosuperior iliac spines. A longitudinal often adherent to the underlying veins which
incision is made into the posterior parietal can be difficult to see, particularly beneath the
302
324
30 August 2006 14:23:27
origin of the internal iliac artery. This is the rectus muscle is exposed and opened below
most hazardous part of the operation. Good the level of the umbilicus, dissection caudal
retraction of the pelvic contents and displace- to the semilunar line of Douglas is performed,
ment of the arteries are needed to visualize and the peritoneal and preperitoneal fat are
the veins. Meticulous dissection with scissors is separated. The peritoneum and its contents are
required to separate the internal iliac vein from reflected to the right (or left), thus exposing the
the artery if they are adherent. Once a plane retroperitoneal structures16.
has been developed between them, a Mixter or
other fine right-angled forceps, or the forceps
designed by Reich and colleagues13 are gently ESSENTIAL SURGICAL
introduced between them. This is best done CONSIDERATIONS
onto the tip of a finger of the opposite hand,
which allows gentle manipulation of the tips of (1) The ureter crosses the common iliac artery
the closed forceps, while feeling if there is still at the level of its bifurcation;
tissue present which requires division by sharp (2) An incision is made inferolaterally and par-
dissection. Simply pushing the forceps between allel to the ureter, which can be identified
the artery and the vein in an uncontrolled visually for safe identification and dissec-
fashion is dangerous. It is also inadvisable to try tion;
separating the artery and the vein by opening
the tips of the forceps forcibly until a path has (3) Following such incision, the peritoneal flap
been found between them. under which the ureter runs is displaced
The peritoneum should be closed with inter- medially and retracted away (the ureter may
rupted 2–0 Vicryl because a continuous suture be controlled with a sling for safety);
can kink the ureter. The procedure on the (4) The internal iliac at the point of its bifurca-
left pelvic wall may be slightly more difficult tion into the anterior and posterior divisions
because it is frequently necessary to mobilize can be seen and palpated with its vein and
the sigmoid flexure at the ‘white line’ to obtain the obturator nerve. It is extremely impor-
adequate exposure. tant not to damage the internal iliac vein.
The main arterial branch of the internal
Extraperitoneal approach iliac is ideal for identification and ligature
by passing a right angle, blunt-ended
The skin incision in the inguinal area parallels eye needle upon which is threaded a
the course of the external oblique muscle. It non-absorbable suture such as silk of 0
runs 10–15 cm in length in a line 3–5 cm medial caliber or vicryl suture of the same caliber
to the anterosuperior iliac spine. After the fat and passed between the artery and the vein.
and subcutaneous tissues are dissected away,
a muscle-splitting incision exposes the perito-
neum. This is gently reflected medially, together
Postoperative care
with the ureter. Ligation is performed as previ-
ously described. Closure is the same as for a Intensive care is necessary because these women
herniorrhaphy and can be time-consuming if a may be moribund and have required huge blood
bilateral approach is to be carried out. transfusion. Large hematomas or collections of
serosanguineous fluid can be drained through
separate stab wounds. Usually, this is unneces-
Mid-line extraperitoneal approach
sary. Antibiotics are not indicated after ligation
(uncommon)
of the arteries. Their use is dictated only by the
A mid-line extraperitoneal approach to the presence of infection. Early ambulation is advis-
aorta has been advocated16. One hospital able in all cases. An indwelling catheter may be
authority extended its use to bilateral ligation of necessary to facilitate adequate assessment of
the hypogastric arteries. A mid-line abdominal urinary output in women who are at risk of
incision is made. After the anterior sheath of the serious morbidity.
303
325
30 August 2006 14:23:27
(a)
(b)
Figure 1 Ligation of the anterior branch of the internal iliac artery with its associated vein.
(a) Demonstrable vulnerability of internal iliac vein and obturator nerve in close proximity; (b) A ‘skeletal’
anatomy, showing proximity of external iliac artery, ureter and anterior branches of sciatic nerve
304
326
30 August 2006 15:03:24
Special clinical considerations normal uterus. It is important to remember

that pituitary necrosis (Sheehan’s syndrome is
The major pitfall associated with ligation of the
discussed in Chapter 38) can affect the ability to
hypogastric artery is delay. When hemorrhagic
reproduce after postpartum hemorrhage, espe-
shock is irreversible, this operation will not
cially if blood replacement has been delayed or
overcome it. Inadequate transfusion is another
inadequate, hemorrhage has been severe, and
pitfall in the therapy of patients with severe
shock profound. Fortunately, this is not a
hemorrhage. Blood loss is often seriously
common occurrence in many modern and
underestimated.
well-equipped obstetric units.
Failure to remember that the vaginal artery
is a separate branch of the hypogastric artery,
rather than a branch of the uterine artery, may Potential failures and consequences18
lead the surgeon into the pitfall of an unneces-
sary and ineffective hysterectomy for control of Occasionally, ligation of the hypogastric arteries
bleeding. Injury to the external iliac artery from fails to stem pelvic hemorrhage. The reason for
retractors or mistaken ligation of this vessel can this is not clear, but some suggestions are:
lead to lower limb amputation. Also, accidental (1) Massive necrosis after infection with
ligature of one or both ureters would lead to destruction of the vessels;
renal function impairment. Accidental incorpo-
ration of the anterior division of the sciatic nerve (2) The presence of large, aberrant branches
may lead to foot drop (Figure 1b). feeding blood to the area;
Most authors consider internal iliac artery (3) Dislodgement of clots when blood pressure
ligation to be a very safe procedure. The avail- rises;
able data suggest that this operation does not
result in necrosis of vital pelvic structures. The (4) Concomitant severe venous bleeding;
only report to the contrary is by Tajes4 who however, this is rare;
cited a case of his own in which this operation (5) Coagulopathy with deranged hematological
resulted in necrosis of the buttocks. Tajes also indices.
reviewed two previously reported cases: in one
case, the bladder mucosa sloughed, in the other,
scrotal necrosis ensued. However, his report Avoiding accidental ligation of the
was 50 years ago. common or external iliac artery
It is essential to identify the internal iliac artery
clearly. Ligation of the common or external iliac
Maintenance of reproductive function
produces an acutely ischemic leg. The classical
It has not always been possible to follow young signs are whiteness or pallor of the foot and
patients for whom this operation has been per- absence of distal pulses – but these may be diffi-
formed. More important, many patients do not cult to assess in a hypotensive, vasoconstricted
understand the exact nature or extent of their patient. If there is concern that the main artery
operation. A patient may remember only that to the lower limb may have been ligated, check
she was sick and bleeding, that she was operated for a pulse in the external iliac artery above the
on and that she recovered. inguinal ligament, beyond the area of ligation;
The incidence of postoperative amenorrhea the femoral pulse in the groin; or the Doppler
is not known. It is common for menses to signals at the ankle, using a hand-held Doppler.
resume after the operation. There have been If the wrong artery has been tied, the ligature
reports of normal pregnancy and delivery occur- should be removed. If this fails to restore a good
ring after bilateral hypogastric artery ligation, pulse (or if the artery has been transected), a
although it is impossible to say how frequently vascular surgeon should be called to repair the
this occurs. It is entirely reasonable to believe vessel (either by end-to-end anastamosis or
that reproductive capacity is not lost after this with the use of a short bypass graft of vein
operation, provided that the patient has a or synthetic material).
305
327
30 August 2006 14:23:40
Damage to the ureter anterior division of the sciatic nerve into any
ligatures17 (Figure 1b).
Damage to either or both ureters should be
avoided by careful visualization and dissection.
In life-threatening surgery or delayed interven-
tion to control massive hemorrhage, accidental USEFUL HINTS
damage to a ureter may occur. Ligature is more
The position of the surgeon relative to the
probable than transection. Prompt diagnosis
patient
and remedial surgery by a urological colleague
are essential. Accidental ligature of one ureter The surgeon should stand where he/she is
may not lead to renal failure but increase most comfortable and this may be influenced
morbidity. by right- or left-handedness. The choice of the
surgeon’s position also depends on the ability
and dexterity of the assistant. If the assistant is
relatively inexperienced, then it may be particu-
Damage to other vessels larly helpful for the surgeon to changes sides
during the procedure in order to deal with each
Damage to the common or iliac vein or one of
internal iliac artery from the opposite side of the
its major tributaries results in brisk hemorrhage.
operating table.
Its source can be difficult to see and to control.
It can threaten the patient’s life, particularly
in the context of pre-existing major blood loss
from postpartum hemorrhage. Checking for thorough control of bleeding
Steps to avoid damage to the iliac veins before closure of abdomen
have been described in detail above: great care
(1) Whilst the patient is in the frog-leg or Lloyd
should be taken when dissecting in the area
Davis position throughout the operation, an
behind the origin of the internal iliac artery and
assistant stands between the legs and swabs
when separating the arteries from the veins. If
the vagina to confirm bleeding has stopped.
sudden venous bleeding does occur, the first
step should be to apply firm pressure to the (2) The abdomen is examined to ascertain that
area. Adequate suction should be prepared – the ligatures have been correctly placed.
two suction tips may be helpful. Swabs
(3) The posterior abdominal wall peritoneum,
mounted on sponge-holding forceps can then
which had been incised to access the poste-
be applied distal and proximal to the site of
rior abdominal wall, may or may not need
damage to compress the veins and allow the
closure.
defect to be visualized. If the venous defect
cannot be seen, deep in the pelvis behind the (4) The abdomen is checked once again thor-
iliac artery, then transaction of the iliac artery to oughly to ensure all instruments, swabs and
expose the vein may solve the problem. The foreign materials have been removed.
artery can then be re-anastamosed. When the
(5) The abdomen is closed according to type of
defect in the vein has been seen, its edges can
the initial incision, i.e. large, pfannenstiel
be held together using atraumatic forceps such
or mid-line. The mid-line incision is
as Stiles, before being sutured.
commonly closed by the mass closure
Repair of the vein is best performed with a
technique.
non-absorbable vascular suture, such as poly-
propylene on a round-bodied needle. For large (6) The sick patient should be quickly trans-
iliac veins, a 3/0 is a reasonable choice: needles ferred directly to a high-care setting such as
smaller than those supplied with 4/0 sutures can ITU for an appropriate length of time, to
be difficult to retrieve during repair of large ascertain hemostasis, to ensure that pulse
veins and present a small danger of becoming and blood pressure have returned to nor-
‘lost’ inside the vein. Finally, it is most impor- mal, and to permit surveillance of the uri-
tant to avoid incorporating branches of the nary systems with a bladder catheter in situ.
306
328
30 August 2006 14:23:40
(7) Counselling for post-traumatic stress, 9. Eastman NJ. Gleanings from maternal mortality
depression, panic attacks and flashbacks reports. Presented in a lecture at Milwaukee
should be provided. County Hospital and the Department of
Obstetrics and Gynaecology of Marquette
University, February 8, 1963
10. Lynch C, Coker Y, Abu J, et al. The B-Lynch
References
surgical technique for the control of massive
1. Burchell RC, Olson G. Internal iliac artery postpartum haemorrhage: an alternative to
ligation: aortograms. Am J Obstet Gynecol 1966; hysterectomy? Five cases reported. Br J Obstet
94:117 Gynaecol 1997;104:372–5
2. Burchell RC. Hemodynamics of the internal iliac 11. Shafiroff BGP, Grillo EB, Baron H. Bilateral
artery ligation. Presented at the 1964 Clinical ligation of hypogastric arteries. Am J Surg 1959;
Congress of the American College of Obstetricians 98:34
and Gynaecologists 12. O’Leary JA. Uterine artery ligation in the control
3. Burchell RC. Internal iliac artery ligation: of postcesarean hemorrhage. J Reprod Med 1995;
hemodynamics. Obstet Gynecol 1964;24:737 40:189–93
4. Tajes RV. Ligation of the hypogastric arteries 13. Reich WJ, Nechtow MJ, Keith L. Supplementary
and its complications in resection of cancer of report on hypogastric artery ligation in the pro-
rectum. Am J Gastroenterol 1956;26:612 phylactic and active treatment of hemorrhage in
5. Kelly H. Ligation of both internal iliac arteries pelvic surgery. Int Surg 1965;44:1
for hemorrhage in hysterectomy for carcinoma 14. Reich WJ, Nechtow HJ, Bogdan J. The iliac
uteri. Bull John Hopkins Hosp 1894;5:53 arteries: a gross anatomic study based on
6. Evans S, McShane P. The efficacy of internal dissection of 75 fresh cadavers. Clinical surgical
iliac artery ligation in obstetric hemorrhage. Surg correlations. J Int Coll Surg 1964;41:53
Gynecol Obstet 1985;160:250–3 15. Burchell RC, Mengert WF. Internal iliac artery
7. Clark SL, Phelan JP, Yeh S-Y, et al. Hypogastric ligation: a series of 200 patients. J Int Fed Obstet
artery ligation for obstetric hemorrhage. Obstet Gynecol 1969;7:85
Gynecol 1985;66:353–6 16. Shumacker HB Jr. Midline extraperitoneal
8. Lane RE, Andleman SL. Maternal mortality exposure of the abdominal aorta and iliac
in Chicago, 1956 through 1960: preventable arteries. Surg Gynecol Obstet 1972;135:791–2
factors and cause of death. Am J Obstet Gynecol 17. Varner M. Obstetric emergencies (post partum
1963;85:61–9 haemorrhage). Crit Care 1991;7:883–97
307
329
30 August 2006 14:23:40
33
THE PELVIC PRESSURE PACK
G. A. Dildy III
When pharmacologic and surgical interventions ‘bowel bag’11 or packing with dry laparotomy
fail to correct postpartum hemorrhage, hyster- packs12. These methods, however, require
ectomy becomes the option of last resort1. The re-laparotomy after initial stabilization and
incidence of hysterectomy on the Louisiana volume control to remove the packing materials.
State University obstetric service at Charity Other recently reported methods for packing,
Hospital of New Orleans during 1975–1981 not requiring re-laparotomy but with limited
was 1.21 per 1000 deliveries2. Contemporary cumulative obstetric experience, include trans-
reports of the incidence of obstetric hyster- cutaneous placement of an inflated condom
ectomy range between 0.33 and 0.70 per 1000 over a 22-Fr catheter13 or ribbon gauze within a
deliveries3–6. Under these circumstances, a Penrose drain14.
moderately busy obstetric unit with 4000 deliv- In 1926, Logothetopoulos described a pack
eries per year may expect to perform as many as for the management of uncontrolled post-
three emergency hysterectomies annually. hysterectomy pelvic bleeding15. This technique
The maternal mortality associated with has subsequently been called the mushroom,
obstetric hysterectomy is significant (4.0–4.5%) parachute, umbrella, pelvic pressure, or Logotheto-
for a number of reasons, not the least of which poulos pack. It is important to note that this
have been outlined elsewhere and much of pelvic pressure pack described is applied post-
which relates to the often moribund condition hysterectomy, and it should not be confused, as it
of the patient when the operation commences, often is, with uterine packing16, or with various
the difficulty of the procedure itself, especially intrauterine balloons17–19 for treatment of post-
in the presence of factors which make the anat- partum hemorrhage due to uterine atony or
omy unclear, and the extent of the bleeding placental site bleeding.
which may accompany the operation4,5. Indeed, The pelvic pressure pack controls hemor-
Clark and colleagues reported an average blood rhage from large raw surfaces, venous plexuses
loss of 3.5 liters during emergency obstetric hys- and inaccessible areas by exerting well-
terectomy7. As recounted in several other chap- distributed pressure, compressing bleeding
ters in this textbook, severe hemorrhage and areas against the bony and fascial resistance
emergency hysterectomy are often accompanied of the pelvis20,21. According to Parente and
by secondary coagulopathy. In the setting colleagues21, several references to the pelvic
of acquired coagulopathy, post-hysterectomy pressure pack appeared in European medical
bleeding may continue despite secure surgical journals during the decades following the origi-
pedicles, much to the consternation of the sur- nal report. The first reported cases appearing in
geon and the members of the operating team. the English literature were not until the 1960s,
Abdominal and pelvic post-surgical packing and these pertained specifically to gynecologic
is an old concept and one that has been used post-hysterectomy hemorrhage20,21. Several
to control hemorrhage from a variety of case reports and a case series for obstetric
sources, including liver trauma8, pre-eclampsia- post-hysterectomy bleeding have since been
induced hepatic rupture9, rectal cancer10, and published22–26. Table 1 summarizes these, 23
gynecologic cancer surgery11. Various packing cases for control of gynecologic and 13 cases for
methods have been described, such as the obstetric post-hysterectomy hemorrhage, with
308
330
30 August 2006 14:23:40
The pelvic pressure pack
Table 1 Summary of contemporary reported unfavorable outcomes. Nonetheless, successful

cases of the pelvic pressure pack for obstetric and control of hemorrhage seems to have been
gynecologic post-hysterectomy hemorrhage. The achieved in the majority of cases.
success rate is defined as the pelvic pressure pack As seen in Figure 1, the pack is constructed
being the last intervention to control bleeding.
by filling a bag (we prefer a sterile X-ray cassette
Modified from Dildy et al.26
drape, but other materials also have been
Gynecology success Obstetric success described) with gauze rolls tied end-to-end
Series rate rate (in this case, five 11.4 cm × 2.8 m Kerlix rolls),
starting at the ‘dome’ of the pack (A), with the
Parente, 196221 14/14 – ‘tail’ of the gauze protruding from the ‘neck’
Burchell, 196820 8/8 – of the pack (B–D). Gauze should be removed,
Cassels, 198522 – 1/1
as visually indicated, from the pack before
Robie, 199023 – 1/1
placement, in order to fit the true pelvis. The
Hallak, 199124 – 1/1
Howard, 200225 – 1/1 pack is introduced transabdominally into the
Dildy, in press26 1/1 7/9 pelvis (Figure 2), and the ‘neck’ is delivered
transvaginally through the introitus by passing a
Total 23/23 (100%) 11/13 (85%) surgical clamp from below through the vagina.
The surgeon should avoid trapping small bowel
behind the pack. Traction and thereby pressure
success rates of 100% and 85%, respectively. are applied to the pack by tying intravenous
Admittedly, accurate success rates are difficult (i.v.) tubing to the neck of the pack and sus-
to determine based on rare cases collected retro- pending a 1-liter i.v. fluid bag off the foot
spectively, with possible under-reporting of of the bed. A 1-liter glass i.v. bottle and mild
Figure 1 Photograph of a pelvic pressure pack, as constructed from an X-ray cassette drape, sterile gauze
rolls, and an intravenous infusion set-up. Please see text for further explanation
309
331
30 August 2006 14:23:43
coagulopathy, and acidosis required early pack-

ing if they were to survive27. Thus, packing
should be considered early on when homeo-
stasis is significantly altered. Febrile morbidity
is very common in these critically ill postopera-
tive patients who have already received massive
blood component therapy and have a foreign
body placed into a contaminated operative
field26. Prophylactic broad-spectrum antibiotics
should be administered whenever a pelvic pres-
sure pack is placed, and this regimen should be
continued after pack removal until the patient
is afebrile at least 24–48 h. Another study
of abdominal trauma patients showed those
packed for ≤ 72 h had lower abscess, sepsis, and
mortality rates than those packed for > 72 h28.
Thus pack removal should be accomplished as
Figure 2 Diagram of the pelvic pressure pack in soon possible following stabilization.
situ. Please see text for further explanation In summary, the pelvic pressure pack is
simple to construct from commonly available
medical materials, and control of hemorrhage is
Trendelenburg position provide additional successfully achieved in the majority of cases.
weight and traction if needed. The i.v. tubing or If the pelvic pressure pack fails to control
a cord can simply be hung over the foot of the bleeding, other medical29, surgical30, or inter-
bed, or over an orthopedic pulley attached to ventional radiology31 approaches will be neces-
the foot of the bed. Compression of the pack sary to ultimately control bleeding. The pelvic
can also be maintained by placing the ‘neck’ of pressure pack should be particularly useful in
the pack through a #80 doughnut pessary (not developing countries where more advanced sur-
shown), applied flush against the perineum with gical skills for pelvic vascular ligation and tech-
a surgical clamp. However, caution must be nologies, such as selective arterial embolization,
taken to avoid perineal pressure necrosis. are not readily available. In developed coun-
We advise placement of an intraperitoneal tries, however, the pelvic pressure pack may
large-gauge closed-system (e.g. Jackson-Pratt) serve as a temporizing measure pending trans-
drain to monitor for postoperative bleeding. An port to a tertiary-care facility. In the majority of
indwelling urinary catheter allows monitoring of instances, the pelvic pressure pack will afford
urine output and avoidance of urinary outflow transfer of the critically ill patient to a post-
obstruction. After stabilization of the patient, an surgical recovery setting, where restoration of
attempt to remove the pack transvaginally is hemodynamic, temperature, hematologic, and
made by slowly removing the gauze rolls under acid-base homeostasis can be accomplished.
intravenous sedation, to allow gradual decom-
pression without inciting bleeding. The optimal
time to leave the pack in situ will vary, but References
extended placement has certain risks (see
below). Usually transvaginal pack removal is 1. Dildy GA, 3rd. Postpartum hemorrhage: new
management options. Clin Obstet Gynecol 2002;
successful, but in some cases the pack will
45:330–44
have to be removed by re-laparotomy or with
2. Plauche WC, Wycheck JG, Iannessa MJ,
laparoscopic assistance. Rousset KM, Mickal A. Cesarean hysterectomy
In one study of trauma patients suffering at Louisiana State University, 1975 through
intra-abdominal hemorrhage, Garrison and 1981. South Med J 1983;76:1261–3
colleagues found that patients who experi- 3. Baskett TF. Emergency obstetric hysterectomy.
enced hypothermia, refractory hypotension, J Obstet Gynaecol 2003;23:353–5
310
332
30 August 2006 14:23:45
The pelvic pressure pack
4. Eniola OA, Bewley S, Waterstone M, Hooper R, massive postpartum haemorrhage. Br J Obstet

Wolfe CD. Obstetric hysterectomy in a popula- Gynaecol 1994;101:259–60
tion of South East England. J Obstet Gynaecol 18. Johanson R, Kumar M, Obhrai M, Young P.
2006;26:104–9 Management of massive postpartum haemor-
5. Kwee A, Bots ML, Visser GH, Bruinse HW. rhage: use of a hydrostatic balloon catheter to
Emergency peripartum hysterectomy: A pro- avoid laparotomy. Br J Obstet Gynaecol 2001;
spective study in The Netherlands. Eur J Obstet 108:420–2
Gynecol Reprod Biol 2006;124:187–92 19. Bakri YN, Amri A, Abdul Jabbar F.
6. Lau WC, Fung HY, Rogers MS. Ten years Tamponade-balloon for obstetrical bleeding.
experience of caesarean and postpartum Int J Gynaecol Obstet 2001;74:139–42
hysterectomy in a teaching hospital in Hong 20. Burchell RC. The umbrella pack to control
Kong. Eur J Obstet Gynecol Reprod Biol 1997;74: pelvic hemorrhage. Conn Med 1968;32:734–6
133–7 21. Parente JT, Dlugi H, Weingold AB. Pelvic
7. Clark SL, Yeh SY, Phelan JP, Bruce S, Paul RH. hemostasis: a new technic and pack. Obstet
Emergency hysterectomy for obstetric hemor- Gynecol 1962;19:218–21
rhage. Obstet Gynecol 1984;64:376–80 22. Cassels JW Jr, Greenberg H, Otterson WN.
8. Feliciano DV, Mattox KL, Burch JM, Bitondo Pelvic tamponade in puerperal hemorrhage.
CG, Jordan GL Jr. Packing for control of hepatic A case report. J Reprod Med 1985;30:689–92
hemorrhage. J Trauma 1986;26:738–43 23. Robie GF, Morgan MA, Payne GG Jr,
9. Smith LG Jr, Moise KJ Jr, Dildy GA 3rd, Wasemiller-Smith L. Logothetopulos pack for
Carpenter RJ Jr. Spontaneous rupture of the management of uncontrollable postpartum
liver during pregnancy: current therapy. Obstet hemorrhage. Am J Perinatol 1990;7:327–8
Gynecol 1991;77:171–5 24. Hallak M, Dildy GA 3rd, Hurley TJ, Moise
10. Zama N, Fazio VW, Jagelman DG, Lavery IC, KJ Jr. Transvaginal pressure pack for life-
Weakley FL, Church JM. Efficacy of pelvic threatening pelvic hemorrhage secondary to
packing in maintaining hemostasis after rectal placenta accreta. Obstet Gynecol 1991;78:938–40
excision for cancer. Dis Colon Rectum 1988;31: 25. Howard RJ, Straughn JM Jr, Huh WK, Rouse
923–8 DJ. Pelvic umbrella pack for refractory obstetric
11. Finan MA, Fiorica JV, Hoffman MS, et al. hemorrhage secondary to posterior uterine
Massive pelvic hemorrhage during gynecologic rupture. Obstet Gynecol 2002;100:1061–3
cancer surgery: ‘pack and go back’. Gynecol Oncol 26. Dildy GA, Scott JR, Saffer CS, Belfort MA.
1996;62:390–5 An effective pressure pack for severe pelvic
12. Ghourab S, Al-Nuaim L, Al-Jabari A, et al. hemorrhage (submitted)
Abdomino-pelvic packing to control severe 27. Garrison JR, Richardson JD, Hilakos AS, et al.
haemorrhage following caesarean hysterectomy. Predicting the need to pack early for severe
J Obstet Gynaecol 1999;19:155–8 intra-abdominal hemorrhage. J Trauma 1996;
13. Luijendijk RW, Jn IJ, Jeekel J, Bruining HA. An 40:923–7; discussion 927–9
inflated condom as a packing device for control 28. Abikhaled JA, Granchi TS, Wall MJ, Hirshberg
of haemorrhage. Br J Surg 1994;81:270 A, Mattox KL. Prolonged abdominal packing for
14. Awonuga AO, Merhi ZO, Khulpateea N. trauma is associated with increased morbidity
Abdominal packing for intractable obstetrical and mortality. Am Surg 1997;63:1109–12;
and gynecologic hemorrhage. Int J Gynaecol discussion 1112–3
Obstet 2006;93:160–3 29. Bouwmeester FW, Jonkhoff AR, Verheijen RH,
15. Logothetopulos K. Eine absolut sichere van Geijn HP. Successful treatment of life-
blutstillungsmethode bei vaginalen und threatening postpartum hemorrhage with
abdominalen gynakologischen operationen. [An recombinant activated factor VII. Obstet Gynecol
absolutely certain method of stopping bleeding 2003;101:1174–6
during abdominal and vaginal operations.] 30. Clark SL, Phelan JP, Yeh SY, Bruce SR, Paul
Zentralbl Gynakol 1926;50:3202–4 RH. Hypogastric artery ligation for obstetric
16. Maier RC. Control of postpartum hemorrhage hemorrhage. Obstet Gynecol 1985;66:353–6
with uterine packing. Am J Obstet Gynecol 31. Vedantham S, Goodwin SC, McLucas B, Mohr
1993;169:317–21; discussion 321–3 G. Uterine artery embolization: an underused
17. Katesmark M, Brown R, Raju KS. Successful method of controlling pelvic hemorrhage. Am J
use of a Sengstaken–Blakemore tube to control Obstet Gynecol 1997;176:938–48
311
333
30 August 2006 14:23:45
34
PERIPARTUM HYSTERECTOMY
T. F. Baskett
HYSTERECTOMY even in countries with low maternal mortality

rates, associated maternal morbidity can be high
Emergency peripartum hysterectomy is an
due to hemorrhage, blood transfusion, dissemi-
unequivocal marker of severe maternal morbid-
nated intravascular coagulation, infection and
ity and ‘near-miss’ mortality1,2. Reviews of pub-
potential injury to the adjacent lower urinary
lished data in the past 25 years show a variable
tract7,10,11. This chapter describes emergency
incidence, from one in 3313 to one in 6978
hysterectomy in the immediate postpartum
deliveries4. In developed countries, the inci-
period following vaginal or Cesarean delivery.
dence is approximately one in 2000 deliveries,
with one population-based study in a Canadian
province showing an incidence of 0.53 per 1000 INDICATIONS
deliveries2.
Because of the increasing Cesarean section By far the most common indication for hyster-
rate world-wide and the concomitant rise in ectomy is hemorrhage associated with the
placenta previa and placenta previa accreta, following conditions7,9–20.
the incidence of emergency peripartum hyster-
ectomy is rising in many countries. For exam-
Abnormal placentation
ple, in Canada from 1991 to 2000 the rate rose
from 0.26/1000 deliveries to 0.46/1000 deliver- In developed countries, placenta previa, with or
ies (relative risk 1.76; 95% confidence interval without associated accreta, is the most common
1.48–2.08)5. Compared to vaginal delivery, indication for hysterectomy. This is due to the
emergency hysterectomy and delivery by rising incidence of these conditions associated
Cesarean section are strongly associated6,7. In with the increasing number of women previ-
addition, a recent study has shown that multiple ously delivered by Cesarean section. Despite the
pregnancy had a six-fold increased risk of fact that numerous other techniques aimed at
emergency peripartum hysterectomy compared preserving the uterus have been proposed and
to singleton pregnancies8. Within this group, are discussed in other chapters in this book, hys-
higher-order multiple pregnancies (triplets and terectomy is used to stem the sometimes fright-
beyond) had an almost 24-fold increased risk of ening hemorrhage associated with placenta
hysterectomy8. It seems logical to conclude that previa or accreta in the majority of hospitals.
the increase in multiple pregnancy rates associ- In addition, on rare occasions, abruptio
ated with assisted reproductive technology placentae, particularly of the concealed variety,
provides a further contribution to the rising may be associated with such a degree of extra-
peripartum hysterectomy rates. vasation of blood into and through the full
Maternal mortality rates associated with thickness of the myometrium (Couvelaire
emergency hysterectomy range from 0 to 30%, uterus) as to make it unresponsive to oxytocic
with the higher rates in regions with limited drugs, so necessitating hysterectomy. It must
medical and hospital resources9. How valid be emphasized, however, that in the majority
these rates are today is unclear, as they were cal- of cases of abruptio placentae with Couvelaire
culated more than a decade ago. Nonetheless, uterus the response to oxytocic drugs is
312
334
30 August 2006 14:23:45
Peripartum hysterectomy
appropriate and the hemorrhage is due to dis- repair. External traumas, such as assault, a fall
seminated intravascular coagulation rather than or motor vehicle accident, are relatively rare
failure of the uterus to contract. causes of uterine perforation and rupture.
Uterine atony Sepsis

As outlined elsewhere in this book (Chapter 27), In the era of modern antibiotics, sepsis is not a
the range of modern oxytocic drugs has greatly common reason for emergency hysterectomy.
improved the management of uterine atony. However, it still may be necessary in cases with
Nonetheless, there are cases in which the uterus extensive uterine sepsis, particularly with
is refractory to all applications of such agents. clostridial infections and myometrial abscess
This is most commonly found in the prolonged, formation, in which antibiotic treatment fails to
augmented and/or obstructed labor: simply control the sepsis. Other septic causes of sec-
stated, the exhausted and infected uterus may ondary postpartum hemorrhage include Cesar-
respond poorly to oxytocic agents. The majority ean scar infection and necrosis, arteriovenous
of these cases occur at the time of Cesarean sec- fistula formation secondary to uterine trauma
tion for dystocia or cephalopelvic disproportion. and infection, and endomyometritis associated
with hemorrhage. All may rarely require
Uterine rupture hysterectomy.
The most common cause of complete uterine

rupture is within a previous Cesarean section SURGICAL PRINCIPLES
scar. If the rupture is extensive and hemorrhage
Although the technique of obstetric hysterec-
cannot be contained by suture of the ruptured
tomy is similar in principle to that of abdominal
area, then hysterectomy may be necessary. In
hysterectomy in gynecology, numerous anatom-
addition, rupture of the intact uterus can occur
ical and physiological changes in pregnancy
in multiparous women in response to inappro-
create potential surgical difficulties.
priate use of oxytocic agents in the first and
second stages of labor. (1) The uterine and ovarian vessels are
enlarged and distended, often markedly
so, and the adjacent pelvic tissues are
Uterine trauma
edematous and friable.
Traumatic rupture, that is, perforation or lacer-
(2) Abdominal entry may have been via
ation of the uterus, can occur with a variety of
Pfannestiel or lower midline incision,
obstetric manipulations, including internal ver-
depending on the urgency and speed
sion and breech extraction in obstructed labor;
required.
instrumental manipulation, such as the classical
application of the anterior blade of Kielland’s (3) Maneuvers to obtain immediate hemo-
forceps; manual exploration of the uterus and stasis will depend on the cause of the
manual removal of the placenta or its fragments hemorrhage. In cases of uterine rupture,
after obstructed labor with a ballooned and thin Green–Armytage clamps or sponge for-
lower uterine segment; and during curettage for ceps can be used to compress the bleeding
secondary postpartum hemorrhage. edges of torn uterine muscle. The uterus
Cesarean section in the second stage of labor should be eventrated from the abdominal
with the fetal head deeply impacted in the wound. The structures of the adnexa on
vagina may be associated with lateral traumatic each side are pulled laterally by an assis-
extension of the lower uterine segment incision tant and the surgeon applies straight
into the major vessels21. On rare occasions, the clamps adjacent to the top sides of the
extent of this tear may necessitate hysterectomy, uterus to include the round ligament, the
especially if one or both uterine arteries is lacer- Fallopian tube and the utero-ovarian liga-
ated and a hematoma obscures the surgical ment. This serves to control the collateral
313
335
30 August 2006 14:23:45
blood flow to the uterus from the ovarian manipulating the bulb of the Foley cathe-
arteries. Using transillumination, the ter to see if it is visible through the bladder
avascular spaces in the broad ligament, wall. The bladder also can be filled with a
roughly opposite the level of a transverse colored fluid such as methylene blue or
lower Cesarean incision, should be identi- sterile milk taken from the neonatal nurs-
fied and a catheter passed through on each ery. The latter is preferable as it does not
side to encircle the lower uterine segment cause permanent staining of the tissues.
just above the cervix. This should be Thus, after repair of any bladder injury, it
twisted tightly closed with a clamp and is easier to see that this has been successful
should serve to compress the uterine arter- with subsequent installation of milk in the
ies. These two maneuvers should occlude bladder. Any tear in the bladder should be
the main collateral ovarian and uterine repaired with two layers of 3/0 polyglactin
artery supply to the uterus. (vicryl) or equivalent suture. Otherwise,
No. 1 polyglactin (vicryl) or equivalent is
(4) The vascular pedicles are thick and edem- used throughout the procedure.
atous and should be double clamped.
Remove the proximal clamp first and (9) If there is any doubt about the integrity
apply a free tie and then replace the distal of the bladder wall or ureters, and after
clamp with a transfixing suture. The prox- repair of any bladder injury, it is wise to
imal free tie should ensure that there is no perform a postoperative cystoscopy to
hematoma formation in the base of the confirm that they are intact. This can be
pedicle. done by observing urine come from each
ureteric orifice; this may be facilitated
(5) If the cervix and paracolpos are not by giving intravenous indigo carmine and
involved as the source of hemorrhage, waiting 10–15 min.
subtotal hysterectomy should be adequate
to achieve hemostasis and is safer, faster (10) Perioperative antibiotic prophylaxis
and easier to perform than total hysterec- should be continued for 24–48 h.
tomy. However, if the lower segment and Thromboprophylaxis with heparin should
paracolpos are involved in the hemor- be instituted as soon as one is satisfied that
rhage, such as in cases of placenta previa hemostasis is secure.
and/or accreta, total hysterectomy will be (11) Detailed notes should be made to include
necessary for hemostasis. the preoperative events, indications for
(6) Avoid the ureters by placing all clamps hysterectomy and the surgical details.
medial to those used to secure the uterine After the initial postoperative recovery, the
arteries. woman should receive a comprehensive
outline of events from an experienced
(7) It can be difficult to identify the cervix, obstetrician.
particularly when the hysterectomy is
being done at full cervical dilatation. If In a number of series, as many as 25% of
there is a Cesarean incision, a finger can women who received an emergency obstetric
be placed through this and the cervical rim hysterectomy were primigravid, for whom the
palpated. It is safest to enter the vagina fertility-ending nature of the procedure can
posteriorly, identify the rim of the cervix be devastating7. Therefore, particularly in this
and then proceed anteriorly. group of women, obstetricians should be famil-
iar with and be prepared to perform alternative
(8) The bladder is particularly vulnerable in procedures to control the hemorrhage. The
cases previously delivered by Cesarean application of other techniques to arrest hemor-
section, as it may be adherent to the lower rhage that can be both life-saving and uterus-
uterine segment and cervix. It is therefore preserving are outlined in several chapters in
essential to check the integrity of the blad- this book. When conditions are recognized in
der intraoperatively. This can be done by the antenatal period that lead to increased risk
314
336
30 August 2006 14:23:46
Peripartum hysterectomy
of severe obstetric hemorrhage, such as placenta 9. Ozumba BC, Mbagwu SC. Emergency obstetric
previa and/or accreta, referral of these cases hysterectomy in Eastern Nigeria. Int Surg 1991;
to hospitals with the equipment and personnel 76:109–11
to provide the alternative techniques to hyster- 10. Bakshi S, Meyer BA. Indications for and out-
comes of emergency peripartum hysterectomy.
ectomy should be undertaken where feasible.
A five-year review. J Reprod Med 2000;45:733–7
Ultimately, however, one has to strike a bal-
11. Engelsen IB, Albrechsten S, Iverson OE. Peri-
ance between spending excessive time on alter- partum hysterectomy – incidence and maternal
native techniques that are proving ineffective, morbidity. Acta Obstet Gynecol Scand 2001;80:
leading to delay, further hemorrhage and proba- 409–12
ble disseminated intravascular coagulation, and 12. Lau WC, Fung HY, Rogers MS. Ten years expe-
moving to the definitive and life-saving hyster- rience of cesarean and postpartum hysterectomy
ectomy. Such is the art of obstetric judgement in a teaching hospital in Hong Kong. Eur J Obstet
in trying circumstances. Gynecol Reprod Biol 1997;74:133–7
13. Stanco LM, Schrimmer DB, Paul RH, Mishell
DR. Emergency peripartum hysterectomy and
References associated risk factors. Am J Obstet Gynecol 1993;
168:879–83
1. Baskett TF, Sternadel J. Maternal intensive care
14. Tuncer R, Erkaya S, Sipahi T, Kara F. Emer-
and ‘near-miss’ mortality in obstetrics. Br J
gency postpartum hysterectomy. J Gynecol Surg
Obstet Gynaecol 1998;105:981–4
1995;11:209–13
2. Baskett TF, O’Connell CM. Severe obstetric
15. Sebitloane MH, Moodley J. Emergency peri-
maternal morbidity: a 15-year population-based
partum hysterectomy. East Afr Med J 2001;78:
study. J Obstet Gynaecol 2005;25:7–9
70–4
3. Korejo R, Jafarey SN. Obstetric hysterectomy –
16. Sheiner E, Levy A, Katz M, Mazor M. Identify-
five years experience at Jinnah Postgraduate
ing risk factors for peripartum cesarean hysterec-
Medical Centre, Karachi. J Pakistan Med Assoc
tomy. A population-based study. J Reprod Med
1995;45:86–8
2003;48:622–6
4. Yamamoto H, Sagae S, Nishik WA, Skuto R.
17. Abu-Hei JA, Jawlad FM. Emergency peripartum
Emergency postpartum hysterectomy in obstet-
hysterectomy at the Princess Badeea Teaching
ric practice. J Obstet Gynecol Res 2000;26:341–5
Hospital in North Jordan. J Obstet Gynaecol Res
5. Wen SW, Huang L, Liston RM, Heaman M,
1999;25:193–5
Baskett TF, Rusen ID. Severe maternal mortal-
18. Bai SW, Lee HJ, Cho JS, Park YW, Kim SK,
ity in Canada, 1991–2001. Can Med Assoc J
Park KH. Peripartum hysterectomy and associ-
2005;173:759–63
ated factors. J Reprod Med 2003;48:148–52
6. Kacmar J, Bhinmai L, Boyd M, Shah-Hosseini
19. Chew S, Biswas A. Caesarean and postpartum
R, Piepert J. Route of delivery as a risk factor
hysterectomy. J Singapore Med 1998;39:9–13
for emergency peripartum hysterectomy: a case-
20. Castaneda S, Karrison T, Ciblis LA. Peripartum
control study. Obstet Gynecol 2003;102:141–5
hysterectomy. J Perinat Med 2000;28:472–81
7. Baskett TF. Emergency obstetric hysterectomy.
21. Allen VM, O’Connell CM, Baskett TF. Mater-
J Obstet Gynaecol 2003;23:353–5
nal and perinatal morbidity of caesarean delivery
8. Francois K, Ortiz J, Harris C, Foley MR, Elliott
at full cervical dilatation compared with caesar-
JP. Is peripartum hysterectomy more common
ean delivery in the first stage of labour. Br J
in multiple gestations? Obstet Gynecol 2005;105:
Obstet Gynaecol 2005;112:986–90
1369–72
315
337
30 August 2006 14:23:46
35
THE MANAGEMENT OF SECONDARY POSTPARTUM
HEMORRHAGE
K. M. Groom and T. Z. Jacobson
INTRODUCTION ETIOLOGY OF SECONDARY

Secondary postpartum hemorrhage is defined
as excessive vaginal bleeding from 24 h after Subinvolution/uterine atony
delivery up to 6 weeks postpartum1. Unlike
The major cause of secondary postpartum
primary postpartum hemorrhage, there is no
hemorrhage is subinvolution of the uterus. This
clear definition for quantity of blood loss and
results in failure of obliteration of blood vessels
this can vary from ‘increased lochia’ to massive
underlying the placental site, leading to pro-
hemorrhage. The diagnosis is therefore subjec-
longed bleeding. The two main causes of this
tive, which may account for the variation
are infection (see Chapter 44) and inflammation
in reported incidence. The reported overall
(endometritis) and retained placental tissue.
incidence of secondary postpartum hemorrhage
Endometritis is more common following pro-
in the developed world varies from 0.47% to
longed rupture of membranes, prolonged labor,
1.44%2,3.
emergency Cesarean section or with a retained
The etiology of secondary postpartum
placenta requiring manual removal. A history of
hemorrhage is diverse and management is
offensive lochia, maternal pyrexia and uterine
dependent on identifying the cause and tailoring
tenderness is often present and retained placen-
treatment appropriately. The published work
tal tissue is more common in women with a pre-
on the management of secondary is limited
vious history of retained placenta or if there
compared with primary postpartum hemor-
were concerns at the time of delivery of incom-
rhage4. However, with falling maternal mor-
plete placenta and/or membranes. It is less likely
tality rates, there is increasing interest
following delivery by Cesarean section. Differ-
and attention to maternal morbidity and the
entiation between the two causes is often diffi-
important topic of management of secondary
cult and both conditions may co-exist.
postpartum hemorrhage. The majority of cases
of secondary postpartum hemorrhage are asso-
ciated with minor morbidities but may still Lower genital tract trauma
require re-admission to hospital, use of anti-
Missed vaginal lacerations and hematomas may
biotics and surgical intervention. In more
present as secondary postpartum hemorrhage.
extreme cases, major morbidity requiring
These are often associated with traumatic
hysterectomy, arterial ligation or radiological
deliveries or those requiring ventouse or
intervention is possible5 and maternal death
forceps. They usually present within the first
may still result from massive secondary post-
few days after delivery. Infected suture lines and
partum hemorrhage despite the use of all avail-
episiotomy sites may lead to wound breakdown
able interventions.
and result in excessive vaginal bleeding.
316
338
30 August 2006 14:23:46
Management of secondary postpartum hemorrhage
Placental abnormalities cases reported, this has lead to massive

postpartum hemorrhage 2–3 weeks after
Placenta accreta, increta and percreta are all
Cesarean section and the need for subtotal
known causes of massive primary postpartum
hysterectomy12. Diagnosis of uterine dehiscence
hemorrhage. When managed conservatively
post-Cesarean section associated with infection
with placental tissue left in situ (with or without
has also been made at hysteroscopy13, although
methotrexate therapy), they can also be associ-
causing less significant postpartum hemorrhage
ated with delayed bleeding and the need for
and only requiring treatment with antibiotics.
hysterectomy6,7 (see Chapter 24).
Choriocarcinoma
Uterine abnormalities
The majority of cases of choriocarcinoma after a
Fibroids are associated with primary post- non-molar pregnancy present with secondary
partum hemorrhage. They cause uterine postpartum hemorrhage or irregular vaginal
enlargement and prevent involution of the bleeding14. In addition, secondary symptoms
uterus, therefore leading to prolonged bleeding of metastatic disease may be present. The diag-
from the placental bed. More rarely, they can be nosis is made by serum β-human chorionic
associated with secondary postpartum hemor- gonadotropin (β-hCG), histological diagnosis
rhage. Fibroids have usually been identified by and radiological imaging including ultrasound,
ultrasound in the antenatal period. plain film X-ray and computed tomography
Abnormalities of uterine vasculature such (CT) scan.
as arteriovenous malformations and false
aneurysms may also lead to secondary postpartum
hemorrhage. Arteriovenous malformations are Bleeding disorders, coagulopathies and
due to an abnormal communication between an use of anticoagulants
artery and vein with proliferation of each vessel Women with congenital hemorrhagic disorders
with interconnecting fistula. It is believed these such as von Willebrand’s disease (quantitative
malformations may result from venous sinuses or qualitative deficiency of von Willebrand
becoming incorporated in scars within the myo- factor), carriers of hemophilia A (factor VIII
metrium after necrosis of the chorionic villi. The deficiency), hemophilia B (factor IX deficiency)
majority are acquired after pregnancy and may and factor XI deficiency are at an increased risk
result from trophoblastic disease, previous uter- of postpartum hemorrhage. Often, these abnor-
ine curettage, uterine or cervical malignancy8,9 malities of the coagulation system are unde-
or Cesarean section10,11. Diagnosis is made tected until challenged by trauma, surgery or
using ultrasound with color Doppler analysis. childbirth and so may be undiagnosed prior to
pregnancy. These women are not at increased
risk of antepartum hemorrhage15 but at sig-
Cesarean section wound dehiscence or
nificant risk of both primary and secondary
surgical injury
postpartum hemorrhage. The risk of secondary
Surgical injury to pelvic blood vessels at postpartum hemorrhage may be even greater
the time of Cesarean section10 usually presents than primary postpartum hemorrhage as the
within 24 h. However, later presentations, in pregnancy-induced rise in maternal clotting fac-
particular those causing broad ligament tors falls after delivery. The reported incidence
hematomas, have been described5 and should for secondary postpartum hemorrhage in these
be considered in women presenting acutely with conditions is 20–28% for von Willebrand’s dis-
signs of intra-abdominal hemorrhage. Delayed ease, 11% for hemophilia carriers and 24% in
presentation of bleeding from non-union/ factor XI deficiency16–19. Postpartum acquired
dehiscence of the Cesarean section uterine scar hemophilia has also been described. This is a
has also been described. This is believed to rare condition but can cause severe hemor-
be due to local infection at the site of uterine rhage. It is caused by antibodies to factor VIII
closure causing erosion of blood vessels. In the which partially or completely suppress factor
317
339
30 August 2006 14:23:46
VIII procoagulant activity in women with previ- output should be maintained throughout resus-
ously normal levels and activity of factor VIII. citation. Blood and blood products should be
Bleeding usually commences within 3 months given according to blood loss, response to initial
of delivery but may be delayed for up to 12 fluid administration and hemoglobin and
months15. coagulation results. If hemorrhage is life-
The use of anticoagulants in the postpartum threatening, transfusion with uncross-matched
period may also cause delayed bleeding. In O rhesus-negative or type-specific blood may
particular, women using warfarin should be need to be considered. Identification of the
carefully monitored and informed of the risks of cause of bleeding should then be made and
hemorrhage. further management planned accordingly.
In cases of less significant hemorrhage, basic
resuscitation should be instigated as appropriate
MANAGEMENT OF SECONDARY
but blood transfusion may be delayed whilst
establishing a cause for the bleeding.
Evidence regarding the management of second-
ary postpartum hemorrhage is limited. A
Cochrane review searched and assessed all ran- Clinical presentation
domized or quasi-randomized comparisons of
Ninety-five percent of women present within
drug therapies, surgical therapies and placebo
the first month after delivery, 19% within 7
or no treatment for secondary postpartum hem-
days, 41% in 8–14 days, 23% in 15–21 days and
orrhage. Forty-five papers were identified, but
12% in 22–28 days2. The amount of blood loss
none met the inclusion criteria, and the review
at presentation varies but most are hemo-
concluded there was no evidence from random-
dynamically stable. A thorough history will pro-
ized trials to show the effects of treatments for
vide information relating to cause and should
secondary postpartum hemorrhage4.
include details regarding parity, labor, mode of
The main aims of treatment are to provide
delivery, third-stage or puerperal complications
basic resuscitation, establish a cause for the
and any relevant medical and family history.
bleeding, and tailor the treatment (medical
Clinical signs and symptoms at the time of pre-
and/or surgical) according to the cause.
sentation may include offensive lochia, abdomi-
nal cramping, uterine tenderness, pyrexia,
Resuscitation enlarged uterus and an open cervical os.
Normal postpartum loss may continue
Approximately 10% of cases of secondary post-
beyond 6 weeks in up to 25% of women, espe-
partum hemorrhage will present with massive
cially if breast-feeding20 and the first period may
hemorrhage20 and require immediate attention.
be heavy, prolonged and painful as a result of an
In these cases, resuscitation should be com-
anovulatory cycle. Women should be given this
menced prior to establishing a cause and should
information during normal postpartum care to
include the involvement of senior staff at the
avoid unnecessary concern and presentation for
earliest opportunity (see Chapter 20).
medical investigation.
Restoration of circulating blood volume
should be achieved by gaining intravenous
access with two large-bore cannulae and admin-
Investigations
istering intravenous fluids initially with physio-
logical saline (up to 2 liters) and then with Baseline blood tests should include full blood
plasma expanders until blood is available. Blood count, coagulation studies, C-reactive protein, a
should be obtained for full blood count, group and hold specimen and serum β-hCG.
coagulation screen and cross-match. High- Vaginal swabs should be taken at the time of
concentration oxygen (10–15 liters per minute) examination for aerobic as well as anaerobic
should be administered by a tight-fitting mask21. bacterial growth, including swabs from
Close observation of vital signs including pulse, episiotomy or vaginal tear sites. In women with
blood pressure, oxygen saturation and urine signs of infection, a mid-stream urine specimen
318
340
06 September 2006 16:29:02
should be collected and, if maternal tempera- Table 1 Causes of secondary postpartum hemor-
ture is > 38°C, blood cultures should be taken. rhage
Ultrasound imaging of the pelvis should be
Subinvolution of the uterus – retained placental
considered if there are concerns of retained pla-
tissue and/or endometritis, fibroid uterus
cental tissue. If this is within 7–14 days of deliv- Lower genital tract lacerations/hematoma
ery, interpretation may be difficult as remaining Surgical injury
blood clots may appear as mixed echogenic Dehiscence of Cesarean section scar
material in a similar manner to retained tissue. Vascular abnormality – arteriovenous malformation
The use of duplex color Doppler helps to Placental abnormality – placenta accreta, percreta
improve diagnostic accuracy in differentiating and increta
clot and tissue22, as retained placental tissue will Choriocarcinoma
often maintain a blood supply unlike necrotic Coagulapathies, bleeding disorders, use of
decidua and clot23. anticoagulants
The over-diagnosis of retained placental
tissue on ultrasound may lead to unnecessary Table 2 A proposed standardized system for
surgical intervention and its potential complica- reporting postpartum ultrasound scan. Adapted
tions. However, ultrasound does have signifi- from Neill et al., 200224
cant benefits, as it has a good negative predictive
1. Normal endometrial cavity
value and therefore is helpful in excluding a
2. Endometrial cavity containing fluid only
diagnosis of retained placental tissue. Neill 3. Endometrial cavity enlarged (anteroposterior
and colleagues assessed 53 women undergoing (AP) depth > 1 cm). Maximum AP dimensions
ultrasound for secondary postpartum hemor- noted
rhage. Definitive diagnosis of retained placental 4. Endometrial cavity containing echogenic foci.
tissue was either made histologically or, in Dimensions of largest foci noted. Doppler
those women managed conservatively, absence evaluation of blood flow in foci
of retained tissue was assumed if bleeding
diminished within 1 week. They demonstrated
Table 3 The management of secondary
the diagnostic value of ultrasound assessment to
postpartum hemorrhage
have a positive predictive value of 46% (95%
confidence interval (CI) 31–70%), a negative Medical Surgical
predictive value of 96% (95% CI 88–100%), Oxytocics Uterine evacuation
with a sensitivity of 93% (95% CI 80–100%) Prostaglandins Uterine tamponade balloon
and a specificity of 62% (95% CI 48–79%)24. Antibiotics Uterine compression sutures
This study also suggested that a standardized Tranexamic acid Hysterectomy
approach to reporting an ultrasound investiga- Vasopressin Pelvic arterial ligation
Clotting factor
tion of secondary postpartum hemorrhage
concentrates
would be helpful. This is shown in Table 2.
Chemotherapy Radiological
Additional imaging should also be consid- Oral contraceptive Selective arterial
ered for specific causes of secondary postpartum pill embolization
hemorrhage such as plain chest film and CT
scanning for metastases in cases of chorio-
carcinoma, magnetic resonance imaging (MRI) should include resuscitation as discussed above,
for placenta accreta25,26 and angiography for the use of uterotonic agents, administration
intractable bleeding of unknown origin10. of antibiotics and consideration of surgical
evacuation of the uterus.
Treatment
Uterotonic agents
The majority of cases of secondary postpartum
hemorrhage are due to subinvolution of the Syntocinon can be administered as an intra-
uterus caused by uterine infection and/or venous or intramuscular bolus (10 units) or in
retained placental tissue. Initial management combination with ergometrine (Syntometrine®)
319
341
06 September 2006 16:29:34
1 ampoule as an intramuscular injection. This hemorrhage in one study, despite only 36%
can be followed by a syntocinon infusion (40 having proven histological evidence of retained
units in 500 ml normal saline at an infusion rate tissue3. This study was unable to find any clear
of 125 ml/h). Prostaglandin F2α (Haemabate®/ association with presence or absence of retained
Carboprost) can be given by intramuscular tissue at the time of evacuation and day of onset
injection at a dose of 250 µg every 15 min, up to of bleeding or morbidity at the time of second-
a total of 2 mg (i.e. 8 doses). Misoprostol can ary postpartum hemorrhage. However, retained
also be given as an alternative prostaglandin tissue was more likely if membranes were
(400–800 µg orally or rectally). incomplete at delivery, primary postpartum
hemorrhage had occurred or if secondary post-
partum hemorrhage was judged to be heavy or
Antibiotics
moderate (compared with light) in volume3. The
Endometritis is likely to play a significant role in use of ultrasound prior to surgical evacuation of
many cases of secondary postpartum hemor- the uterus does not appear to significantly alter
rhage and the majority of women are prescribed the chances of histological diagnosis confirming
antibiotics. In a 3-year study of almost 20 000 retained tissue. In one study, 33% of those with
women, 132 women (0.69%) had a secondary no preoperative scan had retained placental
postpartum hemorrhage and 97% of these were tissue compared to 37% following a scan2.
treated with antibiotics2. However, only three- Retained placental tissue is likely to be
quarters of these women had microbiological associated with infection and, therefore, broad-
specimens collected and a positive culture was spectrum intravenous antibiotics should be
obtained in only 13.5%. In a similar observa- given in conjunction with surgical evacuation.
tional study of 83 women with secondary As serum concentrations of most antibiotics
postpartum hemorrhage, 45% presented with peak 1 h after intravenous administration, these
pyrexia, and 64 had bacteriological swabs taken, should be administered just prior to surgery20;
of which only 12.5% were positive. Organisms however, in women who are hemodynamically
identified included group B streptococcus, stable, it may be appropriate to give 12–24 h
bacteroides, E. coli, Clostridium perfringens and of antibiotic cover prior to consideration of
Lancefield group D streptococcus. Despite the surgery1. At the time of surgery, uterotonic
lack of evidence to support the presence of a agents such as syntocinon, ergometrine and
specific bacterial pathogen, 92% of the women prostaglandins may be given to aid uterine
received antibiotics3. Recommended choices contractility and control hemorrhage.
of antibiotic treatment include amoxycillin There is no clear evidence to support which
with clavulanic acid (Augmentin®)27 and a method of evacuation should be used. Manual
combination of amoxycillin, metronidazole and removal of tissue, use of a suction catheter and
gentamicin3. Endometritis is a major contribu- sharp curettage with a metal curette have all
tor to subinvolution of the uterus. Although been described2. The risk of uterine perforation
infection may not be confirmed in a large popu- is much higher in uterine evacuation post-
lation of cases, we recommend that antibiotics partum and may be even further increased if
are always given for secondary postpartum associated with endometritis. Hoveyda and
hemorrhage (see Chapter 44). colleagues describe uterine perforation in three
of 85 women undergoing the procedure for
secondary postpartum hemorrhage. These were
Uterine evacuation
performed from 4 days to 28 days after delivery
Examination under anesthetic and surgical with both a suction and metal curette. In all
evacuation of the uterus should be considered if cases, the procedures were performed by senior
retained placental tissue is suspected clinically medical staff. One woman went on to require a
or after ultrasound examination. This has good hysterectomy, but the other two were managed
reported success rates, with bleeding stopping conservatively2. Perforation after Cesarean sec-
promptly in all 72 women undergoing evacuation is more likely and, as these women have a
tion of the uterus for secondary postpartum lower risk of retained placental tissue, surgical
320
342
30 August 2006 14:23:47
evacuation in these cases should be very Within an Australian population with an

carefully considered. overall incidence of secondary postpartum hem-
Additional complications include the risk orrhage of 1.44% over 15 years, only nine cases
of Asherman’s syndrome. There is limited required hysterectomy (0.9%)3. However, in a
evidence to ascertain if this risk is increased subgroup of women with massive intractable
for postpartum uterine evacuation; however, obstetric hemorrhage, two out of seven with
in a large study of intrauterine adhesions, secondary postpartum hemorrhage required
21.5% of cases had a postpartum curettage hysterectomy. In one of these cases, hysterec-
as a preceding event28. The need for a second tomy was performed 7 days after delivery due to
procedure due to incomplete evacuation of intractable bleeding from lower genital tract lac-
retained tissue may also occur2. Hysterectomy eration but maternal death still resulted. The
may be required to control bleeding in up to 5% second case had further morbidity following her
of cases20. hysterectomy for secondary postpartum hemor-
In view of these significant complications, rhage with bleeding from wound disunion and
women should always be fully counselled of the sepsis and required bilateral hypogastric artery
risks and informed consent obtained prior to ligation 14 days after delivery5. Hysterectomy in
the procedure. Surgery should be performed by such situations carries significant risks but can
experienced senior medical staff. be life-saving and should be considered early in
cases of massive hemorrhage, whether primary
or secondary.
Other surgical procedures
Pelvic artery ligation may also be considered
In the event of a large secondary postpartum for cases of massive secondary postpartum hem-
hemorrhage, other surgical procedures may orrhage uncontrolled by medical and simple
need to be considered. This includes cases of surgical measures. Lédée and colleagues report
bleeding from an infected placental bed or the use of bilateral hypogastric artery ligation in
placental abnormality such as placenta accreta, 49 of 61 cases of intractable hemorrhage; this
bleeding from retained placental tissue not includes four out of seven cases of secondary
controlled with uterine evacuation, non-union/ postpartum hemorrhage, all of which were suc-
dehiscence of Cesarean section scar, bleeding cessful at arresting bleeding5 (see Chapter 32).
from a surgical injury or uncontrolled bleeding As with primary postpartum hemorrhage, arte-
from a lower genital tract laceration. rial ligation should be performed by an experi-
Insertion of an intrauterine tamponade enced surgeon and his/her involvement should
balloon, such as the Bakri29 or Rüsch balloon30, be considered whilst planning a laparotomy in
has been successfully described for treatment of such cases.
primary postpartum hemorrhage and may be
considered in cases of secondary postpartum
Selective arterial embolization
hemorrhage due to uterine subinvolution/atony
once retained placental tissue has been excluded Pelvic angiography to assess the internal iliac
(see Chapters 28 and 29). Laparotomy may also artery, uterine artery and its vaginal branches
be required which allows further investigation is a helpful tool in the assessment of ongoing
into the cause of bleeding and treatment by the hemorrhage (Figure 1). It also allows the
use of surgical compression sutures, hysterec- introduction of embolization agents to arrest
tomy and pelvic arterial ligation as appropriate. bleeding (see Chapter 30).
The B-Lynch brace suture is well described Pelage and colleagues studied 14 women
for the treatment of primary postpartum hemor- presenting with uncontrollable secondary post-
rhage31 and has now been reported in 72 cases partum hemorrhage at a mean of 16 days after
of secondary postpartum hemorrhage (B-Lynch delivery. Six women (43%) had delivered by
C, personal communication, August 2005). Cesarean section and the remainder by sponta-
The use of a surgical compression suture may neous vaginal delivery. Eight women had evi-
avoid the need for hysterectomy in women dence of endometritis (57%), with four of those
wishing to conserve fertility. associated with histologically proven retained
321
343
30 August 2006 14:23:47
Figure 1 Angiogram demonstrating brisk Figure 2 Angiogram after embolization

hemorrhage from false aneurysm prior to
embolization
anterior division of the left internal iliac artery
was identified (Figure 1). The vessels were
embolized with four coils. There was immediate
placental tissue; a further four women had geni-
cessation of bleeding and the patient’s vital
tal tract lacerations, and the remaining two had
signs normalized (Figure 2). The patient made
no obvious cause for bleeding. Basic resuscita-
a good recovery despite needing 24 units of
tion with use of medical treatments and/or uter-
blood during the postpartum hemorrhage. Sub-
ine curettage were performed. Angiography
sequent histology of the uterus showed acute
found no extravasation in eight women, active
inflammation and subinvolution of the placental
bleeding in three women from uterine and
bed.
vaginal vessels, a false uterine artery aneurysm
in two women, and evidence of an arteriovenous
fistula in one woman. Pledgets of absorbable Other measures
gelatin sponge were introduced to embolize
both uterine arteries in 12 women. Unilateral In cases of massive hemorrhage unsuccessfully
embolization of a false aneurysm and an treated with surgical measures, the use of intra-
arteriovenous fistula were performed for the venous tranexamic acid32, recombinant factor
other two women. External bleeding disap- VIIa33 and local vasopressin34 have been
peared immediately, and hemodynamic stability reported for primary postpartum hemorrhage.
and correction of coagulopathy were obtained There are no reports of their use in secondary
for all cases. There were no general or local postpartum hemorrhage but, if available, it may
complications10. be appropriate to consider their use in combina-
One of the authors (T.J.) recently managed a tion with other therapies and resuscitative
case of massive secondary postpartum hemor- support.
rhage presenting 4 days after a Cesarean sec-
tion. An emergency subtotal hysterectomy was
Chemotherapy
performed with good initial results. Two hours
later, vaginal bleeding restarted. There was The mainstay of treatment for choriocarcinoma
no evidence of significant coagulopathy. Pelvic is chemotherapy. A low-risk chemotherapy
angiography was performed and a bleed from a regimen includes the use of methotrexate
false aneurysm related to a middle branch of the with folinic acid rescue on a 2-weekly cycle35.
322
344
30 August 2006 14:23:51
Medium- and high-risk regimens include the research in the form of a randomized controlled
use of etopside, methotrexate, actinomycin, trial difficult. However, particularly for the
vincristine, cyclophosphomide and 6-mercapto- treatment of hemorrhage due to uterine infec-
purine36,37. Women with choriocarcinoma are tion and/or retained placental tissue, this should
most appropriately treated through specialist be achievable and would provide valuable
trophoblastic disease referral centers14. information to further our understanding of
the management of secondary postpartum
hemorrhage.
Coagulopathies
Women with inherited coagulation disorders
such as von Willebrand’s disease and carriers of References
hemophilia A and B are likely to bleed post-
1. Thompson W, Harper MA. Postpartum
partum if maternal clotting factors are low haemorrhage and abnormalities of the third
(< 50 IU/dl). Prophylactic administration of stage of labour. In Chamberlain G, Steer P,
desmopressin (DDAVP) and clotting factor eds. Turnbull’s Obstetrics, 3rd edn. Edinburgh:
concentrates may prevent postpartum hemor- Churchill Livingstone, 2001;619–33
rhage15. The aim is to raise factor levels above 2. Hoveyda F, MacKenzie IZ. Secondary post-
50 IU/dl during labor and delivery and maintain partum haemorrhage: incidence, morbidity and
these for up to 5 days after delivery. In the event current management. Br J Obstet Gynaecol 2001;
of postpartum hemorrhage15, replacement of 108:927–30
deficient clotting factors should be made and 3. King PA, Duthie SJ, Dong ZG, et al. Secondary
identification and treatment of the cause be postpartum haemorrhage. Aust NZ J Obstet
Gynaecol 1989;29:394–8
instigated. Management should be in close
4. Alexander J, Thomas P, Sanghera J. Treatments
liaison with hematologists and specialist hemo-
for secondary postpartum haemorrhage. The
philia centers as available. In cases of prolonged Cochrane Database of Systematic Review 2002
or intermittent secondary postpartum hemor- Issue 1, Art. No: CD002867. DOI: 10.1002/
rhage15, the use of tranexamic acid (a fibrino- 14651858.CD002867
lytic inhibitor)38 or combined oral contraceptive 5. Lédée N, Ville Y, Musset D, et al. Management
pill has been reported15. in intractable obstetric haemorrhage: an audit
Hemorrhage from postpartum acquired study on 61 cases. Eur J Obstet Gynecol 2001;
hemophilia is treated acutely with factor VIII 94:189–96
(either human, porcine) or recombinant factor 6. Matthews NM, McCowan LME, Patten P. Pla-
VIIa15. Immunosuppressive drugs such as centa praevia accreta with delayed hysterectomy.
Aust NZ J Obstet Gynaecol 1996;36:476–9
corticosteroids, cyclophosphamide and aza-
7. Jaffe R, DuBeshter B, Sherer DM, et al. Failure
thioprine may be used to accelerate the dis-
of methotrexate treatment for term placenta
appearance of factor VIII inhibitors, although percreta. Am J Obstet Gynecol 1994;171:558–9
complete remission is likely to occur spontane- 8. Ggosh H. Arteriovenous malformation of the
ously with time. uterus and pelvis. Obstet Gynecol 1986;
Reversal of bleeding due to anticoagulants 68(Suppl):40–3
should follow normal protocols. Vitamin K 9. Gaylis H, Levine E, van Dongen L, et al. Arterio-
should be considered in women with uncon- venous fistula after gynaecologic operations. Surg
trolled bleeding secondary to warfarin use and Gynecol Obstet 1973;137:655–8
protamine sulfate may be considered if hemor- 10. Pelage J-P, Soyer P, Repiquet D, et al. Secondary
rhage results from the use of heparin, although postpartum haemorrhage: treatment with selec-
tive arterial embolization. Radiology 1999;212:
this has a much shorter half-life.
385–9
Secondary postpartum hemorrhage is an
11. Kelly SM, Belli AM, Campbell S. Arteriovenous
important cause of maternal morbidity and malformation of the uterus associated with
mortality. Basic resuscitation followed by inves- secondary postpartum haemorrhage. Ultrasound
tigation and treatment of the specific cause of Obstet Gynecol 2003;21:602–5
hemorrhage are essential. The diverse nature of 12. Nanda S, Singhal S, Sharma D, et al. Nonunion
its etiology and often acute presentation make of uterine incision: a rare cause of secondary
323
345
30 August 2006 14:23:51
postpartum haemorrhage: a report of 2 cases. 27. Fernandez H, Claquin C, Guibert M, et al. Sus-
Aust NZ J Obstet Gynaecol 1997;37:475–6 pected postpartum endometritis: a controlled
13. Paraskevaides E, Stuart B, Gardeil F. Secondary clinical trial of single-agent antibiotic therapy
postpartum haemorrhage from non-dehisced with Amox-CA (Augmentin®) vs ampicillin-
lower caesarean section scar: a case for hystero- metronidazole ± amnioglycoside. Eur J Obstet
scopy. Aust NZ J Obstet Gynaecol 1993;33:427 Gynecol 1990;36:69–74
14. Tidy JA, Rustin GJS, Newlands ES, et al. Presen- 28. Schenker JG, Margalioth SJ. Intrauterine adhe-
tation and management of choriocarcinoma after sions: an update appraisal. Fertil Steril 1982;37:
non-molar pregnancy. Br J Obstet Gynaecol 1995; 593–610
102:715–19 29. Bakri YN. Amri A, Abdul-Jabar F. Tamponade-
15. Economides DL, Kadir RA. Inherited bleeding balloon for obstetrical bleeding. Int J Gynaecol
disorders in obstetrics and gynaecology. Br J Obstet 2001;74:139–42
Obstet Gynaecol 1999;106:5–13 30. Johanson R, Kumar M, Obhrai M, et al. Man-
16. Kadir RA, Economides DL, Braithwaite J, et al. agement of massive postpartum haemorrhage:
The obstetric experience of carriers of haemo- use of a hydrostatic balloon catheter to avoid
philia. Br J Obstet Gynaecol 1997;104:803–10 laparotomy. Br J Obstet Gynaecol 2001;108:
17. Greer IA, Lowe GDO, Walker JJ, et al. Haemor- 420–2
rhagic problems in obstetrics and gynaecology 31. B-Lynch C, Coker A, Adegboyega HL, et al. The
in patients with congenital coagulopathies. Br J B-Lynch surgical technique for the control of
Obstet Gynaecol 1991;98:909–18 massive postpartum haemorrhage: an alternative
18. Ramsahoye BH, Davies SH, Dasani H, et al. to hysterectomy? Five cases reported. Br J Obstet
Obstetric management in von Willebrand’s dis- Gynaecol 1997;104:372–5
ease: a report of 24 pregnancies and a review of 32. Alok K, Hagen P, Webb JB. Tranexamic acid in
the literature. Haemophilia 1995;1:140–4 the management of postpartum haemorrhage. Br
19. Kadir RA, Lee CA, Sabin CA, et al. Pregnancy in J Obstet Gynaecol 1996;103:1250–1
von Willebrand’s disease or factor XI deficiency. 33. Boehlen F, Morales MA, Fontana P, et al.
Br J Obstet Gynaecol 1998;105:314–21 Prolonged treatment of massive postpartum
20. Neill A, Thornton S. Secondary postpartum haemorrhage with recombinant factor VIIa: case
haemorrhage. J Obstet Gynaecol 2002;22:119–22 report and review of the literature. Br J Obstet
21. Johanson R, Cox C, Grady K, et al. Massive Gynaecol 2004;111:284–7
obstetric haemorrhage. In Managing Obstetric 34. Lurie S, Appelman Z, Katz Z. Subendometrial
Emergencies and Trauma – The MOET Course vasopressin to control intractable placental
Manual. London: RCOG Press, 2003;16: bleeding. Lancet 1997;349:698
151–63 35. Bagshawe KD, Dent J, Newlands SL, et al. The
22. Achiron R, Goldenberg M, Lipitz S, et al. role of low dose methotrexate and folinic acid in
Transvaginal duplex Doppler ultrasonography in gestational trophoblastic tumours. Br J Obstet
bleeding patients suspected of having residual Gynaecol 1989;96:795–802
trophoblastic tissue. Obstet Gynecol 1993;81: 36. Newlands ES, Bagshawe KD, Begent RH, et al.
507–11 Results with EMA/CO (etopside, methotrexate,
23. Zuckerman J, Levine D, McNicholas MM, et al. actinomycin D, cyclophosphamide, vincristine)
Imaging of pelvic postpartum complications. Am regimen in high risk gestational trophoblastic
J Roentgenol 1997;168:663–8 tumours, 1979–1989. Br J Obstet Gynaecol 1991;
24. Neill AC, Nixon RM, Thornton S. A compari- 98:550–7
son of clinical assessment with ultrasound in the 37. Rustin GJS, Newlands ES, Bergent HJ, et al.
management of secondary postpartum haemor- Weekly alternating chemotherapy (EMA/CO)
rhage. Eur J Obstet Gynecol 2002;104:113–15 for treatment of central nervous system
25. Thorp JM, Wells SR, Wiest HH, et al. First- metastases of choriocarcinoma. J Clin Oncol
trimester diagnosis of placenta praevia percreta 1989;7:900–3
by magnetic resonance imaging. Am J Obstet 38. Bonnar J, Guillebrand J, Kasonde JM, et al.
Gynecol 1998;178:616–18 Clinical applications of fibrinolytic inhibition
26. Levine D, Barnes PD, Edelman RR. Obstetric in gynaecology. J Clin Pathol 1980;33(Suppl)
MR imaging. Radiology 1999;211:609–17 14:55–9
324
346
30 August 2006 14:23:52
Section VIII
Consequences of postpartum
hemorrhage
347
30 August 2006 14:23:52
36
PATHOLOGY OF THE UTERUS
P. Kelehan and E. E. Mooney
BACKGROUND AND AIMS GROSS EXAMINATION

Significant postpartum hemorrhage may occur Photography is essential at each step of the
immediately after delivery, or may be delayed dissection, with notes as to what each picture
weeks or months. In either case, a Cesarean or is intended to show. Without a clinical input,
later postpartum hysterectomy may be life- however, much effort may be wasted on
saving. The uterus will normally be sent for documenting features of little relevance at the
laboratory examination. To facilitate a useful expense of missing more important ones. A
surgical pathology report, the pathologist must detailed macroscopic description of sutures,
be given details of the antepartum course and tears, etc. is important and may be medico-
delivery. Considering how uncommon these legally relevant. Our approach is to examine the
specimens are, direct communication between specimen in its fresh state, with photography,
pathologist and clinician is recommended. The and then to open the specimen, avoiding tears
aim of this chapter is to provide a structured and sutures, to permit fixation and further
approach to the analysis of the specimen, examination. It may be opened laterally, but
in order to permit a clinically relevant and more information can be gained by complete
pathologically sound diagnosis. longitudinal anteroposterior section of the
uterus. The approach should be modified to
suit the circumstances as predicted from the
CLINICAL CORRELATION
clinical information. A useful technique that
The parity and gestation should be provided. allows good exposure and photographic demon-
Any abnormality of the clinical course, in partic- stration is the placing of two parallel complete
ular pre-eclampsia or polyhydramnios, may longitudinal anteroposterior sections about
be of relevance. Magnetic resonance imaging 2–3 cm apart on either side of the mid-line.
(MRI) may have been performed for fibroid, How well the uterine cavity has compressed is
placenta creta or congenital abnormality and immediately apparent, contraction band forma-
these images should be reviewed. A history of tion can be demonstrated, and blood clot and
the use of instruments such as forceps is impor- placental tissue fragments can be assessed in the
tant. The clinical appearance of the uterus at lumen.
operation may provide valuable information In the immediate postpartum period, the
on atony. Any therapeutic measures undertaken uterus is characteristically large. It will weigh
such as uterine massage or compression suture 700–900 g and will have substantially reduced
should be noted, along with transfusion and in size and volume from its antepartum state.
fluid replacement. A description of the surgery Clamp marks on the broad and round ligaments
will help the pathologist to interpret the tears should be inspected for residual hematoma,
and sutures that characterize these specimens. remembering that the pathology may be outside
The patient’s postoperative condition will help the clamp. In the fresh specimen with intact
to guide sampling in the event that amniotic vessels, it may be possible to perfuse the vascu-
fluid embolism is a consideration. Finally, the lature for contrast angiography or vascular
placenta must also be available for examination. casting1.
326
348
06 September 2006 16:28:08
Pathology of the uterus
Figure 1 Fixed uterus showing a large anterior and right-sided diverticulum originating in a Cesarean
section scar. The specimen was sutured at operation, but placental villous tissue can be seen adjacent to the
suture
Figure 2 Anteroposterior section of uterus from Figure 1 showing anterior placenta creta
327
349
30 August 2006 14:23:58
Figure 3 H/E section of lower uterine segment showing placenta creta and large vessels in thin
myometrium
Figure 4 Immunohistochemical stain for desmin accentuates the thin myometrial fibers in scar
328
350
30 August 2006 14:24:04
Figure 5 Right lateral endocervical tear at hysterectomy for postpartum hemorrhage
Figure 6 Elastin Van Geisson stain showing torn artery at apex of tear (×10). Arrow, torn elastic artery;
arrowhead, thin fibrin blood clot
329
351
30 August 2006 14:24:10
Figure 7 Amniotic debris in venules (arrows) of cervical stroma following a small endocervical tear in
labor. Postpartum hemorrhage and disseminated intravascular coagulopathy necessitated hysterectomy
(×20)
CERVIX forming and secondary rupture can result

in shock, despite cessation of external vaginal
Important pathologies in the cervix include
hemorrhage.
tears. Small shallow endocervical tears are
In the dilated postpartum cervix, edema,
almost invariably found in the postpartum
hemorrhage and fiber disarray may make it diffi-
uterus, and may be present even in those cases
cult to identify tears on histologic examination.
where there has been a Cesarean section. Signif-
Torn and contracted muscle fibers and torn
icant and deep tears tend to be lateral in loca-
arteries with fibrin plugs and tense hematomas
tion. These tears may penetrate through to the
provide corroboratory evidence of a tear. Histo-
serosa, with or without hematoma formation,
logic sampling should include blocks from
and may extend up into the lower segment or
above the apex and from below the tear for deep
down the cervix into the vagina. Involvement
extension and for identification of large torn
of large uterine arteries should be sought. It
vessels.
is common to find meconium staining of the
Examination of the uterus histologically
mucus of the endocervix with fetal distress, and
following amniotic fluid embolism will show no
meconium may contaminate the tear. A tear
evidence of intravascular disease in most cases.
may have severe consequences: an endocervical
Very occasionally, there may be fibrin clots
tear may cause severe blood loss despite a fully
adherent to vascular endothelium and, rarely,
contracted uterus. Tears are associated with
squames admixed with fibrin have been found
amniotic fluid embolus or with amniotic
in vessels in the body of the uterus. In some
infusion and local defibrination. Bleeding
cases of postpartum hemorrhage, when there
can extend into the broad ligament with
have been no clinical features of amniotic infu-
formation of a large hematoma. Suturing of the
sion but bleeding and unexpected severe onset
tear may not prevent a deep hematoma from
of consumptive coagulopathy, histological
330
352
30 August 2006 14:24:13
(a)
(b)
Figure 8 H/E comparison of (a) normal myometrial fibers and (b) myonecrosis in lower uterine segment
in hysterectomy specimen for postpartum hemorrhage following Cesarean section (×40). Long arrows,
normal viable cell nuclei; short arrows, non-viable necrotic cells
331
353
30 August 2006 14:24:19
(a)
(b)
Figure 9 Desmin comparison of same myometrial fibers accentuates the necrosis. (a) Normal;
(b) myonecrosis (×40). Long arrow, normal myometrial cells with intercellular edema; short arrow, dense,
compacted necrotic myometrial cells at same magnification
332
354
06 September 2006 16:26:12
Figure 10 H/E section showing stitch material in uterine curettings following Cesarean section. Arrow,
absorbable suture; arrowhead, giant cell reaction to suture material (×40)
sections of the endocervix will reveal localized LOWER UTERINE SEGMENT

areas where amniotic debris fills and expands
Important pathologies here involve implanta-
venules and capillaries. This dramatic appear-
tion on a previous Cesarean section scar,
ance is present not just adjacent to the
with abnormal adherence or formation of a
endocervical surface and tears: its presence
diverticulum.
deeper in the stroma distinguishs it from con-
A Cesarean section results in chronic changes
tamination of the surface mucosa by meconium
in the lower uterine segment, including distor-
and amniotic fluid at delivery.
tion and widening, inflammation, giant cell
A subgroup of patients have a lesion of local
reaction and adenomyosis3. In some cases, a
amniotic infusion associated with disseminated
distinctive V-shaped defect of the anterior wall
intravascular coagulopathy and postpartum
(‘tenting’) may be present.
hemorrhage without systemic collapse.
An important cause of weakening of a
Squamous cells may be present in only one
Cesarean section scar is infection. Postoperative
or two sections taken from around the
wound infection is not uncommon following
circumference of the cervix. It is usually on
Cesarean section, particularly emergency sec-
one side. Extensive sampling of the cervix
tion. Prophylactic antibiotics can modify the
may be required to demonstrate amniotic
extent and rate of infection, as can the quality of
debris in cases of suspected amniotic fluid
closure, the amount of local tissue trauma, the
embolism2. It is possible that ongoing blood loss
technique used (one- or two-layer), swelling,
from a tear in this site may occur before the
hematoma and the nature of the organisms
onset of systemic disseminated intravascular
infecting the wound. There may be extensive
coagulopathy, because local thromboplastin
disruption and inflammation in the uterine wall
effect alone of the amniotic debris in the wound
despite a normal healing appearance of the skin
may inhibit hemostasis.
wound. Conservative treatment of the wound
333
355
30 August 2006 14:24:28
is normal, and surgical debridement the excep- Examination of the lateral margins of the defect
tion. Accordingly, the consequences may be may indicate left- or more often right-sided
only appreciated in a subsequent pregnancy. If extension of the bulging diverticulum into
the patient does present before this, hemorrhage parametrial soft tissue of the pelvis. A complete
and/or vaginal discharge may prompt internal section through the anterior lower uterine seg-
examination. A defect may be identified on pal- ment can identify previous Cesarean section
pation. Curettage may be undertaken and may scars with tenting defects and the shape and
retrieve inflammatory exudate, degenerating edges of a recent section. Most importantly,
decidua, polypoid endometrium or fragments of en-face examination of the lateral sides of the
necrotic myometrium that have prolapsed into lower segment will show the cavity and lateral
the endocervical lumen from the internal edge extension of a dehiscence diverticulum, fresh
of the Cesarean section scar. Sometimes, quite tears and/or adherent placenta. The issue of
large pieces of myometrial tissue with edema abnormal adherence is addressed below.
and coagulative necrosis are obtained. This
myonecrosis, or incisional necrosis, is caused by
FUNDUS
local ischemia4. Remodelling of blood vessels
may influence implantation. Implantation on Important pathologies include retained prod-
either a normally healed or on a diseased scar ucts, placenta creta, and subinvolution. Pla-
will not have the protective effect of the decidua centa creta is the name given to abnormally
vera (see below), and so postpartum separation adherent or ingrowing placenta that does not
is less likely to occur. A Cesarean section at detach with full contraction of the uterus after
first birth is associated with increased risks expulsion of the fetus. This term covers pla-
of placenta previa and abruption in second centa accreta (abnormal attachment to the
pregnancies5. wall), increta (extension of villi into the myo-
Implantation in the lower segment (adjacent metrium) and percreta (extension of villi
to the defect) can cause expansion of the defect, through to the serosa). The intimate relation-
dehiscence of the wall and the formation of a ship of villous tissue to myometrium, without
pulsion diverticulum which will further enlarge intervening decidua, is the key to the diagnosis.
and progress with growth of the placenta. If the Descriptions of placenta percreta based on
implantation is fundal, a fortuitous elective illustrations or descriptions of chorionic villi
section may reveal a thin, almost transparent displaced between torn myometrial fibers
anterior lower segment wall. This should be should be evaluated critically.
more easily resected at closure since the scar will MRI may show the loss of zonation associ-
not be excessively vascular. If implantation is ated with penetration rather than invasion of
in the lower segment or in the scar, then there chorionic villi.
is a potential for catastrophic hemorrhage on Full-thickness anteroposterior sections of the
attempt at delivery of the placenta. fundus make it easier to recognize the position
In examining a postpartum hysterectomy of the contracted placental site. It is surprisingly
specimen where there is a history of previous difficult to identify the exact site on inspection
Cesarean section, the points noted above should of the raw decidual surface that is seen if the
be borne in mind. The recently sutured section uterus is opened laterally.
incision should be carefully reopened. Follow- Detachment of the placenta is dependent on
ing photography, the edges and margins should the presence of a normal spongy decidua vera,
be inspected for thinning and scar tissue forma- where shearing of the placenta from the myo-
tion. An enlarged, ragged and open defect of metrium occurs. This soft compressible area is
the anterior lower uterine segment, now tightly not seen when the postpartum uterine lining is
contracted and rigid with formalin fixation, examined histologically, because its many mucous
may be all that is left of a huge, thin-walled, glands are disrupted to facilitate the normal
placenta-filled diverticulum, the result of scar plane of cleavage. It is seen to its full extent in
dehiscence and rupture. It is easy to destroy the tragic case of maternal death prior to labor.
this thin structure with precipitate dissection. Either Alcian blue stain or diastase-PAS to
334
356
30 August 2006 14:24:28
demonstrate mucopolysaccharides in the swol- examination of the lining of the postpartum

len gland crypts can help to identify this layer. uterus, the finding of chorionic villi in clefts in
Deficiency of this layer may be focal or, rarely, the placental bed may be an artefact rubbed
complete. When it is absent, the thinned in following clearance of uterine contents and
Nitabuch’s layer with anchoring villi lies in close is of no diagnostic consequence. Smearing of
proximity to muscle fiber bundles or interstitial DNA due to crush artefact may be helpful in
fibrous Cesarean section scar. An occasional distinguishing this from true extension.
finding is the presence of abundant inter-
mediate trophoblast infiltrating between muscle
RETAINED PRODUCTS OF
fibers beneath a firmly adherent Nitabuch’s
CONCEPTION
layer. Histological examination of multiple
sections can show anchoring villi penetrating The failure of total expulsion of the placenta
Nitabuch’s fibrinoid and ghost villi in dense may lead to postpartum hemorrhage. A frag-
fibrin adherent to muscle. The often described ment of placenta remains, assumes a polypoid
appearance of chorionic villi infiltrating between shape (‘placental polyp’), and undergoes com-
muscle fibers is characteristic only of invasive pression and devitalization. Some viable cells
mole; the key to placenta percreta is absence of may remain in stem villi. Vessels below the
decidua. An increased number of implantation retained fragments may show persistent dilata-
site intermediate trophoblasts has been shown tion. There may be a plasma cell infiltrate in the
in cases of placenta creta compared with adjacent myometrium – this is not diagnostic of
controls6. Retained placental fragments reflect (infective) endometritis in this context. The fre-
some degree of placenta creta and are more quency of detection of retained products varies
common in women with a spectrum of changes from 27 to 88%7, but much of this literature
in previous pregnancies, such as pre-eclamptic is decades old. Retained placental fragments
toxemia, growth restriction, spontaneous abor- are more common in women who have had
tion and retained placental fragments. It has complications such as pre-eclampsia or growth
been hypothesized that these reflect abnormal restriction in previous pregnancies. This has
maternal–trophoblast interaction7. been interpreted as indicative of an abnormal
Placenta creta is therefore due to a deficiency maternal–trophoblast relationship7.
of the decidua. The end result of penetration of
the placenta though a weakened part of the
SUBINVOLUTION
uterine wall includes rupture and secondary
changes, including serosal peritoneal reaction. Subinvolution of the blood vessels of the pla-
Curette penetration may cause secondary infec- cental bed, in the absence of retained placental
tion or hematoma formation and provide the fragments, is an important and distinctive cause
nidus for dehiscence into the adherent bladder of secondary postpartum hemorrhage.
wall, if this had been injured at previous surgery. Normal arterial involution involves a
Placenta creta is only part of the problem decrease in the lumen size, disappearance of
of uncontrolled postpartum hemorrhage. The trophoblast, thickening of the intima, re-growth
thin myometrium, with little muscle, interstitial of endothelium and regeneration of internal
fibrosis and increased intermediate trophoblast elastic lamina. These changes occur within
will contain large dilated arteries of pregnancy 3 weeks of delivery. With subinvolution, arteries
and often widespread extrauterine extension of remain distended and contain red cells or fresh
these changes into the parametrium, as thrombus, and trophoblast persists in a peri-
described on Doppler ultrasound. The degree vascular location8. In some cases, endovascular
of constriction–contraction of the myometrium trophoblast may be present. Hemorrhage from
is insufficient to close off these vessels. Where subinvolution is maximal in the second week
there is severe thinning of the muscle of the postpartum, although it may occur up to several
lower segment with diverticulum formation, months later. It is commoner in older, multi-
abnormal adhesion is not necessary to parous women and may recur in subsequent
sustain bleeding. Conversely, on histological deliveries.
335
357
30 August 2006 14:24:29
Subinvolution is not related to the method anomalies may also result in reduced decidua
of delivery and may be regarded as a specific formation and subsequent postpartum hemor-
entity, possibly due to an abnormal immuno- rhage. Trophoblastic disease has also been
logic relationship between trophoblast and reported in this context.
the uterus8. Despite this, it did not show the
association with markers of such an abnormal
ENDOMETRITIS
relationship seen with retained placental
fragments in another study7. An acute endometritis is reported as a cause of
The changes may be recognized on curettage sepsis and postpartum hemorrhage. It is rela-
specimens. The hysterectomy specimen will tively uncommon in modern obstetric practice
show a uterus that is soft and larger than in the West and may be due to a variety of
expected8. As normally involuted vessels may be organisms. It accounted for < 5% of cases of
present adjacent to subinvoluted ones, multiple delayed postpartum hemorrhage in one series7.
blocks of placental bed should be taken to
exclude this process.
PLACENTAL PATHOLOGY
The placenta should be examined in cases
ATONY
of postpartum hemorrhage. Pre-eclampsia
This is well-recognized obstetric phenomenon, may cause retroplacental hemorrhage: recent
but there may be relatively little to report in and old hemorrhages and infarcts may be seen.
the way of pathology. The diagnosis is one of The characteristic changes of acute atherosis are
exclusion. The uterus is enlarged, edematous only present in 50% of cases of pre-eclamptic
and soft, with edema and hemorrhage apparent toxemia. However, examination of the paren-
microscopically. The diagnosis will depend on chyma will usually show so-called accelerated
clinical information, combined with adequate villous maturation (distal villous hypoplasia) in
histologic sampling to exclude other causes. response to uteroplacental ischemia. Sampling
from the center of the disc is important to avoid
overinterpretation of physiologic changes13.
ARTERIOVENOUS MALFORMATIONS
Uterine arteriovenous malformations (AVMs)
THE AUTOPSY IN POSTPARTUM
are rare and may present with profuse hemor-
HEMORRHAGE
rhage, including hemorrhage in the postpartum
period. Congenital AVMs consist of multiple In data drawn from the Confidential Enquiries
small connections and may enlarge with preg- into Maternal Deaths in the UK for the period
nancy. The more common acquired AVMs are 1970–90, approximately 10% of direct maternal
rare in nulliparous women, and are thought to deaths are due to hemorrhage14. Roughly half
arise following uterine trauma: curettage, myo- were antepartum and half postpartum. Excess
mectomy or even previous uterine rupture9,10. blood loss is more common in older women
AVMs may co-exist with retained products (> 35 or 40 years, depending on the study)15.
of conception or trophoblastic proliferation. Before beginning an autopsy in a case of
Pathologically, vessels of arterial and venous maternal death following postpartum or intra-
caliber are present, along with large vessels of partum hemorrhage, it is critical to plan the pro-
indeterminate nature. cedure and the sequence of the autopsy in the
light of the information received and the sus-
pected cause or causes and mode of death. The
OTHER CAUSES
autopsy must be unhurried and methodical; it is
Lacerations of the inner myometrium have been a fundamental mistake to seek to demonstrate
reported to cause postpartum hemorrhage11. immediately the proposed cause of death.
Women with leiomyomas are at an increased Members of the clinical team should be asked to
risk of postpartum hemorrhage12. Less com- attend the autopsy, but it is unwise to have
monly, endometrial carcinomas and congenital everybody there during what will be a long
336
358
30 August 2006 14:24:29
phase of inspection, measurement and initial uterus is still small, by the characteristic dilated
systematic dissection. When all is ready, the and congested appearance of retroperitoneal
procedure is stopped and members of the team veins. The degree of dilation and turgidity of the
attend. In this way, the history can be reviewed, pelvic veins should be noted at autopsy as they
pre-existing conditions or disease discussed and will be dissected and examined in detail later.
demonstrated, e.g. chronic pyelonephritis, and Retroperitoneal hematoma and broad ligament
the dissection and demonstration of the focus of hematoma should be identified or excluded at
main clinical interest can begin. this stage as these may be less easily assessed
A fundamental aspect of good autopsy prac- and measured following organ removal. The
tice is the confident exclusion of specific dis- uterus may be examined and opened in situ, but
eases and conditions in a systematic approach. it is better to remove adrenal, renal and pelvic
The understandable desire and pressure to skip organs as one complete block.
to the seat of disease must be resisted. The The traditional method of blunt dissection
parametrium, pelvic side-wall and vagina are as along the pelvic side-wall and pubis with
important objects of attention as the uterus. transection at the mid to upper vagina is
At the time of external inspection of the extended in the investigation of postpartum
body, the pathologist must consider in turn each hemorrhage. Following knife separation of the
of the major causes of maternal death. Many symphysis pubis, the legs are externally rotated
require modification of routine techniques, and a knife cut is made along the lower edge of
e.g. air embolism, amniotic fluid embolism, the pubic bone. The pubic bones are forcefully
ruptured aneurysm, and these modifications are separated by 8–10 cm. This, together with the
detailed elsewhere16. Preparation and sampling inguinal femoral incisions, gives good exposure
of blood and fluids for hematology, hemophilia, of the paracervical and paravaginal soft tissues.
toxicology and microbiology should be planned, Lateral vaginal wall tears and hemorrhage can
e.g. sample containers should be pre-labelled be inspected and well demonstrated by this
and set out in sequence. Cardiopulmonary modified technique. The ileofemoral vessels
resuscitation attempt most likely preceded are transected and inspected. The complete
death and therefore the features and sequence urogenital block is placed on a dissection
of sustained unsuccessful resuscitation must board where it can be opened in layers, begin-
be identified and complications and agonal ning with the urethra and bladder, then the
changes interpreted in this context. It is impor- vagina and cervix. Alternatively, the block can
tant from a medicolegal aspect not to allow such be placed in formalin and later dissected after
artefact to be construed as a major factor in the short fixation.
cause of death, e.g. liver or mesenteric tear, The aorta is opened posteriorly and incision
blood in the abdomen, bone marrow embolus. is extended into the branches of the iliac arteries
The traditional Y-shaped autopsy incision for a short distance. The inferior vena cava
should be extended to an abdominal inverted Y is opened from the anterior side, probed and
with the incision continued to the inguinal dissected into the right and left renal veins;
femoral triangle on each side. This allows better the ovarian veins are identified and opened
examination of the ileofemoral vessels and and dissection is continued into the branches of
better exposure of the pelvis. Blood and blood the pelvic veins out to the limits of the excised
clots are removed from the abdomen and the specimen. The intima is examined for evidence
amounts measured. The relative size and posi- of tear or abrasion and for adherent thrombus.
tion of the abdominopelvic organs are assessed. Pieces of tissue containing venous plexus from
The peritoneal lining of the pelvis is inspected, the broad ligament and parametrium are
noting color, texture and degree of congestion. selected for formalin fixation and histological
Patches of peritoneal decidual reaction of examination.
pregnancy can be identified by their gelatinous When the patient has died of hemorrhage
appearance. and where there has been attempt to stem the
In traditional autopsy practice, the state bleeding by hysterectomy and under-sewing of
of pregnancy can be suspected, even when the bleeding sites and pedicles, it may be very
337
359
30 August 2006 14:24:29
difficult to identify the exact sites of bleeding, 7. Khong TY, Khong TK. Delayed postpartum
and ancillary techniques may be helpful. Prior hemorrhage: a morphologic study of causes and
to pelvic dissection, an infusion of saline their relation to other pregnancy disorders.
through an intravenous infusion set and cannula Obstet Gynecol 1993;82:17–22
8. Andrew AC, Bulmer JN, Wells M, Morrison L,
into the clamped abdominal aorta can identify a
Buckley CH. Subinvolution of the uteroplacental
bleeding point. With special preparation and
arteries in the human placental bed.
ligation of all peripheral vessels, post-mortem Histopathology 1989;15:395–405
specimen angiography may be very valuable in 9. Grivell RM, Reid KM, Mellor A. Uterine
selected cases. arteriovenous malformations: a review of the
The most useful of all techniques is the current literature. Obstet Gynecol Surv 2005;
histological examination of carefully selected 60:761–7
blocks of tissue demonstrating vital reaction to 10. Ciani S, Merino J, Vijayalakhsmi, Nogales FF.
injury and the presence or absence of conditions Acquired uterine arteriovenous malformation
predisposing to disease. with massive endometrial stromal component
[letter]. Histopathology 2005;46:234–5
11. Hayashi M, Mori Y, Nogami K, Takagi Y,
References Yaoi M, Ohkura T. A hypothesis to explain the
occurrence of inner myometrial laceration caus-
1. Schaaps JP, Tsatsaris V, Goffin F, et al. Shunting
ing massive postpartum hemorrhage. Acta Obstet
the intervillous space: new concepts in human
Gynecol Scand 2000;79:99–106
uteroplacental vascularisation. Am J Obstet
12. Qidwai GI, Caughey AB, Jacoby AF. Obstetric
Gynecol 2005;192:323–32
outcomes in women with sonographically
2. Cheung ANY, Luk SC. The importance of
identified uterine leiomyomata. Obstet Gynecol
extensive sampling and examination of cervix in
2006;107:376–82
suspected cases of amniotic fluid embolism. Arch
13. Mooney EE, Padfield J, Robboy SJ. Nidation
Gynecol Obstet 1994;255:101–5
and placenta. In Robboy SJ, Anderson MC,
3. Morris H. Surgical pathology of the lower
Russell P, eds. Pathology of the Female Repro-
uterine segment caesarean section scar: is the
ductive Tract. London: Churchill Livingstone
scar a source of symptoms? Int J Gynecol Pathol
2002:721–57
1995;14:16–20
14. Toner PG, Crane J. Pathology of death in preg-
4. Rivilin ME, Carroll CS, Morrison JC. Uterine
nancy. In Anthony PP, MacSween RNM, eds.
incisional necrosis complicating caesarean
Recent Advances in Histopathology, Vol 16. Edin-
section. J Reprod Med 2003;48:687–91
burgh: Churchill Livingstone 1994:189–212
5. Getahun D, Oyelese Y, Salihu H, Anath CV.
15. Ohkuchi A, Onagawa T, Usui R, et al. Effect of
Previous caesarean delivery and risks of placenta
maternal age on blood loss during parturition:
previa and placental abruption. Obstet Gynecol
a retrospective multivariate analysis of 10,053
2006;107:771–8
cases. J Perinat Med 2003;31:209–15
6. Kim KR, Jun SY, Kim JY, Ro JY. Implantation
16. Rushton DI, Dawson IMP. The maternal
site intermediate trophoblasts in placenta cretas.
autopsy. J Clin Pathol 1982;35:909–21
Mod Pathol 2004;17:1483–90
338
360
30 August 2006 14:24:29
37
SEVERE ACUTE MATERNAL MORBIDITY
A. Vais and S. Bewley
INTRODUCTION The UK remains one of the few developed

countries in which every maternal death is
For every woman who dies of postpartum
investigated. This was also the case in the
hemorrhage, a host more suffer short- and
United States (US) after 1930, but the rapid
long-term consequences from postpartum
decline in maternal mortality in the latter part of
hemorrhages or their sequelae, even when
the 20th century diminished the vigor used to
well-managed. During the 1990s, the concept
investigate each individual case. It is not clear
of severe adverse maternal morbidity (SAMM)
how many developed countries have policies
emerged in response to the need for a more
similar to that of the UK. As a result of per-
sensitive marker of quality of obstetric care1,2.
ceived racial discrepancies in maternal mortality
This term has the advantage over maternal
in the US, as well as evidence that not all mater-
death of drawing attention to surviving
nal deaths were reported to the National Vital
women’s reproductive health and lives and is
Statistics System (NVSS)7, a parallel, voluntary
equally applicable in developing as well as
system of reporting was introduced in 1983,
developed countries.
termed the Pregnancy-related Mortality Sur-
In developed countries, maternal death
veillance System (PMSS)8. While the NVSS
from postpartum hemorrhage has become too
collects information from death certificates
rare for adequate surveillance of services. For
alone, the PMSS combines data from maternal
example, the United Kingdom (UK) triennial
death certificates with fetal death certificates,
Confidential Enquiry into Maternal Deaths has
autopsy reports and reports produced by mater-
revealed that, over the past 50 years, the num-
nal mortality review committees8. This has led
ber of maternal deaths from hemorrhage has
to better ascertainment of cause of death, and a
fallen from 40 to 3 per annum3. Currently, the
more accurate maternal mortality rate of 11.88,9
overall maternal mortality rate in the UK is
rather than 7.77,8 per 100 000 live births for the
around 7 per 100 000 maternities4. However,
period 1991–1999.
the same causes of death persist as in the 1950s,
with hypertensive disorders and hemorrhage as
the most common causes of direct obstetric
deaths5. Seventeen out of a total of 106 direct WHAT IS THE DIFFERENCE
obstetric deaths were attributed to hemorrhage BETWEEN A ‘NEAR-MISS’ AND A
during 2000–2002 (i.e. 16%). Of these, ten SAMM?
were due to postpartum hemorrhage3. Com- A ‘near-miss’ used to be thought of as a
pared to the previous report (seven deaths case where a woman had a near brush
in 1997–1999)6, there was a slight rise in with death; she would have died were good
incidence. Although this is not statistically fortune and medical care not on her side. This
significant, it needs to be watched as a possible characterization was also used for women with
trend alongside a rising Cesarean section rate. severe organ dysfunction or organ failure who
A potentially far more worrying factor is that survived10,11, that is, with intensive medical
substandard care was implicated in 80% of the intervention, a maternal death was avoided
cases attributed to hemorrhage3. and turned into a survival12. However, the term
339
361
30 August 2006 14:24:29
‘near-miss’ is no longer used, as the ‘near-miss’ An agreed and accurate definition of SAMM
concept was originally derived from the aviation is not available, as studies in different parts of
industry and referred more to risk management the world use different criteria. The two main
than the effect on the woman. In contrast, definitions of SAMM to date are as follows:
SAMM refers to the morbidity a woman
actually suffers. Essentially, we can think of a (1) An organ system approach10,11,13, e.g.
pyramid of disease in pregnancy (see Figure 1), shock from hemorrhage, severe pre-
the base being the numerically larger general eclampsia or specific organ failure; these are
pregnant population, the ‘tip of the iceberg’ best identified as they occur;
being maternal death and much hidden (2) Management, or process, based12,14, e.g.
morbidity beneath the surface9–11. A clinical admission to a high-dependency unit
insult may be followed by a systemic response (HDU) or intensive care unit (ICU) or
and subsequent organ dysfunction, which leads transfer to another health-care facility
to organ failure and eventual death1,10. Figure 1 (usually higher level); these are usually
illustrates both the severity continuum of retrospective studies.
morbidity as well as factors that move a
woman up or down the pyramid. For example, A diligent unit is more likely to pick up cases via
a faulty ambulance or wrongly cross-matched the organ system approach and carefully record
blood might lead to an anemic woman dying of them; this will translate into a disproportion-
hemorrhage unnecessarily (arrow A). If she is ately higher rate of SAMM11. On the other
well managed and treated promptly, there may hand, a poor-quality unit that does not recog-
be no residual morbidity at all (arrow B). A nize and treat hemorrhage promptly may have
wrongly labelled blood bag that is spotted in more severe sequelae as the natural history
time still constitutes a ‘near-miss’ requiring progresses. Use of the management-based
follow-up of the error, although the woman did approach relies on more easily agreed parame-
not experience a transfusion reaction. ‘Near- ters, but also on the availability of HDU/ICU
miss’ now refers to avoidable risks whereas beds. Units have different policies and thresh-
SAMM retains the concept of the harmed or olds for transfer. There may be an underestima-
damaged mother. tion of the true incidence of SAMM, especially
System factors Timely intervention
Patient factors Skilled personnel

(social exclusion, A Death
poor transport) System factors
SAMM (available drugs &
blood products)
Poor antenatal Moderate morbidity
care Good
Mild morbidity maternity
Antenatal services
anemia
Clinical insult
B
General pregnant population
Figure 1 A representation of the morbidity–mortality continuum
340
362
30 August 2006 14:24:31
Severe acute maternal morbidity
in smaller, more isolated units and in CAUSES OF SAMM

developing countries.
Most cases of SAMM fall into three major
categories:
INCIDENCE OF SAMM
(1) Hemorrhage;
Quantifying SAMM is problematic as there
(2) Hypertensive disorders of pregnancy
is no international definition and recording is
(including eclampsia and HELLP syn-
haphazard at best. Thus, there is a wide varia-
drome);
tion in the estimate of incidence. Tables 1 and 2
summarize studies to date. Wide variations are (3) Sepsis.
present in study settings, definitions and main
Table 1 summarizes both the all-cause inci-
causes. Some studies use admission to ICU14,15,
dence of SAMM as well as the three major
others define the actual conditions responsible
causes. Rates in European countries are similar
for the morbidity10,11, and some list both12.
for the three major causes of severe morbidity
Two methods have been described to address
despite the use of different definitions. Regard-
the relationship between severe morbidity and
less of geographical factors, hemorrhage is the
mortality. These are the mortality-to-morbidity
largest contributor, accounting for one-fifth17
ratio1,13 and the mortality index11,16. The mortal-
to one-half10,18,19 of cases. Hypertensive disease
ity-to-morbidity ratio simply describes the num-
and its consequences account for 10%18 to
ber of severe morbidity cases for each maternal
45%20 of cases of SAMM, whereas morbidity
death1,13. The mortality index, on the other
secondary to sepsis is much lower (1.5%18 to
hand, is defined as the number of maternal
20%10). Other rarer causes of severe morbidity
deaths divided by the sum of women with
include uterine rupture, thromboembolic dis-
SAMM and maternal deaths, and is expressed
ease and psychiatric illness5,21. The Mothers’
as a percentage11,16. Both can be expressed as
Mortality and Severe Morbidity Survey was
totals (all-cause) or by condition. They both
conducted during the 1990s by an international
reflect the impact of a condition on severe
team which spanned 11 European countries.
morbidity and mortality and identify those
There are two parts to the survey: MOMS-A
conditions that are more or less amenable to
and MOMS-B20. MOMS-A collected and com-
intervention. In general, the risk of mortality
pared data on maternal deaths, while MOMS-B
depends on the prior health of the mother, the
identified cases of severe morbidity20. The sur-
severity of the particular condition, the access
vey established that, in European countries with
to skilled help and the availability and quality
the highest SAMM rates, i.e. Belgium, Finland
of medical intervention. Postpartum hemor-
and the UK, most of the difference was due
rhage is common, and has a very favorable
to higher incidence of hemorrhage. However,
morbidity-to-mortality ratio (or low mortality
maternal mortality was no higher than in other
index)1,11,13,16. Stated another way, the condi-
European countries. This suggests either that
tion, at least in developed countries, is very
ascertainment of cases in these three countries
amenable to treatment. More women’s lives can
is more complete or that hemorrhage is not a
be saved with medical interventions than for a
major cause of death; therefore the higher inci-
comparable number of cases of infection or car-
dence of SAMM does not affect overall mortal-
diac disease. Many lives can be, and indeed are
ity data. Alternatively, it may be that mortality
being, saved daily by the provision of adequate
rates are associated more closely to the quality
maternity services world-wide. As hemorrhage
of care than the prevalence of morbidity20. The
is so obviously both avoidable and treatable,
geographical areas chosen in different countries
and because all parturients are at risk, it is tragic
had very different demographics, and this also
that so many women die unnecessarily. Unfor-
may have affected the rates of morbidity;
tunately, complacency in developed countries
Belgium and the UK were represented by
about the daily marvels achieved in childbirth4
Brussels and the South-East Thames region,
has made any sudden unexpected threat to life
areas with significant inner-city and migrant
almost unbelievable and unbearable.
341
363
30 August 2006 14:24:31
Table 1 The incidence of each major cause of morbidity per 1000 deliveries
Incidence Incidence of Incidence of Incidence of
30 August 2006 14:24:32

Study, country and of SAMM hemorrhage hypertension severe sepsis
year of publication (all causes) (% of total) (% of total) (% of total) Additional comments
Stones2, UK, 1991 8.8 3.23 2.77 not available SAMM defined as ‘potentially life-threatening episodes’. Incidence for total
(36.8%) (31.5%) morbidity 267/1000. Incidence of all sepsis 30.5/1000 (severe sepsis not
separated out). Hemorrhage includes antepartum and postpartum if over
2000 ml and also 1 case of secondary PPH due to sepsis which required
hysterectomy
Bouvier-Colle17, 3.1 0.62 0.81 0.14 3rd highest cause of morbidity is embolic events at 0.38/1000. Hemorrhage
France, 1996 (20%).8 (26.2%) (4.36%) includes uterine rupture. Hypertensive disease includes cerebral hemorrhage
Bewley & 4.97 2.3 1.98 0.49 SAMM = ITU admission. Total 30 cases of SAMM. 14 cases classed as
Creighton12, UK, (46.7%) (40%).8 (10%) hemorrhage (blood loss > 2000 ml but a further 2 cases DIC/HELLP so
1997 proportion due to hemorrhage could be > 50%)
Baskett & Sternadel22, 0.72 0.16 0.18 0.1 SAMM = ITU admission
USA, 1998 (22%).8 (25%).8 (14.5%)
342
364
Mantel10, South 10.9 6.1 2.82 2.16 Sepsis incorporates septic abortion, chorioamnionitis and puerperal sepsis.
Africa, 1998 (55.8%) (25.8%) (19.7%) Hemorrhage incorporates antepartum and postpartum hemorrhage and
emergency hysterectomy; PPH alone is 1.8/1000
Prual18, West Africa, 59.8 29.6 6.15 0.9 Obstructed labor is significant cause for severe morbidity (20.5/1000 of
2000 (49.5%) (10.3%) (1.5%) which 1.2/1000 uterine rupture)
Waterstone13 12.0 6.7 4.6 0.35 Clinically based definitions, not including management processes. Estimated
(COSMO), UK, 2001 (55.7%) (38%).8 (2.89%) blood loss > 1500 ml picked up 55% of cases of SAMM due to hemorrhage
Brace19, Scotland, 3.8 1.9 1.15 0.09 Septic shock is the only category for sepsis. Number of SAMM due to

2004 (50%).8 (30%).8 (3%).89 hypertensive disease derived by adding the number of cases with eclampsia,
renal dysfunction and pulmonary edema. Only one-third of patients with
SAMM were admitted to ITU
Zhang20 (MOMS-B), 9.48 4.6 4.33 0.8 Multinational study, rates differing widely between countries. Range of
Europe, 2005 (48.8%) (45.7%) (8.2%) SAMM 6–14.7%, highest in Finland, Belgium and UK; lowest rates in Italy,
Ireland, France
DIC, disseminated intravascular coagulation; SAMM, severe adverse maternal morbidity; ITU = intensive therapy unit; HELLP, hemolysis, elevated liver
enzymes, low platelets; PPH, postpartum hemorrhage
Table 2 The incidence of SAMM and hemorrhage and definitions used to ascertain cases
30 August 2006 14:24:33

Number
Number of of maternal
deliveries, Incidence deaths, MMR
cases of of SAMM Incidence of % of for SAMM
Author, SAMM, or ITU hemorrhage SAMM overall,
country, year Type of study, cases of (/1000 (/1000 due to Definition of severe hemorrhage mortality
of publication type of unit hemorrhage deliveries) deliveries) hemorrhage (additional comments) Perinatal outcome per 100 000
Graham & retrospective, 21 983 1.04 0.05 4.35% 1 case of hemorrhage counted but 5 cases 1 intrauterine death 2.4
Luxton41, general ITU 23 (ITU) (0.23)* (21.7%)* in total (3 abruptions: 1 was the hemorrhage 11.5 : 1
UK, 1989 1 (5)* and 2 cases of DIC). 9 patients showed 9.1.4
some coagulopathy, 5 received > 4 units
transfusion
Mabie & retrospective, 3-bed 22 651 8.82 0.93 10.5% massive hemorrhage not defined not collected not collected
Sibai23, US, dedicated obstetric 200 (ITU)
1990 ITU 21
Stones2, UK, retrospective, single 2164 8.8 3.23 36.8% hemorrhage > 2000 ml or DIC or not collected 0.4
1991 unit 19 hysterectomy
343
365
7 0.4
Killpatrick & retrospective, 8000 4.82 0.5 12.5% hemorrhage/hemodynamic instability. 2 stillbirths delivered 4.4
Matthay14, general ITU 32 (ITU) 52% of postpartum admissions were for on ITU. No neonatal/ 8:1
US, 1992 4 hemodynamic instability fetal deaths after 50.4
admission to ITU
Monaco15, retrospective, ITU 15 323 2.47 0.26 10.5% 2 cases following PPH and 2 cases of perinatal mortality 7.4
US, 1993 admissions 38 (ITU) hematologic dysfunction (local policy is rate 12% (4 of 33 5.4 : 1
4 to admit only for ventilatory support or pregnancies followed 45.7
pulmonary artery catheterization) up)

Bouvier-Colle17, retrospective, 140 323 3.1 0.62 20%.38 hemorrhage not defined but includes stillbirth rate collected 22.4
France, 1996 2 French regions 435 (ITU) uterine rupture only 24.6% in 19.8 : 1
87 hypertension, 17.3% 15.7
in hemorrhage, 33.3%
in sepsis
Bewley & retrospective, 6039 4.97 2.3 46.7%.38 transfer to ITU for blood loss > 2000 ml not collected not collected
Creighton12, general ITU 30 (ITU)
UK, 1997 14
Continued
Table 2 Continued
30 August 2006 14:24:33

Number
Number of of maternal
deliveries, Incidence deaths, MMR
cases of of SAMM Incidence of % of for SAMM
Author, SAMM, or ITU hemorrhage SAMM overall,
country, year Type of study, cases of (/1000 (/1000 due to Definition of severe hemorrhage mortality
of publication type of unit hemorrhage deliveries) deliveries) hemorrhage (additional comments) Perinatal outcome per 100 000
27
Lapinsky , retrospective, ITU 25 000 2.6 0.44 17%.38 hemorrhage requiring ITU admission, perinatal mortality 0.4
Canada, 1997 admissions in 65 (ITU) not defined. 52% of admissions involved rate 11%
tertiary hospital 11 hemodynamic instability; 4 hysterectomies 0.4
25
Tang , Hong retrospective, single 39 354 1.24 0.66 53%.38 blood loss 1000–8500, mean 3500 ml. perinatal mortality 2.4
Kong, 1997 center 49 (ITU) Received blood transfusion (mean 12 rate 10.2% 24.5 : 1
26 units), platelet transfusion or FFP. DIC 5.1
in 34% and mild coagulopathy in 27%
of hemorrhage cases. Hysterectomy in
22 cases (84.6% of all hemorrhage)
344
366
Mantel10, South prospective, 13 429 10.9 6.1 55.8% severe hemorrhage = hypovolemia not collected 30.4
Africa, 1998 multicenter 147 requiring 5 or more units of blood for 4.9 : 1
82 resuscitation or emergency hysterectomy 223.4
Mahutte28, retrospective, 44 340 2.95 0.77 26%.38 hemorrhage causing admission due to not collected 3.4
Canada, 1999 2 tertiary care units 131 (ITU) abnormal placentation, atony, lacerations, 43.7 : 1
with general ITU 34 RPOC, severe coagulopathy. 27 (79%) 6.8
received blood products and 12 (35%)
admitted after Cesarean hysterectomy
Waterstone13,21, prospective, 48 865 12.0 6.7 55.7% blood loss > 1500 ml/peripartum perinatal mortality 5.4
UK, 2001 case–control, 588 hemoglobin drop ≥ 4 g/dl or acute rate 7.5% .4118 : 1

multicenter 327 transfusion ≥ 4 units 10.2
Prual18, West prospective, 20 326 59.8 29.6 49.5% hemorrhage requiring transfusion, not collected 38.4
Africa, 2000 6 countries 1215 hospitalization > 4 days or hysterectomy 32 : 1
601 (only 2.8% of deaths were due to severe 187.4
hemorrhage)
Hazelgrove24, retrospective, 122 850 1.7 0.6 33.3% major hemorrhage not specified; 35% fetal mortality rate 7.4
UK, 2001 multi-unit 210 (ITU) were short admissions (< 2/7) and no 20% (includes fetal 30 : 1
70 specific interventions. 7 patients required losses < 24 weeks 5.7
transfer to specialist ITUs gestation)
Murphy & retrospective cohort, 51 576 0.97 0.23 24%.83 no definition/information on transfusion perinatal mortality 3.4
Charlett29, general ITU in 50 (ITU) given but cause of hemorrhage given; rate 14%. 16.7 : 1
US, 2002 teaching hospital 12 7 hysterectomies 5.8
30 August 2006 14:24:33

Vandecruys16, prospective, tertiary 26 152 11.7 1.68 14.4% definitions not given. Data on hemorrhage not collected 59.4
South Africa, center, phase 1 305 refer to PPH 5.2 : 1

1997–1999 44 225.6
Vandecruys16, prospective tertiary 13 854 8.7 1.66 19%.38 as above. SAMM and mortality declined not collected 26.4
South Africa, center, phase 2 121 compared to the first audit due mainly to 4.7 : 1
2002 (re-audit) 23 reduction in abortion complications 188.4
Pattinson11, prospective, NA NA NA 30.7% condition-based definitions same as not collected 128.4

South Africa, 3 geographic areas 423 Mantel10. Calculates mortality index but 3.3 : 1
2003 (urban and rural) 130 cannot define incidence as number of NA
deliveries not given. Hemorrhage includes
antepartum and postpartum; PPH alone is
18%. PPH is second most common cause
of SAMM but 7th cause of death
Brace19, prospective 51 165 3.8 1.9 50%.38 major hemorrhage = cases transfused at not collected 4.4
Scotland, observational, 196 least 5 units during the acute episode of .449 : 1
2004 (22 consultant-led 98 hypovolemia (13 categories of morbidity 7.8
345
units in Scotland) leading to ITU admission)
367
Gandhi26, prospective, 4 rural 5728 5.41 1.75 32%.38 Mantel’s definitions adapted for use in not collected not disclosed
South Africa, primary hospitals 31 primary hospital with limited resources.
2004 10 Includes antepartum and postpartum
hemorrhage, DIC and hysterectomy
Zhang20 Population based 182 734 9.48 4.6 48.8% blood loss > 1500 ml/blood loss requiring fetal death rate 4.8% 4.4
(MOMS-B), questionnaire, 1734 plasma expanders and/or blood loss 433.5 : 1
Europe, 2005 multi-unit, multi- 847 > 2500 ml in 24 h/blood loss resulting in 2.2
national maternal death. Incidence range 0.7–8.8
according to country

SAMM, severe adverse maternal morbidity; ITU, admissions to intensive therapy unit; (ITU), SAMM cases defined as ITU admissions; MMR, morbidity
to mortality ratio (calculated from rate of SAMM to mortality); PPH, postpartum hemorrhage; DIC, disseminated intravascular coagulation;
HD, high-dependency unit; RPOC, retained products of conception; NA, not available
*Data in parentheses are calculations as applied for five cases
populations, whilst the three regions in France delays in treatment secondary to transfers and
did not include major cities20. also ensures continuity of obstetric care23.
In general, severe hemorrhage and hyper- Obstetric HDU beds are becoming more com-
tension have much higher incidence (range monplace in tertiary centers in the UK12,13,24
0.617–29.618 and 0.1822–6.1518 per 1000 and US14,15,23. Comparisons are more valid
deliveries, respectively) than severe sepsis between studies that have used agreed or similar
(0.0919–2.1610 per 1000). The same low rate for definitions for hemorrhage10,12,18,19,25. Many of
sepsis is observed in West Africa, where the sec- them are prospective10,11,13,19,26 rather than ret-
ond greatest cause of SAMM after hemorrhage rospective14,17,27–29, and they tend to find higher
is obstructed labor18. Uterine rupture has been proportions of hemorrhage: 30–50% rather
combined with data for obstructed labor in than 10–30%, although there is some over-
one study18 and with hemorrhage in another17. lap12,18.The higher rate from prospective
Waterstone and colleagues (2001)13 considered studies is likely to be due to better case
uterine rupture as a separate entity; this is a ascertainment. Rates appear to be very
more accurate way of using the data unless we similar in developed13,19,20 and developing10,18
have clear evidence of the blood loss associated countries when comparable definitions are
with each case13. Few studies have looked at used.
immediate moderate morbidity, although a Emergency obstetric hysterectomy provides
number of studies of the puerperium examine another means of examining SAMM caused by
moderate to minor morbidity2,13. For example, postpartum hemorrhage. It has the advantage
Stones and colleagues (1991) included less of being relatively clearly defined, and is rare
severe cases of morbidity in their analysis: enough to be easy to collect data. The tradi-
anesthetic complications (usually post-spinal tional advice is to perform hysterectomy early to
headache) 0.46%; urinary retention/inconti- avoid mortality6. The threshold for performing
nence 1.2%; late perineal complications hysterectomy clearly varies with operator and
0.65%2. unit, but evidence exists that early hysterectomy
decreases morbidity30. The new United King-
dom Obstetric Surveillance System (UKOSS)
HEMORRHAGE AS A CAUSE FOR
requires units to report cases of specified rare
SAMM: THE EVIDENCE
conditions on a monthly basis. Hysterectomy
Most studies of SAMM to date report severe has been chosen as the exemplar obstetric
hemorrhage as the largest single contributing morbidity, and this large national surveillance
factor. Severe hemorrhage was defined by one should provide further information about best
or a combination of factors: practice in the future.
(1) Estimated blood loss ≥ 1500 ml (or
≥ 2000 ml); RISK FACTORS FOR SAMM AS
APPLIED TO HEMORRHAGE
(2) Hemoglobin drop ≥ 40 g/dl;
Although it is challenging to define the size of
(3) Transfusion of ≥ 4 units of blood.
the problem (i.e. the incidence of SAMM as a
Table 2 outlines the incidence of severe hemor- result of hemorrhage), it is necessary to under-
rhage in a variety of studies to date. The prob- stand the factors that increase the risk of severe
lem of varied definitions is highlighted, making hemorrhage. Table 3 has been adapted from
comparisons between studies difficult. The pro- the findings of a multicenter case–control study
portion of SAMM due to hemorrhage is also in the South East Thames region of the UK
shown. This varies widely but tends to be lower (COSMO)13 and outlines the odds ratios of
in studies that are management-based, as not all having a severe hemorrhage (as defined by
cases require admissions to ICU. Local policies blood loss ≥ 1500 ml, hemoglobin drop
and availability of obstetric HDU beds on labor ≥ 40 g/dl, or blood transfusion ≥ 4 units),
wards has a great influence on the management depending on a wide range of risk factors. The
of massive postpartum hemorrhage as it avoids main predictors are:
346
368
30 August 2006 14:24:34
Table 3 Risk factors for having a severe adverse maternal morbidity, or severe hemorrhage (from
Waterstone et al., 2001)8. Figures are odds ratios (95% confidence interval)
Odds ratios for SAMM Odds ratios for SAMM

Risk factors (all causes) due to hemorrhage
Age > 35 years 1.46 (1.11–1.92) 1.41 (1.03–1.95)

Blood pressure at booking 1.23 (1.12–1.34) 1.18 (1.06–1.31)
Black 1.16 (0.85–1.58) 0.97 (0.66–1.42)
Other race 1.93 (1.24–2.99) 1.82 (1.09–3.03)
Social exclusion 2.64 (1.69–4.11) 2.91 (1.76–4.82)
Smoker 0.68 (0.49–0.93) 0.65 (0.44–0.96)
Previous postpartum hemorrhage 2.41 (1.53–3.77) 2.74 (1.69–4.44)
Hypertension 1.10 (0.63–1.95) 0.82 (0.37–1.80)
Diabetes 1.76 (0.43–7.20) 1.85 (0.38–9.14)
Multiple pregnancy 2.21 (1.24–3.96) 2.29 (1.2–4.37)
Antenatal admission 1.75 (1.37–2.23) 1.85 (1.39–2.47)
Taking iron at booking 5.53 (2.28–13.41) 5.98 (2.28–15.65)
Taking antidepressants at booking 4.3 (0.91–1.88) 10.55 (2.19–50.71)
Taking antiepileptics at booking 5.31 (1.4–20.13) 5.75 (1.28–25.72)
IOL because overdue 1.36 (0.99–1.88) 1.38 (0.95–1.99)
IOL on medical grounds 2.45 (1.68–3.57) 1.33 (0.87–1.07)
Oxytocin augmentation 0.99 (0.76–1.28) 1.61 (1.2–2.15)
Manual removal of placenta 9.60 (5.67–16.28) 13.12 (7.72–22.30)
Emergency Cesarean 4.31 (3.39–5.49) 3.09 (2.29–4.17)
SAMM, severe adverse maternal mortality; IOL, induction of labor
(1) Demographic: age ≥ 35 years, non-white Other studies31,32 find the same trend, with
race, social exclusion (this was a composite death and severe morbidity being more com-
measure of a woman’s social deprivation mon in black women, multiparae and women
beyond the use of her marital or partner’s of ‘low status’32 as defined by poor education,
employment status. It included concealed poverty, low antenatal care attendance, low
pregnancy, age < 16 years, poor housing, contraceptive ever-use and little power to make
‘on income support’ in the notes, previous decisions regarding access to health care. In
minor or child in local authority or state Geller’s study (2004)31, the peak of mortality
care (currently or in the past), in trouble and SAMM occurred in the 20–34-year age
with the law (currently or previously), living group, with three times fewer women aged > 35
alone, unbooked, unwanted pregnancy, years in all categories of the morbidity-to-
currently or previously in foster care, care mortality continuum31. This is more likely
order being considered on potential child, to reflect the age distribution of the study
social worker involved, or drug or alcohol population rather than a true difference
dependency21); between the USA and the UK, and emphasizes
the need for the use of multiple logistic
(2) General medical factors: anemia, depression,
regression to tease out risk factors. Manual
diabetes, hypertension, epilepsy;
removal of placenta had the biggest impact13,
(3) Obstetric factors: previous postpartum in keeping with abnormally adherent placentas
hemorrhage, multiple pregnancy, ante- being a major cause of postpartum hemorrhage.
natal admissions, obstetric interventions Antenatal anemia, Cesarean section, and the
(augmentation with oxytocin, manual use of antidepressants or antiepileptics at
removal of placenta, emergency Cesarean booking also appear to have significant impacts,
section). though their relative importance is difficult
347
369
30 August 2006 14:24:34
to ascertain as the confidence intervals were to develop standardized definitions for severe
wide. Induction of labor increases the risk morbidity and its main causes. Intuitively, a
of postpartum hemorrhage regardless of the condition-based approach will be better, as it
indication13. would allow clinical comparisons to be made.
Recommendations could be more easily applied
to settings where ICU facilities are scarce. Two
OUTCOMES OF WOMEN WHO
groups10,31 have refined this clinical approach
SUFFER SAMM
by classifying SAMM according to the initial
Few studies look at outcomes beyond survival obstetric cause (e.g. hypertensive disorders,
or immediate morbidity. Studies of postnatal hemorrhage or sepsis) as well as the organ
morbidity in general (low- and high-risk women dysfunction which led to the severe illness.
analyzed together) found that the prevalence of Hemorrhage could be further classified by
problems is high (87%) and lasts up to 18 cause (atony, surgical, adherent placenta, dis-
months33. A case–control study of outcome seminated intravascular coagulopathy (DIC),
6–12 months postpartum compared women inverted uterus, etc.).
who had suffered SAMM and women who had In the context of severe hemorrhage, possible
not21. Women who had suffered SAMM were components of an international definition might
twice as likely as controls to attend general prac- include:
titioner services and three times as likely to
● Measured blood loss ≥ 1500 ml at the time
attend Accident and Emergency departments.
of pregnancy outcome (including abortion,
This may have been due in part to some under-
ectopic, vaginal delivery or Cesarean)
lying morbidity and its follow-up but clearly
points to a continuing burden on health services ● Peripartum hemoglobin drop ≥ 4 g/dl
with its personal, family and economic cost.
● Acute transfusion ≥ 4 units of blood
Cases also suffered slightly more postnatal
depression than controls (who were not entirely ● Presence of DIC or shock
‘normal’ women and included women with
● Use of additional, non-obstetric procedures,
operative deliveries and smaller hemorrhages).
e.g. hysterectomy/ laparotomy/interventional
Although this difference was not statistically sig-
radiology
nificant, cases also scored higher on the Edin-
burgh Postnatal Depression Scale. Significantly ● Blood loss resulting in vital organ dysfunc-
more cases (50%) than controls (29%) were tion/ICU admission
reluctant to re-establish sexual relations with
● Blood loss resulting in maternal death
their partners for fear of becoming pregnant,
suggesting that a negative experience in one The first three components are imprecise and
pregnancy may prevent a woman from achiev- early indicators of morbidity. Moreover, the
ing the family she initially intended21. Women amount of blood loss is notoriously inaccurate,
with stillbirths are almost always excluded from blood transfusion is practitioner- and protocol-
postnatal studies33, although a higher propor- dependent and hemoglobin decreases are not
tion of them also suffer SAMM by the nature of measured world-wide. The severity of the
underlying conditions (e.g. abruptions, diabe- impact on the woman’s health will depend on
tes). Only half of the studies of SAMM quoted her prior hemoglobin level and further manage-
give data about perinatal loss. The figures ment of her condition. Blood loss in excess
quoted above are very likely underestimates of of 1500 ml is a sensitive predictor of SAMM.
the true spectrum of postnatal morbidity. Some of the studies discussed so far12,13 identi-
fied 50% of their cases of hemorrhage by using
this measure. The latter categories are clear but
DECREASING SAMM: MOVING
late markers of severity. Potentially, a scoring
FORWARD
system could be devised to increase the accu-
Before designing studies into effective interven- racy of assessment of severity. A woman who
tions for reducing SAMM, it will be necessary loses a large quantity of blood, is adequately
348
370
30 August 2006 14:24:34
transfused with blood and blood products, and substandard care was an issue. Factors identi-
managed in an obstetric HDU is likely to suffer fied were delay in diagnosis, treatment, referral
less long-term morbidity than a woman with the and monitoring. In the UK, substandard care
same blood loss, but also with a hysterectomy was apparent in between 50%12 and 80%3 of
and ICU admission for ventilation and renal cases of hemorrhage. Regular skills drills to train
failure. staff in estimating blood loss more accurately
Based on the risk factors already identified and recognize signs of compromise are benefi-
and presently available treatments, interven- cial3,37. Blood transfusion, although possible
tions can be tested at different levels to reduce in rural South Africa, often fails because of
the rate of SAMM or to convert high-scoring depleted stocks26, even if the need for transfu-
cases to lower-scoring ones. One American sion is recognized26. When misoprostol is used
group has devised a scoring system31 which at home births in Africa, it significantly reduces
contains five categories, including organ system rates of postpartum hemorrhage and does not
failure, ICU admission, extended intubation, require a medically trained attendant36,38. New
transfusion > 3 units and surgical intervention. data39,40 suggest that sublingual administration
The maximum score is 15; women who scored may be most effective, and the women can
more than 8 were classified as near-misses self-administer easily38, further reducing the
whereas those who scored less than 8 were cost of an already cheap intervention.
classified as severe morbidity. The aim was
to refine classification at the most extreme
Secondary and tertiary care
end of the morbidity-to-mortality continuum,
thus enabling identification of the key factors Some of the best data come comes from studies
responsible for moving women along the where cases were identified from daily audit
continuum and targeting interventions that meetings10,11,16. Results of audits and research
shift women more towards morbidity rather should be fed back promptly to staff so that
than mortality9,31. improvements can be implemented. In a two-
To effectively improve outcome, change phase study, a reduction in incidence of SAMM
has to be implemented at various levels, from and maternal mortality was demonstrated over
primary-care providers, through secondary and 4 years, when recommendations from the
tertiary centers to health-care systems. first audit were implemented16. However, in the
same study, the incidence of hemorrhage was
virtually unchanged: the main factor responsible
Basic antenatal care
for decreasing SAMM and mortality was
This cannot be overemphasized, as ample evi- improved care relating to abortion16. Clear
dence shows that antenatal follow-up decreases management protocols and regular skills drills
a woman’s risk when it comes to labor and may both contribute to the maintenance of
delivery26,34. Screening for complications ante- high standards in units3,22. Non-adherence to
natally, treatment and prevention of anemia, guidelines has been identified as a risk factor
cleanliness during delivery, the presence of a for increased maternal morbidity3,37, whereas
skilled birth attendant and active management dissemination of guidelines and skills drills are
of the third stage of labor are all basic require- associated with improved adherence to the
ments advocated by the World Health Organi- agreed protocols and significant reduction in
zation35. Staff attending deliveries in the postpartum hemorrhage37.
primary-care sector need to be trained to recog-
nize postpartum hemorrhage early and have
Access, transport, institutional or
access to simple drugs to treat it (e.g. miso-
organizational change
prostol, ergometrine)36, as well as recognize
when to refer to a more specialized center26. Twenty percent of avoidable SAMM in rural
In rural South Africa, health-worker problems South Africa is due to organizational or admin-
were identified as causing substandard care in istrative causes such as the shortage of essential
35–49% of cases26 out of a total of 65% where drugs, ambulances and lack of recruitment and
349
371
30 August 2006 14:24:35
retaining of experienced staff26. These factors become standard and might lead to significant
are less prominent in the developed world. reductions in severe morbidity1.
However, implementation of guidelines and
issues such as staff training and effective audit
usually occur at organizational levels. Geller CONCLUSION
and colleagues (2004) analyzed the ‘prevent- The triennial UK Confidential Enquiry into
ability’ of events along the continuum of severe Maternal Deaths started in the first half of the
morbidity to near-miss to death and concluded 20th century and witnessed a gradual decline in
that the same factors contributed to the out- maternal mortality. The numbers of maternal
come in all categories31. These were patient- deaths in the developed world are now relatively
factors (13–20%), system factors (33–47%) and few, although still prevalent in developing coun-
provider-related factors (90%), mainly incom- tries. Severe maternal morbidity (SAMM) is
plete or inappropriate management31. Patient prevalent throughout the world, mostly due to
factors are potentially the hardest to rectify, treatable conditions. It is often poor, socially
especially in developing countries where access excluded women that suffer most. For meaning-
to education is limited. System factors figure ful comparisons to be made, standardized, sim-
higher in the US (33–47%)31 than in South ple definitions need to be designed and agreed
Africa (20%)26, possibly because failures of well- on as the benchmark for future research. Condi-
established systems (as in the US) are likely to tion-based definitions are better than manage-
have a greater impact than in settings where trans- ment-based ones9,10,13,31, as the former can be
port or administrative systems are not estab- used in poorly resourced areas26. Hemorrhage
lished in the first place as, for example, in rural accounts for the largest proportion of SAMM,
Africa26. Provider-related factors were more but it is not one of the major causes of maternal
prominent as a cause of substandard care in the mortality, at least in developed countries11. This
US (90%)31 than in primary-care settings in suggests that registering SAMM would be a
South Africa (35–49%)26. This is more likely due valuable way to monitor and improve the
to non-availability of specialist staff in the latter, quality of maternity services. The Scottish pop-
with staff performing to the best of their ability ulation survey19 shows such a register is feasible
in light of the skills they possess. US and Euro- at national level. As the causes of maternal
pean doctors working in obstetrics are specialists, deaths can be very different from the causes of
who work to agreed protocols and participate in SAMM11, it would be most useful to have the
audit and research to maintain high standards31. two enquiries running in parallel. The last trien-
nial report in the UK5, published in 2004, has
Health systems already incorporated a chapter on morbidity,
thus acknowledging the need for SAMM to
Wider factors relating to health systems can be taken into consideration. World-wide, avoid-
move a woman both up and down the risk pyra- able maternal deaths remain a paramount issue
mid for severity of morbidity. Social exclusion of basic women’s rights. Nevertheless, severe
and inequity can be tackled at governmental hemorrhages that women survive are much
level in both developing and developed coun- commoner. Understanding the relationship
tries1,34. Access to contraception, safe legal between morbidity and mortality should lead to
abortion and antenatal care can also be reductions in substandard care and the global
addressed. Special antenatal services for travel- burden of both death and long-term morbidity
lers, teenagers or the mentally ill may be set up from hemorrhage.
by health-service planners1. The health insur-
ance system in the US may play a role, as
women most at risk are often not insured31. In References
the current era of increased migration, espe- 1. Bewley S, Wolfe C, Waterstone M. Severe
cially from deprived areas or as a result of war maternal morbidity in the UK. In Maclean AB,
and conflict, the population in major cities is Neilson J, eds. Maternal Morbidity and Mortality.
changing. Access to 24-h interpreters should London: RCOG Press, 2002:132–46
350
372
30 August 2006 14:24:35
2. Stones W, Lim W, Al-Azzawi F. An investigation 15. Monaco TJ, Spielman FJ, Katz VL. Pregnant
of maternal morbidity with identification of life- patients in the intensive care unit: a descriptive
threatening ‘near-miss’ episodes. Health Trends analysis. South Med J 1993;86:414–17
1991;23:13–15 16. Vandecruys H, Pattinson RC, Macdonald A,
3. Hall M. Haemorrhage. In Department of et al. Severe acute maternal morbidity and mor-
Health, Welsh Office, Scottish Office Depart- tality in the Pretoria Academic Complex: chang-
ment of Health, Department of Health and ing patterns over 4 years. Eur J Obstet Gynaecol
Social Services, Northern Ireland, Why Mothers Reprod Biol 2002;102:6–10
Die. Report on Confidential Enquiries into Maternal 17. Bouvier-Colle MH, Salanave B, Ancel PY, et al.
Deaths in the United Kingdom 2000–2002. Obstetric patients treated in intensive care units
London: Her Majesty’s Stationery Office, 2004 and maternal mortality. Regional teams for the
4. Drife JO. Maternal ‘near-miss’ reports? Br Med J survey. Eur J Obstet Gynaecol Reprod Biol 1996;
1993;307:1087–8 65:121–5
5. Department of Health, Welsh Office, Scottish 18. Prual A, Bouvier-Colle MH, de Bernis L, et al.
Office Department of Health, Department of Severe maternal morbidity from direct obstetric
Health and Social Services, Northern Ireland. causes in West Africa: incidence and case fatality
Why Mothers Die. Report on Confidential Enquiries rates. Bull WHO 2000;78:593–603
into Maternal Deaths in the United Kingdom, 19. Brace V, Penney G, Hall M. Quantifying severe
2000–2002. London: Her Majesty’s Stationery maternal morbidity: a Scottish population study.
Office, 2004 Br J Obstet Gynaecol 2004;111:481–4
6. Department of Health, Welsh Office, Scottish 20. Zhang WH, Alexander S, Bouvier-Colle MH,
Office Department of Health, Department of et al. Incidence of severe pre-eclampsia,
Health and Social Services, Northern Ireland. post-partum haemorrhage and sepsis as a
Why Mothers Die. Report on Confidential Enquiries surrogate marker for severe maternal
into Maternal Deaths in the United Kingdom, morbidity in a European population-based
1994–1996. London: Her Majesty’s Stationery study: the MOMS-B survey. Br J Obstet Gynaecol
Office, 1998 2005;112:89–96
7. Smith JC, Hughes JM, Pekow PS. An assessment 21. Waterstone M, Wolfe C, Hooper R, et al.
of the incidence of maternal mortality in the Postnatal morbidity after childbirth and severe
United States. Am J Public Health 1984;74:780–3 obstetric morbidity. Br J Obstet Gynaecol 2003;
8. Chang J, Elam-Evans LD, Berg CJ, et al. 110:128–33
Pregnancy-related mortality surveillance in the 22. Baskett TF, Sternadel J. Maternal intensive care
United States, 1991–1999. Morbidity and Mortal- and near-miss mortality in obstetrics. Br J Obstet
ity Weekly Report Surveillance Summary 2003;52 Gynaecol 1998;105:981–4
(SS02):1–8 (http://www.cdc.gov/mmwr) 23. Mabie WC, Sibai BM. Treatment in an obstetric
9. Geller SE, Rosenberg D, Cox S, et al. Defining intensive care unit. Am J Obstet Gynecol 1990;
a conceptual framework for near-miss maternal 162:1–4
morbidity. J Am Med Women’s Assoc 2002;57: 24. Hazelgrove JF, Price C, Pappachan VJ, et al.
135–9 Multicenter study of obstetric admissions to 14
10. Mantel GD, Buchmann E, Rees H, et al. Severe intensive care units in southern England. Crit
acute maternal morbidity: a pilot study of a Care Med 2001;29:770–5
definition for a near-miss. Br J Obstet Gynaecol 25. Tang LC, Kwok AC, Wong AY, et al. Critical
1998;105:985–90 care in obstetrical patients: an eight year review.
11. Pattinson RC, Buchmann E, Mantel G, et al. China Med J 1997:110:936–41
Can enquiries into severe acute maternal 26. Gandhi MN, Welz T, Ronsmans C. Severe acute
morbidity act as a surrogate for maternal death maternal morbidity in rural South Africa. Int J
enquiries? Br J Obstet Gynaecol 2003;110:889–93 Gynaecol Obstet 2004;87:180–7
12. Bewley S, Creighton S. ‘Near-miss’ obstetric 27. Lapinsky SE, Kruczynski K, Seaward GR, et al.
enquiry. J Obstet Gynaecol 1997;17:26–9 Critical care management of the obstetric
13. Waterstone M, Bewley S, Wolfe C. Incidence patient. Can J Anaesthesia 1997;4:325–9
and predictors of severe obstetric morbidity: 28. Mahutte NG, Murphy-Kaulbeck L, Le Q, et al.
case-control study. Br Med J 2001;322:1089–93 Obstetric admissions to the intensive care unit.
14. Kilpatrick SJ, Matthay MA. Obstetric patients Obstet Gynecol 1999;94:263–6
requiring critical care. A five year review. Chest 29. Murphy D, Charlett P. Cohort study of
2002:101:1407–12 near-miss maternal mortality and subsequent
351
373
30 August 2006 14:24:35
reproductive outcome. Eur J Obstet Gynaecol 36. Prata N, Mbaruku G, Campbell M, et al.
Reprod Biol 2002;102:173–8 Controlling post-partum haemorrhage after
30. Al-Nuaim LA, Mustafa MS, Abdel Gader AG. home births in Tanzania. Int J Gynaecol Obstet
Disseminated intravascular coagulation and 2005;90:51–5
obstetric haemorrhage. Management dilemma. 37. Rizvi F, Mackey R, Barett T, et al. Successful
Saudi Med J 2002;23;658–62 reduction of massive postpartum haemorrhage
31. Geller SE, Rosenberg D, Cox SM, et al. The by use of guidelines and staff education. Br J
continuum of maternal morbidity and mortality: Obstet Gynaecol 2004;111:495–8
factors associated with severity. Am J Obstet 38. Potts M, Campbell M. Three meetings and
Gynecol 2004;191:939–44 fewer funerals – misoprostol in postpartum
32. Kaye D, Mirembe F, Aziga F, et al. Maternal haemorrhage. Lancet 2004;364:1110–11
mortality and associated near-misses among 39. Vimala N, Mittal S, Kumar S, et al. Sublingual
emergency intrapartum referrals in Mulago misoprostol versus methylergometrine for active
Hospital, Kampala, Uganda. East Afr Med J management of the third stage of labour. Int J
2003;80:144–9 Gynaecol Obstet 2004;87:1–5
33. Glazener CMA, Abdalla M, Stroud P, et al. 40. Hofmeyr GJ, Walraven G, Gulmezoglu AM,
Postnatal maternal morbidity: extent, causes, et al. Misoprostol to treat postpartum haemor-
prevention and treatment. Br J Obstet Gynaecol rhage: a systematic review. Br J Obstet Gynaecol
1995;102:282–7 2005;112:547–53
34. The mother-baby package: WHO’s guide to 41. Graham SG, Luxton MC. The requirement
saving women’s and infant’s lives. Safe Mother for intensive care support for the pregnant
1994;15:4–7 population. Anaesthesia 1989;44:581–4
35. Lazarus JV, Lalonde A. Reducing post-partum
haemorrhage in Africa. Int J Gynaecol Obstet
2005;88:89–90
352
374
30 August 2006 14:24:35
38
SHEEHAN’S SYNDROME
L. Haddock
INTRODUCTION Fifty of the 72 patients diagnosed between

1960 and 1970 were thoroughly studied (Table
In his excellent publications dating from 1938 to
1). Forty-three (86%) had panhypopituitarism,
1968, Sheehan described the natural history,
whereas seven (14%) displayed selective pitu-
clinical signs and pathological findings of the
itary deficiencies. Of the latter group, one had
syndrome which bears his name and results from
isolated thyroid stimulating hormone (TSH)
postpartum necrosis of the anterior lobe of the
deficiency, one had selective human growth
pituitary gland1–10. The exact pathogenesis of the
hormone (HGH) and gonadotropin deficiency,
disease is not well understood, for many women
two lacked HGH and ACTH and three had
who suffer severe hemorrhage at delivery appar-
combined HGH, ACTH, and gonadotropin
ently escape damage to the anterior pituitary.
deficiency.
Although infrequently reported in the US
The age at onset of disease ranged from 20 to
literature, this clinical entity was the most com-
52 years, with an average age of 33.7 years. The
mon cause of hypopituitarism among indigent
age at diagnosis varied from 27 to 65 years, with
women of Puerto Rico in the decade of the
an average of 43.8 years. The duration of pre-
1950s to the late 1960s. During that period, 100
ceding illness at the time of diagnosis ranged
cases were diagnosed in the hospital attached to
from 5 months to 28 years, with an average of
the University of Puerto Rico School of Medi-
10.5 years.
cine. Of these, 72 were diagnosed from 1960 to
1970 and came under our medical supervision.
The clinical and endocrinological evaluations of
50 of the 72 cases which were available to close Table 1 A study in 50 subjects with Sheehan’s
follow-up have been published11. This review syndrome
summarizes these findings and comments on
Number Percentage
the condition.
Panhypopituitarism 43 86
RESULTS Selective hypopituitarism
TSH deficiency 7 14
Clinical data
HGH and gonadotropin 1 2
Of 28 patients diagnosed between 1951 and deficiency
1959, 16 died of cortisol insufficiency precipi- HGH and ACTH deficiency 2 4
tated by concurrent illnesses. In contrast, only HGH, ACTH and gonadotropin 3 6
two patients died in the group diagnosed deficiency
between 1960 and 1970. This marked decrease Age at onset of diagnosis, 20–52 years (mean 33.7
in mortality was secondary to a better and years); age at diagnosis, 27–65 years (mean 43.8
regular follow-up and an improvement in years)
their education regarding the nature and Duration of preceding illness (at time of diagnosis),
life-endangering risks of the disease. 5 months–28 years (mean 10.5 years)
353
375
30 August 2006 14:24:35
The obstetric history of the 50 subjects is previa and silent rupture of the uterus at 8
summarized in Table 2. Twenty-nine patients months’ gestation. Almost half (48%) of the
(58%) delivered at home and 21 (42%) in the patients had postpartum bleeding that led to the
local government hospitals. clinical picture of Sheehan’s syndrome. Eight
Mean parity was 7 ± 3 deliveries. Forty-three patients (16%) had bleeding in earlier pregnan-
(86%) had experienced postpartum bleeding, cies but were able to conceive. Of these eight
most commonly caused by retained placental patients, two had selective ACTH deficiency,
fragments. Of the five patients who had five had a picture of panhypopituitarism at the
antepartum bleeding, one had an abortion at 7 time of study, and one had selective gonadal
months’ gestation (whether spontaneous or insufficiency. A clinical history of shock at deliv-
induced not clear from the records), two had ery was present in 34 cases (68%); seven did not
abruptio placenta at 9 months’ gestation, one develop shock and information was not avail-
had a subarachnoid hemorrhage and Gram- able in nine cases.
negative bacteremic shock at 7 months’ gesta- The salient clinical features are summarized
tion, and the remaining patient had placenta in Table 3. Gonadal insufficiency was present in
94% of the patients, cortisol insufficiency in
96% and thyroid insufficiency in 88%. The ear-
Table 2 Obstetric history of 50 patients with liest sign of pituitary failure was the inability to
Sheehan’s syndrome. Mean parity 7 ± 3
lactate. Three patients had scanty menses for
Number Percentage
Delivery Table 3 Clinical features in 50 patients with

Home 29 58 Sheehan’s syndrome
Local hospital 21 42
Symptoms and signs Percentage
Bleeding in pregnancy
Postpartum 43 86 Gonadal insufficiency 94
Antepartum 5 10 Failure to lactate 86
No history of bleeding 2 4 Loss of libido 84
Amenorrhea 88
Pregnancies complicated by bleeding Breast atrophy 74
One (last) 24 48 Vaginal atrophy 88
More than one and no bleeding 8 16 Uterine atrophy 86
in last
More than one and bleeding in 18 36 Cortisol insufficiency 96
last Anorexia 72
Weight loss 80
Source of bleeding Asthenia, weakness 98
Retained placenta 21 42 Cachexia 6
Placenta abruption 3 6
Placenta previa 2 4 Thyroid deficiency
Abortion 1 2 Cold intolerance 88
Vaginal laceration 1 2 Dry skin 94
Uterine atony 4 8 Hypoactive DTRs 94
Subarachnoid hemorrhage and 1 2 Myxedematous facies 44
septic shock
Secondary sexual characteristics
Information not available 16 32
Loss of body hair 100
Shock Loss of pubic hair 98
Present 34 68 Loss of axillary hair 98
Absent 7 14
Other
Unknown 9 10
Pallor 92
Subsequent pregnancies after 8 16 Polyuria and polydipsia 4
episode causing disease Pigmentation in the face 4
354
376
30 August 2006 14:24:36
Sheehan’s syndrome
several months after the episode of postpartum 11 had poor to no reserve (increase < 6 mg/day)
bleeding, after which they became amenorrheic. and their responses were similar to those of
Cachexia was an infrequent finding; however, Addisonians. Two patients who had limited
when present it was usually due to an associated cortisol reserve (increase in urinary hydroxy-
illness. Of three cachectic patients, two had corticosteroids respectively to 8.8 mg/ day and
pulmonary tuberculosis and one was severely 6.6 mg/day post-ACTH) died, one of cortisol
malnourished. Signs and symptoms of severe insufficiency in another hospital in the commu-
hypothyroidism were present in 22 patients. nity and the other in an accident. Both patients
Severe pallor secondary to ACTH deficiency were receiving a daily maintenance dose of
was the rule in hypopituitarism, although 25 mg of cortisone acetate. The combined
chloasma, for example pigmentation of the weight of both adrenals in one patient was 5.4 g;
face, was seen in two patients. This has been histologically, the glands showed atrophy. The
previously described by Sheehan in 19393. Two adrenals of the other patient were not weighed
patients complained of marked polydipsia and but were described as atrophied, measuring
polyuria immediately following the episode each 1.8 × 1.0 × 0.2 cm. A correlation was
of bleeding. However, symptomatology was made between the duration of illness and the
interpreted as secondary to transient diabetes adrenal reserve found in these patients. Of 14
insipidus from which the patients recovered. patients whose disease had existed for 5 years or
The study of one of these patients was the less, 12 showed an excellent or good adrenal
subject of a previous report12. reserve. Of 36 patients whose disease dated
from 5 to 29 years, 24 showed a limited, poor or
no reserve, and the remaining patients had a
Endocrinological work-up
good or excellent response.
Human growth hormone reserve was studied The aldosterone reserve and ability to
in all the patients. In the basal state, HGH conserve sodium upon sodium deprivation was
was undetectable. With estrogen-priming and studied in 13 patients, seven of whom were
challenging with either insulin-induced hypo- chosen because they had initially shown hypo-
glycemia or arginine infusion, only the patient natremia when first admitted in cortisol insuffi-
with selective TSH deficiency had HGH ciency. All had dilutional hyponatremia which
reserve, as shown by an increase to 20 ng/ml was corrected by fluid restriction and treatment
upon arginine infusion. with intravenous hydrocortisone and intramus-
Pituitary ACTH reserve was also studied in cular cortisone acetate. None of the patients
all patients using the metyrapone test. Only two received desoxycorticosterone (DOCA). Of the
patients, one with selective TSH and the other seven who had hyponatremia, one had excellent
with selective HGH and gonadal insufficiency, cortisol reserve, two had good reserve, three had
had a normal response. limited reserve and one had no reserve. The
Cortisol reserve was studied in all patients remaining six patients were selected because
before replacement therapy. The basal urinary they had either a limited or very little cortisol
hydroxycorticosteroids ranged from undetect- reserve. Additional important clinical data
able values to 1.0 mg/day in the patients who included the presence of old pulmonary tuber-
did not show pituitary reserve. The response to culous lesions in three of the patients, two of
ACTH, 40 units twice a day for 4 days whom showed no cortisol reserve and the other
(ACTHAR gel) was arbitrarily classified as very limited cortisol reserve. These patients
excellent, good, limited or none, according to were kept on a 110 mEq sodium diet for 10 days
the increase inurinary hydroxycorticosteroids and on the subsequent 7–10 days they were
after ACTH administration and the ability of placed on a 8–14 mEq sodium diet. All the
these patients to tolerate stress. Seventeen patients were kept on their maintenance dose
patients had an excellent response (increase of sodium levothyroxine and, in place of their
over 20 mg/day), nine showed a good response maintenance dose of cortisone acetate, they
(rise < 20 but > than 12 mg/day), 13 had a lim- were administered an equivalent dose of
ited reserve (increase < 12 but > 6 mg/day) and dexamethasone. Upon sodium restriction, 12
355
377
30 August 2006 14:24:36
patients attained sodium balance in 2–7 days. non-responders had a severe form of the dis-
One patient was unable to tolerate the low ease. The inadequate response to TSH tended
sodium diet and the study had to be stopped on to correlate better with the severity of the dis-
the 4th day, at which time she had not attained ease than with the duration of the illness. The
sodium balance. No appreciable change in the euthyroid group had a normal response to TSH.
serum sodium level was observed in all patients The PBI pre- and post-TSH was measured in
during sodium restriction. In nine patients, the 24 patients (15 already treated with sodium
aldosterone secretory rate (ASR) was measured levothyroxine in whom it was discontinued 3
on the 7th day of dietary sodium restriction. weeks before testing and six, all hypothyroid,
The ASR ranged from 169 to 535 µg/day. Val- but not treated; three were euthyroid). In the
ues for ASR in healthy subjects on a sodium three euthyroid patients, the increase in PBI
intake of 10–50 mEq of sodium diet are from ranged from 1.4 to 2.6 µg/dl. The six hypothy-
100 to 300 µg/day. In one patient, the ASR was roid patients had never been treated and the
measured on the 4th day of sodium restriction severity of their disease ranged from mild to
and was 283 µg/day. In the remaining patient, it severe. The change in PBI from the basal value
was measured while on the 110 mEq sodium was either decreased or had an insignificant rise
diet and was 160 µg/day. In two patients on a in the three patients with myxedema. In the
low sodium diet in whom ASR testing was not three patients with mild to moderate hypo-
performed, urinary aldosterone was measured thyroidism, as well as in four of the patients
and was considered normal. The urinary receiving treatment, the increase in PBI corre-
aldosterone was low in one but did not correlate sponded to that seen in euthyroid patients; the
with the ASR, which was normal. An adequate remaining 11 had an insignificant or negligible
response of aldosterone excretion or secretory change in PBI post-TSH.
rate therefore was seen in all patients, including
the three in whom the possibility of tuberculosis
Osteoporosis
of the adrenals was considered. The normal
response to sodium restriction and the adequate When these patients were first studied, the tech-
aldosterone excretion or ASR ruled out the nology for bone densitometry was not available.
possibility of coexistent primary adrenal insuffi- When it became available, Aguiló13 proceeded
ciency secondary to tuberculous involvement of to study a group of these patients still under our
the adrenals. care using single photon absorptiometry. Bone
mineral density, measured at the distal third of
the non-dominant arm using a Norland SPA
Thyroid reserve
densitometer, showed in 40 of these patients
The thyroid reserve was determined by measur- that their bone mineral content and bone min-
ing the 24-h radioactive iodine (RAI) uptake eral density were significantly lower than that of
before and after TSH stimulation in 32 subjects age- and sex-matched controls in Puerto Rico.
(29 hypothyroid and three euthyroid) and mea- These patients received thyroid and adrenal
suring the protein-bound iodine (PBI) before physiologic replacement therapy but no estro-
and after stimulation with 10 units of TSH daily gen replacement therapy. Twenty-three of these
for 2 days in 24 patients. The hypothyroid patients were enrolled in a longitudinal bone
group were classified as responders or non- study with the aim of studying changes in bone
responders to TSH. In 21 patients with hypo- mineral content (BMC) with passing time. At a
thyroidism, the difference in the 24-h RAI from mean of 5.5 years, ten (43.5%) had increased
the basal value ranged from 14 to 60%, with a their BMC (Group 1), nine had decreased
mean difference of 26%. The remaining eight BMC (Group 2), and four remained
patients had a response similar to that seen in unchanged. Group 1 had initial BMC
primary hypothyroidism, the difference in the measurements that were significantly lower
post-TSH 24-h RAI ranging from 2 to 9%. (0.578 ± 0.04 g/cm) than those in Group 2
Forty-eight percent of the responders had (0.764 ± 0.03 g/cm). The age of Group 1 was
severe hypothyroidism, whereas 75% of the 65.5 ± 2.6 years and that of Group 2 was
356
378
30 August 2006 14:24:36
65.2 ± 3.3 years. Group 1 was younger at the quantitative manner. Our clinical experience
onset of the disease (29.2 ± 2.1 years vs. in the use of this test is similar to that of
35.9 ± 2.5 years), and the duration of the other investigators. Liddle and associates21,
disease was 7 years longer (36.3 ± 2.6 years vs. using the urinary 17-ketosteroids and 17-
30.4 ± 3.1 years). There was no difference hydroxycorticosteroids as parameters to mea-
regarding race, body mass index, physical activ- sure the response to metyrapone, encountered
ity and sun exposure. Pertinent biochemical and absent pituitary ACTH reserve in the two
hormonal parameters showed no differences patients with Sheehan’s syndrome that they
except for serum alkaline phosphatase, which studied. Kaplan22 studied four patients with
was higher than normal upper limit (115 IU/l) Sheehan’s syndrome, none of whom showed
in both: 131 ± 17 in Group 1 vs. 121 ± 16 in pituitary ACTH reserve as measured by an
Group 2; p < 0.05. increase in the urinary Porter Silber
Upon loss of estrogen support, a rapid phase chromogens and 11-desoxycorticosteroid after
of bone loss ensues (Riggs type 1) followed by a metyrapone administration. Only two of
more gradual age-related loss (Riggs type 2). our 50 patients showed pituitary ACTH
Aguiló concluded that Group 1 resembled that reserve, both of whom had selective tropic
reported by Kruse and Kuhlencordt14 who deficiencies.
found a positive bone balance in 25% of 108 The response to intravenous and intramus-
postmenopausal females, based on histo- cular ACTH has been described as an accurate
morphometric data, and proposed a triphasic means of quantitating adrenocortical reserve in
course of osteoporosis. several disease states23,24. The administration of
ACTH, either intravenously or intramuscularly,
to hypopituitary subjects must be repeated on
COMMENT
several consecutive days in order to reactivate
Panhypopituitarism is usually characterized by a dormant adrenal cortex resulting from pro-
the sequential loss of somatotropic, gonado- longed absence of endogenous ACTH secre-
tropic, adrenocorticotropic and thyrotropic tion. In Addison’s disease, no steroid response
functions15,16. The same order holds true in to ACTH is seen, whereas in hypopituitarism
most of the cases of Sheehan’s syndrome. a gradual increase in urinary steroid excretion
Whereas hypopituitarism associated with pitu- has been described. Adrenal unresponsiveness
itary tumors often presents in an incomplete to ACTH stimulation in hypopituitarism has
form, in Sheehan’s syndrome the opposite is also been documented20,25,26. Chakmakjian20
true. Indeed, 43 (86%) of our 50 studied cases reported no increase of urinary steroid excretion
exhibited a total deficiency of the hormones after the daily intravenous administration of
produced by the anterior pituitary gland. ACTH for 5 days in five patients with Sheehan’s
Studies characterizing human growth reserve syndrome.
in adult hypopituitarism have dealt mainly with Our data support the findings of others in
pituitary neoplasms. Only eight among a total of demonstrating that a lack of adrenal steroid
79 cases included in the largest series17–20 have response to ACTH can be seen in hypo-
been cases of Sheehan’s syndrome. Of these 79, pituitarism and that it does not necessarily
only four had significant levels of HGH, none of means the presence of Addison’s disease.
whom had Sheehan’s syndrome. In this series of In our attempt to correlate the adrenal
50 patients, only one had HGH reserve and this reserve with the duration of illness, we observed
patient also had isolated TSH deficiency. Of that 90% of the patients who had the disease for
all pituitary functions evaluated, that of growth less than 5 years had a good or excellent adrenal
hormone secretion was the most consistently reserve, in contrast to 38% of the patients who
abnormal. Thus, a deficient output of growth had the disease for longer than 5 years. Thus,
hormone represents a sensitive and early index a positive correlation exists between adrenal
of pituitary failure. unresponsiveness and the duration of illness in
With the use of metyrapone, the pituitary approximately two-thirds of the patients, clearly
ACTH reserve in man can be studied in a showing that with time the adrenal cortex
357
379
30 August 2006 14:24:37
becomes progressively atrophied due to severe co-workers32 did in a group of patients with
lack of ACTH. hypopituitarism and steroid suppression. These
Aldosterone secretion is largely independent authors showed a higher urinary sodium excre-
of ACTH regulation, and patients with hypo- tion and a significant delay in increasing the
pituitarism should be able to maintain sodium ASR, although eventually a normal secretion
balance upon sodium restriction. was achieved. The mean ASR in the hypo-
The hyponatremia seen in hypopituitarism pituitary group was significantly less than in
is associated with water retention and mimics the normal or the steroid-suppressed subjects.
the syndrome of inappropriate secretion of Whether the abnormalities in aldosterone
antidiuretic hormone (ADH)27. Ahmed and secretion on sodium restriction and ACTH
colleagues demonstrated increased levels of stimulation play a role in the hyponatremia of
arginine vasopressin in the plasma of patients hypopituitarism is questionable. If this were so,
with either hypopituitarism or Addison’s dis- hyponatremia would not be corrected with fluid
ease28. A decrease in the plasma ADH activity restriction and cortisol administration alone.
and simultaneous improvement of water excre- The concept of inappropriate secretion of
tion were observed in these patients during ADH is favored to explain the hyponatremia of
therapy with glucocorticoids. Turin and col- hypopituitarism. At the time of our study, the
leagues29 demonstrated a reduction in aldo- response of the serum PBI and the RAI to TSH
sterone secretion in patients with inappropriate stimulation was the conventional test to
secretion of ADH. In a study with an untreated differentiate primary from secondary hypo-
patient with hypopituitarism who showed water pituitarism33,38. Taunton and colleagues37
retention, hyponatremia, urinary sodium loss showed that the response in hypopituitary sub-
and a low aldosterone secretory rate upon jects was less the longer they had the disease.
dietary sodium restriction, the same authors30 Sheehan3 was the first to postulate that a dor-
suggested that the decreased secretion of mant thyroid gland unstimulated by TSH might
aldosterone and wasting of sodium in chronic in time become fibrotic and therefore unrespon-
hypopituitarism are related to the persistence of sive to TSH. Fletcher and Berford39 showed
excess ADH. This concept gains strength from that the unresponsiveness to TSH correlated
the studies of Bartter and associates31 who better with the severity of the disease than with
showed that volume expansion produced by the the duration of the disease, as was the case in
administration of vasopressin and water may these patients with Sheehan’s syndrome.
increase the urinary excretion of sodium and The findings in this series of Sheehan’s syn-
decrease the excretion of aldosterone. drome suggest that, due to ACTH and TSH
Our studies favor the latter concept and, deprivation, the adrenals and the thyroid can
in order to avoid the effect of antidiuresis in become atrophied and unresponsive to stimula-
sodium excretion and aldosterone secretion, the tion to the tropic hormones; however, in the
studies of our patients were conducted during majority of the instances, the target glands
glucocorticoid and thyroid replacement. This are dormant rather than atrophied and are
past observation is of clinical and therapeutic responsive to the tropic hormones.
importance in the management of hypo- Recently, Haddock and associates have
natremia in hypopituitarism. In these patients shown an overall morphometric vertebral frac-
with Sheehan’s syndrome, hyponatremia has ture weighted prevalence of 11.2% in a popula-
always been associated with water retention, tion-based study in a female population 50 years
and prompt diuresis and correction of hypo- and older in the city of San Juan40,41. Of the 48
natremia upon cortisol administration are well females out of 398 who had fractures, 19 had an
known. In contradistinction to Addisonians, early menopause at mean age 39 ± 4.7 years.
never in our experience have hypopituitary Although the asssociation did not reach statisti-
patients needed mineralocorticoids for the cal significance, an early menopause at age less
maintenance of sodium balance. In this group than 45 years is an important risk factor for frac-
of patients, the ASR upon ACTH stimulation tures and osteoporosis, more so if patients do
has not been studied as Williams and not receive estrogen replacement. Cooper and
358
380
30 August 2006 14:24:37
associates42 found that women with vertebral Administration, the Government of Puerto
fractures had an earlier menopause, fewer births Rico, and the Department of Health of Puerto
and higher prevalence of clinically diagnosed Rico, and is accredited by the American College
hyperthyroidism. Thus, women with Sheehan’s of Nurse-Midwives. The nurse-midwife is a
syndrome who develop the disease in their registered nurse educated in two disciplines,
reproductive years are more prone to develop nursing and midwifery. The nurse-midwife
osteoporosis and fractures, for they are diag- cares for essentially healthy women before,
nosed late in the course of the disease and have during and after childbirth.
not received hormonal replacement therapy. The nurse-midwife works as an inter-
These patients received replacement therapy dependent health-team member in a setting
with cortisone acetate 25 mg daily divided in that provides physician consultation and
two doses, sodium levothyroxine in the amount referrals for complications44. Thus, the time will
that normalized the PBI, supportive treatment eventually come when deliveries in the indigent
for underlying diseases and education about the population may again be performed by mid-
disease and its life-threatening risks. A letter was wives, as was the case in the first half of the 20th
given to every patient and her immediate family, century, but this time by a skilled health profes-
stating the nature of the disease and what to sional who is part of the health team. The
do in case of illness or if taken unconscious to Department of Health of the Commonwealth of
emergency rooms of government hospitals. Puerto Rico must follow closely all the statistics
regarding prenatal and delivery care so as to
identify all cases with postpartum hemorrhage
CONCLUDING REMARKS
or the deliveries complicated with bleeding in
In the last four decades, we have seldom seen order to identify the potential cases that may
new cases of Sheehan’s syndrome in the Univer- develop Sheehan’s syndrome. Good prenatal
sity Hospital. The main reasons are improve- and delivery care is a must to be able to prevent
ments in the prenatal and delivery care of the Sheehan’s syndrome.
indigent population. Whereas, in the past, 58%
of the patients with Sheehan’s syndrome were
References
delivered at home, from the 1970s onwards all
indigent patients delivered in hospitals. Since 1. Sheehan HL. Postpartum necrosis of the anterior
1993, a Health Reform was implemented in pituitary. J Path Bact 1937;45:189
Puerto Rico based on managed care. All but one 2. Sheehan HL, Murdoch R. Postpartum necrosis
of the secondary hospitals was sold to private of the anterior pituitary; pathological and clinical
aspects. J Obstet Gynaecol Br Commonwealth
enterprise and the main care of patients is now
1938;50:27
in the hands of primary physicians who seldom 3. Sheehan HL. Simmond’s disease due to post-
refer patients to the specialists as it affects their partum necrosis of the anterior pituitary. Q J
income, which is capitated. Fewer and fewer Med 1939;8:277
physicians are practicing Obstetrics because of 4. Sheehan HL, McLetchie NGB. Simmond’s
the high insurance and fewer physicians are disease due to postpartum necrosis of the ante-
selecting Obstetrics and Gynecology for train- rior pituitary. J Obstet Gynaecol Br Commonwealth
ing. Responding to the urgent need of preparing 1943;50:27
a skilled health professional for the delivery 5. Sheehan HL, Summers VK. Syndrome of hypo-
of maternal and infant care services in Puerto pituitarism. Q J Med 1949;18:319
Rico, the Graduate School of Public Health of 6. Sheehan HL. Incidence of postpartum hypo-
pituitarism. Am J Obstet Gynecol 1954;8:202
the University of Puerto Rico Medical Sciences
7. Sheehan HL. Physiopathology of pituitary
Campus has initiated a Nurse Midwifery
insufficiency. Helv Med Acta 1955;22:234
Education Program, which offers a Graduate 8. Sheehan HL. Atypical hypopituitarism. Proc R
Certificate option and a Master of Public Soc Med 1961;54:43
Health/Nurse Midwifery43. This program is 9. Sheehan HL. The repair of postpartum necrosis
supported by the US Department of Health and of the anterior lobe of the pituitary gland. Acta
Human Service, Health Resources and Service Endocrinol 1965;48:40
359
381
30 August 2006 14:24:38
10. Sheehan HL, Davis JC. Pituitary necrosis. Br 25. Landon J, Wynn V, James OHT. The adreno-
Med Bull 1968;24:59 cortical response to insulin induced hypoglycae-
11. Haddock L, Vega LA, Aguiló F, Rodríguez O. mia. J Endocrinol 1963;27:183
Adrenocortical, thyroidal and human growth 26. Case Records of the Massachusetts General
reserve in Sheehan s syndrome. Hopkins Med J Hospital, Case 3-1964. N Engl J Med 1964;270:
1972;131:80 143
12. Aguiló F, Vega LA, Haddock L, Rodríguez O. 27. Bethune JE, Nelson DH. Hyponatremia in
Diabetes insipidus syndrome in hypopituitarism hypopituitarism. N Engl J Med 1965;272:77
of pregnancy. Acta Endocrinol 1969;60(Suppl): 28. Ahmed ABJ, George BC, González-Auvert C,
137 et al. Increased plasma arginine vasopressin in
13. Aguiló F, Mejías NL, Ortiz V. Hypopituitary clinical adrenocortical insufficiency and its
elderly female patients do not decrease bone inhibition by glucocorticoids. J Clin Invest
mineral content uniformly with time. In 1967;46:111
Christiansen C, ed. International Congress Series, 29. Turin MD, Cooke CR, Walker WG.
Osteoporosis. Excerpta Medica, 1990:507–9 Aldosterone secretion in inappropriate ADH
14. Kruse HP, Kuhlecordt F. Pathogenesis and and in altered osmoregulation. Clin Res 1966;
natural course of primary osteoporosis. Lancet 16:66
1980;i:280–2 30. Cooke CR, Lindeman RD, Adler S, et al. Persis-
15. Rabkin MT, Frantz AG. Hypopituitarism: a tent antidiuresis with hypoaldosteronism and
study of growth hormone and other endocrine sodium wasting in hypopituitarism. Am J Med
functions. Ann Intern Med 1966;64:1197 1969;47:963
16. Landon G, Greenwood FC, Stamp TC, et al. 31. Bartter FC, Liddle GW, Duncan LF, et al. The
The plasma sugar, free fatty acid, cortisol and regulation of aldosterone secretion in man. The
growth hormone response to insulin and the role of fluid volume. J Clin Invest 1956;35:1306
comparison of this procedure with other tests 32. Williams GH, Rose, LL, Jagger PI, et al. Role
of pituitary and adrenal function. J Clin Invest of anterior pituitary in aldosterone secretion in
1966;45:437 man. The Endocrine Society Program of the
17. Merimee TJ, Rabinowitz D, Riggs L, et al. Fifty First Meeting, 1969:199
Plasma growth hormone after arginine infusion. 33. Werner CH, Hamilton HB, Leefer E, et al. An
N Engl J Med 1967;276:434 appraisal of radioiodine tracer technique as a
18. Frohman LA, Horton EA, Levobitz HE. Growth clinical procedure in the diagnosis of thyroid
hormone releasing action of a pseudomonas disorder. J Clin Endocrinol 1950;10:1054
endotoxin (pirome). Metabolism 1967;16:57 34. Querido A, Stanbury JB. The response of the
19. Kohler PO, O’Malley PW, Rayford PL, et al. thyroid gland to thyrotropic hormone as an aid in
Effect of pyrogen on blood tests of pituitary the differential diagnosis of primary and second-
hormones. Observation of the usefulness of ary hypothyroidism. J Clin Endocrinol 1950;10:
the growth hormone response in the detection 1192
of pituitary disease. Clin Endocrinol 1967;27: 35. Perloff WH, Levy LM, Despoupolos L. The use
219 of thyrotropic (TSH) hormone in the diagnosis
20. Chakmakjian ZH, Nelson DH, Bethune JE. of myxedema. J Clin Endocrinol 1951;11:1495
Adrenocortical failure in hypopituitarism. J Clin 36. Schneeberg NG, Perloff WH, Levy LM.
Endocrinol 1968;28:259 Diagnosis of hypothyroidism using thyrotropic
21. Liddle G, Estep H, Kendall J, et al. Clinical hormone (TSH). J Clin Endocrinol 1954;14:223
application of a new test of pituitary reserve. 37. Taunton OD, McDaniel HG, Pittman JA. Stan-
J Clin Endocrinol 1959;19:875 dardization of TSH testing. J Clin Endocrinol
22. Kaplan NM. Assessment of pituitary ACTH 1965;25:266
secretory capacity with metopirone. J Clin 38. Skanse B. Value of the TSH-PBI test in the
Endocrinol 1963;23:945 diagnosis of hypothyroidism. Acta Med Scand
23. Renold A, Jenkins D, Forsham PH, et al. The use 1964;175:335
of intravenous ACTH. A study in quantitative 39. Fletcher RF, Berford H. A test of thyroid and
adrenocortical stimulation. J Clin Endocrinol pituitary function using thyrotropin. Clin Sci
1952;12:763 1958;17:113
24. Jenkins D, Forsham PH, Laidlaw JC, et al. Use 40. Haddock L, Suárez E, Pérez C, et al. Prevalence
of ACTH in the diagnosis of adrenal cortical of vertebral fractures and osteoporosis in Puerto
insufficiency. Am J Med 1955;18:3 Rican females: a population based study. Second
360
382
30 August 2006 14:24:38
International Congress of Public Health, 2004 43. Castro de Castañeda M. Project Director,
(abstract) Nurse Midwifery Education Program; Graduate
41. Clark P, Ragi S, Haddock L, et al. and the School of Public Health, University of Puerto
LAVOS (Latin America Vertebral Osteoporosis Rico Medical Sciences Campus (personal
Study) group. J Bone Min Res 2004;19(suppl 1): communication)
abstract F340 44. Brochure. Nurse Midwifery Education Pro-
42. Cooper C, Shah S, Hand DJ, et al. Screening gram. Graduate School of Public Health,
for vertebral osteoporosis using individual risk University of Puerto Rico Medical Sciences
factors. Osteoporosis Int 1991;2:48 Campus
361
383
30 August 2006 14:24:38
39
AN INSTITUTIONAL EXPERIENCE
C. T. Wohlmuth, J. Gumbs and J. Quebral-Ivie
The White Memorial Medical Center In the spring of 1997, the chief obstetric resi-
(WMMC) is a private community hospital dent encountered an instance of postpartum
located in East Los Angeles, California. Estab- hemorrhage, which required life-saving emer-
lished in 1913 by the Seventh-day Adventist gency hysterectomy in a young woman. Moved
Church, the WMMC promotes its mission to by her concern for her patient, by her observa-
provide quality health services, medical and tion of the morbidity associated with post-
health education, and outreach services to the partum hemorrhage, and her intellectual
Los Angeles community, with care and compas- curiosity, she conducted a literature search
sion. With a capacity of 369 beds, WMMC and encountered the publication ‘The B-Lynch
serves a densely populated community, with surgical technique for the control of massive
more than two million people living within a postpartum hemorrhage: an alternative to hys-
5-mile radius of the Medical Center. The demo- terectomy? Five cases reported’, published in
graphics of this service area reflect an ethnic the British Journal of Obstetrics and Gynaecology
homogeneity, whereby 70.7% of residents are in March 1997, by B-Lynch and associates2.
Hispanic and 7.5% Asian. The population can This article was then presented to other
be characterized as low income, with about 62% residents and staff at Journal Club, thus
of households having an annual income of introducing the then new concept of uterine
$25 000 or less1. As a direct result of this compression by brace suture. In the years since
circumstance, government-sponsored health 1997, the operation has been used regularly.
care is relied upon to a great extent and fully
insured coverage is relatively uncommon.
CASE SERIES
The WMMC conducts graduate and contin-
uing medical education activities as an affiliate The study design was as follows: cases with
institution for Loma Linda University School of B-Lynch suture utilization for severe post-
Medicine. Four residency-training programs are partum hemorrhage were identified, from
in place: Obstetrics and Gynecology, Family March 1, 1997 to March 31, 2005, at WMMC.
Medicine, Internal Medicine, and Pediatrics. Case records were reviewed, and postoperative
The Department of Obstetrics and Gynecology follow-up was conducted by telephone inter-
is comprised of 25 medical staff members, seven view and outpatient clinic chart review. The
of whom serve as full-time faculty for the historical characteristics and outcome of these
residency-training program. The remaining patients are described in Table 1.
medical staff members serve as consulting and The B-Lynch suture operation was performed
part-time faculty. The resident house staff par- on 22 patients to control intractable postpartum
ticipates in the care of all patients and its mem- hemorrhage at Cesarean section not responding
bers are an integral part of the health-care team. to uterotonic agents. In 12 instances, the
The obstetric service performs approximately B-Lynch suture was the only surgical inter-
3500 deliveries per year. As is the case in obstet- vention, whereas in ten it was combined with
ric practice of this magnitude, hemorrhage is discrete vessel ligation. The procedure, either
encountered and often is unanticipated. alone or combined with vessel ligation, resulted
362
30 August 2006 14:24:39
Table 1a Biodata and clinical details of women who had B-Lynch suture for postpartum hemorrhage
Case Age Gravidity/ GA Length of Estimated blood

number (years) parity (weeks) Reason for Cesarean section labor (h) loss (ml) Transfusion
1 19 2/0 39 4/7 Arrest of dilation 17 2000

3 25 2/1 41 6/7 Arrest of dilation, unsuccessful VBAC 22 1500 1 unit PRBC

4 27 4/0 40 6/7 Failed induction for pre-eclampsia 15 3500 5 units PRBC, 2 units FFP
5 19 2/1 43 6/7 Elective repeat Cesarean section 0 3000 2 units PRBC

6 27 2/1 42 6/7 Elective repeat Cesarean section 0 2600 4 units PRBC, 2 units FFP
8 20 1/0 41 3/7 Arrest of dilation 20 2200 3 units PRBC

9 19 1/0 38 6/7 Arrest of dilation 12 1800 2 units PRBC
363
385
11 27 2/1 41 5/7 Arrest of dilation, unsuccessful VBAC 18 1500
13 23 3/1 38 3/7 Non-reassuring fetal heart tracing, previous Cesarean section 0 1600
14 27 2/1 40 1/7 Arrest of dilation, unsuccessful VBAC 7 2000
15 38 3/1 40 5/7 Failed induction, unsuccessful VBAC 20 2200
16 33 1/0 41 6/7 Failed induction 12 2300 3 units PRBC
19 26 2/1 38 6/7 Elective repeat Cesarean section 0 2400 6 units PRBC, 1 unit FFP

21 33 4/3 40 1/7 Arrest of descent 17 1250
22 27 3/2 39 6/7 Prior Cesarean section × 2 0 2000 1 unit PRBC
VBAC, vaginal birth after Cesarean section; GA, gestational age; PRBC, packed red blood cells; FFP, fresh frozen plasma
An institutional experience
30 August 2006 14:24:39
Table 1b Biodata and clinical details of women who had B-Lynch suture for postpartum hemorrhage
Case
number Uterotonic drugs given* (IV, IM, IMY) Postpartum hemorrhage management procedures†
1 Oxytocin IV 70 U + IMY 10 U, methylergonovine ×2, carboprost ×5 Uterine artery ligation, ovarian vessel ligation, B-Lynch
2 Oxytocin IV + IMY 40 U, methylergonovine ×2, carboprost ×5, IV + IM Uterine artery ligation, B-Lynch
3 Oxytocin IV 30 U + IMY 40 U, carboprost ×5, IV + IM Uterine artery ligation, ovarian vessel ligation, B-Lynch
4 Oxytocin and carboprost Over-sewing placental bed, uterine artery ligation,
ovarian vessel ligation, B-Lynch, hysterectomy
5 Oxytocin IV 60 U, carboprost IM ×3 Uterine artery ligation, B-Lynch

6 Oxytocin IV + IMY 20 U, methylergonovine IM ×1, carboprost IM ×2 B-Lynch, hysterectomy
7 Oxytocin IV + IMY 20 U, carboprost IM ×1 + IMY ×3 Uterine artery ligation, ovarian vessel ligation, B-Lynch,
hysterectomy
8 Oxytocin IV + IMY 70 U, methylergonovine IM ×4, carboprost IM ×3 + IMY ×3 Uterine artery ligation, ovarian vessel ligation, B-Lynch
9 Oxytocin IV 40 U, carboprost IM ×2 B-Lynch
10 Oxytocin IV 20 U + IMY 20 U, carboprost IM ×1 + IMY ×1 B-Lynch
364
386
11 Oxytocin IV 50 U, carboprost IM ×4 Uterine artery ligation, B-Lynch
13 Oxytocin IV 30 U + IMY 20 U, methylergonovine IM ×1, carboprost IM ×2 B-Lynch
14 Oxytocin IV + IMY 40 U, carboprost ×4, IM + IMY B-Lynch, uterine artery ligation, hysterectomy
15 Oxytocin IV + IMY, methylergonovine IM ×1, carboprost IM ×1 + IMY ×4 B-Lynch
16 Oxytocin IV + IMY 50 U, methylergonovine IM ×2, carboprost IM ×2 + IMY ×4 B-Lynch
17 Oxytocin IV + IMY 30 U, carboprost IM ×4 + IMY ×2 B-Lynch
19 Oxytocin IV 70 U + IMY 40 U, methylergonovine IM, carboprost IM ×2 + IMY ×2, Uterine artery ligation, B-Lynch, supracervical
hysterectomy

20 Oxytocin IV, methylergonovine IM ×2, carboprost IM ×2 + IMY ×3 B-Lynch
21 Oxytocin IV + IMY 40 U, methylergonovine IM ×1, carboprost IM ×3 + IMY ×3 B-Lynch
22 Oxytocin IV, carboprost IM ×2 + IMY ×1 B-Lynch
*Methylergonovine 200 µg per dose, carboprost 250 µg per dose; IV, intravenous; IM, intramuscular; IMY, intramyometrial; †each procedure is listed in
order of execution
in control of bleeding, with uterine preservation (1) B-Lynch technique was the only uterine-
in 77% of the cases. In those instances in which preserving hemostatic surgical procedure
the etiology of postpartum hemorrhage was performed (12 cases);
uterine atony, the B-Lynch suture was success-
(2) Uterine artery ligation (nine cases, five
ful in 85% of the cases. Hysterectomy was thus
with additional ovarian vessel ligation)
avoided in 17 of the 22 cases.
was performed first, followed by B-Lynch
Table 1 describes the biodata and clinical
technique; and
details of the 22 patients. The patients’ ages
ranged from 16 to 40 years, with a mean age of (3) B-Lynch suture was performed first,
26.2 years. Ten patients were para 0, nine para followed by uterine artery ligation.
1, and one was para 6. (‘Para’ refers to parity
at the start of the pregnancy, a reference point In the 12 patients where the B-Lynch was the
that is consistent with other case reports in only surgical procedure performed to achieve
the literature. These patients had delivered hemostasis, 11 resulted in hemorrhage control
when hemorrhage was diagnosed.) The esti- with uterine preservation. The estimated blood
mated gestational ages (EGA) ranged from loss ranged from 1100 to 2600 ml. Five patients
37 to 43 weeks, with an average EGA of 40.1 received transfusion of packed red blood cells.
weeks. Two patients developed dilutional coagulopathy.
All cases were delivered by Cesarean section. Only one B-Lynch suture case required hyster-
Uterine atony was the intraoperative clinical ectomy for intractable uterine hemorrhage. That
working diagnosis for postpartum hemorrhage patient, who had undergone an elective repeat
in all 22 instances. All cases received intra- Cesarean section at term, developed severe
operative uterine compression and medical uterine atony, unresponsive to uterine manual
uterotonic agents. Of the 22, 11 achieved hem- massage and uterotonics. Despite placement of
orrhage control with the B-Lynch suture alone. a B-Lynch suture, brisk bleeding continued,
Six cases achieved control with combined and the patient developed dilutional coagulo-
B-Lynch suture and vessel ligation. The other pathy intraoperatively. In face of continued
five proceeded to hysterectomy for intractable life-threatening hemorrhage and the patient’s
bleeding in spite of placement of the B-Lynch preoperative desire and consent for permanent
suture. Two of these five were found to have sterilization, the surgeon proceeded to hysterec-
focal placenta accreta on histological study. The tomy for intractable hemorrhage. The patient
other three had no specific pathologic finding. received 4 units packed red blood cells and
Antenatal obstetric problems described in 2 units fresh frozen plasma (FFP), and had
these patients included: prior Cesarean section an uncomplicated postoperative course. Histo-
(11 cases), arrest disorder of labor (12 cases), logical study did not reveal placenta accreta.
oxytocin labor induction/augmentation (11 Nine patients first underwent uterine artery
cases), chorioamnionitis (eight cases), pre- ligation (five with ovarian vessel ligation)
eclampsia (five cases), pre-operative magnesium followed by B-Lynch technique. Six of these
sulfate (four cases), macrosomia (seven cases), nine cases resulted in hemorrhage control with
and gestational diabetes (two cases). Obviously, uterine preservation. The estimated blood loss
many patients had more than one problem that in these cases ranged from 1500 to 3500 ml.
preceded hemorrhage. Six patients received transfusion of packed red
All cases received intraoperative uterine blood cells. Two cases of coagulopathy required
compression and medical uterotonic agents FFP transfusion.
(oxytocin, methylergonovine, 15-methyl prosta- In 13 instances, the B-Lynch technique was
glandin F2α3 (carboprost)). Twelve patients did the surgical procedure applied first. Among
not receive methylergonovine, seven of whom these, one case was followed by uterine artery
had hypertensive disease. ligation to achieve hemorrhage control, and one
Three surgical approaches were noted when proceeded to hysterectomy, as previously noted.
the B-Lynch suture was used: The cases in which the surgeons chose to apply
365
first the B-Lynch technique occurred in the COMMENT

latter years of the study (after 1999). This may
In 1997, Christopher B-Lynch introduced a
reflect the surgeons’ preceding experiences of
uterine compression suture for the control
successful B-Lynch cases, thus fostering greater
of postpartum hemorrhage after Cesarean
willingness and less suspicion for the brace
section2. The B-Lynch technique or ‘brace
suture application. A total of seven surgeons
suture’ is performed at laparotomy, with a
and approximately 20 residents managed these
hysterotomy incision in the lower uterine seg-
cases.
ment. Of the five cases presented by B-Lynch,
Of the 22 cases, 17 resulted in hemorrhage
three were delivered by Cesarean section, and
control with uterine preservation. Among the
two by the vaginal route. The patients who
five B-Lynch cases that proceeded to hysterec-
delivered vaginally underwent laparotomy and
tomy, two were noted to have focal placenta
hysterotomy for B-Lynch suture placement.
accreta. Seven patients with uterine preserva-
The etiology of postpartum hemorrhage in
tion developed postoperative endometritis. Five
the original B-Lynch series was variable,
of these cases had preceding chorioamnionitis
including uterine atony, placenta previa and
during labor. There were no cases of postopera-
coagulopathy.
tive pyometrium4. None of the B-Lynch cases
In our case series, all 22 cases involved place-
required readmission for recurrent bleeding.
ment of B-Lynch suture at Cesarean section,
Twelve of 17 B-Lynch patients with uterine
and, therefore, the laparotomy and hysterotomy
preservation were contacted for follow-up 1
prerequisites of suture placement were already
month to 7 years after the operation. Five
in place. All 22 of our cases had the intra-
patients could not be reached due to changes in
operative diagnosis of uterine atony (two cases
address and phone number and, therefore, were
also had diagnosis of placenta accreta).
lost to follow-up. Nine patients were on revers-
Since the B-Lynch technique was intro-
ible contraception. One patient underwent
duced, several case reports and small case series,
laparoscopic tubal sterilization 3 months after
ranging from one to seven cases, have been
B-Lynch procedure. The uterus was described
published5–19. These publications are summa-
as normal in the laparoscopy operative note.
rized in Table 2. Of the 43 cases reported since
One patient is presently pregnant at 28
B-Lynch’s report, 35 were delivered by Cesar-
weeks’ gestation, with an uncomplicated course.
ean section, six were delivered vaginally, and
Conception occurred 1 year after the B-Lynch
two were without specification of delivery route.
procedure.
Uterine atony was the etiology in 36 of these 43
Two patients delivered live births after
cases. Placenta accreta and placenta previa were
receiving the B-Lynch procedure. One patient
the causes of postpartum hemorrhage in four
had a term vaginal delivery (vaginal birth after
instances6,8,12,15. Lower uterine segment bleed-
Cesarean section), 3.5 years after the B-Lynch
ing was reported in two cases16. B-Lynch brace
procedure, resulting in a liveborn female infant,
suture prophylactic application was reported
Apgar scores 9 and 9. A second patient under-
in one case of triplet pregnancy in a Jehovah’s
went a repeat Cesarean section, 2 years after
Witness10. One case of B-Lynch suture place-
B-Lynch procedure, at 24 5/7 weeks’ gestation
ment for uterine atony in the second trimester
for preterm premature spontaneous rupture of
of pregnancy was also reported, with suture
membranes and non-reassuring fetal heart rate
placement after uterine vacuum curettage for
tracing. No uterine anomalies or marks from
intrauterine fetal demise18.
the prior B-Lynch procedure were noted
The cumulative ‘success rate’ (number of
intraoperatively.
cases achieving control of postpartum hemor-
When the B-Lynch procedure was used
rhage with uterine preservation/total number of
alone, it was effective in 92% of cases of post-
cases) is 98%. By combining the literature cases
partum hemorrhage in our institution. When
with our series, 70 cases are identified with a
used in combination with vessel ligation, it was
‘success rate’ of 91%.
effective in 60% of cases.
366
388
30 August 2006 14:24:44
Table 2 B-Lynch suture: literature review
Number ‘Success’
Author Date of cases rate* PPH etiology Delivery route
B-Lynch et al. 1997 5 5/5 uterine atony 1 Cesarean 3

placenta previa 1 vaginal 2
DIC 2
unspecified 1
Ferguson et al. 2000 2 2/2 uterine atony 2 Cesarean 2
Dacus et al. 2000 4 4/4 uterine atony 3 Cesarean 3
placenta accreta 1 vaginal 1
Vangsgaard 2000 1 1/1 uterine atony 1 Cesarean 1
Hayman et al. 2002 3** 3/3 uterine atony 2 Cesarean 2
placenta accreta 1 vaginal 1
Wergeland et al. 2002 5 5/5 uterine atony 5 Cesarean 2
vaginal 1
not stated 2
Kalu et al. 2002 1 1/1 prophylactic 1 Cesarean 1
Danso et al. 2002 1† 1/1 uterine atony 1 Cesarean 1
Mazhar et al. 2003 2 2/2 uterine atony 1 Cesarean 2
placenta previa 1
Smith et al. 2003 7 6/7 uterine atony 7 Cesarean 7
Pal et al. 2003 6 6/6 uterine atony 6 Cesarean 6
Chaudhary et al. 2003 1 1/1 placenta increta 1 Cesarean 1
Holtsema et al. 2004 7 7/7 uterine atony 5 Cesarean 6
lower uterine 2 vaginal 1
segment bleeding
Grotegut et al. 2004 1 1/1 uterine atony 1 Cesarean 1
Hillaby et al. 2004 1 1/1 uterine atony 1 vaginal 1
(curettage for fetal
demise 2nd trimester)
Malibary 2004 1** 1/1 uterine atony 1 vaginal 1
*Number of cases achieving hemorrhage control with uterine preservation/total number of cases; **modified
B-Lynch; †B-Lynch suture combined with intrauterine balloon catheter; DIC, dilatation & curettage
Eight live births are reported after B-Lynch. after placing the B-Lynch suture14. The
Four cases were reported by El-Hamamy and Cesarean hysterotomy management in our
B-Lynch20. Holtsema and associates presented series varied, including initial closure and
two cases of live births delivered by Cesarean reopening for brace suture placement, partial
section after B-Lynch16. The seventh and eighth closure until brace suture placement, or closure
cases of live births after B-Lynch brace suture after brace suture placement.
are within the current series. Additionally, an In 2002, Hayman introduced a ‘modified’
uncomplicated 28-week gestation after B-Lynch B-Lynch suture, whereby brace sutures are
is being followed in our case series. placed without hysterotomy incision8. Hayman
Consistent with the original B-Lynch study, described placing brace sutures in a patient who
subsequently reported cases involved reopening delivered vaginally and subsequently underwent
the Cesarean section hysterotomy wound for laparotomy for postpartum hemorrhage. He
brace suture placement. Pal and associates describes an intact uterus with the lower
reported cases where the hysterotomy was left segment relatively well contracted. Using four
open until hemorrhage control was secured 1-vicryl sutures and passing the needle from
367
389
30 August 2006 14:24:45
front to back in line where a lower segment inci- However, because postpartum hemorrhage
sion would have been, uterine compression was cannot be anticipated and often occurs under
achieved with resultant control of postpartum urgent or emergent life-threatening situations,
hemorrhage. controlling for variables and randomization may
The B-Lynch technique appears to be a be exceedingly difficult to implement and of
safe and efficacious procedure. The cases of highly questionable ethical value.
postoperative endometritis responded to anti- Within the limitation of only having case
biotic treatment, and there were no injuries reports and case series in the literature and the
to bladder, ureters, broad ligament, or pelvic unlikelihood of ever having data from a ran-
side-wall vessels. Patients with long-term domized trial, the authors propose an algorithm
follow-up demonstrated resumption of menses (Figure 1) for the use of the B-Lynch technique
and normal reproductive health practices. at Cesarean section, where the prerequisites of
In 2004, Grotegut and associates reported laparotomy and hysterotomy are part of the
one case of erosion of a B-Lynch suture through delivery process.
the uterine wall17. A 19-year-old primigravida
underwent successful B-Lynch suture place-
USING THE WMMC ALGORITHM
ment at Cesarean section, using No. 0
Maxon (US Surgical, Norwalk, Connecticut, Given that 59 of 70 cases (84%) in the literature
USA). At 6 weeks postpartum, the suture were for postpartum hemorrhage due to uterine
was noted to be protruding through the atony, this diagnosis serves as a template for the
cervical os and was removed without difficulty. B-Lynch procedure. After proceeding with uter-
Sonohysterography performed 6 months after ine massage, close the hysterotomy to minimize
the operation showed a small defect in blood loss from dilated myometrial vessels.
the anterior wall of the lower uterine seg- Administer medical uterotonics: oxytocin,
ment. The authors commented that the effect methyl ergonovine in the non-hypertensive,
of the erosion on future fertility and labor is carboprost, and misoprostol21. Although
unknown. oxytocin is often given as first-line prophylaxis
Danso and Reginald11 presented one case at placental delivery, there is no evidence to
report of B-Lynch suture technique used in support any specific sequence of use. If hemor-
combination with intrauterine balloon catheter rhage persists, continue bimanual compression.
for control of postpartum hemorrhage. A As described in the original article, decreased
38-year-old primigravida underwent Cesarean bleeding with compression serves as an assess-
section at 41 weeks and 3 days gestation for fail- ment tool for the potential success of the
ure to progress. Uterine atony was encountered. B-Lynch technique2. If there is decreased
Uterine massage, oxytocin and carboprost bleeding with compression, proceed with the
uterotonics, and B-Lynch brace suture all B-Lynch suture.
were applied. Although significant reduction There is no evidence that either uterine
in bleeding was noted after these measures, artery ligation or B-Lynch is more effective or
moderate blood loss continued from the vagina. safer than the other. However, in the presence
Through the vagina, a three-way prostatic of uterine atony, the authors suggest that it
balloon catheter was inserted into the uterus makes clinical sense to proceed with surgical
and filled with 70 ml of water, resulting in uterine compression. If the standard B-Lynch
tamponade. Bleeding was reported to have technique is used, the hysterotomy is reopened.
stopped immediately. If the hysterotomy is left intact, the modified
The B-Lynch suture cases at White Memo- B-Lynch by Hayman may be utilized8.
rial Medical Center present the largest series If hemorrhage persists, after manual mas-
to date using this procedure for uterine sage, multiple uterotonic agents and B-Lynch
preservation. Reproductive outcome is also suture, proceed with step-wise uterine
reported. Theoretical limitations of this case devascularization with uterine artery ligation
series study include the small sample size and and ovarian vessel ligation22. If hemorrhage
absence of a controlled, randomized design. remains brisk after these interventions,
368
390
30 August 2006 14:24:45
Figure 1 Postpartum hemorrhage due to uterine atony: management at Cesarean section
hysterectomy must be considered. Generally, Hypogastric artery ligation has a reported

the above measures are performed over time, success rate of 50–60%, but carries with it the
with varying intervals between uterotonic agents associated risks of retroperitoneal surgery23.
and surgical procedures. As blood loss accumu- Moreover, the technique is not familiar to all
lates, continued consideration must be given to obstetric surgeons, and the use of hypogastric
the patient’s hemodynamic stability, presence of artery ligation precludes angiographic emboliz-
coagulopathy, and the potential for further ation as an option. Selective artery embolization
compromise of the patient’s already critical has a reported success rate of 85–95%, but
condition with continued delay rather than pro- requires immediate proximity of an equipped
ceeding to hysterectomy. This is especially true invasive radiology set-up and experienced
if blood replacement supplies are diminishing or personnel24. Intrauterine balloon tamponade is
non-existent. If uterine preservation remains a dependent upon availability of the appropriate
high priority, assess the patient’s stability for device11. At this point, the choice of procedure
selective artery embolization if such facilities for control of postpartum hemorrhage in the
are available, hypogastric artery ligation, or algorithm is a judgment best decided by the
intrauterine balloon tamponade. treating physician.
369
391
30 August 2006 14:25:01
CONCLUSION atony using an analog of prostaglandin F2 alpha.

In the presence of postpartum uterine hemor- 4. Ochoa M, Allaire AD, Stitely ML. Pyometria
rhage, rapid diagnosis and intervention are after hemostatic square suture technique. Obstet
paramount to effective treatment. When preser- Gynecol 2002;99:506–8
vation of the uterus is desired, the surgeon must 5. Ferguson JE II, Bourgeois FJ, Underwood PB.
be knowledgeable in those techniques that will B-Lynch suture for postpartum hemorrhage.
maximize this eventuality. The B-Lynch suture Obstet Gynecol 2000;95:1020–2
is one of a number of new options available 6. Dacus JV, Busowski MT, Busowski JD, et al.
to the obstetrician. It is an alternative uterus- Surgical treatment of uterine atony employing
the B-Lynch technique. J Maternal-Fetal Med
preserving surgical procedure that is effective
2000;9:194–6
in controlling postpartum hemorrhage from
7. Vangsgaard K. B-Lynch suture in uterine atony.
uterine atony. In our experience, the B-Lynch Ugesk Laegar 2000;162:3468
technique is technically simple, easy to learn, 8. Hayman RG, Arulkumaran S, Steer PJ. Uterine
can be performed quickly, and may be used in compression sutures: surgical management of
combination with traditional uterine preserving postpartum hemorrhage. Obstet Gynecol 2002;
procedures. 99:502–6
Recent graduates of the White Memorial 9. Wergeland H, Alagic E, Lokvik B. Use of the
Medical Center’s Obstetrics and Gynecology B-Lynch suture technique in postpartum hemor-
Residency Training Program, trained in the rhage. Tidssk Nor Laegeforen 2002;122:370–2
B-Lynch suture technique, continue to utilize 10. Kalu E, Wayne C, Croucher C, et al. Triplet
pregnancy in a Jehovah’s Witness: recombinant
the brace suture for management of postpartum
human erythropoietin and iron supplementation
hemorrhage in their practices. Their skills
for minimizing the risks of excessive blood loss.
and knowledge are testimony to the impact Br J Obstet Gynaecol 2002;109:723–5
of physicians’ commitment to scholarly activity. 11. Danso D, Reginald P. Combined B-Lynch
Through discovery, dissemination and applica- suture with intrauterine balloon catheter
tion of new approaches and new techniques, the triumphs over massive postpartum hemorrhage.
care of the patient is continually improved and Br J Obstet Gynaecol 2002;109:963
optimized. 12. Mazhar SB, Yasmin S, Gulzar S. Management
of massive postpartum hemorrhage by B-Lynch
brace suture. J Coll Physicians Surg Pak 2003;13:
ACKNOWLEDGEMENTS 51–2
13. Smith K, Baskett TF. Uterine compression
The authors wish to recognize Dr Martin sutures as an alternative to hysterectomy for
Schwartz for his contributions in creating the severe postpartum hemorrhage. J Obstet Gynecol
proposed algorithm for postpartum hemorrhage Can 2003;25:197–200
management. 14. Pal M, Biswae AK, Bhattacharya SM. B-Lynch
This chapter has been adapted from brace suturing in primary post-partum haemor-
Wohlmuth CT, Gumbs J, Quebral-Ivie J. rhaging during cesarean section. J Obstet
B-Lynch suture: a case series. Int J Fertil Gynaecol Res 2003;29:317–20
Women’s Med 2005;50:164–73. 15. Chaudhary P, Sharmas S, Yadav R, et al.
B-Lynch brace suture: an effective method of
conservative surgical management of placenta
increta. Kathmandu University Med J 2003;2:
References
149–51
1. United States Census Bureau, 2000 Census data 16. Holtsema H, Nijland R, Huisman A, et al.
2. B-Lynch C, Coker A, Lawal AH, et al. The The B-Lynch technique for postpartum
B-Lynch surgical technique for the control of haemorrhage: an option for every gynaecologist.
massive postpartum haemorrhage: an alternative Eur J Obstet Gynaecol Reprod Biol 2004;115:
to hysterectomy? Five cases reported. Br J Obstet 39–42
Gynaecol 1997;104:372–5 17. Grotegut CA, Larsen FW, Jones MR, et al.
3. Hayashi R, Castillo M, Noah M. Management of Erosion of a B-Lynch suture through the
severe postpartum hemorrhage due to uterine uterine wall. J Reprod Med 2004;49:849–52
370
392
30 August 2006 14:25:01
18. Hillaby K, Ablett J, Cardozo L. Successful use of postpartum hemorrhage unresponsive to

of the B-Lynch brace suture in early pregnancy. oxytocin and ergometrine: a descriptive study.
J Obstet Gynaecol 2004;24:841–2 Obstet Gynecol 1998;92:212–14
19. Malibary AM. Modified B-Lynch technique for 22. AbdRabbo SA. Stepwise uterine devascularizat-
the control of massive postpartum hemorrhage. ion: a novel technique for management of
An alternative to hysterectomy. Saudi Med J uncontrolled postpartum hemorrhage with
2004;25:1999–2000 preservation of the uterus. Am J Obstet Gynecol
20. El-Hamamy, B-Lynch C. A world wide 1994;171:694–700
review of the uses of the uterine compression 23. Clark SL, Phelan JP, Yeh SY, et al. Hypogastric
suture techniques as alternative to hyster- artery ligation for obstetric hemorrhage. Obstet
ectomy in the management of severe post- Gynecol 1985;66:353–6
partum hemorrhage. J Obstet Gynaecol 2005;25: 24. Hansch E, Chitkara U, McAlpine J, et al. Pelvic
143–9 arterial embolization for control of obstetric
21. O’Brien P, El-Refaey, Gordon A, et al. Rectally hemorrhage: A five-year experience. Am J Obstet
administered misoprostol for the treatment Gynecol 1999;180:1454–60
371
393
30 August 2006 14:25:01
40
FAMILIAL CONSEQUENCES
V. Walvekar and A. Virkud
INTRODUCTION CONSEQUENCES OF POSTPARTUM

HEMORRHAGE
Indian studies on maternal mortality reveal
hemorrhage as the leading cause of maternal Consequences to the mother
death1. What is particularly disturbing is that
The problem is more than one of mortality.
the women who survive postpartum hemor-
Little doubt exists that for every woman
rhage are often not well and that this problem
who dies, there are at least 20–30 who suffer
receives very little attention. Indeed, out of
long-term disabilities.
nearly 120 million women who give birth each
The immediate consequences include:
year, it is estimated that more than 60 million
will develop some kind of complication. Of (1) Failure of or impaired lactation leading
these, 15–20 million will develop moderate- to to an undernourished child prone to infec-
long-term disabilities, however small or large2. tions, especially if born preterm;
Under these circumstances, the global problem
is staggering. As the woman is both a wife and a (2) Anemia leading to susceptibility to post-
mother, she is the center of the family’s exis- partum infections, ranging from simple
tence; her mortality or even morbidity affects endometritis to severe puerperal sepsis,
the functioning of the entire family, as well as which in itself can be a cause of maternal
the nurturing of her children. Of equal impor- death later in the puerperium;
tance, the woman is an important part of a (3) Deterioration of existent diseases, especially
country’s workforce, so her disability has an in mothers who are anemic, or have tuber-
impact on society as a whole. culosis, HIV, and cardiac lesions, all of
which often create disabling complications,
more so if there has been a prior surgical
THE HIDDEN PROBLEM IN intervention for the management of
DEVELOPING NATIONS postpartum hemorrhage;
The sociocultural climate in developing nations (4) Postpartum mental alterations, especially
is such that a woman’s health is given least pri- in the sick mother who cannot fend for
ority. Hence, maternal morbidities receive little her child. Here, the stage is set for a
or no attention, either from the society as a classic postpartum depression or even
whole or from the government and it agencies. psychosis, which may end in dire conse-
In stark contrast, a death causes major distur- quences such as suicide or infanticide in
bance in the family environment, and its poten- some cases.
tial resolution, e.g. a single father or remarriage, Delayed consequences may include:
often complicates a previously bad situation.
Unfortunately, no real evidence-based studies (1) Prolonged collapse with pituitary failure
are available to highlight this hidden problem. It and Sheehan’s syndrome, with resultant
is worthwhile, however, to look at the issue from secondary amenorrhea and infertility (see
a qualitative point of view. Chapter 38);
372
394
30 August 2006 14:25:01
Familial consequences
(2) Premature aging, apathy and mental confu- Consequences to the family and society
sion3;
A mother’s disability profoundly affects the
(3) Chronic and debilitating anemia. Between family and the community at large due to
50 and 90% of pregnant women world- changes in the household responsibilities and
wide, with or without prior postpartum finances:
hemorrhage, suffer from this problem.
(1) The cost of her treatment can cripple the
The causes of anemia include inadequate
family finances;
dietary intake of iron, folic acid, and
vitamin A, and anemic losses due to (2) Her reduced productivity can affect family
parasitic infestations and malaria. Women income and may force the children to leave
with severe anemia are more vulnerable to school, enter the labor force and/or assume
infection during pregnancy and childbirth, domestic responsibilities;
are at increased risk of death due to obstet-
(3) Children often are neglected, undernour-
ric hemorrhage, and are poor operative
ished and have health problems;
risks in the event that Cesarean delivery is
required. World-wide, anemia is considered (4) Some surviving children may be forced into
the most important indirect cause of child prostitution. Of the estimated 2.3 mil-
maternal mortality and morbidity. WHO lion women who make their livelihoods in
data estimate that anemia associated with prostitution, a quarter are minors;
maternal causes in less developed countries
(5) The emotional cost to the family may be
in 2000 alone resulted in a loss of women’s
manifest by psychopathic behavior either in
productivity valued at more than US$5
surviving children or in the father.
billion4.
If such are the potential consequences when the
mother survives, it is logical to ask what happens
when she does not?
Consequences to the children
The same postpartum hemorrhage that Death of the mother
threatens women’s survival can also cause
The consequences of maternal death are
death and disability in newborns. The vast
dramatic, not only for the family but also for
majority of the estimated 8 million perinatal
the medical community and the society at large.
deaths that occur annually in less developed
countries are associated with maternal health
problems or poor management of labor and Emotional cost
delivery5. As an illustration, obstructed and
(1) The family is shattered as the central and
prolonged labor, both important causes
sustaining core is suddenly withdrawn;
of postpartum hemorrhage, asphyxiate an
estimated 3% of newborns, resulting in death (2) The children are suddenly orphans, at the
for nearly 25% of these infants and brain mercy of their relatives and institutions;
damage for another 25%. In addition, women some may become delinquent or street
suffering from severe anemia resulting from children;
postpartum hemorrhage are more likely to
(3) The father is lost, emotionally and finan-
have low birth-weight infants (< 2500 g) in
cially, and may blame the newborn, an
subsequent pregnancies. These low birth-
event which often proves disastrous for the
weight infants are 20–30 times more likely to die
surviving child(ren);
in the first week of life than infants of normal
weight, and those who survive are more likely (4) Medicolegal suits against the doctor and/or
to suffer neurological disabilities including the hospital may come forward out of
cerebral palsy, seizures, and severe learning desperation, anger or even the desire for
disorders2. vengeance.
373
395
30 August 2006 14:25:02
Children pregnancy. Adequate supervision helps to antic-

ipate, diagnose and treat many problems such
Orphan children are more likely to become
as pregnancy-induced hypertension and anemia
juvenile delinquents or wayward members of
before their severity takes a grave turn.
the society, often leading a life of petty and
serious crime or begging. They are also at risk
of physical and/or sexual abuse by family or
Role of the skilled attendant
community members.
The term ‘skilled attendant’ refers exclusively to
people with midwifery skills (for example, doc-
The father/husband
tors, midwives, nurses) who have been trained
(1) He may remarry for the sake of children, to proficiency in the skills necessary to manage
which may or may not be beneficial and normal deliveries and diagnose or refer obstetric
may lead to destruction of the original complications.
family unit; Ideally, skilled attendants live in, and are
part of, the community they serve. They must
(2) He is at risk for depression, reduced income
be able to manage normal labor and delivery,
and dwindling resources. This picture is not
recognize the onset of complications, perform
pleasant but the story goes even further;
essential interventions, start treatment, and
(3) He may initiate medicolegal proceedings supervise the referral of mother and baby for
out of anger or financial need. interventions that are beyond their competence
or not possible in the particular setting6.
Depending on the location, other health-care
Consequences to the society at large
providers, such as auxiliary nurse/midwives,
Today, women form an important world-wide community midwives, village midwives, and
workforce, contributing immensely to the health visitors, may also have acquired appro-
growth and development of nations. This priate skills if they have been specially trained.
prospect is seriously weakened by the long-term These individuals frequently form the backbone
impact of problems following childbirth such of maternity services at the periphery, and preg-
as postpartum hemorrhage. It is very aptly said nancy and labor outcomes can be improved by
that ‘A woman’s health, a nation’s wealth’. making use of their services, especially if they
What is more important is that not only an is are supervised by well-trained midwives.
an effective workforce in place with healthy Home visits also give health workers the
women, but also that the national cost of health chance to educate women about diet and
care can diminish. In India for example, health healthy behaviors and to offer women nutri-
and family welfare ministries in various states tional supplements. This health awareness goes
run and subsidize many public hospitals and a long way. Antenatal care providers should
medical colleges. These hospitals provide medi- inform women about the importance of
cal services at a nominal cost, as the actual cost safe delivery with a skilled birth attendant, the
is subsidized by the government. By reducing warning signs of complications, and how to plan
preventable maladies, the national health-care for emergency care. In developing nations such
cost can diminish by a ripple effect. as India, the importance of a hospital delivery,
which can provide an environment which is
safer for delivery and childbirth, can never be
MEASURES TO REDUCE THE RISK OF
overemphasized (see Chapter 49).
POSTPARTUM HEMORRHAGE AND
ITS IMPACT
Role of the obstetrician Role of the obstetric community
WHO recommends four prenatal visits during The national body of obstetricians, The Federa-
pregnancy as a minimum. The initial visit tion of Obstetricians and Gynecologists of
should be within the first 3 months of India (FOGSI) recognizes this need and has
374
396
30 August 2006 14:25:02
Familial consequences
implemented the following programs (see also 2. Murray C, Lopez A, eds. Health Dimensions
Chapter 49): of Sex and Reproduction, Vol. 3. Global Burden
of Disease and Injury Series. Boston: Harvard
(1) Reproductive and Child Healthcare: University Press, 1998:170–4
under this banner, in collaboration with 3. Barton R, Burkhalter. Consequences of Unsafe
UNICEF, various awareness and training Motherhood in Developing Countries in 2000:
programs for trained birth attendants Assumptions and Estimates from the REDUCE
(TBA), and doctors at primary health Model. In Murray C, Lopez A, eds. Health
centers are conducted to handle emergency Dimensions of Sex and Reproduction. Bethesda,
obstetrics cases; MD: University Research Corporation,
unpublished, 170–4
(2) Emergency Obstetrics Care (EMOC) 4. Murray C, Lopez A. Health Dimensions of Sex
program of FOGSI: in collaboration with and Reproduction; Burkhalter, Consequences of
Macarthur Foundation; FOGSI has initi- Unsafe Motherhood in Developing Countries in 2000;
ated the training of doctors in three states Table 5
of India to deal with complications of 5. Tsui A, Wasserheit JN, Haaga JG, eds. Reproduc-
tive Health in Developing Countries. Washington,
pregnancy and labor in rural areas of India.
DC: National Academy Press, 1997:120–3
In summary, this problem is huge; the efforts 6. Coverage of maternity care. Geneva: World Health
needed are Herculean, the resources inadequate, Organization, 1996 (unpublished document
and the consequences far-reaching. It is only the WHO/FRH/MSM/96.28). http://www.who.int/
persistent will that can minimize the problem, if reproductive-health/publications/reduction_
of_maternal_mortality/reduction_maternal_
not eradicate it!
mortality_chap4.htm
References
1. Daftary SN, Desai SV, eds. Selected Topics in
Obstetrics and Gynecology, Vol 1. Dehli: BI
Publications Pvt. Ltd, 2005:115
375
397
30 August 2006 14:25:02
41
LITIGATION: AN INTERNATIONAL PERSPECTIVE
K. J. Dalton
INTRODUCTION by this time Florence Westrup was dead. The

local police charged the husband with involun-
The history of litigation after postpartum hem-
tary manslaughter, and this was said to have
orrhage spans more than 100 years, but only 34
been committed:
decided cases have been reported in common
law jurisdictions. ‘by wilfully neglecting to furnish his wife . . . with
The LEXIS database includes reported legal such care and attention as were necessary during
cases from the common law jurisdictions, but it her confinement in childbirth, thereby causing
does not include civil law jurisdictions such as her death’.
those that use Napoleonic law. This history was He was tried in Campbell Circuit Court, found
compiled using the following search terms: guilty and sentenced to 8 months imprison-
[(post-partum OR postpartum) AND (haemor- ment. He appealed this decision to the
rhage OR hemorrhage)]. First, databases of Kentucky Court of Appeals, which expressed
English, Commonwealth and Irish, US Federal its own view of the matter1:
and US States case law were searched. Then
‘Those of us who reverence the medical profes-
full-text or abbreviated-text reports of all poten- sion and implicitly trust the learning and skill of
tial cases were searched visually for key words to the family physician . . . [take the view that] . . .
determine the relevance of each for inclusion. postpartum hemorrhage is nearly always fatal
Most were discarded as irrelevant, for example: [and that] . . . the trial judge should have
‘retinal hemorrhage in the postpartum period’; peremptorily instructed the jury to find appellant
after this only 34 relevant cases remained. It is not guilty’.
possible that some cases from lower courts may
Nowadays courts are rarely so deferential to the
have been missed, as no straightforward method
medical profession or to physicians and, as is
exists to retrieve all such cases across all the
shown in numerous other chapters of this book,
jurisdictions studied.
fatality is less likely if physicians are present and
well prepared to treat hemorrhage.
FIRST MATERNAL DEATH LITIGATED
(1905)
UNLAWFUL PRACTICE OF MEDICINE
Half (17) of 34 (i.e. 50%) of the litigated cases
(1907)
involved a maternal death. The first of these
occurred in the US. On 27 February 1905, In 1907, Hannah Porn, a diplomate of the Chi-
Florence Westrup delivered her first child at cago Midwife Institute and a practising midwife
home outside Newport, Kentucky. She had of many years experience, was charged with
‘a great aversion to physicians’, and planned a practising medicine unlawfully. Among the rea-
natural home birth. The birth of the child (at sons cited was the fact that she had used ‘formu-
term) went well, but she began to hemorrhage. lae’ for treating uterine inertia and postpartum
Despite her protests, her husband called the hemorrhage, and also used obstetrical forceps
family physician. He arrived, examined her, and for delivery. These were ‘acts confessedly per-
found a retained placenta. He went home to formed by the defendant’ but she did so only
fetch his bag of instruments and returned, but rarely, and ‘never, if a physician could be called
376
398
30 August 2006 14:25:02
Litigation
in time’. Nevertheless, she was convicted, and this negligence the physician could and would have
on appeal the Supreme Court of Massachusetts reached the plaintiff’s house in time to save the life of
upheld her conviction on the grounds that:2 his wife’. He won his case, and he was awarded
‘The maintenance of a high standard of
$5000 in compensation. The telephone com-
professional qualifications for physicians is of pany appealed the decision to the Court of
vital concern to the public health.’ Appeals of Georgia, but their appeal failed4.
The court held that generally failure of equip-
Here, the Kentucky deference to physicians was ment in the telephone exchange would not be
not afforded to a midwife. negligent, but in this case there was a failure of
diligence on the part of the telephone operator
DANGEROUS SIDEWALK (1908) in that he did not notice the incoming call.
The second maternal death case was heard in

1908. Mollie Short, the wife of an East St Louis ROAD TRAFFIC ACCIDENT (1930)
physician, was 36 weeks pregnant. Out shop- More than 20 years were to pass after the case
ping on the evening of 17 November 1906, she of Mrs Glawson in 1909 before another post-
walked along a wooden sidewalk situated 6 feet partum hemorrhage case reached the courts and
above the ground (i.e. a boardwalk). This had was reported. This was to be the first road traffic
been damaged in the cyclone of 1896, but had accident in pregnancy that was litigated.
not been properly repaired. Her left leg slipped In 1930, only 2 days after Mrs Peterson con-
down a hole, she dislocated her hip, and subse- ceived her second pregnancy, she was involved
quently went into preterm labor. Although the in a road traffic accident near St Paul, Minne-
baby survived, she suffered a postpartum hem- sota. The automobile in which she was travel-
orrhage from which she died. Her husband sued ling overturned. It was said to have been going
the city authority for having a dangerous side- too fast, but the driver claimed that a tire blew
walk, and was awarded damages of $5700. He out. By the end of pregnancy, it was recognized
successfully argued that postpartum hemor- that she had a central placenta previa, in which
rhage was a direct consequence of the preterm the maternal mortality was known to be ‘very
labor, which would not have happened had not high’. Her doctor consulted with another expert.
the sidewalk been dangerous. On appeal, the Rather than carrying out the then relatively rare
trial court’s verdict was affirmed3. operation of Cesarean section, it was advised
that she should be delivered vaginally. Her
doctor used what was termed the ‘Vorhees bag
TELEPHONE PROBLEM (1909)
method’, and he broke through her placenta by
At 3 am on an October morning in 1909 in the vaginal route. The child died, the mother
Georgia, Mrs Glawson started bleeding in a had a postpartum hemorrhage and she died too.
pregnancy of unknown gestational age. Her The driver of the car in which she had been sit-
husband telephoned the local physician who ting 9 months previously was sued for negligence.
was situated 7 miles away. He advised that In court, expert medical evidence said the road
certain remedies be applied, but these did not accident had caused the placenta to be situated
ameliorate the situation. The husband repeat- in a previa position, and this directly led to the
edly tried to make telephonic contact again with mother’s postpartum hemorrhage and death. This
the physician, but the telephone operator did evidence did not convince the jury, however,
not answer for over 2 hours. Eventually, con- who found in favor of the driver. An appeal to
nection was re-established with the physician the Supreme Court of Minnesota failed5.
who set off to visit the home immediately.
By the time he arrived, Mrs Glawson had mis-
IATROGENIC OBSTETRIC INJURY
carried, had a ‘postpartum hemorrhage’, and died.
(1955)
The husband sued the telephone company for
gross negligence in not answering his telephone Occasionally, maternal death has occurred as
call for 2 hours. His lawyer argued that ‘but for a result of unusual management of labor. In
377
399
30 August 2006 14:25:02
1955, Bette Goff had her labor induced by sued the City and County of San Francisco, but
means of pituitrin. During the labor, her doctor he lost his case as the court held that the state
diagnosed a constrictive band of cervical mus- was immune from suit, in a manner akin to sov-
cle, and he incised it just left of the 12 o’clock ereign immunity. The appeal court judges said
position. She delivered vaginally, but the they were unhappy in delivering this decision,
cervical incision was not repaired. She had a but they were bound to follow the precedent of
postpartum hemorrhage over the course of the other cases in which state immunity had been
next few hours, but the two attendant nurses the issue, explaining themselves as follows8:
did not recall the doctor until it was too late,
and the patient died of blood loss. The family ‘This doctrine of non-liability of the state and
its agencies for injuries caused by the negligence
took legal action against the doctor and the hos-
of an employee engaged in the discharge of a
pital as it was vicariously liable for the nurses’ governmental function originated in the fiction
omissions. For legal reasons, the case went to that the king can do no wrong.’
retrial6. Negligence on the part of the doctor
was admitted. As for the nurses, this was evi- [In English law, the Queen is still regarded as
denced from the records. There was no later above the law, but her ministers of state (i.e. the
report on this case, so presumably it settled. government) are not above the law, and often a
court will find against them.]
In 1955, Josephine Gillen delivered at the
HEALTH INSURANCE (1956)
Brooke Army Hospital in Texas. She then had a
Postpartum hemorrhage has occasionally been postpartum hemorrhage and she was given a
at issue in insurance matters. The earliest blood transfusion. Her condition deteriorated,
reported case was that of Juanita Whitten in and 2 days later she died of renal failure. The
1956. Her health insurance policy covered hos- family sued the United States of America,
pitalization for any complication of pregnancy. alleging negligent military medical care which
She had had seven pregnancies: two miscarried included the claim that there had been an
with severe bleeding, and she had a severe post- incompatible transfusion of rhesus O-positive
partum hemorrhage following the delivery of blood into a rhesus O-negative patient, and that
her last child, after which she was sterilized. Her this led to her renal problem. In defence, it was
gynecologist said the sterilization operation was claimed that the patient was in fact rhesus
undertaken to prevent further postpartum hem- O-positive, and she had been given rhesus
orrhage, a complication of pregnancy that was O-negative blood, which would have been a
covered by her insurance policy. However, her group-compatible transfusion. The court found
insurance company and the Court of Appeals of that there had been no incorrect blood transfu-
Alabama disallowed her reimbursement claim, sion, no renal problem arising from this, and no
on the grounds that her policy covered only negligence in the medical care. This finding was
actual complications, and not potential compli- affirmed on appeal9.
cations that might or might not occur in the More than 15 years passed until the case of
future7. Theda Parker in 1972. Her third labor was
induced at 38 weeks gestation at her request.
The birth went well, but she had a postpartum
TRANSFUSION OF THE WRONG
hemorrhage, and her obstetrician had to per-
BLOOD (1951, 1955, 1972)
form a hysterectomy. During the course of the
Three cases involved allegations that the wrong operation, she needed a blood transfusion, but
blood was transfused. unfortunately she was given blood that had been
In 1951, Mrs Madison bled heavily post- cross-matched for another patient. She survived
partum whilst in San Francisco Hospital, a the ordeal, but in the long term she developed
county hospital and a state governmental insti- hematuria due to cystitis, and her marriage
tution. Unfortunately, she was given a blood eventually broke down. In 1976, she and her
transfusion that had been incorrectly cross- husband sued her obstetrician for inducing her
matched, and she died as a result. Her husband labor too soon (for convenience rather than for
378
400
30 August 2006 14:25:02
Litigation
medical reasons) which they said led to the ordered limited disclosure of his medical
postpartum hemorrhage; and for the transfusion records11.
error which they claimed had triggered the In 1981 Matsuko Gaffney, the wife of a US
events that led to their marital breakdown. On naval man, was booked to deliver at the Long
appeal, most of their claims were dismissed, Beach Naval Hospital in California. Her preg-
except that she was awarded $20 000 compen- nancy went overdue by 4 weeks (sic), but her
sation to be paid by the hospital for the negli- cervix was judged unfavorable for induction
gence of its employee in mixing up the bloods10. of labor. She was delivered vaginally, but had
a postpartum hemorrhage for which she was
transfused two units of blood. Various experts
INFECTION FOLLOWING BLOOD
later agreed that, if she had had appropriate fetal
TRANSFUSION (1981, 1982, 1985)
monitoring, fetal distress would have been rec-
Four cases have been litigated where blood- ognized, and she would have been delivered by
borne infection occurred following trans- Cesarean section, without intrauterine death,
fusion for postpartum hemorrhage. Three cases infection, postpartum hemorrhage, and blood
involved HIV, and one hepatitis C. transfusion, all of which she did have. In 1983,
she delivered her next child, a healthy girl, and
then in 1985 she delivered a boy. He proved to
HIV
be a sickly child and was diagnosed with AIDS,
AIDS was recognized in 1982, and the HIV from which he died in 1986. Mrs Gaffney and
virus was identified in 1983. Shortly thereafter, her husband were tested for HIV and both
HIV infection was first reported as a conse- proved positive. She died of AIDS in 1987.
quence of postpartum hemorrhage. In 1984, the After her death, a 1990 Court heard that one
HIV-ELISA test was first marketed as a kit, and of her units of blood came from ‘a donor who
the FDA approved it for sale on 2 March 1985. had engaged in homosexual activity involving the
Only 11 days later, on 13 March, the Belle exchange of bodily fluids’, although he was never
Bonfils Memorial Blood Center in Denver, Col- actually tested for HIV. The Court found that,
orado took delivery of its first testing kit, but its as the United States of America was responsible
staff were not yet trained in its use. On that very for the military hospital, it was liable for the
same day, Mrs KW was admitted to hospital unfortunate train of events that befell Mrs
with a secondary postpartum hemorrhage fol- Gaffney and her family, even though HIV infec-
lowing an apparently uneventful delivery of her tion had not been discovered at the time. It held
baby son 2 weeks earlier. Her bleeding could that the United States was negligent in the treat-
not be stopped and so a hysterectomy was car- ment of Mrs Gaffney, that she needed to be
ried out. Six units of blood were transfused, transfused as a direct result of that negligence,
none of which were tested for HIV. However, and that it was foreseeable in 1981 that a com-
by 1986, donor blood was being routinely tested municable disease could be transmitted through
for HIV, and at this time one of her 1985 donors blood transfusion12.
tested positive. All previous recipients of his In contrast to this was the case of Sheri
blood were tracked and tested, and Mrs KW Traxler, who delivered her baby in 1982. Two
was found to be HIV-positive. She (and her weeks later, she had a major postpartum hemor-
husband and son) sued Belle Bonfils Memorial rhage, for which she was transfused two units
Blood Center on the grounds that the Center of blood. Hysterectomy was considered, but it
had not appropriately identified and excluded proved unnecessary. Eight years later, in 1988,
this donor as ‘not a suitable person’ to donate it emerged that one of her blood donors had
non-infected blood. (Specific testing for HIV, tested positive for HIV, and now she too tested
per se¸ was not an issue in this case.) Most of positive. She sued her 1982 obstetrician on two
the legal arguments in the case revolved around principal grounds: (1) that he had not removed
confidentiality issues regarding access to the her placenta completely, and (2) that she had
donor’s medical records, and so they are not not specifically consented to any blood trans-
relevant here. The Supreme Court of Colorado fusion. His defence was (1) that retention of
379
401
30 August 2006 14:25:03
placental fragments occurs commonly, and (2) of postpartum hemorrhage, as she was to
that her written general consent to treatment become the ‘representative plaintiff’, or lead
provided sufficient authority for him give blood case, in this mass tort action. As her case raised
as she had lost 30–40% of her blood volume. novel legal points that were challenged by the
The lower court held that there had been no CRC, it fell to the Supreme Court of British
negligence at the times of delivery or of the Columbia to grant her membership of this class
postpartum hemorrhage, and that the risk of action. Because the final outcome of her legal
HIV infection could not be foreseen. This deci- action was not reported, it is possible that the
sion was upheld by the Californian Court of matter was settled out of court14.
Appeal13.
DELAY IN TRANSFUSING BLOOD
Hepatitis C (1984, 1988, 2000)
Blood transfusion following postpartum hemor- In several cases it was alleged that there
rhage may cause other blood-borne infections, was unnecessary delay in giving blood after
such as hepatitis C. In 1988, Anita Endean postpartum hemorrhage.
delivered vaginally in British Columbia. She had In 1992, a Saskatchewan court considered
a postpartum hemorrhage, and she was given a the dangers of postpartum hemorrhage in a
transfusion of packed red cells supplied by the rural setting. In 1984, Corrine Naeth had deliv-
Canadian Red Cross (CRC). After she went ered her baby uneventfully in Hospital A, but
home, she had a debilitating flu-like illness. Six her uterus inverted when ‘controlled cord trac-
years later in 1994, she offered to donate blood, tion’ was used to deliver the placenta. Before
but she now tested positive for hepatitis C. replacing the uterus, the delivering doctor tried
Although its short-term effects are transient, to peel the placenta off the inverted uterus, but
hepatitis C carries a long-term risk of cirrhosis the placenta was adherent (placenta accreta).
(10% per annum) and in those patients a Massive hemorrhage ensued, but there was no
further risk of hepatocellular carcinoma (5% per blood transfusion facility in the hospital. She
annum). The CRC carried out a ‘traceback’ was then transferred by ambulance to Hospital
procedure, and found that one of her 1988 B, a traveling distance of 90 min, rather than to
blood donors now tested positive for hepatitis Hospital C, a traveling distance of only 30 min,
C. (Hepatitis C virus (HCV) was first identified but which only had facilities for uncross-
in 1988. An antibody test for HCV was soon matched blood transfusion. During transfer to
developed, but British Columbia did not intro- Hospital B, she lost consciousness in the ambu-
duce widespread testing until 1990. Neverthe- lance, and she was probably brain-dead by the
less, surrogate testing for non-A non-B hepatitis time she arrived there. Hospital B had limited
had been widely available in 1988.) She took no facilities for blood transfusion, but no obstetri-
legal action against her obstetrician, but sued cian in attendance. Here blood was transfused,
the CRC who supplied the blood transfused in and the uterine inversion was corrected using
1988, on the specific grounds that it had neither normal saline as in O’Sullivan’s method. She
tested for HCV nor carried out surrogate test- was then transferred to University Hospital in
ing, and thereby failed to prevent hepatitis C Saskatoon (Hospital D) which had full blood
contamination of its blood supplies. She also transfusion facilities and an obstetrician in
alleged that the CRC had deliberately destroyed attendance. But she was already dead by the
some of her medical records, thus disadvantag- time her ambulance arrived at Hospital D. The
ing her legal action, i.e. a separate tort known as court recognized the additional hazards of deliv-
‘spoliation’. Furthermore, together with many ery in a remote rural setting but, even so, it held
other patients infected with hepatitis C from that in a number of respects ‘the standard of com-
blood transfusions, she joined a class action, or petency, skill and diligence exercised by the delivering
a mass tort action, against the Canadian Red doctor fell below the standard expected of a general
Cross under British Columbia’s Class Proceed- practitioner practising in a rural setting’, and it
ings Act 1995. Hers proved to be a unique case awarded her estate damages of $343 00015.
380
402
30 August 2006 14:25:03
Litigation
In 2000, a Dr Gabaldoni appeared before bled to death. He took legal action against the
the Maryland State Board of Physician Quality medical center, on the grounds that it should
Assurance in connection with his management have carried blood, and it should have trans-
of a patient he had induced at term for pre- fused blood in a timely fashion. The local Ses-
eclampsia. The birth went well, but the mother sions Court found for the defendant hospital.
had a postpartum hemorrhage that was thought The case was appealed to the High Court,
to be due to retained fragments of placenta. She which reversed the decision of the Sessions
deteriorated over the next 48 h and her hemo- Court, and it found for the husband. However,
globin level went as low as 4.7 g/dl. Dr the hospital then went to the Court of Appeal,
Gabaldoni was said to be leisurely in atten- which affirmed the Sessions Court’s rejection
dance, and slow to transfuse blood. However, of expert medical evidence that blood must be
blood transfusion was started at 48 h post- stored before any delivery, as this ‘would result in
partum, but by this time she was in severe an absurd situation when one bears in mind that
respiratory distress, and her condition contin- deliveries are also conducted by midwives in houses
ued to deteriorate. She was admitted to the of the mothers where blood would not be stored before
intensive care unit at 72 h postpartum, but such deliveries’. The Court of Appeal thus
she died there 48 h later. Two days later, Dr reversed the High Court’s decision, as it held
Gabaldoni was said to have made a series of that there was no duty to hold blood for a
undated additions to her notes, which suggested low-risk patient in case she bled. Further, it held
that she had received better care than she did. that in this case the postpartum hemorrhage had
He was said to have made these additional been managed conventionally17.
entries in the same color ink as the original
progress notes, in such a manner that his alter-
OBSTETRICIAN ON VACATION (1961)
ations to the notes would not readily be appar-
ent. The Maryland Board of Physician Quality Obstetricians traditionally hand over the man-
Assurance filed charges under the Maryland agement of a complicated case to a colleague
Medical Practice Act 1995. When this case was when out of town or on vacation. The case may
considered by the Board, there was dispute then go wrong due to the colleague’s negligence,
about when he had seen the patient, when he but the vacationing obstetrician might find him-
had offered a blood transfusion, and whether self sued for negligence. In 1961, this happened
the medical notes as written were correct. After following death from postpartum hemorrhage.
reviewing the evidence, the Board found he had When pregnant with her fifth child, Patricia
‘failed to meet the appropriate standard for delivery Sturm told her obstetrician at 33 weeks that she
of medical care’, and so it issued a reprimand. He no longer felt fetal movements. He could not
appealed, but in a ‘deferential review’ the Court detect any fetal heart beat and, as obstetric ultra-
of Special Appeals of Maryland dismissed his sound had not yet been invented, he advised a
appeal16. conservative approach. He told her that she
In 2000, a Malaysian Court of Appeal con- would probably deliver normally in due course,
sidered whether a medical center had a duty to but he did discuss the possibility of fetal death.
keep blood available for transfusion. In 1988, As she was upset, he did not fully discuss all
Pearly Choo was booked to deliver her first baby the possible complications, but he did test her
in her local medical center, which carried no serum fibrinogen levels intermittently. He told
stored blood. She was healthy, had an uncom- her he would be on vacation at the time of her
plicated pregnancy, and she was considered to delivery, but would arrange for a colleague to
be at low risk. She delivered her baby unevent- look after her. However, she chose not to attend
fully, but she then sustained a major postpartum any further antenatal appointments. At 41
hemorrhage. In keeping with routine practice, weeks’ gestation, when her own obstetrician was
blood was requested from the nearby Kuala away on vacation, she began to bleed vaginally.
Lumpur General Hospital, and her husband She was admitted to hospital, and the colleague
was sent to collect it. By the time the husband delivered her of a stillborn infant. A massive
returned with the blood, his wife had already postpartum hemorrhage followed for which she
381
403
30 August 2006 14:25:03
had an eight-unit blood transfusion and a hys- At 20 days, heavy postpartum bleeding restarted.
terectomy. (The court report says it was carried She was then admitted to a different hospital,
out vaginally, but this may be incorrect.) Unfor- where a different gynecologist diagnosed an
tunately, she died despite the emergency treat- intrauterine infection. Despite a second D&C,
ment. The autopsy report attributed her death her heavy bleeding continued, and a hysterec-
to postpartum hemorrhage due to a clotting tomy had to be carried out. She sued the first
defect that was in turn due to intrauterine doctor and hospital for negligent care. She won
death. The family sued both the delivering doc- her case in the lower court, which held the doc-
tor and the vacationing doctor, on the grounds tor and the hospital jointly and severally liable
that he shared in liability for any perinatal negli- for damages of $100 000. However, the hospital
gence on the part of his deputy. The Supreme appealed the court’s decision on the grounds
Court of Oklahoma rejected this argument, and that the doctor was an independent contractor,
the obstetrician on vacation was exculpated18. and not the hospital’s servant or agent and that,
as the hospital was a governmental unit, it
was immune from tort liability. The Court
UNLICENSED PRACTICE OF
of Appeals upheld the hospital’s appeal, and it
OBSTETRICS (1963)
reversed the lower court’s decision as regards
Only two cases of postpartum hemorrhage have the liability of the hospital. Dr Sheikholeslam
been litigated where a professional attendant at did not appeal, and thus the original liability
delivery was not licensed to practise obstetrics. decision against him remained unchallenged20.
Earlier, the 1907 case of Midwife Porn was dis-
cussed. The only other reported case was in
INADEQUATE STAFFING LEVELS
1963. Bernhardt and Lund were two doctors of
(1981)
chiropractic, but they held themselves out as
competent in the management of childbirth. They In 1981, Stephen Martin was born in Ontario
supervised the delivery of Ladean Stojakovich at by spontaneous vaginal delivery following a
home, but unfortunately she had a postpartum labor complicated by fetal distress. He was in
hemorrhage and she died before she could be poor condition, and later he was diagnosed with
transferred to hospital. They were charged and cerebral palsy. When the case came to trial 17
convicted of breach of the Business and Profes- years later in 1998, Obstetrical Nurse James
sions Code (for practising medicine) and of man- was found guilty of negligence in failing to give
slaughter (for causing a death that was avoidable). appropriate care during labor. In her defence,
Surprisingly, and for complex legal reasons, the she said she was involved with another patient
Court of Appeals of California reversed both who was having a postpartum hemorrhage. This
convictions, and it denied a request for retrial19. was not accepted as a valid excuse as she should
have called for help. She and her hospital were
each found liable for 25% of the damages of
DISCHARGING PATIENT HOME TOO
$250 000 awarded to the claimant21.
SOON (1977)
In 1977, Patricia Hale (aged 20) delivered vagi-
NO AUTOPSY (1982)
nally at term at Fannin County Hospital in Texas,
under the care of Dr Sheikholeslam. Although In 1982, Yong Siew Yin was in labor at term with
she was still bleeding at 30 h after delivery, she her first baby. The labor was prolonged and (on
was discharged home. At 8 days postpartum, one account) she was in labor for over 24 h. She
she was readmitted with continued bleeding. had a small intrapartum hemorrhage. As there
She was given a preoperative injection (presum- was delay in the second stage and fetal distress,
ably of ergometrine) to contract her uterus, a urgent delivery was needed. The fetal head was
blood transfusion and a uterine curettage. After low in the pelvis, and in an occipitoposterior
her operation, she was given no injection and no position, so the baby was delivered ‘face-to
antibiotics. She was discharged home after 36 h, pubes’ by Neville Barnes forceps. Following
although she felt weak and she was still bleeding. this, she had a postpartum hemorrhage, and this
382
404
30 August 2006 14:25:03
Litigation
was attributed to vaginal tears. Whilst these The last dose was given on the morning she
were being repaired she collapsed, and a coagu- was delivered by elective Cesarean section. Her
lation disorder became manifest. She continued obstetrician-gynecologist had recently com-
to bleed heavily. An amniotic fluid embolism pleted his residency training but was not yet
was suspected, but it was never proved. She was board-certified. Moreover, he had not discussed
admitted to the intensive care unit where she her management with any board-certified obste-
died. Surprisingly, there was no autopsy. The trician-gynecologist, and had no other suitably
judge in the lower court found the obstetrician qualified surgeon in attendance. The Cesarean
guilty of negligence, and the hospital vicariously operation was carried out through a low trans-
liable. This verdict was upheld on appeal22. verse abdominal incision, but surgery proved to
be difficult. After the baby was delivered, the
uterus failed to contract, and she hemorrhaged
SUING THE WRONG DOCTOR (1982)
profusely. In these circumstances, it would
Occasionally, a patient may sue the wrong doc- have been usual to give Methergine (methyler-
tor. In 1976, Jean Johnson had a normal vaginal gonovine) and/or Pitocin (oxytocin) to promote
delivery at the Wishard Memorial Hospital in uterine contraction. No Methergine was given;
Indiana. This was followed 2 weeks later by half a dose of Pitocin may have been given, but
a secondary postpartum hemorrhage. She was it was not documented in the medical notes or
seen by the Chief Resident, Dr Deaton, who on the drug chart. A hysterectomy was carried
diagnosed retained products of conception, and out, but the bleeding continued. Her bladder
advised uterine curettage. He checked his diag- was damaged and she developed hematuria.
nosis and treatment plan with Dr Padilla, a staff A urological surgeon was then called, and he
instructor with the Indiana University Medical ligated the left internal iliac (or hypogastric)
School, and the operation was carried out. By artery. This slowed the bleeding considerably,
1982, it had become apparent that Jean Johnson but it did not stop it completely. The tissues
was infertile, and this was attributed to over- were now friable and so the abdomen was
vigorous curettage of the endometrium in 1976 packed and closed, and she was managed
(Asherman’s syndrome). She sued Dr Padilla overnight in intensive care. The abdomen was
for negligent performance of the curettage, but reopened the following day as internal bleeding
did not suggest that the curettage decision itself continued. At the second operation, all bleeding
was negligent. The defence was threefold: (1) was brought under control, but she lost her
Dr Padilla did not carry out the curettage; right ovary. During this episode, she was given a
(2) there was no doctor–patient relationship total of 34 units of blood, 14 of fresh frozen
between Dr Padilla and Jean Johnson; and (3) plasma and 10 of platelets, but she survived.
there was no agency relationship between Dr Postoperatively, she developed a vesico-vaginal
Padilla and Dr Deaton. The Court of Appeals fistula, hepatitis, an extremely short vagina that
of Indiana accepted all three lines of defence, made intercourse impossible, and severe psy-
and dismissed the case against Dr Padilla23. chological problems. As she was managed and
delivered at a naval military hospital in Illinois,
she took legal action against the United States
OBSTETRICIAN WITHOUT
of America. After hearing expert evidence,
SUFFICIENT EXPERIENCE (1986)
the trial judge was critical of: an obstetrician-
In 1986, Christine Steinhagen became pregnant gynecologist who was not board-certified man-
for the third time. She had two previous aging this complicated case without more
Cesarean sections, the second being compli- experienced help; his giving terbutaline immedi-
cated by ‘extreme and profuse bleeding’. In her ately prior to the Cesarean section, thereby
third pregnancy, she had a sudden vaginal bleed inhibiting uterine contraction after delivery;
at about 20 weeks’ gestation, and an anterior his failure to perform the operation through a
placenta previa was diagnosed. She was kept in midline incision which would have minimized
hospital for 18 weeks and throughout this time the risk of bladder damage; his failure to give
given terbutaline to inhibit uterine contractions. Methergine to contract the uterus; and his
383
405
30 August 2006 14:25:03
failure to ligate both hypogastric arteries which ordered a new trial limited to damages, as it
might have avoided the hysterectomy and the considered the jury award excessive26.
loss of an ovary. He awarded her $300 000 in In 1995, Natalie Lomeo was delivered by
compensation24. elective Cesarean section at her local Commu-
nity Medical Center (CMC) in Pennsylvania.
She had an extensive blood loss during the oper-
NO OPERATION NOTE (1992)
ation, and a postpartum hemorrhage followed.
In 1992, Mrs Suchorab was delivered in Sas- Although she exhibited signs of hemorrhagic
katchewan by Cesarean section. Six weeks later, shock, blood was not transfused until much
she had a postpartum hemorrhage and was later in the day. Over the next 3 years, she
readmitted to hospital. Her obstetrician took complained of fatigue, weakness, dizziness, hair
her to the operating theater, where he stabilized loss, amenorrhea, dyspareunia, and vasomotor
her condition. The operation log and the anes- symptomatology. In 1998, the diagnosis of
thetist’s note both record that a dilatation and Sheehan’s syndrome was made. She then took
curettage operation was carried out, but no sur- legal action against her obstetrician and the
gical operation note was ever found to confirm CMC. However, the defendants filed for sum-
this. The following day, she had a further major mary judgment, asserting that her claim was
hemorrhage, and a hysterectomy was carried time-barred under Pennsylvania law, as it had
out. She took legal action against her obstetri- been filed more than 2 years after the allegedly
cian. She argued that his care had been deficient negligent conduct. The Common Pleas Court
as her bleed was due to retained products of denied the motion for dismissal, saying that the
conception, and he had failed to curette her litigation clock only started to run when
uterus as (she claimed) was evidenced by the Sheehan’s syndrome was diagnosed27. What
absence of any operation note. He claimed that happened next was not reported, so the case was
he had curetted her uterus, but he had forgotten probably settled.
to write an operation note. Moreover, he
claimed that her bleed was from a ‘necrotic
MALIGNANT HYPERTENSION (1993)
cervix’, and not from the uterine cavity, and so
no extra harm would have resulted from failure In 1993, Evelyn Dybongco-Rimando had an
to curette the uterus. The court rejected her uneventful spontaneous vaginal delivery of a
claim25. healthy daughter, and she went home shortly
afterwards. Some 8 years later, a judge of the
Superior Court of Justice of Ontario was to say
SHEEHAN’S SYNDROME (1977, 1995)
that her case ‘presents a puzzle with a thousand
In 1977, Mrs Parker delivered her first child. pieces’. The trial started in 1999, and it lasted
Her obstetrician delivered the placenta by for 33 days spread over 3 years. The judge
continuous cord traction. However, she had a described it as ‘a challenge to bench and bar alike’.
uterine inversion and a major postpartum hem- Although her delivery was normal, 7 days later
orrhage followed. She was taken to the operat- she suffered a massive postpartum hemorrhage,
ing theater, and in the operation note it was and she was readmitted to hospital. Over the
recorded that her ‘uterus had resolved itself’. Five next 2 days, she had three operations before
months later, she was found to have ‘an inverted her bleeding could be brought under control:
uterus presenting well down in the vagina’. She had uterine exploration, hysterectomy, and then a
various ongoing symptoms, but it was not until second-look laparotomy. She was given a large
1991 (14 years later) that Sheehan’s syndrome transfusion of blood, and also blood products
was diagnosed. She then took legal action as she developed a coagulation disorder. She
against her obstetrician of 1977. A four-person became profoundly hypotensive, and required
jury awarded her $960 000 in damages. Her inotropic agents (principally dopamine) to sup-
obstetrician appealed the case on both liability port her blood pressure. However, her blood
and quantum. The New South Wales Court of pressure then went too high, and within 33 h
Appeal dismissed his appeal on liability, but it of readmission to hospital she had developed
384
406
30 August 2006 14:25:04
Litigation
malignant hypertension. Dopamine was given and St Joseph’s Health Centre, all of whom
but discontinued when her pressure reached were in Ontario. The Italian court refused to
237/113 mmHg. However, the maximum level hear the case, saying it lacked jurisdiction as
of blood pressure later recorded was 256/ the medical treatment had occurred in Ontario.
126 mmHg. She then had a cerebral hemor- In 2000, she brought a similar legal action in
rhage, and soon after this she died. Her estate Ontario. However, the defendants prevailed, as
started a legal action against 55 defendants, but Ontario law requires an action against a doctor
only three defendants remained shortly after the to be brought within 1 year from when the
trial started in 2000. These were her obstetri- Plaintiff ‘knew or ought to have known’ the mate-
cian, her internal medicine physician, and her rial facts on which the malpractice is alleged,
intensivist. In his final judgment, the judge said and against a hospital or nurse within 2 years of
of the internal medicine physician’s testimony the patient being discharged from hospital or
‘It reflects a triumph of tactics over truth. He is not stopping treatment. Furthermore, the Ontario
credible.’ He found all three defendant doctors Court of Justice also found that in this case
guilty of negligence, and he reserved judgment there was no genuine issue for trial as no expert
on the amount of damages to be awarded to the reports were filed30.
deceased patient’s estate28.
POSTPARTUM HEMORRHAGE INTO
NO EXPERT MEDICAL REPORT (1995) THE PLEURAL CAVITY (1997)
In 1995, Marcia Laidley had a postpartum hem- In 1997, an unusual case of postpartum hemor-
orrhage after delivering her third child. A supra- rhage occurred in California. Martha Guandique
cervical hysterectomy was performed. Later, had severe pre-eclampsia at 38 weeks’ gestation.
she took legal action against her obstetrician. Her signs and symptoms included shortness
However, she failed to provide a timely expert of breath, hypertension, renal malfunction,
medical report in support of her case by the hepatomegaly and pleural effusion. Labor was
court-imposed deadline, and so summary judg- induced and she delivered a male infant. She
ment was awarded against her. She appealed. had a postpartum hemorrhage due to uterine
The Court of Appeals of Ohio held that the trial atony, so she was given Pitocin. Blood clots
court had committed a prejudicial error when it were evacuated from her uterus. Shortly after
granted the defendant’s motion for summary delivery, she had considerable difficulty in
judgment without providing the opportunity for breathing, and back pain. Various physicians
sufficient discovery on the issue29. were called in to see her. Pulmonary embolism
and amniotic fluid embolism were in the differ-
ential diagnosis. Supportive therapy with oxy-
POSTPARTUM HEMORRHAGE IN AN
gen was given and various drugs were used. Her
AIRCRAFT (1997)
hemoglobin fell at first to 9.5 g/dl, and it contin-
In 1997, Gina Paone delivered her baby in ued to fall thereafter. (Subsequent hemoglobin
Ontario, but her placenta had to be removed levels were not recorded in the court report.) A
manually. Her uterine cavity was explored and blood transfusion was started, but 20 min later
considered to be empty. The placenta was she had a cardiopulmonary arrest and then she
judged to be complete. One month later, she died. At autopsy, she was found to have suffered
flew to Italy, but she had abdominal pain and a major postpartum hemorrhage (of 1500 ml)
heavy vaginal bleeding during the flight. On into her right pleural cavity. The pathologist
arrival in Italy, she was admitted to hospital reported that ‘The mechanism of production of this
where she had a uterine curettage. She claims hemorrhage remains unknown in spite of a careful
she was told there was further placental tissue dissection of the blood vessels in the area. . . . That is
recovered from the uterus, but there was no why the mode of this death remains undetermined.’
written confirmation of this. In 1998, she In this case, much of the complicated legal argu-
started legal proceedings in Italy by an Act of ment before the Court of Appeal of California
Citation naming her obstetrician, two nurses focused on which doctors might have been
385
407
30 August 2006 14:25:04
liable for her death, but these legal arguments legally insufficient to support the jury’s verdict,
need not concern us here31. and the State had failed to prove the materiality
of her alleged false statement. The Court of
Appeals of Texas considered her arguments but
DISAPPEARING BABY (1999)
it dismissed her appeal32.
This too represents an unusual case, but I [In the late 1970s, I had a similar case in the
have seen something very similar (see below). In UK: a 14-year-old girl who presented in shock
1999, an unmarried mother was having an adul- with heavy vaginal bleeding. She had a perineal
terous affair with a co-worker. He noticed that midline tear, a widely open cervix, and an
her abdomen was enlarging, and asked whether enlarged uterus, but there was no baby and no
she might be pregnant. She said that she could placenta. Her hemoglobin level was only 4 g/dl,
be. The matter was discussed no further, nei- so she was transfused with blood. Her presenta-
ther with him nor with any other co-workers. A tion was clearly consistent with recent childbirth
few weeks later, she attended her family doctor followed by a major postpartum hemorrhage.
complaining of swollen feet. She told him that Despite the overwhelming evidence, the girl
she was 7 months pregnant. The doctor heard and her parents firmly denied any pregnancy
the fetal heart beat and felt fetal movements, or recent delivery of a baby. The police were
and so he pronounced the fetus healthy. This duly called in. They investigated the matter and
was the only medical care she sought before 12 searched the family home, but no baby was ever
May 1999, when she was admitted to a Texas found. No charges were ever brought.]
hospital with a 2-day history of vaginal bleeding.
She was said to be in shock: she was weak and
ABANDONMENT (2000)
pale, had a low temperature, and a tachycardia.
(Her blood pressure was not mentioned in the In 2000, the New York Bureau of Professional
court report.) She said that she was pregnant, Medical Conduct considered the case of Dr
but she did not know the date of her last men- Wahba, an obstetrician who was charged with
strual period, nor when her baby was due. A professional misconduct in the treatment of seven
blood test showed that she was severely anemic. of his patients. Two of these were at risk of
Her hemoglobin level was not mentioned in the postpartum hemorrhage, and here he was found
court report, but, from comments in the report, guilty of negligence and/or incompetence. In
it was probably around 4–5 g/dl. Four units of both cases, he left the delivery room before the
blood were transfused. An obstetrician was placenta was delivered. The first patient had a
called, and she scanned the uterus with ultra- stillbirth, and so she was at a higher risk of
sound. She found no evidence of a baby, but she postpartum hemorrhage. The second was still
did find a placenta of a size compatible with a hemodynamically unstable; she then hemor-
term baby. The placenta was then delivered, but rhaged but by this time the obstetrician had
it had no cord attached. Both the patient and already left the hospital. Moreover, he refused
her attendant family denied that any baby had the nurse supervisor’s requests to return. After
been born. Therefore the police were called. reviewing his management of all seven patients,
They searched her home, and there they found the Administrative Review Board for Profes-
evidence of extensive blood staining of her bed, sional Medical Conduct revoked his licence to
and of her bathroom – but no baby. A grand practise medicine in the state of New York. He
jury was convened to determine whether any then appealed to the Supreme Court of New
charge, such as homicide, should be brought. York, but his appeal was dismissed33.
Under oath she said that ‘I did not pass a baby’,
and she insisted that she had only passed clots of
POSTPARTUM HEMORRHAGE IN A
blood. She was later charged with aggravated
FEMALE DOG (2006)
perjury before a grand jury, convicted by a jury,
and sentenced to 10 years confinement pro- American courts are well known for leading the
bated for 10 years. She appealed against her way into new areas of litigation. Therefore it
conviction on the grounds that the evidence was may come as no surprise to learn that in
386
408
30 August 2006 14:25:04
Litigation
February 2006 the Court of Appeals of Texas after postpartum hemorrhage (only two cases),
ruled on a case involving the management of and having professional birth attendants
postpartum hemorrhage in a female dog in the who were not licensed to practise obstetrics
Bureau of Animal Regulation and Care in (only two cases, one of which was litigated in
Houston in 1999. This facility takes around 1907).
20–30 000 animals a year. One of their veteri- Apart from the general observation that poor
narians was Dr Levingston. He had made a obstetric practice was a typical feature of many
number of complaints to his employers about of these cases, they were otherwise sporadic in
the inhumane treatment of animals in their care, etiology, with no common cause.
but on one particular occasion they accused him Given the millions of women who have
of the negligent care of animals, and they termi- delivered over the last 100 years across the
nated his employment. They cited his alleged English, Commonwealth, Irish, and American
mismanagement of the care of a female Rott- jurisdictions studied, given that the incidence of
weiler dog who had given birth to nine puppies, postpartum hemorrhage is around 5–10%, and
and who had a postpartum hemorrhage from given that there has been an international
which she exsanguinated and died. They said he increase in litigation for alleged clinical mal-
should have considered the possibilities of hys- practice, it is surprising that there have not been
terectomy or euthanasia. He appealed his termi- many more cases of postpartum hemorrhage
nation of employment and won his case. He was litigated in the courts.
awarded damages in the lower court. His
employers appealed the decision, and the case
went to the Court of Appeals of Texas who References
dismissed their appeal. The court awarded him
1. Westrup v Commonwealth. Court of Appeals of
a total of $1.24 million for past and future
Kentucky. 123 Ky 95; 93 SW 646; 1906 Ky;
lost wages and compensatory damages. This LEXIS 123
amount was to include $194 000 for his law- 2. Commonwealth v Hanna Porn. Supreme Judicial
yers’ fees. If the lawyers’ fees of his employers, Court of Massachusetts, Worcester. 196 Mass
the City of Houston, were of the same order of 326; 82 NE 31; 1907 Mass; LEXIS 1096
magnitude, then the legal bill on this case would 3. US Short, Administrator of the Estate of Mollie
have been around $400 000. Overall, this case Short, Deceased v City of East St Louis. Court of
ran for more than 5 years34. Appeals of Illinois. 4d 140 Ill App 173; 1908 Ill
App; LEXIS 819
4. Southern Bell Telephone & Telegraph Co v Glawson
CONCLUSIONS et al. Court of Appeals of Georgia. 13 Ga App
520; 79 SE 488; 1913 Ga App; LEXIS 247
This account has been international in its scope,
5. Peterson v Langsten. 28,835; Supreme Court of
albeit confined to common law jurisdictions. It Minnesota. 186 Minn 101; 242 NW 549; 1932
is clear that the history of litigation following Minn; LEXIS 844
postpartum hemorrhage stretches for over 100 6. Goff et al. v Doctors General Hospital of San Jose et
years, from Florence Westrup of Newport, al. 9408; Court of Appeal of California, Third
Kentucky in 1905 to the female Rottweiler dog Appellate District. 166 Cal App 2d 314; 333 P2d
of Houston, Texas in 2006. 29; 1958 Cal App; LEXIS 1404
In 17 of 34 cases (50%), a maternal death no 7. Reserve Life Insurance Company v Whitten. Court
doubt prompted the litigation, rather than the of Appeals of Alabama. 38 Ala App 455; 88 So
postpartum hemorrhage itself. 2d 573; 1956 Ala App; LEXIS 208
8. Madison et al. v City and County of San Francisco
After maternal death, the second most
et al. 14410; Court of Appeal of California, First
common reason for litigation was a problem
Appellate District, Division One. 106 Cal App
with the transfusion of blood, such as infection, 2d 232; 234 P 2d 995; 1951 Cal App; LEXIS
delay or possible incompatibility. Such prob- 1738
lems occurred in ten of 34 (29%) of the cases. 9. Gillen v United States of America. 16584; United
Equal third reasons for litigation were States Court of Appeals Ninth Circuit. 281 F2d
having a diagnosis made of Sheehan’s syndrome 425; 1960 US App; LEXIS 4034
387
409
30 August 2006 14:25:05
10. Parker and Parker v St Paul Fire & Marine 85776; 1998 ACWSJ 523416; 81 ACWS (3d)
Insurance Company et al. Court of Appeal of 548
Louisiana, Second Circuit. 335 So 2d 725; 1976 22. Ping and Anor v Woon Lin Sing et al. Rayuan Sivil
La App; LEXIS 3976 No 12-223-92 & 12-225-92. High Court of
11. Belle Bonfils Memorial Blood Center v Denver Dis- Shah Alam, Malaysia. 1998 MLJU; LEXIS
trict Court, Judge Phillips, CW, KW and son RW. 1203; [1998] 583 MLJU 1
88-SA-45; Supreme Court of Colorado. 763 P2d 23. Johnson v Padilla. 2-1280-A-410; Court of
1003; 1988 Colo; LEXIS 174; 12 BTR 1463 Appeals of Indiana, Second District. 433 NE2d
12. Estate of Mutsuko Gaffney; and Gaffney et al. v 393; 1982 Ind App; LEXIS 1122
United States of America. 88-1457-Z; United 24. Steinhagen v United States of America. 89-CV-
States District Court for the District of 72453-DT; US District Court for Eastern
Massachusetts. 1990 US Dist; LEXIS 5184 District of Michigan, Southern Division. 768 F
13. Traxler v Varady. A053098; Court of Appeal Supp 200; 1991 US Dist; LEXIS 8918
of California, First Appellate District, Division 25. Suchorab v Urbanski. Saskatchewan Queen’s
One. 12 Cal App 4th 1321; 16 Cal Rptr 2d 297; Bench. 1997 Sask D; LEXIS 744; [1997] Sask
1993 Cal App; LEXIS 82; 93 Cal Daily Op D. 610.30.50.70–02
Service 747; 93 Daily Journal DAR 1423 26. Fowkes v Parker [1999]. NSWCA 442; Supreme
14. Endean v Canadian Red Cross Society. British Court of New South Wales, Court of Appeal. CA
Columbia Supreme Court. 148 DLR (4th) 158; 40948/98; 1999 NSW; LEXIS 862; BC9908184
1997 DLR; LEXIS 1359 27. Lomeo v Davis. 99-CV-2639; Common Pleas
15. Naeth Estate v Warburton. Saskatchewan Queen’s Court of Lackawanna County, Pennsylvania. 53
Bench. 1992 ACWSJ; LEXIS 33936; 1992 Pa D & C 4th 49; 2001 Pa D & C; LEXIS 95
ACWSJ 569976; 34 ACWS (3d) 1108 28. Dybongco-Rimando Estate et al. v Jackiewicz et al.
16. Gabaldoni v Board of Physician Quality Assurance. Court of Ontario: Superior Court of Justice.
Court of Special Appeals of Maryland. 141 Md 2001 OTC; LEXIS 2442; [2001] OTC 682
App 259; 785 A2d 771; 2001 Md App; LEXIS 29. Laidley v St Luke’s Medical Center et al. 73553;
180. ‘Under Maryland law, the final order of an Court of Appeals of Ohio, Eighth Appellate
administrative agency is subject to deferential review District, Cuyahoga County. 1999 Ohio App;
by the courts. Deferential review prohibits a court LEXIS 2567
from substituting its judgment for that of the agency if 30. Paone v St Joseph’s Health Centre. 00-CV-
substantial evidence exists to support the agency’s 198822CM; Ontario Superior Court of Justice.
decision. The test is ‘reasonableness not rightness’. 2002 ACWSJ; LEXIS 7091; 2002 ACWSJ
17. Arayan et al. v Simon et al. Court of Appeal 10094; 118 ACWS (3d) 46
(Kuala Lumpur); Decided 18 April 2000. [2000] 31. Guandique et al. v Makabali et al. B157844;
3 MLJ 657; Civil Appeal No W-04–71 of 1996 Court Of Appeal Of California, Second
18. Sturm v Green. 40638; Supreme Court of Appellate District, Division Seven. 2004 Cal
Oklahoma. 1965 OK 12; 398 P 2d 799; 1965 App Unpub; LEXIS 6458
Okla; LEXIS 364 32. Steen v State of Texas. 14-00-00429-CR;
19. The People v Bernhardt and Lund. Court of Court of Appeals of Texas, Fourteenth District,
Appeal of California, Second Appellate District, Houston. 78 SW 3d 516; 2002 Tex App; LEXIS
Division Three. 222 Cal App 2d 567; 35 Cal 2306
Rptr 401; 1963 Cal App; LEXIS 1701 33. Wahba v New York State Department of Health et
20. Hale v Sheikholeslam and Fannin County Hospital. al. 86017; Supreme Court of New York, Appel-
83-2047; United States Court of Appeals for late Division, Third Department. 277 AD 2d
the Fifth Circuit. 724 F2d 1205; 1984 US 634; 716 NYS 2d 443; 2000 N Y App Div;
App; LEXIS 25485; 1984 Fed Carr Cas (CCH) LEXIS 12048
P83,141 34. City of Houston v Levingston. 01-03-00678-CV;
21. Martin v Listowel Memorial Hospital. Ontario Court of Appeals of Texas, First District,
Court (General Division). 1998 ACWSJ; LEXIS Houston. 2006 Tex App; LEXIS 859
388
410
30 August 2006 14:25:05
Section IX
Special experiences and unusual
circumstances
411
30 August 2006 14:25:05
42
THE OBSTETRICIAN CONFRONTS POSTPARTUM
HEMORRHAGE
M. E. Setchell
INTRODUCTION HISTORICAL PERSPECTIVE

Postpartum hemorrhage has been recognized as In the middle of the 19th century, maternal
a major cause of maternal death for as long as mortality was around 6 per 1000 live births,
physicians have studied and written about child- and, of those deaths, about one-third were
birth. Until the 20th century, however, little was related to puerperal sepsis, and the remainder
possible in the way of effective treatment, and, were classified as ‘accidents of childbirth’,
as is apparent in many of the chapters of this which included ante- and postpartum hemor-
book, postpartum hemorrhage is still a frequent rhage and deaths from obstructed labor.
cause of death in many parts of the world. Even Table 1 shows birth and death rates in England
in the Western world, significant numbers and Wales from 1847 until 1901. It is evident
of deaths and morbidity from postpartum that there was no real improvement in deaths
hemorrhage continue to plague obstetricians, from sepsis during this period, in contrast to a
despite considerable advances in medical care in relative improvement in the deaths from other
the last half-century. causes.
During the author’s career in Obstetrics The concept of Lying-In Hospitals was first
which has spanned almost 40 years, one of the adopted in the mid-18th century, and by 1904
most striking changes has been the one whereby there were 38 such hospitals in Great Britain.
the individual obstetrician no longer has to deal The stated intention was to provide a safer place
with the problem of postpartum hemorrhage for delivery and postnatal care, but any pur-
alone, but can call on a sophisticated team of ported benefits in better obstetric care were far
helpers, involving a whole range of other spe- outweighed by the risks of death from sepsis,
cialists. A mere glance at the contents of this which, as can be seen in Table 2, amounted to
book confirms that the modern management of 3% in the period of 1838–1860. This appalling
a major postpartum hemorrhage can involve a figure improved considerably during the latter
team of anesthetists, hematologists, vascular part of the 19th century, however, following the
surgeons, gynecologists and radiologists. introduction of Semmelweis’ observations and
Clearly, this change represents an advance teachings on hygiene and antisepsis in 1861.
which has saved and will continue to save Francis Ramsbotham, the first Lecturer and
countless lives, not only in the developed world Obstetric Physician to The London Hospital,
where such teamwork is routine, but also in published ‘The Principles and Practice of
developing nations that are desperately looking Obstetric Medicine and Surgery in reference to
for means to reduce maternal mortality as part the Process of Parturition’ in 1841, and provided
of their efforts to comply with the United some poignant case reports, revealing what the
Nations Millennium Development Goals by the practice of Obstetrics was like at that time. The
year 2015. case of a rich patient in the City of London,
390
412
30 August 2006 14:25:05
The obstetrician confronts postpartum hemorrhage
Table 1 Mortality in childbirth in England and Wales 1847–1901 (a period of 55 years), in General
Lying-in Hospital, London
Deaths Death rate to 1000 children born alive, from
Registered births Puerperal septic Puerperal Accidents Puerperal septic Puerperal Accidents
of children diseases and septic of diseases and septic of
Year born alive accidents of childbirth diseases childbirth accidents of childbirth diseases childbirth
1847 539 965 3226 784 2442 5.97 1.45 4.52

1848 563 059 3445 1365 2080 6.12 2.42 3.70
1849 578 159 3339 1165 2174 5.78 2.02 3.76
1850 593 422 3252 1113 2139 5.48 1.88 3.60
1851 615 865 3290 1009 2281 5.34 1.64 3.70
1852 624 012 3247 972 2275 5.20 1.56 3.64
1853 612 391 3060 792 2268 5.00 1.30 3.70
1854 634 405 3009 954 2055 4.74 1.50 3.24
1855 635 043 2979 1079 1900 4.69 1.70 2.99
1856 657 453 2888 1067 1821 4.39 1.62 2.77
1857 663 071 2787 836 1951 4.20 1.26 2.94
1858 655 481 3131 1068 2063 4.78 1.63 3.15
1859 689 881 3496 1238 2258 5.07 1.79 3.28
1860 684 048 3173 987 2186 4.64 1.44 3.20
1861 696 406 2995 886 2109 4.30 1.27 3.03
1862 712 684 3077 940 2237 4.32 1.32 3.00
1863 727 417 3588 1155 2433 4.93 1.59 3.34
1864 740 275 4016 1484 2532 5.43 2.00 3.43
1865 748 069 3823 1333 2490 5.11 1.78 3.33
1866 753 870 3682 1197 2485 4.88 1.59 3.29
1867 768 349 3412 1066 2346 4.44 1.39 3.05
1868 786 858 3503 1196 2307 4.45 1.52 2.91
1869 773 381 3283 1181 2102 4.24 1.53 2.71
1870 792 787 3875 1492 2383 4.89 1.88 3.01
1871 797 428 3935 1464 2471 4.98 1.81 3.09
1872 825 907 3803 1400 2403 4.60 1.70 2.90
1873 829 778 4115 1740 2375 4.96 2.10 2.86
1874 854 956 5927 3108 2819 6.93 3.63 3.30
1875 850 607 5064 2504 2560 5.95 2.94 3.01
1876 887 968 4142 1746 2396 4.66 1.97 2.69
1877 888 200 3443 1444 1999 3.88 1.63 2.25
1878 891 906 3300 1415 1885 3.70 1.59 2.11
1879 880 359 3340 1464 1876 3.79 1.66 2.13
1880 881 643 3492 1659 1833 3.94 1.88 2.08
1881 883 642 4227 2287 1940 4.78 2.58 2.20
1882 889 014 4524 2564 1960 5.09 2.89 2.20
1883 890 722 4508 2616 1892 5.06 2.94 2.12
1884 906 750 4647 2468 1879 4.79 2.72 2.07
1885 874 970 4449 2420 2029 4.98 2.71 2.27
1886 903 866 3877 2078 1799 4.72 2.39 1.99
1887 886 331 4160 2450 1710 4.69 2.80 1.90
1888 879 868 4160 2386 1774 4.73 2.49 2.01
1889 885 944 3585 1852 1733 4.05 2.09 1.95
1890 869 937 4255 1956 2299 4.89 2.24 2.62
1891 914 157 4787 1973 2814 5.24 2.15 3.06
continued
391
413
30 August 2006 14:25:05
Table 1 Continued
Deaths Death rate to 1000 children born alive, from
Registered births Puerperal septic Puerperal Accidents Puerperal septic Puerperal Accidents
of children diseases and septic of diseases and septic of
Year born alive accidents of childbirth diseases childbirth accidents of childbirth diseases childbirth
1892 897 957 5194 2356 2838 5.78 2.62 3.16

1893 914 542 5950 3023 2927 6.51 3.30 3.19
1894 890 289 4775 2167 2608 5.36 2.43 2.92
1895 922 291 4219 1849 2370 4.57 2.00 2.56
1896 915 309 4561 2053 2508 4.98 2.24 2.74
1897 921 693 4250 1836 2414 4.61 1.99 2.62
1898 923 265 4074 1707 2367 4.41 1.84 2.56
1899 928 646 4326 1908 2418 4.66 2.05 2.63
1900 927 062 4454 1941 2514 4.81 2.09 2.71
1901 927 807 4394 2079 2315 4.73 2.24 2.49
Table 2 Number of deliveries, deaths and death already produced such a degree of compression
rates during different time periods in the General as I have rarely witnessed, with its concomitant
Lying-in Hospital, London symptoms. Upon a vaginal examination a little
after six, I detected the Placenta to be placed
Time Average death rate
immediately over the Os Uteri; some discharge
period Deliveries Deaths from all causes
was still oozing away, but there was no tendency
1838–1860 5833 180 1 in 32.5 or 30.85 to pain. The urgency of the haemorrhage
per 1000 appeared therefore to be at present somewhat
1861–1879 3773 64 1 in 57.875 or 16.96 abating; and the lady for a short time seemed
per 1000 disposed to revive; but presently the flooding
1880–1887 2585 16 1 in 161.5 or 6.18 returned with its original violence. Anxiously
per 1000 watching its progress for a short time, and
1888–1892 2364 9 1 in 262.67 or 3.80
observing no diminution in the discharge, I
per 1000
determined on delivery; but previously I
requested my professional friend to satisfy
himself that the Placenta was presenting. Being
described below, illustrates how little could really answered in the affirmative, I proceeded with-
be done for intra- and postpartum hemorrhage. out further loss of time to empty the Uterus.
The Os Uteri was but little opened, yet it was
relaxed, and permitted the passage of my hand
‘Case C1V’
with ease into the Uterus; but that organ
‘I was summoned to a private patient near the showed at the moment no disposition to active
Mansion House, who had been, a few minutes contraction; having brought down the breech,
before, attacked with a sudden flooding in the the child was found to be alive; I therefore pro-
eighth month of pregnancy, while sitting with ceeded gently in its extraction; and after the
her family at tea, in the drawing-room. Upon child was born, the Placenta was thrown off,
proceeding up stairs, tracks of blood were and was soon withdrawn. The uterine tumour
perceptible upon every step. In the bedroom, I proved now to be irregularly contracted, and
found a neighbouring professional gentleman, fell flaccid under the hand. For a short time,
who had been also called by the servants in this lady appeared comfortable; the discharge
their alarm at the state of their mistress; and, ceased, and she expressed her warmest thanks
although this unfortunate occurrence had not for my prompt assistance; but by-and-by she
happened a quarter of an hour before, it had began to complain of her breath: ‘Oh! my
392
414
30 August 2006 14:25:06
breath! my breath!’ was her urgent exclamation. laparotomy and hemostatic suturing, ligation of
My patient continued to sink, and expired soon vessels or embolization.
after seven o’clock; so that in less than two The author’s first ‘lone’ experience of post-
hours, from an apparent state of perfect health, partum hemorrhage occurred whilst working as
her valuable life was sacrificed to a sudden a new Registrar at the University Hospital of the
attack of haemorrhage, in spite of the most West Indies in Jamaica. Having just successfully
prompt assistance. The child was lively, and conducted a very straightforward twin delivery,
promised to do well.’ including completion of the third stage of labor
with a standard dose of syntometrine, my state
of calm was interrupted by a sudden gush of
THE LONELINESS OF THE blood of such proportion that it seemed then
OBSTETRICIAN (and even now) as if an old-fashioned bath tap
had been turned on full pelt. The sound and
Fifty years ago, and for the ensuing 20 years at sight of that hemorrhage will never leave my
least, ‘Practical Obstetric Problems’ by the late memory; it was a moment of absolute panic
Professor Ian Donald, Professor of Midwifery in and helplessness. Miraculously, something took
the University of Glasgow, was the essential and over, and decisions and actions were taken as if
valued textbook for all young obstetricians of they were automatic, probably because Profes-
that generation. Nowhere is the famous dedica- sor Ian Donald had been read, and re-read, in
tion in the frontispiece more relevant than in preparation for such an event. Bimanual com-
relation to postpartum hemorrhage: pression, intravenous ergometrine administered
‘To all those who have known doubt, perplexity by a much more experienced midwifery sister,
and fear as I have known them, who then made up a bottle of intravenous
To all who have made mistakes as I have, Syntocinon almost without being asked, and the
To all whose humility increases with their situation was quickly under control. The young
knowledge of this most fascinating subject, obstetrician grew significantly in maturity and
This book is dedicated.’
experience in those few minutes, grateful that
The sense of helplessness, loneliness and fear simple actions had averted what had seemed a
that Dr Ramsbotham must have felt as he potential disaster.
watched his patient expire in spite of all his good During the remaining years of my training,
work and intentions is something that none of other dramatic postpartum hemorrhages also
us ever wish to experience in our career. occurred, but the range of available interven-
As modern obstetricians, we no longer per- tions was limited. Intravenous or intramuscular
form our tasks in isolation; we practice in hospi- ergometrine, intravenous Syntocinon infusions,
tals which, in the majority of instances, are well bimanual compression, or packing the uterus
or relatively well equipped, are surrounded by with enormous packs (one teacher described
midwives, junior or senior colleagues, and know putting a pillow case into the uterus first, and
that various other specialists are standing by then filling it with as many packs as one could
in support. Nevertheless, in dealing with post- get hold of) were the only effective treatments.
partum hemorrhage, there comes a moment One had occasionally seen the need for post-
when our decisions and actions (or lack thereof) partum hysterectomy and internal iliac artery
are going to determine the sequence of events. ligation, but, in those circumstances, there had
Even in complex cases of more prolonged always been the welcome presence of a more
hemorrhage, when all the support of the senior colleague.
laboratory hematologists, the blood transfusion It is not only the trainee obstetrician who
service, the anesthetic intensivist and other sup- may still be faced with hard decisions. Some-
porting clinicians has been called in, there will times, the presence and involvement of a large
come a time when the only the attending team lead to confusion of leadership. Whilst
obstetrician, using his or her best and most protocols, guidelines and practice ‘drills’ may
considered judgements, has to make a decision help to coordinate teamwork and familiarize
about radical treatments such as hysterectomy, staff in how to deal with these unusual
393
415
30 August 2006 14:25:06
situations, there remain numerous times when uterus closed without difficulty. A drain was left
the obstetrician has to take command and make in the abdomen. An hour later, it was evident
rapid or difficult decisions. In a lengthy career, that there was major intra-abdominal hemorr-
one may be faced with a situation that is hage. The drainage bottle had filled and been
unique and has not been met with before. A emptied twice, and the abdomen was distended,
few such cases which have faced the author are tense and tender. Unfortunately, the general
now discussed. surgeon had departed for the weekend and
A patient had been admitted at 34 weeks with was not contactable. When the obstetrician
severe abdominal pain, a tense abdomen and returned, the patient was in a desperate condi-
absent fetal heart tones. Signs of shock and the tion, with major cardiovascular collapse. The
tense, tender abdomen suggested a placental anesthetist had inserted a subclavian line in
abruption, and the cardiovascular and respira- order to obtain good venous access, and in
tory collapse was of such severity that she was doing so had inadvertently caused a pneumo-
immediately transferred to the Intensive Care thorax. He was therefore inserting a chest
Unit (ITU), with a presumed diagnosis of pla- drain. Once this had been accomplished and
cental abruption. Despite massive blood trans- transfusion had restored the blood pressure, a
fusion, her condition deteriorated, and, despite laparotomy was carried out by the obstetrician.
ventilation, it was difficult to maintain her PO2. A small arterial bleeder was found at the ileo–
The ITU team felt that attempts to induce colic anastomosis and was easily dealt with. The
labor needed to be delayed until her condition patient, who was the wife of a solicitor, made
improved. Eventually, ventilation resistance was an uncomplicated recovery. The obstetrician
so great that the ITU team was of the opinion expected that he might find a legal suit impend-
that death was imminent. The obstetrician ing, but instead received a case of champagne
was therefore asked to consider carrying out a and letter of thanks from the solicitor husband.
laparotomy and delivery of the dead baby in the This lady also subsequently went on to have a
hope that this might improve the situation. As successful pregnancy.
the patient was deemed too ill to leave ITU, the On yet another occasion, the author was
operation was performed on an ITU bed. On called in at 3 a.m. by a consultant colleague
entering the abdomen, a massive hemoperito- because a patient who had had a vaginal delivery
neum was encountered, and the first thought with a very extensive vaginal and perineal lacer-
was of a ruptured uterus. However, the uterus ation was still bleeding heavily after more than
was found to be intact, and, upon further an hour of attempted suturing of the tear, and
exploration, it became obvious that the source no fewer than 18 units of blood had been trans-
of the intra-abdominal hemorrhage had been a fused. The operating theater looked like a bat-
ruptured liver. A general surgeon was called, tlefield theater, and the vaginal tissues appeared
who was able to secure hemostasis with several like wet blotting paper, with no identifiable
large hemostatic liver sutures, and the patient anatomical layers. By then, the patient had
made a slow recovery. During the postoperative major clotting deficiencies, and anesthetists and
period, however, it became apparent that she hematologists were busy attempting to correct
also had HELPP syndrome. A stormy recovery that. Attempts were made at packing the vagina
ensued, but a year later the patient was pregnant and applying pressure, but to no avail. A
again and delivered a healthy baby. gynecological oncology colleague was contacted
Another once-in-a-lifetime experience con- to discuss internal iliac artery ligation, and he
cerned a late vaginal termination at 18 weeks for advised that this should be done forthwith. The
a major chromosomal abnormality. During the author had not participated in such a procedure
procedure, it was apparent that the uterus had for something like 20 years, and, although the
been perforated and a laparotomy was therefore gynecological oncologist said he would come in,
carried out. A small tear was found in the he advised that time should not be wasted in
caecum and a general surgeon called in. He rec- getting on with the procedure. To the author’s
ommended partial right colectomy, which was relief, the requisite details of the anatomy and
elegantly performed, and the perforation of the necessary procedure were retrieved from the
394
416
30 August 2006 14:25:06
cerebral archive almost automatically. By the chapters of this book. However, decisions about
time the oncologist arrived, the hemorrhage was which intervention to try, and after how much
almost completely under control, and it was blood loss, remain difficult, and are influenced
then possible to complete hemostasis with a few by the likely future reproductive wishes of
additional vaginal sutures. After a short period the woman, as well as the facilities or lack
of intensive care, the young woman recovered thereof available in the particular obstetric unit.
well, as did the anatomy of the vagina and Whilst much progress has been achieved in the
perineum. last few decades, there remain many parts of the
A final case involved a collapse at 36 weeks, world where treatment options either are not
with abdominal distension and extreme pain much greater than they were 50 or more years
and tenderness. The fetal heart tones were still ago in more developed countries or are even
present, and the presumed diagnosis was pla- less, being hampered by the logistic consider-
cental abruption. The patient was immediately ations detailed in still other chapters in this
taken to theater for Cesarean section. On open- volume.
ing the peritoneum, a massive hemoperitoneum The major challenge in the 21st century
gushed forth, but the uterus was perfectly soft in this field is to narrow the inequalities of
and normal in color. A Cesarean section was health-care provision in childbirth. It is hoped
carried out and a healthy baby delivered. It was that this textbook, the first ever to discuss the
assumed that the source of bleeding could be a topic of postpartum hemorrhage in a compre-
splenic artery aneurysm accident, and a four- hensive manner, will go a long way in helping
quarter exploration of the abdomen carried out. health-care providers to achieve this goal, for it
The upper abdomen revealed no bleeding what- should be obvious, even to the most neophyte
soever, and eventually an arteriovenous malfor- reader, that the problems related to postpartum
mation at the brim of the pelvis was found to be hemorrhage are not confined to one country or
bleeding. A vascular surgeon was called in to to one region. They are indeed world-wide, and
check that hemostasis was satisfactory. After an their control will be facilitated by collaborations
8-unit blood transfusion, the patient and baby and partnerships, as seen in this textbook in
did well. which several chapters present details of what is
being done in the developing as well as the
developed world.
CONCLUSION
The plethora of interventions available to the Further reading
obstetrician now includes many different drugs
Donald I. Practical Obstetric Problems. London: Lloyd
to promote uterine contraction and hemostasis, Luke Ltd, 1969
a complex range of hematological products, and Williams W. Deaths in Childbed. London: H. K.
surgical interventions, including the B-Lynch Lewis, 1904
stitch, the use of intrauterine pressure balloons, Ramsbotham F. The Principles & Practice of Obstetric
and early resort to hysterectomy or radiological Medicine & Surgery in Reference to the Process of
embolization. All are described in detail in other Parturition. London: Churchill, 1941
395
417
30 August 2006 14:25:07
43
THE MIDWIFE CONFRONTS POSTPARTUM HEMORRHAGE
A. M. Ward
INTRODUCTION PREVENTION OF POSTPARTUM

HEMORRHAGE
As repeatedly stated earlier in this book, post-
partum hemorrhage is a major killer of women Antenatal prevention
throughout the world1 and is the second leading
Prevention of postpartum hemorrhage should
cause of admission of women to high-dependency
begin in the antenatal period. Midwives should
units in the Western world2,3. Postpartum hem-
assess women’s risk factors at every antenatal
orrhage also causes significant morbidity for
visit and then, in partnership with the women,
women in the Third and Western worlds1,4,5.
plan care that identifies the most appropriate
Waterstone and colleagues6 noted that two-
lead health-care professional10. The antenatal
thirds of severe maternal morbidity is related to
risk factors, all within the midwives’ domain to
severe hemorrhage. It stands to reason that any
determine, that most commonly are reported
reduction in the frequency of postpartum hem-
for postpartum hemorrhage follow11:
orrhage would impact the lives of women and
their families throughout the world1. Given ● Body mass index > 30 kg/m2
these circumstances, it is essential that mid-
wives, as first-line staff, be able to prevent, ● Previous postpartum hemorrhage
identify early and provide appropriate manage- ● Antepartum hemorrhage
ment during a postpartum hemorrhage7,8.
Midwives practising in the United Kingdom ● Placental abruption
(UK) are fortunate to work in a country with a ● Placenta previa
relatively low maternal mortality rate1. At first
sight, the role of midwives in the management ● Multiple pregnancy
of a postpartum hemorrhage may seem obvious, ● Macrosomic infant
that is, they should diagnose the bleed, call for
help and instigate emergency treatment9. The ● Previous uterine surgery
reality of the management of a postpartum hem- ● Antenatal anticoagulation
orrhage is much more complex than this, how-
ever, and involves an ability to work effectively Other risk factors include anemia, poly-
within a multidisciplinary team and to possess hydramnios, maternal age, uterine fibroids and
an indepth knowledge of the social, psychologi- a history of retained placenta7. Nulliparity has
cal and physiological processes that surround recently been identified as a possible risk factor
pregnancy and childbirth. Midwives should be for postpartum hemorrhage, rather than grand
central to the prevention, identification and multiparity12. This is important, and it could
management of postpartum hemorrhage and well be that this group of women has not
these precepts will form the focus of this chap- previously been identified as being at significant
ter. The degree to which midwives can achieve risk of postpartum hemorrhage. In the past, the
these goals will obviously vary with local cus- management of such women may have been
toms, resources and practices, but the goals sub-standard as postpartum hemorrhage was
should remain the same regardless. not anticipated12. The above-mentioned risk
396
418
30 August 2006 14:25:07
The midwife confronts postpartum hemorrhage
factors focus totally on the physical aspects of between women and midwives and reduces the
pregnancy. To ensure the optimum safety of potential for conflict when the safest manage-
women and their babies and to ensure holistic ment of care conflicts with women’s wishes for
care, these risk factors need to be assessed in their childbirth experience15–18.
conjunction with other risk factors for severe
maternal morbidity; these include maternal age
> 34 years, social exclusion and non-white Intrapartum prevention
ethnicity6.
Intrapartum prevention of postpartum hemor-
Risk assessments undertaken by midwives
rhage should begin in the antenatal period with
need to carefully consider social and psycho-
the aim of helping women to be as healthy as
logical aspects of women’s lives, as there is clear
possible, both physically and emotionally, and
evidence that women from poor areas, socially
should include preparation for childbirth, focus-
excluded groups and ethnic minorities have
ing on strategies to keep the process normal19.
poorer health outcomes than other groups
Throughout the intrapartum period, midwives
of women1,13,14. Midwives particularly need to
need to be with women supporting them,
focus care on women who book late, are poor
encouraging them to be mobile and offering
attendees or who do not access antenatal care at
alternative methods of pain relief that are less
all, as these are key indicators of poorer out-
likely to interrupt the progress of labor20,21.
comes13. This requires effective communication
Labor causes a great deal of insensible fluid loss
links with other groups such as Public Health
and women need to be kept well hydrated to
Nurses, General Practitioners and Social Ser-
ensure adequate circulating volumes at delivery
vices to ensure these special women are identi-
to enable them to cope with any excessive blood
fied as being pregnant as early as possible and
loss22. Women should also be provided with a
provided care in an environment appropriate for
quiet, private environment where they feel safe
them and tailored to meet their social, cultural
and protected to reduce the need for interven-
and psychological needs1,13.
tion during the process of labor21,23. All this is
The National Institute for Clinical Excel-
even more vital in areas where there is no direct
lence (NICE) has produced guidelines for ante-
access to intravenous fluids in the event of a
natal care of healthy pregnant women in the
postpartum hemorrhage.
UK10. These are useful in honing effective use
Midwives need an indepth understanding of
of resources, but midwives need to be mindful
intrapartum risk factors and need to constantly
that the guidelines are intended to guide the
reassess the woman for risk throughout labor24.
care of healthy pregnant women. The NICE
Intrapartum risk factors for postpartum hemor-
document15 clearly states that women should
rhage include:
have a plan of care that is relevant to their indi-
vidual physical, social and psychological needs, ● Prolonged labor > 12 h
and the World Health Organization (WHO)1
● Prolonged third stage > 30 min
further indicates that this also needs to be
culturally specific to women’s backgrounds if it ● Retained placenta
is to be truly effective.
● Febrile illness
Although midwives clearly need to know the
risk factors for postpartum hemorrhage, identi- ● Instrumental delivery
fying risk factors is not enough if appropriate
● Cesarean section, especially emergencies in
care is not then instigated13. Once identified,
late first or second stage of labor
risk factors need to be acted upon. Even where
women have strong views about the type of ● Amniotic fluid embolism
childbirth experience they desire, open, frank
● Placental abruption
discussion of identified risk factors and their
implication for women and their babies, The first four conditions are most likely to cause
with time to assimilate and consider the infor- an atonic uterus, whereas operative deliveries
mation provided, leads to stronger relationships are the main cause of uterine, cervical or vaginal
397
419
30 August 2006 14:25:07
trauma; embolisms and abruptions are common loss retrospectively. As blood loss may not be
causes of coagulopathy, although this is the least entirely revealed, its estimation is notoriously
common reason for postpartum hemorrhage11. inaccurate and difficult26.
The debate on whether to manage the third Healthy, young women can compensate for
stage of labor actively could fill an entire text routine post-delivery blood loss very effectively,
itself when considering practice in the UK and and this toleration is increased even further if
other developed countries. In the Third World, there has been a healthy increase in blood vol-
however, this is a different matter and routine ume during pregnancy22. Normally, plasma vol-
active management of the third stage of labor ume increases by 1250 ml and the red cell mass
could save many women’s lives as well as saving also increases, resulting in women being able to
many more from the abject misery of severe tolerate a drop in their pre-delivery blood vol-
morbidity brought about by a postpartum hem- ume of up to 25% and remain hemodynamically
orrhage1,5,6,12. This treatment needs to be stable22. In practice, this means that midwives
carried out in conjunction with having in need to be encouraged to ignore machines and
place trained birth attendants that understand use their clinical skills of observation. They
women’s specific cultural issues and are aware need to be alert to signs of earlier stages of
of when pregnancy and labor are not progress- shock – pallor, sweating and muscle weakness
ing normally1. characterized by severe and rapid fatigue22.
The type of management used for the third When women become restless and confused,
stage of labor may be of no real consequence in shock is advancing rapidly and immediate,
a well-nourished, healthy population, but it is aggressive treatment is needed if not already
vitally important that midwives can clearly instigated22 (see also Chapter 8).
identify those women at increased risk of a There are only two definitions for post-
postpartum hemorrhage, as well as understand- partum hemorrhage, primary (occurring within
ing and carrying out expectant and active the first 24 h after birth) or secondary (occur-
management of the third stage of labor25. ring after 24 h and before 6 weeks postpartum).
Table 1 describes the main components of each In contrast, experienced health-care practi-
management option for the third stage of labor. tioners will recognize that, in practice, there are
three different presentations of postpartum
hemorrhage:
DIAGNOSIS OF POSTPARTUM
HEMORRHAGE AND POSTPARTUM (1) Rapid loss of blood at or just shortly after
PREVENTION delivery;
Definitions in themselves may not be useful, (2) Constant heavy lochia that persists for a
as they often involve measurement of blood significant length of time after delivery;
Table 1 Options for the management of the third stage of labor
Active management Expectant management

Oxytocic drug given at delivery of anterior shoulder No oxytocic drug given
Cord clamped and cut immediately Cord not clamped until pulsation ceased, then
only clamped at baby’s umbilicus
When uterus central and well contracted, controlled No cord traction
cord traction applied Signs of separation awaited:
● Rise in fundus
● Lengthening of cord
● Trickle of blood at introitus
Midwife delivers placenta and membranes Maternal effort delivers placenta and membranes
398
420
30 August 2006 14:25:07
(3) Bleeding after the first 24 h following child- hemorrhage is effective fluid resuscitation8,22
birth. (see also Chapter 5). Midwives may be con-
cerned about which fluids are best, but their
It is the second type of bleeding that can cause
focus needs to be on ensuring fluid is adminis-
problems for health-care practitioners, because
tered quickly and is not cold. Where available,
it is often the type of bleeding that is missed.
fluid warmers and pressure bags must be uti-
Women will experience heavy lochia that
lized. Every 1 ml of blood lost needs to be
they report. Their sanitary protection will be
replaced with 3 ml of fluid until blood is
changed and then, a little while later, the same
available8,22. To ensure fluid can be delivered
will happen and they will report it again, but
as quickly as possible, two wide-bore, short
this may be to another member of staff who
cannulae need to be used, as the volume that
is unaware of the previous loss. Midwives and
can be infused through a cannula is propor-
midwifery assistants need to be encouraged to
tional to the diameter and inversely propor-
quantify the amount of blood lost and record
tional to its length22. Midwives may also be
this in the maternal notes, keeping a running
concerned about commencing intravenous flu-
total of the amount of blood lost to alert them to
ids without prescription or written order. How-
women who are bleeding significantly but still
ever, postpartum hemorrhage is an emergency
compensating adequately22.
situation and, as such, midwives can administer
resuscitative fluids without a prescription first9.
Women need to be kept warm as hypothermia is
MANAGEMENT OF POSTPARTUM
a consequence of hypovolemic shock8,22. As the
HEMORRHAGE
assessment of renal function is an essential part
As any postpartum hemorrhage has the poten- of management once the bleed is controlled, an
tial to cause maternal collapse with loss of con- indwelling urinary catheter should be inserted,
sciousness, midwives need to be competent with using strict aseptic techniques to avoid infection
basic life support (ABC algorithm)8,22,27. The in women who are already compromised as a
first principle of which to be aware is that a result of the postpartum hemorrhage22.
single individual cannot effectively manage an
emergency situation, and help must be urgently
CARE FOLLOWING A POSTPARTUM
requested prior to commencing any treat-
HEMORRHAGE
ment27. Midwives need to constantly ensure
that women have patent airways and are breath- Women who have sustained a significant post-
ing adequately; here, expensive technology is partum hemorrhage need to be receive one-to-
not required. If women do not respond when one care to facilitate close monitoring4-6,12.
spoken to, then they potentially cannot manage Initially, the focus of care will be on the
their own airway and an individual with the woman’s physical condition, observing and
appropriate skills and training needs to do this. monitoring urinary output, fluid intake, vital
Until the airway and breathing are effectively signs and subsequent blood loss. Ideally, such
brought under control, there is little point care is best provided in an obstetric high-
undertaking any other task, as hypoxia can dependency unit if available. Any women
kill women much faster than hypovolemia22. requiring mechanical ventilation should be
Proper airway management needs to ensure that cared for in an intensive care unit4-6,12.
oxygen therapy is optimally utilized to ensure Intensive monitoring often means that other
depleted hemoglobin is as well oxygenated as aspects of care important to women following
possible to prevent cell death22. Once sufficient childbirth are neglected3. Care provided by
members of the team are present, then they can midwives also needs to include the psychologi-
move onto maintaining the circulatory system cal well-being of women and the integration of
and determining the cause of the postpartum the family unit who may be bewildered by the
hemorrhage (see Chapter 13). goings-on after the delivery3,24. Women who are
The key to reducing morbidity and mortal- conscious need to have contact with their babies
ity in the management of a postpartum and feel central in any decision-making around
399
421
30 August 2006 14:25:08
the care of their babies3. Skin–skin contact is a ● Printing name and qualification alongside
simple procedure that can be carried out even the first signature in any records;
for the sickest women and can be beneficial to
● Writing legibly.
women as well as their babies; it assists in the
effective introduction of breastfeeding and has Documentation of vital signs and urine output
relaxing properties for women and babies is essential following a significant postpartum
alike28. hemorrhage, but documentation itself will not
Given the traumatic nature of a postpartum ensure effective management of sick women. It
hemorrhage, women will need support long into is vital to ensure that trends in all important
the postnatal period as they recover physically physical parameters, especially respiration, are
and emotionally29. Initial debriefing may not be being acted upon effectively because they can
beneficial and may, in fact, be detrimental to indicate the effectiveness of any treatment as
these women. Later debriefing may discuss, well as when women are deteriorating2,3. Scor-
among other things, the risk of recurrent ing tools can be developed that assist practitio-
postpartum hemorrhage. After the crisis has ners to identify women who are not responding
passed, these women need effective long-term to treatment and therefore require the expertise
follow-up. In larger units, it may be appropriate of senior obstetricians and anesthetists and
to have a lead midwife and obstetrician to run admission to an intensive care setting.
combined postnatal clinics for these women,
where recovery can be monitored and any
COMMUNICATING EFFECTIVELY
concerns about subsequent pregnancies can
be discussed with relevant health-care In any emergency health care, professionals are
professionals29. relieved when help arrives, but the larger the
team the more complex communication and the
more difficult it can be to manage the situation
DOCUMENTATION effectively and utilize the team efficiently31,32.
Someone needs to take charge, stand back,
Accurate documentation is crucial during an
observe and then direct the working of the
emergency procedure and the leader of the
team31,33. The role of this lead individual is
emergency team needs to task someone by
also to constantly evaluate the effectiveness
name to record events as they occur, including
of treatments instigated and to constantly be
the times team members enter and leave the
re-thinking the potential causes of postpartum
room, as well as the timing of any procedures
hemorrhage when the treatment instigated is
and drugs administered, including route and
not being effective in controlling the bleeding34.
dose30. Good records are an indication that the
Historically, this has been the most senior
quality of care given to women was of a good
obstetrician on duty in an obstetric maternity
standard30. Midwives have a professional duty
unit. Both obstetricians and midwives recognize
to ensure records are kept as contemporane-
that the person co-ordinating the team at an
ously and accurately as possible9,30. Good
emergency should be the most experienced
practice is to ensure that the documentation
clinician available31,33. In some circumstances,
completed by the named scribe is included in
this may be the senior midwife who will be more
the maternal records and not disposed of once
experienced than the house officers.
individual health-care practitioners have used
An emergency situation is no time for hierar-
them to complete their own notes. Accurate
chy. Communication needs to be precise, with
record-keeping is vital to reduce the risk of
tasks directed to a named individual (not Mr or
successful litigation, but it is also vital in the
Mrs Somebody) and feedback requested from
active debriefing of all team members31 (see also
that individual at regular intervals. Training of
Chapter 13). Simple factors can dramatically
teams, within individual units or the community
improve the quality of record-keeping and only
setting, needs to be multidisciplinary, realistic
take seconds30. These include:
to the work environment, scenario-driven and
● Dating and timing all new entries; based on real timing and action to make it as
400
422
30 August 2006 14:25:08
realistic as possible33. For example, if simulating efficiencies and can be achieved with effective
a postpartum hemorrhage in a home setting, timetabling and allocation36. Anecdotally, the
then paramedics need to be involved and the training improves communication and team
setting needs to reflect the equipment that work, but needs to be audited against unit
would be available to midwives in those situa- guidelines considering maternal outcomes and
tions. For midwife-led units not attached to focusing on morbidity and mortality rates, as
obstetric units, the training should involve well as adherence to the guidelines themselves.
paramedics and the ambulance service and not
include the management schemes using drugs
DEBRIEFING
and techniques that would not be available to
those midwives. Part of ensuring a team learns from stressful
clinical incidences is a review of their perfor-
mance as close to the event as possible. The
TRAINING
purpose of this ‘debriefing’ session should be to
Team sports have recognized for decades that, focus on what was done well. It can be used
to ensure that a team functions efficiently and to identify what needs to be shared with team
effectively, its members must train together; members not involved in the emergency, to aid
such training must focus on utilizing individual their development and learning, as well as to
skills to their greatest potential for the good of provide a forum where those involved in the
the team. In the NHS, individual professional emergency can vocalize how they feel in a pro-
bodies have trained their own practitioners tective environment. This will enable learning
largely in isolation of other health-care profes- whilst at the same time offering professional and
sionals and then they have been expected to emotional support, recognizing that health-care
work as a well-oiled machine in times of great professionals are caring individuals who can be
stress, with minimal understanding of each profoundly affected by traumatic situations37.
others’ strengths and weaknesses31,35. Happily, Debriefing is a useful tool to help team mem-
this trend is changing and the benefit of multi- bers recognize that they are valued and the role
disciplinary training is being recognized33. they play in the effective running of the team,
In the Yorkshire Region, this has been taken all of which can help increase job satisfaction
one step further with many maternity units and reduce the number of professionals leaving
adopting a regional training program aimed midwifery and obstetrics37.
at managing the first 20–30 min of obstetric
emergencies effectively. As medical trainees
CONCLUSION
rotate around the region, there is a systematic
approach to the training for management of Midwives are central to the effective prevention,
obstetric emergencies that they are expected to recognition and treatment of postpartum hem-
complete as early as possible into their time in a orrhage. They need to be aware of the risk
new unit. Units in the region that have adopted factors for postpartum hemorrhage and take
the training have made it mandatory for anyone appropriate action when they are identified.
involved in the intrapartum care of women, They should also be skilled in basic life support
from health-care assistants to consultants. and have an understanding of the pathophysio-
Scenarios are run real-time using manne- logy of hypovolemic shock. This knowledge
quins, and participants are expected to carry out must be used in conjunction with an under-
procedures as if it were a real emergency. This is standing of women’s social, cultural and
then videoed and the participants on the day psychological well-being.
debrief themselves, with a facilitator assisting Training as multidisciplinary teams can be
them to focus on issues of leadership, control effective in improving outcomes for women and
and communication, all of which have been their families. The Yorkshire model may be
highlighted as factors in suboptimal care13. beneficial in units that have trainees who
Dedicated time is given for this training, which rotate throughout their region. Effective com-
has been shown to improve outcomes and munication and leadership are vital in the
401
423
30 August 2006 14:25:08
management of any obstetric emergency and 17. Guiver D. The epistemological foundation of
scenario-based training can be used to highlight midwife-led care that facilitates normal birth.
issues of control and communication. Evidence Based Midwifery 2004;2:28–34
18. Hunter B. Conflicting ideologies as a source of
emotion work in midwifery. Midwifery 2004;20:
References 261–72
19. Eames C. Midwives’ role in preparing women for
1. The World Health Organization. The World
birth. Br J Midwifery 2004;12:447–50
Health Report 2005 – Make Every Mother and
20. Yogev S. Support in labour: a literature review.
Child Count. http://www.who.int/whr/2005/en/
MIDIRS Midwifery Digest 2004;14:486–92
index.html. Accessed 20th December 2005
21. Oudshoorn C. The art of midwifery, past,
2. Okafor UV, Aniebu U. Admission pattern and
present and future. MIDIRS Midwifery Digest
outcome in critical care obstetric patients. Int J
2005;15:461–8
Obstet Anesthesia 2004;13:164–6
22. Hofmeyr GJ, Mohlala BKF. Hypovolaemic
3. Goebel N. High dependency midwifery care –
shock. Best Practice Res Clin Obstet Gynaecol
does it make a difference? MIDIRS Midwifery
2001;15:645–62
Digest 2004;14:221–6
23. Ryan M, Roberts C. A retrospective cohort study
4. Paruk F, Moodley J. Severe obstetric morbidity.
comparing the clinical outcomes of a birth centre
Curr Opin Obstet Gynecol 2001;13:563–8
and labour ward in the same hospital. Aust Mid-
5. Waterstone M, Wolfe C, Hooper R, Bewley S.
wifery J 2005;18:17–21
Postnatal morbidity after childbirth and severe
24. Simpson KR. Failure to rescue: implications for
obstetric morbidity. Br J Obstet Gynaecol 2003;
evaluating quality of care during labour and
110:128–33
birth. J Perinat Neonat Nursing 2005;19:24–34
6. Waterstone M, Bewley S, Wolfe C. Incidence
25. Rogers J, Wood J, McCandlish R, Ayres S,
and predictors of severe obstetric morbidity:
Truesdale A, Elbourne D. Active versus expec-
case-control study. Br Med J 2001;322:1089–94
tant management of third stage of labour: the
7. Selo-Ojeme DO. Primary postpartum haemor-
Hichingbrooke randomised controlled trial.
rhage. J Obstet Gynaecol 2002;22:463–9
Lancet 1998;351:693–9
8. Clarke J, Butt M. Maternal collapse. Curr Opin
26. Prasertcharoensuk W, Swadpanich U,
Lumbiganon P. Accuracy of blood loss
9. NMC. Midwives Rules and Standards. London:
estimation in the third stage of labour. Int J
NMC, 2004
10. NICE. Antenatal Care. Routine Care for the
27. Resuscitation Council (UK). Resuscitation
Healthy Pregnant Woman. London: NICE, 2003
Guidelines 2005. [online]. http://www.resus.org.
11. McLintock C. State-of-the-art lectures: Post-
uk/pages/guide.htm. Accessed 20th December
partum Haemorrhage. Thrombosis Res 2005;
2005
1155:65–8
28. Carfoot S, Williamson P, Dickson R. A
12. Hazra S, Chilaka VN, Rajendran S, Konje JC.
randomised controlled trial in the north of
Massive postpartum haemorrhage as a cause of
England examining the effects of skin-to-skin
maternal morbidity in a large tertiary hospital.
contact on breast feeding. Midwifery 2005;21:
J Obstet Gynaecol 2004;24:519–20
71–9
13. CEMACH. Why Mothers Die 2000–2002.
29. Kline CR, Martin DP, Deyo RA. Health Conse-
London: RCOG, 2004
quences of Pregnancy and Childbirth as Per-
14. Doran T, Denver F, Whitehead M. Is there a
ceived by Women and Clinicians. Obstet Gynecol
north-south divide in social class inequalities in
1998;92:842–8
health in Great Britain? Cross sectional study
30. NMC. Guidelines for Records and Record Keeping.
using data from 2001 census. Br Med J 2004;
London: NMC, 2005
328:1043–5
31. Brownlee M, McIntosh C, Wallace E, Johnston
15. Graham WJ, Hundley V, McCheyne AL, Hall
F, Murphy-Black T. A survey of inter-
MH, Gurney E, Milne J. An investigation of
professional communication in a labour suite.
women’s involvement in the decision to deliver
Br J Midwifery 1996;4:492–5
by caesarean section. Br J Obstet Gynaecol 1999;
32. Duff E. No more ‘quarrelling at the mother’s
106:213–20
bedside’: inter-professional approaches can help
16. Buckley SJ. Undisturbed birth – nature’s hor-
to stop women dying. MIDIRS Midwifery Digest
monal blueprint for safety, ease and ecstasy.
2004;14:35–6
J Perinatal Psychol Health 2003;17:261–88
402
424
30 August 2006 14:25:09
33. Cro S, King B, Paine P. Practice makes perfect: the early 20th century. In Kirkham M, ed. Super-
maternal emergency training. Br J Midwifery vision of Midwives. Cheshire: Books for Midwives
2001;9:492–6 Press, 1996
34. Mousa HA, Walkinshaw S. Major postpartum 36. Sabey A, Jacobs K. Live and learn. Health Service
haemorrhage. Curr Opin Obstet Gynaecol 2001; J 2003;16:32–3
13:595–603 37. Gamble J, Creedy D. Content and processes of
35. Heagerty BV. Reassuring the guilty: The Mid- postpartum counseling after a distressing birth
wives Act and the control of English midwives in experience: a review. Birth 2004;31:213–18
403
425
30 August 2006 14:25:09
44
SEPSIS AND POSTPARTUM HEMORRHAGE
B. Das and S. Clark
INTRODUCTION may consist of peptostreptococci, bacteroides,

streptococci, enterococci and E. coli. Group A
Sepsis and postpartum hemorrhage are linked
streptococcal endometritis, a rare cause in
by common predisposing factors, especially
developed countries, usually occurs in early-
considering that secondary postpartum hemor-
onset disease (within the first 48 h of delivery),
rhage can follow infection of retained placenta
often with high temperature > 39°C (102.2°F).
or endometrium. Depending on the extent and
In contrast, Chlamydia trachomatis is involved
severity of the condition, postpartum uterine
with late-onset disease (from 2 days up to
infection is designated as postpartum endo-
6 weeks postpartum) in patients who deliver
metritis, endomyometritis or parametritis.
vaginally.
Postpartum endometritis may be divided into
early-onset disease, occurring within the first
48 h, and late-onset disease, presenting up to CLINICAL FEATURES AND
6 weeks postpartum. This chapter reviews the INVESTIGATIONS
causes, pathogenesis and management of
uterine sepsis. Postpartum endometritis is diagnosed by signif-
icant pyrexia associated with uterine tenderness
or abnormal lochia in absence of other obvious
CLINICAL RISK FACTORS sources of infection. Significant pyrexia is
defined as oral temperature of 38.5°C
The most critical factor is the route of delivery. (101.3°F) or higher in the first 24 h after deliv-
After vaginal delivery, the incidence of post- ery or 38°C (100.4°F) or higher, for at least 4
partum endometritis varies between 0.9 and consecutive hours, in the first 24 or more hours
3.9%, but can increase to 12–51% after Cesar- after delivery. The first manifestation of fever
ean section. Factors such as duration of labor, may occur at night2,3. Uterine sepsis associated
bacterial vaginosis and vaginal interventions with late-onset disease and secondary post-
are secondary predictors of post-Cesarean partum hemorrhage usually presents as fever
endometritis. Early rupture of the membranes, on days 10–12 after delivery.
mid-forceps delivery, poor maternal health and Patients with suspected postpartum endo-
soft tissue trauma act as ‘relative risk factors’ for metritis should have early clinical evaluation
uterine sepsis, although they are not present including bimanual pelvic examination to deter-
in most patients with such infections1. Indi- mine size, consistency and tenderness of the
gent parturients are at higher risk of developing uterus and to detect any adnexal mass (ultra-
postpartum endometritis. sound study may help, if available). Cesarean
section/episiotomy wounds should be assessed
for evidence of surgical site infection. Unremit-
ETIOLOGICAL AGENTS
ting pain at the operative site may indicate
Postpartum uterine sepsis is thought to arise necrotizing fasciitis, wherein urgent debride-
from an ascending infection caused by coloniz- ment is life-saving3. A distant site of infection,
ing vaginal flora. Etiological agents include both e.g. urinary or respiratory tract, should be ruled
aerobic and anaerobic micro-organisms and out.
404
426
30 August 2006 14:25:09
Sepsis and postpartum hemorrhage
Laboratory investigations (where facilities are been used, albeit with greater failure rates than
available) include full blood count, transcervical the combination of gentamicin and clindamy-
cultures (aerobic and anaerobic) and one set of cin4. Alternative antibiotic regimens are shown
blood cultures, remembering that only 10–20% in Table 1. Intravenous clindamycin and intra-
of patients with postpartum endometritis have venous once-daily gentamicin are the cheapest
bacteremia. The presence of bacteremia does of the antibiotic regimen options, an important
not predict severity of infection or prolonged issue in countries with restricted resources.
recovery. Transcervical cultures, although diffi- Parental therapy is continued until the
cult to interpret because of contamination with patient is pain-free, afebrile for 24–48 h, the
vaginal flora, are helpful in those patients in leukocyte count returns to normal, and oral
whom initial therapy fails. Whenever possible, liquids and solids are tolerated. There is no
culture/antigen tests for chlamydia should be need to continue with oral antibiotics after stop-
performed in patients with late-onset disease or ping parental treatment. Patients with positive
those who are at high risk for acquisition of such cultures for chlamydia should receive a 7-day
infections. course of azithromycin or doxycycline, even if
there is good response to the initial empirical
antibiotic regimen. Azithromycin and doxy-
ANTIBIOTIC THERAPY AND cycline, although good antichlamydial agents,
FURTHER MANAGEMENT are bacteriostatic drugs and should not be
The aim of the antibiotics should be to provide used as first-line antimicrobial agents to treat
bactericidal cover for aerobic Gram-positive endometritis.
cocci, Gram-negative bacilli and β-lactamase- Failure to respond to the initial antibiotic
producing anaerobes. Those antibiotics which regimen in 48 h or clinical deterioration
have been used for prophylaxis should be requires further clinical evaluation and investi-
avoided. Empirical treatment should be com- gations to rule out another site of infection and
menced as soon as possible. Parental treatment complications (see Figure 1). The antibiotic
with once-daily intravenous gentamicin and regimen needs to be altered, preferably after
intravenous clindamycin is an effective combi- reviewing transcervical culture and sensitivity
nation, especially in post-Cesarean section results (see Table 2).
patients and those awaiting surgical inter-
ventions, including removal of retained
PREVENTION OF UTERINE SEPSIS
placenta. Gentamicin levels need to be moni-
tored. However, other alternatives, including Strategies to prevent uterine sepsis include
extended-spectrum penicillins or second- improved obstetric care and the use of prophy-
generation cephalosporins (cefoxitin), have lactic antibiotics in high-risk patients, as well as
Table 1 Initial antibiotic therapy1,2,4
Day 1
1. Clindamycin 900 mg 8-hourly + intravenous gentamicin 5 mg/kg body weight* once daily
or
2. Intravenous piperacillin–tazobactam 4.75 g 6-hourly
or
3. Intravenous metronidazole 500 mg 8-hourly + intravenous gentamicin 5 mg/kg body weight* once daily
or
4. Ampicillin–sulbactam 3.1 g 6-hourly + intravenous gentamicin 5 mg/kg body weight* once daily
*Monitor gentamicin level
405
427
30 August 2006 14:25:09
Day 1
Specimen collected: Empirical antibiotics, e.g. Surgical intervention if

transcervical swab for i.v. gentamicin + i.v. clinically relevent
aerobic organism and clindamycin
chlamydia* – blood
culture
*Specimen for Chlamydia trachomatis to be obtained on patients with:
(a) late onset of disease or

(b) high risk for acquisition of this infection
Day 3
Patient on empirical antibiotic
Afebrile Febrile
Continue i.v. antibiotics ·Clinical evaluation
until:
·Rule out other source of infection
·Patient afebrile × 24–48 h
·Urine, wound swab for culture
·Pain-free and sensitivity (if facility available)
·Normal cell count
·Imaging ± surgical intervention
Altered antibiotic regimen (see

Table 2) or specific antimicrobial
therapy as per culture results, e.g.
i.v. benzyl penicillin and i.v.
clindamycin for Group A
streptococci
Figure 1 Flow chart of treatment regimens for patient with suspected uterine sepsis and postpartum
hemorrhage1,2,4
coverage of planned or emergency surgical minimizing the number of vaginal examinations

interventions. In areas with limited resources, all play an important role in reducing uterine
education with the emphasis on a clean environ- infection.
ment and simple infection control measures like The risk of infection increases with post-
hand-washing, cleaning the genital area, prefer- partum hemorrhage especially if the blood loss
ably with mild detergents/disinfectant, and is greater than 1 liter. If uterine sepsis occurs,
406
428
30 August 2006 14:25:10
Sepsis and postpartum hemorrhage
Table 2 Altered antibiotic regimen1,4. In all cases, the therapy should ideally be guided by culture results
If the patient deteriorates or Day 3 onwards if patient continues to be febrile on initial regimen:
Initial antibiotic regimen + another antimicrobial agent = altered regimen
1. Intravenous clindamycin + intravenous gentamicin* + intravenous ampicillin 1–2 g 6-hourly
or
2. Intravenous piperacillin–tazobactam + intravenous gentamicin 5 mg/kg* once daily)
or
3. Intravenous metronidazole + intravenous gentamicin* + intravenous ampicillin 1–2 g 6-hourly
or
4. Intravenous ampicillin–sulbactam + intravenous gentamicin intravenous metronidazole 500 mg/8 hourly
The choice of antibiotics in the altered regimen may be determined by transcervical isolates (where culture
facilities are available), cost and availability of antimicrobial agents.
*Monitor gentamicin levels
such hemorrhagic consequences can be devas- blood and a B-Lynch compression suture was
tating: collapse can lead to death, as discussed inserted to stay the continual bleeding. The
elsewhere in this book. In many developing patient received a total of 7 units of blood and 4
countries, the majority of deliveries do not units of fresh frozen plasma. The patient contin-
occur in a facility with a skilled attendant. ued to be pyrexial, her white blood cell rose
Traditional birth attendants (TBAs) need to from 10 000 to 25 000; lochia was offensive and
recognize the consequences of delayed referral. a transcervical swab grew E. coli and anaerobes.
Local and international organizations that aim She was therefore administered an intravenous
to provide resources to educate TBAs, increase clindamycin and once-daily intravenous genta-
access to skilled attendants and to facilities for micin regimen for uterine sepsis. Gentamicin
prompt care of postpartum hemorrhage and levels were regularly monitored and the patient
sepsis all help to decrease maternal mortality in was discharged after 8 days intravenous therapy,
these countries. having being ambulant, afebrile and pain-free
for 48 h. The baby remained well.
CASE STUDY
A 28-year-old primigravida presented at 41/40
References
weeks with a history of prolonged latent phase
of labor. She underwent a Cesarean section as 1. Mead PB. Infections of the female pelvis. In
she failed to respond to 8-h oxytocin infusion, Mandel GL, Bennet JE, Dolin R, eds. Principles
commenced after artificial rupture of the and Practice of Infectious Disease, 5th edn. Philadel-
phia: Churchill Livingstone 2000:1235–43
membranes. Prior to the procedure, the patient
2. Ledger WJ. Post partum endometritis diagnosis
received intravenous cefuroxime and intra-
and treatment: a review. J Obstet Gynaecol Res
venous metronidazole as she was found to be 2003;29:364–73
pyrexial (38.7°C). At Cesarean section, she had 3. Goepfert AR, Guinn AA, Andrews WW, et al.
offensive grade 1 meconium liquor; the placen- Necrotising fasciitis after caesarean delivery.
tal membranes were found to be adherent but Obstet Gynecol 1997;89:409–12
were successfully removed. A live baby with 4. French LM, Smaill FM. Antibiotic regimens for
good Apgar score was delivered; however, the endometritis after delivery. Cochrane review,
patient had primary postpartum hemorrhage 2004. Cochrane Library. Chichester: John Wiley
due to uterine atony. The patient lost 6 liters of & Sons
407
429
30 August 2006 14:25:10
45
THE SINGLE-UNIT TRANSFUSION IN THE BLED-OUT
OBSTETRIC PATIENT
V. Nama, M. Karoshi, M. Wac, L. G. Keith and S. A. Mujeeb
THE HISTORICAL PRACTICE OF widespread practice of single-unit transfusions

SINGLE-UNIT TRANSFUSION and the scrutiny to which they were subjected
during the 1960s.
The first reported blood transfusion took place
The debate on the usefulness of single-unit
in Rome in 1492. Pope Innocent VIII suffered
red blood cell transfusions continued with vigor
an apoplectic stroke, became weak and lapsed
in the following decades. In 1985, Grindon and
into coma. His physicians advised a blood trans-
associates6 condemned the scrutiny of single-
fusion in hopes that it would help their patient.
unit transfusions advocated in 1962 by the Joint
Employing the crude methods of the day, the
Blood Council, stating that the ‘administration
Pope failed to benefit from this intervention and
of one unit of blood more often reflects appro-
died by the end of that year. Since then, many
priate use than misuse’. One year later, in 1986,
advances have been made, blood groups have
an observational study reported that most
been discovered and transfusion practices
single-unit transfusions were administered
refined. Presently, blood is part of the everyday
during surgery and that the indications for 62%
armamentarium used by physicians to treat
of these were questionable7. Very shortly there-
countless diseases and conditions.
after, however, a case report published in the
In their 2005 retrospective analysis evaluat-
Journal of the American Medical Association
ing the role of single-unit red blood cell trans-
demonstrated that a single-unit transfusion
fusion, Ma and colleagues noted that, in the
increased the hematocrit to a safe level, espec-
1960s, single units were deemed insufficient to
ially in patients with low body mass index8.
correct anemia and, therefore, useless1. These
investigators also retold a clinical maxim from
that time, i.e. the patient whose transfusion
EVENTS LEADING TO ALTERATION
requirements could be met with one unit of red
OF BLOOD TRANSFUSION
blood cells was no more in need of a transfusion
PRACTICES
than the donor who gave 500 ml of blood.
Although the origins of this maxim are unclear, The conflicting opinions were so numerous that
the prevailing attitude in the medical commu- the US government decided to address this
nity was rather obvious. In the years following issue. However, this effort only provided
the 1962 Joint Blood Council call for scrutiny of tangential guidelines rather than ending the
blood transfusion practices in hospitals having a debate. In 1988, the National Institutes of
predominance of single-unit transfusions, one Health (NIH) formulated a Consensus Confer-
study found that 60–70% of these interventions ence Statement, entitled ‘Perioperative Red Cell
were not indicated2; in addition, two studies Transfusion’. This document recommended
found that all of the single-unit transfusions the threshold of hemoglobin concentration for
assessed were unnecessary or questionable3,4, transfusion to be lowered from 10 g/dl to a value
and yet another study found this practice ques- between 7 and 10 g/dl, depending on the clini-
tionable in 38% of assessed cases5. The very cal assessment, laboratory data, and volemia of
existence of these investigations documents the individual patients. At the same time, it was
408
430
30 August 2006 14:25:11
Single-unit transfusion
deemed advisable that the number of units of possibility that patients could be undertrans-
blood administered should be kept to a mini- fused, was repeated in 1998 by a study that
mum, mostly to reduce the number of transmit- pointed out the dangers of lowering transfusion
ted infections9. thresholds16.
The safety concerns expressed by the NIH Shortly thereafter, the 1999 multicenter
were amplified by additional reports associating Transfusion in Critical Care (TRICC) trial
allogenic blood transfusions with infections, randomized intensive-care patients to receive
transfusion-related and allergic reactions as ‘restricted’ or ‘liberal’ red blood cell trans-
well as adverse immunomodulatory effects10,11. fusions in order to analyze overall 30-day
A review published in the British Medical Journal mortality rates in patients who might be under-
in 199012 sought to bring an end to single-unit transfused17. Patients received transfusions
transfusions. Simply stated, this article opined when their hemoglobin concentrations dropped
that the single-unit transfusion significantly below 7 g/dl in the restricted treatment group or
increased the risks of viral infection while, at 9 g/dl in the liberal treatment group. Mortality
the same time, offered little or no therapeutic rates in this Canadian trial were similar in the
benefit. In the immediate aftermath of this pub- two groups, but mortality was significantly
lication, a study conducted in 1992 in a West lower among patients who were less acutely ill in
African city, also published in the British Medi- the restricted treatment group.
cal Journal, estimated the risk of HIV infection A more recent study (2003) conducted to
to be between 5.4 and 10.6 per 100 units of assess transfusion practices in a large Scottish
blood administered, a substantial threat even in hospital concluded that hospital clinicians
cases of single-unit transfusions in developing administered transfusions when their patients’
countries13. hemoglobin concentrations were between 7 and
9 g/dl18. Not only did the clinicians not follow
the available TRICC trial protocol, which pro-
RE-EMERGENCE OF
posed transfusions only when hemoglobin con-
CONSIDERATION OF THE
centrations fell below 7 g/dl, but the study’s
SINGLE-UNIT TRANSFUSION
authors reported that the Scottish practices
Not surprisingly, the 1990 British Medical Jour- were consistent with the findings of other
nal publication and others condemning single- recently published studies19-21. In 2004, hoping
unit red blood cell transfusion did not bring the to finally close the argument fuelled by the
debate to a halt. At the same time, the studies TRICC trial, a Canadian review reaffirmed the
from the 1960s to the 1980s warned against 1988 NIH recommendation regarding thresh-
administering single-unit red blood cell trans- olds by stating, ‘The quest for a universal trans-
fusions, and clinical guidelines suggested ever- fusion trigger, i.e. one that would be applicable
lower thresholds for transfusion as a means to to patients of all ages under all circumstances,
preserve blood resources and increase safety14. must be abandoned. All RBC (red blood cell)
Whereas some physicians became convinced transfusions must be tailored to the patient’s
that single-unit blood transfusions had no place needs as it arises’22. This statement is of particu-
in the treatment of anemia, others came to lar relevance to obstetricians, who commonly
believe, somewhat paradoxically, that individual deal with anemic parturients and occasionally
units of blood should be given as needed, but deal with bled-out postpartum mothers.
only when the patient’s hemoglobin concentra- In 2005, Ma and collaborators1 analyzed the
tion fell below a specified threshold, which var- results of single-unit transfusions for thresholds
ied across guidelines. As a counterplea to those that began at 7 g/dl, and were raised to 9 g/dl
who were opposed to single-unit transfusions by increments of 0.5 g/dl. These investigators
and in favor of low, specified thresholds of demonstrated that, for most patients, the trans-
hemoglobin concentration for transfusion, one fusion of one unit of red blood cells could raise
1992 study concluded that transfusion practices the hemoglobin concentration sufficiently to
should be audited for undertransfusion as well avoid the need for a second unit. When the goal
as overtransfusion15. This suggestion, i.e. the of red blood cell transfusion was to maintain the
409
431
30 August 2006 14:25:11
hemoglobin concentration above a threshold greater risk of mortality from anemia than other
considered as safe, they concluded, ‘the single- groups of individuals23. Figure 1 shows the case
unit transfusions may not only be appropriate, fatality rate in relation to maternal anemia.
but preferable.’ Unfortunately, the circum- Anemia during pregnancy increases the risk
stances leading to compliance with the premises of death, as it may lead to rapid cardiac
of using a threshold level before administering decompensation, even without the additional
a transfusion do not apply in all parts of the stress of a true postpartum hemorrhage. Under
world and are particularly restrictive in terms of such circumstances, the loss of less than 500 ml
bled-out parturients in the developing world. of blood could represent a fatal insult in a
severely anemic woman. When the hemoglobin
concentration is < 8 g/l, compensatory mecha-
SINGLE-UNIT TRANSFUSIONS IN THE nisms fail, lactic acid accumulates and patients
BLED-OUT PATIENT OF THE become breathless at rest. Cardiac failure may
DEVELOPING WORLD occur when the hemoglobin concentration is
The only subgroup analyzed in the 2005 study < 4 g/l, especially with twin pregnancies or with
by Ma and collaborators1 consisted of orthope- splenomegaly.
dic patients, and the authors did not mention Based on available evidence, the single-unit
whether their analysis included pregnant transfusion should remain a viable therapeutic
women who experienced postpartum hemor- option in selected obstetric patients, especially
rhage or were anemic. In 1992, the World in the developing world, where many women
Health Organization (WHO) published a finish their pregnancies in moderate or severe
tabulation of its 1990 estimates of the global anemic states. Depending on a variety of
death burden from all forms of anemia. Women circumstances, such patients may die within
of reproductive age were determined to be at a relatively short time after a postpartum
60
50
40
Case fatality (%)
30
20
10
5 10 15
Hb (g/dl)
Case fatality with relation to maternal hemoglobin (g/dl)
Figure 1 An analysis of anemia and pregnancy-related maternal mortality. Modified from Brabin BJ,
Hakimi M, Pelletier D. J Nutr 2001;131:604S–14S
410
432
30 August 2006 14:25:12
Single-unit transfusion
hemorrhage. If they do survive and if they are References

fortunate enough to be transported to a place
1. Ma M, Eckert K, Ralley F, Chin-Yee I. A retro-
where transfusion is available along with other spective study evaluating single-unit red blood
resuscitative measures, a single-unit transfusion cell transfusions in reducing allogeneic blood
may make the difference between whether or exposure. Transfus Med 2005;15:307–12
not the woman can be considered sufficiently 2. Diethrich EB. Evaluation of blood transfusion
well to return home for aftercare. therapy. Transfusion 1965;5:82–8
Clearly, each case must be evaluated on its 3. Crispen JF. The single-unit transfusion. A con-
own merit. Sometimes, care-givers will be suffi- tinuing problem. Pa Med 1966;69:44–8
ciently satisfied with the effect of one unit that 4. Reece RL, Beckett RS. Epidemiology of single-
they may hold off a second planned transfusion, unit transfusion. A one-year experience in a com-
munity hospital. JAMA 1966;195:801–16
especially if they think that the patient has suffi-
5. Morton JH. Surgical transfusion practices, 1967.
ciently recovered from the insult of postpartum
Surgery 1969;65:407–16
hemorrhage. It is unlikely, however, that the 6. Grindon AJ, Tomasulo PA, Bergin JJ, Klein HG,
effect of single-unit transfusion in women who Miller JD, Mintz PD. The hospital transfusion
have sustained postpartum hemorrhage will committee. Guidelines for improving practice.
ever be examined in a randomized, controlled JAMA 1985;253:540–3
trial. In addition, it is highly unlikely that such 7. Domen RE. The single-unit transfusion. J Fla
a trial, if ever proposed to most western Med Assoc 1986;73:855–7
world institutional review boards, would be 8. Cass RM, Blumberg N. Single-unit blood trans-
considered ethical. fusion: doubtful dogma defeated. JAMA 1987;
257:628–9
9. Perioperative Red Cell Transfusion. NIH
SUMMARY Consensus Development Conference Consensus
Statement. Online 1988 June 27–29 [cited year
In summary, numerous studies have shown that month day];7(4), 1–19
single-unit transfusion can ameliorate symp- 10. Brunson ME, Alexander JW. Mechanisms of
toms of chronic anemia. Its value in obstetric transfusion-induced immunosuppression. Trans-
emergencies when the anemia is acute has never fusion 1990;30:651–8
been tested. In countries where resources are 11. Conrad ME, Knodell RG, Bradley EL Jr,
poor and the majority of women have anemia at Flannery EP, Ginsberg AL. Risk factors in trans-
the onset of their pregnancies, even the slightest mission of non-A, non-B posttransfusion hepati-
deviation from normality during labor and tis. The role of hepatitis B antibody in donor
delivery may lead to excessive obstetric hemor- blood. Transfusion 1977;17:579–85
12. Davies SC, Brozovic M. ABC of transfusion.
rhage that can put a woman’s life at risk. In
Transfusion of red cells. Br Med J 1990;300:
these cases, after stabilizing the patient with 248–52
whatever resuscitative measures are available, 13. Savarit D, De Cock KM, Schutz R, Konate S,
transfer to a suitable center would be ideal. Lackritz E, Bondurand A. Risk of HIV infection
There, blood should be given, preferably typed from transfusion with blood negative for HIV
and cross-matched and screened for infections. antibody in a west African city. Br Med J 1992;
Unfortunately, such an eventuality is rare in the 305:498–502
developing world, and patients, if they survive, 14. Weiskopf RB. Do we know when to transfuse red
arrive at the referral center in a moribund state. cells to treat acute anemia? Transfusion 1998;38:
Some of these women, especially those who are 517–21
younger and have stopped bleeding, may bene- 15. Lenfant C. Transfusion practice should be
audited for both undertransfusion and over-
fit from a single unit of blood along with other
transfusion. Transfusion 1992;32:873–4
appropriate therapeutic measures. If, in the 16. Valeri CR, Crowley JP, Loscalzo J. The red cell
opinion of the clinician, this appears to be a rea- transfusion trigger: has a sin of commission now
sonable course of action, then the single unit become a sin of omission? Transfusion 1998;38:
should not be denied simply in order to comply 602–10
with what presently appears to be an outdated 17. Hebert PC, Wells G, Blajchman MA, et al.
dictum24. A multicenter, randomized, controlled clinical
411
433
30 August 2006 14:25:12
trial of transfusion requirements in critical care. 20. Rao MP, Boralessa H, Morgan C, et al. Blood
Transfusion Requirements in Critical Care component use in critically ill patients. Anaesthe-
Investigators, Canadian Critical Care Trials sia 2002;57:530–4
Group. N Engl J Med 1999;340:409–17 21. Vincent JL, Baron JF, Reinhart K, et al. Anemia
18. Chohan SS, McArdle F, McClelland DB, and blood transfusion in critically ill patients.
Mackenzie SJ, Walsh TS. Red cell transfusion JAMA 2002;288:1499–507
practice following the transfusion requirements 22. Hardy JF. Current status of transfusion triggers
in critical care (TRICC) study: prospective for red blood cell concentrates. Transfus Apher
observational cohort study in a large UK inten- Sci 2004;31:55–66
sive care unit. Vox Sang 2003;84:211–18 23. Murray CJ, Lopez AD. Global mortality, disabil-
19. French CJ, Bellomo R, Finfer SR, Lipman J, ity, and the contribution of risk factors: Global
Chapman M, Boyce NW. Appropriateness of Burden of Disease Study. Lancet 1997;349:
red blood cell transfusion in Australasian 1436–42
intensive care practice. Med J Aust 2002;177: 24. Mujeeb SA. Single unit blood transfusion, a bad
548–51 clinical practice? Transfus Today 1998;36:5–7
412
434
30 August 2006 14:25:12
46
OUT-OF-HOSPITAL DELIVERIES
E. Sheiner, I. Ohel and A. Hadar
INTRODUCTION when bleeding is slow and steady or in the pres-

ence of concomitant intra-abdominal bleeding18.
Out-of-hospital deliveries can be divided into
Moreover, the clinical signs of blood loss, such
planned, generally in a prepared setting and
as decrease in blood pressure and increased
attended by medical personnel, and
heart rate, tend to appear late, and only when
unplanned1. Unplanned out-of-hospital deliv-
the amount of blood loss reaches 1500 ml,
ery or delivery en route to the hospital occurs
mainly due to the high blood volume of preg-
when the woman is entering the active phase of
nant women (see Chapter 4 on assessing loss).
labor rapidly; it can be a stressful and sometimes
Recently, our group performed a large
even hazardous experience. Unplanned out-of-
population-based study of risk factors for early
hospital deliveries also carry an increased risk
postpartum hemorrhage16. Although this was
for adverse maternal and perinatal outcomes
not the first such study19–22, we were stimulated
and specifically hemorrhage and perinatal
to characterize women at risk that warrant spe-
mortality2–14.
cial attention after birth and consultation about
Bateman and colleagues3 reported that
the advisability of out-of-hospital delivery. Early
patients who delivered out of hospital in the
postpartum hemorrhage complicated 0.43%
USA were more likely to be African-American,
(n = 666) of all singleton deliveries included in
multigravid and to have had no or poor prenatal
this study (n = 154 311). Independent risk
care. Similarly, other ethnic minorities (Asians
factors for early postpartum hemorrhage,
living a long way from the hospital in Europe)
which can be of major importance during out-
are also at risk for out-of-hospital deliveries and
of-hospital deliveries, are presented in Table 1.
for adverse pregnancy outcome4–6. Indeed,
These risk factors were drawn from a multi-
reducing out-of-hospital deliveries has been
variate analysis and included retained placenta,
cited as an explicit health goal in the USA15.
labor dystocia, placenta accreta, severe lacera-
The two major forms of postpartum hemor-
tions, large-for-gestational age newborn and
rhage are early postpartum hemorrhage, which
hypertensive disorders16.
occurs within 24 h from delivery, and late
One of the largest studies regarding out-of-
postpartum hemorrhage, which takes place
hospital deliveries derives from a tertiary medical
between 24 h and 6 weeks after delivery16. Defi-
center, located in the Negev region, in Israel12,13.
nitions of each class are provided in Chapter 2;
In this area, most deliveries occur in the hospi-
however, the validity of the definitions has not
tal, and virtually all newborns and their mothers
been shown by rigorous evaluation and the pro-
are brought to the hospital, when delivered
cesses of estimation of blood loss are suspected
outside. This is done mainly because hospital
(see Chapters 5 and 6).
deliveries are entitled to a birth payment from
In one often-quoted article, approximately
the government, which is also given to new-
5% of all women who underwent vaginal deliv-
borns who are brought within 24 h to the hos-
ery without complications lost more than 1000
pital. The incidence of unplanned, accidental
ml of blood17. Having acknowledged this, it is
out-of-hospital deliveries was 2% (2328/
generally agreed that the objective evaluation of
114 938). Perinatal mortality was significantly
bleeding after labor may be difficult, specifically
higher among out-of-hospital deliveries (odds
413
435
30 August 2006 14:25:13
ratio (OR) 2.01, 95% confidence interval (CI) and remote regions of developing countries,
1.4–2.9), as compared to in-hospital deliveries. out-of-hospital deliveries occur mainly due to
Parturients who gave birth out of hospital had limited access to health services. Often, there
higher rates of perineal tears and retained pla- is limited access not only to referral health
centa, as compared to patients delivered in hos- facilities, but even to basic life-saving measures
pital (Table 2). In addition, patients delivered within the home and community. Such out-of-
out of hospital had a higher rate of delayed dis- hospital deliveries are associated with high rates
charge from hospital, as compared to controls. of perinatal morbidity and mortality2–14,23.
In a report by the Pan American Health
Organization (PAHO), currently, 79% of deliv-
GLOBAL RATES OF
eries in the Region of the Americas take place
OUT-OF-HOSPITAL DELIVERIES
in institutional settings, with only a few
The number of out-of-hospital deliveries in countries in the Region reporting institutional
the world is not well documented (Table 3). It deliveries below 50%24. Unfortunately, the
is important to distinguish between accidental trend of increasing institutional deliveries in
out-of-hospital deliveries and those intended the Americas has not resulted in a correspond-
and planned to take place out of hospital, with ing decrease in maternal and perinatal mortal-
the attendance of medical personnel. In rural ity. In fact, there are even greater variations
Table 1 Independent risk factors for early Table 3 Rates of planned and unplanned
postpartum hemorrhage, which can be of major out-of-hospital deliveries in the world
importance during out-of-hospital deliveries. Results
from a multiple logistic regression model. Data are Country Reference Rate (%)
expressed as odds ratio, 95% confidence interval
(CI) and p values for statistical significance. Adapted Planned out-of-hospital deliveries
from Sheiner et al. J Matern Fetal Neonatal Med United States and Canada Johnson et al.29 1
2005;18:149–5416 Netherlands Anthony et al.32 33
Home births in developing countries
Odds 95%
ratio CI p Ethiopia-southern Sibley et al.33 90
India Kodkany34 50
Retained placenta 3.5 2.1–5.8 < 0.001
Unplanned out-of-hospital deliveries
Labor dystocia, second stage 3.4 2.4–4.7 < 0.001
Israel, Negev region Sheiner et al.13 2
Placenta accreta 3.3 1.7–6.4 < 0.001
UK Scott et al.27 0.3
Lacerations 2.4 2.0–2.8 < 0.001
Finland Viisainen et al.7 0.1
Large for gestational age 1.9 2.4–1.6 < 0.001
Scotland catchment Rodie et al.28 0.6
Hypertensive disorders 1.6 2.1–1.2 < 0.001
Table 2 Pregnancy and labor complications of patients delivered out of hospital in comparison to patients
delivered in hospital. Adapted from Sheiner et al. J Reprod Med 2002;47:62–3013
Out of hospital In hospital

Characteristics (n = 2328) (n = 114 938) p
Lack of prenatal care 809 (34.8%) 10 822 (9.4%) < 0.001

Perineal tear grade 1–2 435 (18.7%) 16 178 (14.1%) < 0.001
Perineal tear grade 3–4 4 (0.2%) 77 (0.1%) < 0.056
Retained placenta 27 (1.2%) 693 (0.6%) < 0.001
Small for gestational age 233 (10.0%) 6 809 (5.9%) < 0.001
Large for gestational age 145 (6.2%) 11 774 (10.2%) < 0.001
Perinatal mortality 29 (1.2%) 718 (0.6%) < 0.001
Delayed discharge from hospital 911 (39.7%) 35 343 (31.1%) < 0.001
414
436
30 August 2006 14:25:13
Out-of-hospital deliveries
in neonatal and maternal mortality among the United States and Canada during the year
countries with high rates of institutional 2000 were assessed. The rate of planned home
delivery. This is perhaps due to unnecessary delivery was 1.6%29.
interventions, such as Cesarean section and In the Netherlands, approximately one-third
episiotomy, which may lead to increased of births are planned home deliveries, attended
morbidity and even mortality25,26. Efforts are by midwives. In this cross-sectional study,
being made in order to promote evidence-based maternal demographics associated with home
interventions in these countries24. birth included multiparity, age above 25 years
In a few reports from developed countries, and living in small as opposed to large cities32.
the incidence of accidental out-of-hospital The condition is quite different in undevel-
deliveries varied from 0.1 to 2%7,13,27–29. Fac- oped countries. In these areas, home birth with
tors associated with accidental out-of-hospital unskilled attendants is the norm, and maternal
deliveries include multiparity and lack of prena- and neonatal mortality rates are high. Unfortu-
tal care, which by themselves might increase the nately, the rates and outcomes of these out-of-
risk for adverse perinatal outcome30,31. hospital births are grossly underreported. The
In a large population-based study in the causes for this situation include inadequate
Negev region in Israel, the incidence of acciden- emergency care and home-based care by atten-
tal out-of-hospital deliveries was 2% (2328/ dants who are poorly equipped or educated to
114 938)13. These deliveries were described respond to emergencies, leading to inappropri-
as unattended, as opposed to deliveries that ate or delayed action. For example, in rural
were out of hospital but attended by skilled southern Ethiopia, over 90% of births take
personnel. place at home in the presence of unskilled
A report from a District General Hospital in attendants33.
the UK indicated a low incidence of 0.31% of In most parts of the world, postpartum
unplanned out-of-hospital deliveries occurring hemorrhage accounts for 35–55% of maternal
over a 3-year period27. Women were multi- deaths. In India, maternal mortality rates are
parous, and 11 of 14 deliveries (78.6%) estimated at 560/100 000 live births. In rural
occurred during the night, between the hours India, at least 50% of births occur at home34.
of 20.00 and 08.00, suggesting difficulties in These figures are reported in a recent study
access to the hospital. which presents a design for a randomized,
In a study from Finland, a trend was found placebo-controlled, clinical trial conducted in
towards a decrease in accidental out-of-hospital four primary health center areas of Belgaum
deliveries between 1963 and 1973: from 1.3 to District, Karnataka, India. The main goal of
0.4 per 1000 births. Nevertheless, this trend was this study was to assess the effectiveness of
changed by the 1990s when the figures rose up misoprostol 600 µg orally in reducing the
to 1/1000. This change was attributed to the incidence of acute postpartum hemorrhage.
closing of small hospitals in remote parts of Misoprostol would be administrated by mini-
the country, leading to inconvenient access to mally trained midwives to women randomized
perinatal facilities7. to receive misoprostol or placebo immediately
In a retrospective case-control study, women post-delivery of the infant. A secondary goal of
who delivered accidentally out of hospital in the this study was to test the international and com-
catchment area in Scotland during a given study munity collaborations necessary for the conduct
period were identified. Of all deliveries, 117 of this study, so that it could serve as a model
women delivered 121 babies accidentally out of for future studies in rural settings throughout
hospital. The rate was 0.6% of all deliveries the developing world for reducing maternal
registered at the hospital28. mortality and morbidity.
Examples for planned home births are found In conclusion, the number of out-of-hospital
in the following two studies. In a prospective deliveries in the world is not well documented.
study designed to evaluate the safety of home Although it is widely accepted that the quality of
births in North America, all home births maternity care is a main determinant of mater-
involving certified professional midwives across nal and fetal morbidity and mortality rates35,
415
437
30 August 2006 14:25:13
the lack of statistical information on out-of- Anemia is rarely detected or treated during
hospital deliveries is a severe limitation for pregnancy and often exacerbated by malarial
further evaluation of the relationship between and other parasitic diseases39.
out-of-hospital deliveries and maternal morbid- Although the vast majority of patients with
ity and mortality in general and specifically postpartum hemorrhage have no identifiable
postpartum hemorrhage. risk factor, young age at marriage40,41 and low
contraceptive use among many women in the
developing world result in high total fertility
OUT-OF-HOSPITAL DELIVERY AND
rates, which result in more grand multiparous
women giving birth in low-resource countries
Our group14 compared maternal and neonatal compared with more developed countries42.
outcomes in out-of-hospital vs. in-hospital A retrospective study from Ghana compared
deliveries in a prospective study. Unplanned active vs. expectant management in a rural
out-of-hospital deliveries resulted in a statisti- setting at Holy Family Hospital in Berekum43.
cally significant higher rate of postpartum The study found that postpartum hemorrhage
hemorrhage (OR 8.4; 95% CI 1.1–181.1; (blood loss = 500 ml) occurred less often in
p = 0.018) (Table 4). the actively managed group (OR 0.8; 95%
Postpartum hemorrhage due to uterine atony CI 0.7–0.9). McCormick and colleagues44
is the primary direct cause35 of maternal published a systematic review of studies that
mortality globally. Management strategies in assessed the efficacy of active management of
developed countries involve crystalloid fluid the third stage in low-resource settings. Active
replacement, blood transfusions, and surgery. management of the third stage of labor, espe-
Such definitive therapies are often not accessi- cially the administration of uterotonic drugs,
ble in developing countries, particularly in reduces the risk of postpartum hemorrhage due
out-of-hospital deliveries. The lack of skilled to uterine atony without increasing the inci-
attendants at delivery who can provide even the dence of retained placenta or other serious
minimum of care, long transport times to facili- complications. Oxytocin is preferred over
ties that can manage uterine atony or severe lac- syntometrine, but misoprostol also can be
erations of the genital tract, and unattended used to prevent hemorrhage in situations
obstructed labor leading to a ruptured uterus, where parenteral medications are not available
elevate postpartum hemorrhage to its position (see Chapters 16–19).
as the number one killer of women during child A 2003 Cochrane Review of active versus
birth36. These factors are exacerbated by the expectant management of the third stage of
prevalence of anemia, which is estimated to labor45 included five randomized, controlled tri-
affect half of all pregnant women in the world37, als and found that, for all women, including
with this figure rising to 94% in Papua New women deemed to be at low risk for postpartum
Guinea, 88% in India, and 86% in Tanzania38. hemorrhage, active management decreased the
Table 4 Maternal outcomes of patients with unplanned out-of-hospital deliveries and the control group.
Adapted from Hadar et al. J Reprod Med 2005;50:832–614
Unplanned out-of-hospital deliveries Control group

Characteristics (n = 151) (n = 151) p
Vaginal tears 27 (17.9%) 18 (12.0%) 0.087

Postpartum hemorrhage 8 (5.3%) 1 (0.6%) 0.018
Postpartum endometritis 2 (1.3%) 0 0.157
Antibiotic treatment 2 (1.3%) 0 0.157
Sutures of vaginal tears 25 (16.6%) 18 (12.0%) 0.249
Revision of uterus cavity 6 (4.0%) 0 0.013
Hospitalization (days) 3.2 ± 0.9 2.95 ± 0.6 0.111
416
438
30 August 2006 14:25:14
incidence of postpartum hemorrhage (both of morbidities because many hospital deliveries

500–1000 ml and > 1000 ml), shortened the (which may have a higher proportion of
third stage of labor, decreased the amount problems) were not studied.
of maternal blood loss, decreased the need for Another randomized, controlled trial was
blood transfusion, and decreased the need for carried out to determine whether suckling
additional therapeutic uterotonic agents. The immediately after birth reduces the frequency
incidence of postpartum hemorrhage of 500 ml of postpartum hemorrhage47. The trial partici-
or more was reduced in the actively managed pants were attended by traditional birth atten-
group (relative risk 0.38; 95% CI 0.32–0.46). dants. Traditional birth attendants recorded
These figures mean that for every 12 women blood loss in the early suckling group and in the
who are actively managed rather than expec- control group in 2104 and 2123 deliveries of
tantly managed, one case of postpartum hemor- liveborn singletons, respectively. The frequency
rhage (defined as blood loss = 500 ml) will be of postpartum hemorrhage (loss greater than
averted, whereas the number needed to treat for 500 ml) was 7.9% in the suckling group and
averting blood loss greater than 1000 ml would 8.4% in the control group. Prual and colleagues
be 57. Women who were actively managed lost reported the frequency of such morbidity as
less blood, with a weighted mean blood loss of revealed in a population-based survey of a
79.33 ml less than those who were expectantly cohort of 20 326 pregnant women in six West
managed. The third stage was an estimated African countries48. The main direct causes of
9.77 min shorter in actively managed women. severe maternal morbidity were hemorrhage
The use of routine uterotonic agents to prevent (3.05 per 100 live births); in this report 23 cases
postpartum hemorrhage can reduce maternal involved uterine rupture (0.12 per 100). Case
mortality by 40%46. fatality rates were high for hemorrhage and var-
The data on the types and incidences of ied from 1.9% for antepartum or peripartum
maternal morbidities in communities with hemorrhage to 3.7% for placental abruption.
limited access to health services are scarce23. The high case fatality rates of several complica-
Bang and colleagues found, in their prospective tions reflected a poor quality of obstetric care.
observational study in Gadchiroli, India, Walraven and colleagues49, in their double-
that the incidence of maternal morbidity was blind, randomized, controlled trial, sought to
52.6%, 17.7% during labor and 42.9% during evaluate the impact of oral misoprostol on post-
puerperium. The most common intrapartum partum hemorrhage compared with standard
morbidities were prolonged labor (10.1%), treatment in the home birth situation in rural
prolonged rupture of membranes (5.7%), Gambia, with measured blood loss, postpartum
abnormal presentation (4.0%) and primary hemoglobin, change in hemoglobin level
postpartum hemorrhage (3.2%). The post- between the last antenatal care visit and 3–5
partum morbidities included secondary days postpartum as outcome measures. The
postpartum hemorrhage (15.2%). In their study was carried out in 26 primary health-care
study, mothers and the neonates were prospec- villages of the North Bank East Health Division,
tively observed at home in 39 villages without The Gambia, West Africa. Seventy-two percent
interventions. It included a population of of births occur at home and maternal mortality
approximately one million parturients. Most in the study area has been estimated at
deliveries in the area were conducted by 424/100 000 live births in a reproductive age
traditional birth attendants and family mem- mortality survey, with postpartum hemorrhage
bers. This is the first reported study in a rural as the most important direct cause of maternal
setting in a developing country where labor and mortality. There were two maternal deaths in
the puerperium were prospectively observed at the study population (maternal mortality ratio
home in a systematic and objective manner to for study population of 163 per 100 000 live
measure the incidence of maternal morbidities. births; 95% Poisson CI 20–595), both in
This study had certain limitations which should the misoprostol group. These deaths were
be kept in mind while extrapolating the results. attributed to postpartum hemorrhage (mea-
Their sample may underestimate the incidence sured blood loss 2200 ml) and disseminated
417
439
30 August 2006 14:25:14
Table 5 The risk for postpartum hemorrhage in 4. Hinds MW, Bergeisen GH, Allen DT. Neonatal
out-of-hospital deliveries outcome in planned vs. unplanned out-of-
hospital births in Kentucky. JAMA 1985;253:
Postpartum 1578–82
hemorrhage in out- 5. Goldenberg RL, Hale CB, Houde J,
Country Reference of-hospital deliveries Humphrey JL, Wayne JB, Boyd BW. Neonatal
deaths in Alabama. III. Out-of-hospital births,
West Africa Prual et al.48 3.1% 1940–1980. Am J Obstet Gynecol 1983;147:
Malawi Bullough et al.47 8.4% 687–93
Ghana Geelhoed et al.7 17.4% 6. Bhoopalam PS, Watkinson M. Babies born
India Bang et al.23 3.2% before arrival at hospital. Br J Obstet Gynaecol
Israel Hadar et al.14 5.3% 1991;98:57–64
Israel Sheiner et al.13 3.2% 7. Viisainen K, Gissler M, Hartikainen AL,
Hemminki E. Accidental out-of-hospital
births in Finland: incidence and geographical
intravascular coagulation due to malaria (mea- distribution 1963–1995. Acta Obstet Gynecol
sured blood loss 300 ml), respectively. Scand 1999;78:372–8
Table 5 summarizes the existing, limited 8. Moscovitz HC, Magriples U, Keissling M,
data regarding the association between out-of- Schriver JA. Care and outcome of out-of-
hospital deliveries and postpartum hemorrhage. hospital deliveries. Acad Emerg Med 2000;7:
757–61
9. Verdile VP, Tutsock G, Paris PM, Kennedy RA.
CONCLUSIONS Out-of-hospital deliveries: a five-year experience.
Prehospital Disaster Med 1995;10:10–13
The fact that so many women deliver in domi- 10. Chen CC, Huang CB, Chung MY. Unexpected
ciliary conditions clearly affects their risk of delivery before arrival at hospital: an observation
having postpartum hemorrhage. It is widely of 18 cases. Changgeng Yi Xue Za Zhi 2000;23:
accepted that the quality of maternity care is a 205–10
main determinant of maternal mortality rates. 11. Walraven GE, Mkanje RJ, Roosmalen J, van
Our research has, for the first time, established Dongen PW, Dolmans WM. Perinatal mortality
an odds ratio of 8.414. This number represents in home births in rural Tanzania. Eur J Obstet
an urgent call to the medical community to Gynecol Reprod Biol 1995;58:131–4
change this circumstance whenever possible, as 12. Sheiner E, Hershkovitz R, Shoham-vardi I, Erez
O, Hadar A, Mazor M. A retrospective study
is detailed in other chapters of this book. All
of unplanned out-of-hospital deliveries. Arch
births should be attended by adequately trained
Gynecol Obstet 2004;269:85–8
personnel. More effective strategies are needed 13. Sheiner E, Shoham-vardi I, Hadar A, Sheiner
to convince women with high-risk pregnancies EK, Hershkovitz R, Mazor M. Accidental out-
to deliver in a hospital which has access to of-hospital delivery as an independent risk factor
emergency referral services. for perinatal mortality. J Reprod Med 2002;47:
625–30
14. Hadar A, Rabinovich A, Sheiner E, Landau D,
References Hallak M, Mazor M. Obstetrics characteristics
1. Zur M, Hadar A, Sheiner E, Mazor M. Out of and neonatal outcome of unplanned out of hos-
hospital deliveries: incidence, obstetrical charac- pital term deliveries: a prospective case–control
teristics and perinatal outcome. Harefuah 2003; study. J Reprod Med 2005;50:832–6
142:38–41 15. National Center for Health Statistics. Prevention
2. Burnett CA 3rd, Jones JA, Rooks J, Chen CH, profile, health, United States, 1989. Hyattsville,
Tyler CW Jr, Miller CA. Home delivery and neo- Md: US Public Health Services, 1990:35
natal mortality in North Carolina. JAMA 1980; 16. Sheiner E, Sarid L, Levy A, Seidman DS,
244:2741–5 Hallak M. Obstetric risk factors and outcome
3. Bateman DA, O’Bryan L, Nicholas SW, of pregnancies complicated with early post-
Heagarty MC. Outcome of unattended out-of- partum hemorrhage: A population-based study.
hospital births in Harlem. Arch Pediatr Adolesc J Matern Fetal Neonatal Med 2005;18:
Med 1994;148:147–52 149–54
418
440
30 August 2006 14:25:14
17. Pritchard JD, Baidwin RM, Dickey JC, Wiggins 31. Sheiner E, Hallak M, Twizer E, Mazor M,
KM. Blood volume changes in pregnancy and Katz M, Shoham-Vardi I. Lack of prenatal care
the puerperium. II. Red blood cell loss and in two different societies living in the same region
changes in apparent blood volume during and sharing the same medical facilities. J Obstet
and following vaginal delivery, cesarean section, Gynaecol 2001;21:453–8
and cesarean section plus total hysterectomy. Am 32. Anthony S, Buitendijk SE, Offerhaus PM,
J Obstet Gynecol 1962;84:1271–82 Dommelen P, Pal-de Bruin KM.Maternal
18. Norris CT. Management of postpartum factors and the probability of a planned home
hemorrhage. Am Fam Physician 1997;55: birth. Br J Obstet Gynaecol 2005;112:748–53
635–40 33. Sibley L, Buffington ST, Haileyesus D.The
19. Combs CA, Murphy EL, Laros RK. Factors American College of Nurse-Midwives’ Home-
associated with postpartum hemorrhage with Based Lifesaving Skills Program: A Review of the
vaginal birth. Obstet Gynecol 1991;77:69–76 Ethiopia Field Test. J Midwifery Womens Health
20. Combs CA, Murphy EL, Laros RK. Factors 2004;49:320–8
associated with hemorrhage in cesarean deliver- 34. Kodkany BS, Derman RJ, Goudar SS, et al. Initi-
ies. Obstet Gynecol 1991;77:77–82 ating a novel therapy in preventing postpartum
21. Magann EF, Evans S, Hutchinson M, Collins R, hemorrhage in rural India: a joint collaboration
Howard BC, Morrison JC. Postpartum hemor- between the United States and India. Int J Fertil
rhage after vaginal birth: an analysis of risk Womens Med 2004;49:91–6
factors. South Med J 2005;98:419–22 35. World Health Organization. Mother–Baby Pack-
22. Magann EF, Evans S, Chauhan SP, Lanneau G, age (WHO/RHT/MSM/94.11, Rev1). Geneva:
Fisk AD, Morrison JC. The length of the third World Health Organization, 1998
stage of labor and the risk of postpartum 36. Miller S, Lester F, Hensleigh P. Prevention
hemorrhage. Obstet Gynecol 2005;105:290–3 and treatment of postpartum hemorrhage: new
23. Bang RA, Bang AT, Reddy MH, Deshmukh advances for low-resource settings. J Midwifery
MD, Baitule SB, Filippi V.Maternal morbidity Womens Health 2004;49:283–92
during labour and the puerperium in rural homes 37. Levy A, Fraser D, Katz M, Mazor M, Sheiner E.
and the need for medical attention: A prospective Maternal anemia during pregnancy is an
observational study in Gadchiroli, India. Br J independent risk factor for low birthweight and
Obstet Gynaecol 2004;111:231–8 preterm delivery. Eur J Obstet Gynecol Reprod Biol
24. Health in the Americas, 2002 edn, Vol I. Pan 2005;122:182–6
American Health Organization 38. Brabin BJ, Hakimi M, Pelletier D. An analysis
25. Lydon-Rochelle M, Holt VL, Martin DP, of anemia and pregnancy-related maternal
Easterling TR. Association between method of mortality. J Nutr 2001;131:604–15S
delivery and maternal rehospitalization. JAMA 39. Steketee RW. Pregnancy, nutrition and parasitic
2000;283:2411–16 diseases. J Nutr 2003;133:1661–7S
26. Srp B, Velebil P. Proportion of caesarean 40. Mahler K, Rosoff JI. Into a New World: Young
sections and main causes of maternal mortality Women’s Sexual and Reproductive Lives. New
during 1978–1997 in the Czech Republic. Ceska York: Alan Guttmacher Institute, 1998
Gynekol 1999;64:219–23 41. Miller S, Lester F. Married young first time
27. Scott T, Esen UI. Unplanned out of hospital mothers: Meeting their special needs. Proceedings
births – who delivers the babies? Ir Med J 2005; of the WHO/Population Council Technical
98:70–2 Meeting on Married Adolescents. Geneva,
28. Rodie VA, Thomson AJ, Norman JE. Accidental Switzerland, December 9–12, 2003
out-of-hospital deliveries: an obstetric and 42. Trussell J, Pebley AR. The potential impact of
neonatal case control study. Acta Obstet Gynecol changes in fertility on infant, child, and maternal
Scand 2002;81:50–4 mortality. Stud Fam Plann 1984;15:267–80
29. Johnson KC, Daviss BA. Outcomes of planned 43. Geelhoed D, Visser L, Agordzo P, et al. Active
home births with certified professional midwives: versus expectant management of the third stage
large prospective study in North America. Br of labor in rural Ghana. Acta Obstet Gynecol
Med J 2005;330:1416 Scand 2002;81:171–3
30. Twizer E, Sheiner E, Hallak M, Mazor M, Katz 44. McCormick ML, Sanghvi HC, Kinzie
M, Shoham-Vardi I. Lack of prenatal care in B, McIntosh N. Preventing postpartum
a traditional society: is it an obstetrical hazard? hemorrhage in low-resource settings. Int J
J Reprod Med 2001;46:662–8 Gynaecol Obstet 2002;77:267–75
419
441
30 August 2006 14:25:14
45. Prendiville WJ, Elbourne D, McDonald S. trial in deliveries by traditional birth attendants.
Active versus expectant management in the Lancet 1989;2:522–5
third stage of labour (Cochrane Review). In 48. Prual A, Bouvier-Colle MH, de Bernis L,
The Cochrane Library. Chichester: John Wiley & Breart G. Severe maternal morbidity from direct
Sons, 2003, Issue 4 obstetric causes in West Africa: incidence and
46. Prendiville WJ, Harding JE, Elbourne DR, case fatality rates. Bull WHO 2000;78:593–602
Stirrat GM. The Bristol third stage trial: active 49. Walraven G, Blum J, Dampha Y, et al.
versus physiological management of third stage Misoprostol in the management of the third
of labour. Br Med J 1988;297:1295–300 stage of labour in the home delivery setting in
47. Bullough CH, Msuku RS, Karonde L. Early rural Gambia: a randomized controlled trial. Br J
suckling and postpartum hemorrhage: controlled Obstet Gynaecol 2005;112: 1277–83
420
442
30 August 2006 14:25:15
47
INTRAOPERATIVE AUTOLOGOUS BLOOD TRANSFUSION
S. Catling and D. Thomas
INTRODUCTION particular, its evolving use in obstetrics with

direct reference to postpartum hemorrhage.
Life-saving transfusion using human blood was
first described by James Blundell in 1818. He
performed ten transfusions, five of which were DEFINITION
successful; of these, four were in women suffer-
ing postpartum hemorrhage. He typically used Autologous blood salvage is the collection of
donor blood from the patient’s husband, thus spilt blood resulting from surgical or traumatic
showing that the technique of blood injection bleeding that can be undertaken intra-
with a syringe-infusion was safe1. In one operatively or postoperatively. The collected
account, he is credited with the re-infusion of blood can be filtered and re-infused or filtered,
autologous blood2. It is entirely appropriate, washed and then re-infused.
therefore, that the subject of intraoperative
autologous transfusion be described in this text-
METHODS
book on postpartum hemorrhage. Blundell’s
original report described a reasonable outcome The quality and constitution of re-infused blood
considering the crude understanding of blood vary depending on whether unwashed or
transfusion techniques in existence almost a washed systems are used. In the absence of
century before Landsteiner’s identification of automated cell-washing devices, simple
the ABO blood groups3. collection, filtration and re-infusion during
Some of the earliest reports of intraoperative postpartum hemorrhage have been described
blood salvage described the life-saving tech- and continue to be used in some areas in the
nique of simply collecting spilt blood from the world. This technique is not ideal. However, the
abdominal cavity, filtering it through a gauze use of unwashed blood (particularly for postop-
swab, and re-infusing what remained. In the erative collection and re-infusion using a sealed
ensuing years, techniques to collect, filter and postoperative collection unit with a filter) has
wash blood lost at the time of surgery have been used extensively in total knee surgery
become commonplace, although refinements of and seems safe and effective. It is interesting
the method vary widely and depend not only that a recent report suggested that the use
upon the nature of the surgical procedure but of unwashed blood might have properties that
also the availability of technical resources. improve the recipient’s immune response4.
As might be expected, expensive apheresis The more widely applied intraoperative cell
machines are lacking in many, if not most parts salvage is conducted with an apparatus that
of the world, where operative obstetrics is rou- has the ability to collect spilt blood at the time
tine. Nevertheless, the problem of postpartum of operation and anticoagulate it at the tip of
hemorrhage is so common and remains such a the suction apparatus with citrated solution
clinical challenge that, in these circumstances, or heparinized saline (25 000 IU per liter of
the technique as originally described is still used normal saline) (Figure 1). Collected blood is
out of necessity. This chapter describes various then transferred to a centrifugal bowl, where
methods of autologous blood salvage and, in spinning at 5500 revolutions per second moves
421
443
30 August 2006 14:25:15
Figure 1 A diagrammatic representation of intraoperative cell salvage. (Adapted from an original drawing
with the kind permission of Haemonetics Inc., Baintree). The dotted line represents infusion sent back to
the patient. In the case of a Jehovah’s Witness, this is primed with saline before starting to complete
continuity of the circuit
the heavier red cells to the outer periphery of cases of massive hemorrhage, the cell salvage
the bowl. As the bowl fills, the accumulation of devices help to recycle transfused allogeneic
red cells forces the plasma, platelets and other blood as well as autologous blood.
cellular debris out of a central exit, discarding As few platelets and minimal clotting factors
waste products of the process. Special sensors are present in these re-infused red cells,
identify when the bowl is full of red cells, and careful assessment of coagulation parameters is
the fully automated machine begins to wash the required in cases of excessive bleeding where
collected and concentrated erythrocytes with massive transfusion is required. Nowadays, this
normal saline. This process further cleanses the is the case in patients with massive hemorrhage
red blood cells. The resultant concentrate is anyway, as the provision of red cells suspended
then suspended in normal saline, producing a in a mixture of saline, adenine, glucose and
solution with a hematocrit of 60%. Most of mannitol (SAGM) means that only packed
the platelets and clotting factors will have been allogeneic red cells are being infused, and so
washed away at this point, however, and the similar provisos apply. Early consideration
fluid for re-infusion consists of autologous red therefore needs to be given to platelet and fresh
cells suspended in normal saline5. frozen plasma administration.
In the presence of brisk bleeding, any of the
commercially available automated cell-washing
HISTORICAL COMPLICATIONS
devices can produce a unit of red cell concen-
trate in 5–10 min. The volume of lost blood that Current machines have an extremely good
can be processed is infinite, and reports of cell safety record, but it is worthwhile dispelling
salvage in major trauma describe its successful some misconceptions about the technique that
use, the process providing approximately 50% persist today. Air embolism is not a problem
of the required red cell transfusion6. Of course, with modern equipment when it is used cor-
in such situations, the use of cell salvage only rectly. Free hemoglobin is almost completely
minimizes the demand for allogeneic blood. In removed, and the very small amounts that
422
444
30 August 2006 14:25:19
Intraoperative autologous blood transfusion
remain have no significant clinical effect. Plate- SAFETY OF CELL SALVAGE IN

lets are activated during salvage, but the major- OBSTETRICS
ity are removed during the process. Leukocytes,
Two theoretical problems attend the use of cell
complement and kinins are also activated
salvage at the time of Cesarean section. First,
during salvage, but systemic inflammatory
in a Rh-negative mother, there is a risk of Rh
responses have not been reported as clinically
immunization if the fetus is Rh-positive. As the
relevant.
cell saver cannot distinguish fetal from adult red
cells, any fetal red cells suctioned from the oper-
POSSIBLE CONTRAINDICATIONS ative field will be processed and re-infused with
the maternal red cells. In practice, studies show
Following a seminal report7 supporting this
that the degree of contamination with fetal red
technology, it now is accepted that three areas
cells during cell salvage at Cesarean section is
exist where the process of red cell salvage needs
between 1 and 19 ml11–13. Applying the stan-
to be used with caution and following necessary
dard Kleihauer calculation, this would require
risk–benefit analysis, depending on the clinical
between 500 and 2500 units (1–5 ampules) of
urgency of the situation. These involve the use
Anti-D to avoid Rh immunization. As all
of red cell salvage when spilt operative blood
Rh-negative patients require Anti-D after
may contain malignant cells, or be heavily con-
Cesarean section, patients receiving salvaged
taminated with bowel bacteria. Another area of
blood may simply require an increased dose.
caution is the use of red cell salvage when con-
The second theoretical problem is contami-
taminated by amniotic fluid. It is accepted that,
nation with amniotic fluid, raising the specter of
in the presence of any of these preconditions,
iatrogenic amniotic fluid embolus (AFE). This
cell salvage is not used unless considered
theoretical complication has been investigated
necessary.
by several workers, and has not been found to
The non-availability of a safe allogeneic
be a problem in practice12–16. The difficulty is
blood supply is clearly a situation when the use
that the precise elements of amniotic fluid,
of cell salvage is justified in an attempt to pre-
which cause the rare, and unpredictable
serve the patient’s own blood and help oxygen
‘anaphylactoid syndrome of pregnancy’ (as
carriage In the UK, current blood conservation
AFE is more correctly called), remain unknown.
recommendations promote the use of cell sal-
To conduct a prospective, randomized, con-
vage8. The current drive for blood conservation
trolled trial with an 80% power to demonstrate
is multifactorial, but the most topical reason is
that cell salvage does not increase the incidence
the potential decrease in the availability of
of AFE by five-fold would require up to 275 000
donor blood resulting from the introduction of a
patients, a number so enormous that the effort
test for the presence of abnormal prion protein.
is unlikely ever to be undertaken. To demon-
However, reduced numbers of donors is a prob-
strate the absolute safety of a technique without
lem that had its inception prior to the present
randomized, controlled trials requires careful
testing concerns, as the presence of HIV and
clinical audit of a large number of cases,
other viral pathogens have also restricted the
supported by robust in vitro evidence.
number of potential donors.
It is against this backdrop that consideration
of cell salvage in postpartum hemorrhage was
IN VITRO STUDIES OF AMNIOTIC
made, and the remainder of this chapter exam-
FLUID CLEARANCE:
ines the use of intraoperative cell salvage during
postpartum hemorrhage. Fortunately, the wide- In vitro studies have examined the clearance
spread use of such devices has confirmed the of α-fetoprotein14, tissue factor15, trophoblastic
safety of this process, providing there is no tissue12, fetal squames and lamellar bodies13
technical failure and the correct procedure for from maternal blood by the cell salvage process.
machine operation is practiced. The use of such Small molecules are removed in the plasma frac-
devices is endorsed by national guidelines and tion by the centrifuge and wash process alone,
Government directives9,10. and particulate material is removed by the use
423
445
30 August 2006 14:25:19
of specialized leukodepletion filters. Using the patient who is seriously ill with HELLP syn-
combination of cell salvage and these special- drome and who refuses platelet and coagulation
ized filters, every element of amniotic fluid that factor transfusion is unlikely to survive, and
has been studied so far has been effectively that, under such circumstances, her death
removed from salvaged blood prior to should logically not be related to the use of cell
re-transfusion12–16. salvage, but rather to her refusal to accept blood
component therapy.
Cell salvage in obstetrics was introduced in
CLINICAL CASES
the UK in 1999, and its use is growing rapidly,
Prior to 1999, approximately 300 cases in which with most major obstetric units now advocating
cell-salvaged blood was administered to patients the technique in selected circumstances. The
had been reported world-wide16. No obstetric Confidential Enquiry into Maternal and Child
clinical or physiological problems were encoun- Health 2000–2002 (CEMACH)20 stated that
tered, despite the fact that filters were not used ‘. . . (cell salvage) may be used in any case of obstet-
at this time. This means that each of these ric haemorrhage, not just women who refuse blood
patients had some exposure to amniotic fluid, transfusion’ and described the technique as ‘a
and with no ill effects. Waters and colleagues new development which will prove helpful in the
shed some light on this topic13 by describing not future’. It further stated that ‘the risk of causing
only the complete clearance of squamous cells coagulopathy by returning amniotic fluid to the
and phospholipid lamellar bodies from filtered, circulation is thought to be small’. Subsequent
cell-salvaged blood, but also by clearly demon- to this, the 2005 revised Guidelines for Obstet-
strating the presence of both these amniotic ric Anaesthetic Services were published jointly
fluid markers circulating in the maternal central by the UK Obstetric Anaesthetists Association
venous blood at the time of placental separa- (OAA) and the Association of Anaesthetists of
tion. In 100% of patients in this trial, amniotic Great Britain and Ireland (AAGBI)21, stating
fluid was demonstrated in the circulation of that ‘an increasing shortage of blood and blood
healthy parturients undergoing elective Cesarean products and growing anxiety about the use of donor
section. It is therefore probable that amniotic blood are leading to an increasing interest in the
fluid routinely enters the maternal circulation use of cell salvage in obstetrics. Staff will have to
and does no harm in the vast majority of cases. be suitably trained, and equipment obtained and
This exposure may trigger the syndrome of AFE maintained. . .’
due to an anaphylactoid reaction to an as-yet In November 2005, the UK National Institute
unidentified endogenous mediator in a very for Clinical Excellence (NICE) reported on Cell
small number of women, the incidence of Salvage in Obstetrics22, describing cell salvage as
which varies between 1 in 8000 and 1 in 80 000 ‘an efficacious technique for blood replacement,
patients17. [Editor’s note: since it has never well established in other areas of medicine’ and
been studied, there is no evidence to state that pointing out the theoretical concerns when used
entry does not occur in an unknown number in obstetrics. NICE goes on to recommend that
of cases of vaginal parturition.] Clearly, re- clinicians using it in the UK should report any
infusion of cell-salvaged blood, even if contami- side-effects to the UK Department of Health
nated with traces of amniotic fluid, presents no Regulatory Authority (MHRA), that patients
extra risk to the woman from whom that blood should be fully informed prior to its use, and that
has come, as she has already been exposed to it. cell salvage in obstetrics should be performed by
In 1999, a single report appeared describing multidisciplinary teams that have developed
a seriously ill Jehovah’s Witness woman with regular experience in its use.
severe pre-eclampsia complicated by HELLP
syndrome (hemolysis, elevated liver enzymes,
PRACTICAL USE OF CELL SALVAGE
low platelets) who died in Holland, after having
IN OBSTETRICS
received cell-salvaged blood18. It has been
quoted as a ‘death due to obstetric cell sal- There presently exists a substantial experience
vage’19. It should be noted, however, that a with the use of cell salvage in obstetrics in the
424
446
30 August 2006 14:25:19
Intraoperative autologous blood transfusion
UK; cases include major hemorrhage due to should try to suction blood from ‘pools’
placenta previa, placenta accreta, ruptured rather than ‘dabbing’ tissue surfaces with
uterus, extrauterine placentation, massive the suction tip, as this minimizes erythro-
fibroids and placental abruption, as well as cyte damage.
routine use in Jehovah’s Witnesses to avoid
(6) Blood from swabs can be gently washed
postoperative anemia14.
with saline and salvaged from a sterile
The following guidelines are in use for cell
bowl into the main reservoir.
salvage in obstetric use in the Swansea NHS
Trust Hospitals, UK: (7) Suction pressure should be kept as low
as practicable (< 300 mmHg) to avoid
(1) It may be used for any situation in which red cell damage, although higher vacuum
allogeneic blood is used, but in practice can be safely used if necessary with
this has so far been confined to Cesarean only a minimum increase in red cell
sections and uterine re-exploration or damage.
laparotomy following postpartum hemor-
rhage. There is no reason why vaginal (8) It is advisable to use a leukocyte depletion
blood loss could not be collected and filter (Leukoguard RS Pall) in the re-
cell-salvaged, as fears about infection have transfusion circuit if there is any risk
proved unfounded in abdominal gunshot of amniotic fluid contamination. This is
wounds as long as the patients are on currently the only filter that has been
antibiotics – but the technical problem shown to remove all particulate elements
with physically collecting vaginal blood of amniotic fluid (fetal squames, lamellar
loss has yet to be solved! [Editor’s note: bodies). This filtration process will neces-
the routine and planned use of the sarily slow down the rate at which blood
BRASSS technique described in Chapter can be infused, but it is permissible to
4 would be useful to overcome this prob- pressurize the bag of salvaged red cells
lem as well as underestimation of loss.] up to 200 mmHg after having ensured
there is no air in the bag (otherwise it
(2) The machine is set up and operated may burst!), or to use a large-volume
according to standard operating proce- syringe and three-way tap. In situations
dure, with an ‘in-continuity’ set-up for when hemorrhage is rapid, it is possible
Jehovah’s Witnesses (this means that the to connect more than one suction
whole circuit is run through with saline nozzle to the reservoir, and two filters
and the re-transfusion bag connected to and a dual giving-set to the re-infusion
the intravenous cannula before starting bag.
the salvage suction, thereby establishing a
continuous circuit between the blood lost (9) As with any transfusion, the patient
and the recipient vein). should be carefully monitored, preferably
in an obstetric ‘critical care’ facility
(3) In cases where there is doubt about the for 24 h. Coagulation tests should be
extent of expected blood loss, it is eco- obtained post-transfusion, and repeated if
nomical to set up the aspiration and reser- abnormal or if clinically indicated.
voir kit only – the decision to process and
re-transfuse can be made when the degree (10) If the patient is Rh-negative, a Kleihauer–
of hemorrhage has become clear (e.g. Braun–Betke test should be performed
‘expected’ bleeding from placenta previa). and Anti-D administered as appropriate
within 72 h.
(4) Where practicable, amniotic fluid should
be removed by separate suction prior to Units that use obstetric cell salvage should
starting cell salvage. keep careful records for Audit reporting in
due course – with any problems also being
(5) Suction should be via the wide-bore reported to the MHRA as per NICE
suction nozzle in the kit, and the surgeon Guidelines.
425
447
30 August 2006 14:25:19
SUMMARY 11. Fong J, Gurewitsch ED, Kump L, Klein R.

Clearance of fetal products andsubsequent
The use of intraoperative cell salvage is a safe immunoreactivity of blood salvaged at Cesarean
method of conserving operative blood loss and delivery. Obstet Gynecol 1999;93:968–72
minimizing the need for allogeneic transfusion. 12. Catling SJ, Williams S, Fielding AM. Cell sal-
In an environment where allogeneic blood is in vage in obstetrics: an evaluation of the ability of
limited supply or the demands for blood trans- cell salvage combined with leucocyte depletion
fusion are so great, as in the case of massive filtration to remove amniotic fluid from operative
postpartum hemorrhage, the use of intra- blood loss at caesarean section. Int J Obstet
operative cell salvage may be life-saving and its Anesth 1999;8:79–84
13. Waters JH, Biscotti C, Potter PS, Phillipson E.
use in this area is gaining clinical acceptance.
Amniotic fluid removal during cell salvage in the
Cesarean section patient. Anaesthesiology 2000;
References 92:1531–6
14. Thornhill MI, O’Leary AJ, Lussos SA,
1. Blundell J. Experiments on the transfusion of Rutherford C, Johnson MD. An in vitro
blood by the syringe. Med Chirg Trans 1818;9: assessment of amniotic fluid removal from
57–92 human blood through cell saver processing.
2. Allen JG. Discussion. Ann Surg 1963;158:137 Anaesthesiology 1991;75:A830
3. Landsteiner K. Ueber Agglutinationser- 15. Bernstein HH, Rosenblatt MA, Gettes M,
scheinungen normalen menschlichen Blutes. Lockwood C. The ability of the Haemonetics
Wien Klin Wochenschr 1901;14:1132–4 4 cell saver to remove tissue factor from
4. Gharehbaghian A, Haque KM, Truman C, et al. blood contaminated with amniotic fluid. Anesth
Effect of autologous blood on postoperative Analgesia 1997;85:831–3
natural killer cell precursor frequency. Lancet 16. Catling SJ, Freites O, Krishnan S, Gibbs R.
2004;363: 1025–30 Clinical experience with cell salvage in obstetrics:
5. Tawes RL, Duvall TB. The basic concepts of an 4 cases from one UK centre. Int J Obstet Anesthes
autotransfusor: the cell saver. In Tawes RL, ed. 2002;11:128–34
Autotransfusion. Michigan: Gregory Appleton, 17. Morgan M. Amniotic fluid embolism. Anaesthe-
1997 sia 1979;34:20–32
6. Hughes LG, Thomas DW, Wareham K, et al. 18. Oei SG, Wingen CBM, Kerkkamp HEM
Intra-operative blood salvage in abdominal (letter). Int J Obstet Anesth 2000;9:143
trauma: a review of 5 years’ experience. Anaesthe- 19. Controversies in Obstetric Anaesthesia Meeting,
sia 2001;56:217–20 London UK, March 2004
7. Council on Scientific Affairs. Autologous blood 20. Confidential Enquiry into Maternal and Child
transfusions. JAMA 1986;256:2378–80 Health (CEMACH) 2000–2002. The 6th report
8. A National Blood Conservation Strategy for of the Confidential Enquiries into Maternal
NBTC and NBS. Compiled by Virge James on Deaths in the UK
behalf of the NBS Sub-Group ‘Appropriate Use 21. AAGBI Guidelines for Obstetric Anaesthetic
of Blood’, January 2004 Services, Revised Edition 2005
9. NHS Executive. Better Blood Transfusion: Appro- 22. Intra-operative blood cell salvage in obstetrics.
priate Use of Blood. London: Department of National Institute for Health and Clinical Excel-
Health, 2002 (Health Service Circular 2002/009) lence, November 2005
10. Peri-operative Blood Transfusion for Elective
Surgery. http://www.sign.ac.uk
426
448
30 August 2006 14:25:20
48
TREATING HEMORRHAGE FROM SECONDARY ABDOMINAL
PREGNANCY: THEN AND NOW
N. A. Dastur, A. E. Dastur and P. D. Tank
INTRODUCTION hemorrhage, so the peritoneal cavity was packed

under pressure with a large bed sheet as a last
Abdominal pregnancy is an unusual but real
resort. She was stable for the first 6 h postopera-
cause of postpartum hemorrhage. The high
tively, but then developed hypovolemic shock
maternal morbidity and mortality associated
from intraperitoneal hemorrhage and died on
with abdominal pregnancy are a function of
the first postoperative day.
abnormal placentation which leads to intra-
abdominal hemorrhage or the aftermath of
retention of large amounts of dead tissue.
CASE 2
Presently, no evidence-based guidelines have
been published on this subject. This chapter The second case occurred 4 years later at the
begins with a series of four cases treated at same institute. A Cesarean delivery was under-
the Nowrosjee Wadia Maternity Hospital in taken to deliver a 30-year-old multiparous
Mumbai, India, which are illustrative of the woman with no progress in labor. On opening
available treatment options. Wadia Hospital is a the peritoneum, the amniotic sac was encoun-
tertiary-care center with a wide referral base, tered directly. A 2400-g female child was deliv-
both inside the city and throughout the sur- ered. The placenta covered the lateral pelvic
rounding areas. This is followed by a discussion wall and posterior surface of the uterus. The
on the technical aspects of the surgical interven- senior consultant was called and an attempt at
tion and a review of the literature on modern placental separation was made. This effort was
treatment options. soon abandoned in view of the difficulty in sepa-
ration and ensuing hemorrhage. The cord was
then cut short and tied, the placenta left in situ
CASE 1 and the abdomen closed. The abdomen was
packed under pressure with large abdominal
In 1970, a primigravida aged 24 years was
packs for control of the hemorrhage. However,
referred to the hospital with an abnormal pre-
the patient developed a disseminated intra-
sentation. The senior author (NAD) was prac-
vascular coagulopathy and died within 48 h of
ticing as a junior trainee. At that time, it was
the surgery.
routine to confirm the diagnosis of abnormal
presentation with abdominal radiography.
Because the radiograph was suspicious of an
CASE 3
abdominal pregnancy, the senior consultant
planned an exploratory laparotomy to deliver In 1980, the senior author was involved in the
the woman. A male child weighing 2700 g third case of abdominal pregnancy. A 20-year-
was delivered in good condition. However, the old primigravida was referred to the hospital at
placenta was attached to the mesentery, and full term with abdominal pain thought to be of
an attempt to separate it set off massive a surgical cause. There was a strong clinical
hemorrhage. Local measures such as ligation of suspicion of acute appendicitis which did
vessels and compression failed to reduce the not respond to conservative treatment. A
427
449
30 August 2006 14:25:20
laparotomy was performed. A full-term abdomi- pregnancies may have risen over the years, con-
nal pregnancy was found with the sac just below sidering that the risk factors such as ectopic
the peritoneum. A female child weighing 2600 g pregnancy, infertility from tuberculosis and
was delivered in good condition. The placenta endometriosis, pelvic infections and infertility
was firmly adherent to the right pelvic side-wall. treatments are more common today. Regard-
No attempt was made to remove it. The cord less, an obstetrician practicing alone may never
was cut short and tied and the abdomen was come across an abdominal pregnancy in a career
closed with a pelvic drain. The postoperative spanning decades. In the singular instance
course was complicated by fever for the first where he/she does have the need to treat such a
10 days in spite of antibiotics. She continued patient, it may be in circumstances far from
to have abdominal pain for 6 months after ideal. Although unusual, obstetricians should
delivery. This patient had sequelae of a retained be aware of this potentially fatal condition, a
placenta but survived the pregnancy. circumstance amply illustrated by the first two
cases described above.
CASE 4
Although this is not a case of an abdominal
DIAGNOSIS
pregnancy, it is used to illustrate the manage-
ment of abnormal placentation. In 2001, the A primary abdominal pregnancy presents in the
senior author performed a Cesarean section for first trimester in much the same fashion as an
a 25-year-old primigravida at term. She was ectopic pregnancy. An advanced secondary
diagnosed to have an anterior placenta previa abdominal pregnancy, on the other hand, is
with accreta. Blood vessels were seen invading much more difficult to diagnose. Presenting
into the bladder wall on color Doppler. After complaints may include abdominal pain (rang-
delivering a 2500-g male child in good condi- ing from mild discomfort to unbearable pain),
tion, no attempt at placental separation was painful or absent fetal movements, nausea,
initiated. Rather, a decision was made to leave vomiting, abdominal fullness, flatulence,
the placenta in situ followed by methotrexate diarrhea and general malaise. On examination,
therapy. The woman was monitored in hospital there may be an abnormal lie (15–20% of
for 3 weeks after delivery and administered cases), easily palpable fetal parts, a closed unef-
a prolonged course of antibiotics. She had an faced cervix on vaginal examination, and the
uneventful course. Further follow-up was pro- failure to stimulate contractions with oxytocin
vided on an outpatient basis with color Doppler or prostaglandins on attempting an induction of
and serum β-hCG levels. The placental mass labor3. Obviously, these symptoms and circum-
gradually involuted over a period of 5 months stances are far from specific. Taken together,
and the patient resumed menstruation 7 months however, they may (and should) raise a question
after delivery. about the location of the pregnancy. On review-
ing the laboratory findings, one may also find
an unexplained transient anemia in early preg-
INCIDENCE
nancy corresponding to the time of tubal rup-
Abdominal pregnancies are rare events. In the ture or abortion. The serum α-fetoprotein value
United States, it is estimated that it occurs once may be abnormally elevated without explana-
in 10 000 live births and once also for every tion. Early diagnosis has been described in
1000 ectopic pregnancies1. A more recent Afri- response to evaluation of abnormal biochemical
can report provides a much higher estimate of screening results4.
4.3% of ectopic pregnancies, which is probably The diagnosis can be established with far
a reflection of referral patterns in that region as greater certainty by imaging studies. Ultrasound
well as a higher baseline rate of inherent tubal is ubiquitously used in pregnancy, but it does
disease in the patient base of the hospital catch- not always provide an unequivocal diagnosis.
ment area2. However, it also may be reasonable Even under ideal conditions, the diagnosis is
to presume that the incidence of abdominal missed on ultrasound in more than half of
428
450
30 August 2006 14:25:20
Treating hemorrhage from secondary abdominal pregnancy
cases3. Akhan and colleagues5 report the follow- reduced bleeding at laparotomy is not substanti-
ing criteria suggestive of abdominal pregnancy: ated. Surgical intervention is mandated if any of
the above conditions are present.
(1) Visualization of the fetus separate from the
Some clinicians argue that, if there is an
uterus;
ongoing abdominal pregnancy greater than 24
(2) Failure to visualize the uterine wall between weeks, a conservative approach should be taken
the fetus and the maternal urinary bladder; to allow fetal maturity and improve chances of
survival6. However, even after 30 weeks, fetal
(3) Close approximation of fetal parts to the
survival is only 63%, and 20% of fetuses have
maternal abdominal wall;
deformations (craniofacial and various joint
(4) Eccentric position (relation of fetus to abnormalities) and malformations (central
uterus) or abnormal fetal attitude (relation nervous system and limb deficiencies)7. With
of fetal parts to one another) and visualiza- advancing gestation, one also has to contend
tion of extrauterine placental tissue. with the growing placenta and greater risk of
bleeding. In our opinion, it would very rarely be
In the past, radiography was commonly used to
justified to manage an abdominal pregnancy
establish or at least point to this diagnosis. Fea-
conservatively.
tures such as absence of uterine shadow around
the fetus, maternal intestinal shadow intermin-
gling with fetal parts on anteroposterior view, PREOPERATIVE PREPARATIONS
and overlapping of the maternal spine by fetal
The major risk with surgery is torrential hemor-
small parts in a lateral view were all described.
rhage. When a diagnosis of abdominal preg-
Today, however, radiography is largely sup-
nancy is established in advance, the opportunity
planted by magnetic resonance imaging and
to be prepared should not be lost. At least six
computed tomography. Both these techniques,
units of blood should be cross-matched and
with their ability to produce images in different
read to transfuse in the operating room, and
planes, have much greater accuracy and speci-
other blood products should also be available.
ficity than ultrasound. There is little to choose
Two intravenous infusion systems capable of
between the two imaging modalities in cases of
delivering large volumes of fluids rapidly should
fetal demise. If the fetus is alive, magnetic reso-
be established. A mechanical bowel preparation
nance imaging may be preferable since ionizing
should be affected if time permits. A MAST
radiations are avoided.
(medical antishock garment) suit has been
utilized successfully in controlling intractable
hemorrhage with an abdominal pregnancy8,
TIMING OF INTERVENTION
but these garments are not always available
Maternal mortality is about 7.7 times higher (see Chapter 14 for a full discussion). Kerr
with an abdominal pregnancy as compared to a and colleagues9 have advocated preoperative
tubal ectopic pregnancy and 90 times higher as transfemoral catheterization and embolization
compared to an intrauterine pregnancy1. These of selective vessels before surgical intervention.
risks are thought to be chiefly related to the This intervention was used successfully in three
delay in diagnosis and mismanagement of the cases and the catheters can be left in place for
placenta. To minimize the risk from sudden, their potential help in treating postoperative
life-threatening intra-abdominal bleeding, it bleeding as well. The operating team should
seems prudent to time intervention as soon be an experienced one, and preferably should
as feasible after the diagnosis is confirmed. include a general, vascular and genitourinary
There is no controversy if there is maternal surgeon. The anesthesia team should be com-
hemodynamic instability, the fetus is dead or prised of senior consultants and their assistants.
pre-viable (less than 24 weeks pregnancy), has The operating room and nursing staffs should
oligohydramnios or gross abnormalities on be fully aware of the nature of the diagnosis and
ultrasound. The hypothesis that fetal death its implications and schedule extra personnel in
will bring about placental involution and hence the room and as ‘runners’.
429
451
30 August 2006 14:25:20
SURGICAL APPROACH hemorrhage and shock2. The hemorrhage from

the placenta is now torrential and rapid surgical
A mid-line vertical approach is preferential, as it
action is essential. Various local techniques such
can easily be extended above the umbilicus if
as compression of the bleeding site, ligating the
necessary. The amniotic sac may be adherent to
vascular pedicles, lavage with cold saline, and
the abdominal wall and viscera. It should be
local and/or systemic coagulation promoting
dissected free and opened in an avascular
agents (tranexamic acid, plasminogen deriva-
area away from the placenta. The fetus should
tives, absorbable gelatin sponge, etc.) have been
be removed in such a manner as to minimize
described. Repair of placental lacerations may
placental manipulation and avoid bleeding. If
be required. The removal of the organ to which
the pregnancy has been retained for a long
the placenta is adherent (hysterectomy and/or
period after fetal death, the fetus will have
salpingoophorectomy, resection of the bowel
undergone suppuration. Bacterial contamina-
and/or bladder) may be justified to control
tion and abscess formation are highly likely,
the hemorrhage. If a hysterectomy has been
especially if the placenta is adherent to the
performed and bleeding continues, a Logo-
intestines. There may be frank pus upon enter-
thetopoulos pack brought out through the
ing the peritoneal cavity. Rarely, the fetus may
vaginal cuff can be used to exert pressure
be mummified and calcified into a lithopedion
on the pelvic side-walls and bleeding vessels
or become converted into a yellow greasy mass
(see Chapter 33 for complete details). As a
called adipocere formation.
last resort, the abdomen may be packed tight
with abdominal sponges and closed partially.
The packs can be removed 48 h postoperatively
MANAGEMENT OF THE PLACENTA
or sooner if directed by hemodynamic
The torrential hemorrhage that often ensues instability.
with surgery for abdominal pregnancy is related
to the lack of constriction of the hypertrophied
POSTOPERATIVE CARE
opened blood vessels after placental separation.
Usually, the placenta is firmly attached to the Even when the placenta is left in situ, compli-
parietal peritoneum, mesentery and bowel and cations such as infection, abscesses, bowel
there is no bleeding if it is left alone. The umbilical obstruction secondary to adhesions or wound
cord should be ligated close to the placenta, dehiscence occur in about one-half of the
excess membranes trimmed away and the abdo- patients11,12. Although the problems associated
men closed with drainage. Only very rarely is with an abdominally retained placenta may
the placental implantation limited to the repro- be distressing and lead to subsequent repeat
ductive organs by a single pedicle, so that it can laparotomy, they are potentially less disastrous
be easily removed10. than an ill-advised attempt at removing the
In some instances, the placenta may separate placenta. Prophylactic antibiotics should be
spontaneously, simulating an abruption, but administered so as to cover a substantial part
the situation in which hemorrhage becomes of the postoperative course. Less common
uncontrollable is more likely to arise from failed complications of the retained placenta include
attempts at placental removal. Some clinicians reversible maternal hydronephrosis13 and pro-
advocate routine placental removal3,8, but these longed persistent postpartum pre-eclampsia14.
papers were written before the obstetrics com- To hasten placental resorption, methotrexate
munity appreciated the value of methotrexate as a single dose of 50 mg/m2 can be used. This
in such instances. Placental separation requires too is not without its specific problems, how-
complete ligation of the blood vessels supplying ever. In a series of ten cases, accelerated placen-
the placenta and manipulating it at its insertion. tal destruction led to accumulation of necrotic
More importantly, placental separation is not tissue and abscess formation15. It is difficult to
always straightforward and fails in 40% of attribute this to methotrexate therapy alone, as
cases3. This is where the blood supply cannot these complications arise even without adminis-
be completely ligated, resulting in massive tration of methotrexate.
430
452
30 August 2006 14:25:20
Treating hemorrhage from secondary abdominal pregnancy
The patient with a retained placenta is moni- 6. Hage ML, Wall LL, Killam A. Expectant
tored with clinical evaluation, ultrasound, color management of abdominal pregnancy. A report
Doppler and serum β-hCG levels. Hormonal of two cases. J Reprod Med 1988;33:407–10
parameters drop rapidly in the postoperative 7. Stevens CA. Malformations and deformations in
abdominal pregnancy. Am J Med Genet 1993;47:
period as most live cells will be destroyed early.
1189–95
The physical mass of the placenta is resorbed
8. Sandberg EC, Pelligra R. The medical anti-
slowly over an average period of 6 months. A gravity suit for management of surgically uncon-
resorption period of 5 years has been reported16, trollable bleeding associated with abdominal
although this is highly unusual. pregnancy. Am J Obstet Gynecol 1983;146:
519–25
9. Kerr A, Trambert J, Mikhail M, Hodges
CONCLUSION L, Runowicz C. Preoperative transcatheter
Secondary abdominal pregnancy is an uncom- embolization of abdominal pregnancy: Report of
mon and exceedingly dangerous variant of three cases. J Vasc Interv Radiol 1993;4:733–5
ectopic pregnancy. It is usually not diagnosed 10. Noren H, Lindblom B. A unique case of abdom-
inal pregnancy: what are the minimal require-
until laparotomy which leaves the obstetrician
ments for placental contact with the maternal
little preparation to face the prospect of torren-
vascular bed? Am J Obstet Gynecol 1986;155:
tial postpartum hemorrhage, albeit not from the 394–6
usual sources. In this situation, minimizing pla- 11. Bergstrom R, Mueller G, Yankowitz J. A
cental handling and leaving it in the abdominal case illustrating the continued dilemmas in
cavity can be life-saving. treating abdominal pregnancy and a potential
explanation for the high rate of postsurgical
febrile morbidity. Gynecol Obstet Invest 1998;46:
References 268–70
1. Atrash HK, Friede A, Hogue CJR. Abdominal 12. Martin JN Jr, McCaul JF 4th. Emergent
pregnancy in the United Status: frequency and management of abdominal pregnancy. Clin
maternal mortality. Obstet Gynecol 1987;69: Obstet Gynecol 1990;33:438–47
633–7 13. Weiss RE, Stone NN. Persistent maternal
2. Ayinde OA, Aimakhu CO, Adeyanju OA, hydronephrosis after intra-abdominal pregnancy.
Omigbodun AO. Abdominal pregnancy at the J Urol 1994;152:1196–8
University College Hospital, Ibadan: a ten-year 14. Piering WF, Garancis JG, Becker CG, Beres JA,
review. Afr J Reprod Health 2005;9:123–7 Lemann J Jr. Preeclampsia related to a function-
3. Costa SD, Presley J, Bastert G. Advanced ing extrauterine placenta: Report of a case and
abdominal pregnancy. Obstet Gynecol Surv 1991; 25-year follow-up. Am J Kidney Dis 1993;21:
46:515–25 310–13
4. Bombard AT, Nakagawa S, Runowicz CD, 15. Rahman MS, Al-Suleiman SA, Rahman J,
Cohen BL, Mikhail MS, Nitowsky HM. Early Al-Sibai MH. Advanced abdominal pregnancy –
detection of abdominal pregnancy by maternal observations in 10 cases. Obstet Gynecol 1982;59:
serum AFP+ screening. Prenat Diag 1994;14: 366–72
1155–7 16. Belfar HL, Kurtz AB, Wapner RJ. Long-term
5. Akhan O, Cekirge S, Senaati S, Besim A. follow-up after removal of an abdominal preg-
Sonographic diagnosis of an abdominal ectopic nancy: ultrasound evaluation of the involuting
pregnancy. Am J Radiol 1990;155:197–8 placenta. J Ultrasound Med 1986;5:521–3
431
453
30 August 2006 14:25:21
454
30 August 2006 14:25:21
Section X
National experiences
455
30 August 2006 14:25:21
49
COMBATING POSTPARTUM HEMORRHAGE IN INDIA:
MOVING FORWARD
D. S. Shah, H. Divakar and T. Meghal
INTRODUCTION One of the critical bottlenecks for providing

more high-quality emergency obstetric care
The World Health Organization (WHO)
(EOC) was a serious shortage of specialist staff
estimates that, of the 529 000 maternal deaths
such as obstetricians and anesthesiologists at
occurring every year, 136 000 or 25.7% take
various levels in rural areas. This deficiency was
place in India, where two-thirds of maternal
accentuated by the limited capacity for transfu-
deaths occur after delivery, postpartum hemor-
sion outside of the more sophisticated urban
rhage being the most commonly reported
areas.
complication and the leading cause of death
The present strategies to prevent maternal
(29.6%)1. The unacceptably high maternal
mortality in India focus on building a better and
death ratio (540/100 000 live births)1 in India
more fully functioning primary health-care
during the last few decades remains a major
system, from first referral level facilities to the
challenge for health systems.
community level. It is unfortunate that emer-
According to the same WHO estimates, for
gency obstetric care is not yet available for all
every maternal death about 20 women suffer
patients in labor and this should be the main
from harm to general and reproductive health.
focus of the government as well as the medical
In India, around 70% of the population lives in
profession.
villages. Out of an estimated 25 million deliver-
ies each year, 18 million take place in peripheral
areas where maternal and perinatal services are
Effective interventions for reducing the
either poor or non-existent. India’s stated goal is
incidence of postpartum hemorrhage
to reduce maternal mortality (MMR) from 437
deaths per 100 000 live births that was recorded Although training programs for traditional birth
in 1991 to 109 by 2015. The MMR for 1998 is attendants (TBAs) are designed to improve the
407. Along with this improvement, the propor- routine care for mothers and newborns at deliv-
tion of births attended by skilled health person- ery, these interventions have proved ineffective
nel has increased from 25.5% in 1992–1993 to in reducing maternal deaths2–5. Neither trained
39.8% in 2002–2003, thereby reducing the TBAs nor any other category of minimally
chances of occurrence of maternal deaths1. trained community health worker can prevent
The efforts to improve maternal health and the vast majority of obstetric complications
reduce maternal mortality have been continu- from occurring. Once a complication occurs,
ous in India since 1960 under the public health there is almost nothing TBAs, by themselves,
program of Primary Health Care – specifically can do to reduce the chance of morbidity or
under the Maternal and Child Health (MCH) death that can ensue.
program. In various policy documents, the gov- As women at high risk for postpartum hem-
ernment of India has listed the reduction of orrhage account for only a small percentage of
maternal mortality as one of its key objectives. all maternal deaths, the vast majority of deaths
Unfortunately, progress has been less than occur in women with no known risk factors.
hoped for several reasons. Stated another way, risk screening programs
434
456
30 August 2006 14:25:21
Combating postpartum hemorrhage in India
have had little impact on overall maternal deliver far fewer babies. This affects their com-
mortality levels6–9. petence, because specific skills, such as manual
Recognizing these flaws in the early recom- removal of the placenta, require regular practice
mendations of the Safe Motherhood Initiative, in order to be maintained. In Indonesia, for
the present-day clear international consensus example, where tens of thousands of commu-
is that scarce resources should not be spent in nity midwives have been trained and deployed
trying to predict which women will have life- to villages around the country, each typically
threatening complications (Safe Motherhood delivers fewer than 36 babies a year. Assess-
Initiative). Rather, maternal mortality reduction ments within 3 years of placement found that
programs should be based on the principle confidence and competency-based skills were
that every pregnant women is at risk for life- exceedingly low, with only 6% scoring above
threatening complications. In order to reduce 70, the minimum level considered necessary for
the maternal mortality ratio dramatically, all competence11.
women must have access to high-quality care at In addition to being properly trained for
delivery. That care has three key elements: conducting routine deliveries, a second and
more promising way in which skilled attendants
(1) A skilled attendant at delivery;
can reduce the incidence of postpartum hemor-
(2) Access to emergency obstetric care (EOC); rhage is by actively managing the third stage of
labor in every delivery12 (see Chapters 11 and
(3) A functional referral system.
13). However, the same techniques of active
management that can prevent some postpartum
hemorrhages can also cause serious damage
SKILLED ATTENDANTS AT DELIVERY
if performed incorrectly. This is not just a
Evidence concerning the effect of skilled theoretical risk. Incorrect use of oxytocic drugs,
attendants at delivery is somewhat confused for example, can cause the uterus to rupture,
by different definitions and by variations across which, in the absence of surgical intervention,
countries. The training of midwives and the can lead to death.
regulations governing the procedures they are
permitted to perform vary considerably. In
The EOC Project in India
2004, WHO, the International Confederation
of Midwives, and the International Federation A project is being established to develop the
of Gynecology and Obstetrics issued a joint capacity of general practitioners and non-
statement with a revised definition of skilled specialist medical officers to provide high-
attendant: ‘A skilled attendant is an accredited quality EOC services in rural areas where
health professional – such as a midwife, doctor skilled obstetricians are not available to prevent
or nurse – who has been educated and trained maternal mortality and morbidity13.
to proficiency in the skills needed to manage The Federation of Obstetrics and Gyneco-
normal (uncomplicated) pregnancies, childbirth logical Societies of India (FOGSI) has estab-
and the immediate postpartum period, and in lished five EOC training centers in rural India
the identification, management and referral of that will improve the provision of EOC services
complications in women and newborns.’ by medical officers, with the ultimate goal of
Wide variation exists in the extent to which reducing maternal mortality and morbidity.
skilled attendants are supported and supervised The project has been funded by the MacArthur
in the broader health system. This is also true Foundation, Baltimore, USA and the AMDD
for the number of deliveries that skilled atten- (Averting Maternal deaths and Disability),
dants perform annually. In a country such as Columbia University, New York. JHPIEGO (an
Malaysia, which dramatically lowered its mater- international health organization affiliated with
nal mortality in the 1960s and 1970s, midwives Johns Hopkins University) assists FOGSI in its
became the backbone of the program, each endeavor to assess and strengthen selected
delivering 100–200 babies per year10. However, EOC training sites, train selected trainers and
in many other countries, birth attendants strengthen FOGSI’s capacity in the area of
435
457
30 August 2006 14:25:21
monitoring and evaluation. During Phase 2, work places. A Certificate will be issued at the
FOGSI and JHPIEGO will also work together end. Advocacy with the government and NGO
to orient key stakeholders to the value of heads is being negotiated to ensure that the
these innovations in EOC training and service trainee’s facility is functional and to establish
delivery for feedback in order to gain consensus one training center in each state of India.
among stakeholders for scale-up of the
approaches and technical interventions.
Expected outcomes
FOGSI members who have a keen interest in
training doctors and midwives for rural areas Five tertiary training centers and 20 district
will run these training centers. Each center will centers are well equipped to start the EOC
have a coordinator and three to four faculty Training Certification Course. Three tertiary
members. These are all staff of medical colleges centers and eight district centers have already
or well-known consultants. The District Train- started training, whilst two tertiary centers and
ing Centers will have one obstetrician func- 12 district centers will start functioning by the
tioning as the District Trainer. end of October 2006. A total of 162 doctors will
be trained during the pilot project of 2 years for
three centers established by FOGSI, MacArthur
Design and methods policy
and JHPIEGO. FOGSI plans to develop, in
Training centers will be set up in medical a phased manner, one center per state in the
colleges where there are dedicated doctors inter- future. It is expected that this pilot effort will
ested in rural women’s health. All master train- be replicated by the government. The policy
ers will be trained in EOC at the nodal center advocacy efforts will help in this direction to
by doctors trained by JHPIEGO. Four master convince government and other stakeholders
trainers at medical colleges and four at district to support and develop the program so as to
level hospitals will provide the training in a uni- provide 24-h EOC services in rural areas.
form manner. Each training center will offer two
types of courses: a short course of 3 weeks for
Upscaling the program
upgrading the skills of doctors already working
in rural or under-served areas but not possess- The advocacy efforts of FOGSI have resulted
ing sufficient knowledge of EOC, and a long in a significant change in the priorities of the
course of 16 weeks to provide comprehensive government of India for phase II of the Repro-
skills including training in performing a Cesar- ductive and Child Health Care program. Very
ean section. This latter course will be composed recently, the Indian government committed
of 6 weeks of training in medical college by four itself to the EOC training project of FOGSI.
master trainers and 10 weeks of practical train- According to the preliminary discussions with
ing in a district-level hospital. Courses will be the government, FOGSI has been entrusted
competency-based and finalized in consultation with the task of developing 20 tertiary training
with the Department of Health and Family Wel- centers and 160 district training centers wherein
fare. These courses will be open to any doctor 2000 medical officers will be trained for 16
working in rural and under-served areas, from weeks of comprehensive emergency obstetric
the government, NGO or private sectors. care. These medical officers will provide a
The roles of FOGSI/ICOG will be, first, to skilled high-quality comprehensive EOC
coordinate with medical colleges and govern- through the network of first referral units and
ment hospitals to make arrangement for community health care centers at subdistrict
training, and, second, to regularly monitor the and Taluka places (a Taluka is an administra-
master trainers, the training program and the tive block consisting of 80–100 contiguous
quality of training centers and to formalize villages). The whole program has been planned
the end assessment and certification. At the within a time frame of 5 years. During the same
end of each course, follow-up and support time period, the government will upgrade these
activities will ensure that the trainees start to centers with the necessary infrastructure such as
offer EOC services after going back to their an operating theater, equipment, blood storage
436
458
30 August 2006 14:25:21
facilities and persons trained in anesthesia. This It is expected that the participants of each
conceptual change in providing EOC at under- individual institute will be able to state and
served places will take EOC to the areas where it demonstrate the standard protocol for active
is most needed and will bring about a significant management of the third stage, will practice
reduction in the maternal mortality ratio. active management of the third stage and
have an updated knowledge and skills for the
management of postpartum hemorrhage.
The AOFOG PPH initiative
The Asia Oceania Federation of Obstetrics and ACCESS TO EMERGENCY OBSTETRIC
Gynaecology (AOFOG) has launched a pro- CARE
gram called the AOFOG PPH Initiative14. This
program focuses on the active management of Even under the very best of circumstances, with
the third stage of labor in areas with skilled birth adequate nutrition, high socioeconomic status
attendants and in areas where misoprostol is and good health care, approximately 15% of
available but without skilled birth attendants. pregnant women experience potentially fatal
This effort is in support of the FIGO/ICM joint complications. Fortunately, virtually all obstet-
statement on the management of the third stage ric complications can be successfully treated if
of labor to prevent postpartum hemorrhage. EOC is universally accessible and appropriately
The focus is on training of trainers in the utilized. United Nations guidelines recommend
national societies of those countries whose a minimum of one comprehensive facility and
maternal mortality ratio exceeds 100/100 000 four basic EOC facilities per 500 000 popula-
live births. tion. To reduce maternal mortality ratios
by 75%, high-mortality countries must
substantially improve access to emergency care.
Objectives
The objectives of the AOFOG PPH initiative Solution exchange for maternal and child
are: health practitioners in India
(1) To disseminate a standard protocol for India is a vast, powerful storehouse of
active management of the third stage of knowledge. While ‘expert’ knowledge is well
labor and to ensure uniform and safe documented, valuable knowledge gained
institutional practice; through practitioner experience is typically lost
or ignored. Furthermore, practitioners cannot
(2) To train the service providers (doctors,
always access the knowledge they need, such as
midwives, nurses, family welfare visitors) in
whether a particular idea was tried before or
the institutes to perform active manage-
where to turn when facing a bottleneck. To har-
ment of the third stage of labor for all
ness this knowledge pool and help practitioners
women giving birth;
avoid reinventing the wheel, the United Nations
(3) To inform the medical and nursing profes- offices in India created the Solution Exchange –
sion about the rational use of uterotonic a free, impartial space where professionals are
drugs, such as oxytocin and ergometrine, welcome to share their knowledge and experi-
and the role of misoprostol for preventing ence15. Members represent a wide range of
postpartum hemorrhage; perspectives from government, NGOs, donors,
the private sector and academia. They are
(4) To discuss, demonstrate and to train
organized into Communities of Practice built
the service providers regarding the
around the framework of the Millennium
evidence-based management for post-
Development Goals. Members interact on an
partum hemorrhage;
ongoing basis, building familiarity and trust,
(5) To develop an action plan to be imple- gaining in knowledge that helps them contribute
mented in respective institutes and to more effectively – individually and collectively –
monitor the outcome. to development challenges.
437
459
30 August 2006 14:25:21
Communities begin with the Solution Andhra Pradesh, Chidambaram in Tamil

Exchange’s personalized ‘Research Service’. Nadu, Loni, Solapur and Amravathi in
Here individual members post questions on the Maharashtra, Bijapur and Shimoga in
Community’s web-based platform about the Karnataka, Kota and Ajmer in Rajasthan,
development challenges they face; other mem- Jabalpur and Sagar in Madhya Pradesh, to
bers respond to these questions and the moder- name just a few16.
ation team provides research into them. The The take-home messages from these work-
tacit knowledge and expert knowledge are shops were, first, that actively managed and
brought together in a summarized ‘Consoli- supervised labor has a better outcome with a
dated Reply’ which is circulated to the Commu- decreased incidence of operative deliveries, and,
nity, normally within 10 working days. second, that an actively managed third stage
The Maternal & Child Health (MCH) decreases the blood loss and incidence of
Community, facilitated by WHO, UNICEF postpartum hemorrhage.
and UNFPA country offices in India, focuses on
implementation issues facing the attainment of
REFERRAL SYSTEMS
the development goals and targets in the Tenth
Five-Year Plan of India, the National Popula- Widely available, good-quality EOC is neces-
tion Policy 2000, Rural Health Mission and sary but not sufficient by itself to reduce the
Phase II of the Reproductive and Child Health incidence of postpartum hemorrhage. Appro-
Programme, which correspond most closely to priate utilization is also necessary. A helpful way
the universally endorsed Millennium Develop- to analyze the barriers to utilization is through
ment Goals and targets leading to reduction of the ‘three delays model’17. Once a complication
maternal and child mortality. occurs, the key to saving a woman’s life is to
The main focuses of the MCH Community provide her adequate care in time. The delays
are to improve maternal health and reduce leading to death can be divided into three
maternal mortality, and to improve child health categories:
and reduce infant and child mortality. The
(1) Delay in deciding to seek care;
MCH Community has now been in action for
almost a year, with membership growing from (2) Delay in reaching care;
130 to 725 during this time, representing 28
(3) Delay in getting treatment at the facility.
states and union territories of India and a few
members from outside India as well. Discus- One important element of strategies to reduce
sions have ranged from skilled attendance at delays is the strengthening of the referral sys-
birth, setting up a telemedicine center, exclusive tem. Widespread ‘failures’ in referral systems
breast-feeding and complementary feeding, are often present, particularly for the poor and
operationalizing urban Integrated Child Devel- marginalized. The recent review by Murray and
opment Services, medical termination of Pearson18 found significant gaps in understand-
pregnancies, etc. ing how referral systems are currently function-
ing in addition to highlighting a fundamental
problem in the literature, that is, that many
Safe motherhood initiative from FOGSI
studies rely on a conceptualization of an ideal
‘Optimizing Labor workshops’ were held in 66 referral system that has a dangerously tenuous
societies across the country, and four Work- relationship to realities on the ground.
shops on postpartum hemorrhage were spon- Maternity referral systems were first con-
sored by AOFOG. The Federation was able to ceived at a time when risk screening was
involve doctors from the government service thought to be an appropriate maternal mortality
and nurses practicing in rural areas in the work- reduction strategy, even for high-mortality
shops along with its members. Workshops were countries. This conception assumed a stepwise
held in the Societies that cater to large rural hierarchy of increasingly sophisticated facilities,
populations such as Kalyani in Bengal, Gawhati and it assumed that high-risk women would
in Assam, Rajmundhry and Vijaywada in be referred up the ladder as their pregnancy
438
460
30 August 2006 14:25:22
progressed. Today, however, maternal mortality complications. They find that fear of losing
strategies concentrate on emergencies, because prestige and future business often gets in the
it is acknowledged that time is critical. An ele- way.
gant model of referral from facility to facility Maternal mortality strategies should focus
could be worse than inefficient, it could be on building a functioning primary health-care
deadly! system, from first referral level facilities to
Although organized ambulance services the community level. Emergency obstetric care
appear to be part of the referral system in every must be accessible for all women who experi-
country that has achieved major maternal mor- ence complications in pregnancy and child
tality reductions, access to transport is only one birth. Skilled birth attendants, whether based in
part of a far more complex problem. Maternal facilities or communities should be the back-
mortality strategies that address the ‘second bone of the system. Skilled attendants for all
delay’ simply by funding and organizing deliveries must be integrated with a functioning
transport fail to grapple with perhaps even district health system that supplies and supports
more critical systemic issues. them adequately.
First and foremost is the need for referral
facilities that provide 24-h 7-day-a-week care
Achievements of the health department
within a reasonable distance of where people
live. Murray and Pearson conclude that ‘Exten- The government of the state of Tamil Nadu is
sive pyramidal structures of referral systems committed to providing good-quality medical
with multiple tiers of facilities would seem to care to the people in the rural areas. To achieve
offer little benefit in the majority of cases for this, 105 primary health centers have been
maternity care and simply delay treatment’18. upgraded to 30-bed hospitals20. These hospitals
In most countries, attention should be concen- have been equipped with X-ray machines,
trated on referral within the district-level sys- ECG, ultrasonography, operation theaters
tem. From the perspective of a district health and laboratories. Another 180 primary health
system as a whole, it is the strength of the centers provide 24-h delivery care.
referral facilities and associated supervision and In addition, 62 Comprehensive Emergency
referral systems that should determine the level Obstetric and Newborn Care (CEONC) cen-
of skill that birth attendants must have in order ters have been established for providing 24-h
to avert maternal deaths, not vice versa. Murray maternal and child health-care services, includ-
and Pearson provide the example of Yunnan, ing Cesarean sections. These centers have been
China, where accessible referral facilities, a so located as to be accessible within an hour’s
well-functioning referral system, and a strong travel from anywhere in Tamil Nadu. In the
and very active supervision system meant that second phase, more hospitals will be upgraded
semi-skilled village doctors could successfully as CEONC centers so as to reduce the time to
conduct normal births, recognize problems, sta- 30 min.
bilize patients, and refer them onward for more For the first time in India, a birth companion
complex treatment of emergencies. With this scheme has been introduced, permitting one
system, Yunnan reduced its maternal mortality female attendant to stay with the antenatal
ratio from 149 to 101 in the 1990s11. mother during labor in the labor room of
Unfortunately, however, such results have all government health institutions to provide
not been documented for TBAs. A stated goal psychological support.
of many training programs for TBAs is to In this state, maternal deaths have been
improve their referral of women experiencing reduced by 25% during the last 4 years
obstetric emergencies to facilities that can man- (2001–2004). An excellent network of blood
age them. A recent meta-analysis of studies banks and blood storage centers has been
evaluating training programs designed to established in the government health institu-
improve referral practices of TBAs found little tions to ensure the supply of blood and its com-
effect19. Other recent studies explore why TBAs ponents (86 blood banks and 26 blood storage
often fail to refer even patients with obvious centers).
439
461
30 August 2006 14:25:22
COMMENTS home, very often attended by family members

or neighbors, TBAs or other kinds of minimally
In the safe motherhood community today, the
trained community health workers. The health
question is often posed as whether to give high-
system is so weak that there is no hope of pro-
est priority to training a cadre of workers with
viding emergency obstetric care or even a true
midwifery skills who can attend every birth or to
skilled birth attendant in rural areas at any time
focus on strengthening emergency obstetric care
in the foreseeable future: therefore the strategy
services (including the human resources neces-
should be to provide some additional training to
sary to staff them) in order to treat the approxi-
community health workers or traditional birth
mately 15% of pregnant women who experience
attendants, making them, in effect, semi-skilled
complications. Under the strategy of emergency
attendants.
obstetric care first, therefore, emergency ser-
The enormous pressure that concerned
vices need to be accessible to all (albeit not used
policy-makers feel to do something for the mil-
by all). In theory, the two interventions – skilled
lions of women who give birth in these circum-
attendants for all births and emergency obstetric
stances is recognized. It is also recognized that a
care for complicated ones – do not contradict
semi-skilled worker may have the potential to
each other. But, as strategies in resource-
save a substantial number of newborns who
constrained settings, they fit together less easily.
otherwise would die. But it must be clearly
Ultimately, both interventions appear to be nec-
stated that a strategy of training tens of thou-
essary to reach very low maternal mortality lev-
sands of semi-skilled workers who will not be
els: in every country with a maternal mortality
backed up by a supervision system, a supply sys-
ratio of less than 50 – or even less than 100 – a
tem, or a referral system, is not a strategy that
high proportion of births are attended by skilled
will significantly reduce maternal mortality. In
health personnel and access to emergency
fact, the proliferation of unsupported, unsuper-
obstetric care is widespread. Be that as it may,
vised, semi-skilled workers (‘certified’ after
the reality in high-mortality countries today is
short training courses to manage deliveries) who
that policymakers are indeed confronted with a
are deployed in the context of policies effectively
choice between the two interventions, at least as
that marketize and privatize health care has the
a matter of emphasis or priority setting. Where
potential to increase the dangers for pregnant
should they put their scarce financial, human,
and delivering women. In some cases where
and managerial resources? How should they
such a strategy is being considered, the explicit
sequence these interventions?
objective is to train such workers on the
To look for an answer, we should look to
assumption that they will set up their own
contemporary cases of the few countries or sub-
private practices21. Such private provision will
national units in which maternal mortality ratios
be quite outside any government supervision,
of less than 100 have been achieved. In Malaysia
any effective regulatory system, or even any
and Sri Lanka, a step-by-step approach, starting
self-policing professional body.
with coverage of basic facilities that can deliver
It is not suggested that highly trained special-
emergency obstetric care, followed by a focus on
ists are not necessary to reduce maternal mor-
utilization and quality, went hand in hand with
tality. Many categories of health personnel can
the professionalization of midwifery and a gov-
be taught to provide various health services – as
ernmental commitment to ensuring universal
long as effective systems of support, supervision
access to health services, including access by
and supplies are established.
the poor and people in rural areas10. Over the
All the interventions necessary to save
course of several decades, both countries
women’s lives can be delivered in a district
reduced the incidence of postpartum hemor-
health system – at the primary care and first
rhage and thus halved their maternal mortality
referral levels. This does not mean that women
ratios every 6–12 years, going from more than
must give birth in facilities, nor does it mean
500 in 1950 to less than 30 by the early 1990s.
that TBAs and other private providers have
In a country like India, the vast majority of
no place in a delivery system. The case studies
births (often more than 80%) take place at
of countries that have substantially reduced
440
462
30 August 2006 14:25:22
maternal mortality demonstrate that success is 10. Pathmanathan I, Liljeastrand J, Martins J,

possible with multiple combinations of home et al. Investing in Maternal Health in Malaysia
and institutional births, attended by different and Sri Lanka. Washington, DC: World Bank,
categories of health workers, as long as women 2003
11. Koblinsky M, Campbell O. Factors affecting the
have access to emergency obstetric care staffed
reduction of maternal mortality. In Koblinsky
by skilled health personnel11.
M, ed. Reducing Maternal Mortality: Learning
from Bolivia, China, Egypt, Honduras, Indonesia,
Jamaica and Zimbabwe. Washington, DC: World
References
Bank, 2003
1. Lynn P, Freedman RJ, Waldman H de Pinho, 12. McCormick M, Sanghvi H, Kinzie B, McIntosh
Wirth ME. Who’s got the power? Transforming N. Preventing postpartum hemorrhage in low-
health systems for women and children. UN resource settings. Int J Gynaecol Obstet 2002;77:
Millenium Project Task Force on Child Health 267–75
& Maternal Health, 2005:77–95 13. Abstract of proceedings submitted by Dr
2. Rosenfield A, Maine D. Maternal mortality – a Prakash Bhatt, Vice President FOGSI on
neglected tragedy: where’s the M in Mch? Lancet personal communication
1985;2:83– 5 14. AOFOG PPH Initiative, FOGSI memories
3. Greenwood AM, Bradley AK, Byass P, et al. 2005. Publication from Federation of Obstetric
Evaluation of a primary care programme in the & Gynecological Societies of India
Gambia: the impact of traditional birth atten- 15. Solution Exchange for Maternal & Child
dants on the outcome of pregnancy. J Trop Med Health Practitioners in India. Personal
Hygiene 1990;93:58–66 communication by Dr. Meghendra Banerjee.
4. Goodburn EA, Chowdhury M, Gazi R, et al. mch@solutionexchange-un.net.in
Training traditional birth attendants in clean 16. FOGSI memories 2005. Publication from
delivery does not prevent postpartum infection. Federation of Obstetric & Gynecological Societies of
Health Policy Planning 2000;15:394–9 India
5. Smith JB, Coleman NA, Fortney JA, et al. The 17. Thaddeus S, Maine D. Too far to walk: maternal
impact of traditional birth attendant training on mortality in context. Soc Sci Med 1984;38:
delivery complications in Ghana. Health Policy 1091–110
Planning 2000;15:326–31 18. Murray SF, Pearson S. Maternity referral
6. Danel I, Rivera A. Honduras, 1990–1997. In systems in developing countries: challenges and
Koblinsky M, ed. Reducing Maternal Mortality: next steps. A scoping review of current knowl-
Learning from Bolivia, China, Egypt, Honduras, edge. Background paper commissioned by the
Indonesia, Jamaica and Zimbabwe. Washington, UN Millenium Project Task Force on Child
DC: World Bank, 2003 Health and Maternal Health and the World
7. McCaw-Binns A. Jamaica, 1991–1995. In Health Organization. New York, 2004
Koblinsky M, ed. Reducing Maternal Mortality: 19. Sibley L, Sipe TAT, Koblinsky M. Does tradi-
Learning from Bolivia, China, Egypt, Honduras, tional birth attendant training improve referral of
Indonesia, Jamaica and Zimbabwe. Washington, women with obstetric complications: a review of
DC: World Bank, 2003 the evidence. Soc Sci Med 2004;59:1757–68
8. Maine D. Safe Motherhood Programs: Options and 20. Tamil-Nadu Government Publication on World
Issues. New York: Center for Population and Health Day, 2006. Times of India, April 7th,
Family Health, Columbia University, 1991 2005
9. Greenwood AM, Greenwood BM, Bradley AK, 21. Mavalankar D. Auxiliary nurse midwifes’
et al. A prospective study of the outcome of preg- (ANM) changing role in India: Policy issues
nancy in a rural area of the Gambia. Bull WHO for reproductive and child health. Ahmedabad:
1987;65:635–43 Indian Institute of Management, 1997
441
463
30 August 2006 15:09:46
50
ELIMINATING MORTALITY: LESSONS FROM LUBLIN
PROVINCE IN POLAND
J. Oleszczuk, B. Leszczynska-Gorzelak, D. Szymula, M. Grzechnik, G. Pietras,
J. Bartosiewicz and J. J. Oleszczuk
INTRODUCTION succeeded, in eradicating maternal mortality

from postpartum hemorrhage in one of the
Every year, over half a million women die of
Polish provinces. We are of the opinion that
pregnancy, delivery and postpartum complica-
the findings from our province can be applied
tions – equivalent to the death toll of 15
around the world and have immense impact on
September 11th tragedies in a single year!
reducing unnecessary deaths.
Postpartum hemorrhage is almost always the
number one cause of mortality, and in Poland it
is no different. In the 10 years between 1991
and 2000, a total of 135 women died of THE SYSTEM
postpartum hemorrhage, accounting for about The regionalization system presently in place in
35% of all maternal mortality. In Lublin Lublin Province is based on a very tight network
Province (2 181 018 inhabitants) in the of heads of obstetric and neonatal units
south-eastern section of the country, a well- throughout the Province. The Obstetrician-in-
functioning regionalization system, based on Chief of the Province, who is currently heading
three levels of perinatal care, introduced in one of the perinatal centers, leads the network.
1993, has led to a marked reduction in perinatal The postpartum hemorrhage pathway is
mortality. A total of 25 obstetric units are part based upon a centralized support system in
of the system – 18 in level I, five in level II and which the Provincial Obstetrician-in-Chief acts
two in level III – the latter being the perinatal as at all times as a ‘last resort’ for the most
centers. The organizational structure is com- severe postpartum hemorrhage cases. If such a
prised of the heads of obstetric and neonatal case occurs and the local obstetric unit decides
units all of whom report to the Provincial that an intervention of this senior obstetrician is
Obstetrician-in-Chief who currently is the required, the unit pages the Obstetrician-in-
Head of the Department of Obstetrics and Chief asking for immediate support. If the
Perinatology of the Medical University in Obstetrician-in-Chief is unavailable (which
Lublin. Since 2002, no maternal death due to happened four out of 33 times in the time under
postpartum hemorrhage has been reported in study), the next most senior person in the
Lublin Province. postpartum hemorrhage SWAT team is paged
This chapter describes the regionalization and attends to the patient. An ambulance is sent
system in Lublin Province, along with a specific to pick-up a postpartum hemorrhage ‘rescue
pathway that exists for all postpartum hemor- kit’ (containing recombinant factor VIIa,
rhage cases. In addition, the system is critically NovoSeven®, Novo Nordisk, and a set of faster
evaluated, and potential approaches to replicat- absorption profile sutures for the B-Lynch oper-
ing this system elsewhere are provided. This ation) from the hospital of the Obstetrician-
effort can be viewed as a population-based, in-Chief and then takes him directly to the local
multicentric, prospective, controlled trial of obstetric unit. As the farthest unit is approxi-
an organizational system that aimed, and mately 130 km away from the perinatal center
442
464
30 August 2006 14:25:22
Eliminating mortality: lessons from Poland
and the transport takes up to 1.5 hours in switch to surgical management (laparotomy or
extreme cases, the average time from initiating repeat laparotomy). In all cases, the Obstetri-
the call and delivering the Obstetrician-in-Chief cian-in-Chief was paged and took over further
to the unit takes ~90 minutes. management. (See Chapter 22 for a US
The Obstetrician-in-Chief then takes charge hospital-based approach to reducing mortality.)
of the local obstetric team, evaluates the status Several types of cases can be described,
of the patient and makes a decision about the depending on the status of the patient at the
most appropriate management approach. After local obstetric unit as determined by the Obste-
the intervention, the patient usually remains in trician-in-Chief when he arrived on the scene
the local obstetric unit (or is taken to the local (see Figure 1):
intensive care unit) to which she was admitted
● Patient undergoing surgery with hemor-
but rarely is transferred to the perinatal center.
rhage, difficult to manage but prior to hyster-
During recovery, the Obstetrician-in-Chief
ectomy;
then provides telephone consultations to the
obstetric and intensive care unit teams.
RESULTS
A total of 86 237 births were recorded in Lublin
Province between January 1, 2002 and March
31, 2006. During this time, no maternal
mortality due to postpartum hemorrhage was
reported. The numbers of maternal deaths from
other direct obstetric causes are summarized
in Table 1. No deaths were caused by indirect
obstetric factors or non-obstetric factors.
Between January 1, 2003 and March 31,
2006, 33 cases of postpartum hemorrhage were
managed in the collaborative fashion described
above. In all instances, the local obstetric units
did not manage to control the hemorrhage Figure 1 Level of the local obstetric unit in the 33
pharmacologically, and a decision was made cases of postpartum hemorrhage managed through
to change the pharmacologic approach or to the regionalization system between 2003 and 2006
Table 1 Causes of maternal mortality in Lublin Province between 2002 and 2006
Year
2006
2002 2003 2004 2005 (1.01–31.03) Total
Deliveries 20 260 20 337 19 896 20 598 5146 86 237

Maternal deaths from:
Postpartum hemorrhage 0 0 0 0 0 0
Infection 0 0 0 1 0 1
Embolism 0 0 0 0 0 0
Hypertensive disorders 1 1 0 0 0 2
Indirect obstetric factors 0 0 0 0 0 0
Non-obstetric factors 0 0 0 0 0 0
Total 1 1 0 1 0 3
443
465
30 August 2006 14:25:38
● Patient undergoing surgery with hemor- world-wide, 75 000–125 000 lives could be
rhage, difficult to manage after hysterectomy; saved every year. In all 33 cases, patient status
after surgery was satisfactory and they quickly
● Patient after Cesarean section but repeat
recovered and were discharged home with no
laparotomy needed (to perform hysterectomy
neurologic or other post-hemorrhagic complica-
or save the uterus);
tions. It is important to underline that these
● Patient after delivery – conservative manage- patients experienced the most severe post-
ment unsuccessful and a decision was partum hemorrhage in which the local obstetric
required to switch to other conservative team, usually very well trained, was helpless and
approaches or decide to operate. required support from the Provincial Obstetri-
cian-in-Chief. The other cases of postpartum
Interventions were performed in six cases of
hemorrhage which occurred in the province
vaginal delivery and in 27 cases of Cesarean
were less severe and responded to a variety
section (Figure 2). Table 2 shows the various
of interventions without the need for outside
management approaches used in the 33 cases of
assistance.
severe postpartum hemorrhage described in this
The regionalization system was critical in our
chapter.
success in eradicating maternal mortality due
to postpartum hemorrhage in Lublin Province.
The system in principle aimed at ensuring that
DISCUSSION the most complicated cases are transferred
Using coordinated and well-planned efforts, it is antenatally to the perinatal center, wherever
possible to ‘eradicate’ maternal mortality from such forecasting was possible (e.g. in cases of
postpartum hemorrhage in a large population. placenta previa in patients after prior Cesarean
Even if half of these deaths could be prevented section). In acute cases, however, when patient
Figure 2 Underlying pathology in the 33 cases of severe postpartum hemorrhage between 2003 and 2006
in the Lublin Province
444
466
30 August 2006 14:26:11
Eliminating mortality: lessons from Poland
Table 2 Management approaches in the 33 cases The latter method should, however, only
of severe postpartum hemorrhage in Lublin Province be performed by the highest skilled surgeons
between 2003 and 2006 who are comfortable with retroperitoneal space
surgery, as these approaches carry a high risk of
Number
vascular or ureteral complications. For exam-
Management approach of cases
ple, in one of the cases, the hypogastric vein was
Laparotomy 3 damaged and subsequently required suture clo-
Repeat laparotomy 9 sure. In addition, if these conservative surgical
Total hysterectomy 14 methods are not successful, hysterectomy is the
Cervical stump excision after supracervical 1 method of choice, and it is critical to time this
hysterectomy decision appropriately. In such cases, the uterus
Unilateral adnexectomy due to bleeding or 4 is excised with the cervix (total hysterectomy)
hematoma
but without the adnexa.
Bilateral adnexectomy due to septic shock 1
We see two potential risks with our approach
(with total hysterectomy)
Retroperitoneal hematoma evacuation 2 and potential replicas of our approach else-
Uterine artery ligation 8 where: reimbursement and legal/malpractice.
Ligation of the uterine branches of ovarian 8 In Poland, reimbursement is on a quasi-DRG
arteries (diagnosis-related groups) basis, but the full
Bilateral hypogastric artery (internal iliac) 20 payment goes to the admitting hospital, without
ligation specific breakdown of doctor fees from hospital
Unilateral hypogastric artery ligation 1 fees. Thus, our entire system is essentially per-
Repair of cervical laceration 1 formed on a pro bono basis by the postpartum
B-Lynch suture 1 hemorrhage SWAT team. Unfortunately, this
Hayman suture 1
is not sustainable for the long term, and the
NovoSeven 7
hospital administration of the perinatal center is
Uterus saved 8
currently negotiating appropriate remuneration
for these services with the Polish national payor.
Legal/malpractice is another risk. In Poland,
physicians are covered by a hospital malpractice
transport was not possible, it was critical that an insurance contract, but theoretically this covers
appropriately trained senior obstetrician from services provided only within the premises of
the perinatal center be taken to the patient at the hospital. Thus, our postpartum hemorrhage
the remote location, along with specialist SWAT team is not covered by malpractice
supplies that the local hospital did not have. insurance while performing the intervention in a
In order to provide appropriate coverage at all remote location. Again, this is not sustainable
times every day of the year, a team of highly on a long-term basis, as these cases are the most
trained and skilled obstetricians is ready and difficult ones and legal proceedings are more
available in the perinatal center (a postpartum likely than after a physiologic delivery. Attempts
hemorrhage SWAT team). Because severe post- are now being made to resolve this issue
partum hemorrhage is rare, every member of the and introduce a malpractice insurance scheme
postpartum hemorrhage SWAT team should similar to that of the ambulance services or the
take every opportunity to observe and/or per- Good Samaritan Act in the United States.
form most, if not all, of these operations as well
as the simpler interventions to get the appropri-
ate training and familiarity with the surgical CONCLUSIONS
technique.
(1) It is possible to ‘eradicate’ maternal mortal-
With regard to management approaches,
ity from postpartum hemorrhage in a large
a number of methods were used, including a
population.
combination of the well-known surgical ligation
methods of the uterine artery, uterine branch of (2) Programs aiming to ‘eradicate’ maternal
the ovarian artery and the hypogastric artery. mortality from postpartum hemorrhage in
445
467
30 August 2006 14:26:11
large populations should encompass the 3. Allam MS, B-Lynch C. The B-Lynch and other
following: uterine compression suture techniques. Int J
Gynaecol Obstet 2005,89:236–41
● medical staff – one to several highly 4. Hayman R, Arulkumaran S, Steer P. Uterine
experienced surgeons compression sutures: surgical management of
postpartum hemorrhage. Obstet Gynecol 2002;99:
● Availability at all times every day of the
502–6
year of an ambulance or other means of 5. Troszynski M, Kowalska B, Jaczynska R, Filipp E.
medical transport. Zgony matek w okresie ciazy, porodu i pologu w
● Therapeutic aids – recombinant Factor Polsce w dziesiecioleciu 1991–2000 wg czterech
glównych przyczyn. [Maternal mortality from
VIIa (e.g. NovoSeven®) and faster
pregnancy, labor and post-partum complications
absorption profile sutures for the
in Poland between 1991 and 2000 by four major
B-Lynch operation in a ready ‘Post- causes of death]. Gin Pol 2003 LXXIV;269:Suppl
partum hemorrhage kit’. II
● Appropriate support – reimbursement 6. Sobieszczyk S, Breborowicz GH. Rekomendacje
postêpowania w krwotokach poporodowych
contract and malpractice.
Czesd I. Protokól postepowania [PPH manage-
ment recommendations. I. Clinical pathway].
References Klin Perinatol Gin 2004;40:60–3
7. Crombach G. Operative Behandlung schwer-
1. AbouZahr C, Wardlaw T. Maternal mortality in wiegender postpartaler Blutungen. Gynaekologe
2000: estimates developed by WHO, UNICEF 2000;33:286–7
and UNSPA, 2003 8. Roman A, Rebarber A. Seven ways to control
2. World Health Organization. Reduction of Maternal postpartum hemorrhage. Contemp Obstet Gynecol
Mortality. Geneva: World Health Organization, 2003;48:34–53
1999
446
468
30 August 2006 14:26:11
51
POSTPARTUM HEMORRHAGE IN IRAN
P. Rezai, A. Beigi and A. Jamal
Iran, located in the Middle East, is one of the intrapartum or postpartum periods, highlight-
world’s oldest cultures. Although Iran is pres- ing the importance of essential obstetric care.
ently considered to be amongst the so-called These observations are entirely compatible with
developing countries, it was once a renowned those from other developing countries, where
center of wisdom and science, boasting such poor availability and access to medical services
famous names as Avecenna whose thoughts and are considered the primary causes of maternal
works influenced much of the world for mortalities.
centuries. Between 1997 and 2000, the annual num-
In 2005, the official ‘Statistical center of bers of reported obstetric deaths were 162, 170,
Iran’ cites Iran’s population as 60 055 488. The 214 and 212, respectively. Because these num-
number of registered live births 961 572 in bers are considerably less than those reported in
20051. During these years, approximately the 1996 survey, they suggest, but do not prove,
two-thirds of births took place in urban areas a possible inadequacy of that survey system and
and one-third in rural communities, villages or the potential for underreporting. In any event,
in the countryside. The 1996 live birth rate per the main causes of deaths had equal proportions
100 000 population was 37.4, placing Iran in a in the years under consideration.
transitional zone between developing countries Despite these limitations, the surveys yielded
(200 per 100 000) and industrialized nations the following findings:
(20 per 100 000).
(1) During 2000, 24.5% and 31.5% of deaths,
Due to the desire of Iran’s government to
respectively, occurred among women older
achieve the United Nations Millennium Devel-
than 35 years of age and those with at least
opment Goals and to identify the main causes of
four pregnancies.
mortality among neonates and mothers which
would affect strategies toward public health (2) As many as 57% of the women who died
promotion, a National Committee of Maternal in 2000 had either basic or no education
& Neonatal Mortality Reduction was formed in whatsoever, a circumstance that was more
1995. In addition, a Reproductive Age Mortal- prominent in rural populations.
ity Survey (RAMOS) was designed by the
(3) Approximately two-thirds of deaths during
Maternal Health Unit of Iran’s Ministry of
2000 took place among the rural popula-
Health & Medical Education to deal with every
tion, a figure which accentuates the impor-
reported case of death related to obstetric com-
tance of creating health-care facilities in
plications. Some of the information collected
underprivileged regions.
and disseminated by this committee is pre-
sented below. (4) Despite a decrease from 44% in 1996 to
During 1996, a total of 382 deaths were 19.5% in 2000, delivery by inexperienced
recorded as being directly related to obstetric and less than fully trained birth personnel
complications. The main causes for these (‘Ghabele’ in the local language) remains an
deaths were hemorrhage, eclampsia, cardio- important factor associated with maternal
vascular disorders and puerperal infections. mortality and morbidity during this time
Most hemorrhagic events occurred during the interval.
447
469
30 August 2006 14:26:11
(5) Whereas 43% of deaths occurred in the available, are rare and are primarily due to
hospitals in 1996, this number reached delayed referral from underprivileged areas.
68.5% in 2000, a change which might One more important factor in the areas with
be indicative of either declining quality of poor access or unavailability of medical services
services in hospitals or increased hospital is the high prevalence of puerperal infections
admissions. as a leading cause of maternal death. Delayed
diagnosis and limited access to antibiotics are
The shortcomings of the survey system between possible contributing factors. Moreover, in a
1996 and 2000 led to the evolution of a revised survey between 1998 and 2000, tetanus vacci-
national system called the National Maternal nation coverage among pregnant women was
Mortality Surveillance System in 2000, which only 80%2. So, in the deprived areas, it is esti-
was a dynamic surveillance system of inputs, mated that maternal mortality due to infectious
interpretations and feedbacks. The main char- etiologies is only second to hemorrhagic events
acteristics of this system included a standard and surpasses eclampsia. Moreover, these
definition of pregnancy-associated deaths, patients usually carry a more unfavorable
regional surveillance systems, reliance on prognosis as compared to those with hemor-
national registration systems that recorded rhagic events due to late referral and limited
every death, questionnaires regarding the cir- therapeutic options.
cumstance of every death, acquiring data from Hemorrhagic events have been the most
multiple sources, timely gathering of informa- common cause of maternal mortality in Iran
tion, precise recognition of causes of death and the postpartum hemorrhage is the most fre-
(especially preventable), dissemination of quent type. Unfortunately, there are no official
results and feedback and, finally, the design population-based statistics about postpartum
and use of appropriate interventions to resolve hemorrhage, and only two cross-sectional
shortcomings. studies in university-based hospitals described
Utilizing the new surveillance system, the postpartum hemorrhage. Although these stud-
total number of reported deaths between 2001 ies might not reflect an exact view of post-
and 2004 showed a significant increase com- partum hemorrhage in Iran, they might be
pared to previous reports, a change which may helpful in some aspects.
be indicative of a higher adequacy and reliability In the first study by Sadeghi and colleagues,
of the new system. During 2001, 2002, 2003 among 18 134 women undergoing delivery,
and 2004, a total of 222, 308, 332 and 278 cases 141 patients with postpartum hemorrhage
of maternal mortality were reported. Table 1 were identified in Tehran’s Akbarabadi and
shows the main results of this survey. Firoozgar hospitals in 1993. This represents a
The main causes of maternal deaths were frequency of 1%. Of these occurrences, 90%
as follow: hemorrhage (34.8%), eclampsia occurred in patients undergoing normal sponta-
(16.7%), embolic events (10%), infection neous vaginal delivery and 10% in patients
(9.1%), cardiovascular events (6.2%), other undergoing Cesarean section. About two-thirds
causes (12.5%), and unknown (10.7%)2. of cases of postpartum hemorrhage occurred in
As mentioned previously, these statistics the first and second pregnancies. While 91% of
(extracted from Maternal Health Unit of cases were early cases of postpartum hemor-
Iran’s Ministry of Health & Medical Education rhage, 9% were late. Approximately two-thirds
publications) only represent registered cases of of cases of postpartum hemorrhage were mild,
maternal mortality and the actual rates might be and one-third was either moderate or severe.
higher. Moreover, deaths due to hemorrhagic The etiologies of postpartum hemorrhage in
events are primarily attributable to poor access this study were uterine atony (38%), retained
or unavailability of health-care facilities or products of conception (38%), lacerations
delayed referral, which usually happens in the (8%), prolonged stage 3 of labor (4%), puer-
rural and underprivileged areas. Deaths due to peral infection (1.4%), uterine rupture (1.4%),
obstetric hemorrhagic events in urban areas, placenta accreta/increta (1.4%), hematoma
where appropriate medical services are readily (1.4%), etc.
448
470
30 August 2006 14:26:11
Postpartum hemorrhage in Iran
Table 1 Characteristics of maternal mortality in Iran, 2000–2004
Percentage of women
2001 2002 2003 2004
Age
> 18 and < 35 years 76 71 77 71
< 18 years 1 2 3 2
> 35 years 20 25 20 27
Residency of mother
Urban 42 44 45 44
Rural 58 56 55 56
Timely report of death 36 80 92 88
Proposing a plan to prevent similar deaths 32 76 62 57
Applying the plan to prevent similar deaths n/a 53 45 50
Identifying preventable causes of death n/a 76 66 63
Number of pregnancies ≥ 5 29 30 20 24
Pregnancy interval < 3 years 10 16 20 23
High-risk mother since start of pregnancy 60 73 67 71
Death during pregnancy 15 16 18 9
Death during delivery 4 8 4 9
Death after delivery 80 75 76 78
Mother delivered at home 36 22 20 15
Mother delivered at hospital 61 73 78 78
Maternal death at home 19 11 9 9
Maternal death on way to hospital 10 11 13 10
Maternal death in hospital 69 75 73 80
Mother delivered by obstetrician 46 56 56 69
Mother not delivered by obstetrician 36 36 40 25
Mother delivered by herself 17 7 4 6
Mother delivered by normal spontaneous vaginal delivery 57 48 52 43
Mother delivered by Cesarean section 42 50 48 55
Delayed family decision causing death 19 38 37 36
Delayed referral causing death 5 23 29 32
Delayed hospital treatment causing death 74 37 39 37
No predisposing factor was found in 62% uterine rupture, one case with uterine atony
of patients, but in 13% and 9%, respectively, and one case with placenta accreta and uterine
induction and multiparity were recognized as rupture3.
predisposing factors, especially in those with It is important to note that this study was
uterine atony. Approximately two-thirds of conducted 13 years ago; at that time, ultra-
patients received more than one treatment. sonography was not available in the centers, and
Most patients with uterine atony were success- those suspected of having retained products
fully treated with uterine massage and oxytocin, of conception would undergo D&C without
whereas most of those with retained products imaging documentation. In addition, poor
of conception were treated with dilatation management of the third stage of labor may
and curettage (D&C). Six patients (4.2%) who have been a contributing factor.
underwent emergency hysterectomy included The second study by Beigi and colleagues
two cases of placenta accreta, two cases of investigated 74 cases of early moderate to severe
449
471
30 August 2006 14:26:12
postpartum hemorrhage among 5601 patients referred patients usually carried a more unfavor-
undergoing delivery in Tehran’s Arash hospital able prognosis4.
between 2001 and 20024. Here, a frequency Whereas it is recognized that neither of these
of 1.3% postpartum hemorrhage was observed studies reflect an exact picture of postpartum
over 3 years. The most common etiologies were hemorrhage in Iran, they reflect the presence
uterine atony (60%), lacerations (23%) and of postpartum hemorrhage in three university-
retained products of conception (16%). Of based hospitals of the capital. In this regard,
those patients with postpartum hemorrhage, they complement literature from other parts of
77% and 70%, respectively were between 20 the world where population-based statistics are
and 35 years of age and between 38 and 40 absent. They also reflect the actual problems
weeks of gestational age. Nulliparity and that exist in developing countries in the recent
multiparity were almost evenly distributed, and past that are continuing to the present day.
two-thirds of patients underwent normal spon-
taneous vaginal delivery, in contrast to 32%
References
who underwent Cesarean section. Four
patients delivered macrosomic babies (5.4%). 1. http://www.sci.org.ir/farsi/default.htm
Multiple gestation was present in two patients 2. Delavar B, Jalilvand P, Azemikhah A, et al.
(2.7%) but polyhydramnios was present in National Maternal Mortality Surveillance System.
only one of them. One patient had a previous Tehran: Iran’s Ministry of Health & Medical
Education, Family Health & Population Office,
history of postpartum hemorrhage. The dura-
Maternal Health Unit, 2002: 1–9
tion of labor was normal in 88% of patients,
3. Sadeghi F. An evaluation of postpartum hemor-
prolonged in 11% and short in 1%. One-third of rhage in Firoozgar and Akbarabadi Hospital in
patients were induced by oxytocin and no 1993. Iran University of Medical Sciences
induction method was used in the remaining 4. Beigi A, Nooroozi A, Zarrinkoub F, et al. An
two-thirds. investigation on early moderate to severe post-
Approximately 40% of the cases of postpartum hemorrhage: frequency, etiologies and
partum hemorrhage in this study were referred risk factors in Tehran’s Arash Hospital between
patients, primarily from satellite districts. 2001 and 2003. Tehran University of Medical
Although no maternal mortality was observed, Sciences, 2005
450
472
30 August 2006 14:26:12
52
POSTPARTUM HEMORRHAGE AND MATERNAL
MORTALITY IN NIGERIA
I. A. O. Ujah and I. S. Ejeh
INTRODUCTION FACTORS CONTRIBUTING TO

POSTPARTUM HEMORRHAGE IN
Although specific studies on postpartum
NIGERIA
hemorrhage in Nigeria are scanty, the contri-
bution of postpartum hemorrhage to maternal In Nigeria, as in other parts of the world, post-
mortality is well documented. Almost partum hemorrhage is most commonly caused
four decades ago, a study conducted by by uterine atony. Other causes include retention
Balachandran1 in Kaduna, Northern Nigeria of placenta or placental fragments, trauma to
documented postpartum hemorrhage as the the genital tract, prolonged second stage of
most common cause of maternal mortality, labor, multiple gestations or hydramnios, past
accounting for 25% of all maternal deaths. history of postpartum hemorrhage, antepartum
Many patients were admitted in a moribund hemorrhage, uterine fibroids, mismanaged third
state having delivered their babies several hours stage of labor, and Cesarean section. All are
previously at home. Adewunmi’s 1986 study recounted in detail in later chapters of this book.
from Ibadan reported that postpartum hemor- However, poverty, illiteracy, and unavailability
rhage contributed 18.7% to maternal mortal- of trained medical personnel combine to accen-
ity2. A more recent study from Eastern Nigeria tuate these problems in Nigeria, as do dwindling
reported 2.72% incidence and a case fatality health resources as a result of bad governance.
rate of 3.25% for postpartum hemorrhage3.
In the most recent report (2003), Ijaiya and
MANAGEMENT OF POSTPARTUM
colleagues reported an incidence of 4.5% for
HEMORRHAGE
postpartum hemorrhage in Ilorin. The risk fac-
tors in this study included advanced maternal The prevention of postpartum hemorrhage is
age of over 35 years and grand multiparity4. predicated on its anticipation, and active man-
Uterine atony was the most common cause agement of the third stage of labor. Several
of postpartum hemorrhage, accounting for 183 strategies have prevented or reduced post-
(53.8%) of the cases, the same as was noted in a partum blood loss and decreased the incidence
study from the same center a decade earlier5. of severe postpartum hemorrhage and hence
Similar data are reported from diverse loca- maternal mortality. Active management of the
tions within the country. At the University of third stage of labor with the use of oxytocics
Nigeria Teaching Hospital, Enugu, south-east with or without ergometrine is beneficial. The
Nigeria, hemorrhage was second only to use of oxytocic drugs reduces postpartum hem-
obstructed labor as the cause of maternal mor- orrhage by about 40%. The effective use of con-
tality6, and the most recent study from north traception also reduces the risk of high parity
central Nigeria revealed that hemorrhage was and consequently reduces the incidence of
responsible for 34.6% of maternal mortality7,8. maternal deaths due to postpartum hemorrhage.
451
473
30 August 2006 14:26:12
Here also, these concepts are dealt with in detail to uterine atony. A medical audit of cases of
in chapters that follow. Ideally, every woman in postpartum hemorrhage should be introduced
labor must be closely monitored after childbirth with the aim of identifying the factors associated
for symptoms and/or signs of postpartum with postpartum hemorrhage, in order to deter-
hemorrhage, although this is not yet possible mine the preventive measures necessary.
in Nigeria. In addition, steps should be taken In many ways, Nigeria exemplifies many
to eliminate the unnecessary procedures that other countries in the developing world where
contribute to the high incidence of postpartum the factors working against the hoped for reduc-
hemorrhage such as episiotomy or operative tions in maternal mortality outnumber those
vaginal delivery without clear indications. Apart that actually reduce the problem.
from medical management of postpartum
hemorrhage, the surgical approach is well
documented and discussed in detail elsewhere.
References
1. Balachandran V. Maternal mortality in Kaduna.
PREVENTION OF POSTPARTUM Nigerian Med J 1975;5:366–70
HEMORRHAGE 2. Adewunmi OA. Maternal mortality in Ibadan
Because postpartum hemorrhage is unpredict- City – 1982. West Africa J Med 1986;5:121–7
3. Anya AE, Anya SE. Trends in maternal mortality
able, it is pertinent in countries such as Nigeria
due to haemorrhage at the Federal Medical
to advocate for the promotion of the routine
Centre, Umuahia, Nigeria. Trop J Obstet Gynaecol
active management of the third stage of labor 1999;16:1–5
with oxytocin and the availability and use of 4. Ijaiya MA, Aboyeji AP, Abubakar D. Analysis
misoprostol when oxytocin is not available. of 348 consecutive cases of primary postpartum
Training and re-training of skilled birth atten- haemorrhage at a tertiary hospital in Nigeria.
dants on active management of labor will help J Obstet Gynaecol 2003;23:374–7
to reduce maternal hemorrhage-related morbid- 5. Adetoro OO. Primary post-partum haemorrhage
ity and mortality. Extensive governmental cam- at a university hospital in Nigeria. West Afr J Med
paign efforts should be directed at sensitizing 1992;11:172–8
the community to institutional deliveries where 6. Chukudeblu WO, Ozumba BC. Maternal
mortality at the University of Nigeria Teaching
adequate monitoring is ensured such that
Hospital, Enugu: a 10-year survey. Trop J Obstet
prolonged labor is avoided
Gynaecol (Special Edition) 1988;1:23–6
7. Ujah IAO, Aisien OA, Mutihir JT, Vanderjagt DJ,
CONCLUSION Glew RH, Uguru VE. Factors contributing to
maternal mortality in North-Central Nigeria
Postpartum hemorrhage remains a major cause (1985–2001). Afr J Reprod Health 2006;9:27–40
of obstetric morbidity and mortality in Nigeria. 8. Ujah IAO, Aisien OA, Mutihir JT, Vanderjagt DJ,
Active management of the third stage of labor Glew RH, Uguru VE. Maternal mortality among
for high-risk pregnancies is advocated to reduce adolescent women in Jos, North-Central, Nigeria.
the incidence of postpartum hemorrhage due J Obstet Gynaecol 2005;25:3–6
452
474
30 August 2006 14:26:12
53
POSTPARTUM HEMORRHAGE IN ASIAN COUNTRIES:
AVAILABLE DATA AND INTERPRETATION
S. J. Duthie
INTRODUCTION domestic product of Hong Kong increased

14-fold between 1966 and 19853, while the
Asia is one of the world’s largest continents,
maternal mortality ratio dropped nine-fold
stretching from the Arctic Ocean to the Equator
between 1961 and 19854.
and from Sumatra and Borneo in the south-east
In September 2000, representatives from 189
to the Suez Canal in the south-west. Its north-
nations met at the United Nations Millennium
western boundary comprises the Ural Moun-
Summit in New York and endorsed the
tains and the Ural River. As such, Asia is home
Millennium Declaration with its eight goals.
to approximately two-thirds of the world’s pop-
The Millennium Declaration represents a global
ulation, contains its two largest nations (China
agenda for the start of the twenty-first century
and India), is home to three of the longest rivers
to promote human development and to reduce
in the world (Ob, Yangtze and Amur) and
global inequalities. Goal number 5 was to
comprises deserts as well as paddy fields
improve maternal health. Some Asian countries,
and coconut plantations.
such as Thailand and China, have already pub-
Within Asia, postpartum hemorrhage is a sig-
lished progress reports on their achievement of
nificant cause of maternal mortality and mor-
the Millennium Development goals5,6. Since
bidity, although sharp differences exist across
that time, there have been encouraging signs in
Asian countries in the maternal mortality ratio,
the management of postpartum hemorrhage
which is itself a measure of socioeconomic
in Asia. Two issues are very clear. First, Asian
well-being1. The major causes of maternal mor-
countries are making strident efforts to control
tality and the manner in which the maternal
maternal mortality due to postpartum hemor-
mortality ratio has fallen also vary between
rhage. Second, it is clear that both the Asian
Asian countries. Obstetric hemorrhage is often
gross domestic product and population num-
the most common cause of maternal deaths
bers will continue to rise. How these obviously
within Asia. The recognition of this fact and
intertwined phenomena will affect maternal
improvements in care, targeted at the manage-
deaths from hemorrhage will depend on factors
ment of women with hemorrhage, are both
other than the gross domestic product alone.
important challenges facing governments as
well as medical authorities.
A striking feature about Asia in recent
KEY ISSUES
decades is the increase in the gross domestic
product of most nations, and the fact that this The definition of postpartum hemorrhage
advance was not always accompanied by a varies. The standard definition is blood loss
decrease in maternal deaths from hemorrhage. in excess of 500 ml or more within 24 h after
Even in Japan, which has a modern and mature delivery. However, Global Burden of Disease
economy, obstetric hemorrhage is the most 2000 defines postpartum hemorrhage only when
common cause of maternal mortality, and inad- the blood loss exceeds 1000 ml or more7. In
equate obstetric services have been blamed for Thailand, Prasertcharoensuk and colleagues
maternal deaths2. In stark contrast, the gross studied 228 pregnant women and reported that
453
475
30 August 2006 14:26:12
the actual incidence of primary postpartum loss was proportional to the measured blood
hemorrhage (defined as a blood loss in excess of loss (Figure 1). This study highlights the fact
1000 ml) was 3.51% by direct measurement of that blood loss during a so-called normal deliv-
the blood loss8. Other authors, however, report ery may be considerable. Using the standard
that visual estimation of blood loss would have definition of primary postpartum hemorrhage,
identified postpartum hemorrhage correctly in 11 out of 62 patients (17.4%) had unnoticed
only 0.44% of patients8. An earlier report primary postpartum hemorrhage and six
from Asia demonstrated significant differences women (10%) developed postpartum anemia.
between measured and estimated blood loss Studies on postpartum hemorrhage are lim-
at normal deliveries9. In a study on normal ited by variations in the definition, differences
patients who received standard antenatal care in between visual estimation and measured blood
Hong Kong, blood loss during normal delivery loss, and (in many countries) the sheer difficulty
was measured in 37 primiparas and 25 multi- of collecting data from widespread and remote
paras. In primigravidas, the mean estimated areas. Many papers report the incidence and
blood loss was 260 ml and the mean measured management of obstetric hemorrhage, but
blood loss was 401 ml. In multiparas, the mean include antepartum hemorrhage, postpartum
estimated blood loss was 200 ml and the mean hemorrhage and secondary postpartum hemor-
measured blood loss was 319 ml. In both rhage. The widespread lack of a confidential
groups, the mean estimated blood loss was system of enquiry into maternal death with
significantly lower (p < 0.05) than the mean published findings adds to the difficulty of
measured blood loss. The size of the discrep- ascertaining accurate figures in many Asian
ancy between measured and estimated blood countries.
600
Difference between measured and estimated blood loss (ml)
500
+2 SD
400
300
200
mean
100
-100
-200 -2 SD
-300
100 200 300 400 500 600 700 800 900 1000
Measured blood loss (ml)
Figure 1 Plot of the difference between measured and estimated blood loss against the measured blood
loss. Reprinted from Duthie SJ et al. Eur J Obstet Gynaecol Reprod Biol 1990;38:119–24, with kind
permission of Elsevier9
454
476
30 August 2006 14:26:13
Postpartum hemorrhage in Asian countries
ASIAN DATA 200311. More importantly, as a percentage of

causes of maternal mortality, postpartum hem-
As postpartum hemorrhage continues to be a
orrhage increased from 4.7% to 24.6% between
major cause of maternal death in the developing
1995 and 2003 (Figure 2). During the same
and the developed world10, this section details
period of time, obstetric embolism (amniotic
available data (in no particular order).
fluid, air and septic embolism as well as pulmo-
nary embolism) decreased in incidence as a
cause of maternal mortality.
Japan
Nagaya and colleagues studied 219 cases
Japan has the largest and most sophisticated of maternal death in Japan between 1991
economy in Asia and a medical system that was and 19922. The purposes of this study were to
substantially revised after World War II. Des- identify causes of maternal mortality, examine
pite these advantages, the number of maternal attributes of treating facilities associated with
deaths from postpartum hemorrhage increased maternal mortality and assess the presence of
in number from four to 17 between 1995 and preventable factors. Of the 230 maternal deaths

0OSTPARTUMHEMORRHAGEISNOWTHEOUTSTANDING
CAUSEOFMATERNALMORTALITYIN*APAN
#AUSESOFDEATH)#$

/THERDIRECTOBSTETRICCAUSES
/BSTETRICEMBOLISM

0OSTPARTUMHEMORRHAGE

0REPARTUMHEMORRHAGENOT
ELSEWHERECLASSIFIED

0LACENTAPREVIAANDPREMATURE
.UMBEROFDEATHS
SEPARATIONOFPLACENTA
ABRUPTIOPLACENTAE
%DEMAPROTEINURIAAND
HYPERTENSIVEDISORDERSIN
PREGNANCYCHILDBIRTHANDTHE
PUERPERIUM

%CTOPICPREGNANCY

9EAR
Figure 2 Maternal deaths and percentages by main causes, 1995–2003 (Statistics and Information
Department, Minister’s Secretariat, Ministry of Health, Labor and Welfare of Japan)
455
477
30 August 2006 14:26:13
which were identified, 197 women died in the 640,641,666). Of the 438 maternal deaths dur-
hospital and had medical records available for ing the study period, 150 (34%) were due to
review, 22 died outside of a medical facility but hemorrhage. This compares with 89 (20%) due
medical records were also available, and, for 11 to gestational proteinuric hypertension and 60
women, no records were available. Hemorrhage (14%) due to ectopic pregnancy. Comparison
was identified as the most common cause of of the distribution of maternal deaths by cases
these deaths, occurring in 86 (39%) of the 219 between two periods (1961–1965 (inclusive)
women for whom records were available. The and 1981–1985 (inclusive)) showed an 86%
overall maternal mortality ratio was 9.5 per reduction in deaths due to hemorrhage4.
100 000 total births, a figure more than double Pulmonary embolism was not a major cause
that reported from Hong Kong for the period of maternal mortality between 1981 and 1985.
1986–1990 (4 per 100 000 total births12). Further data from Hong Kong on obstetric
Nagaya and colleagues found that 37% of the hemorrhage and pulmonary embolism reveal an
maternal deaths occurring in health-care facili- interesting observation (Figure 3). During the
ties were deemed preventable and another 16% time period 1986–1990 (inclusive), the most
possibly preventable. Among the preventable common cause of maternal death was pulmo-
deaths, most were attributed to the fact that nary embolism12. Although hemorrhage during
only one physician worked both as the obstetri- pregnancy and childbirth accounted for the
cian and anesthetist. Nagaya and colleagues same proportion (34%) of maternal deaths
conclude that inadequate obstetric services are during both time periods, it was no longer the
associated with maternal mortality in Japan. most common cause of maternal mortality in
Hong Kong. However, it is still a cause for
concern that the proportion of deaths due to
Hong Kong
hemorrhage had not diminished.
Data from Hong Kong covering the period These are important lessons for the rest of
1961–1985 show that the major cause of Asia, and indeed for the rest of the world. The
maternal death was hemorrhage during preg- experience in Hong Kong between 1961 and
nancy and childbirth (ICD Ninth Revision 1985 clearly shows that, as the gross domestic
89 150
5
60
8
16
2
123
Maternal deaths 1961–1985 Maternal deaths 1986–1990
Hemorrhage Hypertension Ectopic pregnancy

Sepsis Pulmonary embolus Abortive outcome
Other causes
Figure 3 There were no maternal deaths from either hypertension or ectopic pregnancy between 1986
and 1990 in Hong Kong. This is a significant achievement compared with the previous 25 years.
Hemorrhage during pregnancy and childbirth (ICD 640, 641, 666) remains a significant cause of maternal
mortality, whereas the most common cause of maternal mortality during the study period was obstetric
pulmonary embolism (ICD 673). Reproduced from Duthie SJ et al. Br J Obstet Gynaecol 1994;101:906–7,
with the kind permission of the Royal College of Obstetricians and Gynaecologists
456
478
30 August 2006 14:26:13
product of Hong Kong rose, the maternal mor- out of 105 (25.4%). However, in the succeed-
tality ratio fell. The exact number of deaths ing quinquennium, 1992–1997, sepsis only
due to pulmonary embolism during 1961–1985 accounted for 22 out of 107 deaths (13.5%) and
was unavailable4, but it was recognized as a sig- hemorrhage increased to 45 out of 107 (28%).
nificant cause of maternal mortality in Hong Hemorrhage was also shown to be the leading
Kong, indeed more common than hemorrhage cause of maternal mortality at a hospital in West
between 1986 and 199012. It is important to ask India, where 24.6% of maternal deaths were
why pulmonary embolism has overtaken hemor- due to hemorrhage16. Data from north-eastern
rhage as the main cause of maternal mortality in India, covering a 5-year time period between
Hong Kong. Part of the answer may well be due June 1998 and June 1993, also identified hem-
to a change in civil practice. Since 1986, all orrhage as the most common cause of maternal
maternal deaths were investigated by a coro- mortality, accounting for 27.65% of deaths17. In
ner’s post-mortem. In many parts of Asia, this a detailed and critical analysis, Mukherji and
is not possible, and thus it is very difficult to colleagues show that 58.0% of the cases of
obtain a perspective on individual cases of fatal maternal death due to hemorrhage were actually
postpartum hemorrhage. due to postpartum hemorrhage, mainly stem-
ming from the lack of provision of emergency
transport at community level17.
India
Chhabra and Sirohi studied trends in maternal
Pakistan
mortality specifically due to hemorrhage over 20
years in rural India13. Obstetric hemorrhage was Evidence from Pakistan identifies hemorrhage
the contributory cause of maternal mortality in as being the most common cause of maternal
19.9% of the cases, and the leading cause of mortality in many regions. In one study, direct
fatal hemorrhage was postpartum hemorrhage causes of maternal death accounted for 78.1%
due to an atonic uterus. Other causes included of cases and the most common cause was hem-
ruptured uterus, placental abruption and orrhage18. A study in a tertiary care hospital in
retained placenta. It is reasonable to ask why so Pakistan also identified hemorrhage as the most
many women died due to an atonic uterus. The common cause of maternal mortality19. One-
answer could be as simple as the high number third of the cases of maternal mortality were due
of home births (60%) and the lack of medical to hemorrhage (nine out of 26), and the authors
attention and/or use of uterotonic agents. conclude that most maternal deaths were
A study carried out in Calcutta, India preventable.
between 1995 and 1997 (inclusive) identified a
maternal mortality ratio of 686.67 per 100 000
Thailand
births, and hemorrhage was the second most
common cause (16.75%), with toxemia as the Data from Thailand show that the maternal
leading cause (53.2%) of maternal mortality14. mortality ratio in a government hospital in
Several reports from India describe hemor- Thailand was 19.18 per 100 000 births between
rhage as the leading cause of maternal death in 1985 and 1998 (inclusive). The most common
specific regions of the country. Comparison of cause of death was identified as hemorrhage and
two quinquennia, 1987–1991 and 1992–1997, the authors concluded that a significant number
shows that the number of deaths due to of maternal deaths were preventable20.
hemorrhage has increased and hemorrhage has
increased as a proportion of direct causes of
Indonesia
maternal mortality15. The data from a govern-
ment hospital in Peninsular India show that the Between 1995 and 1999, a district-based audit
main causes of maternal death were hemorrhage of maternal deaths in South Kalimantan, Indo-
and sepsis. Between 1987 and 1991, sepsis nesia demonstrated that 41% of cases of mater-
accounted for 47 out of 105 maternal deaths nal death were due to hemorrhage21. Supratikto
(31%), whereas hemorrhage accounted for 36 and colleagues provide a searching discussion of
457
479
30 August 2006 14:26:14
some of the key issues. Their audit of maternal The leading causes of maternal deaths were
death in South Kalimantan led to changes in the hemorrhage followed by postpartum infections
quality of obstetric care in the district, by devel- and pregnancy-induced hypertension. Over the
oping the concept of accountability of both 3 years between 1985 and 1988, improvements
health providers and policy-makers. Supratikto in the health-care system were achieved in terms
and colleagues also underlined the need to of strengthening referrals between village health
incorporate scientific evidence to the review stations and township hospitals, establishment
process. These are important points as there is of case management procedures for caring for
ample evidence from some parts of Asia that women with postpartum hemorrhage, severe
maternal mortality is not always investigated pregnancy-induced hypertension, amniotic
robustly22. fluid embolism, shock and neonatal asphyxia.
These improvements were followed by signifi-
cant reductions in the maternal mortality ratio
Malaysia
in pilot areas25.
A report from Malaysia with a 10-year study
period between 1981 and 1990 demonstrated
Saudi Arabia
a maternal mortality rate of 74 per 100 000
births23. The most common cause of maternal In Saudi Arabia, the leading cause of maternal
death was hemorrhage. Other causes included death was hemorrhage in 43.75% of patients at
hypertension, embolism and sepsis during the a university hospital between 1983 and 200226.
decade between 1981 and 1990. Risk factors The authors carried out a detailed analysis of
from maternal deaths were lack of antenatal the underlying cause of each maternal death and
care, maternal age above 40 years, grand multi- analyzed potentially avoidable factors. Risk fac-
party, and Indian ethnic origin. A postmortem tors for maternal death were maternal age in
examination was performed in only 8.2% of the excess of 35 years, a parity of 5 or greater, and
women who died. iron deficiency anemia26. The main avoidable
Similarly, a report from Malaysia describes factors were identified as the failure of patients
postpartum hemorrhage, obstetric embolisms, to seek timely medical advice and to follow
trauma and hypertensive disorders of pregnancy medial advice.
as the main causes of 131 sudden maternal
deaths23. The sudden maternal deaths com-
Sri Lanka
prised 20.6% of all maternal deaths. Twenty
mothers died after a Cesarean section and the Although hemorrhage is the leading cause of
need for training in the emergency care of maternal death in several Asian reports, it must
women who collapse is emphasized. be emphasized that it is not always the major
cause of maternal mortality. In Sri Lanka, for
example, a study covering an 11-year period
China
identified 103 maternal deaths and the main
Data from China also underline the contri- causes were genital tract sepsis (26%), hyper-
bution of postpartum hemorrhage to maternal tension in pregnancy (24%) and obstetric hem-
mortality. A case-controlled study of maternal orrhage (20%)27. Care was considered to be
mortality that was conducted in two rural prov- substandard in 79% of the deaths, and the sub-
inces of China (Henan and Jiangsu) identified standard care was felt to have influenced the
postpartum hemorrhage as the major cause of outcome of maternal deaths in 7% of the cases.
maternal death in both provinces24. Here, the
large proportion of deaths occurred during
Bangladesh
the journey between the woman’s home and the
health-care facility, a fact not well emphasized Here also, hemorrhage did not head the list of
in reports from other countries. A study carried causes of maternal death. Eclampsia (34.3%)
out in Myiun County in China estimated that was first and hemorrhage (27.9%) second in a
27.3% of maternal deaths were unreported25. study of 8562 maternal deaths28.
458
480
30 August 2006 14:26:14
Singapore whereas the population size and the number of

legal abortions rose over the period 1961–1985.
The situation in Singapore is encouraging, in
Therefore, parity among Hong Kong women
that a close scrutiny of maternal deaths in that
fell during this period4. Studies carried out in
highly industrialized and compact country
Hong Kong showed that maternal mortality
shows that most maternal deaths were not due
rates vary greatly by parity, i.e. 41 per 100 000
to the traditional direct causes of hemorrhage,
total births if the parity was zero, and 82 per
sepsis, embolism or hypertensive disease29.
100 000 total births if the parity was more than
Over a 7-year period between 1986 and 1992,
530. These observations support the speculation
the maternal mortality rate at the National Uni-
that the fall in parity contributed to the fall in
versity Hospital in Singapore was 34.4 per
maternal mortality rate, but the available data
100 000 births, including incidental deaths.
cannot show that it was the high-parity women
Most of the women who died had underlying
who were the major group who died from
medical disorders that placed them at high risk
hemorrhage in 1961, and who were no longer
of mortality.
present in 19854.
Once a maternal death takes place, signifi-
INTERPRETATION cant differences in practice are present in differ-
ent parts of Asia. In Hong Kong, for example,
Postpartum hemorrhage is a significant cause of
all maternal deaths have been investigated by a
maternal mortality in Asia, regardless of occa-
coroner’s post-mortem since 198612. By con-
sional exceptions. The numbers are large and
trast, Abdullah and Raj-Hasim pointed out that
generally on a downward trend. However, some
a post-mortem examination (of any type) was
regions still have an extremely high maternal
carried out in only 8.2% of cases of maternal
mortality ratio. A direct comparison between
death in a university obstetric unit in Malaysia
different countries and the regions within each
over a 10-year study period between 1981 and
country is not feasible for the following reasons:
199022. It is also important to note that the data
(1) There is a variation in the definition of from many countries are already several years
postpartum hemorrhage; old. Medical practice, changes in gross domestic
product and the health of individual women
(2) Different data are not necessarily matched
have altered in recent years and all these issues
over the same time frame;
have an impact on postpartum hemorrhage and
(3) There are major differences in the sizes of its consequences.
the study group in different papers; A common theme in most of the Asian
papers is that postpartum hemorrhage and its
(4) Obstetric hemorrhage is not divided into
sequelae are preventable. Education of both the
antepartum, postpartum and secondary
health-care workers and the women themselves
postpartum hemorrhage in many papers;
is emphasized. This leads to two questions:
(5) Some studies rely on indirect observations
and verbal autopsies, whereas other studies (1) How can matters be improved further?
involve detailed analysis of individual (2) Once postpartum hemorrhage is controlled
maternal deaths; as a cause of maternal death, will other
(6) The depth of discussion on preventable causes of maternal death become more
factors varies between reports. significant?
The study from Japan2 investigated prevent- In order to reduce the incidence and conse-
ability of maternal deaths as determined by a quences of postpartum hemorrhage, it is essen-
42-member panel of medical specialists. This tial to obtain precise data. It is of concern that a
was a well-organized and robust study. On the significant number of women die outside the
other hand, several reports rely on estimates and medical facility and a significant number of
speculation. In Hong Kong, the total number maternal deaths did not have records available
of births per annum and birth rate dropped, for scrutiny in a country as technologically
459
481
30 August 2006 14:26:14
advanced as Japan2. Japan has a low maternal maternal death is certified, investigated and dis-
mortality ratio, a high gross domestic product cussed. Analysis of death certificates is useful,
and a high rate of female literacy. Although the but is simply not enough. The practice in Hong
data indicate that Japan is ahead of most other Kong whereby the coroner investigates mater-
Asian countries, there is still room for improve- nal deaths is an example to follow. Once an ade-
ment. The rise in maternal deaths due to quate investigation is carried out, each maternal
postpartum hemorrhage in Japan is a major death (whether due to postpartum hemorrhage
cause for concern. In Hong Kong and Singa- or other causes) must be the subject of a
pore, the maternal mortality rates have fallen to multidisciplinary meeting. The appropriate
admirably low levels. However, a system of con- conclusions and areas for improvement must be
fidential enquiries into maternal deaths is long disseminated within the health-care system.
overdue in Hong Kong12. Journals that publish articles on maternal mor-
Thailand reported a large reduction in mater- tality and postpartum hemorrhage must be
nal mortality ratio between 1990 and 19996. encouraged to do so only if the authors meet
The proportion of deaths due to hemorrhage strict criteria of definition, sources of data and
has also fallen sharply. Similarly, China has an analysis of preventable factors with robust
reported a sharp reduction in the maternal recommendations for change. There is no
mortality ratio between 1990 and 2001, thereby doubt that the problem of postpartum hemor-
demonstrating China’s progress to achieving rhage has been tackled throughout Asia. How-
Goal number 55. Nevertheless, the report from ever, there remains a need for education,
China highlights the differences in the maternal certification of deaths, uniformity of definition
mortality ratio between western China, with a and critical incident review. There is a signifi-
relatively low stage of development, and the cant rise in the gross domestic product of most
more advanced eastern provinces, including Asian counties and the benefits should be seen
Hong Kong. In India, the maternal mortality in terms of a reduction in the incidence and fatal
ratio and proportion of deaths due to post- consequences of postpartum hemorrhage. The
partum hemorrhage remain unacceptably high. progress achieved so far shows that we need not
In Hong Kong, hemorrhage was the most despair but we should not be complacent.
common cause of maternal mortality between
1961 and 1985. However, pulmonary embolism
was the most common cause of maternal mor- References
tality between 1986 and 199012. This leads to
the interesting question as to whether or not 1. Hogberg U. Maternal mortality – a worldwide
problem. Int J Gynaecol Obstet 1985;23:463–70
other Asian countries that follow Hong Kong’s
2. Nagaya K, Fetters MD, Ishikawa M, et al.
development would face a similar change in the
Causes of maternal mortality in Japan. JAMA
causation of maternal mortality. 2000;283:2661–714
3. Census and Statistics Department. Estimates of
RECOMMENDATIONS Gross Domestic Product 1966 to 1986. Hong Kong:
Government Printer, 1987
It is crucial that progress be made beyond the 4. Duthie SJ, Ghosh A, Ma HK. Maternal mortal-
stage of simply acknowledging the problem, ity in Hong Kong 1961–1985. Br J Obstet
describing it and emphasizing that the problem Gynaecol 1989;96:4–8
must be resolved. In practical terms, the first 5. Office of the United Nations Resident Coordina-
step is to provide adequate training and educa- tor. Millennium development goals – China’s
progress. 2003
tion of health-care personnel in each country.
6. Office of the United Nations Resident Coordina-
The systems that provide emergency obstetric
tor. Thailand Millennium development goals
care to women with postpartum hemorrhage report. 2004
must be in place within the frame work of clini- 7. Dolea C, Abouzahr C, Stein C. Global burden of
cal governance. Professional responsibility and maternal haemorrhage in the year 2000. Evi-
accountability must be established. The next dence and Information for Policy (EIP). Geneva:
step would be to ensure that each and every World Health Organization, 2003
460
482
30 August 2006 14:26:15
8. Prasertcharoensuk W, Swadpanich U, 20. Kovavisarach E, Sathiraleela B. Maternal

Lumbiganon P. Accuracy of the blood loss mortality in Rajavithi Hospital 1984–1998:
estimation in the third stage of labor. Int J analysis of the cause of death. J Med Assoc
Gynaecol Obstet 1990;71:69–70 Thailand 2001;84:763–7
9. Duthie SJ, Yung GLK, Dong DZ, Chan SYW, 21. Supratikto G, Wirth ME, Achadi E, Cohen S,
Ma HK. Discrepancy between laboratory deter- Ronsmans C. A district-based audit of the causes
mination and visual estimation of blood loss dur- and circumstances of maternal deaths in South
ing normal delivery. Eur J Obstet Gynaecol Reprod Kalimantan, Indonesia. Bull WHO 2002;80:
Biol 1990;38:119–24 228–34
10. Abouzahr C. Antepartum and postpartum haem- 22. Abdullah R, Raj-Hasim R. Reproductive Health
orrhage. In Murray CJL, Lopez AD, eds. Health in Asia and Pacific: Some Facts. Kuala Lumpur,
Dimensions of Sex and Reproduction: the Global Malaysia: Asian-Pacific Resource and Research
Burden of Sexually Transmitted Diseases, Maternal Centre for Women, 1994:14–20
Conditions, Perinatal Disorders, Congenital Anoma- 23. Jagasothy R. Sudden maternal deaths in Malay-
lies. Geneva: World Health Organization, 1998 sia: a case report. J Obstet Gynaecol Res 2002;
11. Mothers’ & Children’s Health Organization. 28:186–93
Maternal and Child Health Statistics of Japan 24. Li Q, Fottler MD. Determinants of maternal
2004. Tokyo: Mothers’ & Children’s Health & mortality in rural China. Health Services Manage-
Welfare Association, 2005:78 ment Research 1996;9:45–54
12. Duthie SJ, Lee CP, Ma HK. Maternal mortality 25. Xu Z. China: lowering maternal mortality in
in Hong Kong 1986–1990. Br J Obstet Gynaecol Miyun County, Beijing. World Health Statistics
1994;101:906–7 Quarterly – Rapport Trimestriel de Statistiques
13. Chhabra S, Sirohi, R. Trends in maternal mor- Sanitaires Mondiales 1995;48:11–14
tality due to haemorrhage: two decades of Indian 26. Al-Suleiman SA, Al-Sibai MH, Al-Jama FE,
rural observations. J Obstet Gynaecol 2004;24: et al. Maternal mortality: a twenty-year survey at
40–3 the King Faisal University Hospital, Al-Khobar,
14. Majhi AK, Mondal A, Mukherjee GG. Safe Eastern Saudi Arabia. J Obstet Gynaecol 2004;24:
motherhood – a long way to achieve. J Ind Med 259–63
Assoc 2001;99:132–7 27. Wagaarachchi PT, Fernando L. Trends in
15. Jayaram VK. Review of maternal mortality. maternal mortality and assessment of substan-
J Obstet Gynaecol Ind 2001;51:80–2 dard care in a tertiary care hospital. Eur J Obstet
16. Sharma N. Maternal mortality – a retrospective Gynaecol Reprod Biol 2002;101:36–40
study of ten years. J Obstet Gynaecol Ind 2001; 28. Rahman MH, Akhter HH, Khan CME, Yusuf
51:60–2 HR, Rochat RW. Obstetric deaths in Bangla-
17. Mukherji J, Ganguly RP, Saha SK. Maternal desh, 1996–1997. Int J Gynaecol Obstet 2002;77:
mortality due to haemorrhage with emphasis on 161–9
post partum haemorrhage. J Obstet Gynaecol Ind 29. Loh FH, Arulkumaran S, Montan S, Ratnam
2001;51:130–3 SS. Maternal mortality: evolving trends. Asia-
18. Jafarey SN. Maternal mortality in Pakistan – Oceania J Obstet Gynaecol 1994;20:301–4
compilation of available data. J Pak Med Assoc 30. Yam A, Ghosh A, Ma HK. Maternal mortality
2002;52:539–44 yet to be minimized. Asia-Oceania J Obstet
19. Begum S, Aziz-un-Nisa, Begum I. Analysis of Gynaecol 1986;12:79–87
maternal mortality in a tertiary care hospital to
determine causes and preventable factors. JAMC
Abbottabad 2003;15:49–52
461
483
30 August 2006 14:26:15
Index
abdominal pregnancy, antepartum hemorrhage 27

case studies 427 anti-compartment syndrome 139
diagnosis 428 anti-shock garments 136–45
incidence 428 antibiotics,
management of placenta 430 use in secondary postpartum hemorrhage 320
postoperative care 430 use in sepsis 405
preoperative preparations 429 anticoagulant therapy,
surgical approach 430 during pregnancy 219–21
timing of intervention 429 in the peripartum period 219–21
treating hemorrhage from 427–31 antiphospholipid syndrome 212
acquired hemophilia 218 AOFOG postpartum hemorrhage initiative 437
active management of third stage of labor 98–104 arteries,
cf expectant management 103 arteriovenous malformations as a pathological finding
definition 98 336
early cord clamping 103 branch patterns of pelvic 277
for midwives 398 embolization of 277–85
joint ICM/FIGO statement and action plan on 6, 7 selective for supralevator hematomas 200
options for 398 torn in uterus 329
postpartum hemorrhage and 19 Asia,
uterotonic agents (see also under specific agents), mortality risk in 3
ergot alkaloids 101 recommendations for management in 460
oxytocin 98 statistics from 453–60
prostaglandins 102 assault and blunt abdominal trauma 76
syntometrine 100 atony, see uterine atony
acute fatty liver of pregnancy 216 autopsy after postpartum hemorrhage mortality 336–8
acute uterine inversion 70–4
adherent placenta, see placenta accreta
advance directives in Jehovah’s Witnesses 152 B-Lynch brace suture compression technique,
Afghanistan, mortality risk in 4 biodata of women undergoing 363, 364
Africa, case series of use in USA 362–70
misoprostol administration by TBAs in Tanzania 158 cf other techniques 365
morbidity in 22 clinical points 294
mortality in 3, 21 cumulative success rate 366
in Nigeria 451–2 literature review 367
age and postpartum hemorrhage 25 method of 288–92
algorithms, normal MRI after 293
adult basic life support algorithm 176 use in secondary postpartum hemorrhage 321
advanced life support for cardiac arrest in adults 177 world-wide reports of 294
for conservative management of placenta accreta 205 Bakri balloon use in uterine tamponade 264
management of hemorrhage from lower genital tract balloon tamponade 268
198 as a test 269
WMMC algorithm for use in B-Lynch suture Bangladesh, statistics on postpartum hemorrhage in 458
technique 368 Benedetti’s classification of hemorrhage 12
amniotic fluid clearance 423 Blakemore–Sengstaken tube 199
AMTSL, see active management of third stage of labor use in uterine tamponade 264
analgesia 29 bleeding from lower genital tract 194–201
anesthetics, choice of in postpartum hemorrhage 175 blood loss,
antenatal care, and reduction in severe adverse maternal assessment of 18, 35–42
morbidity 349 comparison of techniques 40–2
462
484
05 September 2006 12:14:09
Index
direct collection 37 diabetes 26

normal 35 disseminated intravascular coagulopathy 174, 216–18
physiological adaptations to during pregnancy 36 Doppler velocimetry, intrapartum 58
visual assessment 37 drills,
accuracy of 55 labor ward drills 127–35
cf direct measurement 18 preparations for 129
blood product use in resuscitation 172 running the drill 130
blood reinfusion 421 drug-induced thrombocytopenia 213
blunt abdominal trauma 76–8
body mass index and postpartum hemorrhage 25
BRASSS-V drape 39 education,
basic equipment list for hemorrhage training 132
drills 129
Canada, obstetric hemorrhage equipment tray use in 179 for midwives 401
carboprost, use in uterine atony 258 importance in hospital systems for management of
cardiopulmonary resuscitation 176 major hemorrhage 183
catheters used in embolization 282 labor ward drills 127–35
cervical tears, practical teaching 128
as pathological findings 330 principles of adult learning 127
in fixed uterus 329 sample scenario for postpartum hemorrhage with
see also hemorrhage from lower genital tract uterine atony 134
Cesarean section, sample scenario for postpartum hemorrhage without
incidence 188 uterine atony 135
postpartum hemorrhage and previous sections 27 scenario teaching 131
wound dehiscence and secondary postpartum skills 131
hemorrhage 317 Egypt, pilot study with non-pneumatic anti-shock
chemotherapy, use in secondary postpartum hemorrhage garment 142
322 embolization 277–85
China, statistics on postpartum hemorrhage in 458 arterial anatomy for 278
Cho multiple square sutures 295, 296 case study 280
chorioamnionitis 30 catheters used in 282
choriocarcinoma 317, 322 change of protocols needed for use 285
circulating blood volume, complications 284
assessment of 46 in case of double uterus 280
management of changes in 48 materials used for 282, 283
physiology 45 selective arterial embolization in secondary
replenishing 50 postpartum hemorrhage 321
classification of postpartum hemorrhage 11–15, 36 targets of 282
based on causative factors 12, 13 technical aspects 279
based on clinical signs and symptoms 13 technique 280–2
based on quantification of blood loss 12 endometritis 316
conventional temporal classification 11 as a pathological finding 336
pitfalls of 13 EOC project 435
proposed 14 epidemiology, confounding factors in 19
coagulation and placental separation 65 episiotomy,
coagulation factors, changes in during pregnancy 209 postpartum hemorrhage and 30
coagulopathies 26, 173, 317 risk of postpartum hemorrhage and 196
rare 226 see also hemorrhage from lower genital tract
treatment of in secondary postpartum hemorrhage Eptacog-alfa, mode of action 237
323 equipment tray for use in hospitals 179–82
combined deficiency of factors II, VII, IX and X 227 ergometrine,
common iliac artery, accidental ligation 305 isolation of 256
communication, use in uterine atony 258
importance in hospital systems 183, 400 ergot alkaloids,
congenital disorders of hemostasis 221–4 cf oxytocin 99
connective tissue disorders 26 use in active management of third stage of labor 101
consent in Jehovah’s Witnesses 147–52 use through history 256
consumptive coagulopathy 216 ethnicity 25
controlled cord traction, ICM/FIGO joint statement 6 etiology 23
cord clamping 103 EUPHRATES study 109–13
cryoprecipitate 173 Europe,
morbidity in 22
mortality in 20, 21
delivery, out-of-hospital 413–18 European Consensus on Prevention and Management of
method and postpartum hemorrhage 29 postpartum hemorrhage 109–13
determination of blood loss 18 external iliac artery, accidental ligation 305
463
485
05 September 2006 12:14:09
factor VII deficiency 226 resuscitation strategies and techniques in 170–7

factor VIII inhibitors 218 systems for management of major hemorrhage
factor X deficiency 227 183–91
factor XI deficiency 228 hydrostatic condom catheter, use in uterine tamponade
factor XIII deficiency 228 266
fertility after conservative management of placenta accreta hypovolemia, recognition of 171
206 hysterectomy,
fibrinogen deficiency 226 avoidance using postpartum equipment tray 180,
fibroids 27 182
Foley catheter, use in uterine tamponade 265 for placenta accreta 203
fresh frozen plasma 172 for supralevator hematomas 200
fundus, pathological findings in 334 indications for 312
mortality linked to obstetric 308
pathology of uterus after 326–38
gestational thrombocytopenia 210 peripartum 312–15
guidelines, surgical principles of 313
European Consensus on prevention and management use in secondary postpartum hemorrhage 321
of postpartum hemorrhage 109–13
for consent to elective surgery in Jehovah’s Witnesses
150 ICM/FIGO, joint statement and action plan 6
for life-threatening bleeding in unconscious adult idiopathic thrombocytopenic purpura 210
Jehovah’s Witnesses 152 iliac artery (internal), ligation of 299–307
joint ICM/FIGO statement and action plan on active incidence of postpartum hemorrhage 3
management of third-stage labor 6, 7 India,
access to emergency care in 437
achievements of health department 439
Hayman uterine compression suture 295, 296 EOC project 435
health status, postpartum hemorrhage and 3 misoprostol administration in 160
HELLP syndrome 215, 394 progress in combating postpartum hemorrhage in
hematoma, 434–41
management 200 referral systems in 438
paravaginal 195 statistics on postpartum hemorrhage in 457
postpartum 194 Indonesia, statistics on postpartum hemorrhage in 457
hemolytic uremic syndrome 215 induction of labor 27
hemophilia 224 after fetal death 119
acquired 218 postpartum hemorrhage risk and 27
hemorrhage from lower genital tract 194–201 using misoprostol 118
classification 194 infection,
diagnosis 196 after uterine packing 266
incidence 194 as a cause of uterine atony 316
prevention 196 pelvic pressure pack use and 310
risk factors for 196 sepsis
treatment, and postpartum hemorrhage 404–7
algorithm for management 198 as an indication for hysterectomy 313
cervical tear 199 infralevator hematomas, management of 200
infralevator hematomas 200 internal iliac artery,
perineal tear 197 anastomoses of 300
supralevator hematomas 200 illustration of ligation 304
vaginal tear 197 indications for ligation 301
hemostatic disorders, ligation 299–307
acquired 210 surgical techniques,
congenital disorders of 221–4 extraperitoneal approach 301
in pregnancy and puerperium 209–29 general considerations 301
heparin, use in pregnancy 219–21 midline extraperitoneal approach 301
heparin-induced thrombocytopenia 213 postoperative care after 303
hepatitis C, litigation for infection following transfusion surgeon position for 306
380 transabdominal approach 301
history 2 intraoperative autologous blood transfusions 421–6
perspective on postpartum hemorrhage through intraoperative cell salvage 422
history 390 intravenous fluids used in shock 51
HIV, litigation for infection following transfusion 379 Iran, management of postpartum hemorrhage in 447–50
HIV-associated thrombocytopenia 213
Hong Kong, statistics on postpartum hemorrhage in 456
hospitals, Japan, statistics on postpartum hemorrhage in 455
impact of changes systems in outcomes 187 Jehovah’s Witnesses,
labor ward drills 127–35 acceptable blood products 147
postpartum hemorrhage equipment tray 179–82 guidelines for consent to elective surgery in 150
464
486
05 September 2006 12:14:10
Index
guidelines for life-threatening bleeding in unconscious management,

adults 152 hospital systems for management of major
management of elective surgery in 150 hemorrhage 183–91
management of unconscious patient 152 in Nigeria 451
outcome in non-obstetric patients 148 Mexico, pilot study with non-pneumatic anti-shock
outcome in obstetric patients 148 garment 143
use of recombinant factor VIIa in 148 midwives,
antenatal prevention of postpartum hemorrhage
396
Kanga 3 care following postpartum hemorrhage by 399
development of midwifery 2
diagnosis and prevention by 398
labor, intrapartum prevention of postpartum hemorrhage
Doppler velocimetry during 58 397
drills in labor wards 127–35 management of postpartum hemorrhage by 399
duration and postpartum hemorrhage risk 28 response in cases of postpartum hemorrhage
induction and postpartum hemorrhage risk 27 396–402
induction of, role in reducing postpartum hemorrhage and its
after fetal death 119 impact 374
using misoprostol 118 military anti-shock trousers 138
misoprostol, misoprostol 114–21
for induction 118 administration routes 115
in third stage 119 adverse effects 116
multigate spectral Doppler analysis during 59 availability 121
labor ward drills 127–35 in community setting 159
Latin America, mortality risk in 4 restrictions on 5
ligation, cf oxytocin 102
of internal iliac artery 299–307 cf placebo/no uterotonic 102
of uterine and ovarian artery 180 for postpartum hemorrhage prevention 119
litigation cases, for postpartum hemorrhage treatment 4, 119
abandonment (2000) 386 other uses 121
dangerous sidewalk (1908) 377 overview of postpartum hemorrhage studies 162–8
delay in transfusing blood (1984, 1988, 2000) 380 pharmacokinetics 114
disappearing baby (1999) 386 structure 114
discharging patient home too soon (1977) 382 trials in low-resource countries 5
first maternal death to be litigated (1905) 376 use in active management of third-stage labor 102
health insurance (1956) 378 use in first trimester 117, 118
iatrogenic obstetric injury (1955) 377 use in practice 156
inadequate staffing levels (1981) 382 use in second trimester 118
infection following transfusion (1981, 1982, 1985) 379 use in the community setting 158, 160
international perspective 376–87 use in third stage of labor 119
malignant hypertension (1993) 384 use in third trimester for labor induction 118
no autopsy (1982) 382 use in uterine atony 259
no expert medical report (1995) 385 morbidity 21–3
no operation note (1992) 384 consequences for family and society 373
obstetrician on vacation (1961) 381 delayed 372
obstetrician without sufficient experience (1986) 383 effects on children 373
postpartum hemorrhage in an aircraft (1997) 385 severe acute 339–50
postpartum hemorrhage in female dog (2006) 386 the morbidity–mortality continuum 340
postpartum hemorrhage in the pleural cavity (1997) mortality,
385 1847 to 1901 391–2
road traffic accident (1930) 377 consequences for society 374
Sheehan’s syndrome (1977, 1995) 384 effects on children 374
suing the wrong doctor (1982) 383 elimination in Poland 442–5
telephone problem (1909) 377 emotional cost 373
transfusion of wrong blood (1951, 1955, 1972) 378 global 20
unlawful practice of medicine (1907) 376 in Africa 21
unlicensed practice of obstetrics (1963) 382 in France 21
Logothetopoulos pack 308 in Iran 449
low-resource countries, familial consequences of in UK 20
postpartum hemorrhage 372–5 incidence in childbirth 2
lower genital tract trauma in secondary postpartum linked to obstetric hysterectomy 308
hemorrhage 316 reasons for 170
the morbidity–mortality continuum 340
multigate spectral Doppler analysis 59
macrosomia (fetal) 26 multiple pregnancy 26
Malaysia, statistics on postpartum hemorrhage in 458 myonecrosis 331, 332
465
487
05 September 2006 12:14:10
national professional organizations, role of 7 placenta,

Nigeria, abnormal placentation as an indication for
maternal mortality in 451–2 hysterectomy 312
pilot study with non-pneumatic anti-shock garment abnormalities and secondary postpartum hemorrhage
143 317
postpartum hemorrhage management in 451 abnormalities of 80–92
non-pneumatic anti-shock garment 136–45 abruption after blunt abdominal trauma 77
case histories in Pakistan 141 management in abdominal pregnancy 430
pilot studies, manual removal 5
in Egypt 142 pathological findings in 336
in Mexico 143 retained 5, 24, 335
in Nigeria 143 separation 62
potential benefits in low-resource settings 139 placenta accreta 24, 67, 90–1
suggested protocol for use 140 as an indication for hysterectomy 312
North America, conservative management 204
hospital systems for management of major algorithm for 205
hemorrhage in 183–91 complications of 206
morbidity in 22 follow-up after 204
optimal adjuvant therapy 206
prenatal identification 206
obstetricians, diagnosis 203
decision-making by 393 management 203–7
historical perspective on postpartum hemorrhage prenatal identification 206
390–5 preparation for hemorrhage in 185, 186
role in reducing postpartum hemorrhage and its risk of recurrence 206
impact 374 placenta creta in fixed uterus 327
osteoporosis in Sheehan’s syndrome 356 placenta increta 90–1
out-of-hospital deliveries 413–18 placenta membranacea 92
incidence of 414 placenta percreta 90–1
maternal outcomes in 416 placenta previa 67, 80–6
planned and unplanned 414 as an indication for hysterectomy 312
postpartum hemorrhage in 416 diagnosis 81
risk factors in 414 grades of 80
ovarian artery ligation 180 incidence of 81
oxytocin 4, 5 management 83–5
agonists 100 preparation for hemorrhage in 185
cf ergot alkaloids 99 risk factors 81
cf misoprostol 102 placental abruption 86–90
cf no uterotonics 99 as an indication for hysterectomy 312
discovery of 256 clinical picture 87
role in uterine activity 64 complications of 89
use in active management of third-stage labor 98 diagnosis 87
use in low-resource countries 5 grading of 88
use in uterine atony 257 incidence 86
intrauterine fetal death in 90
management 88–9
Pakistan, risk factors 86
case series of non-pneumatic anti-shock garment use platelet disorders, congenital 221
141 platelets 172
statistics on postpartum hemorrhage in 457 pneumatic anti-shock garment 138
panhypopituitarism 357 Poland,
paramedical workers, misoprostol administration by mortality elimination 442–5
160 postpartum hematoma 194
paravaginal hematoma 195 postpartum hemorrhage,
parity 25 antenatal risk factors 25
pathophysiology of postpartum hemorrhage 62–8 choice of anesthetic in 175
patient safety teams 184 classification of 11–15
pelvic pressure pack 308–10 common causes 287
pelvic vasculature, injury to during surgery 317 confrontation by midwife 396–402
pelvis, confrontation by obstetrician 390–5
major pelvic anastomoses 300 definition 3
vasculature of 299 by blood loss 17
percutaneous transcatheter arterial embolization diagnosis of 36
277–85 equipment tray for use in hospitals 179–82
perineal tears, see hemorrhage from lower genital tract etiology and precipitating factors 23
pituitary gland, postpartum necrosis of anterior lobe, in out-of-hospital delivery 413–18
see Sheehan’s syndrome incidence 3
466
488
05 September 2006 12:14:10
Index
incidence in mortality 2 intrapartum 27–30

litigation concerning 376–87 road traffic accidents and blunt abdominal trauma 76
management 112 ruptured uterus 74–5
at community level 158, 160 after blunt abdominal trauma 78
measurement of 19 Rüsch hydrostatic urological balloon 199
ongoing initiatives for prevention 5 use in balloon tamponade as a test 270
pathophysiology of 62–8 use in uterine tamponade 264
prevention at Cesarean delivery 111
prevention at vaginal delivery 110
prevention of 4 safe motherhood initiative 438
religious beliefs and management 147–53 Saudi Arabia, statistics on postpartum hemorrhage in 458
resuscitation 170 secondary autoimmune thrombocytopenia 212
secondary, management of 316–23 secondary care, and reduction in severe adverse maternal
staging scheme for 172 morbidity 349
pre-eclampsia 214 secondary postpartum hemorrhage,
precipitatiing factors 23 causes of 319
pregnancy, clinical presentation 318
adverse effects of misoprostol in 116 investigations 318
anticoagulant therapy during 219–21 management of 316–23
multiple 26 resuscitation for 318
prolonged 26 treatment for 319
risk of hemophilia in 224–6 selective arterial embolization 200
previous history 27 Sengstaken–Blakemore tube, use in uterine tamponade
prolonged pregnancy 26 199, 264
prostaglandins, sepsis,
gemeprost, use in uterine atony 260 antibiotic therapy and management 405
in postpartum hemorrhage treatment 4 as an indication for hysterectomy 313
prostaglandin E2, use in uterine atony 260 case study 407
prostaglandin F2α, use in uterine atony 260 clinical risk factors for 404
role in uterine activity 64 etiological agents 404
see also misoprostol flow chart for treatment regimens 406
use in uterine atony 260 investigations for 404
prevention of 405
severe adverse maternal morbidity 339–50
recombinant factor VIIa, administrative causes 349
cf plasma-derived factor VIIa 236 causes of 341
experience with, cf a ‘near-miss’ 339
in Poland 239 decreasing incidence 348
world-wide 240–50 health systems and 350
mechanism of hemostatic action 235 hemorrhage as a cause of 346
monitoring clinical effect of 238 in rural areas 349
peculiarities of obstetric hemorrhage 233 incidence 341, 342–5
pharmacokinetic studies 234 outcomes after 348
production using recombinant DNA techniques 234 risk factors for 346
safety of 239 Sheehan’s syndrome 353–9
three-dimensional molecular structure 235 clinical features 354
use in Jehovah’s Witnesses 148, 149 endocrinological work-up in 355
use in obstetric hemorrhage 233–50 litigation concerning 384
use in uterine atony 260 osteoporosis in 356
red cells 172 thyroid reserve 356
religion, treatment 357
religious beliefs and postpartum hemorrhage shock,
management 147–53 definition 45
use of embolization as an alternative to transfusion management of 48
278 non-pneumatic anti-shock garment 136–45
resuscitation, obstetric warning system 47
adult basic life support algorithm 176 stages of 46
cardiopulmonary 176 staging scheme for 172
for secondary postpartum hemorrhage 318 Singapore, statistics on postpartum hemorrhage in 459
general considerations 170 single-unit transfusion
retained products of conception 5, 24, 335 historical practice of 408
risk factors, re-emergence as a consideration 409
antenatal 25 use in developing world 410
for hemorrhage from lower genital tract 196 Sri Lanka, statistics on postpartum hemorrhage in 458
for placenta previa 81 subinvolution 316
risk factors for postpartum hemorrhage, as a pathological finding 335
antenatal 25–7 supralevator hematomas, management of 200
467
489
05 September 2006 12:14:10
surgical management, ureter, damage to on ligation of internal iliac artery 306

B-Lynch suture compression technique 288–92 uterine, lower segment, pathological findings in 333
Cho multiple square sutures 295, 296 uterine activity 62–3
conservative 287–97 uterine artery,
Hayman uterine compression suture 295, 296 branch patterns of 277
hysterectomy 312–15 ligation of 180
ligature placement in stepwise devascularization 295, uterine atony 23, 66
296 as a pathological finding 336
post-surgical follow-up 292 as an indication for hysterectomy 313
symptoms related to blood loss 14 bimanual massage for 257
synechia 203 carboprost 258
syntocinon use in secondary postpartum hemorrhage 319 comparison of therapies 261
syntometrine, ergometrine 258
cf oxytocin 101 management at Cesarean section 369
use in active management of third-stage labor 100 misoprostol 259
systemic lupus erythematosus 212 oxytocin 257
risk factors for 257
standard medical therapy for 256–61
Taj Mahal 2 subinvolution, causes of 316
tamponade, see uterine tamponade uterine blood flow, Doppler evaluation of 58–60
TBA, see traditional birth attendants uterine compression sutures 181
tertiary care and reduction in severe adverse maternal uterine evacuation 320
morbidity 349 uterine inversion, acute 70–4
Thailand, statistics on postpartum hemorrhage in 457 uterine massage, ICM/FIGO joint statement and action
thrombin, coagulation failures in postpartum hemorrhage plan on 7
25 uterine packing 268
thrombocytopenia 210 use in uterine tamponade 266
drug-induced 213 uterine rupture 74–5
gestational 210 as an indication for hysterectomy 313
heparin-induced 213 uterine tamponade,
HIV-associated 213 Bakri balloon for 264
idiopathic purpura 210 balloon tamponade as a diagnostic test 268–75
secondary autoimmune 212 care after successful 266
with microangiopathy 214 Foley catheter for 265
thrombotic thrombocytopenic purpura 215 hydrostatic condom catheter for 266
tissue factor, role in hemostasis 236 principles of 263
tocolytics, use after blunt abdominal trauma 78 Rüsch hydrostatic urological balloon 264
traditional birth attendants, Sengstaken–Blakemore tube for 199, 264
misoprostol administration by 158 test 272
role in reducing postpartum hemorrhage and its the tamponade test 263
impact 374 use in secondary postpartum hemorrhage 321
training of 434 uterine packing for 266
training, uterine trauma 24
for midwives 401 as an indication for hysterectomy 313
labor ward drills 127–35 uteropyosis 203
tranexamic acid, use in uterine atony 260 uterotonic agents,
transfusions, ICM/FIGO joint statement and action plan on 6
alternatives to 175 use in active management of third-stage labor 98–102
changes in practice 408 uterus,
complications of 174 abnormalities and secondary postpartum hemorrhage
considerations for 171 317
intraoperative autologous blood transfusions 421–6 gross examination 326
intraoperative cell salvage for 422 pathology of 326–38
products for 172
single-unit transfusions 408–11
trauma 24 vaginal tears, see hemorrhage from lower genital tract
acute uterine inversion 70 visual assessment 37
as an indication for hysterectomy 313 accuracy of 55
blunt abdominal 76–8 cf other assessment methods 40–2
of lower genital tract 316 vitamin K metabolism, combined deficiency of factors II,
uterine rupture 74–5 VII, IX and X 227
von Willebrand disease 221
classification of 222
ultrasound, pregnancy and 223
diagnosis of placenta previa by 81 secondary postpartum hemorrhage and 317
see also Doppler von Willebrand factor 222
umbilical cord, early clamping of 103 vulval tears, see hemorrhage from lower genital tract
468
490
05 September 2006 12:14:10

A Textbook of Postpartum Hemorrhage

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Textbook of Postpartum Hemorrhage

Uploaded by

Copyright:

Available Formats

Color profile: Generic CMYK printer profile

Composite Default screen

This copy is presented free of charge by courtesy of

Louis G. Keith MD, PhD

André B. Lalonde MD, FRCSC, FRCOG

Mahantesh Karoshi MBBS, MD

With a Special Message from HRH The Princess Royal

This book, and other elements in the

Abigail Bloomer, 1970 –2001

Typeset by AMA DataSet Limited, Preston, Lancashire, UK

Special Message from HRH The Princess Royal iii

Section I: Demographic and logistical considerations

Section II: Causation

Section III: General preventive measures

Section IV: Special preventive measures: misoprostol in action

Section V: Hospital preparation

Section VI: Therapy for non-atonic conditions

Section VII: Therapy for atony

Section VIII: Consequences of postpartum hemorrhage

Section IX: Special experiences and unusual circumstances

43 The midwife confronts postpartum hemorrhage 396

Section X: National experiences

André B. Lalonde, MD, FRCSC, FRCOG, Mahantesh Karoshi, MBBS, MD

A. Acosta, MD D. H. Bevan, FRCP, FRCPath

F. Breathnach, MRCOG, MRCPI, K. Choji

K. J. Dalton, LLM, PhD, FRCOG, FCLM, C. Dell’Aquila

G. S. Eglinton, MD R. V. Ganchev, MB, ChB

B. S. Kodkany, MD, DGO, FICS, FICOG C. A. Ludlam, PhD, FRCP, FRCPath

S. A. Mujeeb, MBBS, MPhil M. Paidas, MD

N. Prata, MD, MSC S. C. Robson, MD, MRCOG

E. Sheiner, MD P. D. Tank, MD, DNBE, FCPS, DGO, DFP,

A. Vais, MB, BS, BSc C. S. Waldmann, MA, MB, BChir, FRCA,

The consequences of postpartum hemorrhage for maternal mortality and morbidity

Arnaldo Acosta, MD André B. Lalonde, MD

Christopher B-Lynch Louis G. Keith

Maternal mortality in the developing world

be implemented in resource-poor settings. This text book provides such options.

Dr Luis Gomes Sambo

Postpartum hemorrhage today

DEFINITION AND INCIDENCE POSTPARTUM HEMORRHAGE:

expectancy at birth strongly correlate with EXISTING EVIDENCE FOR

Postpartum hemorrhage today

in Central America, into the health-care team, ● Administration of uterotonic agents,

Postpartum hemorrhage today

Postpartum hemorrhage today

INTRODUCTION clinical outcome. Therefore, a prognostic classi-

Classification based on quantification of Table 1 Benedetti’s classification of hemorrhage15

Blood loss at delivery is estimated using various 1 900 15

Definitions and classifications

Table 2 Classification of postpartum hemorrhage (PPH) according to causative factors

Causes of primary PPH

Adapted from Wac et al. Female Patient 2005;30:19

Table 3 Symptoms related to blood loss with postpartum hemorrhage

10–15 500–1000 normal palpitations, dizziness, tachycardia

Definitions and classifications

Table 4 Proposed classification. Adapted from Benedetti15

0 (normal loss) < 500 < 10 none

Need observation ± replacement therapy

References 8. Alexander J, Thomas P, Sanghera J. Treatments

INTRODUCTION in significant morbidity, one might argue that

DIFFICULTIES OF COMPARING blood pigment converted to acid or alkaline

Conduct of third stage of labor higher rates of age-related medical diseases,

Table 1 Denominators used in calculating maternal mortality and morbidity