Anestheology Latest PDF

balt6/z7i-anesth/z7i-anesth/z7i00812/contents panickes S12 7/2/12 22:40 Art: toc Input-citi
January 2015
Volume 122, Number 1
ISSN 0003-3022
ON The
THE Cover:
COVER:
David
Mackey,
has organized
fascinating
serieson
ofdetermining
commentaries
national
leaders
The focus
of theM.D.,
anesthesiologist
hasalong
been largely
the by
most
appropriate
in
the construction
and use
data registries
regulatory
agencies
and physician
thresholds
for transfusion
in of
theclinical
perioperative
setting. by
This
months issue
of Anesthesiology
practices,
firstoftwo
of which
appear in this months
includes a the
series
articles
and accompanying
Editorial issue:
Views that detail risks associated with
perioperative
transfusion
that
were previously
ignored,
including
the risk
of Systems-based
transfusion-related
Mackey: Can
We Finally
Conquer
the Problem
of Medical
Quality?
The
acute
lung
injury
(TRALI),
transfusion-associated
circulatory
overload
(TACO),
and the risks and
Opportunities of Data Registries and Medical Teamwork, p. 225
benefits of transfusion in surgical oncology patients.
Jain et al.: A Public-Private Strategy to Advance the Use of Clinical Registries, p. 227
Clifford et al.: Characterizing the Epidemiology of Postoperative Transfusion-related Acute
Lung Injury, p. 12
The Adverse Effects of Transfusion
et al.: Characterizing the Epidemiology of Perioperative Transfusion-associated

Circulatory Overload, p. 21
Clifford
Pinheiro de Almeida et al.: Transfusion Requirements in Surgical Oncology Patients:

THIS MONTH IN ANESTHESIOLOGY
A Prospective, Randomized Controlled Trial, p. 29
9A
Simmons
EDITORIAL VIEWS
and Pittet: Revealing the Real Risks of Perioperative Transfusion: Rise of the
Machines!, p. 1
Cata:
Perioperative Anemia and Blood Transfusions in Patients with Cancer: When the Problem,
thethe
Solution,
and Their
Combination
Are EachThe
Associated
with Poor Outcomes, p. 3
. Can We Finally Conquer
Problem
of Medical
Quality?
Systems-based
Opportunities of Data Registries and Medical Teamwork
225
David C. Mackey
. A Public-Private Strategy to Advance the Use of Clinical Registries
227
1A
THIS
MONTH
IN ANESTHESIOLOGY
Sachin
H. Jain,
Patrick H. Conway,
Science, Medicine, and the Anesthesiologist
230
21A
Infographics in Anesthesiology
. Making Sedation Safer: Is Simulation the Answer?
232
23A
and Donald M. Berwick
. What Are We Looking For? The Question of Resident Selection

Lee A. Fleisher, Alex S. Evers, Jeanine Wiener-Kronish, and John A. Ulatowski
Carl E. Rosow
Pay Toll on the Bridge from Innate Immunity to Ventilator-induced

Diaphragm Atrophy?
Revealing
the Real Risks of Perioperative Transfusion: Rise of the Machines!
Maria T. Kuipers, Catharina W. Wieland, and Marcus J. Schultz

Do We Need
EDITORIAL
VIEWSto
J.W.Simmons and J.-F.Pittet
Ultrasound versus Fluoroscopy in Image-guided Pain Treatment: Use Caution

Perioperative Anemia and Blood Transfusions in Patients with Cancer: When the Problem, the
Asokumar Buvanendran and James P. Rathmell
Solution, and Their Combination Are Each Associated with Poor Outcomes
234
1
236
3
J. P. Cata
SPECIAL ARTICLES
Anaphylaxis to Neuromuscular-blocking Drugs: All Neuromuscular-blocking Drugs

Ebenezer
Hopkins
Are
Not the
Same Frost (1824 1866): William T.G. Mortons First Identified Patient
and
Why
He
WasG.W.Volcheck
Invited to the Ether Demonstration of October 16, 1846
P. M.Mertes and
Ryan LeVasseur and Sukumar P. Desai
Big Brain, Small World?
Ebenezer Hopkins Frost was the first identified patient of William T.G. Morton to receive anesthesia. We suggest why
E.Olofsen
and A.Dahan
he was invited to attend the ether demonstration of October 16, 1846.
Refers to This Month in Anesthesiology

See Supplemental Digital Content
CME
Supplemental Digital Content
Refers
This Month
SeeArticle
Refers
to to
Editorial
Viewsin Anesthesiology

CME
SeeRefers
to
Editorial
Views
page2anesthesiology.org for content related to this article Article
5
238
8
contents

PERIOPERATIVE MEDICINE
CLINICAL SCIENCE
Characterizing the Epidemiology of Postoperative Transfusion-related Acute Lung Injury
12
L.Clifford, Q.Jia, A.Subramanian, H.Yadav, G.A.Wilson, S.P.Murphy, J.Pathak,

D.R.Schroeder, and D.J.Kor
A retrospective cohort analysis from one institution documented that perioperative transfusion-related acute lung injury occurs
approximately 1.4 to 3% in surgical patients, with higher rates in patients who received larger volumes of blood component therapies.
Characterizing the Epidemiology of Perioperative Transfusion-associated Circulatory

Overload
21
L.Clifford, Q.Jia, H.Yadav, A.Subramanian, G.A.Wilson, S.P.Murphy, J.Pathak,

D.R.Schroeder, M.H.Ereth, and D.J.Kor
This retrospective cohort study evaluated 2,162 and 1,908 patients who received intraoperative transfusions during
noncardiac surgery in 2004 and 2011, respectively. A total of 119 patients (5.5%) in 2004 and 57 patients (3%) in
2011 met criteria for transfusion-associated circulatory overload. The incidence of transfusion-associated circulatory
overload increased with the volume of blood product transfused, advanced age, and total intraoperative fluid balance.
SUPPLEMENTAL DIGITAL CONTENT IS AVAILABLE IN THE TEXT

Transfusion Requirements in Surgical Oncology Patients: A Prospective, Randomized
Controlled Trial
29
J. Pinheiro deAlmeida, J.-L.Vincent, F. R. B. GomesGalas, E.PintoMarinhodeAlmeida,

J.T.Fukushima, E.A.Osawa, F.Bergamin, C.LeePark, R.ElyNakamura, S.M.R.Fonseca,
G.Cutait, J.InacioAlves, M.Bazan, S.Vieira, A.C.VieiraSandrini, H.Palomba, U.Ribeiro,
Jr., A.Crippa, M.Dalloglio, M.delPilarEstevezDiz, R.KalilFilho, J.OtavioCostaAuler,
Jr., A.Rhodes, and L.AbrahaoHajjar
In 198 patients randomly assigned to red cell transfusions at a hemoglobin concentration of 7 or 9g/dl. Major
complications were nearly twice as common in patients managed with the restrictive approach as in those managed with
the liberal approach (36% vs. 20%). This study supports a more liberal transfusion strategy in major cancer surgery.
Anaphylaxis Is More Common with Rocuronium and Succinylcholine than with

Atracurium
39
J.I.Reddy, P.J.Cooke, J.M.vanSchalkwyk, J.A.Hannam, P.Fitzharris, and S.J.Mitchell

Search of a database containing more than 400,000 anesthetic records identified 92,858 new patient exposures to
neuromuscular-blocking drugs between 2006 and 2012. Twenty-one of 89 patients referred to the Anesthetic Allergy
Clinic had anaphylaxis attributed to muscle relaxants. Use of credible numerator and denominator data found similar
rates of anaphylaxis after succinylcholine and rocuronium administration, rates that were nearly an order of magnitude
higher than those for atracurium and other neuromuscular-blocking drugs. SUPPLEMENTAL DIGITAL CONTENT
IS AVAILABLE IN THE TEXT
Postoperative Bladder Catheterization Based on Individual Bladder Capacity:

A Randomized Trial
46
T.A.Brouwer, P.F.W.M.Rosier, K.G.M.Moons, N.P.A.Zuithoff, E.N.vanRoon, and C.J.Kalkman

In a prospective trial involving 1,840 patients, maximum bladder capacity was determined before surgery. Using predetermined
maximum bladder capacity, the authors demonstrated that a reduction in the need for postoperative bladder catheterization could be
achieved.
Accuracy of Malignant Hyperthermia Diagnoses in Hospital Discharge Records
55
T.Pinyavat, H.Rosenberg, B.H.Lang, C.A.Wong, S.Riazi, J.E.Brady, L.S.Sun, and G.Li

In review by an expert panel of International Classification of Diseases coding for malignant hyperthermia over a
3-yr period, approximately 70% of coded cases were considered to be malignant hyperthermia susceptible. The most
common reason for inaccurate coding was high fever unrelated to anesthesia.
Relationship between Chronic Intermittent Hypoxia and Intraoperative Mean Arterial

Pressure in Obstructive Sleep Apnea Patients Having Laparoscopic Bariatric Surgery
A.Turan, J.You, C.Egan, A.Fu, I.Gazmuri, A.Khanna, Y.Eshraghi, R.Ghosh, S.Bose,
S.Qavi, L.Arora, D.I.Sessler, and A.G.Doufas
Recurrent nocturnal hypoxemia in obstructive sleep apnea is not a risk marker for intraoperative hypotension in patients
undergoing laparoscopic bariatric surgery.
64
contents
BASIC SCIENCE
Propofol Attenuated Acute Kidney Injury after Orthotopic Liver Transplantation via
Inhibiting Gap Junction Composed of Connexin 32
72
C.Luo, D.Yuan, X.Li, W.Yao, G.Luo, X.Chi, H.Li, M.G.Irwin, Z.Xia, and Z.Hei
Anesthetized rats underwent autologous orthotopic liver transplantation in the absence or presence of treatments with a
selective Cx32 inhibitor, 2-aminoethoxydiphenyl borate, or propofol. Propofol inhibited Cx32 function and attenuated
postautologous orthotopic liver transplantation acute kidney injury. SUPPLEMENTAL DIGITAL CONTENT IS
AVAILABLE IN THE TEXT
Long-term Effects of Single or Multiple Neonatal Sevoflurane Exposures on Rat

Hippocampal Ultrastructure
87
L.G.Amrock, M.L.Starner, K.L.Murphy, and M.G.Baxter

Repeated exposure to sevoflurane led to a greater loss of synapses in comparison to a single exposure. Anesthetic
exposure led to a reduction in the number of synaptic terminals with mitochondria. Interestingly, this reduction was
correlated to total anesthetic exposure rather than frequency of exposure. These data suggest that a brief anesthetic
exposure might sensitize the brain to subsequent anesthetic induced injury.

CRITICAL CARE MEDICINE
CLINICAL SCIENCE
Assessment of Neutrophil Gelatinase-associated Lipocalin in the Brain-dead Organ Donor
to Predict Immediate Graft Function in Kidney Recipients: A Prospective, Multicenter Study 96
L. Muller, A. Nicolas-Robin, S. Bastide, O. Martinez, G. Louart, J.-C. Colavolpe, F. Vachiery,
S. Alonso, J.-Y. Lefrant, and B. Riou; for AzuRea Group
Despite the ability to predict acute renal failure earlier than serum creatinine rises in critically ill patients, neutrophil
gelatinase-associated lipocalin measurements in blood samples obtained from brain-dead donors before kidney graft
harvesting failed to predict either delayed or normal graft function in kidney recipients.
BASIC SCIENCE
Modulation of Stress versus Time Product during Mechanical Ventilation Influences

Inflammation as Well as Alveolar Epithelial and Endothelial Response in Rats
106
P.M.Spieth, P.L.Silva, C.S.N.B.Garcia, D.S.Ornellas, C.S.Samary, L.Moraes,

M.Bentes, M.M.Morales, M.Kasper, A.Gldner, R.Huhle, T. Koch, P. Pelosi,
M. Gama de Abreu, and P. R. M. Rocco
In a mild acute lung inflammation model in rats, using mechanical ventilation with an inspiratory-to-expiratory ratio
of 1:1 minimized lung damage, whereas an inspiratory-to-expiratory ratio of 2:1 led to increased gene expression of
inflammatory mediators and markers of alveolar epithelial cell injury. SUPPLEMENTAL DIGITAL CONTENT IS
AVAILABLE IN THE TEXT
Volatile Organic Compounds during Inflammation and Sepsis in Rats: A Potential Breath
Test Using Ion-mobility Spectrometry
117
T. Fink, A. Wolf, F. Maurer, F. W. Albrecht, N. Heim, B. Wolf, A. C. Hauschild, B. Bdeker,

J. I. Baumbach, T. Volk, D. I. Sessler, and S. Kreuer
Exhaled gas from rats given endotoxin compared with the gas from rats with bacterial sepsis was found to be
significantly different and different from rats who were in hemorrhagic shock. Breath analysis appears to be able to
distinguish inflammation from infection.
Extracellular Histones Play an Inflammatory Role in Acid Aspiration-induced Acute

Respiratory Distress Syndrome
Y. Zhang, Z. Wen, L. Guan, P. Jiang, T. Gu, J. Zhao, X. Lv, and T. Wen

Extracellular histones were significantly elevated in the bronchoalveolar lavage from mice with acid-induced lung injury
versus sham mice and in human patients who died from acute lung injury compared to survivors with acute lung injury.
Extracellular histones may be causal, and targeting histones may be a reasonable therapeutic strategy. SUPPLEMENTAL
DIGITAL CONTENT IS AVAILABLE IN THE TEXT
127
contents

PAIN MEDICINE
CLINICAL SCIENCE
Disruption of Cortical Connectivity during Remifentanil Administration Is Associated

with Cognitive Impairment but Not with Analgesia
140
A. Khodayari-Rostamabad, S. S. Olesen, C. Graversen, L. P. Malver, G. P. Kurita, P. Sjgren,

L. L. Christrup, and A. M. Drewes
Remifentanil altered graph-theoretical measures of the electroencephalography, characterized by an increase in
path length in the alpha and low beta frequency ranges. Changes in path length were correlated to continuous
reaction time, a measure of sedation. However, a correlation between electroencephalography measures and pain
perception was not apparent. Remifentanil alters functional network connectivity in the brain, and the changes in the
electroencephalography have the potential to serve as markers of remifentanil-induced sedation but not analgesia.
Immediate Rescue Designs in Pediatric Analgesic Trials: A Systematic Review and

Meta-analysis
150
J. Kossowsky, C. Donado, and C. B. Berde

The investigators performed a meta-analysis of pediatric trials with four classes of analgesics, using rescue/opioid-sparing
designs. Average pain scores were low and similar in control and experimental analgesic groups, confirming the ethical
basis of opioid-sparing rescue designs. Opioid-sparing designs also showed good assay sensitivity. SUPPLEMENTAL
DIGITAL CONTENT IS AVAILABLE IN THE TEXT
A Randomized Control Trial of Bupivacaine and Fentanyl versus Fentanyl-only for

Epidural Analgesia during the Second Stage of Labor
172
M. G. Craig, E. N. Grant, W. Tao, D. D. McIntire, and K. J. Leveno

In 310 nulliparous women with epidural analgesia randomized at the onset of second stage to receive epidural fentanyl
alone or with bupivacaine, there was no difference in duration of second stage, degree of motor block, or instrumental
delivery. To achieve similar degrees of analgesia, women receiving epidural fentanyl without bupivacaine required a
fivefold increased dose of fentanyl.
BASIC SCIENCE
Brain Serotonin Content Regulates the Manifestation of Tramadol-induced Seizures in
Rats: Disparity between Tramadol-induced Seizure and Serotonin Syndrome
178
Y. Fujimoto, T. Funao, K. Suehiro, R. Takahashi, T. Mori, and K. Nishikawa

Tramadol-induced seizure thresholds were reduced by serotonin depletion and increased by serotonin augmentation.
Serotonin antagonists also reduced seizure threshold. The results suggest that tramadol-induced seizures are not related
to serotonin uptake inhibition and that these seizures are distinct from the serotonin syndrome.

EDUCATION
IMAGES IN ANESTHESIOLOGY
Inversion of the Right Hemidiaphragm due to Massive

Hemothorax after Central Line Placement
190
A. F. Simpao, J. A. Galvez, A. Jay Schwartz, and M. A. Rehman

SUPPLEMENTAL DIGITAL CONTENT IS AVAILABLE IN THE TEXT
Clinical Concepts and Commentary
Pretransfusion Testing and Transfusion of Uncrossmatched Erythrocytes
191
M. L. Boisen, R. A. Collins, M. H. Yazer, and J. H. Waters

Pretransfusion testing is reviewed for the anesthesiologist, with an emphasis on the electronic crossmatch and transfusion
of uncrossmatched erythrocytes when testing is incomplete.
Review Article
Regulation of Cerebral Autoregulation by Carbon Dioxide

L. Meng and A. W. Gelb
Both perfusion pressure and nonperfusion pressure processes regulate cerebral blood flow. The integrated effect of carbon dioxide
and perfusion pressure on cerebral circulation, or the regulation of cerebral autoregulation by carbon dioxide, is discussed.
196
contents
MIND TO MIND
Electrocardiogram
206
A. Shafer

CORRESPONDENCE
Is the Standard Supplied by the Association for the Advancement of Medical

Instrumentation the Measure of All Things for Noninvasive Continuous
Hemodynamic Devices?
208
J. Fortin, K. Lerche, D. Flotzinger,and T. OBrien
In Reply
M. Cannesson, J. Rinehart, and S.-H. Kim
Inotrope Use in Cardiac Surgery: ACause of Worse Outcomes, or Just a Marker

of Patients Who Are at Risk?
210
B. G. Maxwell, J. O. Wasey, and E. S. Heitmiller
In Reply
D. V. Nielsen, S. P. Johnsen, M. K. Hansen, and C.-J. Jakobsen
Lung Ultrasonography for the Detection of Anesthesia-induced Lung Atelectasis
213
M. Girard, V.Gnreux, and A.Monastesse
In Reply
G. Tusman, C. M. Acosta, and S. H. Bohm
Early Childhood Anesthetic Neurotoxicity and Unmeasured Covariates: Theres the RUB
216
J.C.Drummond
In Reply
C.H.Ing, C.J.DiMaggio, E.Malacova, A.J.Whitehouse, M.K.Hegarty, T.Feng,
J.E.Brady, B.S.vonUngern-Sternberg, A.J.Davidson, M.M.Wall, A.J.J.Wood, G.Li,
and L.S.Sun
Old Guidelines or Methods Cannot Insure Quality or Progress
218
P. M. Kempen
In Reply
J.L.Lockman and A. J.Schwartz
In Reply
W.R.Hand and M.D.McEvoy
ANESTHESIOLOGY REFLECTIONS FROM

THE WOOD LIBRARY-MUSEUM
Katz Oxygen Treatment for Catarrh
George S. Bause
Laughing Gas from Knoxvilles Dr. H. F. Huffaker
11
George S. Bause
Figuiers Forlorn Figure: Horace Wells and the Humbug Affair

George S. Bause
54
contents

REVIEWS OF EDUCATIONAL MATERIAL
222
ANNOUNCEMENTS
224
CAREERS & EVENTS
228
INSTRUCTIONS FOR AUTHORS

The most recently updated version of the Instructions for Authors is available at
www.anesthesiology.org. Please refer to the Instructions for the preparation of any material for
submission to ANESTHESIOLOGY.
Manuscripts submitted for consideration for publication must be submitted in electronic format. The preferred method is via the Journals
Web site (http://www.anesthesiology.org). Detailed directions for submissions and the most recent version of the
Instructions for Authors can be found on the Web site (http://www.anesthesiology.org). Books and educational materials
should be sent to Alan Jay Schwartz M.D., M.S.Ed., Director of Education, Department of Anesthesiology and Critical
Care Medicine, The Childrens Hospital of Philadelphia, 34th Street and Civic Center Blvd., Room 9327, Philadelphia,
Pennsylvania 19104-4399. Requests for permission to duplicate materials published in ANESTHESIOLOGY should be submitted
in electronic format, to the Permissions Department (journalpermissions@lww.com). Advertising and related correspondence should
be addressed to Advertising Manager, ANESTHESIOLOGY, Lippincott Williams & Wilkins, Two Commerce Square, 2001 Market Street,
Philadelphia, Pennsylvania 19103 (Web site: http://www.wkadcenter.com/). Publication of an advertisement in ANESTHESIOLOGY does
not constitute endorsement by the Society or Lippincott Williams & Wilkins, Inc. of the product or service described therein or of any
representations made by the advertiser with respect to the product or service.
ANESTHESIOLOGY (ISSN 00033022) is published monthly by Lippincott Williams & Wilkins, 16522
Hunters Green Parkway, Hagerstown, MD 21740-2116. Business office: Two Commerce Square,
2001 Market Street, Philadelphia, PA 19103. Periodicals postage paid at Hagerstown, MD, and at
additional mailing offices. Copyright 2014, the American Society of Anesthesiologists, Inc.
Annual Subscription Rates: United States$799 Individual, $1566 Institution, $319
In-training. Rest of World$843 Individual, $1739 Institution, $319 In-training. Single copy rate
$145. Subscriptions outside of North America must add $52 for airfreight delivery. Add state
sales tax, where applicable. The GST tax of 7% must be added to all orders shipped to Canada
(Lippincott Williams & Wilkins GST Identification #895524239, Publications Mail Agreement
#1119672). Indicate in-training status and name of institution. Institution rates apply to
libraries, hospitals, corporations, and partnerships of three or more individuals. Subscription
prices outside the United States must be prepaid. Prices subject to change without notice.
Subscriptions will begin with currently available issue unless otherwise requested. Visit us
online at www.lww.com.
Individual and in-training subscription rates include print and access to the online version.
Online-only subscriptions for individuals ($257) and persons in training ($257) are available
to nonmembers and may be ordered by downloading a copy of the Online Subscription
FAXback Form from the Web site, completing the information requested, and faxing the
completed form to 301-223-2400/44 (0) 20 7981 0535. Institutional rates are for print only;
online subscriptions are available via Ovid. Institutions can choose to purchase a print and
online subscription together for a discounted rate. Institutions that wish to purchase a print
subscription, please contact Lippincott Williams & Wilkins, 16522 Hunters Green Parkway,
Hagerstown, MD 21740-2116; phone: 1-800-638-3030 (outside the United States 301-2232300/44 (0) 20 7981 0525); fax: 301-223-2400/44 (0) 20 7981 0535. Institutions that
wish to purchase an online subscription or online with print, please contact the Ovid Regional
Sales Office near you or visit www.ovid.com/site/index.jsp and select Contact and Locations.
Address for non-member subscription information, orders, or change of address:

Lippincott Williams & Wilkins, 16522 Hunters Green Parkway, Hagerstown, MD 217402116; phone: 1-800-638-3030 (outside the United States 301-223-2300/44 (0) 20 7981
0525); fax: 301-223-2400/44 (0) 20 7981 0535; email: customerservice@lww.com.
In Japan, contact LWW Japan Ltd., 3-23-14 Hongo, Bunkyo-ku, Tokyo 113, Japan; phone:
81-3-5689-5400; fax: 81-3-5689-5402; email: bclaim@lwwis.co.jp. In Bangladesh, India, Nepal,
Pakistan, and Sri Lanka, contact Globe Publications Pvt. Ltd., B-13 3rd Floor, A Block, Shopping
Complex, Naraina, Vihar, Ring Road, New Delhi 110028, India; phone: 91-11-25770411; fax:
91-11-25778876; email: info@globepub.com.
Address for member subscription information, orders, or change of address: Members of the American Society of Anesthesiologists receive the print and online journal with
their membership. To become a member or provide a change of address, please contact the
American Society of Anesthesiologists, 1061 American Lane, Schaumburg, Illinois
60173-4973; phone: 847-825-5586; fax: 847-825-1692; email: membership@ASAhq.
org. For all other membership inquiries, contact Lippincott Williams & Wilkins Customer
Service D
epartment, P.O. Box 1580, Hagerstown, MD 21741-1580; phone: 1-800-638-3030
(outside the United States 301-223-2300/44 (0) 20 7981 0525); fax: 301-223-2400/44
(0) 20 7981 0535; email: memberservice@lww.com.
Postmaster: Send address changes to ANESTHESIOLOGY, P.O. BOX 1550, Hagerstown, MD 21740.
Advertising: Please contact Michelle Smith, Senior Account Manager, Advertising, Lippincott
Williams & Wilkins, 333 Seventh Avenue, 19th Floor, New York, NY 10001; tel: (646) 674-6537,
fax: (646) 607-5479, e-mail: Michelle.Smith@wolterskluwer.com. For classified advertising:
Keida Spurlock, Recruitment Advertising Representative, Lippincott Williams & Wilkins, Two
Commerce Square, 2001 Market Street, Philadelphia, PA 19103; tel: (215) 521-8501, fax:
(215) 689-2453. e-mail: Keida.Spurlock@wolterskluwer.com.
Mission: Promoting scientific discovery and knowledge in perioperative, critical care, and pain medicine to
advance patient care.
EDITOR-IN-CHIEF
COVER ART
James C. Eisenach, M.D.

Editor-in-Chief, Anesthesiology
Department of Anesthesiology
Wake Forest University School of Medicine
Medical Center Boulevard
Winston-Salem, NC 27157
Tel: 1-800-260-5631
E-mail: editorial-office@anesthesiology.org
James P. Rathmell, M.D.

Boston, Massachusetts
Annemarie B. Johnson, C.M.I.
Medical Illustrator
Winston-Salem, North Carolina
PAST EDITORS-IN-CHIEF, Anesthesiology
Henry S. Ruth, M.D., 19401955

Ralph M. Tovell, M.D., 19561958
James E. Eckenhoff, M.D., 19591962
Leroy D. Vandam, M.D., 19631970
Arthur S. Keats, M.D., 19711973
Nicholas M. Greene, M.D., 19741976
C. Philip Larson, Jr., M.D., 19771979
John D. Michenfelder, M.D., 19801985
Lawrence J. Saidman, M.D., 19861996
Michael M. Todd, M.D., 19972006
CME EDITORS
Leslie C. Jameson, M.D.

Dan J. Kopacz, M.D.
EDITORIAL OFFICE
Vicki Tedeschi, Managing Editor

E-mail: managing-editor@anesthesiology.org
Allison Akeley
Vicky J. Farrington-Howrey
Angel R. Marsh
Karen Parks
Anesthesiology Journal
1061 American Lane
Schaumburg, IL 60173-4973
Tel: (847) 268-9296
E-mail: editorial-office@anesthesiology.org
LWW PUBLICATION STAFF
Druanne Martin, Publisher

Joseph Albrecht, Journal Production Editor
Debbie Moody, Journal Production Associate
Michelle Smith, Senior Account Manager, Advertising
Keida Spurlock, Classified Advertising Representative
Silvia Serra, Director, Rights, Licensing & Permissions
ASA OFFICERS
John P. Abenstein, M.D., President

Daniel J. Cole, M.D., President-Elect
Jane C. K. Fitch, M.D., Immediate Past President
Jeffrey Plagenhoef, M.D., First Vice-President
Beverly K. Philip, M.D., Vice-President for Scientific Affairs
All articles accepted for publication are done so with the understanding that
they are contributed exclusively to this Journal and become the property
of the American Society of Anesthesiologists, Inc. Statements or opinions
expressed in the Journal reflect the views of the author(s) and do not represent
official policy of the American Society of Anesthesiologists unless so stated.
Mission: Promoting scientific discovery and knowledge in perioperative, critical care, and pain medicine to
advance patient care.
EDITOR-IN-CHIEF
ASSOCIATE EDITORS
James C. Eisenach, M.D.

Julien Amour, M.D., Ph.D., Paris, France

Takashi Asai, M.D., Ph.D., Osaka, Japan
Brian Thomas Bateman, M.D., Boston, Massachusetts
George S. Bause, M.D., M.P.H., Cleveland, Ohio
Beatrice Beck-Schimmer, M.D., Zurich, Switzerland
Chad Michael Brummett, M.D., Ann Arbor, Michigan
John Butterworth, M.D., Richmond, Virginia
Jaume Canet, M.D., Ph.D., Barcelona, Spain
Carol Wiley Cassella, M.D., Seattle, Washington
Steven P. Cohen, M.D., Baltimore, Maryland
Deborah J. Culley, M.D., Boston, Massachusetts
Albert Dahan, M.D., Ph.D., Leiden, The Netherlands
Andrew J. Davidson, M.D., M.B.B.S., F.A.N.Z.C.A.,
Melbourne, Australia
Holger K. Eltzschig, M.D., Ph.D., Aurora, Colorado
Pamela Flood, M.D., San Francisco, California
Amanda A. Fox, M.D., M.P.H., Dallas, Texas
Admir Hadzic, M.D., Ph.D., New York, New York
Quinn H. Hogan, M.D., Milwaukee, Wisconsin
Ru-Rong Ji, Ph.D., Durham, North Carolina
Yandong Jiang, M.D., Ph.D., Boston, Massachusetts
Sachin Kheterpal, M.D., M.B.A., Ann Arbor, Michigan
Kate Leslie, M.B.B.S., M.D., M.Epi., F.A.N.Z.C.A.,
Parkville, Australia
Ronald S. Litman, D.O., F.A.A.P., Philadelphia, Pennsylvania
Martin J. London, M.D., San Francisco, California
Edward Mascha, Ph.D., Cleveland, Ohio
George A. Mashour, M.D., Ph.D., Ann Arbor, Michigan
Paul S. Myles, M.B., B.S., M.P.H., M.D., F.F.A.R.C.S.I.,
F.A.N.Z.C.A., Melbourne, Australia
Peter Nagele, M.D., M.Sc., St. Louis, Missouri
Mark D. Neuman, M.D., M.Sc., Philadelphia, Pennsylvania
Warren S. Sandberg, M.D., Ph.D., Nashville, Tennessee
Alan Jay Schwartz, M.D., M.S.Ed., Philadelphia, Pennsylvania
Nikolaos J. Skubas, M.D., New York, New York
Jamie W. Sleigh, M.D., Hamilton, New Zealand
Balachundhar Subramaniam, M.B.B.S., M.D., M.P.H.,
Boston, Massachusetts
Marcos F. Vidal Melo, M.D., Ph.D., Boston, Massachusetts
Hannah Wunsch, M.D., M.Sc., New York, New York
EXECUTIVE EDITORS
Michael J. Avram, Ph.D., Chicago, Illinois

Charles D. Collard, M.D., Houston, Texas
Jerrold H. Levy, M.D., F.A.H.A., F.C.C.M.,
Durham, North Carolina
James P. Rathmell, M.D., Boston, Massachusetts
David S. Warner, M.D., Durham, North Carolina
STATISTICAL EDITOR
Timothy T. Houle, Ph.D.

EDITORS
J. David Clark, M.D., Ph.D., Palo Alto, California

Hugh C. Hemmings, Jr., M.D., Ph.D., New York, New York
Shiroh Isono, M.D., Chiba, Japan
Brian P. Kavanagh, M.B., B.Sc., M.R.C.P., F.C.A.R.C.S.I.,
Toronto, Canada
Jean Mantz, M.D., Ph.D., Clichy, France
Piyush M. Patel, M.D., F.R.C.P.C., San Diego, California
David M. Roth, M.D., Ph.D., San Diego, California
Daniel I. Sessler, M.D., Cleveland, Ohio
Access to articles published in ANESTHESIOLOGY is available to nonsubscribers

through pay-per-view at the journal Web site, www.anesthesiology.org. Authors
may order quantity print or electronic reprints of their own articles using the order
form supplied with their proofs and posted to the journal Web site. For all other bulk
purchases, please contact reprintsolutions@wolterskluwer.com in North America, Carlos Moreyra (Carlos.Moreyra@wkglobal.com) in Latin America, and Jean Jones
(PS-Reprints@wolterskluwer.com) in Europe and elsewhere.
Anesthesiology is abstracted or indexed in Index Medicus/MEDLINE, Science Citation
Index/SciSearch, Current Contents/Clinical Medicine, Current Contents/Life Sciences,
Reference Update, EMBASE/Excerpta Medica, Biological Abstracts (BIOSIS), Chemical
Abstracts, Hospital Literature Index, and Comprehensive Index to Nursing and Allied
Health Literature (CINAHL).
The affiliations, areas of expertise, and conflict-of-interest disclosure statements for each
Editor and Associate Editor can be found on the Journals Web site (www.anesthesiology.org).
balt6/z7i-anesth/z7i-anesth/z7i00812/thismonth panickes S245 6/29/12 22:41 Art:
Input-ebh
THIS MONTH IN
THIS MONTH IN
Factors Affecting Admission to Anesthesiology

243
Residency in the United States: Choosing the Future of
Innate Immune Dysfunction in 411
29Transfusion Requirements
Our Specialty in Surgical Oncology Patients: A Prospective,
Trauma Patients: From
Randomized
Controlled
Trial
Pathophysiology to Treatment
The proportion of anesthesiology residents from U.S. medical schools has more than dou(Clinical Concepts and
since 1995.
Thissurgery
retrospective
cohort study
evaluated
the 2010
and 2011
The decision to transfusebled
patients
having
for cancer
should
take into
account
risksresidency
of anemia
Commentary)
applicants
to determine
and adverse effects of blood
transfusion.
A the
composite primary outcome of death from all causes or
Recent insights into posttraumatic

immune
dysfunction
factors
associated
with aof
sucsevere
clinical
complications
within
30 days
randomization was compared in 198 high-risk abdominal
have defined new targets for immunointervention that
cessfulrandomly
admissionassigned
to resi- to a restrictive or a liberal transfusion strategy, who
oncological
surgery
patients
hold promise for improving outcomes in such critically ill
dency
training
programs.
received an erythrocyte unit
when
their hemoglobin
concentration decreased to less than 7 g/dl or 9 g/
patients.
Theintensive
sample represented
58%
dl, respectively, during their
care unit
stay. The composite primary outcome occurred in 19.6%
of the total
applicant
of the liberal strategy patients
andnational
in 35.6%
of the restrictive strategy patients. See the accompanying
66% of M.J.
the applicants
Editorial View on page 3. pool;
(Summary:
Avram. Image: A. Johnson/Vivo Visuals.)
successfully matched to anesthesiology. The odds for a successful
match were higher for of Perioperative Transfusion-associated
Characterizing
the Epidemiology
High Intraoperative21
Inspired
271
applicants from U.S. medical
Oxygen Does Not Increase
Circulatory Overload
schools, those with United
Postoperative Supplemental
States overload
Medical (TACO)
Licensingis the second leading cause of transfusion-related death, but
Transfusion-associated
circulatory
Oxygen Requirements
Examination
scores
greaterPerioperative TACO incidence was determined by analyzing
perioperative
is poorly understood.
High inspired oxygen may beitsreasonable
in lowerepidemiology
risk
thanmedical
210, younger
applidata
abstracted
from
electronic
records
of 4,070 adults who underwent noncardiac surgery under general
surgery to improve wound oxygenation.
cants,
and
females.
Prior transfusions in 2004 or 2011. TACO occurred within 6 h of
anesthesia and received intraoperative blood product
or peerthe last intraoperative bloodgraduate
producteducation
transfusion
at rates of 5.5% in 2004 and 3.0% in 2011. Although increased
reviewed
publications
did notpatient
offer any
advantage.
Thisvolume,
study suggests
the potential
age balrates of TACO were associated
with surgical
specialty,
age,
transfusion
and total
operativeforfluid
and characteristic
gender bias in the
selection
process. See
accompanying
View
on page 2004
230. and
ance, no patient or transfusion
could
fully account
forthe
the
decreasedEditorial
incidence
between
2011. See the accompanying Editorial View on page 1. (Summary: M.J. Avram. Image: Shutterstock.)
Accuracy of Ultrasound-guided 347

What Factors Affect Intrapartum Maternal Temperature?
302
Nerve Blocks of the Cervical
A
Prospective
CohortofStudy:
Maternal Intrapartum
12
Characterizing
the
Epidemiology
Postoperative
Transfusion-related
Zygapophysial Joints
Temperature
Ultrasound imaging was an accurate technique

cervi- Injury
AcuteforLung
cal zygapophysial joint nerve blocks in volunteers. See the
The cause of rises in intrapartum maternal temperature is not known. In this prospective
accompanying Editorial View onTransfusion-related
page 236.
acutestudy
lungofinjury
(TRALI)
is the
cause hourly
of transfusion-related
death,
but its
81 women
scheduled
forleading
labor induction,
oral temperatures were
recorded
perioperative epidemiology
poorly understood.
perioperative
andisanalyzed
based on race, The
bodycombined
mass index,incidence
duration ofoflabor,
and time toTRALI/possiepidural.
ble TRALI was determined
by
analyzing
data
abstracted
from
electronic
medical
records
of 3,379
adults
Overall, temperature rose in a significant linear trend over time. Positive temperature
trends
who underwent noncardiac
surgery
under
general anesthesia
blood
product
were
associated
with significantly
longer time and
fromreceived
membraneintraoperative
rupture to delivery
and higher
transfusions in 2004 or 2011,
and Temperature
after introduction
of TRALI
mitigation
strategies.
bodybefore
mass index.
slopes did
not differ
before compared
with TRALI/possible
after epidural
TRALI occurred within 6 analgesia.
h of the This
last study
intraoperative
product
rates the
of 1.3%
in 2004
suggests thatblood
epidural
analgesiatransfusion
alone does notatincrease
risk of high
Estimation
of
the
Contribution
353
and 1.4% in 2011. Increased rates were associated
with
increased
volumes
of
transfused
blood
product.
See
the
accompanying
temperatures in intrapartum women.
of Norketamine
to1.KetamineEditorial
View on page
(Summary: M.J. Avram. Image: J.P. Rathmell.)
induced Acute Pain Relief and

Postoperative QT Interval Prolongation in Patients
Neurocognitive Impairment in
321
Control Trial
of Bupivacaine
and Fentanyl
Noncardiac
Surgery under
General versus
Healthy Volunteers 172A RandomizedUndergoing
Norketamine has an effect opposite to that
of ketamine
Anesthesia
Fentanyl-only
for Epidural Analgesia during the Second Stage of Labor
on pain relief.
Electrocardiograms
(ECG) childbirth
can identifymay
abnormal
cardiac repolarization
by observation
Although epidural analgesia
for pain relief during
not increase
cesarean delivery
rates,ofita can
prolonged
QT interval.
QTIninterval
is often
caused
by drugs and
can result
in
slightly prolong first and second
stages
of labor.
orderprolongation
to determine
whether
eliminating
local
anesthetic
suddenthe
cardiac
death.
In this
ancillary
studyitstolength,
the Vitamins
in Nitrous Oxide
trial,with
serial
from epidural analgesia during
second
stage
would
reduce
310 nulliparous
women
labor
postoperative
were0.125%
obtainedbupivacaine
from 469 patients
major noncardiac
epidurals were randomly assigned
to 12-lead
receiveECG
either
andundergoing
2 g/ml fentanyl
or 10 g/ml
surgerycatheter.
under general
anesthesia.
Eighty
percent
patients of
experienced
a significant
fentanyl alone via the epidural
The groups
did
not differ
in of
duration
the second
stage ofQT
labor,
approximately
halfor
hadmaternal
increases greater
than 440 outcomes.
ms at the end(Sumof
Severe Emergence degree
Agitation
of motor399
blockade,interval
modeprolongation,
of delivery, and
relief
of labor pain,
and neonatal
One patient developed torsade de pointes. Drugs associated with prolonged QT
after Myringotomy mary:
in a 3-yr-old
M.J. Avram. Image:surgery.
J.P. Rathmell.)
interval included isoflurane, methadone, ketorolac, cefoxitin, zosyn, unasyn, epinephrine,
Child (Case Scenario)
ephedrine, and calcium. Although the exact cause of the association between perioperatively
Emergence agitation, the associated risk factors, and its
administered drugs and QT interval prolongation is not known, further study is warranted to
prevention and treatment are discussed.
determine the clinical relevance.
Anesthesiology, V 122 No 1
January 2015
46Postoperative Bladder Catheterization Based on Individual Bladder Capacity: A Randomized Trial

Untreated postoperative urinary retention can lead to bladder wall overdistention and damage to the
detrusor muscle. Nonetheless, most patients prefer to avoid bladder catheterization, which has its
own risks. Eighteen hundred forty patients undergoing operations with general or spinal anesthesia
were randomly assigned to have 500 ml or their maximum bladder capacity (MBC) as the threshold
for catheterization if the patient was unable to void spontaneously. MBC was the maximum voided
volume at home plus the residual volume measured by ultrasound at preoperative assessment. The
incidence of catheterization was 11.8% in the control group and 8.6% in the MBC group. The average
MBC in all patients was approximately 600 ml. (Summary: M.J. Avram. Image: Hellerhoff [own work] [CC-BY-SA-3.0] [http://
creativecommons.org/licenses/by-sa/3.0, via Wikimedia Commons].)
140Disruption of Cortical Connectivity during Remifentanil Administration Is

Associated with Cognitive Impairment but Not with Analgesia
Electroencephalographic graph theory decomposes the multiple cortical connectivity features extracted
from multichannel electroencephalography into composite measures of the overall brain network performance and functionality: characteristic path length, measuring functional integration; mean clustering
coefficient, measuring functional segregation; and relative small-worldness, reflecting the balance of local
segregation and global integration. These graph-theoretical measures were obtained from functional connectivity network measures in resting state electroencephalographic data from 21 volunteers in a double-blind, placebo-controlled, crossover study of remifentanil. Remifentanil reduced overall efficiencies of
cortical networks in and 1 frequency ranges. Disruptions of the complex cortical networks subserving normal brain function were
associated with loss of stability of sustained attention but not with analgesic effect. See the accompanying Editorial View on page 8.
(Summary: M.J. Avram. Image: Thinkstock.)
196Regulation of Cerebral Autoregulation by Carbon Dioxide (Review Article)

Cerebral autoregulation protects the brain from ischemia and overperfusion in the face of fluctuating perfusion pressure. Cerebral blood flow remains stable between the lower and the upper cerebral perfusion pressure limits and is pressure passive at the cerebral perfusion pressures below the
lower limit and above the upper limit. The integrated effect of carbon dioxide and perfusion pressure
on cerebral circulation is discussed on the basis of published large animal data as well as human
data, with speculation on the aspects for which there are no supportive data. (Summary: M.J. Avram.
Image: J.P. Rathmell.)
191Pretransfusion Testing and Transfusion of Uncrossmatched Erythrocytes

(Clinical Concepts and Commentary)
The level of pretransfusion testing ordered can be based on a variety of considerations, including an
institutions maximum surgical blood order schedule, which recommends the extent of pretransfusion
testing for common surgical procedures that will reduce unnecessary testing and costs. Pretransfusion testing aimed at avoiding potentially fatal hemolytic transfusion reactions is reviewed, including
computer or electronic crossmatching, on the basis of which any ABO and RhD type-specific unit can
be issued to a patient with a negative antibody screen and no historical antibodies. Indications for the
emergency release of uncrossmatched group O erythrocytes and risks and special considerations
associated with their use are reviewed. (Summary: M.J. Avram. Image: J.P. Rathmell.)
January 2015
SCIENCE, MEDICINE, AND THE ANESTHESIOLOGIST
Input-ebh
Key Papers from the Most Recent Literature Relevant to Anesthesiologists

THIS MONTH IN
Jean Mantz, M.D., Ph.D., Editor

243
Our Specialty
Global, regional, and national
levels and causes of maternal mortality during
Pathophysiology to19902013:
Treatment A systematic
analysis
the Global
Burden
of Disease
Study
2013.
The proportion offor
anesthesiology
residents
from U.S.
medical schools
has more
than dou(Clinical Concepts Lancet
and
2014; 384:9801004.
bled since 1995. This retrospective cohort study evaluated the 2010 and 2011 residency
Commentary)
applicants to determine the
Thisimmune
worldwide
survey used
robust
methodology
Recent insights into posttraumatic
dysfunction
factors
associated
with a suc- to analyze trends in all-cause maternal mortality ratio
between that
1990 andcessful
2013.admission
The totalto
annual
have defined new targets for (MMR)
immunointervention
resi- number of maternal deaths decreased from 376,034
(95% in
uncertainty
343,483407,574)
in 1990 to 292,982 (261,017327,792) in 2013. Between
hold promise for improving outcomes
such critically illintervaldency
training programs.
patients.
2003 and 2013, the annual rate of change in MMR was greater than 1%, reaching 3.3% for 201213.
The sample represented 58%

The most likely underlyingofetiologies
are complications
of anesthesia, embolism (air, amniotic fluid, and
the total national
applicant
blood clot). Global rates of change suggest that pool;
only 66%
16 countries
will achieve the fifth Millennium Development Goal target by
of the applicants
2015, which requires careful consideration for regions that are making slow progress, such as west and central Africa. (Summary:
successfully matched to anesJ. Mantz. Image: J.P. Rathmell.)
thesiology. The odds for a successful match were higher for
High Intraoperative Inspired
271
applicants
medical
Effect of fenoldopam and
usefrom
of U.S.
renal
replacement therapy among patients with
schools,
those
with
United A randomized clinical trial. JAMA 2014;
acute kidney injury after cardiac surgery:
States Medical Licensing
doi: 10.1001/jama.2014.13573
[Epub ahead of print].
Oxygen Requirements
Examination scores greater
High inspired oxygen may be reasonable in lower risk
than
210,
younger
appli- complications of cardiac surgery and is associated
Acute kidney injury is one of the most
frequent
cants,Inand
Prior randomized controlled trial, the effect of fenoldopam
with morbidity and mortality.
this females.
multicenter
graduate education
or peer(a selective dopamine receptor
D1 agonist
that induces vasodilation of the renal, mesenteric, and
reviewed publications
anywas
advantage.
This study
suggests
the potential
age
coronary arteries) via continuous
infusiondid
fornot
4 offer
days
compared
with
placebo
in 667forcardiac
and gender
bias in the selection
the accompanying
on pagecare
230. unit.
surgical patients with early
postoperative
acute process.
kidneySeeinjury
admittedEditorial
to the View
intensive
The need for renal replacement therapy was the primary outcome. Fenoldopam infusion did not reduce the need for renal
replacement
therapy
or 30-day mortality
with acute kidney injury after cardiac surgery. However, fenoldoAccuracy of
Ultrasound-guided
347 among patients
Factors Affect
Intrapartum
Maternal
302
pam
wasBlocks
associated
with an increased rateWhat
of hypotension.
(Summary:
J. Mantz.
Image:Temperature?
J.P. Rathmell.)
Nerve
of the Cervical
A Prospective Cohort Study: Maternal Intrapartum

Temperature
Ultrasound imaging was an accurate
technique 1
foryear
cervi- after thrombus aspiration for myocardial infarction.
Outcomes
cal zygapophysial joint nerve blocks in volunteers. See the
N Engl J Med 2014; 371:111120.
accompanying Editorial View on page 236.
study of 81 women scheduled for labor induction, hourly oral temperatures were recorded
and analyzed
based on before
race, body
mass index,
duration ofcoronary
labor, andintervention
time to epidural.
The use of intracoronary thrombus
aspiration
primary
percutaneous
(PCI) in
temperature
rose ininfarction
a significant
linear trend
over time.
Positiveoftemperature
patients with ST-segmentOverall,
elevation
myocardial
(STEMI)
remains
a matter
debate. Intrends
this ranwere
associated
with significantly
longer
time
from membrane
rupture
to delivery
and higher
domized, registry base trial,
7,244
patients
with STEMI
were
allocated
to undergo
manual
thrombus
aspibody
mass index.
slopes
did that
not differ
before compared
after epidural
ration followed by PCI or to
undergo
PCI Temperature
alone. It was
found
a strategy
of routinewith
manual
thrombus
analgesia. Thiswith
study
suggests
that
epidural
analgesiaall-cause
alone doesmortality
not increaseorthe
riskcomposite
of high
aspiration before PCI, as compared
PCI
alone,
did
not reduce
the
Estimation of the Contribution
353
of death from any
cause, rehospitalization
for myocardial
temperatures in intrapartum
women. infarction, or stent thrombosis up to 1 yr. (Sumof Norketamine to mary:
KetamineJ. Mantz. Image: J.P. Rathmell.)
induced Acute Pain Relief and

Postoperative QT Interval Prolongation in Patients
Neurocognitive Impairment in
321
basis
for nicotine
as a gateway
drug. N Engl J Med 2014;
Undergoing
Noncardiac
Surgery
under General
Healthy VolunteersShattuck lecture. A molecular
371:93243.
Norketamine has an effect opposite
to that of ketamine
Anesthesia
on pain relief.
Epidemiologic studies have

suggested that
nicotine
use is a
gateway
to the
use of marijuana
and cocaine
Electrocardiograms
(ECG)
can identify
abnormal
cardiac
repolarization
by observation
of a
in human populations, butprolonged
the mechanisms
phenomenon
unclear.
QT interval. of
QTthis
interval
prolongationare
is often
caused In
bythis
drugsilluminating
and can resultarticle
in
written by prestigious authors
in cardiac
the domain
and medicine,
reader
will find
sudden
death. of
In molecular
this ancillaryneuroscience
study to the Vitamins
in Nitrous the
Oxide
trial, serial
a description of how nicotine
produces
its ECG
effects
the brain
of469
mice.
Theundergoing
conclusions
applied to
postoperative
12-lead
wereinobtained
from
patients
majorare
noncardiac
public health concerns that
are being
as the popularity
of electronic
with emphasurgery
under raised
general anesthesia.
Eighty percent
of patientscigarettes
experiencedrises,
a significant
QT
of the potential
to society
of translating
epidemiologic
findings
health
interval
prolongation,
and approximately
half had increases
greaterinto
than public
440 ms at
the endpolicy.
of
Severe Emergencesis
Agitation
399 benefits
J. Mantz. Image:
Nicotinic
Nature
Publishing
Used
permission.)
surgery.
One channel-receptor,
patient developed torsade
de pointes.
Drugs Group.
associated
withwith
prolonged
QT
after Myringotomy (Summary:
in a 3-yr-old
Child (Case Scenario)
Emergence agitation, the associated risk factors, and its
prevention and treatment are discussed.
January 2015
SCIENCE, MEDICINE, AND THE ANESTHESIOLOGIST

Key Papers from the Most Recent Literature Relevant to Anesthesiologists
Sepsis Severity Score: An internationally derived scoring system from the surviving
sepsis campaign database. Crit Care Med 2014; 42:196976.
Sepsis remains a major perioperative public health challenge. A better understanding of the risk factors predicting mortality can help to improve the level of care in severe sepsis. In this study including a database of
23,482 observations, a 34-item Sepsis Severity Score was generated and exhibited an excellent ability to discriminate between survivors and nonsurvivors. The most significant limitation is that of age and comorbidity.
As with all severity scoring models, the probabilities generated in these models should be used for gaining
further understanding of populations, not for decision making with regard to individual patients. (Summary:
J. Mantz. Image: J.P. Rathmell.)
Impact of closure at the first take back: Complication burden and potential
overutilization of damage control laparotomy. J Trauma 2011; 71:150311.
Damage control laparotomy (DCL) has been used as a lifesaving technique that minimizes coagulopathy,
hypothermia, and acidosis. However, the morbidity carried by overutilization of this technique remains
unknown. In this retrospective cohort including 925 trauma patients who underwent immediate DCL, early
fascial closure was found to be an independent predictor of reduced complications in DCL patients. This
may represent an overutilization of this valuable technique, exposing patients to increased complications.
(Summary: J. Mantz. Image: J.P. Rathmell.)
Acupuncture for chronic knee pain: A randomized clinical trial. JAMA 2014;
312:131322.
Acupuncture is a popular approach to pain control. However, few studies in this area have been well
powered, and adequate control groups have often not been included. In their recent randomized trial
involving 282 patients, Hinman et al. evaluated acupuncture on chronic knee pain at the end of treatment
(12 weeks) and at 1 yr. The study was remarkable for its use of a control (no treatment) and sham treatment group as well as groups receiving each of two alternative forms of acupuncture. The investigators
found that compared with sham treatment the traditionally and laser acupuncturetreated patients did
not achieve significant improvement in average knee pain or function, the primary outcomes in this trial.
This study does not support the use of acupuncture for knee pain, and demonstrates the need for powerful study designs to be
applied to alternative as well as allopathic pain therapies. (Summary: J.D. Clark. Image: Thinkstock.)
Doing todays work superbly welltreating Ebola with current tools.

N Engl J Med 2014; 371:15656.
News of the Ebola outbreak across West Africa is taking over the lay press and scholarly medical publications. Articles describing this disease evoke fear among the public and clinicians as well as a concern
that, if there is no treatment, nothing can be done. In this superb perspective, the authors point out a most
important educational message for every care provider: reduction of mortality does not only rely on new
high-tech antiviral drugs, but also on excellent routine clinical care with simple interventions (i.e., fluid
loading to prevent deaths attributable to hypovolemia, rational use of basic routine biology, strict isolation
measures of suspect cases) that will help to break transmission chains. (Summary: J. Mantz. Image: Colorized transmission electron micrograph of the Ebola virus, Centers for Disease Control and Prevention.)
January 2015
INFOGRAPHICS IN ANESTHESIOLOGY
Input-ebh
Complex Information for Anesthesiologists Presented Quickly and Clearly

THIS MONTH IN

Pathophysiology to Treatment
(Clinical Concepts and
Commentary)
Recent insights into posttraumatic immune dysfunction
have defined new targets for immunointervention that
hold promise for improving outcomes in such critically ill
patients.
High Intraoperative Inspired

Oxygen Requirements
271
High inspired oxygen may be reasonable in lower risk

Accuracy of Ultrasound-guided 347

Nerve Blocks of the Cervical
Ultrasound imaging was an accurate technique for cervical zygapophysial joint nerve blocks in volunteers. See the
accompanying Editorial View on page 236.

Our Specialty
243
The proportion of anesthesiology residents from U.S. medical schools has more than doubled since 1995. This retrospective cohort study evaluated the 2010 and 2011 residency
applicants to determine the
factors associated with a successful admission to residency training programs.
The sample represented 58%
of the total national applicant
pool; 66% of the applicants
successfully matched to anesthesiology. The odds for a successful match were higher for
applicants from U.S. medical
schools, those with United
States Medical Licensing
Examination scores greater
than 210, younger applicants, and females. Prior
graduate education or peerreviewed publications did not offer any advantage. This study suggests the potential for age
and gender bias in the selection process. See the accompanying Editorial View on page 230.
What Factors Affect Intrapartum Maternal Temperature?

A Prospective Cohort Study: Maternal Intrapartum
Temperature
302
study of 81 women scheduled for labor induction, hourly oral temperatures were recorded
and analyzed based on race, body mass index, duration of labor, and time to epidural.
Overall, temperature rose in a significant linear trend over time. Positive temperature trends
were associated with significantly longer time from membrane rupture to delivery and higher
body mass index. Temperature slopes did not differ before compared with after epidural
analgesia. This study suggests that epidural analgesia alone does not increase the risk of high
temperatures in intrapartum women.
Estimation of the Contribution 353

of Norketamine to KetamineAcute
PainofRelief
andtransfused by service directly from the 2011 National Blood Collection and Utilization Survey Report.* The mediWeinduced
extracted the
percentage
erythrocytes
Postoperative
QT Interval
Prolongation
Patients
Neurocognitive
Impairment
inhematologyoncology
321
cine
and pediatric (med/peds)
services and
(hem/onc) services
were grouped
together in notin
having
the potential for anesthesiologist
Undergoing
Noncardiac
Surgery
under
General
Healthy
Volunteers
involvement, whereas combined surgical departments (surgery), trauma/emergency room, medical and surgical intensive care units (ICU), obstetrics/gyneNorketamine
has and
an effect
opposite
to that
of ketamine outpatient)
Anesthesia
cology
(OB/GYN),
other
services
(anesthesiology,
were deemed areas in which anesthesiologists may be involved in decisions to transfuse.
on pain relief.
Infographic created by Jonathan P. Wanderer, VanderbiltElectrocardiograms
University School (ECG)
of Medicine,
and James
P. Rathmell,
Massachusetts
General Hospital/
can identify
abnormal
cardiac repolarization
by observation
of a
Harvard Medical School. Illustration by Annemarie Johnson,
Vivo Visuals.
Address
to Dr.is Wanderer:
prolonged
QT interval.
QTcorrespondence
interval prolongation
often causedjon.wanderer@vanderbilt.edu.
by drugs and can result in
sudden cardiac death. In this ancillary study to the Vitamins in Nitrous Oxide trial, serial
1. Clifford L, Jia Q, Yadav H, Subramanian A, Wilson GA, Murphy SP, Pathak J, Schroeder DR, Ereth MH, Kor DJ: Characterizing the epidemiology
postoperative 12-lead ECG were obtained from 469 patients undergoing major noncardiac
of perioperative transfusion-associated circulatory overload. Anesthesiology 2015; 122:21-8
surgery under general anesthesia. Eighty percent of patients experienced a significant QT
2. Clifford L, Jia Q, Subramanian A, Yadav H, Wilson GA, Murphy SP, Pathak J, Schroeder DR, Kor DJ: Characterizing the epidemiology of postinterval prolongation, and approximately half had increases greater than 440 ms at the end of
Severe
Emergence
Agitation
399
operative transfusion-related acute lung injury. Anesthesiology 2015; 122:12-20
surgery. One patient developed torsade de pointes. Drugs associated with prolonged QT
after
Myringotomy
in
a
3-yr-old
3. de Almeida JP, Vincent J-L, Galas FRBG, de Almeida EPM, Fukushima JT, Osawa EA, Bergamin F, Park CL, Nakamura RE, Fonseca SMR,
Child
(Case
Scenario)
Cutait G, Alves JI, Bazan M, Vieira S, Vieira Sandrini AC, Palomba H, Ribeiro U Jr, Crippa A, Dalloglio M, del Pilar Estevez Diz M, Filho
Emergence
agitation,
the
associated
risk
factors,
and
its
RK, Costa Auler JO Jr, Rhodes A, Haijar LA: Transfusion requirements in surgical oncology patients: A prospective, randomized controlled trial.
prevention
and
treatment
are
discussed.
Anesthesiology 2015; 122:29-38
* The 2011 National Blood Collection and Utilization Survey Report. Available at: http://www.hhs.gov/ash/bloodsafety/2011-nbcus.pdf. Accessed
October 29, 2014.
January 2015
ANESTHESIOLOGY CME PROGRAM
Instructions for Obtaining ANESTHESIOLOGY Continuing

Medical Education (CME) Credit
CME Editors: Leslie C. Jameson, M.D., and Dan J. Kopacz, M.D.
Anesthesiologys Journal CME is open to all readers. To take part
in Anesthesiology Journal-based CME, complete the following
steps:
1.Read the CME information presented on this page.
2.Read this months article designated for CME credit (listed on
the right) in either the print or online edition.
3.
Register at http://education.asahq.org/2015-journal-cme.
Nonmembers will be asked to provide payment.
4.Achieve a score of at least 50% correct on the six-question online journal CME quiz and complete the evaluation.
5.Claim credit in 15-minute increments, for a maximum of 1
AMA PRA Category 1 Credit per journal article.
Cancer: When the Problem, the Solution, and Their Combination Are
Each Associated with Poor Outcomes on page 3 of this issue.
Learning Objectives
After successfully completing this activity, the learner will be able
to apply the risks of anemia to perioperative cancer patients, apply
the risks of blood transfusion to perioperative cancer patients, and
recognize the differences in the risk of blood transfusion between
specific patient groups.
Disclosures
Editor-in-Chief: James C. Eisenach, M.D., receives consulting
fees from Aerial BioPharma LLC and Cubist Pharmaceuticals, Inc.
CME Information & Disclosure

Purpose: The focus of Anesthesiology Journal-based CME is to
educate readers on current developments in the science and clinical
practice of anesthesiology.
Target Audience: Anesthesiology Journal-based CME is intended for anesthesiologists. Researchers and other health care professionals with an interest in anesthesiology may also participate.
Accreditation: The American Society of Anesthesiologists is accredited by the Accreditation Council for Continuing Medical
Education to provide continuing medical education for physicians.
CME Designation Statement: The American Society of Anesthesiologists designates this Journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit. Physicians should claim
only the credit commensurate with the extent of their participation
in the activity.
Rates
CME Editors: Leslie C. Jameson, M.D., receives honoraria from

GE Medical International and Masimo Corporation. Dan J.
Kopacz, M.D., has an equity position in SoloDex, LLC.
Authors: Juliano Pinheiro de Almeida, M.D., Jean-Louis Vincent,
M.D., Ph.D., Filomena Regina Barbosa Gomes Galas, M.D., Ph.D.,
Elisangela Pinto Marinho de Almeida, M.D., Julia T. Fukushima,
M.Sc., Eduardo A. Osawa, M.D., Fabricio Bergamin, M.D., Clarice
Lee Park, M.D., Rosana Ely Nakamura, M.D., Silvia M. R. Fonseca,
M.D., Guilherme Cutait, M.D., Joseane Inacio Alves, R.N., Mellik Bazan, P.T., Silvia Vieira, R.N., Ana C. Vieira Sandrini, L.D.N.,
Henrique Palomba, M.D., Ph.D., Ulysses Ribeiro, Jr., M.D., Ph.D.,
Alexandre Crippa, M.D., Marcos Dalloglio, M.D., Ph.D., Maria
del Pilar Estevez Diz, M.D., Ph.D., Roberto Kalil Filho, M.D.,
Ph.D., Jose Otavio Costa Auler, Jr., M.D., Ph.D., Andrew Rhodes,
M.B., B.S., and Ludhmila Abrahao Hajjar, M.D., Ph.D., have reported no financial relationships with commercial interests.
Author: Juan P. Cata, M.D., has reported no financial relationships
with commercial interests.
Two options are available:

ASA Member
Annual Fee
Per Issue
$0
$0
Non-member
$120
$20
Payment may be made using Visa or MasterCard.

Please direct any questions about Journal-based CME to:
EducationCenter@asahq.org
Date of Release: December 2014
Termination Date: December 2017
This Months Anesthesiology Journal-based

CME Article
Read the article by Pinheiro de Almeida et al. entitled Transfusion Requirements in Surgical Oncology Patients: A Prospective, Randomized
Controlled Trial on page 29 and the accompanying editorial by Cata
entitled Perioperative Anemia and Blood Transfusions in Patients with
ASA Staff: Kari L. Lee, Editorial Manager, and Ginger Yarger, Editor,
have reported no financial relationships with commercial interests.
Resolution of Conflicts of Interest

In accordance with the ACCME Standards for Commercial Support of CME, the American Society of Anesthesiologists has implemented mechanisms, prior to the planning and implementation of
this Journal-based CME activity, to identify and resolve conflicts of
interest for all individuals in a position to control content of this
Journal-based CME activity.
Disclaimer
The information provided in this CME activity is for continuing
education purposes only and is not meant to substitute for the independent medical judgment of a health care provider relative to diagnostic and treatment options of a specific patients medical condition.
EDITORIAL VIEWS
Revealing the Real Risks of Perioperative Transfusion

Rise of the Machines!
Jeffrey W. Simmons, M.D., Jean-Francois Pittet, M.D.
pproximately
21,000,000 blood components were transfused in 2011 in
the United States. Almost half of
those were transfused in the operating room or perioperatively according to the 2011 National Blood
Collection and Utilization Survey
Report.* Consequently, the anesthesiologist assumes a substantial
role in mitigating adverse transfusion reactions. In fact, one adverse
reaction per 414 units transfused
was reported in the above survey,
accounting for 50,570 total.* Many
transfusion reactions can be identified with predictable laboratory
diagnostic tests. However, when
no specific test exists or concrete
definitions are not established, the
diagnosis is often missed or underappreciated. As a result,
the transfusion reaction remains in obscurity; if proper safeguards are not implemented, the possibility of occurring again
is not deterred. This is the case in transfusion-related acute
lung injury (TRALI) and transfusion-associated circulatory
overload (TACO). In this months edition of Anesthesiology,
Clifford et al. have taken an enormous step in providing better detection of these common adverse reactions.1,2
Why is reporting a seemingly obvious reaction that is temporally related to a visible intervention so difficult? On paper,
a patient without lung injury developing acute respiratory
distress within hours of receiving a blood transfusion would
seem likely to raise red flags. Unfortunately, the clinical picture is often blurred with multiple confounders. Dyspnea
is expected in patients presenting with preexisting cardiac
conditions or underlying pulmonary disease. Isolating one
culprit among multiple interventions occurring in surgical
patients is often not possible or may not have clinical relevance when choosing the next therapy. Indeed, the treatment
for acute respiratory failure from

adult respiratory distress syndrome and circulatory overload
is similar regardless of etiology. In
addition, reporting depends on
the specialty of the physicians and
how they diagnose TRALI and
TACO.3 Personal reasons may
also factor into reporting. Fear
of reprisal, apathy, or unawareness of the reporting mechanisms
must be taken into account.
Finally, to steal a sports phrase,
the concept of no harm, no
foul may be a reason for underreporting. To address these problems, Clifford et al. minimized
the human element and used the
robust databases of Mayo Clinic
to examine noncardiac surgical
patients over two different years. In this retrospective analysis,
83,204 patients were electronically screened for signs and symptoms consistent with TRALI or TACO. The precision of this
study was augmented by natural language-processing software
developed at the Mayo Clinic. This electronic data-miner
evaluates radiographic reports looking for explicit or descriptive
terms used to identify TRALI and TACO. The importance of
this initial screening tool cannot be underestimated. Thousands
of hours would be required for detailing such records making
a study of this size implausible for many institutions. In addition, the fidelity of the natural language-processing software is
high, producing a specificity and sensitivity of 92.5 and 93.6%,
respectively.1 Once limited to manageable numbers, the positively screened patients were evaluated by independent physicians for actual cases of TRALI or TACO.
The Mayo Clinic studies focused on patients transfused in
the perioperative period in an effort to categorize this unique
population. As such, postoperative incident rates of 1.4% for
TRALI and 4.3% for TACO may not be generalizable to the
The potential for this type of

technology is extraordinary.
Image: Shutterstock.
Corresponding articles on page 12 and page 21.
Accepted for publication September 3, 2014. From the Department of Anesthesiology ( J.W.S.) and Departments of Anesthesiology, Surgery, and Cell and Developmental and Integrative Biology ( J.-F.P.), University of Alabama at Birmingham, Birmingham, Alabama.
* The 2011 National Blood Collection and Utilization Survey Report. Available at: http://www.hhs.gov/ash/bloodsafety/2011-nbcus.pdf.
Accessed October 29, 2014.
Copyright 2014, the American Society of Anesthesiologists, Inc. Lippincott Williams & Wilkins. Anesthesiology 2015; 122:1-2
January 2015
Editorial Views
entire transfused population. The incidence rate of TRALI in

the critically ill population, as might be common at the Mayo
Clinic, has been estimated to be between 5.1 and 15%.4,5
Also, by combining highly probable cases with actual cases,
the incident rates that were calculated have the potential to be
artificially high. However, requiring imputability or identification of transfusion as the sole causal factor in acute respiratory
failure from adult respiratory distress syndrome and circulatory
overload results in underappreciation of the disease. Coinciding with underrecognition, the Mayo Clinic researchers limited
their results to 6h posttransfusion. This was intentionally done;
however, expanding their definition of the perioperative period
may have identified an even greater incidence. The postoperative
period can range from 6 to 24h depending on the disease studied. In an effort to generalize these results, multicenter efforts
using the Mayo Clinic natural language-processing software
could be used, and the definition of the postoperative period
standardized. Highlighted in this study is the use of computer
automation to examine thousands of records in an electronic
database. In this study, language-processing software retrospectively identified low-incident, high-morbidity processes with
astounding fidelity. Will it be possible for this same software
to perform concurrent searches during a patients hospital stay
to detect disease patterns and assist in decision making? The
potential for this type of technology is extraordinary.
Transfusion-related acute lung injury and TACO are the
number one and two causes of transfusion-related deaths in
the United States. In 2013, the Federal Drug Administration
reported that 38 and 24% of the deaths after transfusion were
attributable to TRALI and TACO. Transfusion could not be
ruled out as the cause in an additional 32% of deaths (approximately 21 of 71 cases in FY2013).6 Thus, recognition of the
adverse reaction and accurate reporting are crucial in mitigating risk to patients. Through identification of patients with
TRALI and TACO, risk stratification regimens can be better
constructed. Epidemiological studies such as these studies pave
the way for tailored transfusion practices for patients at high risk
for TRALI or TACO. Importantly, recognition of blood components that have caused TRALI will allow for future exclusion
of those donors. As well, blood component traits can be identified that put patients at higher risk for these disease processes.
Whether transfusion-related adverse reactions are not
reported due to human error or underrecognition, the risk to
patients remains high. Through use of intelligent automation,
Clifford et al. have identified the overall risk of TRALI at 1.4%
and TACO at 4.3% in the perioperative patient. For TRALI,
dramatic underrecognition of the disease was identified. In
addition, increasing volume transfused and increasing age was
clinically predictive of developing TRALI. The Mayo Clinic
study identified increased risk of TACO with certain surgical
procedures, increased transfusion volume, and total operative
fluid balance. This coincides with know risk factors published
by the TRALI Study Group.7 When transfusion-related adverse
reactions occurred, hospital stay was prolonged and mortality
was augmented (odds ratio, 15.6 for TRALI).
The landscape of hemotherapy is ever-changing. Clinicians

should seek to better recognize adverse transfusion reactions.
The goal of minimizing transfusions to limit patient exposure
to adverse reactions is reasonable and practical. To accomplish
this, patient blood management strategies are being widely
implemented and have been successful in reducing blood transfusions in tertiary care centers.8 The principles of patient blood
management are based on the three pillars concept: optimizing preoperative erythropoiesis, reducing operative blood loss,
and harnessing the patients physiological tolerance of anemia.9
Anesthesiologists should recognize this changing landscape as
we are often on the front lines of blood transfusion.
Acknowledgments
Supported by the National Institutes of Health, Bethesda,
Maryland (grant no. RO1 GM086416, to Dr. Pittet).
Competing Interests
The authors are not supported by, nor maintain any financial interest in, any commercial activity that may be associated with the topic of this article.
Correspondence
Address correspondence to Dr. Pittet: pittetj@uab.edu
References
1. Clifford L, Jia Q, Subramanian A, Yadav H, Wilson GA,
Murphy SP, Pathak J, Schroeder DR, Kor DJ: Characterizing
the epidemiology of postoperative transfusion-related acute
lung injury. Anesthesiology 2015; 122:1220
2. Clifford L, Jia Q, Yadav H, Subramanian A, Wilson GA, Murphy
SP, Pathak J, Schroeder DR, Ereth MH, Kor DJ: Characterizing
the epidemiology of perioperative transfusion-associated circulatory overload. Anesthesiology 2015; 122:218
3. Vlaar AP, Wortel K, Binnekade JM, van Oers MH, Beckers E,
Gajic O, Schultz MJ, Juffermans NP: The practice of reporting
transfusion-related acute lung injury: A national survey among
clinical and preclinical disciplines. Transfusion 2010; 50:44351
4. Benson AB, Austin GL, Berg M, McFann KK, Thomas S,
Ramirez G, Rosen H, Silliman CC, Moss M: Transfusionrelated acute lung injury in ICU patients admitted with gastrointestinal bleeding. Intensive Care Med 2010; 36:17107
5. Vlaar AP, Juffermans NP: Transfusion-related acute lung
injury: A clinical review. Lancet 2013; 382:98494
6. Research CfBEa: Fatalities Reported to FDA Following Blood
Collection and Transfusion: Annual Summary for Fiscal Year
2013. Administration USFaD, ed2013
7. Murphy EL, Kwaan N, Looney MR, Gajic O, Hubmayr RD,
Gropper MA, Koenigsberg M, Wilson G, Matthay M, Bacchetti
P, Toy P; TRALI Study Group: Risk factors and outcomes in
transfusion-associated circulatory overload. Am J Med 2013;
126:357.e2938
8. Leahy MF, Roberts H, Mukhtar SA, Farmer S, Tovey J, Jewlachow
V, Dixon T, Lau P, Ward M, Vodanovich M, Trentino K, Kruger
PC, Gallagher T, Koay A, Hofmann A, Semmens JB, Towler
S; Western Australian Patient Blood Management Program: A
pragmatic approach to embedding patient blood management
in a tertiary hospital. Transfusion 2014; 54:113345
10. Leahy MF, Mukhtar SA: From blood transfusion to patient
blood management: A new paradigm for patient care and
cost assessment of blood transfusion practice. Intern Med J
2012; 42:3328
2
J. W. Simmons and J.-F. Pittet
Perioperative Anemia and Blood Transfusions in

Patients with Cancer
When the Problem, the Solution, and Their Combination
Are Each Associated with Poor Outcomes
Juan P. Cata, M.D.
This article has been selected for the Anesthesiology CME Program. Learning objectives
and disclosure and ordering information can be found in the CME section at the front
of this issue.
N this months issue of

Anesthesiology, de Almeida
etal.1 elegantly investigated
postoperative administration of
packed erythrocytes in patients
who were admitted to the surgical
intensive care unit after abdominal cancer surgery. The study
was a controlled, parallel-group,
double-blind superiority trial in
which patients were randomized
to receive blood transfusions using
restrictive (hemoglobin <7
g/dl)
or liberal (hemoglobin <9
g/dl)
criteria. The authors found that
patients randomized to the liberal
strategy had a lower absolute risk
of the 30-day primary composite
endpoint that was a composite of
all-cause mortality, cardiovascular
complications, acute respiratory
distress syndrome, acute kidney
injury requiring renal replacement
therapy, septic shock, or reoperation. Patients assigned to the liberal transfusion strategy also had a
lower 60-day mortality.
The prevalence of anemia in
patients with cancer ranges from 30 to 90%.2 Perioperative
anemia in this population is multifactorial and can be the
result of the cancer per se, its treatment, or chronic kidney disease. Preoperative anemia in patients with cancer usually is the
result of blood loss due to advanced cancer or bone marrow
suppression secondary to inflammation or myelotoxicity. In
contrast, intraoperative and postoperative anemia results from
surgical blood loss or administration of fluids (dilutional).
Recently, Loor et al.3 developed the concept of three
evils with respect to anemia and transfusion to explain the
association of each and the combination of both with poor
outcomes after cardiac surgery.

This concept can also be applied
in patients with cancer; however,
patients with cancer are exposed to
different risk than cardiac patients
and the overall surgical population. Treating moderate-to-severe
perioperative anemia (first evil)
should be considered in patients
with cancer for two reasons. First,
anemia per se can increase the
risk of cancer recurrence. Low
hemoglobin concentrations can
render cancer cells more aggressive through several mechanisms,
including tumoral hypoxia and
inducing the release of interleukin-6.4 Second, patients with cancer appear to tolerate anemia less
than the general population. The
hemoglobin concentration that is
associated with an increase in early
postoperative morbidity and mortality in cancer surgery patients
appears to be about 8g/dl. The
odds of death increase 2.5-fold for
every 1-g decrease in postoperative hemoglobin concentrations
below 8g/dl.5 And the mortality risk sharply increases in
patients with hemoglobin concentrations less than 6g/dl.
Increased mortality can result from failure of adaptive and
compensatory mechanisms to hypoxia to improve oxygen
delivery to organs; an exaggerated inflammatory response
to severe acute anemia, erythropoietin release, and surgery
leading to worsening endothelial dysfunction and organ
ischemia; or a poor bone marrow angiogenic and regenerative response (i.e., an impaired endothelial progenitor cell
response) resulting from very low hemoglobin concentrations and inflammation.
The hemoglobin concentration that is associated

with an increase in early postoperative morbidity and mortality in cancer
surgery patients appears to
be about 8g/dl.
Image: A. Johnson.
Corresponding article on page 29.
Accepted for publication September 4, 2014. From the Department of Anesthesiology and Perioperative Medicine, The University of Texas
MD Anderson Cancer Center, Houston, Texas.
January 2015
Editorial Views
Other important points to consider in terms of the

potential impact of anemia on postoperative outcomes are
the patients age, rapidity of anemia onset, and the hemoglobin-duration deficit product (or duration below a critical
hemoglobin value).6 Very low hemoglobin concentrations
are generally well tolerated in young individuals, but older
patients appear to be at a high risk for anemia-related mortality, possibly owing to age-associated limited organ reserve.
The duration below a critical hemoglobin value also deserves
attention because a delay in administering erythrocytes to
improve oxygen delivery is directly related to mortality in
patients whose hemoglobin concentrations are below 8g/dl.7
Blood transfusion is the mainstay therapy for moderate-tosevere perioperative anemia; however, administration of blood
products is associated with poor outcomes in patients with
and without cancer. In those with cancer, blood transfusions
can directly (via soluble factors) and indirectly (via transfusion
related immune modulation) induce proliferation and spread
cancer cells present at sites of minimal residual disease (tumor
margins and dormant tumors).8 After transfusions, there is an
increase in the concentrations of local (tumor microenvironment) and circulating proinflammatory cytokines and prostaglandin E, which tilt the balance toward immune suppression.
At the cellular level, there is a reduction in the function of natural killer cells; a decrease in the proliferation of CD4+, CD8+
T cells, and B lymphocytes; induction of T regulatory cells;
and a decrease in maturation and antigen-presenting activity of dendritic cells, which also contributes to a diminished
immune function.9 Furthermore, administration of blood
products may facilitate proliferative and metastatic properties
of cancer cells via angiogenic and oncogenic factors leaked
from stored erythrocytes, a phenomenon associated with the
so called storage lesion.10 Hence, patients with cancer are at a
higher risk of tumor recurrence after receiving blood transfusions, which constitutes the second evil.
Last, the combination of intraoperative or postoperative anemia and blood transfusion, the third evil, may
cause the most harm.3 In the context of an ongoing insult
as occurs during surgery, the summation of the response
to moderate-to-severe anemia (first hit) followed by the
response to the blood transfusion (second hit) may cause
an exaggerated systemic inflammatory and immune suppressive response, and endothelial dysfunction.11 In the
general surgical population, this can translate into a higher
risk for early postoperative morbidity and mortality. In
patients with cancer, the consequences of inflammation
and immune paralysis can be seen as an increased risk for
recurrence or cancer-specific mortality.
In conclusion, moderate-to-severe anemia and blood transfusions induce marked changes in the homeostasis of several
physiological processes, including endothelial function, complement activation, the inflammatory response, and immune
function, all of which have been linked to the pathogenesis of
end-organ ischemia (particularly in older patients and those
with advanced atherosclerotic disease). They have also been
linked to cancer recurrence. Until oxygen delivery can be

monitored precisely, perioperative physicians must rely on a
hemoglobin value to safely transfuse or not patients during
and after surgery. The work by de Almeida et al.1 supports
previous studies, indicating that perioperative anemia is the
predictor of mortality in patients with cancer. Their work
indicates that maintaining a hemoglobin concentration above
9g/dl is prudent in cancer surgery patients. In those patients
who are at risk of developing significant anemia during or after
surgery (hemoglobin <9g/dl), preoperative administration of
blood transfusions, administration of iron supplements, and
minimally invasive surgical techniques may prove helpful,
although none has been specifically tested in this context.
Competing Interests
The author is not supported by, nor maintains any financial
interest in, any commercial activity that may be associated
with the topic of this article.
Correspondence
Address correspondence to Dr. Cata: jcata@mdanderson.org
References
1. de Almeida JP, Vincent J-L, Galas FRBG, de Almeida EPM,
Fukushima JT, Osawa EA, Bergamin F, Park CL, Nakamura
RE, Fonseca SMR, Cutait G, Alves JI, Bazan M, Vieira S, Vieira
Sandrini AC, Palomba H, Ribeiro U Jr, Crippa A, Dalloglio M,
del Pilar Estevez Diz M, Filho RK, Auler JOC Jr, Rhodes A,
Hajjar LA: Transfusion requirements in surgical oncology patients: A prospective, randomized controlled trial.
Anesthesiology 2015; 122:2938
2. Knight K, Wade S, Balducci L: Prevalence and outcomes of
anemia in cancer: A systematic review of the literature. Am J
Med 2004; 116(suppl 7A):11S26S
3. Loor G, Rajeswaran J, Li L, Sabik JF 3rd, Blackstone EH,
McCrae KR, Koch CG: The least of 3 evils: Exposure to red
blood cell transfusion, anemia, or both? J Thorac Cardiovasc
Surg 2013; 146:14801487.e6
4. Semenza GL: HIF-1 and tumor progression: Pathophysiology
and therapeutics. Trends Mol Med 2002; 8(4 suppl):S627
5. Carson JL, Noveck H, Berlin JA, Gould SA: Mortality and morbidity in patients with very low postoperative Hb levels who
decline blood transfusion. Transfusion 2002; 42:8128
6. Gilreath JA, Stenehjem DD, Rodgers GM: Diagnosis and treatment of cancer-related anemia. Am J Hematol 2014; 89:20312
7. Mackenzie CF, Moon-Massat PF, Shander A, Javidroozi M,
Greenburg AG: When blood is not an option: Factors affecting survival after the use of a hemoglobin-based oxygen carrier in 54 patients with life-threatening anemia. Anesth Analg
2010; 110:68593
8. Acheson AG, Brookes MJ, Spahn DR: Effects of allogeneic
red blood cell transfusions on clinical outcomes in patients
undergoing colorectal cancer surgery: A systematic review
and meta-analysis. Ann Surg 2012; 256:23544
9. Cata JP, Wang H, Gottumukkala V, Reuben J, Sessler DI:
Inflammatory response, immunosuppression, and cancer recurrence after perioperative blood transfusions. Br J
Anaesth 2013; 110:690701
10. Glynn SA: The red blood cell storage lesion: A method to the
madness. Transfusion 2010; 50:11649
11. Bilgin YM, Brand A: Transfusion-related immunomodulation:
A second hit in an inflammatory cascade? Vox Sang 2008;
95:26171
4
Juan P. Cata
Anaphylaxis to Neuromuscular-blocking Drugs

All Neuromuscular-blocking Drugs Are Not the Same
Paul Michel Mertes, M.D., Ph.D., Gerald W. Volcheck, M.D.
n the current issue of Anesthe, Reddy et al.1 report a

two-hospital, retrospective, observational, cohort study confirming
that anaphylaxis is more common
with rocuronium and succinylcholine than with atracurium, a topic
that is difficult to assess and was first
highlighted in this journal in 2003.2
Although any medication can potentially cause perioperative anaphylaxis, neuromuscular-blocking drugs
(NMBDs), antibiotics, latex, and
chlorhexidine are the most likely to
do so. Regional differences regarding
the relative risk of allergic reactions
to NMBDs do exist. NMBDs represent the dominant causes of anaphylaxis in several countries and regions
such as France,24 Norway,5 Spain,6
and Australasia,7 whereas other
agents may be primarily involved in
other countries.8 Nevertheless, allergic reactions to NMBDs remain a
serious concern for anesthesiologists
because death may occur even when
reactions are rapidly and adequately
treated.9 The reported incidence of
perioperative anaphylaxis is quite
varying, ranging between 1:3,500
and 1:20,000. Part of the variability
is likely due to difficulty in determining the exact exposures to the numerous drugs, blood
products, and agents used in the operative setting. The number of documented cases of intraoperative anaphylaxis is
typically reported in aggregate for a large population, leaving
the specifics of the total amount and type of medications the
population was exposed to in question.
In the study by Reddy et al., the authors take the advantage of their ability to retrieve detailed information concerning new patient exposure to each NMBD from electronic
anesthetic records available in the two participating centers
over 7 yr. This allowed a more precise estimate of the number of patients exposed as the denominator when calculating the relative risk of allergic reactions associated with the
use of each NMBD. This method

helps eliminate the primary concern
with data based on drug sales, which
have the potential to overestimate
the exposure resulting in a potential underestimation of anaphylaxis
rate. Interestingly, the authors findings are similar to the estimates of
allergic reactions to NMBDs based
on drug sales. This study confirms
the increased relative risk of allergic reaction to succinylcholine and
rocuronium in countries where a
high rate of reaction to NMBDs is
reported.
The surveillance of intraoperative
adverse drug reactions still represents a clinical and statistical challenge10 because these reactions are
rare, random, and mostly independent from the repeated exposure of
patients to anesthesia. In addition,
possible biases and underreporting
make comparison between drugs
relatively difficult. Another weakness of any reporting system is that
responsible physicians seem to have
little understanding of which drug
is actually causing the anaphylactic reaction when several drugs are
simultaneously administered during anesthesia induction due to a
lack of a single confirmatory test.11 With thorough review
in this study, it was noted that 9 of the 21 cases of identified NMBD anaphylaxis did not meet the standard skin test
criteria for positivity but correctly warranted inclusion based
on clinical picture and adjunct testing.
Because identification of the anaphylactic mechanism,
of the responsible drug, and of the alternative safe agents
is not always straightforward, a standard use of tryptase
measurements in case of suspected allergic reactions and
investigation of these reactions in compliance with established guidelines12 by allergists trained in the field of drug
allergy working in close collaboration with anesthesiologists should be promoted.13,14 Reddy et al. confirm that
siology
There are many

factors
that will influence the
choice of a specific NMBD,
depending on the clinical situation, [including] the likely increased
allergic risk
associated
with succinylcholine and
rocuronium....
Image: Thinkstock.
Accepted for publication September 15, 2014. From the Strasbourg Medical School, Hpitaux Universitaires de Strasbourg, Nouvel Hpital
Civil, Strasbourg, Cedex, France (P.M.M.); and Mayo Medical School, Rochester, Minnesota (G.W.V.).
January 2015
Editorial Views
patients.7,22 There are many factors that will influence the

choice of a specific NMBD, depending on the clinical
situation, but the likely increased allergic risk associated
with succinylcholine and rocuronium, and the relatively
low risk associated with atracurium and even more so with
cisatracurium must be part of the clinical reasoning when
considering the use of a NMBD.
allergic reactions are associated with greater tryptase and

greater severity than nonallergic cases. Because no predictive test can help us to identify patients at risk before any
reaction, reduction of the risk of perioperative anaphylaxis
can only be based on secondary prevention.12 This report
provides a strong motivation for a thorough and systematic
investigation of any hypersensitivity reactions occurring
during the perioperative period15 to avoid any undesirable
subsequent exposure to an offending agent toward which
one is already sensitized. This necessity is further supported
by the small number of minor reactions diagnosed in this
study, probably related to under-referral of mild reactions
to all agents, a reality clearly demonstrated in the literature.4,15 The authors were not able to determine the number of reactors who were receiving anesthesia for the first
time, had a history of multiple anesthetic exposure or even
history of previous reaction. This information would be
helpful in future studies in determining sensitization patterns. Going forward, studies of intraoperative anaphylaxis
should include a standard definition of anaphylaxis, uniform skin testing, specific immunoglobulin E drug testing,
tryptase measurements, and review by an allergist in conjunction with an anesthesiologist.
The risk of allergic reactions is not the only drug characteristic that anesthesiologists must take into account
when making their clinical choice. In view of the number
of side effects associated with the use of succinylcholine,
a controversy exists concerning replacing this old drug by
rocuronium for rapid sequence induction.16 Nevertheless, because of their rapid onset of effect, both drugs will
remain essential in the anesthesiologists armamentarium.
Another interesting point that must be considered is that
rocuronium can be rapidly reversed by sugammadex, a
possibility that can make rocuronium a drug of choice
in countries where sugammadex is available.17 Sugammadex has also recently been proposed to improve recovery in case of anaphylaxis to rocuronium18; however, its
ability to play a role in reaction reversal remains controversial.19,20 Moreover, hypersensitivity reactions, either
allergic or not, have been reported with sugammadex,21
and this drug has not been approved in the United States
at present.
Due to the amount of vecuronium exposures, Reddy et
al. were not able to provide specific information concerning the risk associated with its use. This drug has been
shown to have a lower risk of anaphylaxis than rocuronium
in large epidemiologic studies22 and its effect can also be
effectively reversed by sugammadex.23 They considered
atracurium to be a safe alternative but were not able to
comment on the relative risk associated with cisatracurium
because this drug is not in use in Australasia. Cisatracurium has been shown to have the lowest risk of hypersensitivity reactions, either allergic or not, in large cohort
studies,3,22 and has also been shown to have the lowest
rate of cross-sensitization with other NMBDs in allergic
Competing Interests
Correspondence
Address correspondence to Dr. Mertes: paul-michel.mertes@
chru-strasbourg.fr
References
1. Reddy JI, Cooke PJ, van Schalkwyk JM, Hannam JA, Fitzharris
P, Mitchell SJ: Anaphylaxis is more common with rocuronium
and succinylcholine than with atracurium. Anesthesiology
2015; 122:3945
2. Mertes PM, Laxenaire MC, Alla F: Anaphylactic and anaphylactoid reactions occurring during anesthesia in France in
19992000. Anesthesiology 2003; 99:53645
3. Dong SW, Mertes PM, Petitpain N, Hasdenteufel F, Malinovsky
JM; GERAP: Hypersensitivity reactions during anesthesia.
Results from the ninth French survey (20052007). Minerva
Anestesiol 2012; 78:86878
4. Mertes PM, Alla F, Trchot P, Auroy Y, Jougla E; Groupe
dEtudes des Ractions Anaphylactodes Peranesthsiques:
Anaphylaxis during anesthesia in France: An 8-year national
survey. J Allergy Clin Immunol 2011; 128:36673
5. Harboe T, Guttormsen AB, Irgens A, Dybendal T, Florvaag E:
Anaphylaxis during anesthesia in Norway: A 6-year singlecenter follow-up study. Anesthesiology 2005; 102:897903
6. Lobera T, Audicana MT, Pozo MD, Blasco A, Fernndez E,
Caada P, Gastaminza G, Martinez-Albelda I, GonzlezMahave I, Muoz D: Study of hypersensitivity reactions and
anaphylaxis during anesthesia in Spain. J Investig Allergol
Clin Immunol 2008; 18:3506
7. Sadleir PH, Clarke RC, Bunning DL, Platt PR: Anaphylaxis
to neuromuscular blocking drugs: Incidence and cross-reactivity in Western Australia from 2002 to 2011. Br J Anaesth
2013; 110:9817
8. Gurrieri C, Weingarten TN, Martin DP, Babovic N, Narr BJ,
Sprung J, Volcheck GW: Allergic reactions during anesthesia
at a large United States referral center. Anesth Analg 2011;
113:120212
9. Reitter M, Petitpain N, Latarche C, Cottin J, Massy N,
Demoly P, Gillet P, Mertes PM; French Network of Regional
Pharmacovigilance Centres: Fatal anaphylaxis with neuromuscular blocking agents: A risk factor and management
analysis. Allergy 2014; 69:9549
10. Laake JH, Rttingen JA: Rocuronium and anaphylaxisA statistical challenge. Acta Anaesthesiol Scand 2001; 45:1196203
11. Krigaard M, Garvey LH, Menn T, Husum B: Allergic reactions in anaesthesia: Are suspected causes confirmed on subsequent testing? Br J Anaesth 2005; 95:46871
12. Mertes PM, Malinovsky JM, Jouffroy L, Aberer W, Terreehorst
I, Brockow K, Demoly P; Working Group of the SFAR and
SFA; ENDA; EAACI Interest Group on Drug Allergy: Reducing
the risk of anaphylaxis during anesthesia: 2011 updated
guidelines for clinical practice. J Investig Allergol Clin
Immunol 2011; 21:44253
6
P. M. Mertes and G. W. Volcheck
EDITORIAL VIEWS
19. Clarke RC, Sadleir PH, Platt PR: The role of sugammadex in
the development and modification of an allergic response to
rocuronium: Evidence from a cutaneous model. Anaesthesia
2012; 67:26673
20. Leysen J, Bridts CH, De Clerck LS, Ebo DG: Rocuroniuminduced anaphylaxis is probably not mitigated by sugammadex: Evidence from an in vitro experiment. Anaesthesia 2011;
66:5267
21. Tsur A, Kalansky A: Hypersensitivity associated with sugammadex administration: A systematic review. Anaesthesia
2014; 69:12517
22. Mertes PM, Aimone-Gastin I, Guant-Rodriguez RM,
Mouton-Faivre C, Audibert G, OBrien J, Frendt D, Brezeanu
M, Bouaziz H, Guant JL: Hypersensitivity reactions to
neuromuscular blocking agents. Curr Pharm Des 2008;
14:280925
23. Malinovsky JM, Plaud B, Debaene B, Mertes PM: [Do we
know all indications and side effects of sugammadex?]. Ann
Fr Anesth Reanim 2011; 30:70910
13. Mertes PM, Demoly P, Malinovsky JM: Hypersensitivity reactions in the anesthesia setting/allergic reactions to anesthetics. Curr Opin Allergy Clin Immunol 2012; 12:3618
14. Volcheck GW, Mertes PM: Local and general anesthetics

immediate hypersensitivity reactions. Immunol Allergy Clin
North Am 2014; 34:52546, viii
15. Malinovsky JM, Decagny S, Wessel F, Guilloux L, Mertes

PM: Systematic follow-up increases incidence of anaphylaxis during adverse reactions in anesthetized patients. Acta
Anaesthesiol Scand 2008; 52:17581
16. Perry JJ, Lee JS, Sillberg VA, Wells GA: Rocuronium versus
succinylcholine for rapid sequence induction intubation.
Cochrane Database Syst Rev 2008: CD002788
17. Aceto P, Perilli V, Modesti C, Ciocchetti P, Vitale F, Sollazzi L:
Airway management in obese patients. Surg Obes Relat Dis
2013; 9:80915
18. McDonnell NJ, Pavy TJ, Green LK, Platt PR: Sugammadex in
the management of rocuronium-induced anaphylaxis. Br J
Anaesth 2011; 106:199201
ANESTHESIOLOGY REFLECTIONS FROM THE WOOD LIBRARY-MUSEUM
Just before World War I, the company of Chicagos Samuel Katz peddled his Oxygen Treatment for
Catarrh as an oxygenating panacea. He advertised that his cure-all contained as much Oxygen as 86
times its weight in food and drink (left). Katz reminded his readers that if they placed any living thing in a
vacuum, without oxygen it will die (right). In 1917 another Chicago-based organization, the American
Medical Association (AMA) published analyses of Katz Oxygen Treatment revealing it to consist of four
discrete boxes, consisting chiefly of (1) aloes, (2) magnesium dioxide, magnesium carbonate and
calcium salts, with acacia, (3) sodium perborate and tartaric acid, and (4) cotton soaked in menthol.
So ironically, Chicago provided a home to promoters (Katz and Company) and discreditors (the AMA) of
the Katz Oxygen Treatment. (Copyright the American Society of Anesthesiologists, Inc.)
George S. Bause, M.D., M.P.H., Honorary Curator, ASAs Wood Library-Museum of Anesthesiology,
Schaumburg, Illinois, and Clinical Associate Professor, Case Western Reserve University, Cleveland,
Ohio. UJYC@aol.com
EDITORIAL VIEWS
19. Clarke RC, Sadleir PH, Platt PR: The role of sugammadex in
the development and modification of an allergic response to
rocuronium: Evidence from a cutaneous model. Anaesthesia
2012; 67:26673
20. Leysen J, Bridts CH, De Clerck LS, Ebo DG: Rocuroniuminduced anaphylaxis is probably not mitigated by sugammadex: Evidence from an in vitro experiment. Anaesthesia 2011;
66:5267
21. Tsur A, Kalansky A: Hypersensitivity associated with sugammadex administration: A systematic review. Anaesthesia
2014; 69:12517
22. Mertes PM, Aimone-Gastin I, Guant-Rodriguez RM,
Mouton-Faivre C, Audibert G, OBrien J, Frendt D, Brezeanu
M, Bouaziz H, Guant JL: Hypersensitivity reactions to
neuromuscular blocking agents. Curr Pharm Des 2008;
14:280925
23. Malinovsky JM, Plaud B, Debaene B, Mertes PM: [Do we
know all indications and side effects of sugammadex?]. Ann
Fr Anesth Reanim 2011; 30:70910
13. Mertes PM, Demoly P, Malinovsky JM: Hypersensitivity reactions in the anesthesia setting/allergic reactions to anesthetics. Curr Opin Allergy Clin Immunol 2012; 12:3618
14. Volcheck GW, Mertes PM: Local and general anesthetics

immediate hypersensitivity reactions. Immunol Allergy Clin
North Am 2014; 34:52546, viii
15. Malinovsky JM, Decagny S, Wessel F, Guilloux L, Mertes

PM: Systematic follow-up increases incidence of anaphylaxis during adverse reactions in anesthetized patients. Acta
17. Aceto P, Perilli V, Modesti C, Ciocchetti P, Vitale F, Sollazzi L:
Airway management in obese patients. Surg Obes Relat Dis
2013; 9:80915
18. McDonnell NJ, Pavy TJ, Green LK, Platt PR: Sugammadex in
the management of rocuronium-induced anaphylaxis. Br J
Anaesth 2011; 106:199201
Just before World War I, the company of Chicagos Samuel Katz peddled his Oxygen Treatment for
Catarrh as an oxygenating panacea. He advertised that his cure-all contained as much Oxygen as 86
times its weight in food and drink (left). Katz reminded his readers that if they placed any living thing in a
vacuum, without oxygen it will die (right). In 1917 another Chicago-based organization, the American
Medical Association (AMA) published analyses of Katz Oxygen Treatment revealing it to consist of four
discrete boxes, consisting chiefly of (1) aloes, (2) magnesium dioxide, magnesium carbonate and
calcium salts, with acacia, (3) sodium perborate and tartaric acid, and (4) cotton soaked in menthol.
So ironically, Chicago provided a home to promoters (Katz and Company) and discreditors (the AMA) of
the Katz Oxygen Treatment. (Copyright the American Society of Anesthesiologists, Inc.)
Ohio. UJYC@aol.com
Big Brain, Small World?

Erik Olofsen, M.Sc., Albert Dahan, M.D., Ph.D.
HE electroencephalogram
is the scalp recording of the
electrical activity of the brain and
reflects the activity of millions of
neurons on and just below the surface of the brain.
The
electroencephalogram
is used extensively as surrogate
marker of anesthetic effect. The
rhythmic pattern of the electroencephalogram is complex and while
some patterns are easily recognized
by the eye, such as sleep spindles
during stage 2 sleep, sophisticated analyses techniques have
been developed to disentangle the
complexities of electroencephalographic behavior. In this issue of
Anesthesiology, Khodayari-Rostamabad et al.1 present data on the
effect of the opioid remifentanil on
the electroencephalogram. They
first performed a functional connectivity analysis to establish the
characteristics of the cortical network and subsequently analyzed
the network with a technique
based on graph theory. Functional
connectivity represents statistical
covariation between brain regions
(i.e., brain regions with similar
rhythmic activity), however, without any implications regarding causality or mechanism. To
construct the network, Khodayari-Rostamabad et al.1 calculated the coherence matrix of the electroencephalogram
in the standard bands, assuming that the electrodes serve as
nodes to the network. Apart from the electroencephalogram,
other techniques, such as functional magnetic imaging, may
be used to define functional brain connectivity.2
in the United States. In one of his

studies, the Nebraska study, the
number of connections ranged
from 2 to 10 with a median of
6, very close to an earlier prediction by Frigyes Karinthy in 1929,5
who predicted that a network of
just five acquaintances of 1.5 billion people suffices to connect two
inhabitants of our planet. We are
all aware of the complex networks
in our lives. Another intuitive
example of a small-world network
is the existing network of airports.
To get around the world, we fly to
a highly connected hub airport
that creates a shortcut throughout
the network and allows us to travel
most efficiently to our destination.
A major step forward to further understanding small-world
networks was made by Watts and
Strogatz.6 They explored networks
that varied in complexity from
a regular network to a network
with an increased amount of disorder to a randomly connected
network (fig.1).6 The networks
are described in terms of nodes
(vertices) and connecting paths
(edges) with characteristic path
length L (fig.2). Three connected
vertices (e.g., positioned in a triangle) make up a cluster with
clustering coefficient C (note that clusters may be defined
in alternative ways). Clustering coefficient C is defined as
the average across vertices of the ratio of existing triangles
and possible triangles; the characteristic path length L as the
average shortest distance between vertices (one step between
vertices equals a path length of 1).6,7 As recently explained,
for brain networks, the characteristic path length represents
the global efficiency in information transmission in the brain
network; the clustering coefficient represents a measure of
local functional segregation of brain activities.8,9 The highly
regular network of Watts and Strogatz6 has high values for
C and L, indicative of a network in which the vertices are
highly connected but the flow of information between vertices is slow. The total random network of Watts and Strogatz6
has low values for C and L, which indicates that the vertices
The graph theoretical description of these networks

has various advantages as
it improves insight and understanding of these systems that seem unimaginably complex at first sight
but do harbor an underlying systematic.
Graph Theory
Graph theory can be described as the mathematical analysis of networks in which the graph represents a set of network nodes.3 A well-known example of a graph theoretical
problem was published by Stanley Milgram in 19674 in his
famous paper entitled The small-world problem. He studied the number of intermediate sequential acquaintances
(nodes) that are required to connect two random individuals
Image: Thinkstock.
Accepted for publication August 29, 2014. From the Department of Anesthesiology, Leiden University Medical Center, Leiden, The Netherlands.
January 2015
EDITORIAL VIEWS
The network depicted in figure3G represents a typical

small-world experience. Persons 3 and 4 initially did not
know each other, but find out that they share a common
friend (no. 7 in the network). Interestingly, as they start talking, a connection between them is being established, giving
rise to a change in small-worldness. The changes in C, L, and
S due to the newly formed connection are given in the figure.
Small-worldness of the Brain
Fig. 1. The networks described by Watts and Strogatz.6 The

networks vary in complexity from a regular network (C=0.5,
L=2.89, S = 0.173) to a small-world network with an increased
amount of disorder (C = 0.453, L = 2.58, S=0.175) to a randomly connected network (C = 0.343, L = 2.35, S=0.146).
P is the probability of reconnection of edges (see ref. 6 for
further explanation).
Khodayari-Rostamabad et al.1 show that remifentanil at an

approximate effect-site concentration of 2ng/ml had marked
effects on the graph theoretical measures of the electroencephalogram, especially in the 1 and to a lesser extent in
the bands. Remifentanil increased the characteristic path
length L and decreased the clustering coefficient C and relative small-worldness. As stated, these effects are suggestive
of a disruption of the complex cortical network subserving
normal brain function. In terms of our simple examples,
network 3C reverts to network 3B. As correctly mentioned
by the authors, it is important to realize that while the
observed opioid-induced changes suggest a change of the
balance between integration and segregation of the brain
networks that make up the electroencephalogram, they give
us no indication of the underlying neurobiological, mechanistic, or functional changes within the brain.
The results of Khodayari-Rostamabad et al.1 contrast
earlier reports on the observation that anesthesia-induced
unconsciousness is not associated with changes in brain
small-worldness.1113 In animal and human studies, induction, maintenance, and recovery from isoflurane or propofol anesthesia had no effect on spatial organization of brain
networks, as defined by resting-state functional magnetic
imaging or multichannel electroencephalography. In this
light, the findings of Khodayari-Rostamabad et al.1 seem
surprising. The authors suggest that their findings are opioid-specific, a remarkable postulation, given the fact that
the observed changes in small-worldness were associated
with changes in vigilance and attention rather than with
antinociception (a surrogate marker of analgesia). Evidently,
the findings of Khodayari-Rostamabad et al.1 deserve confirmation in future studies.
Networks are present in a multitude of systems in our daily
lives. The graph theoretical description of these networks has
various advantages as it improves insight and understanding of these systems that seem unimaginably complex at first
sight but do harbor an underlying systematic. Currently,
studies focus on possibly one of the most complex networks,
the brain functional network. Khodayari-Rostamabad et al.1
tested the effect of the opioid remifentanil on cortical smallworldness and showed a disruptive effect. The presented outcome is difficult to interpret in terms of clinical utility. We
look forward to further (confirmatory) studies that increase
our insight in the behavior of small-work networks during
the administration of centrally acting drugs and determine
applicability in clinical practice.
Fig. 2. Graph illustrating a network with multiple nodes (vertices, numbered from 1 to 9) and connections (edges) between
nodes. There are two clusters present (i.e., triangle networks).
One possible path is given in red with path length 4 (1-3-7-6-4).
are locally segregated but information between vertices travels fast. Intermediate, there is a network with high C and
low L, which Watts and Strogatz6 describe as small-world
(fig.1). In their article, Khodayari-Rostamabad et al.1 assess
the effect of remifentanil by quantitation of the ratio S,
where S = C/L, which is the relative small-worldness and in
fact, a measure of the balance between global integration (L)
and local segregation (C).7
To explain the calculation of C, L, and S, we give a series
of examples in figure3 in which the vertices or nodes represent people. People who are connected know each other,
similar to experiments of Milgram.4 The constructed networks represent the flow of information, in these cases as
simple as a person-to-person conversation. See the legend
of figure3 for an explanation of the calculations of C, L,
and S. Evidently, for more complex networks, the analysis
requires potent computing power. Sometimes alternative
measures of small-worldness are calculated, for example,
by normalizing C and L by values obtained from a random
graph.10 Khodayari-Rostamabad et al.1 use a weighted graph,
in which edges are not present or absent, but their existence
is a weight between 0 and 1. The weights are given by the
correlation between electrodes in the usual frequency bands.
E. Olofsen and A. Dahan
Editorial Views
Fig. 3. Seven example graphs with their graph theoretical properties based on ref. 7. C is the clustering coefficient (the average
across vertices of the ratio of existing triangles and possible triangles), L is the characteristic path length (the average shortest
distance between vertices; each step between vertices equals 1), and S is the relative small-worldness where S = C/L. (A) Persons 1 and 2 know each other and are therefore connected by the arrowed line. In this, simple 1 on 1 network C is undefined, L
is 1 because the average shortest path length = (12 + 21)/2 = (1+1)/2 = 1, and relative small-worldness, C/L, is undefined as
well. (B) Individual 3 has two friends that do not know each other: C = 0 (since no triangle network is present), L 1.33 (the average path length = (31 + 13 + 21 + 12 + 132 + 321)/6 = (1+1+1+1+2+2)/6 1.33), and S = 0/1.33 = 0. (C) This
is a triangle network. C = 1 since for each of the three vertices (people) there is just one possible triangle connection. Hence, C
= (1/1 + 1/1 +1/1)/3 = 1. The average path length equals (12 + 21 + 13 + 31 + 23 + 32)/6 = 6/6 = 1. Relative smallworldness = 1. (D) The square network the relative small-worldness equals zero as C = 0 as for none of the people a triangle is
present, although for each of the people two triangles connections would be possible (e.g., 1-2-3-1). Hence, C = (0/2 + 0/2 + 0/2
+0/2)/4 = 0. (E) We need to calculate C by counting the number of possible and existing triangles in the network for each vertex
(person). For persons 1 and 2, there is one triangle connection of one possible; the network of person 3 has one existing triangle
connection (3-2-1-3), but there are two additional possible triangles (3-4-1-3 and 3-4-2-3); in other words, for person 3, there is
one triangle of three possible ones; finally for person 4, there are no triangles of a two possible; the average C = (1/1 + 1/1 + 1/3
+ 0/2)/4 0.58. The average shortest path length, calculated as presented before, equals 1.33, so that S 0.44. (F and G) The
calculations for these more complex networks follow from the earlier examples. In G, the broken blue arrow is an emerging connection between persons 3 and 4. Values for C, L, and S differ for the state in which 3 and 4 are not connected compared to the
state in which they are. Parameter values are given for both the unconnected state (S 0.34) and the connected state (S 0.39).
Competing Interests
Correspondence
Address correspondence to Dr. Dahan: a.dahan@lumc.nl
References
1. Khodayari-Rostamabad A, Olesen SS, Graversen C, Malver LP,
Kurita GP, Sjgren P, Christrup LL, Drewes AM: Disruption
of cortical connectivity during remifentanil administration is
Anesthesiology 2015; 122:8-11 10
associated with cognitive impairment but not with analgesia.

2. Niesters M, Khalili-Mahani N, Martini C, Aarts L, van Gerven J,
van Buchem MA, Dahan A, Rombouts S: Effect of subanesthetic
ketamine on intrinsic functional brain connectivity: A placebocontrolled functional magnetic resonance imaging study in
healthy male volunteers. Anesthesiology 2012; 117:86877
3. Stam CJ, Reijneveld JC: Graph theoretical analysis of complex
networks in the brain. Nonlinear Biomed Phys 2007; 1:3
4. Milgram S: The small-world problem. Psychol Today 1967; 1:
617
5. Karinthy F: Lncszemek [Chains]. Published in Minden
maskppen van [Everything Is Different]. Budapest,
Atheneum Irodai es Nyomdai R. - T. Kiadasa, 1929 (translation available at: http://electriceye.org.uk/sites/default/files/
Chains.pdf. Accessed October 30, 2014)
EDITORIAL VIEWS
6. Watts DJ, Strogatz SH: Collective dynamics of small-world

networks. Nature 1998; 393:4402
7. Rubinov M, Sporns O: Complex network measures of brain
connectivity: Uses and interpretations. Neuroimage 2010;
52:105969
8. Lee H, Mashour GA, Noh GJ, Kim S, Lee U: Reconfiguration
of network hub structure after propofol-induced unconsciousness. Anesthesiology 2013; 119:134759

9.
Mashour GA, Avidan MS: Postoperative delirium:
Disconnecting the network? Anesthesiology 2014; 121:2146
10. Humphries MD, Gurney K: Network small-world-ness: A
qualitative method for determining canonical network equivalence. PloS One 2008; 3: 110
11. Lee U, Oh G, Kim S, Noh G, Choi B, Mashour GA: Brain

networks maintain a scale-free organization across consciousness, anesthesia, and recovery: Evidence for
adaptive reconfiguration. Anesthesiology 2010; 113:
108191
12. Schrter MS, Spoormaker VI, Schorer A, Wohlschlger
A, Czisch M, Kochs EF, Zimmer C, Hemmer B, Schneider
G, Jordan D, Ilg R: Spatiotemporal reconfiguration of
large-scale brain functional networks during propofolinduced loss of consciousness. J Neurosci 2012; 32:
1283240
13. Liang Z, King J, Zhang N: Intrinsic organization of the anesthetized brain. J Neurosci 2012; 32:1018391
A registered dentist, Dr. Henry Floyd Huffaker (18701946) shifted his practice location around
Knoxville, Tennessee, at least twice before settling by 1914 on the fancy McTownlee Building for
professionals. To attract dental patients, Dr. Huffaker made a point of advertising that he supplied
laughing gas. The attractive lady on his advertising postcard was photographed, not there in
Knoxville, but apparently over 700 miles to the northeast, in Rochester, New York. This postcard is
part of the Wood Library-Museums Ben Z. Swanson Collection. (Copyright the American Society
of Anesthesiologists, Inc.)
Ohio. UJYC@aol.com.
EDITORIAL VIEWS

52:105969

9.
Mashour GA, Avidan MS: Postoperative delirium:
Disconnecting the network? Anesthesiology 2014; 121:2146
qualitative method for determining canonical network equivalence. PloS One 2008; 3: 110

networks maintain a scale-free organization across consciousness, anesthesia, and recovery: Evidence for
adaptive reconfiguration. Anesthesiology 2010; 113:
108191
12. Schrter MS, Spoormaker VI, Schorer A, Wohlschlger
A, Czisch M, Kochs EF, Zimmer C, Hemmer B, Schneider
G, Jordan D, Ilg R: Spatiotemporal reconfiguration of
large-scale brain functional networks during propofolinduced loss of consciousness. J Neurosci 2012; 32:
1283240
A registered dentist, Dr. Henry Floyd Huffaker (18701946) shifted his practice location around
Knoxville, Tennessee, at least twice before settling by 1914 on the fancy McTownlee Building for
professionals. To attract dental patients, Dr. Huffaker made a point of advertising that he supplied
laughing gas. The attractive lady on his advertising postcard was photographed, not there in
Knoxville, but apparently over 700 miles to the northeast, in Rochester, New York. This postcard is
part of the Wood Library-Museums Ben Z. Swanson Collection. (Copyright the American Society
of Anesthesiologists, Inc.)
Ohio. UJYC@aol.com.
Characterizing the Epidemiology of Postoperative

Transfusion-related Acute Lung Injury
Leanne Clifford, B.M., Qing Jia, M.D., Arun Subramanian, M.B.B.S., Hemang Yadav, M.B.B.S.,
Gregory A. Wilson, R.R.T., Sean P. Murphy, B.S., Jyotishman Pathak, Ph.D.,
Darrell R. Schroeder, M.S., Daryl J. Kor, M.D.
ABSTRACT
Background: Transfusion-related acute lung injury (TRALI) is the leading cause of transfusion-related death in the United
States; however, it remains poorly characterized in surgical populations. To better inform perioperative transfusion practice,
and to help mitigate perioperative TRALI, the authors aimed to better define its epidemiology before and after TRALI mitigation strategies were introduced.
Methods: This retrospective cohort study examined outcomes of adult patients undergoing noncardiac surgery with general
anesthesia who received intraoperative transfusions during 2004 (n = 1,817) and 2011 (n = 1,562). The demographics and
clinical characteristics of transfusion recipients, blood transfusion descriptors, and combined TRALI/possible TRALI incidence rates were evaluated. Univariate analyses were used to compare associations between patient characteristics, transfusion
details, and TRALI mitigation strategies with TRALI/possible TRALI incidence rates in a before-and-after study design.
Results: The incidence of TRALI/possible TRALI was 1.3% (23 of 1,613) in 2004 versus 1.4% (22 of 1,562) in 2011
(P = 0.72), with comparable overall rates in males versus females (1.4% [23 of 1,613] vs. 1.2% [22 of 1,766]) (P = 0.65).
Overall, thoracic (3.0% [4 of 133]), vascular (2.7% [10 of 375]), and transplant surgeries (2.2% [4 of 178]) carried the highest
rates of TRALI/possible TRALI. Obstetric and gynecologic surgical patients had no TRALI episodes. TRALI/possible TRALI
incidence increased with larger volumes of blood product transfused (P < 0.001).
Conclusions: Perioperative TRALI/possible TRALI is more common than previously reported and its risk increases with
greater volumes of blood component therapies. No significant reduction in the combined incidence of TRALI/possible TRALI
occurred between 2004 and 2011, despite the introduction of TRALI mitigation strategies. Future efforts to identify specific
risk factors for TRALI/possible TRALI in surgical populations may reduce the burden of this life-threatening complication.
(Anesthesiology 2015; 122:12-20)
OR the past decade, transfusion-related acute lung injury

(TRALI) has consistently been the leading cause of transfusion-related fatalities reported to the U.S. Food and Drug
Administration.1 Indeed, recent estimates note that TRALI was
implicated in 38% of all fatalities.1 Perhaps more concerning
is the fact that passive reporting has repeatedly been shown to
significantly underestimate the true burden of this syndrome.2,3
For example, Kopko et al.2 noted that only 13.3% of cases with
symptoms suggestive of TRALI were reported to the blood
bank by the clinical service as a possible transfusion reaction.
These findings were replicated in a recent study carried out by
this investigative group, in which just 14.1% of cases had been
reported by the responsible clinical team.3 In addition to the
potential impact on the delivery of appropriate care, this lack
of recognition and reporting of TRALI may result in the failure to prevent subsequent reactions to blood products of implicated donors who were not identified after the initial TRALI
What We Already Know about This Topic

Transfusion-related acute lung injury is the leading cause of
transfusion-related death in the United States
What This Article Tells Us That Is New

A retrospective cohort analysis from one institution documented that perioperative transfusion-related acute lung injury occurs approximately 1.4 to 3% in surgical patients, with
higher rates in patients who received larger volumes of blood
component therapies
episode. Moreover, the poorly defined epidemiology of TRALI

has undoubtedly contributed to the lack of studies testing
potential preventive or treatment strategies, as well as to uncertainty regarding the attributable burden of this syndrome on
patient-important outcomes and healthcare resource utilization.
This article is featured in This Month in Anesthesiology, page 1A. Corresponding article on page 1. Presented at the American Society
of Anesthesiology Annual Congress, San Francisco, California, October 13, 2013.
Submitted for publication November 13, 2013. Accepted for publication May 8, 2014. From the Department of Internal Medicine (L.C.)
Department of Anesthesiology (Q.J., A.S., G.A.W., D.J.K.), Division of Pulmonary and Critical Care Medicine (H.Y.), Department of Information Technology (S.P.M.), and Division of Biomedical Statistics and Informatics ( J.P., D.R.S.), Mayo Clinic, Rochester, Minnesota.
Anesthesiology, V 122 No 1 12
January 2015
Previous estimates of the incidence of TRALI have ranged

from 0.04 to 8.0% per transfused patient.48 However, many
of these studies used passive reporting strategies, and thus are
believed to dramatically underestimate the true incidence of
TRALI. Additionally, most studies investigating the epidemiology of TRALI have focused on patients in the critical care
setting. Although these investigations have provided valuable
information about TRALI in critically ill patients, substantial
differences frequently exist in patient characteristics and baseline risk factors for lung injury when the noncardiac surgical population is compared with the critically ill population.
Therefore, the available data on the critically ill population
may not be generalizable to surgical cohorts. Furthermore,
the paucity of studies focusing on surgical patients has precluded the identification of potentially modifiable perioperative risk factors specific to this unique population.
Importantly, approximately 50% of all blood product
transfusions take place in the perioperative environment.9
Consequently, anesthesiologists are well positioned to make
meaningful contributions to improving transfusion-related
outcomes for surgical patients. To this end, a critical first
step is to establish a clear understanding of the epidemiology
of transfusion-related complications. Therefore, the overall
goal of this investigation was to clearly define the epidemiology of TRALI after intraoperative blood product administration in a large cohort of patients undergoing noncardiac
surgery with general anesthesia. Of note, previous reports
have shown a decline in the incidence of TRALI after the
introduction of universal leukoreduction10 and the maleonly donor policy for plasma11 (occurring in 2005 and 2008,
respectively, at our institution). Therefore, in order to evaluate the impact of these interventions in surgical patients, we
conducted a before-and-after study, evaluating the outcomes
of all patients transfused intraoperatively during the calendar
years 2004 and 2011.
Materials and Methods

After obtaining Mayo Clinic Institutional Review Board
approval, we conducted a retrospective cohort study evaluating the incidence of TRALI/possible TRALI in all transfused patients undergoing noncardiac surgery with general
anesthesia during the calendar years 2004 and 2011. All
patients had given previous consent for the use of their medical records in research. The STROBE (Strengthening the
Reporting of Observational Studies in Epidemiology) guidelines were followed in the conduct of this study and in the
reporting of its results.12
Study Population
All noncardiac surgical patients receiving general anesthesia
were identified from an institutional database (the perioperative data mart).13 This database captures high-resolution,
near real-time data from all monitored care environments
within our institution. Specifically, details relating to vital
signs, laboratory tests, electrocardiogram monitoring, ventilator settings, fluid therapies, radiology results, medications,
transfusion therapies, and procedures are stored within an
open-access connectivity database accessed via JMP statistical software (SAS Institute Inc., Cary, NC). All noncardiac
surgical patients receiving general anesthesia and allogeneic
blood products in the operating room during the calendar
years 2004 and 2011 were eligible for inclusion. Criteria for
exclusion were (1) absence of research authorization, (2) age
younger than 18 yr, (3) inclusion in study already, (4) preoperative respiratory failure, (5) preoperative diffuse bilateral infiltrates on chest radiograph, (6) intraoperative death,
(7) postoperative requirement for extracorporeal membrane
oxygenation, and (8) receipt of only autologous blood products (fig.1). Patients who received only autologous blood
products were not considered at risk for TRALI and were
therefore not included in this investigation.
Fig. 1. Patient flowchart. ECMO = extracorporeal membrane oxygenation; TACO = transfusion-associated circulatory overload;
TRALI = transfusion-related acute lung injury.
Anesthesiology 2015; 122:12-20 13 Clifford et al.
Outcome Adjudication
In the current study, cases adjudicated as TRALI or as possible TRALI according to the 2004 Canadian consensus
criteria14 (table1) were considered as a single composite outcome. The rationale for combining these two outcomes is
that many high-risk surgeries (e.g., aortic vascular, lung/
esophageal resection) are known to be associated with a high
rate of postoperative lung injury, thereby necessitating the
adjudication of possible TRALI.15,16 Excluding this group
from our analyses would therefore have resulted in the exclusion of a large number of potential cases with important deleterious outcomes. Thus, for purposes of this study, we refer
to this composite outcome as TRALI.
To identify cases of TRALI in this population, we used
a three-step approach. First, to enhance the feasibility of
this large study, and to facilitate the accurate detection of
TRALI cases, we utilized our recently developed natural
language processing (NLP)based electronic screening algorithm (fig.2).17 This algorithm was developed in an alternate
cohort of patients that included subjects known to have had
a TRALI episode, as well as in complication-free transfused
controls. The algorithm variables and their cutoff thresholds
were developed using classification and regression tree analysis, as previously described.17 In brief, this algorithm evaluates chest radiograph reports for the presence of terms and
phrases believed to be consistent with a diagnosis of TRALI
(table2). In total, there were 4 explicit terms, 8 descriptive
terms, 16 location terms, and 128 term combinations that
were considered for NLP chest radiograph case identification using the open-source clinical text analysis and knowledge extraction system.18 In addition, vital status data stored
within the electronic health record (including partial pressure
of arterial oxygen [Pao2], Pao2 to fraction of inspired oxygen
[Fio2] ratio [Pao2:Fio2], and oxygen saturation [Spo2]) were
evaluated to identify potential TRALI cases. We have previously documented the diagnostic accuracy of this TRALI
case detection algorithm with a sensitivity and specificity of
92.5% (95% CI, 84.6 to 96.7%) and 93.6% (95% CI, 85.0
to 97.6%), respectively.17
Second, two independent physicians (L.C. and Q.J.)
manually reviewed the electronic health records of all patients
who screened positive for TRALI and allocated a diagnosis
based upon the 2004 Consensus Criteria (table1).14 Due to
our interest in the epidemiology of TRALI after intraoperative blood product administration, we limited our outcome
assessment to 6h after the last intraoperative blood product
transfusion. Third, in situations in which these two physicians disagreed, a panel of three senior critical care physicians reviewed each case to adjudicate a final outcome.
* CDC. The National Healthcare Safety Network (NHSN) Manual:
Biovigilance Component Protocol [Internet]. Atlanta, GA: Division
of Healthcare Quality Promotion, National Center for Emerging
and Zoonotic Infectious Diseases, Centers for Disease Control and
Prevention; January 2014 [cited April 15, 2014]. Available at: http://
www.cdc.gov/nhsn/PDFs/Biovigilance/BV-HV-protocol-current.pdf.
Accessed November 13, 2014.
Table 1. TRALI and Possible TRALI Definitions from the 2004

Canadian Consensus Statement
1. TRALI
a) ALI
i. Acute onset
ii. Hypoxemia (Pao2:Fio2 300 mmHg or Spo2 <90% on room
air [or other clinical evidence of hypoxemia])
iii. B
ilateral infiltrates on frontal chest radiograph
iv. No evidence of left atrial hypertension (i.e., circulatory
overload) as the sole explanation for the critical findings*
b) No preexisting ALI before transfusion
c) Onset during or within 6h of transfusion
d) No temporal relationship to an alternative risk factor for ALI
2. Possible TRALI
a) ALI
b) No preexisting ALI before transfusion
c) During or within 6h of transfusion
d) A clear temporal relationship with an alternative risk factor
for ALI
Adapted, with permission, from Kleinman et al. Transfusion 2004; 44:
177489.14 Adaptations are themselves works protected by copyright. So in
order to publish this adaptation, authorization must be obtained both from the
owner of the copyright in the original work and from the owner of copyright in
the translation or adaptation.
* Data evaluated to inform decisions regarding the presence or absence of
left atrial hypertension/fluid overload (when available) included: fluid balance, patient weight, echocardiographic data, pulmonary artery catheter
hemodynamic data, central venous pressures, and the clinical context
(comorbidities and acute physiology).
ALI = acute lung injury; Fio2 = fraction of inspired oxygen; Pao2 = partial
pressure of arterial oxygen; Spo2 = oxygen saturation; TRALI = transfusion-related acute lung injury.
Fig. 2. Screening algorithm. Natural language processing

(NLP)based transfusion-related acute lung injury (TRALI)
screening algorithm. Asterisk indicates that, in the absence
of a Pao2:Fio2 ratio, surrogate markers of hypoxemia used in
order of availability were Pao2 117mmHg followed by Spo2
97%. CXR = chest radiograph; plus sign (+) = positive.
Physicians were able to adjudicate a diagnosis of TRALI

using the 2004 Consensus Criteria, as described in table1;
of transfusion-associated circulatory overload (TACO)
based on the 2014 National Healthcare Safety Network criteria* (table3); of both TRALI and TACO (if there was evidence for both TRALI and TACO but neither was thought
to fully explain the clinical picture); or of neither TRALI
nor TACO (where neither the TRALI nor the TACO case
definitions were met). Since the primary aim of the current investigation was to better describe the incidence of
TRALI in a surgical population, data from patients with a
diagnosis of only TACO were not analyzed as part of this
investigation (fig.1).
Table 2. Variation in Descriptive Terms Corresponding to a

Diagnosis of TRALI Included in Natural Language Processing
Location Terms
Alveolar
Perihilar
Bilateral
Bilaterally
Basilar
Bibasilar
Bibasilar
Diffuse
Extensive
Airspace
Air space
Air-space
Parenchymal
Pulmonary
Lung
Lungs
Description Terms
Explicit Terms
Opacification
Opacity
Opacities
Infiltrate
Infiltrates
Hazy
Fluffy
Patchy
Edema
ALI
ARDS
CHF
ALI = acute lung injury; ARDS = acute respiratory distress syndrome;

CHF = congestive heart failure; TRALI = transfusion-related acute lung injury.
Table 3. TACO from the CDC National Healthcare Safety

Network Biovigilance Component 2013
New onset or exacerbation of three or more of the following
within 6h of cessation of transfusion:
Acute respiratory distress (dyspnea, orthopnea, cough)
Increased BNP
Increased CVP
Evidence of left heart failure
Evidence of positive fluid balance
Radiographic evidence of pulmonary edema
Adapted from the U.S. Centers for Disease Control, The National Healthcare Safety Network (NHSN) Manual: Biovigilance Component Protocol
[Internet]. Atlanta (GA): Division of Healthcare Quality Promotion, National
Center for Emerging and Zoonotic Infectious Diseases, Centers for Disease
Control and Prevention; January 2014 [cited 2014 Apr 15]. http://www.cdc.
gov/nhsn/PDFs/Biovigilance/BV-HV-protocol-current.pdf.
BNP = brain type natriuretic peptide; CDC = Centers for Disease Control;
CVP = central venous pressure; TACO = transfusion-associated circulatory
overload.
Data Sources and Collection

Data required to apply our electronic algorithm fit into two main
categories: 1) structured data (e.g., vital signs) and 2) unstructured data (e.g., radiology reports processed using NLP). The
required structured data points included Pao2; Pao2:Fio2 ratio;
and Spo2 (fig.2). To ensure completeness, we collected these
data from three existing databases: the perioperative data mart
(as described above in the Study Population subsection of the
Materials and Methods section); the intensive care unit (ICU)
data mart, a near real-time database similar to the perioperative
data mart that captures multiple data points directly from the
electronic health record for patients in the ICU13; and the Mayo
Clinic Life Sciences System,19 a sophisticated data warehouse
that contains a normalized replica of the electronic health record
derived from multiple original clinical data sources, including
vital signs for all patients. For the current study, the Mayo
Clinic Life Sciences System was used to obtain structured data
for patients located in nonmonitored care environments such as

the general surgical floor. Baseline demographic data were also
extracted from these three databases.
Extensive efforts have been undertaken to ensure the validity and reliability of these databases, and these efforts have been
described in detail previously,13,19 To maximize the sensitivity
of our algorithm, we identified the worst or the most extreme
values for each patient within the time period of interest after
blood product transfusion. For example, we used the lowest
Pao2:Fio2 ratio. Errors were minimized by manually verifying the most extreme values in the electronic health record.
Unstructured data used in this algorithm (e.g., chest radiograph reports) were obtained from the enterprise data trust,
an existing database of clinical, research, and administrative
data.19 NLP was applied to the reports to identify terms consistent with a diagnosis of TRALI as described previously.17
Detailed transfusion data were collected using the perioperative information tool. This user interface application was developed using Microsoft.NET (Microsoft Corp., Redmond, WA)
technology and contains detailed information pertaining to all
transfusions, including exact transfusion times, blood products,
and volumes, and it allows users to retrieve multiple perioperative patient data points in an efficient manner. Data sources for
this application include direct transmission from the perioperative data mart, as well as from the Amalga Unified Intelligence System (Microsoft Corp.).20 This latter resource captures
a broader range of near real-time patient data. The automated
nature of these complex databases ensures accuracy of data to
the same extent that the original medical record is accurate.
Considerable effort has been exerted to ensure the validity of
these databases. Indeed, previously developed screening algorithms for other critical care syndromes (using this same technology) at our institution have proven highly successful.2124
Statistics
In the current study, we included all eligible patients who
underwent operations during the index years 2004 and 2011.
Baseline characteristics are summarized by frequency (percentage) or median (interquartile range [IQR]) for categorical and
continuous variables, respectively. Comparison of characteristics between calendar years was carried out using the t test and
the chi-square test test, respectively. We evaluated the frequency
of TRALI after intraoperative blood product transfusion by calculating event rates both overall and separately for the calendar
years 2004 and 2011. Thereafter, age and transfusion volume
were categorized by quintiles, and age-specific and transfusion
volumespecific rates were calculated and tested statistically
using the Cochran-Armitage trend test. Sex-specific rates were
calculated and tested using the chi-square test, and surgical specialtyspecific rates were calculated and tested using the MonteCarlo estimate of the exact test. Agreement between physicians
adjudicating outcome was assessed using coefficients. Finally,
an unadjusted comparison of outcomes was carried out within
this cohort between patients determined to have TRALI versus
complication-free transfused patients (no evidence of TRALI or
TACO). Specifically, ICU length of stay and hospital length of

stay (median, IQR in days) were compared using the t test. Inhospital mortality (frequency [percentage]) was compared using
the chi-square test, and an odds ratio (95% CI) of in-hospital
death was generated. All statistical analyses were carried out
using SAS version 9.1 (SAS Institute Inc.), with P values < 0.05
considered significant.
Results
A total of 83,204 unique patients underwent surgical procedures at our institution during the index calendar years
(41,490 in 2004 and 41,714 in 2011). Of these, 3,379
adult patients undergoing noncardiac surgery with general
anesthesia were eligible for inclusion in this study (1,817
in 2004 and 1,562 in 2011) (fig.1). Baseline characteristics are summarized in table4. In brief, the median age for
this transfused cohort was 66 yr (range, 54 to 76 yr), with
patients being marginally older in 2004. About half of the
patients (1,614 [47.8%]) were male. The median intraoperative transfusion volume was 660ml. Both age and median
transfusion volume were comparable between calendar years.
Overall, most patients underwent abdominal surgeries (944
of 3,379 [27.9%]), followed by orthopedic surgeries (673 of
3,379 [19.9%]), with the fewest patients undergoing neurologic surgery (99 of 3,379 [2.9%]). There was a statistically
significant difference in the distribution of patients among
surgical specialties between 2004 and 2011 (P < 0.001).
Overall, 512 (15.2%) patients electronically screened
positive for TRALI, 324 in 2004 and 188 in 2011. Manual
review showed that 45 of these 512 screen-positive patients
(8.8%) had experienced TRALI/possible TRALI (agreement by statistic, 0.22). The overall rate of TRALI occurring within 6h of the last intraoperative blood product
transfusion was therefore 1.3% (95% CI, 1.0 to 1.8%) (45

of 3,379) (table5). Specifically, 13 cases (0.4% [95% CI,
0.2 to 0.7%]) were deemed to be TRALI, whereas 32 cases
(0.9% [95% CI, 0.7 to 1.3%]) were deemed to be possible
TRALI due to the presence of alternate acute lung injury risk
factors (table5).14 Of the 13 true TRALI cases, 5 occurred in
2004 (0.3% [95% CI, 0.1 to 0.6%]) and 8 occurred in 2011
(0.5% [95% CI, 0.3%-1.0%]). Of the 32 possible TRALI
cases, 18 occurred in 2004 (1.0% [95% CI, 0.6 to 1.6%])
and 14 occurred in 2011 (0.9% [95% CI, 0.5 to 1.5%]).
Product-specific rates of TRALI and possible TRALI are
also shown in table6. The composite rate of TRALI did not
change significantly between 2004 (23 of 1,817) and 2011
(22 of 1,562): 1.3% (95% CI, 0.8 to 1.9%) versus 1.4%
(95% CI, 0.9 to 2.1%) (P = 0.72) (table5).
The incidence of TRALI was comparable among male
and female patients in both calendar years (1.4 vs. 1.2%)
(P = 0.65) (table5). Overall, thoracic (2 of 45 [3.0%]), vascular (10 of 45 [2.7%]), and transplant surgeries (4 of 45
[2.2%]) carried the highest rates of TRALI/possible TRALI.
In fact, the rate of TRALI in thoracic surgical patients
was approximately three times that of patients undergoing
abdominal or orthopedic surgeries, and more than three
times the risk observed in patients undergoing spinal or urologic surgeries. Obstetric and gynecologic surgical patients
had no TRALI episodes during either study year. This variability in the rate of TRALI across surgical specialties failed
to reach statistical significance (P = 0.07). There was some
increased incidence of TRALI/possible TRALI with increasing age that was not significant (P = 0.06). Specifically, the
rate of TRALI in patients aged 80 yr was almost twice the
rate observed in all groups aged 69 yr, and almost 1.5
times the rate observed in patients aged 70 to 79 yr. The rate
Table 4. Baseline Characteristics of Transfusion Recipients by Age, Sex, Transfusion Volume, and Surgical Specialty in 2004 and
2011
Characteristics
Age, median (IQR), yr*
Male sex*
Intraoperative transfusion volume,
median (IQR), ml
Surgical specialty
Abdominal
OB/GYN
Neurologic
Orthopedic
Spine
Thoracic
Transplant
Urology
Vascular
Other
Overall
(N = 3,379)
66 (54 to 76)
1,614 (47.8)
660 (350 to 1,200)
2004
(n = 1,817)
2011
(n = 1,562)
67 (54 to 76)
858 (47.2)
639 (350 to 1,050)
65 (54 to 75)
756 (48.4)
660 (330 to 1,028)
481 (26.5)
157 (8.6)
45 (2.5)
401 (22.1)
114 (6.3)
59 (3.2)
81 (4.5)
154 (8.5)
248 (13.6)
77 (4.2)
463 (29.6)
136 (8.7)
54 (3.5)
272 (17.4)
135 (8.6)
74 (4.7)
97 (6.2)
125 (8.0)
127 (8.1)
79 (5.1)
P Value*
.004
.049
.070
<0.001
994 (27.9)
293 (8.7)
99 (2.9)
673 (19.9)
249 (7.4)
133 (3.9)
178 (5.3)
279 (8.3)
375 (11.1)
156 (4.6)
Values are numbers (percentage) unless indicated otherwise.

*P values compare medians and frequencies between 2004 and 2011 using the t test and the chi-square test, respectively.
IQR = interquartile range; OB/GYN = obstetrics and gynecology.
Table 5. Incidence Rates of Perioperative Transfusion-related Acute Lung Injury Overall and by Age, Sex, Transfusion Volume, and
Surgical Specialty in 2004 and 2011
Variable
Overall
Age, yr

49
50 to 59
60 to 69
70 to 79

80
Sex
Male
Female
Surgical specialty
Abdominal
OB/GYN
Neurologic
Orthopedic
Spine
Thoracic
Transplant
Urology
Vascular
Other
Transfusion volume, ml

350
351 to 700
701 to 1,050
1,051 to 1,400

1,401
2004 No./Total (%)
2011 No./Total (%)
Overall No./Total (%)
23/1,817 (1.3)
22/1,562 (1.4)
45/3,379 (1.3)
3/318 (0.9)
2/296 (0.7)
4/414 (1.0)
8/486 (1.6)
6/303 (2.0)
3/277 (1.1)
3/288 (1.0)
6/421 (1.4)
5/370 (1.4)
5/206 (2.4)
6/595 (1.0)
5/584 (0.8)
10/835 (1.2)
13/856 (1.5)
11/509 (2.2)
12/857 (1.4)
11/960 (1.1)
11/756 (1.5)
11/806 (1.4)
23/1,613 (1.4)
22/1,766 (1.2)
5/481 (1.0)
0/157 (0.0)
1/45 (2.2)
4/401 (1.0)
1/114 (0.9)
2/59 (3.4)
0/81 (0.0)
1/154 (0.6)
7/248 (2.8)
2/77 (2.6)
6/463 (1.3)
0/136 (0.0)
1/54 (1.9)
3/272 (1.1)
1/135 (0.7)
2/74 (2.7)
4/97 (4.1)
1/125 (0.8)
3/127 (2.4)
1/79 (1.3)
11/944 (1.2)
0/293 (0.0)
2/99 (2.0)
7/673 (1.0)
2/249 (0.8)
4/133 (3.0)
4/178 (2.2)
2/279 (0.7)
10/375 (2.7)
3/156 (1.9)
2/606 (0.3)
2/635 (0.3)
3/200 (1.5)
3/130 (2.3)
13/246 (5.3)
4/512 (0.8)
5/485 (1.0)
2/187 (1.1)
2/101 (2.0)
9/277 (3.2)
6/1,118 (0.5)
7/1,120 (0.6)
5/387 (1.3)
5/231 (2.2)
22/523 (4.2)
P
Value
0.72
0.06
0.65
0.07
<0.001
OB/GYN = obstetrics and gynecology.
Table 6. Rates of TRALI and Possible TRALI by Type of

Transfusion Product for 2004 and 2011
TRALI, No.
Product
Erythrocyte only
FFP only
Platelets only
Mixed
Possible
TRALI, No.*
2004
(n = 5)
2011
(n = 8)
2004
(n = 18)
2011
(n = 14)
1
0
0
4
3
0
0
5
6
0
0
12
5
2
1
6
*Thirty-two cases were deemed as possible TRALI due to the presence of

sepsis (n = 7), trauma (n = 6), shock (n = 6), aspiration (n = 3), pneumonia
(n = 1), or surgical procedures with high risk for ALI/ARDS (n = 9). Some
patients fit into multiple categories.
FFP = fresh frozen plasma; TRALI = transfusion-related acute lung injury.
of TRALI also increased with greater volumes of transfused

product (P < 0.001; table5).
Importantly, of the 45 cases of TRALI identified during
these two calendar years, none were reported to the transfusion medicine service. However, our comparison of outcomes
for patients identified as having TRALI with outcomes for
complication-free transfused patients revealed a number of
important differences. Namely, for the 2,073 patients who
were admitted to the ICU postoperatively (61.3% of the

study cohort), the median length of ICU stay for patients
with TRALI was 11.2 days (IQR, 5.7 to 27.9 days), compared
with 5.9 days (IQR, 2.7 to 15.2 days) for transfused controls
(P < 0.001). Patients with TRALI also had a significantly longer median hospital length of stay at 12.3 days (IQR, 6.1 to
23.0 days), compared with 6.6 days (IQR, 4.5 to 10.0 days)
for complication-free transfused patients (P < 0.001). Finally,
the rate of in-hospital mortality for the 45 TRALI cases was
28.9% (n = 13), compared with 2.5% (n = 81) for the 3,193
complication-free transfused patients. This resulted in an odds
ratio of 15.6 (95% CI, 7.9 to 30.8) for death in TRALI cases,
compared with transfused controls (P < 0.001).
Discussion
This large-scale cohort study of noncardiac surgical patients
undergoing general anesthesia offers insights into the incidence of TRALI after intraoperative blood product transfusion. Using a novel NLP-based algorithm, followed by a
rigorous two-phase manual review strategy, we conducted a
detailed epidemiologic analysis of postoperative TRALI. Our
findings confirmed that TRALI is dramatically underrecognized and underreported in clinical practice; specifically, its
1.3% incidence is greater than many previous estimates.

Notably, increasing transfusion volume was a potent predictor of postoperative TRALI, whereas increasing incidence
with age fell just short of statistical significance. Importantly,
the median duration of hospitalization was markedly prolonged for patients with TRALI versus controls (ICU, 11.2
vs. 5.9 days; hospital overall, 12.3 vs. 6.6 days). Furthermore,
in-hospital mortality was augmented (odds ratio, 15.6) in
TRALI patients versus complication-free transfused controls.
Despite targeted mitigation strategies, TRALI incidence was
unchanged between 2004 and 2011 (1.3% vs. 1.4%) (P =
0.72).
Despite its underrecognition,2,3,25,26 TRALI remains the
leading cause of transfusion-related fatalities in developed
countries.1 Beyond the prognostic impact on transfusion
recipients, accurate reporting of TRALI is critical for identificationand possible exclusionof implicated, potentially alloimmunized donors. As highlighted by Kopko et
al.,2 unrecognized TRALI may lead to the administration of
blood products from risky donors, precipitating additional
TRALI episodes in vulnerable recipients.
In this study, we determined rates of TRALI specific to
surgical patients transfused intraoperatively. A large proportion of blood product transfusions (about 50%) takes place
in operating rooms,9 and a substantial proportion (41%)
of TRALI is experienced after intraoperative transfusion.3
Nonetheless, most studies investigating TRALI rates have
focused on critically ill patients, with an incidence historically estimated at 0.02 to 0.05% per unit transfused5,26 and
at 2.2 to 8.2% per patient transfused in the ICU.4,27,28
Although the importance of TRALI in surgical populations
has been recognized since the 1990s,29 data are sparse concerning its occurrence. Early investigations of perioperative
TRALI were limited by passive reporting, markedly underestimating its true incidence.30 Several recent isolated studies have
focused on high-risk surgical populations, which, coupled
with differences in methodology and outcome adjudication,
limits generalizability to the wider surgical population.31,32
In 2012, Tsai et al.33 conducted a large retrospective
study of postoperative TRALI in which they identified only
15 cases in 14,441 patients (0.1%). Notably, they retrospectively identified previously adjudicated TRALI diagnoses
from an existing database that defined TRALI as hypoxemia
(Pao2:Fio2 ratio <300 mmHg), with evidence of pulmonary
edema. However, the absence of an available Pao2:Fio2 ratio
in these patients precluded the diagnosis of TRALI. The
authors acknowledged that inadequate retrospective documentation may have discounted less severe forms of TRALI.
We overcame this limitation by including other parameters
of hypoxemia (Spo2 and Pao2 [fig.2]).
Tsai et als33 exclusion of patients with other risk factors
for acute lung injury would also have excluded all cases of
possible TRALI. In contrast, we grouped TRALI and possible TRALI as a single outcome, likely accounting for
much of the difference in our incidence rates. Tsai et al.33
also excluded patients undergoing thoracic surgery, whereas

we included them and found them to be at highest risk for
postoperative TRALI (3.0%). Finally, Tsai et al.33 included
patients receiving exclusively regional anesthesia (none experienced TRALI), whereas we evaluated postoperative TRALI
in patients undergoing general anesthesia only.
Undoubtedly, uncertainty remains about the true incidence of TRALI; however, numerous explanations may
account for the variable previous estimates. Before the 2004
consensus criteria for TRALI,14 its definition in the medical literature was inconsistent. Since then, explicit criteria
applied under study conditions have varied.34,35 Furthermore, while some authors reported TRALI rates,33,36 others
grouped TRALI and possible TRALI37 (as we did).
Of equal importance are the different methods by which
cases are identified. Some investigators have used manual retrospective chart review35; others have conducted prospective
active surveillance36; but few have used electronic algorithms
in case identification.38 As demonstrated previously, passive
identification strategies are prone to underrecognition, particularly of less severe cases. Additionally, the populations in
each study range from all transfused patients36 to the critically
ill4 and, rarely, to surgical subsets,31,32 with differing underlying risks for TRALI, thus limiting comparability. Finally, compounding these study design differences is the impact of the
passage of time. Not only has TRALI recognition improved,
but there have also been changes in blood management and
donor procurement strategies that may have reduced the rate
of TRALI.10,11,37 Although our study did not demonstrate
this hypothesized decline in incidence after TRALI mitigation strategies were introduced (as discussed in the next paragraph of this Discussion), a substantial and growing body of
medical literature supports this effect.11,37 Consequently, we
believe that early data before the mid-2000s are unlikely to be
representative of the current syndrome burden.
Interestingly, we did not observe the expected decline in
TRALI rates from 2004 to 2011, with the introduction of
TRALI mitigation strategies, perhaps because we grouped
TRALI and possible TRALI as a combined outcome.
Although previous reports indicate that TRALI mitigation
strategies have resulted in a significant decline in the rate of
true TRALI events, their impact on the occurrence of possible TRALI remains unclear.11 We believe this previously
observed decline in TRALI incidence with TRALI mitigation
strategies to be a true effect, but our detailed breakdown of
results by year, implicated product, and TRALI versus possible TRALI (table6) does not support this hypothesis. However, given the low overall event rate (45 cases in 2 yr), our
study was not sufficiently powered to detect such a difference.
Another noteworthy observation is the association
between TRALI incidence and transfusion volume. Indeed,
a positive association between dose of transfusion and
TRALI incidence has been reported.36 Although TRALI is
believed to be largely immune mediated, the risk of receiving
an alloimmunized donor unit undoubtedly increases with
more transfusions. Indeed, patients experiencing TRALI

have often received a proportionally greater number of units
than their complication-free transfused counterparts. In
addition, in the current study, physician reviewers adjudicated cases as TRALI, TACO, both, or neither. Patients who
had both TRALI and TACO (n = 23) were those in whom
the evidence of volume overload did not fully explain the
extent of respiratory distress. These patients were included in
the current analyses, which may partly explain the increased
average volume transfused in TRALI patients.
Notably, several study limitations merit discussion. First,
in light of the excellent sensitivity of our electronic screening algorithm, we limited manual review to screen-positive
transfusion recipients. Hence, we may have overlooked some
cases that screened false-negative. Without manual review
of all screen-negative patients, we are unable to corroborate
this possibility. However, we again stress that the algorithm
achieved a sensitivity of 92.5% (95% CI, 84.6 to 96.7%) in
our derivation cohort.17 Although deficiencies may still exist,
our NLP-based algorithm has been shown to significantly
improve upon manual detection alone,3 as well as upon nonNLPbased electronic algorithms.17 For a sensitivity analysis, we tested the assumption that our algorithm may have
performed at the lower boundary of the CI of its previously
demonstrated sensitivity (84.6%). In this circumstance, we
determined that there could have been up to 53 cases of
TRALI in this population of 3,379 patients as opposed to
the 45 identified in our primary analyses. The rate of TRALI
would therefore marginally increase to 1.6% as opposed to
the 1.3% reported in the primary analyses. We also emphasize that there were no cases of TRALI reported to the blood
bank during our study period among the screen-negative
population. This further supports the accuracy of the TRALI
screening algorithm. Second, the nature of the study population at our single-center, tertiary-care referral center limits the
generalizability of our results. We surmise that patients may
have had a greater prevalence of comorbid conditions and
may have been more acutely unwell than surgical populations
elsewhere. Thus, the incidence of TRALI in our practice may
be above average. Finally, our investigation was also limited
by well-recognized difficulties pertaining to making a TRALI
diagnosis. Indeed, diagnosing TRALI, TACO, or acute lung
injury can be challenging in the best clinical settings. This
diagnostic dilemma was further pronounced in our attempt
to retrospectively identify cases of TRALI in a postoperative cohort that frequently experiences atelectasis. However,
to overcome this dilemma, we implemented robust methodology whereby patients were first electronically screened.
Thereafter, two independent physicians with previous training manually reviewed cases. When discrepancies arose, their
review was followed by a panel review by three senior critical
care physicians. We believe that these procedures represent
a marked improvement over those used in past studies aiming to characterize TRALI epidemiology using only passive
reporting or single-reviewer outcome adjudication.
In conclusion, our findings document a contemporary

rate of TRALI/possible TRALI of 1.4% after intraoperative
transfusion for noncardiac surgery. This rate was consistent
for the 2 yr investigated (2004 and 2011) and was not meaningfully impacted by TRALI mitigation strategies. These
results should help increase awareness among perioperative
healthcare providers, and thereby improve case recognition, reporting, and, ultimately, transfusion practice. Future
efforts should focus on identification of pertinent and potentially modifiable risk factors for postoperative TRALI.
Acknowledgments
This study was performed with the support of funds from
the Mayo Clinic Center for Translational Science Activities
High-Impact, Pilot and Feasibility Award (HIPFA-2012Dr.
Pathak; 94164003), Rochester, Minnesota, and of funds from
the Department of Critical Care Independent Multidisciplinary Program, Mayo Clinic, Rochester, Minnesota.
Competing Interests
The authors declare no competing interests.
Correspondence
Address correspondence to Dr. Kor: Department of Anesthesiology, Mayo Clinic, 200 First Street, SW, Rochester,
Minnesota 55905. kor.daryl@mayo.edu. This article may be
accessed for personal use at no charge through the Journal
Web site, www.anesthesiology.org.
References
1. Fatalities Reported to FDA Following Blood Collection and
Transfusion: Annual summary for fiscal year 2013: 2013 US
Department of Health and Human Services. Rockville (MD),
US Food and Drug Administration, 2013
2. Kopko PM, Marshall CS, MacKenzie MR, Holland PV,
Popovsky MA: Transfusion-related acute lung injury: Report
of a clinical look-back investigation. JAMA 2002; 287:196871
3. Clifford L, Singh A, Wilson GA, Toy P, Gajic O, Malinchoc
M, Herasevich V, Pathak J, Kor DJ: Electronic health record
surveillance algorithms facilitate the detection of transfusion-related pulmonary complications. Transfusion 2013;
53:120516
4. Gajic O, Rana R, Winters JL, Yilmaz M, Mendez JL, Rickman
OB, OByrne MM, Evenson LK, Malinchoc M, DeGoey SR,
Afessa B, Hubmayr RD, Moore SB: Transfusion-related acute
lung injury in the critically ill: Prospective nested case-control study. Am J Respir Crit Care Med 2007; 176:88691
5. Popovsky MA, Moore SB: Diagnostic and pathogenetic
considerations in transfusion-related acute lung injury.
Transfusion 1985; 25:5737
6. Gajic O, Moore SB: Transfusion-related acute lung injury.
Mayo Clin Proc 2005; 80:76670
7. Silliman CC, Boshkov LK, Mehdizadehkashi Z, Elzi DJ, Dickey
WO, Podlosky L, Clarke G, Ambruso DR: Transfusion-related
acute lung injury: Epidemiology and a prospective analysis
of etiologic factors. Blood 2003; 101:45462
8. Wallis JP, Lubenko A, Wells AW, Chapman CE: Single hospital
experience of TRALI. Transfusion 2003; 43:10539
9. Van Dijk PM, Kleine JW: The transfusion reaction in anesthesiological practice. Acta Anaesthesiol Belg 1976; 27:24754
10. Blumberg N, Heal JM, Gettings KF, Phipps RP, Masel D,
Refaai MA, Kirkley SA, Fialkow LB: An association between
decreased cardiopulmonary complications (transfusionrelated acute lung injury and transfusion-associated circulatory overload) and implementation of universal
leukoreduction of blood transfusions. Transfusion 2010;
50:273844
11. Wiersum-Osselton JC, Middelburg RA, Beckers EA, van

Tilborgh AJ, Zijlker-Jansen PY, Brand A, van der Bom JG,
Schipperus MR: Male-only fresh-frozen plasma for transfusion-related acute lung injury prevention: before-and-after
comparative cohort study. Transfusion 2011; 51:127883
12. von Elm E, Altman DG, Egger M, Pocock SJ, Gtzsche PC,
Vandenbroucke JP; STROBE Initiative: The Strengthening
the Reporting of Observational Studies in Epidemiology
(STROBE) statement: Guidelines for reporting observational
studies. Ann Intern Med 2007; 147:5737
13. Herasevich V, Kor DJ, Li M, Pickering BW: ICU data mart: A
non-iT approach. A team of clinicians, researchers and informatics personnel at the Mayo Clinic have taken a homegrown
approach to building an ICU data mart. Healthc Inform 2011;
28:42, 445
14. Kleinman S, Caulfield T, Chan P, Davenport R, McFarland

J, McPhedran S, Meade M, Morrison D, Pinsent T, Robillard
P, Slinger P: Toward an understanding of transfusionrelated acute lung injury: Statement of a consensus panel.
15. Gajic O, Dabbagh O, Park PK, Adesanya A, Chang SY, Hou P,
Anderson H III, Hoth JJ, Mikkelsen ME, Gentile NT, Gong MN,
Talmor D, Bajwa E, Watkins TR, Festic E, Yilmaz M, Iscimen
R, Kaufman DA, Esper AM, Sadikot R, Douglas I, Sevransky
J, Malinchoc M; U.S. Critical Illness and Injury Trials Group:
Lung Injury Prevention Study Investigators (USCIITG-LIPS):
Early identification of patients at risk of acute lung injury:
evaluation of lung injury prediction score in a multicenter
cohort study. Am J Respir Crit Care Med 2011; 183:46270
16. Kor DJ, Warner DO, Alsara A, Fernndez-Prez ER, Malinchoc
M, Kashyap R, Li G, Gajic O: Derivation and diagnostic accuracy of the surgical lung injury prediction model.
17. Clifford L, Wilson GA, Gajic O, Toy P, Herasevich V, Murphy
S, Pathak J, Kor DJ: Natural language processing of chest
radiograph reports improves the identification of transfusion-related pulmonary complications [abstract]. Am J Respir
Crit Care Med 2013; 187:A2218
18. Savova GK, Masanz JJ, Ogren PV, Zheng J, Sohn S, KipperSchuler KC, Chute CG: Mayo clinical Text Analysis and
Knowledge Extraction System (cTAKES): architecture, component evaluation and applications. J Am Med Inform Assoc
2010; 17:50713
19. Chute CG, Beck SA, Fisk TB, Mohr DN: The Enterprise Data
Trust at Mayo Clinic: A semantically integrated warehouse of
biomedical data. J Am Med Inform Assoc 2010; 17:1315
20. Plaisant C, Lam S, Shneiderman B, Smith MS, Roseman D,
Marchand G, Gillam M, Feied C, Handler J, Rappaport H:
Searching electronic health records for temporal patterns in
patient histories: A case study with Microsoft amalga. AMIA
Annu Symp Proc 2008; 6015
21. Alsara A, Warner DO, Li G, Herasevich V, Gajic O, Kor DJ:
Derivation and validation of automated electronic search
strategies to identify pertinent risk factors for postoperative
acute lung injury. Mayo Clin Proc 2011; 86:3828
22. Cartin-Ceba R, Kojicic M, Li G, Kor DJ, Poulose J, Herasevich
V, Kashyap R, Trillo-Alvarez C, Cabello-Garza J, Hubmayr
R, Seferian EG, Gajic O: Epidemiology of critical care syndromes, organ failures, and life-support interventions in a
suburban US community. Chest 2011; 140:144755
23. Schmickl CN, Li M, Li G, Wetzstein MM, Herasevich V, Gajic
O, Benzo RP: The accuracy and efficiency of electronic
screening for recruitment into a clinical trial on COPD.

Respir Med 2011; 105:15016
24. Singh B, Singh A, Ahmed A, Wilson GA, Pickering BW,

Herasevich V, Gajic O, Li G: Derivation and validation of
automated electronic search strategies to extract Charlson
comorbidities from electronic medical records. Mayo Clin
Proc 2012; 87:81724
25. Wallis JP: Transfusion-related acute lung injury (TRALI)

Under-diagnosed and under-reported. Br J Anaesth 2003;
90:5736
26. Popovsky MA, Chaplin HC Jr, Moore SB: Transfusion-related
acute lung injury: A neglected, serious complication of
hemotherapy. Transfusion 1992; 32:58992
27. Wallis JP: Transfusion-related acute lung injury (TRALI):

Presentation, epidemiology and treatment. Intensive Care
Med 2007; 33(suppl 1):S126
28. Palfi M, Berg S, Ernerudh J, Berlin G: A randomized controlled trial oftransfusion-related acute lung injury: Is plasma
from multiparous blood donors dangerous? Transfusion
2001; 41:31722
29. Florell SR, Velasco SE, Fine PG: Perioperative recognition,
management, and pathologic diagnosis of transfusion-related
acute lung injury. Anesthesiology 1994; 81:50810
30. Weber JG, Warner MA, Moore SB: What is the incidence

of perioperative transfusion-related acute lung injury?
31. Wright SE, Snowden CP, Athey SC, Leaver AA, Clarkson JM,
Chapman CE, Roberts DR, Wallis JP: Acute lung injury after
ruptured abdominal aortic aneurysm repair: The effect of
excluding donations from females from the production of
fresh frozen plasma. Crit Care Med 2008; 36:1796802
32. Vlaar AP, Hofstra JJ, Determann RM, Veelo DP, Paulus F,

Kulik W, Korevaar J, de Mol BA, Koopman MM, Porcelijn L,
Binnekade JM, Vroom MB, Schultz MJ, Juffermans NP: The
incidence, risk factors, and outcome of transfusion-related
acute lung injury in a cohort of cardiac surgery patients:
A prospective nested case-control study. Blood 2011;
117:421825
33. Tsai HI, Chou AH, Yang MW: Perioperative transfusion
related acute lung injury: A retrospective analysis. Acta
Anaesthesiol Taiwan 2012; 50:96100
34. Eder AF, Herron R, Strupp A, Dy B, Notari EP, Chambers LA,
Dodd RY, Benjamin RJ: Transfusion-related acute lung injury
surveillance (2003-2005) and the potential impact of the
selective use of plasma from male donors in the American
Red Cross. Transfusion 2007; 47:599607
35. Ozier Y, Muller JY, Mertes PM, Renaudier P, Aguilon P,

Canivet N, Fabrigli P, Rebibo D, Tazerout M, Trophilme C,
Willaert B, Caldani C: Transfusion-related acute lung injury:
Reports to the French Hemovigilance Network 2007 through
2008. Transfusion 2011; 51:210210
36. Toy P, Gajic O, Bacchetti P, Looney MR, Gropper MA,

Hubmayr R, Lowell CA, Norris PJ, Murphy EL, Weiskopf RB,
Wilson G, Koenigsberg M, Lee D, Schuller R, Wu P, Grimes B,
Gandhi MJ, Winters JL, Mair D, Hirschler N, Sanchez Rosen
R, Matthay MA; TRALI Study Group: Transfusion-related
acute lung injury: Incidence and risk factors. Blood 2012;
119:175767
37. Arinsburg SA, Skerrett DL, Karp JK, Ness PM, Jhang J,
Padmanabhan A, Gibble J, Schwartz J, King KE, Cushing
MM: Conversion to low transfusion-related acute lung injury
(TRALI)-risk plasma significantly reduces TRALI. Transfusion
2012; 52:94652
38. Finlay HE, Cassorla L, Feiner J, Toy P: Designing and testing
a computer-based screening system for transfusion-related
acute lung injury. Am J Clin Pathol 2005; 124:6019
Characterizing the Epidemiology of Perioperative

Transfusion-associated Circulatory Overload
Leanne Clifford, B.M., Qing Jia, M.D., Hemang Yadav, M.B.B.S., Arun Subramanian, M.B.B.S.,
Gregory A. Wilson, R.R.T., Sean P. Murphy, B.S., Jyotishman Pathak, Ph.D., Darrell R. Schroeder, M.S.,
Mark H. Ereth, M.D., Daryl J. Kor, M.D.
ABSTRACT
Background: Transfusion-associated circulatory overload (TACO) is a leading cause of transfusion-related fatalities, but its
incidence and associated patient and transfusion characteristics are poorly understood. To inform surgical transfusion practice
and to begin mitigating perioperative TACO, the authors aimed to define its epidemiology.
Methods: In this retrospective cohort study, the medical records of adult patients undergoing noncardiac surgery with general
anesthesia during 2004 or 2011 and receiving intraoperative transfusions were screened using an electronic algorithm for
identification of TACO. Those patients who were screened as high probability for TACO underwent rigorous manual review.
Univariate and multivariate analyses evaluated associations between patient and transfusion characteristics with TACO rates
in a before-and-after study design.
Results: A total of 2,162 and 1,908 patients met study criteria for 2004 and 2011, respectively. The incidence of TACO was
5.5% (119 of 2,162) in 2004 versus 3.0% (57 of 1,908) in 2011 (P < 0.001), with comparable rates for men (4.8% [98 of
2,023]) and women (3.8% [78 of 2,047]) (P=0.09). Overall, vascular (12.1% [60 of 497]), transplant (8.8% [17 of 193]),
and thoracic surgeries (7.2% [10 of 138]) carried the highest TACO rates. Obstetric and gynecologic patients had the lowest
rate (1.4% [4 of 295]). The incidence of TACO increased with volume transfused, advancing age, and total intraoperative
fluid balance (all P < 0.001).
Conclusions: The incidence of perioperative TACO is similar to previous estimates in nonsurgical populations. There was a
reduction in TACO rate between 2004 and 2011, with incidence patterns remaining comparable in subgroup analyses. Future
efforts exploring risk factors for TACO may guide preventive or therapeutic interventions, helping to further mitigate this
transfusion complication. (Anesthesiology 2015; 122:21-8)
INCE the 1940s, circulatory overload after blood

product transfusion has been reported in the medical
literature.1 Despite this recognition, transfusion-associated circulatory overload (TACO) was neither defined nor
investigated as a separate entity until the 1990s.2 With the
initiation of passive reporting to the U.S. Food and Drug
Administration in 2005, TACO has become increasingly
recognized, and recent estimates suggest incidence rates as
high as 11%.35 During this time, TACO has accounted for
2 to 27% of the transfusion-related fatalities reported to the
U.S. Food and Drug Administration, making it the second
leading cause of transfusion-related death after transfusionrelated acute lung injury (TRALI).*
Notably, most investigators evaluating the epidemiology
of TACO have concentrated on patients in the intensive care
unit (ICU).3,5,6 As a result, the epidemiology of TACO and
its resultant burden in surgical populations remains incompletely defined. Moreover, TACO is frequently overlooked

Transfusion-associated circulatory overload is a leading cause
of transfusion-related fatalities; however, its epidemiology after
noncardiac surgery is not well characterized

This retrospective cohort study evaluated 2,162 and 1,908
patients who received intraoperative transfusions during noncardiac surgery in 2004 and 2011, respectively
A total of 119 patients (5.5%) in 2004 and 57 patients (3%) in
2011 met criteria for transfusion-associated circulatory overload
The incidence of transfusion-associated circulatory overload
increased with the volume of blood product transfused, advanced age, and total intraoperative fluid balance
in clinical practice and is infrequently reported to the transfusion medicine service.7,8 Thus, the actual incidence of
perioperative TACO remains poorly defined and is likely
much greater than that currently reported in other patient
This article is featured in This Month in Anesthesiology, page 1A. Corresponding article on page 1. Supplemental Digital Content is
available for this article. Direct URL citations appear in the printed text and are available in both the HTML and PDF versions of this article.
Links to the digital files are provided in the HTML text of this article on the Journals Web site (www.anesthesiology.org). Presented at the
American Society of Anesthesiology Annual Congress, San Francisco, California, October 13, 2013.
Submitted for publication November 13, 2013. Accepted for publication August 29, 2014. From the Department of Anesthesiology (L.C.,
Q.J., A.S., M.H.E., G.A.W., D.J.K.), Division of Pulmonary and Critical Care Medicine (H.Y.), Department of Information Technology (S.P.M.),
and Division of Biomedical Statistics and Informatics ( J.P., D.R.S.), Mayo Clinic, Rochester, Minnesota.
* Available at: http://www.fda.gov/BiologicsBloodVaccines/SafetyAvailability/ReportaProblem/TransfusionDonationFatalities/ucm391574.
htm. Accessed May 9, 2014.
21
January 2015
Perioperative TACO
populations. Not only does this failure to appreciate TACO

events contribute to our incomplete understanding of
TACO epidemiology, but it may also result in suboptimal
care delivery and unfavorable outcomes for patients.
Current data demonstrate that approximately 50% of
all blood product transfusions take place in the perioperative setting.9 Perioperative care providers are thus in an ideal
position to recognize transfusion risks, which may allow for
more appropriate transfusion therapies and for more timely
and appropriate interventions in patients with transfusionrelated complications such as TACO. Thus, it is imperative
to more clearly define the epidemiology of perioperative
TACO. To this end, we sought to define the incidence of
TACO in a large cohort of patients who had undergone noncardiac surgery. In addition, we aimed to further describe
the incidence of TACO as it relates to specific characteristics of patients and transfusion situations. Finally, in light
of a recent report postulating that universal leukoreduction
reduces the rate of TACO,7 we assessed the impact of this
intervention by evaluating the incidence of TACO in all
patients who received transfusions before (2004) and several
years after (2011) its implementation in 2005.

Study Design
In this retrospective cohort study evaluating the incidence of
TACO, we analyzed the data abstracted from the electronic
medical records (EMRs) of adult noncardiac surgery patients
treated with general anesthesia at Mayo Clinic who had
received intraoperative blood product transfusions in 2004
or 2011. All patients had previously given signed consent
for the use of their medical records for research purposes,

and the Mayo Clinic Institutional Review Board approved
the study before its onset. The Strengthening the Reporting
of Observational Studies in Epidemiology guidelines were
followed in the conduct of this study and in the reporting
of its results.10
Study Population
Adult patients (aged 18 yr) who had undergone noncardiac surgical procedures with general anesthesia at our
institution (Mayo Clinic, Rochester, Minnesota) during
the calendar years 2004 or 2011 were identified from an
institutional database (the perioperative data mart).9 This
database contains a near real-time duplicate of all patient
data from monitored care environments. Data are stored
and accessed via an open-access connectivity database
using JMP statistical software (SAS Institute Inc., Cary,
NC). The accuracy of this database has been previously
described and validated.9,1113
All adult noncardiac surgical patients who received
intraoperative blood product transfusions were eligible for
inclusion. Exclusion criteria included (1) denial of research
authorization, (2) age younger than 18 yr, (3) previous inclusion in the study (i.e., patients who had multiple eligible
surgeries were included only once), (4) evidence of preoperative respiratory failure, (5) preoperative diffuse bilateral
infiltrates evident on chest radiographs, (6) death intraoperatively, (7) receipt of extracorporeal membrane oxygenation commenced intraoperatively before blood product
transfusion, and (8) receipt of nongeneral anesthesia instead
of general anesthesia (fig.1). Notably, all subjects in the current study were specifically identified for inclusion in this
Fig. 1. Patient flowchart. ECMO=extracorporeal membrane oxygenation; TACO=transfusion-associated circulatory overload;

TRALI=transfusion-related acute lung injury.
investigation and underwent outcome adjudication using

the methods described herein. None of the subjects included
in this study were identified from previous studies evaluating
the incidence of TRALI and TACO.
Outcome Adjudication
All study participants were screened for the development of
TACO using our recently developed natural language processing (NLP)based electronic screening algorithm (Supplemental Digital Content 1, http://links.lww.com/ALN/B112:
TACO screening algorithm).14 In brief, this screening tool
identifies patients with evidence of hypoxemia within 6h of
blood product transfusion and abnormalities consistent with
TACO on chest radiographs. The algorithm can detect cases
of TACO with a sensitivity and a specificity of 100% (95%
CI, 91.4 to 100.0%) and 93.6% (95% CI, 85.0 to 97.6%),
respectively.7,14 The EMRs of all patients who were screened as
high probability for TACO by this algorithm were then manually reviewed by two independent physicians (L.C. and Q.J.).
A final diagnosis of TACO was determined if patients met
at least three of the 2014 criteria established by the National
Healthcare Safety Network (table1). Our interest in the
epidemiology of TACO after intraoperative blood product
administration led us to limit outcomes assessments to the 6h
after the last blood product transfusion in the operating room.
If the two reviewing physicians disagreed on a diagnosis,
three senior critical care physicians reviewed the case to adjudicate a final outcome. During adjudication, physicians could
allocate a diagnosis of TACO as described previously (table 1);
TRALI as based on the 2004 Canadian consensus criteria,15
both, with evidence for both TACO and TRALI, but neither
considered sufficient to explain the clinical picture fully, or
neither (when case definitions were not met for either TACO
or TRALI). Because our primary aim in this investigation was
to describe the incidence of TACO in surgical patients, we do
not describe cases adjudicated as TRALI here. Of note, these
TRALI cases were included in the data set denominator when
TACO incidence rates were calculated but were excluded from
the analyses evaluating outcomes for TACO cases compared
with complication-free transfused patients. Patients considered to have both TRALI and TACO were included in our
incidence calculations. Finally, to verify the cases of TACO
identified in our study and to evaluate whether our methodology missed any known cases of TACO, we cross-referenced
all suspected transfusion reactions reported to the transfusion
medicine service with our study population using the FileNet
(IBM, Armonk, NY) system maintained at our institution for
all suspected transfusion reactions.
Data Sources and Collection
To execute our NLP-based algorithm,14 we first extracted
data used in the algorithm from existing institutional
databases. Specifically, unstructured data required for this
Available at: http://www.cdc.gov/nhsn/PDFs/Biovigilance/BVHV-protocol-current.pdf. Accessed May 9, 2014.
Table 1. TACO Definition from the Biovigilance Component of

the CDC National Healthcare Safety Network
2010 TACO Criteria
New onset or exacerbation of 3 of the following within 6h of
transfusion:
Acute respiratory distress (dyspnea, orthopnea, cough)
Evidence of positive fluid balance
Increased BNP
Radiographic evidence of pulmonary edema
Evidence of left heart failure
Increased CVP
Adapted from Division of Healthcare Quality Promotion; Centers for Disease Control and Prevention. The National Healthcare Safety Network
(NHSN) Manual: Biovigilance Component (Internet). Atlanta, GA (2010
July). Available at: http://www.cdc.gov/nhsn/PDFs/Biovigilance/BV-HVprotocol-current.pdf. Accessed May 9, 2014.
BNP=brain natriuretic peptide; CDC=Centers for Disease Control and
Prevention; CVP = central venous pressure; TACO = transfusion-associated circulatory overload.
algorithmchest radiograph reports obtained within 24h

of the last intraoperative blood product transfusion
were extracted from the enterprise data trust,12 an institutional warehouse of operational, research, and education
databases. Our recently developed NLP procedures were
applied to these reports to identify phrases consistent with
a diagnosis of TACO. Structured data points that were collected included partial pressure of arterial oxygen (Pao2),
Pao2 to fraction of inspired oxygen ratio (Pao2:Fio2), oxygen
saturations (Spo2), and respiratory rate (Supplemental Digital Content 1, http://links.lww.com/ALN/B112: TACO
screening algorithm).7,14 To ensure completeness, we collected these data from three databases: (1) the perioperative data mart (described above), (2) the ICU data mart9a
near real-time database that captures data directly from the
EMR for patients in the ICU,9 and (3) the Mayo Clinic Life
Sciences System11,12a repository of replicated data from
the EMR of patients admitted to the general surgical floor
postoperatively. The sensitivity of our algorithm was maximized by identifying the worst or most extreme values for
the patients up to 6h after blood product transfusion. The
most physiologically deranged values were manually verified in the EMR to minimize error. Baseline demographic
data were also extracted from the three databases, which
have been validated by numerous previous procedures.
Patient and Transfusion Characteristics
We collected detailed transfusion data using the perioperative information tool, a Microsoft.Net (Microsoft Corp.,
Redmond, WA) application that contains detailed transfusion information (e.g., exact transfusion times, types of
blood products, and volumes). Patient data were extracted
from the perioperative data mart, including age, sex, type
of surgery, ICU and hospital lengths of stay, and in-hospital
mortality status. The automated nature of these validated
data extraction procedures helps to ensure optimal data
accuracy. This technology has been used previously at our
institution to develop screening algorithms for other critical
care syndromes with great success.1517
Perioperative TACO
Statistics
To evaluate the frequency of TACO after intraoperative blood
product transfusion in all eligible patients operated on during
2004 or 2011, we calculated event rates (number of TACO
cases/number of encounters) overall and separately for each
year. Baseline characteristics are summarized by frequency
(percentage) or median (interquartile range [IQR]) for categorical and continuous variables, respectively. Comparison
of characteristics between calendar years was carried out using
the Wilcoxon rank sum test and the chi-square test, respectively. To assess the relation of various patient and transfusion
characteristics on the incidence of TACO, we categorized age,
transfusion volume, and fluid balance by quintiles; thereafter,
we calculated age-specific, sex-specific, surgical specialtyspecific, transfusion volumespecific, and fluid balancespecific
rates of TACO for each category (number of TACO cases/
number of encounters). Separate univariate logistic regression analyses were performed for each of these characteristics
to assess whether the rate of TACO was associated with the

given characteristic after adjustment for calendar year. Finally,
to assess the clinical significance of a TACO diagnosis in
patients compared with other patients in this cohort who did
not develop a pulmonary transfusion complication, ICU and
postoperative hospital lengths of stay were compared using
the t test. Patients considered to be transfused controls in this
cohort were those who were transfused intraoperatively and
did not go on to develop TACO or TRALI. Only patients
who went to the ICU were included in the analysis of length
of ICU stay. In addition, in-hospital mortality was compared
using the chi-square test. Statistical analyses were conducted
using the statistical software SAS version 9.1 (SAS Institute
Inc.), and P values less than 0.05 were considered significant.
Results
We identified 4,070 patients who were eligible for inclusion
in this study (2,162 in 2004 and 1,908 in 2011). Table2
Table 2. Baseline Characteristics of 4,070 Patients by Age, Sex, Transfusion Volume, Fluid Balance, and Surgical Specialty Overall
and by Index Year
Characteristics
Overall (N=4,070)
2004 (n=2,162)
Age, median (IQR), yr

Male sex, No. (%)
ASA status, No. (%)
I
II
III
IV
V
Emergency, No. (%)
Intraoperative transfusion
Volume, median (IQR), ml
Intraoperative erythrocytes
Intraoperative FFP
Intraoperative platelets
Intraoperative cryo
Intraoperative fluid balance, median (IQR), ml
Surgical specialty, No. (%)
Abdominal
OB/GYN
Neurologic
Orthopedic
Spine
Thoracic
Transplant
Urology
Vascular
Other
65 (5375)
2,023 (49.7)
66 (5376)
1,057 (48.9)
104 (2.6)
1,268 (31.2)
2,272 (55.8)
399 (9.8)
26 (0.6)
477 (11.7)
4,070 (100)
650 (3301,025)
3,875 (95.2)
660 (330996)
540 (13.3)
725 (4901,326)
487 (12.0)
300 (227562)
134 (3.3)
208 (194387)
4,180 (2,6996,154)
49 (2.3)
601 (27.8)
1,308 (60.5)
196 (9.1)
7 (0.3)
231 (10.7)
2,162 (100)
600 (3501,050)
2,095 (96.9)
600 (3501,000)
226 (10.4)
860 (5001,418)
210 (9.7)
300 (253600)
51 (2.4)
212 (197410)
4,632 (3,1426,469)
993 (24.4)
295 (7.2)
101 (2.5)
1,005 (24.7)
404 (9.9)
138 (3.4)
193 (4.7)
285 (7.0)
497 (12.2)
159 (3.9)
513 (23.7)
157 (7.3)
45 (2.1)
540 (25.0)
205 (9.5)
62 (2.9)
88 (4.1)
157 (7.3)
316 (14.6)
79 (3.6)
2011 (n=1,908)
64 (5374)
966 (50.6)
55 (2.9)
667 (35.0)
964 (50.5)
203 (10.6)
19 (1.0)
246 (12.9)
1,908 (100)
660 (330996)
1,780 (93.3)
660 (330993)
314 (16.5)
607 (4471,199)
277 (14.5)
292 (225518)
83 (4.4)
205 (184382)
3,655 (2,2005,712)
P Value*
0.004
0.27
<0.001
0.03
0.31
0.72
0.83
<0.001
0.05
<0.001
0.03
<0.001
0.30
<0.001
<0.001
480 (25.2)
138 (7.2)
56 (2.9)
465 (24.4)
199 (10.4)
76 (4.0)
105 (5.5)
128 (6.7)
181 (9.5)
80 (4.2)
Values are number (%) or median (IQR), unless indicated otherwise. For transfusion volumes, the values presented correspond to the median (IQR) volume
transfused in those who received a transfusion of the given product.
* P values compare medians and percentages between 2004 and 2011 using the Wilcoxon rank sum test and the chi-square test, respectively. Percentages
for all surgical specialties overall total <100% due to rounding; percentages for all surgical specialties for 2004 total >100% due to rounding.
ASA=American Society of Anesthesiologists; cryo=cryoprecipitate; FFP=fresh-frozen plasma; IQR=interquartile range; OB/GYN=obstetrics and gynecology.
summarizes their baseline characteristics. The median age

was 65 yr (IQR, 53 to 75 yr), and patients were marginally older in 2004 (66 vs. 64 yr; P=0.004). About half
(49.7% [2,023]) were men, which was comparable between
calendar years (P=0.27). The median intraoperative transfusion volume was 650ml (IQR, 330 to 1,025ml), which
was also comparable between calendar years (P=0.31). The
median intraoperative fluid balance was 4,180ml (IQR,
2,699 to 6,154ml) overall, which was significantly reduced
from 2004 (4,632ml [IQR, 3,142 to 6,469 ml]) to 2011
(3,655ml [IQR, 2,200 to 5,712 ml]) (P < 0.001). Most
patients had orthopedic surgeries (24.7% [n=1,005]) or
abdominal surgeries (24.4% [n=993]); neurologic surgery
(2.5% [n=101]) was the least common. The distribution of
patients among surgical specialties between 2004 and 2011
was significantly different (P < 0.001); most noticeably, there
was a 5.1% reduction in the number of patients undergoing
vascular surgeries (14.6% in 2004 vs. 9.5% in2011).
The electronic algorithm identified 510 (12.5%) of the
4,070 patients as having a high probability of TACO (321
in 2004 and 189 in 2011). Manual review showed that 176
(34.5%) of these 510 patients had experienced TACO (agreement by statistic, 0.46). One hundred nineteen TACO
cases occurred in 2004 (14 of which were considered to have
evidence of both TRALI and TACO), and 57 occurred in
2011 (8 of which were considered to have evidence of both
TRALI and TACO). Overall, TACO occurred within 6h of
the last intraoperative blood product transfusion at a rate of
4.3% (176 of 4,070 [95% CI, 3.7 to 5.0%]). The incidence
of TACO decreased significantly from 5.5% (95% CI, 4.6
to 6.5%) in 2004 to 3.0% (95% CI, 2.3 to 3.9%) in 2011
(P < 0.001). This reduction in the rate of TACO was not
fully explained by any of the stratified analyses. Specifically,
multiple logistic regression analyses revealed a statistically significant reduction in TACO from 2004 to 2011 even after
adjustment for variation in age (P < 0.001), sex (P < 0.001),
type of surgery (P=0.002), transfusion volume (P < 0.001),
and fluid balance (P=0.005; table3). Multivariable logistic
regression models evaluating product-specific rates of TACO
between 2004 and 2011 demonstrate significantly different
rates of TACO by both blood product (P < 0.001) and year
(P < 0.001) (table4). Overall, the highest incidence rate of
TACO occurs in patients receiving mixed blood products
(73 of 564 [12.9%]), closely followed by patients receiving
fresh-frozen plasma only (11 of 93 [11.8%]). In 2004, freshfrozen plasma had the highest incidence rate of TACO (7 of
34 [20.6%]), with an absolute reduction in number of events
between the 2 calendar years; this was surpassed by mixed
blood products in 2011 (35 of 321 [10.9%]) (table 4).
After adjustment for the reduction in TACO between
2004 and 2011, the difference in the rate of TACO between
men and women was no longer significant (4.8% [95% CI,
4.0 to 5.9%] vs. 3.8% [95% CI, 3.1 to 4.7%]; P=0.09).
However, the rates of TACO by type of surgical procedure remained significant (P < 0.001). Overall, vascular
(12.1%[95% CI, 9.5 to 15.2%]), transplant (8.8% [95%

CI, 5.6 to 13.7%]), and thoracic surgeries (7.2% [95% CI,
4.0 to 12.8%]) had the highest TACO rates, whereas obstetric and gynecologic surgical patients had the lowest rates
(1.4% [95% CI, 0.5 to 3.4%]). The rate of TACO increased
with increasing age (P < 0.001), with the rate in patients
aged 80 yr or older (7.4% [95% CI, 5.2 to 9.6%]) almost
quadruple that observed in patients aged 49 yr or younger
(2.0% [95% CI, 1.1 to 3.1%]), double that in patients aged
60 to 69 yr (4.2% [95% CI, 2.9 to 5.4%]), and 1.5 times
that in patients aged 70 to 79 yr (5.2% [95% CI, 3.8 to
6.6%]). We also observed an increased rate of TACO with
increasing amount of volume transfused (P < 0.001) and
increasing total fluid balance (P < 0.001).
Importantly, of the 176 cases of TACO identified during
the 2-yr study period, only 3 were included in the transfusion
medicine services database of potential transfusion reactions.
In all three cases, the investigation into a potential transfusion
reaction was unrelated to blood products administered in the
operating room. Furthermore, none were identified as a TACO
reaction. This once again highlights the lack of clinical recognition of TACO and the rationale for this study. However,
we did observe important differences in outcomes of patients
who developed TACO compared with their complication-free
transfused counterparts. Specifically, TACO cases had a median
postoperative ICU length of stay of 10.8 days (IQR, 4.2 to
30.8 days) compared with just 5.7 days (IQR, 2.7 to 14.8 days)
for transfused controls (P < 0.001). Similarly, for postoperative
hospital length of stay, TACO cases had a median of 11.6 days
(IQR, 7.3 to 22.3 days) compared with just 6.2 days (IQR, 4.2
to 9.3 days) for transfused controls (P < 0.001). Finally, evaluation of in-hospital mortality showed that, of the 176 patients
with TACO, 15 died (8.5%) compared with 93 of the 3,894
transfused controls (2.4%). This resulted in an odds ratio of
death for TACO cases compared with transfused controls of
3.8 (95% CI, 2.2 to 6.7) (P < 0.001).
Discussion
Our findings provide detailed descriptive data on the incidence rates and characteristics of patients with perioperative
TACO. The use of our recently developed NLP-based algorithm enabled us to capture what we believe to be a more representative sample of surgical patients experiencing TACO
and therefore to better define its true incidence in this population. This retrospective cohort study of 4,070 noncardiac
surgical patients confirms an overall rate of TACO of 4.3%
(95% CI, 3.7 to 5.0%). Although this overall incidence for
both years of the study is consistent with that reported previously in other patient populations,3,5 the incidence from
2004 to 2011 did decrease significantly (from 5.5 to 3.0%;
P < 0.001). Increased rates of TACO were associated with
surgical specialty, increased transfusion volume, and total
operative fluid balance (all P < 0.001). However, the declining incidence of TACO was not fully accounted for by any
patient or transfusion characteristics.
Perioperative TACO
Table 3. Incidence Rates of Perioperative Transfusion-associated Circulatory Overload by Subgroup Overall and by Index Year
Overall
Age, yr
49
5059
6069
7079
80
Sex
Male
Female
Surgical specialty
Abdominal
OB/GYN
Neurologic
Orthopedic
Spine
Thoracic
Transplant
Urology
Vascular
Other
Transfusion volume, ml
350
351700
7011,050
1,0511,400
1,401
Fluid balance, ml
2,000
2,0014,000
4,0016,000
6,0018,000
8,001
2004 No./Total (%)
2011 No./Total (%)
P Value
176/4,070 (4.3)
119/2,162 (5.5)
57/1,908 (3.0)
<0.001*
16/779 (2.1)
23/699 (3.3)
42/1,005 (4.2)
53/1,019 (5.2)
42/568 (7.4)
98/2,023 (4.8)
78/2,047 (3.8)
10/414 (2.4)
13/345 (3.8)
22/500 (4.4)
39/572 (6.8)
35/331 (10.6)
62/1,057 (5.9)
57/1,105 (5.2)
6/365 (1.6)
10/354 (2.8)
20/505 (4.0)
14/447 (3.1)
7/237 (3.0)
36/966 (3.7)
21/942 (2.2)
42/993 (4.2)
4/295 (1.4)
4/101 (4.0)
22/1,005 (2.2)
6/404 (1.5)
10/138 (7.2)
17/193 (8.8)
7/285 (2.5)
60/497 (12.1)
4/159 (2.5)
27/513 (5.3)
4/157 (2.5)
3/45 (6.7)
18/540 (3.3)
5/205 (2.4)
6/62 (9.7)
8/88 (9.1)
4/157 (2.5)
42/316 (13.3)
2/79 (2.5)
15/480 (3.1)
0/138 (0.0)
1/56 (1.8)
4/465 (0.9)
1/199 (2.3)
4/76 (5.3)
9/105 (8.6)
3/128 (2.3)
18/181 (9.9)
2/80 (2.5)
22/1,353 (1.6)
44/1,279 (3.4)
19/487 (3.9)
17/289 (5.9)
74/662 (11.2)
16/682 (2.3)
32/719 (4.5)
16/262 (6.1)
13/172 (7.6)
42/327 (12.8)
6/671 (0.9)
12/560 (2.1)
3/225 (1.3)
4/117 (3.4)
32/335 (9.6)
13/601 (2.2)
32/1,325 (2.4)
43/1,065 (4.0)
28/558 (5.0)
60/521 (11.5)
8/186 (4.3)
23/682 (3.4)
31/656 (4.7)
20/340 (5.9)
37/298 (12.4)
5/415 (1.2)
9/643 (1.4)
12/409 (2.9)
8/218 (3.7)
23/223 (10.3)
<0.001
<0.001*
0.09
<0.001*
<0.001
0.001*
<0.001
<0.001*
<0.001
0.005*
Multiple independent logistic regression analyses were performed for each patient characteristic to evaluate the difference in rate of transfusion-associated
circulatory overload.
* P value for the difference in the rate of transfusion-associated circulatory overload by year (from 2004 to 2011), after adjustment for the variation in rate by
each characteristic. P value for the difference in the rate of transfusion-associated circulatory overload by characteristic, after adjustment for the variation
in rate by year (from 2004 to 2011).
OB/GYN=obstetrics and gynecology.
Table 4. Product-specific Incidence Rates of Perioperative Transfusion-associated Circulatory Overload Overall and by Index Year
Erythrocytes only
FFP only
Platelets only
Mixed products
2004 No./Total (%)
2011 No./Total (%)
90/3,322 (2.7)
11/93 (11.8)
2/89 (2.2)
73/564 (12.9)
74/1,853 (4.0)
7/34 (20.6)
0/32 (0.0)
38/243 (15.6)
16/1,469 (1.1)
4/59 (6.8)
2/57 (3.5)
35/321 (10.9)
Multivariable logistic regression analyses demonstrated that the frequency of transfusion-associated circulatory overload was related to both blood product
group (P < 0.001) and calendar year (P < 0.001). Two patients were excluded from this secondary analysis (one received only fresh whole blood and one
received only cryoprecipitate; neither experienced transfusion-associated circulatory overload).
FFP=fresh-frozen plasma.
To better understand the epidemiology of TACO in

patients cared for by anesthesia providers, we studied only
those surgical patients who were transfused in the operating room. Our primary reasons for selecting this population
are that a large proportion of blood product transfusions

take place in the operating room environment18 and that a
substantial proportion of those who experience TACO will
do so after intraoperative transfusion (70% of all TACO
episodes).7 Notwithstanding the findings of a few studies,2,16

most investigations to date have focused on the epidemiology of TACO in the critically ill.3,5 How well these results
apply to most patients encountered in the operating room
is unclear. Historically, incidence rates have ranged from less
than 1 to 11%.3,5 One of the earliest studies investigating
TACO in surgical populations in the mid-1990s found it
to be a frequent and serious event in orthopedic surgery,
occurring at a rate of 1.05%.2 A decade later, the same
authors reported rates of TACO ranging from 1 to 8% in
patients undergoing hip or knee surgery.16 More recently, a
2011 report from a single-center prospective cohort study
documented an incidence of 6% in the ICU.3
The reasons for this variation in TACO incidence rates
are likely multifactorial, explained in part by differing populations, definitions of TACO, case adjudication strategies,
and study designs. However, none of these studies addressed
the specific objective of our studyto define the incidence
of TACO in noncardiac surgical populations. We believe
that the use of our novel, highly sensitive NLP-based screening algorithm, combined with the detailed manual review
of a large number of transfused surgical cases from 2 different calendar years, produced results that likely provide a
more accurate reflection of the true incidence of TACO and
are thus also broadly generalizable to noncardiac surgical
populations.
Interestingly, in the current study, we observed a significant reduction in the rate of TACO between 2004 and 2011.
Moreover, this decline in incidence was not fully explained
by any of the patient and transfusion characteristics evaluated in this study. Although the introduction of male-only
plasma has been accepted as resulting in a reduced rate of
TRALI during this time period,18 until recently no hypotheses existed to explain similar reductions in the incidence of
TACO as perhaps resulting from changes in blood product
procurement strategies. The pathophysiologic mechanisms
underlying TACO have largely been considered to be due to
volume overload. In this context, there is little biologic rationale to explain why the exclusion of female plasma donors
would affect the rates of TACO. Interestingly, additional
changes in blood management have also taken place during this interval. For example, our institution implemented
universal leukoreduction in 2005. Blumberg et al.19 recently
described a 49% reduction in the rate of TACO after universal leukoreduction. This finding was largely attributed
to a reduction in TACO related to erythrocyte units rather
than other components. Blumberg et al.19 hypothesized
that TACO may therefore not be solely explained by volume overload but instead may also have an inflammatory/
immunologic component. They highlighted the possible
role of leukocyte-derived microparticles and other mediators
of capillary leak that accumulate during storage. Although
data from this observational study cannot be used to imply
causality, our findings suggest that this potential mechanism
may warrant further investigation. Future studies should be
designed and powered to evaluate the potential role of leukoreduction while accounting for other potential confounding
variables and changes in clinical practice that have occurred
during the same period.
Although the sensitivity of our electronic screening algorithm was previously demonstrated to be 100%,14 the decision to limit manual review to patients who screened as high
probability for TACO with the electronic algorithm may
have resulted in missed cases. For example, a patient with
three or more U.S. Centers for Disease Control and Prevention criteria for TACO may have been missed because of the
absence of infiltrates on chest radiographs or the hypoxemia
criteria used in our screening algorithm. This in addition to
our exclusion of patients with preexisting respiratory failure and/or pulmonary infiltrates on chest radiographs may
have resulted in an underestimation of the true incidence of
TACO. Nonetheless, although our estimates may be conservative, we believe that the strategies implemented for TACO
adjudication in this investigation likely remain a significant
improvement upon past efforts aiming to detect TACO,
which have primarily relied on manual detection alone.7 In
addition, we would also emphasize that the algorithm did
not miss any cases of TACO that were reported to the transfusion medicine service during the defined study period.
In addition to the limitations noted above, we further
acknowledge that our study population was derived from a
single tertiary care referral center. It is possible that referral bias may affect the reported incidence of TACO because
we selected patients with a higher prevalence of comorbid
conditions than that of surgical populations seen in nonreferral centers. Similarly, results obtained in the current
study cannot be generalized to patients receiving nongeneral
anesthesia, as this subset of patients was not included in our
study cohort. Indeed, external validation of our findings will
require further study.
Finally, this study was designed as a descriptive cohort
study with the aims of more accurately defining the incidence of TACO after noncardiac surgery and describing it
in the context of a limited number of patient and transfusion characteristics. The investigation was not designed to
extensively evaluate specific risk factors for TACO or its
attributable burden. Thus, although associations were noted
between such characteristics and the development of TACO,
more robust and definitive evaluations of TACO risk factors will require future study. Similarly, although associations
were noted between TACO and adverse patient-important
outcomes such as prolonged ICU and hospital length of stay
and increased mortality, these findings were not adjusted for
severity of illness of other potentially confounding variables.
These findings are presented primarily to highlight the fact
that cases of TACO identified in this study but missed by the
clinical practice do not appear to be clinically insignificant.
However, as with the evaluation of risk factors for TACO,
a more comprehensive understanding of the true impact of
TACO on patient-important outcomes as compared with
Perioperative TACO
outcomes for complication-free transfused patients will

require additional study, with fully adjusted analyses.
In conclusion, there remain significant barriers to the accurate characterization of the epidemiology of TACO, particularly in surgical populations. Herein, we demonstrate a recent
estimate of the incidence of TACO in surgical populations to
be 3.0%, with a significant reduction in the rate of TACO
from 2004 and 2011, that was not fully explained by either
patient or transfusion characteristics evaluated in this study.
This robust characterization of the epidemiology of TACO
in surgical populations should (1) heighten the awareness of
transfusion risks among perioperative healthcare providers,
(2) facilitate improved decision making regarding transfusion strategies in at-risk patients, and (3) enhance the timely
implementation of appropriate treatment interventions for
this serious transfusion-related pulmonary complication.
Future studies will more completely define the true attributable burden, the underlying mechanisms, and the risk factors
for TACO. Furthermore, with improved understanding of the
risk factors underlying TACO, we would expect to be able
to refine the electronic algorithm used to screen patients in
this study, developing a real-time prediction and/or prevention model for TACO that may help to mitigate its incidence.
Acknowledgments
The authors thank Lisa M. Vasgaard, M.T., S.B.B. (A.S.C.P.),
and Brenda J. Bendix, M.L.S. (A.S.C.P.) S.B.B., Division of
Transfusion Medicine, Mayo Clinic, Rochester, Minnesota,
for their assistance in gathering data related to reported
cases of transfusion-associated circulatory overload (TACO).
This study was conducted with funds from the
Mayo Clinic Center for Translational Science Activities
High-Impact Pilot and Feasibility Award (HIPFA-2012; grant
no. 94164003, to Dr. Pathak) and by funds from the Mayo
Clinic Critical Care Integrated Multidisciplinary Practice,
Rochester, Minnesota.
Competing Interests
Correspondence
Address correspondence to Dr. Kor: Department of Anesthesiology, Mayo Clinic, 200 First St SW, Rochester, Minnesota 55905. kor.daryl@mayo.edu. This article may be accessed
for personal use at no charge through the Journal Web site,
www.anesthesiology.org.
References
1. Drummond R: Transfusion reactions and fatalities due to circulatory overloading. Br Med J 1943; 2:31922
2. Popovsky MA, Audet AM, Andrzejewski C Jr: Transfusionassociated circulatory overload in orthopedic surgery patients:
A multi-institutional study. Immunohematology 1996; 12:879
3. Li G, Rachmale S, Kojicic M, Shahjehan K, Malinchoc M,
Kor DJ, Gajic O: Incidence and transfusion risk factors for
transfusion-associated circulatory overload among medical

intensive care unit patients. Transfusion 2011; 51:33843
4. Popovsky MA: Transfusion and the lung: Circulatory overload and acute lung injury. Vox Sang 2004; 87(suppl 2):625
5. Rana R, Fernndez-Prez ER, Khan SA, Rana S, Winters JL,
Lesnick TG, Moore SB, Gajic O: Transfusion-related acute
lung injury and pulmonary edema in critically ill patients: A
retrospective study. Transfusion 2006; 46:147883
6. Li G, Kojicic M, Reriani MK, Fernndez Prez ER, Thakur L,
Kashyap R, Van Buskirk CM, Gajic O: Long-term survival and
quality of life after transfusion-associated pulmonary edema
in critically ill medical patients. Chest 2010; 137:7839
7. Clifford L, Singh A, Wilson GA, Toy P, Gajic O, Malinchoc
M, Herasevich V, Pathak J, Kor DJ: Electronic health record
surveillance algorithms facilitate the detection of transfusion-related pulmonary complications. Transfusion 2013;
53:120516
8. Narick C, Triulzi DJ, Yazer MH: Transfusion-associated circulatory overload after plasma transfusion. Transfusion 2012;
52:1605
9. Herasevich V, Kor DJ, Li M, Pickering BW: ICU data mart: A
non-iT approach. A team of clinicians, researchers and informatics personnel at the Mayo Clinic have taken a homegrown
approach to building an ICU data mart. Healthc Inform 2011;
28:42, 445
10. von Elm E, Altman DG, Egger M, Pocock SJ, Gtzsche PC,
Vandenbroucke JP; STROBE Initiative: The Strengthening
the Reporting of Observational Studies in Epidemiology
(STROBE) statement: Guidelines for reporting observational
studies. Ann Intern Med 2007; 147:5737
11. Alsara A, Warner DO, Li G, Herasevich V, Gajic O, Kor DJ:
Derivation and validation of automated electronic search
strategies to identify pertinent risk factors for postoperative
acute lung injury. Mayo Clin Proc 2011; 86:3828
12. Chute CG, Beck SA, Fisk TB, Mohr DN: The Enterprise Data
Trust at Mayo Clinic: A semantically integrated warehouse of
biomedical data. J Am Med Inform Assoc 2010; 17:1315
13. Schmickl CN, Li M, Li G, Wetzstein MM, Herasevich V, Gajic O,
Benzo RP: The accuracy and efficiency of electronic screening for recruitment into a clinical trial on COPD. Respir Med
2011; 105:15016
14. Clifford L, Wilson GA, Gajic O, Toy P, Herasevich V, Murphy
S, Pathak J, Kor DJ: Natural language processing of chest
radiograph reports improves the identification of transfusion-related pulmonary complications (abstract). Am J Respir
Crit Care Med 2013; 187:A2218
15. Kleinman S, Caulfield T, Chan P, Davenport R, McFarland

J, McPhedran S, Meade M, Morrison D, Pinsent T, Robillard
P, Slinger P: Toward an understanding of transfusionrelated acute lung injury: Statement of a consensus panel.
16. Popovsky MA: Pulmonary consequences of transfusion:

TRALI and TACO. Transfus Apher Sci 2006; 34:2434
17. Arinsburg SA, Skerrett DL, Karp JK, Ness PM, Jhang J,
Padmanabhan A, Gibble J, Schwartz J, King KE, Cushing
MM: Conversion to low transfusion-related acute lung injury
(TRALI)-risk plasma significantly reduces TRALI. Transfusion
2012; 52:94652
18. Van Dijk PM, Kleine JW: The transfusion reaction in anesthesiological practice. Acta Anaesthesiol Belg 1976; 27:24754
19. Blumberg N, Heal JM, Gettings KF, Phipps RP, Masel D,
Refaai MA, Kirkley SA, Fialkow LB: An association between
decreased cardiopulmonary complications (transfusionrelated acute lung injury and transfusion-associated circulatory overload) and implementation of universal
leukoreduction of blood transfusions. Transfusion 2010;
50:273844
Transfusion Requirements in Surgical Oncology Patients

A Prospective, Randomized Controlled Trial
Juliano Pinheiro de Almeida, M.D., Jean-Louis Vincent, M.D., Ph.D.,
Filomena Regina Barbosa Gomes Galas, M.D., Ph.D., Elisangela Pinto Marinho de Almeida, M.D.,
Julia T. Fukushima, M.Sc., Eduardo A. Osawa, M.D., Fabricio Bergamin, M.D., Clarice Lee Park, M.D.,
Rosana Ely Nakamura, M.D., Silvia M. R. Fonseca, M.D., Guilherme Cutait, M.D.,
Joseane Inacio Alves, R.N., Mellik Bazan, P.T., Silvia Vieira, R.N., Ana C. Vieira Sandrini, L.D.N.,
Henrique Palomba, M.D., Ph.D., Ulysses Ribeiro, Jr., M.D., Ph.D., Alexandre Crippa, M.D.,
Marcos Dalloglio, M.D., Ph.D., Maria del Pilar Estevez Diz, M.D., Ph.D., Roberto Kalil Filho, M.D., Ph.D.,
Jose Otavio Costa Auler, Jr., M.D., Ph.D., Andrew Rhodes, M.B., B.S.,
Ludhmila Abrahao Hajjar, M.D., Ph.D.
This article has been selected for the Anesthesiology CME Program. Learning objectives
and disclosure and ordering information can be found in the CME section at the front
of this issue.
ABSTRACT
Background: Several studies have indicated that a restrictive erythrocyte transfusion strategy is as safe as a liberal one in critically ill
patients, but there is no clear evidence to support the superiority of any perioperative transfusion strategy in patients with cancer.
Methods: In a randomized, controlled, parallel-group, double-blind (patients and outcome assessors) superiority trial in the intensive
care unit of a tertiary oncology hospital, the authors evaluated whether a restrictive strategy of erythrocyte transfusion (transfusion
when hemoglobin concentration <7g/dl) was superior to a liberal one (transfusion when hemoglobin concentration <9g/dl) for
reducing mortality and severe clinical complications among patients having major cancer surgery. All adult patients with cancer having major abdominal surgery who required postoperative intensive care were included and randomly allocated to treatment with the
liberal or the restrictive erythrocyte transfusion strategy. The primary outcome was a composite endpoint of mortality and morbidity.
Results: A total of 198 patients were included as follows: 101 in the restrictive group and 97 in the liberal group. The primary composite
endpoint occurred in 19.6% (95% CI, 12.9 to 28.6%) of patients in the liberal-strategy group and in 35.6% (27.0 to 45.4%) of patients
in the restrictive-strategy group (P = 0.012). Compared with the restrictive strategy, the liberal transfusion strategy was associated with an
absolute risk reduction for the composite outcome of 16% (3.8 to 28.2%) and a number needed to treat of 6.2 (3.5 to 26.5).
Conclusion: A liberal erythrocyte transfusion strategy with a hemoglobin trigger of 9g/dl was associated with fewer major
postoperative complications in patients having major cancer surgery compared with a restrictive strategy. (Anesthesiology
2015; 122:29-38)
OR many patients with solid tumors, surgery remains

the mainstay of therapy. For these patients, a complication-free operative procedure is vital to maximize the chances
that oncological treatment is successful. The care of patients
having abdominal oncologic surgery is challenging because
of the unusually long duration of the surgical procedures,
the significant fluid and blood losses that can occur, and
the inherent increased operative risk related to the cancer
diagnosis.13
Perioperative anemia occurs in 25 to 75% of patients
having surgery for cancer, largely because of blood loss
during long and complex surgical procedures, but poor

It remains unknown whether a liberal or restrictive transfusion
strategy is superior in patients having major cancer surgery

In 198 patients randomly assigned to erythrocyte transfusions
at a hemoglobin concentration of 7 or 9g/dl
Major complications were nearly twice as common in patients
managed with the restrictive approach as in those managed
with the liberal approach (36 vs. 20%)
This study supports a more liberal transfusion strategy in major
cancer surgery
This article is featured in This Month in Anesthesiology, page 1A. Corresponding article on page 3.
Submitted for publication January 9, 2014. Accepted for publication August 8, 2014. From the Surgical Intensive Care Unit and Department of Anesthesiology, Cancer Institute, Hospital das Clinicas da Faculdade de Medicina da Universidade de Sao Paulo, Sao Paulo, Brazil
( J.P.d.A., F.R.B.G.G., E.P.M.d.A., J.T.F., E.A.O., F.B., C.L.P., R.E.N., S.M.R.F., J.I.A., M.B., S.V., A.C.V.S., H.P., R.K.F., J.O.C.A., L.A.H.); Department
of Intensive Care, Erasme Hospital, Universit Libre de Bruxelles, Brussels, Belgium ( J.-L.V.); Department of Intensive Care Medicine, St.
Georges Healthcare NHS Trust, London, United Kingdom (A.R.); Department of Surgery, Cancer Institute, Hospital das Clinicas da Faculdade
de Medicina da Universidade de Sao Paulo, Sao Paulo, Brazil (G.C., U.R., A.C., M.D.); and Department of Oncology, Cancer Institute, Hospital
das Clinicas da Faculdade de Medicina da Universidade de Sao Paulo, Sao Paulo, Brazil (M.d.P.E.D.).
January 2015
Transfusion Requirements in Surgical Oncology
nutritional condition, tumor stage, cancer-related anemia,

and previous chemotherapy may also contribute.4 To treat
or prevent anemia, these patients often receive allogeneic
erythrocyte transfusions.57 Recent evidence from critically
ill patients has led to renewed consideration of the potential benefits and risks of erythrocyte transfusion. In patients
with cancer, in particular, immunomodulation and the
inflammatory consequences of transfusion may outweigh
the possible advantages of improved oxygen delivery and
tissue perfusion.79 The decision to transfuse these patients
should, therefore, carefully take into account the potential risks of anemia versus known adverse effects of blood
transfusion.
Several previous studies have indicated that a restrictive
erythrocyte transfusion strategy is as safe and effective as a
liberal one in critically ill patients.10 In addition, a restrictive
strategy reduces costs and results in less exposure to allogeneic blood.8 However, patients with cancer are not included
in most trials investigating this issue.
The aim of this study was, therefore, to compare mortality rates and severe clinical complications in high-risk
abdominal oncological surgery patients managed using a
restrictive or a liberal erythrocyte transfusion strategy. Our
hypothesis was that a restrictive strategy of erythrocyte transfusion would be associated with improved outcomes in these
patients with cancer having major surgery.

This study was approved by the Ethics and Research Committee of the University of Sao Paulo (Comit de tica
em Pesquisa da Faculdade de Medicina da Universidade
de So Paulo, number 002/11). Patients enrolled in the
study were admitted to the intensive care unit (ICU) of a
tertiary oncology university hospital in Sao Paulo, Brazil,
between January 2012 and July 2012. Written informed
consent was obtained from all the participants in the study.
The study protocol was registered at ClinicalTrials.gov
(NCT01502215).
All adult patients who had a major surgical procedure
for abdominal cancer and required postoperative care in
the ICU because of physiological instability and had an
expected ICU stay of more than 24h were included. We
excluded all patients with any of the following characteristics: age less than 18 yr, hematological malignancy, a
Karnofsky score less than 50, preexisting anemia (defined
as a preoperative hemoglobin concentration <9g/dl), preexisting thrombocytopenia (defined as a platelet count
<50,000/mm3), preexisting coagulopathy (defined as a
prothrombin time >14.8 s) or anticoagulation therapy,
active or uncontrolled bleeding, expected death within
24h of ICU admission, end-stage renal failure requiring
renal replacement therapy, pregnancy, a do-not-resuscitate
order, inability to receive transfusion of blood components, or refusal to participate in the study.
Major surgical procedures for abdominal cancer included

esophagectomy, gastrectomy, gastroduodenopancreatectomy
(Whipple procedure), hepatectomy or bile duct resection,
radical cystectomy, partial or total colectomy, retroperitoneal
tumor resection, cytoreductive surgery with heated intraperitoneal chemotherapy, radical hysterectomy, or emergency
laparotomy.
Randomization and Masking
All patients were assessed for eligibility at the time of ICU
admission by a member of the medical staff. After signing
informed consent, patients were randomized in a 1:1 ratio
to one of two erythrocyte strategies: a restrictive erythrocyte transfusion strategy or a liberal erythrocyte transfusion
strategy. After consent, the medical staff contacted the study
randomization center to register the patient and to be told
which group the patient was allocated to. To avoid loss of
concealment, the group to which the patient was allocated
could only be accessed after registration in the study randomization center. Allocation numbers were derived from a
random number table prepared by the chief statistician and
were placed in opaque envelopes and opened sequentially to
determine the treatment group of the patient. The patients
and the study investigators who classified outcomes and
those who conducted the follow-up telephone assessments
were blinded to the study-group assignments and had no
access to transfusion data.
Treatment Protocol
The patients in the restrictive and liberal erythrocyte transfusion strategy groups received one erythrocyte unit each
time their hemoglobin concentration decreased to less than
7 or 9g/dl, respectively, during their ICU stay. Physicians
were instructed to administer transfusions 1 unit at a time
and to measure hemoglobin concentration after each transfused unit. In both groups, no further units were given if
the goal hemoglobin concentration was obtained (7g/dl
for the restrictive strategy and 9g/dl for the liberal strategy). Hemoglobin levels were measured in all patients at
least twice a day while patients were in the ICU. After ICU
discharge, hemoglobin levels were measured as clinically
indicated.
If the treating clinicians transfused an additional erythrocyte unit outside the protocol, it was recorded as a protocol deviation. After ICU discharge, the decision to transfuse
was left to the discretion of the physician in charge of the
patient clinical care. During the 30-day follow-up period, if
a patient returned to the ICU for any reason, the allocated
transfusion strategy was maintained. An intention-to-treat
analysis was performed and considered the patients in their
originally assigned groups. The erythrocytes were leukodepleted and stored in a citrate solution. The erythrocyte unit
volume ranged from 250 to 350ml, with a hematocrit of
70%. The policy of our hospitals blood bank is to use blood
that has been stored for less than 35 days.
Pinheiro de Almeida et al.
During the ICU stay, all other monitoring and treatment

was given according to established ICU protocols.
Data Collection
After randomization, we recorded baseline demographic
characteristics, the American Society of Anesthesiology
physical status classification, and specific data on the type
and status of the neoplasm and its previous treatment. The
values of two scores used to assess the degree of functional
impairment, the Karnofsky Performance Status scale11
(from 0 [most impaired] to 100 [least impaired]) and the
Eastern Cooperative Oncology Group scale12 (from 0 [most
impaired] to 5 [least impaired]), were also recorded. We
also evaluated the Charlson comorbidity index, a score used
to predict outcome according to the weighted presence of
various comorbid conditions, which can also be adjusted by
patient age (higher score, higher risk of death).13
We also collected data related to characteristics of the surgical procedure, including type of surgery, amount and types
of fluids, blood transfusions, and laboratory data during the
intraoperative period.
During the ICU stay, all patients were assessed daily by a
team of three blinded investigators and clinical and laboratory data were recorded. The lowest hemoglobin concentration and the use of erythrocyte transfusions were recorded
every day. Leukocyte count, C-reactive protein, platelet
count, prothrombin time and activated partial thromboplastin time, serum creatinine, troponin, creatine kinase-MB,
bilirubin, arterial lactate concentration, and central venous
oxygen saturation were collected daily. A Simplified Acute
Physiology Score 3 was calculated using the worst value
within the first 24h of the ICU stay. An electrocardiogram
was performed daily during the ICU stay.
During the hospital stay, data were collected regarding hemoglobin concentrations and the use of erythrocyte
transfusions.
During a 30-day follow-up period, we evaluated the
incidence of stroke (including transient ischemic attack),
acute myocardial infarction, pulmonary embolism, congestive heart failure, cardiac arrest, septic shock, acute
kidney injury (AKI), renal replacement therapy, acute
respiratory distress syndrome (ARDS), mesenteric ischemia, peripheral ischemia, reoperation, and infectious
complications. Acute myocardial infarction was defined
as at least one of the following findings associated with
clinical symptoms suggestive of myocardial ischemia: an
increase or decrease in cardiac troponin I, with at least
one value above the 99th percentile of the upper reference
limit; electrocardiographic changes, such as new Q waves,
ST-elevation, or a new left branch block; or image-based
evidence of new loss of viable myocardium.14 Infectious
complications included new infection, sepsis, severe sepsis, and septic shock, which was defined using standard
criteria15; wound infection with positive cultures; and
pneumonia, including signs of sepsis associated with a
new, persistent, or progressive lung infiltrate on a chest

radiograph and purulent endotracheal secretions, with a
Gram stain showing more than 25 neutrophils and fewer
than 10 epithelial cells per field.
Renal function was evaluated daily using the risk, injury,
failure, loss, and end-stage classification16; patients who were
classified as injury or more were defined as having AKI. Mesenteric arterial ischemia was defined as an occlusive or nonocclusive impairment of intestinal blood flow leading to bowel
ischemia diagnosed by arteriography or surgery. Peripheral
vascular ischemia was defined as a sudden decrease in limb
perfusion characterized by an absence of arterial pulses, pale
extremities, cyanosis or ischemic skin lesions, and absence
of arterial blood flow on Doppler examination. ARDS was
defined by the usual criteria.17 Stroke was characterized by
a new focal deficit with a compatible image on computed
tomography.
Primary Outcome Measures
The primary outcome was a composite endpoint of death
from all causes or severe clinical complications within 30
days after the randomization. Severe clinical complications
included major cardiovascular complications (defined as
acute myocardial infarction, pulmonary embolism, congestive heart failure, cardiac arrest, acute mesenteric ischemia,
stroke [including transient ischemic attack], and acute
peripheral vascular ischemia), septic shock, AKI requiring
renal replacement therapy, ARDS, and reoperation.
Fig. 1. Study flow chart. ITT = intention to treat.

Table 1. Baseline Demographic Data and Preoperative Characteristics of Patients

Variables
Patients, No.
Age, mean (SD), yr
Male (%)
Body mass index, mean (SD)
Hypertension (%)
Diabetes (%)
Dyslipidemia (%)
Baseline creatinine >1.5mg/dl (%)
Tobacco use (%)
COPD (%)
Peripheral arterial disease (%)
Cerebrovascular disease (%)
Congestive heart failure (%)
Arterial coronary disease (%)
History of dementia (%)
Age-adjusted Charlson comorbidity index
American Society of Anesthesiologists risk score (%)
II
III
IV
V
Nutritional risk screening (2002) score
1
2
3
4
5
6
Cachexia score
1
2
3
Hemoglobin, g/dl (SD)
Serum albumin, g/dl (SD)
Simplified Acute Physiology Score 3
C-reactive protein at ICU admission, mg/l
Liberal Strategy
Restrictive Strategy
97
64 (14)
55 (56.7)
25 (5)
51 (53.1)
20 (20.6)
10 (10.3)
11 (11.3)
42 (43.3)
5 (5.2)
1 (1.0)
2 (2.1)
3 (3.1)
6 (6.2)
1 (1.0)
6 (49)
101
64 (12)
55 (54.5)
25 (5)
46 (45.5)
26 (25.7)
10 (9.9)
8 (7.9)
41 (40.6)
9 (8.9)
1 (1.0)
8 (7.9)
6 (5.9)
8 (7.9)
1 (1.0)
7 (59)
57 (58.8)
30 (30.9)
9 (9.3)
1 (1.0)
67 (66.3)
25 (24.8)
8 (7.9)
1 (1.0)
20 (20.6)
21 (21.6)
21 (21.6)
23 (23.7)
8 (8.2)
4 (4.1)
18 (17.8)
25 (24.8)
33 (32.7)
17 (16.8)
3 (3.0)
5 (5.0)
47 (49.0)
11 (11.5)
38 (39.6)
12.4 (1.7)
2.6 (0.6)
34 (2945)
79 (45116)
52 (51.5)
5 (5.0)
44 (43.6)
12.6 (1.8)
2.6 (0.5)
37 (3146)
73 (49117)
COPD = chronic obstructive pulmonary disease; ICU = intensive care unit.
Secondary Outcome Measures

Secondary outcomes included the 30-day incidence of
infection, development of AKI, need and duration of
mechanical ventilation, duration of vasopressor therapy,
ICU readmission rate, ICU and hospital lengths of stay,
and 60-day mortality. Other secondary outcomes included
need for erythrocyte transfusion and the number of units
transfused.
Statistical Analysis
This was a superiority trial, and we predicted a 40% incidence of the primary outcome based on a previous study
showing rates of 40 to 60%.18 A minimum enrollment of
164 patients was required to provide 80% statistical power
to detect a difference of 20% with a two-sided = 0.05.
We added 20% to the sample size to compensate for subject
attrition, yielding a final requirement for 197 patients.
We compared the baseline characteristics, follow-up measures, and clinical outcomes on an intention-to-treat basis
according to randomized study-group assignment. Continuous variables were compared using a t test or the MannWhitney U test, and categorical variables were compared using
Pearson chi-square test, Fisher exact test, or a likelihood ratio
test. We compared hemoglobin levels during the ICU stay
between groups using a mixed-design ANOVA model because
many patients died or were discharged from the ICU at different times. The model was constructed using the lowest daily
average hemoglobin concentrations during the ICU stay.
Results are expressed as means with SDs or medians
with interquartile ranges (IQRs). We calculated unadjusted
30-day KaplanMeier survival estimates, dividing patients
according to the transfusion strategy and the number of
transfused erythrocyte units. A two-sided P value less than
0.05 was considered to be statistically significant. The
Table 2. Characteristics Related to the Underlying Malignancies of Patients and Types of Surgical Procedure
Variables (%)
Liberal Strategy (N = 97)
Type of tumor
Upper gastrointestinal
Lower gastrointestinal
Pancreas
Liver and biliary tract
Urogenital
Other
Extent of cancer
Localized
Metastatic
Karnofsky performance status
ECOG performance status
0
1
2
3
4
Chemotherapy in the 4wk before ICU admission
Type of procedure
Esophagectomy
Gastrectomy
Gastroduodenopancreatectomy
Liver resection
Biliary duct resection
Colectomy
Peritonectomy with intraperitoneal chemotherapy
Abdominoperineal rectal amputation
Resection of retroperitoneal tumor
Emergency laparotomy
Pelvic exenteration
Radical cystectomy
Radical hysterectomy
Duration of surgery (min)
Type of anesthesia
General
Spinal
Spinal + general
Intraoperative fluid (l)
Restrictive Strategy (N = 101)
20 (20.6)
33 (34.0)
6 (6.2)
1 (1.0)
27 (27.8)
11 (11.3)
20 (19.8)
45 (44.6)
5 (5.0)
3 (3.0)
22 (21.8)
7 (6.9)
65 (67.0)
32 (33.0)
90 (80100)
62 (61.4)
39 (38.6)
90 (80100)
38 (40.0)
41 (43.2)
7 (7.4)
7 (7.4)
2 (2.1)
6 (6.2)
45 (45.0)
41 (41.0)
10 (10.0)
4 (4.0)
0 (0)
8 (7.9)
5 (5.2)
12 (12.4)
3 (3.1)
9 (9.3)
2 (2.1)
23 (23.7)
2 (2.1)
2 (2.1)
3 (3.1)
9 (9.3)
1 (1.0)
17 (17.5)
9 (9.3)
355 (250493)
7 (6.9)
8 (7.9)
7 (6.9)
13 (12.9)
2 (2.0)
18 (17.8)
7 (6.9)
7 (6.9)
2 (2.0)
13 (12.9)
1 (1.0)
9 (8.9)
7 (6.9)
323 (188476)
33 (34.0)
1 (1.0)
63 (64.9)
4.5 (3.56.0)
32 (31.7)
2 (2.0)
67 (66.3)
4.5 (3.56.5)
ECOG = Eastern Cooperative Oncology Group; ICU = intensive care unit.
statistical analyses were performed using SPSS version 18.0

(SPSS Inc., Chicago, IL).
Results
Study Population
A total of 1,521 patients were screened for eligibility and 234
met the inclusion criteria (fig.1). After exclusions for medical reasons or lack of consent, 198 patients were enrolled in
the study; of whom, 97 were randomized to the liberal group
and 101 to the restrictive group. All patients completed the
study and were followed up for outcome criteria. Baseline
characteristics were well balanced between the study groups
(table1). The majority of the patients had a good performance status and localized disease and had an elective surgical procedure for gastrointestinal cancer (table2).
Primary Outcome
The primary composite endpoint at 30-daysall-cause mortality, cardiovascular complication, ARDS, AKI requiring
renal replacement therapy, septic shock, or reoperation
occurred in 19 patients (19.6%) in the liberal-strategy group
and in 36 patients (35.6%) in the restrictive-strategy group
(P = 0.012). This represents an absolute risk reduction for the
liberal strategy of 16% (95% CI, 3.8 to 28.2) and a number
needed to treat of 6.2 (95% CI, 3.5 to 26.5) to avoid the
composite outcome.
Secondary Outcomes
In total, 31 patients (15.7%) died during the 30-day follow-up. The 30-day mortality rate was lower in the liberalstrategy group than in the restrictive group (8 [8.2%] vs. 23
Fig. 2. KaplanMeier curves showing the probability of 30day survival in patients randomized to a restrictive strategy
of erythrocyte transfusion (transfusion when hemoglobin concentration <7g/dl) and those randomized to a liberal strategy
(transfusion when hemoglobin concentration <9g/dl). The P
value was calculated with the use of the log-rank test.
[22.8%]; P = 0.005; fig.2). At 60 days, we also observed a

lower mortality rate in the liberal group compared with the
restrictive group (11 [11.3%] vs. 24 [23.8%]; P = 0.022). At
30 days, the most frequent cause of death was septic shock
and multisystem organ failure (24 patients), followed by
noninfectious circulatory shock (3 patients), cancer-related
complication (3 patients), and respiratory failure (1 patient).
At 30 days, the number of patients with major cardiovascular events was lower in the liberal than in the restrictive
group (5 [5.2%] vs. 14 [13.9%]; P = 0.038). There were no
differences between groups in the incidence of each major
cardiovascular complication separately (table3).
There was a higher incidence of intraabdominal infection
in the restrictive group than the liberal group (15 [14.9%]
vs. 5 [5.2%]; P = 0.024), but no differences in the incidence
of other infections. There was no difference in the incidence
of septic shock in the restrictive group compared with the
liberal group (13.4 vs. 21.8%; P = 0.122). There were no
statistically significant differences between the liberal and
restrictive-strategy groups in the occurrence of ARDS,
incidence of AKI or dialysis, requirement and duration of
mechanical ventilation, incidence of reoperation, use of vasopressors, and lengths of time in ICU or hospital (table3).
Hemoglobin Concentrations and Transfusion
Hemoglobin concentrations (mean, 11.0g/dl [1.6] vs.
11.2g/dl [1.8]; P = 0.46) and the number of patients transfused before randomization were similar in both groups
(table4). The average hemoglobin concentration before
transfusion was higher in the liberal-strategy group than
in the restrictive-strategy group (7.9g/dl [0.5] vs. 6.8g/dl
[0.5]; P < 0.001; table 4). Hemoglobin concentrations were
statistically significantly higher in the liberal-strategy group
than in the restrictive-strategy group during the ICU stay
(P < 0.001; fig.3).
More patients received an erythrocyte transfusion in the

liberal-strategy group than in the restrictive-strategy group during the ICU stay (42.3 vs. 20.8%; P = 0.005). Among transfused patients, the liberal-strategy group received a median of 2
erythrocyte units (IQR, 1 to 3), whereas the restrictive-strategy
group received a median of 1 unit (IQR, 1 to 3; P = 0.17).
Most transfusions were given after the third day of the ICU
stay. During the hospital stay, more patients in the liberal-strategy group received an erythrocyte transfusion than did patients
in the restrictive-strategy group (48.5 vs. 32.7%; P = 0.024).
The average hemoglobin concentration before transfusion in
the regular ward was similar between groups (7.5g/dl [0.6] vs.
7.5g/dl [0.9]; P = 0.99)
There was no difference in the age of erythrocyte units
between the liberal- and the restrictive-strategy groups
(median, 10 days [IQR, 12 to 15] vs. 13 days [IQR, 9 to
16]; P = 0.74; table 4).
Attending physicians could administer erythrocyte transfusions outside the rules of the protocol if they considered
the patient status to be life threatening, as in hemorrhagic
or other forms of circulatory shock; such an event was considered a protocol deviation. There were 13 cases of protocol
deviation in the liberal group; in all cases, patients did not
receive erythrocyte transfusion when the hemoglobin concentration was less than 9g/dl. In the restrictive group, there
were seven cases of protocol deviation; in all cases, patients
received erythrocyte transfusion when the hemoglobin concentration was greater than 7g/dl.
Discussion
In this trial involving 198 critically ill patients who were
admitted to a surgical ICU after major surgery for abdominal cancer, a blood transfusion strategy using a hemoglobin trigger of 9.0g/dl was superior to a transfusion
strategy using a trigger of 7.0g/dl in terms of the primary
outcome (30-day mortality or severe clinical complications). The restrictive erythrocyte transfusion strategy was
also associated with an increased rate of severe complications, including intraabdominal infections, cardiovascular
complications, and 60-day mortality, compared with the
liberal group.
Our study enrolled patients with active cancer who had
a high risk of postoperative complications. The aim was to
assess the trade-off between the complications of postoperative anemia and the benefits of blood transfusion in
this high-risk population using two different blood transfusion strategies and a composite endpoint of cardiovascular events, severe surgical complications, infection, organ
failure, and death.
In two previous randomized controlled trials in highrisk surgical patients, one concerning cardiac surgery (the
Transfusion Requirements After Cardiac Surgery study)19
and the other major orthopedic surgery (the Transfusion
Trigger Trial for Functional Outcomes in Cardiovascular
Table 3. Outcome Measures

Variable, % (95% CI)
Primary outcome
Death or severe complication at 30 d
Secondary outcomes
Mortality from all causes at 30 d
Acute respiratory distress syndrome
Septic shock
Acute kidney injury
Renal replacement therapy
Cardiovascular complications
Myocardial infarction
Stroke or transient ischemic attack
Mesenteric ischemia
Peripheral arterial ischemia
Unexpected cardiac arrest
Congestive heart failure
Pulmonary embolism
Reoperation
New infection
Source of infection
Abdomen
Lung
Urinary tract
Wound
Mediastinum
Blood stream
Unidentified
Need for mechanical ventilation during
ICU stay
Duration of mechanical ventilation, median
IQR, d
Need for vasopressor during ICU stay
Duration of vasopressor, median IQR, d
ICU readmission
ICU length of stay, median IQR, d
Hospital length of stay, median IQR, d
60-d mortality from all causes
Liberal Strategy (N = 97)
Restrictive Strategy (N = 101)
P Value
19.6 (12.928.6)
35.6 (27.045.4)
0.012
8.2 (4.215.4)
0 (03.8)
13.4 (8.021.6)
45.4 (35.855.3)
2.1 (0.67.2)
5.2 (2.211.5)
0 (03.8)
0 (03.8)
0 (03.8)
1.0 (0.25.6)
1.0 (0.25.6)
2.1 (0.67.2)
1.0 (0.25.6)
10.3 (5.718.0)
21.6 (14.630.8)
22.8 (15.731.9)
2.0 (0.56.9)
21.8 (14.930.8)
43.6 (34.353.3)
3.0 (1.08.4)
13.9 (8.421.9)
1.0 (0.25.4)
3.0 (1.08.4)
1.0 (2.05.4)
2.0 (0.56.9)
4.0 (1.69.7)
5.0 (2.111.1)
1.0 (2.05.4)
16.8 (10.825.3)
30.7 (22.540.3)
0.005
0.498
0.122
0.799
1.00
0.038
1.00
0.247
1.00
1.00
0.369
0.445
1.00
0.181
0.148
5.2 (2.211.5)
7.2 (3.514.2)
3.1 (1.18.7)
4.1 (1.610.1)
1.0 (0.25.6)
4.1 (1.610.1)
1.0 (0.25.6)
30.9 (22.640.7)
14.9 (9.223.1)
7.9 (4.114.9)
3.0 (1.08.4)
3.0 (1.08.4)
2.0 (0.56.9)
3.0 (1.08.4)
0 (03.7)
39.6 (30.149.4)
0.024
0.851
1.00
0.717
1.00
0.717
0.490
0.202
2 (13)
2 (12)
58.8 (48.868.0)
2 (24)
15.5 (9.624.0)
4 (37)
13 (1020)
11.3 (6.519.2)
0.803
56.4 (46.765.7)
2 (14)
17.8 (11.626.4)
4 (38)
14 (1022)
23.8 (16.532.9)
0.740
0.476
0.656
0.758
0.686
0.022
ICU = intensive care unit; IQR = interquartile range.
Patients Undergoing Surgical Hip Fracture Repair study),20

no benefit in terms of reduced severe postoperative complications or mortality was demonstrated when comparing
a liberal transfusion strategy with a restrictive strategy. We
formulated our hypothesis based on these studies that demonstrated that a restrictive strategy of blood transfusion was
safe in surgical patients.19,20 However, our results showed
the opposite effect: We found that patients in the restrictive-strategy group were more likely to achieve the primary
endpoint particularly in relation to cardiovascular events
and death. There are several differences between these studies and the current study that may, in part, explain the different results. First, the Transfusion Requirements After
Cardiac Surgery study19 enrolled only patients who were
having elective cardiac surgery, whereas we included both
elective and emergency surgical patients. Second, in the
Transfusion Requirements After Cardiac Surgery study, a
higher trigger was used to define the restrictive group (a
hematocrit of 24% and a hemoglobin level of approximately 8.0g/dl) compared with the hemoglobin threshold
of 7.0g/dl used in the current study. In the Functional
Outcomes in Cardiovascular Patients Undergoing Surgical
Hip Fracture Repair study,20 the hemoglobin threshold was
again 8.0g/dl or symptoms of anemia; as a result of this less
restrictive strategy, 41% of the patients in the Functional
Outcomes in Cardiovascular Patients Undergoing Surgical
Hip Fracture Repairrestrictive group received erythrocyte
transfusions compared with 21% in our study. Furthermore, only 3% (approximately 60) of the patients in the
Functional Outcomes in Cardiovascular Patients Undergoing Surgical Hip Fracture Repair study were transferred to
the ICU. Conversely, all of the participants in our study
were critically ill.
Other factors may also help explain some of the differences
in our results compared with those of previous studies. First,
we used leukodepleted blood in all transfused patients. The
Table 4. Hemoglobin Concentrations and Erythrocyte Transfusions

Variable
Hemoglobin concentration at ICU admission, g/dl (SD)
Hemoglobin concentration before transfusion during ICU stay, g/dl (SD)
Hemoglobin concentration before transfusion in regular wards, g/dl (SD)
Transfusion before randomization
0 unittotal No. (%)
1 or more unitstotal No. (%)
Total No. of units
Transfusion during ICU stay
0 unittotal No. (%)
1 unittotal No. (%)
2 unitstotal No. (%)
Transfusion during hospital stay
1 or more unitstotal No. (%)
Total no. of units
Protocol deviation (%)
Age of transfused erythrocytes (d)
Liberal Strategy
(N = 97)
Restrictive Strategy
(N = 101)
11.0 (1.6)
7.9 (0.5)
7.5 (0.6)
11.2 (1.8)
6.8 (0.5)
7.5 (0.9)
70 (72.2)
27 (27.8)
2 (13)
76 (75.2)
25 (24.8)
1 (12)
56 (57.7)
14 (14.4)
8 (8.2)
13 (13.4)
6 (6.2)
80 (79.2)
12 (11.9)
3 (3)
2 (2)
4 (4)
50 (51.5)
47 (48.5)
134
13 (13.4)
10 (1215)
68 (67.3)
33 (32.7)
88
7 (6.9)
13 (916)
P Value
0.458
<0.001
0.99
0.622
0.105
0.005
0.024
0.131
0.743
ICU = intensive care unit.
leukocyte component of transfused blood has been associated

with development of acute lung injury, immunosuppressive
effects, postoperative infection, and systemic inflammatory
response syndrome. Second, our erythrocyte units had a shorter
storage time compared with previous studies.20,21 Older erythrocytes have increased adherence to the endothelium and reduced
capacity for tissue oxygen transport because of the higher affinity for oxygen when compared with younger erythrocytes, thus
potentially limiting the benefits of blood transfusion.
Fig. 3. The mean lowest hemoglobin concentration per day

during the first 14 days in patients randomized to a restrictive
strategy of erythrocyte transfusion (transfusion when hemoglobin concentration <7g/dl) and those randomized to a liberal strategy (transfusion when hemoglobin concentration <9g/
dl). The P value was calculated using a mixed-model ANOVA
to verify differences in hemoglobin concentrations over time
between groups.
Another possible explanation for the different finding

is that patients with cancer receiving restrictive transfusions may be more susceptible to altered oxygen delivery
and impaired tissue oxygenation during the postoperative
period, leading to higher rates of complications and death.
Jhanji et al.22 reported that patients having major abdominal
surgery who had impaired microvascular flow after surgery
experienced a higher rate of postoperative complications
than did patients with normal microvascular flow (measured
with sublingual capillaroscopy). Abnormalities in microvascular flow can occur when hemoglobin levels decrease
less than 8.0g/dl. In a study of patients with trauma with
an average hemoglobin of 7.5g/dl and impaired capillary
perfusion, Weinberg et al.23 demonstrated that erythrocyte
transfusion improved microvascular flow. However, no
changes were observed after blood transfusions in patients
with normal capillary perfusion. Several small trials have
also demonstrated that strategies based on maximizing oxygen delivery after major surgery may reduce postoperative
complications.18,2428
A relationship between anemia and postoperative mortality has been described in other studies.2931 Carson
et al.29 reported that pre- and postoperative anemia was independently associated with 30-day mortality, particularly in
patients with cardiovascular disease. In a propensity score
matched retrospective study, Wu et al.31 showed that erythrocyte transfusion in patients with hematocrit levels less than
24% (approximately equal to hemoglobin concentrations
of 8.0g/dl) was associated with reduced 30-day postoperative mortality in elderly patients having major, noncardiac
surgery. Similar findings were reported by Sakr et al.21 who
conducted a large prospective study of 5,925 patients in a

surgical ICU and reported that anemia (hemoglobin concentration <9.0g/dl) was common and was associated with
higher morbidity and mortality. In a further analysis, after
propensity score matching and adjusting for possible confounders, higher hemoglobin concentrations and the receipt
of a blood transfusion were independently associated with a
lower risk of hospital mortality.21
The main cause of mortality in our study was multiple
organ failure as a result of septic shock, particularly in patients
who developed surgical complications and intraabdominal
infection. Although there is no evidence from randomized
clinical trials that a restrictive strategy increases mortality
in patients with severe sepsis or septic shock, recent studies
have suggested a potential benefit of blood transfusion in
this population. Park et al.32 reported in a propensity score
matched cohort that blood transfusion was independently
associated with a lower risk of 7-day, 28-day, and in-hospital
mortality in patients with severe sepsis and septic shock.
Similar findings were also reported by Sakr et al.21 in surgical
patients. In their study, blood transfusion was independently
associated with a lower risk of in-hospital death, especially in
patients aged from 66 to 80 yr, in patients admitted to the
ICU after noncardiovascular surgery, in patients with higher
severity scores, and in patients with severe sepsis.21
Our study has limitations inherent to a single-center
study. We performed the study in a tertiary university hospital for patients with cancer within the public health system.
Our results may, therefore, reflect the characteristics of our
center and generalization of our findings to other centers may
be limited. However, we observed a 30-day mortality rate of
15.5%, similar to that reported in a recent South American
multicenter cohort of patients having surgery for cancer who
were admitted to the ICU.33 In a recent large European survey of patients having noncardiac surgery for whom ICU
admission was required, 14% died before hospital discharge.3
Another limitation of our study is the lack of long-term follow-up of patients, and thus we could not address any potential negative effects of transfusion on cancer recurrence.
Conclusion
In this controlled, randomized trial of patients admitted to
the ICU after major surgery for abdominal cancer, a liberal
erythrocyte transfusion strategy using a hemoglobin threshold of 9.0g/dl was superior to a restrictive strategy with a
hemoglobin threshold of 7.0g/dl. Our findings are highly
relevant because a restrictive erythrocyte transfusion policy
has been advocated for surgical patients with cancer because
of the potential association between erythrocyte transfusion
and cancer recurrence. The association of a restrictive postoperative blood transfusion strategy with poorer short-term
outcomes, even from a single-center study, should alert physicians to the possibility that a restrictive strategy, based on a
hemoglobin concentration of 7.0g/dl, may not be as safe as
previously perceived.
Acknowledgments
Support was provided solely from institutional and/or departmental sources.
Competing Interests
Correspondence
Address correspondence to Dr. Vincent: Erasme University
Hospital, Route de Lennik 808, B-1070 Brussels, Belgium.
jlvincen@ulb.ac.be. This article may be accessed for personal use at no charge through the Journal Web site, www.
anesthesiology.org.
References
1. Law S, Wong KH, Kwok KF, Chu KM, Wong J: Predictive
factors for postoperative pulmonary complications and
mortality after esophagectomy for cancer. Ann Surg 2004;
240:791800
2. Masoomi H, Kang CY, Chen A, Mills S, Dolich MO, Carmichael
JC, Stamos MJ: Predictive factors of in-hospital mortality in
colon and rectal surgery. J Am Coll Surg 2012; 215:25561
3. Pearse RM, Moreno RP, Bauer P, Pelosi P, Metnitz P, Spies C,
Vallet B, Vincent JL, Hoeft A, Rhodes A; European Surgical
Outcomes Study (EuSOS) Group for the Trials Groups of
the European Society of Intensive Care Medicine and the
European Society of Anaesthesiology: Mortality after surgery
in Europe: A 7 day cohort study. Lancet 2012; 380:105965
4. Weber RS, Jabbour N, Martin RC II: Anemia and transfusions
in patients undergoing surgery for cancer. Ann Surg Oncol
2008; 15:3445
5. Acheson AG, Brookes MJ, Spahn DR: Effects of allogeneic
red blood cell transfusions on clinical outcomes in patients
undergoing colorectal cancer surgery: A systematic review
and meta-analysis. Ann Surg 2012; 256:23544
6. Dicato M, Plawny L, Diederich M: Anemia in cancer. Ann
Oncol 2010; 21(suppl 7):vii16772
7. Koch M, Antolovic D, Reissfelder C, Rahbari NN, Holoch
J, Michalski I, Sweiti H, Ulrich A, Bchler MW, Weitz J:
Leucocyte-depleted blood transfusion is an independent predictor of surgical morbidity in patients undergoing elective
colon cancer surgery-a single-center analysis of 531 patients.
Ann Surg Oncol 2011; 18:140411
8. Amato A, Pescatori M: Perioperative blood transfusions for
the recurrence of colorectal cancer. Cochrane Database Syst
Rev 2006;CD005033
9. Vamvakas EC, Blajchman MA: Deleterious clinical effects of
transfusion-associated immunomodulation: Fact or fiction?
Blood 2001; 97:118095
10. Hbert PC, Wells G, Blajchman MA, Marshall J, Martin C,

Pagliarello G, Tweeddale M, Schweitzer I, Yetisir E: A multicenter, randomized, controlled clinical trial of transfusion
requirements in critical care. Transfusion Requirements
in Critical Care Investigators, Canadian Critical Care Trials
Group. N Engl J Med 1999; 340:40917
11. Schag CC, Heinrich RL, Ganz PA: Karnofsky performance status revisited: Reliability, validity, and guidelines. J Clin Oncol
1984; 2:18793
12. Oken MM, Creech RH, Tormey DC, Horton J, Davis TE,

McFadden ET, Carbone PP: Toxicity and response criteria of
the Eastern Cooperative Oncology Group. Am J Clin Oncol
1982; 5:64955
13. Charlson ME, Pompei P, Ales KL, MacKenzie CR: A new method
of classifying prognostic comorbidity in longitudinal studies:
Development and validation. J Chronic Dis 1987; 40:37383
14. Thygesen K, Alpert JS, White HD; Joint ESC/ACCF/AHA/WHF

Task Force for the Redefinition of Myocardial Infarction:
Universal definition of myocardial infarction. J Am Coll
Cardiol 2007; 50:217395
15. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D,
Cohen J, Opal SM, Vincent JL, Ramsay G; SCCM/ESICM/ACCP/
ATS/SIS: 2001 SCCM/ESICM/ACCP/ATS/SIS International
Sepsis Definitions Conference. Crit Care Med 2003; 31:12506
16. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P; Acute
Dialysis Quality Initiative workgroup: Acute renal failure
Definition, outcome measures, animal models, fluid therapy
and information technology needs: The Second International
Consensus Conference of the Acute Dialysis Quality Initiative
(ADQI) Group. Crit Care 2004; 8:R20412
17. Bernard GR, Artigas A, Brigham KL, Carlet J, Falke K, Hudson
L, Lamy M, Legall JR, Morris A, Spragg R: The AmericanEuropean Consensus Conference on ARDS. Definitions,
mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med 1994; 149(3 Pt 1):81824
18. Pearse R, Dawson D, Fawcett J, Rhodes A, Grounds RM,

Bennett ED: Early goal-directed therapy after major surgery
reduces complications and duration of hospital stay. A randomised, controlled trial [ISRCTN38797445]. Crit Care 2005;
9:R68793
19. Hajjar LA, Vincent JL, Galas FR, Nakamura RE, Silva CM,
Santos MH, Fukushima J, Kalil Filho R, Sierra DB, Lopes
NH, Mauad T, Roquim AC, Sundin MR, Leo WC, Almeida JP,
Pomerantzeff PM, Dallan LO, Jatene FB, Stolf NA, Auler JO Jr:
Transfusion Requirements After Cardiac Surgery: The TRACS
randomized controlled trial. JAMA 2010; 304:155967
20. Carson JL, Terrin ML, Noveck H, Sanders DW, Chaitman BR,
Rhoads GG, Nemo G, Dragert K, Beaupre L, Hildebrand K,
Macaulay W, Lewis C, Cook DR, Dobbin G, Zakriya KJ, Apple
FS, Horney RA, Magaziner J; FOCUS Investigators: Liberal or
restrictive transfusion in high-risk patients after hip surgery.
N Engl J Med 2011; 365:245362
21. Sakr Y, Lobo S, Knuepfer S, Esser E, Bauer M, Settmacher U,
Barz D, Reinhart K: Anemia and blood transfusion in a surgical intensive care unit. Crit Care 2010; 14:R92
22. Jhanji S, Vivian-Smith A, Lucena-Amaro S, Watson D, Hinds
CJ, Pearse RM: Haemodynamic optimisation improves tissue
microvascular flow and oxygenation after major surgery: A
randomised controlled trial. Crit Care 2010; 14:R151
23. Weinberg JA, MacLennan PA, Vandromme-Cusick MJ, Angotti
JM, Magnotti LJ, Kerby JD, Rue LW III, Barnum SR, Patel
RP: Microvascular response to red blood cell transfusion in
trauma patients. Shock 2012; 37:27681
24. Flancbaum L, Ziegler DW, Choban PS: Preoperative intensive care unit admission and hemodynamic monitoring in
patients scheduled for major elective noncardiac surgery:
A retrospective review of 95 patients. J Cardiothorac Vasc
Anesth 1998; 12:39
25. Lobo SM, Salgado PF, Castillo VG, Borim AA, Polachini CA,
Palchetti JC, Brienzi SL, de Oliveira GG: Effects of maximizing oxygen delivery on morbidity and mortality in high-risk
surgical patients. Crit Care Med 2000; 28:3396404
26. Pearse R, Dawson D, Fawcett J, Rhodes A, Grounds RM,

Bennett ED: Changes in central venous saturation after
major surgery, and association with outcome. Crit Care 2005;
9:R6949
27. Pearse RM, Belsey JD, Cole JN, Bennett ED: Effect of dopexamine infusion on mortality following major surgery:
Individual patient data meta-regression analysis of published
clinical trials. Crit Care Med 2008; 36:13239
28. Sakr Y, Chierego M, Piagnerelli M, Verdant C, Dubois MJ, Koch
M, Creteur J, Gullo A, Vincent JL, De Backer D: Microvascular
response to red blood cell transfusion in patients with severe
sepsis. Crit Care Med 2007; 35:163944
29. Carson JL, Duff A, Poses RM, Berlin JA, Spence RK, Trout
R, Noveck H, Strom BL: Effect of anaemia and cardiovascular disease on surgical mortality and morbidity. Lancet 1996;
348:105560
30. Carson JL, Noveck H, Berlin JA, Gould SA: Mortality and morbidity in patients with very low postoperative Hb levels who
decline blood transfusion. Transfusion 2002; 42:8128
31. Wu WC, Schifftner TL, Henderson WG, Eaton CB, Poses RM,
Uttley G, Sharma SC, Vezeridis M, Khuri SF, Friedmann PD:
Preoperative hematocrit levels and postoperative outcomes
in older patients undergoing noncardiac surgery. JAMA 2007;
297:24818
32. Park DW, Chun BC, Kwon SS, Yoon YK, Choi WS, Sohn JW,
Peck KR, Kim YS, Choi YH, Choi JY, Kim SI, Eom JS, Kim
HY, Cheong HJ, Song YG, Choi HJ, Kim JM, Kim MJ: Red
blood cell transfusions are associated with lower mortality in
patients with severe sepsis and septic shock: A propensitymatched analysis. Crit Care Med 2012; 40:31405
33. Soares M, Caruso P, Silva E, Teles JM, Lobo SM, Friedman
G, Dal Pizzol F, Mello PV, Bozza FA, Silva UV, Torelly AP,
Knibel MF, Rezende E, Netto JJ, Piras C, Castro A, Ferreira
BS, Ra-Neto A, Olmedo PB, Salluh JI; Brazilian Research in
Intensive Care Network (BRICNet): Characteristics and outcomes of patients with cancer requiring admission to intensive care units: A prospective multicenter study. Crit Care
Med 2010; 38:915
Anaphylaxis Is More Common with Rocuronium

and Succinylcholine than with Atracurium
Jeffrey I. Reddy, F.A.N.Z.C.A., Peter J. Cooke, F.A.N.Z.C.A., Johan M. van Schalkwyk, F.R.A.C.P.,
Jacqueline A. Hannam, Ph.D., Penny Fitzharris, M.D., Simon J. Mitchell, Ph.D., F.A.N.Z.C.A.
ABSTRACT
Background: Intraoperative anaphylaxis is a rare but serious occurrence, often triggered by neuromuscular-blocking drugs
(NMBDs). Previous reports suggest that the rates of anaphylaxis may be greater for rocuronium than for other NMBDs, but
imprecise surrogate metrics for new patient exposures to NMBDs complicate interpretation.
Methods: This was a retrospective, observational cohort study of intraoperative anaphylaxis to NMBDs at two hospitals
between 2006 and 2012. Expert anesthetic and immunologist collaborators investigated all referred cases of intraoperative
anaphylaxis where NMBDs were administered and identified those where a NMBD was considered responsible. New patient
exposures for each NMBD were extracted from electronic anesthetic records compiled during the same period. Anaphylaxis
rates were calculated for each NMBD using diagnosed anaphylaxis cases as the numerator and the number of new patient
exposures as the denominator.
Results: Twenty-one patients were diagnosed with anaphylaxis to an NMBD. The incidence of anaphylaxis was 1 in 22,451
new patient exposures for atracurium, 1 in 2,080 for succinylcholine, and 1 in 2,499 for rocuronium (P < 0.001).
Conclusions: In Auckland, the rate of anaphylaxis to succinylcholine and rocuronium is approximately 10-fold higher than
to atracurium. Previous estimates of NMBD anaphylaxis rates are potentially confounded by inaccurate proxies of new patient
exposures. This is the first study to report anaphylaxis rates using a hard denominator of new patient exposures obtained
directly from anesthetic records. (Anesthesiology 2015; 122:39-45)
NTRAOPERATIVE anaphylaxis is a rare but serious

event that may cause significant morbidity and mortality.13 Neuromuscular-blocking drugs (NMBDs) are
common causative agents during anesthesia.2,48 There is
controversy whether the incidence of anaphylaxis is higher
with rocuronium than with other NMBDs. Evidence that
this might be so has been reported from France,1,6 Norway,5,9
and some parts of Australia,4,7 whereas no difference has been
found from the limited data available for the United States.10
Such comparisons are complicated by difficulties in
obtaining accurate numerator and denominator data with
which to calculate an incidence for the various drugs. Deriving accurate numerators relies on capture of all relevant
anaphylaxis cases and thorough and consistent case investigation. Denominators based on cases actually exposed to
each agent are even harder to obtain because of the difficulties associated with retrieval of administration records from
many thousands of anesthetics. For the latter reason, relevant denominators have usually been estimated from sales
data or similar metrics that fail to account for confounders
such as vials opened but not used, discarded date-expired
vials, and repeat administrations or infusions. These problems, combined with the previously mentioned potential

Neuromuscular-blocking drugs are common causative agents
of intraoperative anaphylaxis
Comparisons of neuromuscular-blocking drug anaphylaxis
rates are complicated by difficulties in obtaining accurate numerator and denominator data

Search of a database containing more than 400,000 anesthetic records identified 92,858 new patient exposures to
neuromuscular-blocking drugs between 2006 and 2012
Twenty-one of 89 patients referred to the Anaesthetic Allergy
Clinic had anaphylaxis attributed to muscle relaxants
Use of credible numerator and denominator data found similar
rates of anaphylaxis after succinylcholine and rocuronium administration, rates that were nearly an order of magnitude higher than
those for atracurium and other neuromuscular-blocking drugs
for geographical variation, result in divergent estimates of

anaphylaxis incidence for the same drug. For example, the
reported incidence of anaphylaxis to rocuronium varies from
approximately 1:3,500 to 1:445,000.5,11
We undertook a 7-yr retrospective review of the incidence
of intraoperative anaphylaxis to NMBDs in Auckland, New
Corresponding article on page 5. Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text
and are available in both the HTML and PDF versions of this article. Links to the digital files are provided in the HTML text of this article on
the Journals Web site (www.anesthesiology.org).
Submitted for publication March 18, 2014. Accepted for publication August 11, 2014. From the Department of Anaesthesia and Perioperative Medicine, North Shore Hospital, Auckland, New Zealand ( J.I.R.); Department of Anaesthesia and Perioperative Medicine (P.J.C., J.M.v.S.)
and Department of Immunology (P.F.), Auckland City Hospital, Auckland, New Zealand; and Department of Anaesthesiology ( J.A.H., S.J.M.),
University of Auckland, Auckland, New Zealand.
January 2015
Zealand. All cases of intraoperative anaphylaxis in the city

were referred to a single clinic for investigation, facilitating capture of cases. Moreover, hospital catchment areas are
strictly defined and maintained, and two of the three large
hospitals in the city used an electronic system to record all
anesthetics during this 7-yr period. The associated database
contains more than 400,000 anesthetic records which can be
searched for administration of particular drugs. These local
practices provide accurate numerators and denominators for
the calculation of anaphylaxis rates for anesthetic drugs. We
compared anaphylaxis rates for various NMBDs, with the
null hypothesis being that there is no difference in anaphylaxis rates between agents.

This retrospective, observational cohort study was approved
by the Health and Disability Ethics Committee (Reference:
12/NTA/65) and institutional approval was granted by
Auckland District Health Board (Auckland, New Zealand)
and Waitemata District Health Board (Auckland, New Zealand). Approval was also granted to release the full, anonymized dataset.
Denominator Data
Auckland City Hospital and North Shore Hospital are the
two principal public hospitals within Auckland District
Health Board and Waitemata District Health Board, respectively. Both hospitals use the SAFERsleep electronic anesthetic record keeping and safety system (SAFERSleep: Safer
Sleep LLC, Nashville, TN).12 This system was fully implemented in all theaters before the study period from January
1, 2006 to December 31, 2012. All drug administrations
during an anesthetic are entered by the anesthesiologist using
either bar code scanning of specific drug labels on syringes or
manual entry (via a keyboard) and are permanently recorded
by the system.
SAFERsleep maintains anesthetic records in a secure
database. Using relevant search criteria in Structured Query
Language, we identified all anesthetics in which NMBDs
were used. For each record, we extracted the patients unique
National Health Index number, sex, age, name of NMBD
used, total number of administrations of NMBD, and use of
infusions. After excluding duplication of patients undergoing multiple procedures, we calculated the number of new
patient exposures to each NMBD. A new patient exposure
was defined as the administration of an NMBD to a patient,
for the first time (during the study period). That is, if the
same patient received the same NMBD during one or more
anesthetics, a single new exposure was considered to have
occurred during the period of analysis.
Numerator Data
The Auckland Anesthetic Allergy Clinic is a multidisciplinary
clinic staffed by anesthesiologists, immunologists, and immunology technologists. Case referrals listed all medications and
substances administered before the episode of anaphylaxis,

the clinical features, and details of treatment. The anesthetic
record was also attached. Patients were seen at the clinic
approximately 6 weeks after receipt of referral for consultation and skin testing. The consultation elucidated any other
relevant history and established the patients fitness and consent for skin testing. Skin testing was carried out according to
the clinics protocol which is based on the methodology first
described by Fisher and Bowey.13 The clinical features, serial
tryptase results, specific immunoglobulin E testing, and skin
testing were then used to confirm the diagnosis and identify the likely causative agent. All medications administered
before the anaphylaxis were tested. All patients had skin testing for chlorhexidine (skin prick test 2% aqueous) and latex.
Intradermal skin testing was generally performed on
the patients back. A volume of 0.02ml of each drug was
injected intradermally, and the size of the wheal was measured with calipers after 15min for comparison with the size
of the wheal produced by the injection of 0.02ml of 0.9%
saline (negative control). The test was regarded as positive if
the wheal diameter obtained with the drug was larger than
the negative control wheal by 3mm or more. The test was
regarded as equivocal if the wheal diameter increased by 1
to 2mm, with a surrounding flare. A skin prick test with
histamine 10mg/ml was used as a positive control. The drug
dilutions used for intradermal testing of muscle relaxants are
provided in table1. If any muscle relaxant had been administered, skin tests were carried out with the full range of muscle relaxants available to detect cross-sensitization.
We included all patients from the two hospitals who
were referred after intraoperative anaphylaxis and who had
received NMBDs during the study period. Relevant correspondence, referral forms, anesthetic records, and the skin
testing results were examined. All the cases were reviewed
independently by an anesthetist and an immunologist and
then discussed. In the cases confirmed as NMBD-induced
allergic anaphylaxis, severity grading was made according to
the guidelines published by Mertes et al. (table2).14 Peak
serum tryptases were also recorded.
Diagnostic classification of the patients was based on
clinical consensus on all of the following points:
1. Whether or not the patient had one or more manifestations of anaphylaxis as described by Mertes et al.14;
2. The temporal relation between the administration of an
NMBD and the onset of anaphylaxis;
3. The supporting laboratory evidence of allergic anaphylaxis to the relevant NMBD based on intradermal
testing with NMBDs, the serum tryptase result, and
specific immunoglobulin E testing when available;
4. Ensuring that skin testing had been carried out for
other substances or medications that may have caused
the anaphylaxis.
Where skin testing (described earlier in this section) was
equivocal, all the above features were used to determine
Anesthesiology 2015; 122:39-45 40 Reddy et al.
Table 1. Drug Dilutions Used for Intradermal Testing

Drug
Concentration
Succinylcholine
Mivacurium
Atracurium
Pancuronium
Vecuronium
Rocuronium
Saline (negative
control)
0.05mg/ml
0.0002mg/ml
0.001mg/ml
0.002mg/ml
0.004mg/ml
0.01mg/ml
0.9%
1:1,000 dilution of 100mg in 2 ml

Table 2. Clinical Grading of Anaphylaxis

Grade
1
2
3
4
Symptoms
Cutaneous signs: generalized erythema, urticaria,
angioedema
Measurable but not life-threatening symptoms: cutaneous signs, hypotension, tachycardia
Respiratory disturbances: cough, difficulty inflating
Life-threatening symptoms: collapse, tachycardia or
bradycardia, arrhythmias, bronchospasm
Cardiac and/or respiratory arrest
Adapted from Mertes et al. J Investig Allergol Clin Immunol 2011; 21:442
53.14 Adaptations are themselves works protected by copyright. So in
order to publish this adaptation, authorization must be obtained both from
the owner of the copyright in the original work and from the owner of copyright in the translation or adaptation.
the cause, as is appropriate in clinical decision making. It

is important to emphasize that no one test unequivocally
allows diagnosis of anaphylaxis to NMBDs. It is difficult to
use a rigid case definition with the sensitivity and specificity
of all the available tests being incompletely understood.
The rate of anaphylaxis to NMBDs was calculated using
confirmed cases of anaphylaxis to each drug as the numerator and the number of new patient exposures to the drug as
the denominator. Fisher exact test was used to compare the
incidence of anaphylaxis to the various NMBDs during the
entire interval. CI (95%) were calculated based on the Poisson distribution. A P value of less than 0.05 was considered to
indicate statistical significance. All analyses were conducted
in R, version 2.15.1 (R Foundation for Statistical Computing, Vienna, Austria). Because of the large numbers in the
denominator, P values for Fisher exact test were computed
using Monte Carlo simulation with 107 replicates. Additional
details of our statistical analysis are given in Supplemental
Digital Content 1, http://links.lww.com/ALN/B110.
Results
During the 7-yr period from January 1, 2006 to December 31,
2012, there were 92,858 new patient exposures to NMBDs.
Database queries and analyses are available in Supplemental
* Available at: http://www.anaesthetist.com/R/allergy2014/nmba_
anaphylaxis_supplement_3.zip. Accessed June 19, 2014.
Digital Content 2, http://links.lww.com/ALN/B111. The

full (deidentified) dataset is available online.* Eighty-nine of
these patients were referred to the Anesthetic Allergy Clinic
for follow-up investigation of an intraoperative event that
was thought to be anaphylaxis (table3).
Two referred cases did not attend the clinic and were
lost to follow-up. In five cases, we excluded anaphylaxis
with a high level of certainty on historical grounds, and in
36 cases with negative skin testing, a diagnosis of nonallergic (nonimmunoglobulin E mediated) anaphylaxis was
made. In 25 cases, the causative agent was identified as a
substance other than a muscle relaxant (chlorhexidine 8,
cefazolin 7, Gelofusine 5, latex 1, tramadol 1, diclofenac
1, paracetamol 1, and protamine 1). Twenty-one cases of
allergic anaphylaxis were attributed to muscle relaxants.
Table 3 summarizes these cases and lists all use of muscle
relaxants in all cases, including those cases lost to follow-up
and those considered either due to nonallergic anaphylaxis
or not to represent anaphylaxis at all.
Demographics and clinical features of these 21 cases are
shown in table4. The average age of patients was 59 yr and
females accounted for 17 of 21 (81%) cases. Four cases were
categorized as clinical grade 2, 12 as grade 3, and 5 as grade
4. The median peak tryptase level was 59 g/l (range, 7.8 to
>200 g/l), with only one patient (20) having a tryptase of
Table 3. Classification of Patients Referred to the Anesthetic
Allergy Clinic and Muscle Relaxants Received
Count
Nonallergic anaphylaxis
Atracurium
Succinylcholine
Succinylcholine and
atracurium
Pancuronium
Vecuronium
Rocuronium
Total
Did not attend clinic
Atracurium
Succinylcholine
Total
Allergic anaphylaxis to a muscle relaxant
Atracurium
Rocuronium
Succinylcholine
Total
Allergic anaphylaxis to
drugs that are not muscle
relaxants:
Not allergy
Atracurium
Succinylcholine
Succinylcholine and
atracurium
Rocuronium
Total
Grand total
11
12
2
1
2
8
36
1
1
2
3
6
12
21
25
1
2
1
1
5
89
Table 4. Clinical Features of Cases with Anaphylaxis to Neuromuscular-blocking Drugs

Wheal/Flare
Size (mm)
Saline
Relaxant
Positive
by Skin
Test
Criteria
Index
Drug
Age
(yr)
Sex
Grade
Tryptase
(peak:
g/l)
sux
64
49
7/0
No
roc
69
37.2
10/25
Yes
sux
70
>200
12/29
Yes
roc
78
147
10/32
No
roc
70
7/10
No
sux
HYT, R
sux
41
127
10/35
Yes
roc
roc
46
154
9/31
No
sux
56
63
9/59
Yes
atrac
67
76
8/12
No
HYT, BSM,
TACHY
HYT, BSM,
TACHY,
HYPOX, R
HYT, HYPOX,
TACHY
HYT, R
10
11
12
sux
sux
roc
65
49
96
F
M
F
3
2
4
174
22.8
30.4
7
7
6
11/196
12/81
8/23
Yes
Yes
No
13
sux
36
38
13/53
Yes
14
sux
69
16.5
7/56
No
roc
HYT, BSM,
HYPOX, R
HYT, BSM,
HYPOX, R,
urticaria
15
16
17
sux
sux
atrac
31
65
32
F
F
F
3
3
3
58.5
79.3
39.6
8
8
7
12/35
14/40
9/35
Yes
Yes
No
vec
HYT, BSM, FS
HYT, TACHY, R
HYT, TACHY, R
18
19
20
sux
sux
roc
66
50
50
F
F
F
4
3
3
>200
67.8
7.8
7
8
6
16/32
13/32
12/125
Yes
Yes
Yes
21
atrac
65
59
6/0
No
Case
16.3
Crosssensitivity
Prominent
Clinical Features
HYT, BSM,
HYPOX, R
vec
vec, sux
miv, atrac
vec, panc
sux, vec
HYT, BSM, R,
arrhythmia
HYT, BSM, R,
BRADY
HYT, BSM,
HYPOX, R
HYT, BSM, R
HYT, BSM, R
HYT, TACHY, R
HYT, R
HYT, R
HYT, BSM,
TACHY, R
HYT, TACHY,
flushing,
ventricular
fibrillation
Notes
+IgE succinylcholine, negative
ID test
ID equivocal,
strongly positive
clinical features
On retest rocuronium 8/30 saline
6mm, equivocal
+IgE
succinylcholine
ID equivocal,
multiple other
positives
Systemic
mastocytosis
ID equivocal,
multiple other
positives
+IgE
succinylcholine
+IgE succinylcholine, ID for
rocuronium
positive
ID equivocal,
strongly positive
clinical features
Severe anaphylaxis 2 related

to atracurium,
retested 6/0
atrac = atracurium; BRADY = bradycardia; BSM = bronchospasm; FS = facial swelling; HYPOX = hypoxemia; HYT = hypotension; ID = intradermal; IgE =
immunoglobulin E; miv = mivacurium; panc = pancuronium; R = rash; roc = rocuronium; sux = suxamethonium; TACHY = tachycardia; vec = vecuronium.
less than 12 g/l. This compared with a median peak tryptase

of 7.5 g/l (range, 1 to 33.2 g/l) in the group with nonallergic anaphylaxis, which also exhibited lower severity scores
(4 were grade 1, 20 were grade 2, and 12 were grade 3).
Nine of the 21 cases of allergic anaphylaxis did not meet
the standard skin test criteria but were nevertheless considered
to warrant inclusion on careful consideration of the clinical
picture and relevant tests. The notes on the right hand side of
table4 give an indication as to why this diagnosis was made,
despite the absence of a wheal increase of 3mm or more in

these nine cases. Two succinylcholine cases (1, 14) showed
the presence of immunoglobulin E antibodies specific for
succinylcholine with one (14) also showing cross-sensitization to rocuronium (which had not been administered).
Three rocuronium patients (5, 7, and 12) with equivocal
skin tests were cross-sensitized to various other NMBDs (not
administered). One patient (21) with negative skin tests to
atracurium experienced further anaphylaxis on reexposure
Table 5. Intraoperative Incidence of Neuromuscular-blocking Drug-related Anaphylaxis
Anaphylaxis
95% CI (Poisson)
Exposure
Rate
Range (from CI)
Succinylcholine
Rocuronium
Atracurium
Other
12
621
24,960
1:2,079
1:1,1904,030
6
213
14,995
1:2,498
1:1,1506,810
3
09
67,354
1:22,450
1:7,680109,000
0
04
15,042
1:4,080
to atracurium. One rocuronium (4) and two atracurium (9,

17) patients showed a 1 to 2-mm wheal increase, with flare,
no other cause, positive tryptase and timing consistent with
the NMBD being causative. Cross-sensitization was demonstrated overall in 9 of the 21 cases (43%).
The number of new patient exposures, number of confirmed cases of anaphylaxis, and rates of confirmed anaphylaxis to succinylcholine, rocuronium, atracurium, and
a composite of other NMBDs (vecuronium, pancuronium,
and mivacurium) are shown in table5. These data suggest
that there is a large (10-fold) difference between the rate for
atracurium and the rates for succinylcholine and rocuronium
(P < 0.001). Unsurprisingly, individual 22 comparisons
reveal that the differences reside in the rates of anaphylaxis
to succinylcholine and rocuronium compared with the other
agents. For example, the P value for rocuronium versus atracurium is approximately 0.002.
To test the robustness of our results, we performed two
main sensitivity analyses. In the first analysis, we assumed a
worst-case scenario: anaphylaxis to NMBDs in all patients
who either did not attend or are labeled as nonallergic anaphylaxis in table3. Rates of anaphylaxis to succinylcholine,
rocuronium, atracurium, and other agents are then 1:920,
1:1,070, 1:5,000, and 1:4,000, respectively. We did not
observe any cases of anaphylaxis to vecuronium in our dataset (0 of 9,585 new exposures). Application of Fisher test as
before still results in rejection of the null hypothesis at a P
value of 6107.
The second restrictive sensitivity analysis took the
opposite approach, rejecting all cases in table4 that do not
strictly conform to standard criteria and abandoning the
clinical judgment of the anesthesiologist and immunologist
who assessed the cases. Even here, a P value of 2106 mandates rejection of the null hypothesis although the difference
is then mainly due to succinylcholine.
Discussion
The principal finding of this study was that in the Auckland region, the use of succinylcholine and rocuronium was
associated with a substantially higher rate of intraoperative
anaphylaxis compared with atracurium and other NMBDs.
There was similarity between the incidence of anaphylaxis to
rocuronium and succinylcholine (approximately 1:2,500 and
1:2,000, respectively); in contrast, the rate of anaphylaxis to
atracurium was substantially lower (1:22,000). This difference is unlikely to be an artifact due to the large numbers
in the denominators, and this observation is supported by

several large European studies.1,6,8 No cases of anaphylaxis
were observed for vecuronium (0 of 9,585 new exposures).
The proportion of anaphylaxis events during anesthesia
resulting from sensitization to NMBDs (46%) is similar to
that reported in France, Norway, Spain, and Australia.2,48,15
The characterization of our patient series, with 56% of anaphylaxis cases being found to be allergic and associated with
higher tryptase and greater severity than nonallergic cases, is
similar to that in other published studies.1,2,4,8
The study provides direct calculation of comparative rates
of anaphylaxis based on actual measurement of denominator data. Previous studies have used surrogate denominators
based on metrics such as drug sales data, which are prone
to inaccuracies. The use of drug sales as an index of patient
exposures is confounded by the discarding of expired drugs,
multiple administrations, and infusions in long cases. Wastage of NMBDs can be substantial, suggesting that denominators based on drug sales or supply may substantially
overestimate exposure, resulting in a potential underestimation of anaphylaxis rates. Notwithstanding such concerns,
other studies have reported a higher rate for anaphylaxis to
rocuronium than to other nondepolarizing NMBDs,1,47 in
agreement with the results from our region.
This finding will likely give anesthetists pause to consider
the place of rocuronium in their clinical armamentarium. It
is a popular drug for a variety of reasons, not least of which
is that it exhibits the fastest onset of all the nondepolarizing NMBDs and it can be used as an acceptable substitute
for succinylcholine in a rapid sequence induction. A further
reason to use rocuronium in the latter application, despite
slower onset when compared with succinylcholine,16 may be
that its effect can be rapidly reversed by sugammadex. Other
than to discourage the selection of rocuronium over succinylcholine on the basis of a lower risk of anaphylaxis, our finding
is unlikely to change the use of rocuronium in rapid sequence
inductions because there are still many reasons why succinylcholine may be contraindicated or why anesthetists may prefer to avoid it. The present authors would have no hesitation
in using rocuronium under such circumstances. In contrast,
our findings suggest that when all other factors are equal, it
may be prudent to reconsider the use of rocuronium in routine cases where it is not being used for any of its particular
properties, at least in those regions where there is some evidence that sensitivity is prevalent. Atracurium seems a safer
alternative, and although we cannot comment on the basis
of our data which contained too few vecuronium exposures,

others have also shown vecuronium to be safer.1,6,7
There are several potential limitations to our study. First,
the data are limited to the Auckland region of New Zealand
and the results can only be extrapolated to other regions and
nations with caution. Geographical differences in sensitivity to NMBDs are likely to be real and may be based on
regional differences in exposure to other sensitizers such as
pholcodine.17,18
Second, studies of this nature are vulnerable to any systematic error that leads to an unequal likelihood of identifying cases due to one drug relative to others. In our study,
such errors would be possible either in the selection of cases
for referral to the regional anesthetic allergy clinic or in the
clinical evaluation of the cause of anaphylaxis.
Regarding potential referral errors, despite the single anesthetic allergy clinic in the Auckland region, it is possible that
some patients with anaphylactic reactions were not referred
from study hospitals for evaluation at the clinic. However,
this would represent a serious departure from mandated
practice at these institutions (or indeed from accepted anesthetic practice anywhere). Moreover, there is no convincing reason to suspect that any such departures would favor
one drug. One possible concern is that the well-understood
propensity for atracurium to cause histamine release may
have inclined anesthesiologists to overlook anaphylaxis of a
minor degree related to this agent, but the severity grading
of the identified reactions (table3) appears balanced across
agents and therefore does not support this hypothesis. We
do acknowledge that there may have been underreferral of
minor reactions to all agents, as there were few grade 1 reactions diagnosed in the study.
In respect of potential evaluation errors, the evaluation
of referred cases at the clinic followed a standard protocol
(outlined earlier) including application of a consistent case
definition, and the determination of causation for each case
reported in this study was independently reviewed by an
anesthetic allergy specialist and an immunologist. Although
it is acknowledged that there is variation in the way in which
the clinical histories, skin testing results, and other tests are
evaluated, there is widespread acceptance that all features
should be considered in diagnosis.19 All the relevant clinical features and available test results have been transparently
provided. It was not always possible to identify the agent on
skin testing even though there was a convincing history of
anaphylaxis. In 46 cases, the drug or substance was identified
on skin testing, but in 36 cases (44%), it was not. This reaction rate is consistent with other large surveys.1,3
Third, there is a small potential for inaccuracies in the
denominator data and the other associated data gathered
from the electronic database of anesthetic records. For example, the records rely on anesthesiologists entering details such
as the drug type and dose. These details are checked intraand postoperatively and a printout requiring a signature
by the anesthesiologist certifies this as a legal record of the
procedure. Previous research in our institution, which took

place during the period of the current study, demonstrated a
rate of omission of drug administration from the electronic
record of 2.31 per 100 drug administrations.20 Finally, we
did not formally account for the increased risk of family-wise
error rate by correcting our P values for multiple testing; this
topic is further explored in Supplemental Digital Content 1,
http://links.lww.com/ALN/B110.
In conclusion, we have used credible numerator and
denominator data to demonstrate similar rates of anaphylaxis
after administration of succinylcholine and rocuronium
these rates were approximately an order of magnitude
higher than those for atracurium and other nondepolarizing
NMBDs. Rocuronium remains a useful alternative to succinylcholine in rapid sequence inductions where succinylcholine is contraindicated, but its routine use as a muscle
relaxant in preference to other NMBDs deserves careful consideration, particularly, in regions where presensitization is
thought to be common.
Acknowledgments
The authors thank Ann Wills, F.A.N.Z.C.A., Department
of Anaesthesia and Perioperative Medicine, Auckland City
Hospital, Auckland, New Zealand, for her work in initiating
data collection at the Anaesthetic Allergy Clinic; and Roy
The, B.Sc., Immunopathology Department, Auckland City
Hospital, for preparing the dilutions and administering the
skin tests.
Competing Interests
Correspondence
Address correspondence to Dr. Reddy: Department of Anaesthesia, North Shore Hospital, 124 Shakespeare Road,
Takapuna, Auckland 0620, New Zealand. jeffrey.reddy@
waitematadhb.govt.nz. Information on purchasing reprints
may be found at www.anesthesiology.org or on the masthead page at the beginning of this issue. Anesthesiologys articles are made freely accessible to all readers, for personal
use only, 6 months from the cover date of the issue.
References
1. Mertes PM, Laxenaire MC, Alla F: Anaphylactic and anaphylactoid reactions occurring during anesthesia in France in
19992000. Anesthesiology 2003; 99:53645
2. Mertes PM, Aimone-Gastin I, Guant-Rodriguez RM, MoutonFaivre C, Audibert G, OBrien J, Frendt D, Brezeanu M,
Bouaziz H, Guant JL: Hypersensitivity reactions to neuromuscular blocking agents. Curr Pharm Des 2008; 14:280925
3. Gibbs NM, Sadleir PH, Clarke RC, Platt PR: Survival from
perioperative anaphylaxis in Western Australia 20002009.
Br J Anaesth 2013; 111:58993
4. Brereton A, Russell WJ: Anaphylaxis to muscle relaxants: An
audit of ten years of allergy testing at the Royal Adelaide
Hospital. Anaesth Intensive Care 2012; 40:8616
5. Laake JH, Rttingen JA: Rocuronium and anaphylaxisA statistical challenge. Acta Anaesthesiol Scand 2001; 45:1196203
6. Laxenaire MC, Mertes PM; Groupe dEtudes des Ractions

Anaphylactodes Peranesthsiques: Anaphylaxis during
anaesthesia. Results of a two-year survey in France. Br J
Anaesth 2001; 87:54958
7. Sadleir PH, Clarke RC, Bunning DL, Platt PR: Anaphylaxis
to neuromuscular blocking drugs: Incidence and cross-reactivity in Western Australia from 2002 to 2011. Br J Anaesth
2013; 110:9817
8. Mertes PM, Alla F, Trchot P, Auroy Y, Jougla E; Groupe
dEtudes des Ractions Anaphylactodes Peranesthsiques:
Anaphylaxis during anesthesia in France: An 8-year national
survey. J Allergy Clin Immunol 2011; 128:36673
9. Florvaag E, Johansson SG, Oman H, Venemalm L, Degerbeck
F, Dybendal T, Lundberg M: Prevalence of IgE antibodies to
morphine. Relation to the high and low incidences of NMBA
anaphylaxis in Norway and Sweden, respectively. Acta
10. Bhananker SM, ODonnell JT, Salemi JR, Bishop MJ: The risk
of anaphylactic reactions to rocuronium in the United States
is comparable to that of vecuronium: An analysis of food
and drug administration reporting of adverse events. Anesth
Analg 2005; 101:81922
11. Guttormsen AB: Allergic reactions during anaesthesia
Increased attention to the problem in Denmark and Norway.
Acta Anaesthesiol Scand 2001; 45:118990
12. Merry AF, Webster CS, Mathew DJ: A new, safety-oriented,
integrated drug administration and automated anesthesia
record system. Anesth Analg 2001; 93:38590
13. Fisher MM, Bowey CJ: Intradermal compared with prick testing in the diagnosis of anaesthetic allergy. Br J Anaesth 1997;
79:5963
14. Mertes PM, Malinovsky JM, Jouffroy L, Aberer W, Terreehorst

I, Brockow K, Demoly P; Working Group of the SFAR and
SFA; ENDA; EAACI Interest Group on Drug Allergy: Reducing
the risk of anaphylaxis during anesthesia: 2011 updated
guidelines for clinical practice. J Investig Allergol Clin
Immunol 2011; 21:44253
15. Lobera T, Audicana MT, Pozo MD, Blasco A, Fernndez E,
Caada P, Gastaminza G, Martinez-Albelda I, GonzlezMahave I, Muoz D: Study of hypersensitivity reactions and
anaphylaxis during anesthesia in Spain. J Investig Allergol
Clin Immunol 2008; 18:3506
17. Johansson SG, Florvaag E, Oman H, Poulsen LK, Mertes
PM, Harper NJ, Garvey LH, Gerth van Wijk R, Metso T,
Irgens A, Dybendal T, Halsey J, Seneviratne SL, Guttormsen
AB: National pholcodine consumption and prevalence
of IgE-sensitization: A multicentre study. Allergy 2010;
65:498502
18. Florvaag E, Johansson SG: The pholcodine story. Immunol
Allergy Clin North Am 2009; 29:41927
19. Ebo DG, Fisher MM, Hagendorens MM, Bridts CH, Stevens
WJ: Anaphylaxis during anaesthesia: Diagnostic approach.
Allergy 2007; 62:47187
20. Merry AF, Webster CS, Hannam J, Mitchell SJ, Henderson R,
Reid P, Edwards KE, Jardim A, Pak N, Cooper J, Hopley L,
Frampton C, Short TG: Multimodal system designed to reduce
errors in recording and administration of drugs in anaesthesia: Prospective randomised clinical evaluation. BMJ 2011;
343:d5543
Postoperative Bladder Catheterization Based

on Individual Bladder Capacity
A Randomized Trial
Tammo A. Brouwer, M.D., Peter F. W. M. Rosier, M.D., Ph.D., Karel G. M. Moons, Ph.D.,
Nicolaas P. A. Zuithoff, Ph.D., Eric N. van Roon, Ph.D., Cor J. Kalkman, M.D., Ph.D.
ABSTRACT
Background: Untreated postoperative urinary retention can result in permanent lower urinary tract dysfunction and can be
prevented by timely bladder catheterization. The author hypothesized that the incidence of postoperative bladder catheterization can be decreased by using the patients own maximum bladder capacity (MBC) instead of a fixed bladder volume of
500ml as a threshold for catheterization.
Methods: Randomized parallel-arm and single-blinded comparative effectiveness trial conducted in 1,840 surgical patients,
operated under general or spinal anesthesia without an indwelling urinary catheter. Patients were randomized to either use
their individual MBC (index) or a fixed bladder volume of 500ml (control) as a threshold for postoperative bladder catheterization. Preoperatively, the MBC was determined at home by voiding in a calibrated bowl. All other bladder volumes were
measured by ultrasound. Postoperatively, bladder catheterization was performed when spontaneous voiding was impossible,
and the ultrasound measurement exceeded the threshold for the group in which the patient was randomized (500 or MBC).
The primary outcome was the incidence of bladder catheterization.
Results: The average MBC in the control group was 582ml (199ml) and in the index group 611ml (209ml). The incidence of catheterization decreased from 11.8% (107 of 909 patients) in the control group to 8.6% (80 of 931) in the index
group (relative risk 0.73, 95% CI 0.55 to 0.96, P = 0.025). There were no adverse events in either group.
Conclusions: In patients undergoing surgery under general or spinal anesthesia using the MBC rather than a fixed 500ml
threshold for bladder catheterization is a safe approach that significantly reduces the incidence of postoperative bladder catheterizations. (Anesthesiology 2015; 122:46-54)
FTER almost any surgical procedure, the recurrence

of spontaneous voiding can be impaired. If the delay
of spontaneous voiding results in bladder distension beyond
the maximum bladder capacity (MBC), postoperative urinary retention (POUR) is present.1 In-and-out bladder
catheterization is the standard treatment for POUR.2,3 The
incidence of POUR is unknown; estimates vary between 5
and 70%, depending on the definition applied. Definitions
of POUR, as part of clinical protocols for postoperative
bladder catheterization, are often based on expert opinion
and vary between hospitals.46 If POUR is left untreated,
overdistension of the bladder wall may occur, which can
damage the detrusor muscle, potentially leading to a complete inability to void with the need for lifelong intermittent
self-catheterization.79 This is especially the case if the duration of the overdistension exceeds 2 to 3 hours.10,11

Postoperative urinary retention is a potential complication of
surgery and anesthesia
It is unclear at what volume we should consider bladder catheterization

In a prospective trial involving 1,840 patients, maximum bladder capacity was determined before surgery
Using predetermined maximum bladder capacity, the authors
demonstrated that a reduction in the need for postoperative
bladder catheterization could be achieved
In the Western world alone, more than 70 million people

undergo surgery each year. Even if the risks of serious permanent sequelae of POUR are very low, it would still affect
This article is featured in This Month in Anesthesiology, page 1A. Presented, in part, at the 15th World Federation of Societies of Anaesthesiologists World Congress of Anesthesiologists, March 28, 2012, Buenos Aires, Argentina.
Submitted for publication May 13, 2013. Accepted for publication August 12, 2014. From the Department of Anesthesiology, Medical
Center Leeuwarden, Leeuwarden, The Netherlands (T.A.B.); Department of Functional Urology, University Medical Center Utrecht, Utrecht,
The Netherlands (P.F.W.M.R.); Division of Perioperative Care and Emergency Medicine, Julius Center for Health Sciences and Primary Care,
University Medical Center Utrecht, Utrecht, The Netherlands (K.G.M.M.); Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands (N.P.A.Z.); Department of Pharmacotherapy and Pharmaceutical Care, University of Groningen,
Groningen, The Netherlands, and Department of Clinical Pharmacy and Pharmacology, Medical Center Leeuwarden, Leeuwarden, The Netherlands (E.N.v.R.); and Division of Anesthesiology, Intensive Care, and Emergency Medicine, University Medical Center Utrecht, Utrecht, The
Netherlands (C.J.K.).
January 2015
thousands of patients annually. However, reliable data on

how often patients develop permanent postoperative bladder
damage are scarce and mostly rely on case reports. On the
other hand, preemptive in-and-out catheterization in every
surgical patient to prevent overdistension is a procedure most
patients prefer to avoid. Bladder catheterization costs money,
nursing time, and bears its own risks (trauma, infection).12,13
To prevent bladder overdistension, portable bedside ultrasonography can be used to determine bladder volumes postoperatively.1416 Surprisingly, there is no evidence as to what
constitutes a safe maximum bladder volume in surgical
patients, but is it likely to vary between patients. Bladder volumes between 400 and 600ml are commonly used as thresholds for bladder catheterization to prevent POUR.46,17
In a previous study, measuring maximum bladder volumes
in volunteers, large bladder volumes with large interindividual variations (200ml to more than 1 l)independent of
age, sex, or heightwere observed (average bladder volume
567190ml).18 Therefore, we hypothesized that many surgical patients are catheterized too early when an arbitrary, but
commonly accepted limit for a maximum bladder volume of
500ml is used, instead of the patients individual MBC. But,
on the other hand, patients may also be catheterized too late
if their MBC is smaller than 500ml. Knowledge of a patients
individual MBC combined with postoperative serial ultrasound measurements of actual bladder volume might reduce
the incidence of POUR and will result in a more selective
use of perioperative bladder catheterization. Such practice
potentially reduces both overtreatment (unnecessary catheterizations and its attendant risk) and undertreatment (late
catheterization with the risk of an overdistended bladder).
The current randomized trial investigated the superiority
of using the patients individual MBC (index group) as a
threshold for postoperative bladder catheterization, as compared to a fixed volume of 500ml (control group).

Participants
Written informed consent was obtained from all patients.
All consecutive patients who visited the preanesthesia assessment clinic (PAC) at the Medical Center Leeuwarden (a large
referral hospital in The Netherlands) were invited to participate in this study. Eligible patients were clinical patients
undergoing elective surgical procedures and patients undergoing surgical procedures performed in day care. They were
18 yr or older and were planned for surgical intervention
under general or spinal anesthesia without the anticipated
need for an indwelling catheter preoperatively. Consenting
patients agreed to follow the instructions for measuring their
individual MBC at home, to be randomized to one of the
two parallel study arms, and to complete the questionnaires
as required by the study protocol.
This study was conducted in compliance with Helsinki
Guidelines (2008 revision) and started after approval from
the Ethical Review Board of the Medical Center Leeuwarden (protocol no. TPO 523) and after approval from the
Central Committee for Human Studies (Centrale Commissie voor Mensgebonden Onderzoek [CCMO] registered trial
database no: NL 21058.099.07 www.ccmmo.nl, The Hague,
The Netherlands). The study is registered in the Current
Controlled Trials database no: ISRCTN97786497 (http://
www.controlled-trials.com/ISRCTN97786497).
Measurements
Following the advice from our consulting urologist consenting patients were asked to completely empty their bladder at
their preoperative visit at the PAC to be able to measure their
residual bladder volumes by ultrasound (BladderScan; Verathon Inc., Bothell, WA). Ultrasound residual urine volume
measurements were performed by the PAC-nursing staff
and research assistants.19,20 All consenting patients received
a measuring bowl (Glaswarenfabrik Karl Hecht KG, 97647
Sondheim/Rhne, Bayern-Germany) and instructions how
to measure their maximum voided volumes at home. They
were asked to measure the maximum voided bladder volume
at least three times in the weeks before the operation: two
times during the day and once immediately after waking up
in the morning. They were instructed to postpone voiding
until a strong urge appeared. The feel of a strong urge is different for each patient, but we considered this voided volume
as the maximum bladder volume that could be determined
noninvasively in a stress-free environment and would still be
safe, without any burden for the patient. The MBC was then
calculated as the reported maximum voided volume at home
(in the calibrated bowl) plus the residual volume measured
at the PAC (by ultrasound).
It was unknown if catheterization based on the expected
larger bladder volume limit might harm bladder function. For
this safety reason, all patients were asked to complete two standard lower urinary tract symptoms questionnaires at home.
Both questionnaires had to be answered before surgery, but
were repeated also a day, a week, and a month after surgery
by telephone or e-mail. Any disturbanceschanges in voiding patternthat could have been caused by reaching larger
bladder volumes or by catheterization could then be recorded.
The first questionnaire was the International Prostate
Symptoms Score (IPSS) with questions about possible voiding complaints such as hesitancy, incomplete emptying, or
nighttime voiding. Each answer was recorded on a scale from
0 (no, does not happen) to 5 or 7 (happens all the time). The
second questionnaire was about the possible change, preand postoperatively, of lower urinary tract symptoms on the
Quality of Life (QoL).21,22
Randomization
A computer program randomly assigned 2,500 case record
form numbers in a 1:1 ratio to the control group or to the
index group (1,250 numbers for each group). Patients received
an information folder with their case record form number
Anesthesiology 2015; 122:46-54 47 Brouwer et al.
Individualized Postoperative Bladder Catheterization
printed, and a calibrated bowl to measure the voided volume.

The folder contained information about the study, including
the two questionnaires, instructions on how to perform the
MBC measurements at home, together with a form on which
to record the MBC measurements and an informed consent
form. After surgery, upon arrival at the postanesthesia care unit
(PACU), the patients case record form number was coupled to
either the control or to the index group according to the preoperatively performed randomization. The research assistant was
the only person aware of the group randomization.
Control Treatment
In the control group, a fixed volume of 500ml was set as
the threshold volume, above which catheterization was performed if the patient was unable to void spontaneously. Postoperatively, after arriving at the PACU, the patients bladder
was scanned by the research assistant. If the patient was not
able to void spontaneously or had no urge to void and the
scanned bladder volume was less than the 500ml, a subsequent scan was performed after 1h. The patients bladder
was scanned every hour at the PACU and on the way to the
surgical ward. If the patient was able to void spontaneously
before the 500ml limit was reached, prevoiding volume and
subsequently the residual volume was measured. However, if
a patient had a strong painful urge less than the bladder volume limit and was unable to void spontaneously, catheterization was performed to relieve the patient. When the scanned
bladder volume was larger than 500ml, POUR was diagnosed and patients were encouraged to void spontaneously.
If spontaneous voiding was not possible, bladder in-andout catheterization was performed, regardless if the patient
expressed sensation of urge.
Index Treatment
In the index group, the calculated individual MBC (maximum voided volume at home plus residual volume measured
at the PAC) was set as a threshold volume for catheterization. Postoperatively, the same procedure was used for the
control group, except for the different threshold definition.
In both study, arms anesthesia technique, type of local
anesthetic, and volume of fluids infused during the perioperative period were at the discretion of the attending anesthesiologist who was unaware of the group assignment. All
data were recorded in the case record form.
Outcome
The primary outcome was the incidence of bladder catheterization in each treatment group. Secondary outcomes
included incidence of POUR according to the bladder volume thresholds per group (control >500ml; index >MBC)
and the postoperative IPSS and QoL scores.
Another secondary outcome was the number of patients in
whom catheterization could have been avoided, which could
retrospectively only be determined in the control group.
Avoidable catheterization was defined as catheterization
in the presence of a measured bladder volume larger than

500ml, but smaller than the patients individual MBC.
Those patients would possibly not have been catheterized if
they were randomized in the index group where their MBC
would be used as threshold instead of 500ml. Finally, we
also calculated the number of patients in the control group
with MBCs less than 500ml in which bladder catheterization happened too late compared to their MBC, and the
number of patients in the MBC group with a MBC less than
500ml who would not have been catheterized if they had
been randomized in the 500-ml group.
In a pilot study, measuring ultrasound bladder volumes at
the PACU for a 1-month period, we recorded how often
patients had a measured bladder volume larger than 500ml
(this is the bladder volume limit used in our hospital protocol for POUR). Approximately 20% of the patients had
a measured bladder volume 500ml or greater. However,
the incidence of postoperative bladder catheterization at
the PACU differed per day (incidence varying between 0
and 10% per day) and depended not only on the measured
bladder volume, but also on the nurse or anesthesiologist
who took care of the patient. We did not register how often
patients were able to void spontaneously after reaching the
volume limit of 500ml or were postoperatively catheterized
at the surgical wards. This is the first study about the incidence of postoperative bladder catheterization, comparing a
fixed volume limit versus a variable volume limit (MBC) and
consisting of measurements at the PACU and at the wards.
But for our sample size estimation, we used only the data of
the incidence of reaching a bladder volume 500ml or greater
at the PACU. Accordingly, using the MBC approach for
POUR, with expected larger maximum allowable bladder
volumes, the incidence of POUR in the intervention group
was expected to decrease by half (difference 10%). To be able
to detect a more conservative absolute difference of 5%, a
level of significance of 0.05 (two sided), and power of 80%,
906 patients per group were needed to test the hypothesis.
All analyses were performed according to intention to
treat. The incidence of postoperative bladder catheterization
at the PACU (primary outcome) in the index and control
groups was compared by estimating the relative risk (RR)
with 95% CI and corresponding P value using the log-binomial regression model. Differences in IPSS and QoL scores
(secondary outcome) were estimated with median interquartile ranges at 1 day, 1 week, and 1 month after surgery separately. MannWhitney tests were used to calculate P values
for the differences between the groups.
Preplanned subgroup analyses consisted of two possible
risk factors for bladder catheterization and POUR: sex (male
vs. female) and anesthesia technique (general vs. spinal).
Post hoc subgroup analysis consisted of age, duration of surgery, and total fluid volume (both infused and taken orally
by the patient). P values for differences in treatment effects
in subgroups were estimated by adding both the risk factor

and the interaction between treatment and the risk factor to
the log-binomial regression model. Subsequently, RRs with
95% CIs for each subgroup were estimated.
Analyses were performed using SPSS version 17 (IBM
Corporation, Somer, NY).
Blinding
Only the research assistant, who performed all ultrasound
measurements, knew in which group the patient was randomized. During the entire study, anesthesia team, nursing
staff, patients were not aware of the group allocation and
the result of the ultrasound measurements (single blinded).
The PACU nurseor after discharge from the PACU, the
ward nursewho was responsible for the patient, was asked
by the research assistant to perform bladder catheterization
when the patients scanned bladder volume was larger than
500ml (control group) or when the bladder volume was
larger than the patients MBC (index group) and the patient
was unable to void spontaneously.
Results
Recruitment
Between May 2008 and June 2009, 4,500 consecutive surgical patients were asked to participate in the study, after
receiving approval from the Ethical Review Board. Ultimately, 1,840 patients who fulfilled the inclusion criteria
and who gave informed consent participated in the study
(fig.1). They were analyzed for the primary endpoint bladder catheterization. Fewer patients (n = 1,792) were available
to compare the incidence of POUR. Twenty-six patients in
the control group and 22 patients in the index group had
to be excluded because their bladder volumes could not be
measured, as they had voided before the bladder was scanned
(n = 19 in the control group and n = 18 in the index group,
respectively). In another 11 patients scanning was impossible
because the abdominal region above the pubic bone could
not be reached (n = 7 vs. n = 4 patients, respectively). Thirty
patients could not be reached by telephone or e-mail for collecting the secondary outcomes IPSS en QoL scores (n = 14
vs. n = 16 patients, respectively) leaving 1,762 to be analyzed
for this outcome. Demographic and clinical parameters of
the patients were comparable in the two groups (table1).
Outcomes
Bladder Capacity. In the control group, the average MBC
was 582ml with an SD of 199ml. The average MBC in the
index group was 611ml with an SD of 209ml.
Primary Outcome
The incidence of bladder catheterization in the control group
was 11.8% (107 of 909 patients) compared to 8.6% (80 of
931 patients) in the index group (RR, 0.73; 95% CI, 0.55 to
0.96; P = 0.025) (table2). The risk difference was 3.2 (11.8
to 8.6) and the number needed to treat was 31.
In the control group 43% (n = 376) of the patients

reached the 500ml bladder volume threshold (POUR
present) of which 72.6% voided spontaneously and 27.4%
were catheterized because of the inability to void spontaneously. In the index group 32% (n = 288) of the patients
had POUR, of which 77.4% voided spontaneously and
22.6% had to be catheterized (RR, 0.82; 95% CI, 0.63 to
1.08; P = 0.160) (table3).
In evaluating this outcome across the above-mentioned
subgroups, all tests for interactions were nonsignificant
(table 2). However, male patients in the control group were
catheterized more often compared to the index group, 14.5
versus 9.5%. For female patients this was 9.6% versus 7.8%,
respectively (P = 0.48). In patients receiving spinal anesthesia, the overall incidence of bladder catheterization decreased
from 24.4% in the control group to 16.3% in the index
group. For patients receiving general anesthesia this was 5.9
versus 5.0%, respectively.
In post hoc analyses also no significant interaction
between treatment and age (P = 0.10), duration of surgery
(0.89), and total volume infused (0.64) was observed. Of
note, when scanned bladder volume on arrival at the PACU
was greater than 250ml (196 patients in the control and 214
in the index group), the likelihood of subsequent bladder
catheterization was higher (33% in the control vs. 18% in
the index group) than when bladder volume was less than
250ml (5.5 vs. 5.6%, respectively).
Secondary Outcomes
Avoidable Catheterizations. In the control group, 37.1%
(39of 105 of the patients) would not have been catheterized
if they were randomized in the MBC group. Their measured
bladder volume was smaller than their MBC (but >500ml).
Also in the control group, 25 (7.9%) patients with MBC
smaller than 500ml were catheterized too late according to
their MBC. Finally, in the MBC group, 36 (12.3%) patients
had a MBC smaller than 500ml and were catheterized because
they had reached their MBC value and were unable to void
spontaneously despite a scanned bladder volume smaller than
500 ml. These patients would likely not have been catheterized if they had been randomized in the 500-ml group.
International Prostate Symptoms and QoL Score
Median IPSS and QoL scores of the control and index group
at 1 day, 1 week, and 1 month after surgery were very similar
(table4). No significant differences were observed.
Discussion
This large randomized study is the first to address POUR
and the need for bladder catheterization from the perspective of the patients individually measured MBC in a large
cohort of surgical patients. Using the MBC (index group)
instead of a fixed bladder volume of 500ml (control group),
the absolute incidence of bladder catheterization decreased
from 11.8 to 8.6% (absolute difference 3.2) or a number
needed to treat of 31.
Fig. 1. Flow diagram of the patients through the phases of the randomized trial. IPSS = International Prostate Symptoms Score;
MBC = Maximum Bladder Capacity; POUR = Post Operative Urinary Retention; QoL = Quality of Life Score.
In the subgroup analyses, we found no significant effect

modification by sex or anesthesia technique. Nonetheless, we
observed some differences in bladder catheterization across
subgroups that may be clinically meaningful, but these will
need further study to confirm or refute their relationship with
the intervention. For example, using the MBC as threshold
appeared to decrease the incidence of bladder catheterization
more often in male patients and in patients receiving spinal
anesthesia. Adequately powered studies in selected subgroups
of surgical patients are needed to determine if these differences are statistically and clinically relevant.
This study also shows that most patients who had
reached their volume limit were able to void spontaneously, although in the index group less patients reached
the threshold than in the control group (table 3). However,

the proportion of patients who were able to void spontaneously was similar in both groups (approximately 75%).
As observed earlier,15 patients arriving at the PACU with
a bladder volume 250ml or greater were at higher risk for
POUR and bladder catheterization.
In the current study, the average MBC (approximately
600ml) for all included patients was larger than the normally used threshold volume for catheterization of 500ml,
suggesting that the current 500ml limit may be too conservative. More than 66% of the included patients had a MBC
larger than 500ml (fig.2). Male patients had a slightly larger
average MBC than female patients (619 vs. 579ml). However, this sex difference can likely be considered clinically
Table 1. Demographic and Clinical Characteristics of Patients

across the Two Treatment Groups (1,840 Patients)
Control
Group
n = 909
MBC Group
n = 931
Patient Data
Women, No. (%)
508 (56)
490 (53)
Age, mean (SD), yr
48.5 (15)
47.9 (15)
Height, mean (SD), cm
175 (10)
176 (10)
Weight, mean (SD), kg
80.1 (16)
81.4 (17)
BMI, mean (SD), kg/m2
26.2 (5)
26.3 (5)
Type of Surgery, No. (%)
Head/neck
219 (24)
203 (22)
Thoracic/breast
85 (9)
75 (8)
Spine
31 (3)
36 (4)
Abdominal
231 (25)
272 (29)
Extremities
343 (38)
345 (37)
Study Data
MBC, mean (SD), ml
582 (199)
611 (209)
Residual volume, mean (SD), ml
33 (61)
33 (53)
Voided before surgery, No. (%)
854 (94)
874 (94)
Time before surgery, mean (SD), min
59 (59)
59 (48)
Volume at holding, mean (SD), ml
44 (69)
52 (81)
General anesthesia, No. (%)
622 (68)
636 (68)
Spinal anesthesia, No. (%)
287 (32)
295 (32)
Articaine, No. (%)
213 (74)
234 (79)
Bupivacaine, No. (%)
74 (26)
61 (21)
Total volume infused, mean (SD), ml 1,475 (580) 1,492 (647)
Procedure time, mean (SD), ml
61 (37)
61 (40)
BMI = body mass index; MBC = maximum bladder capacity.
irrelevant, bearing in mind the 7% underestimation of the

ultrasound device.18 The average maximum scanned postoperative bladder volume for patients who had reached the
POUR criteria was even larger: 660ml in the control group
and 648ml in the index group. Obviously, fewer patients
would have been catheterized if we had selected 600ml as
the threshold in the control group.
For both the control and index group, the need for postoperative bladder catheterization after spinal anesthesia
was substantially higher than after general anesthesia (table
2). In patients who had received bupivacaine, the absolute
probability of being catheterized postoperatively decreased
from 63% in the control group to 44% in the index group
(RR = 0.71; table5). It is known that local anesthetics used
for spinal anesthesia can impair bladder function for several
hours.23,24 Still, no prospective randomized controlled trials
have compared the incidence of POUR and bladder catheterization after general versus spinal anesthesia. The current
study shows that for both groups, the probability of being
catheterized after spinal anesthesia, using local anesthetics
without adding opioids, is about two times higher than after
general anesthesia when using the short acting local anesthetic articaine and about 10 times higher when using the
long acting local anesthetic bupivacaine.
The large absolute difference between the control and
index group after spinal anesthesia (8.1%) can be explained
by the fact that in the index group bladder volume thresholds were typically larger than 500ml. This larger volume
limit gave patients more time to reach the threshold and
allowed more time for regression of the spinal block. The
ability to void spontaneously may then have returned before
the MBC was reached, leading to a reduction in the number
of bladder catheterizations.
None of the differences in IPSS and QoL scores reached
statistical significance (table 4). Patients who reported a
relatively high IPSS preoperatively also had high scores
postoperatively and vice versa. Bladder damage is likely to
occur only when the bladder has been distended for a longer
period of time (2 to 3h).911 None of our patients had a
distended bladder for more than an hour, and we postulate
that no serious complication of the lower urinary tract could
or had occurred in any of the studied patients because we
used a strict protocol of measuring bladder volumes each
Table 2. Incidence of Bladder Catheterization in the Preplanned and Post Hoc Subgroup Analyses
Females
Males
General
Spinal
Age 60 yr
Age<60 yr
Volume 1.5 l
Volume < 1.5 l
OR time 60 min
OR time < 60 min
Control Group
n = 909
Index Group
n = 931
Catheterization
Catheterization
RR (95% CI)
P Value*
49/508 (9.6)
58/401 (14.5)
37/622 (5.9)
70/287 (24.4)
42/236 (17.8)
65/673 (9.7)
46/381 (12.1)
61/528 (11.6)
67/378 (17.7)
40/531 (7.5)
38/490 (7.8)
42/441 (9.5)
32/636 (5.0)
48/295 (16.3)
39/230 (17.0)
41/701 (5.9)
33/404 (8.2)
47/527 (8.9)
48/371 (12.9)
32/560 (5.7)
0.80 (0.541.21)
0.66 (0.540.96)
0.85 (0.531.34)
0.67 (0.480.93)
0.95 (0.641.42)
0.61 (0.420.88)
0.68 (0.441.03)
0.77 (0.541.11)
0.73 (0.521.03)
0.76 (0.481.19)
0.477
0.411
0.104
0.643
0.894
Numbers are expressed as numbers and percentages.

*P values for interaction.
Control group = threshold bladder volume 500ml; index group = threshold bladder volume maximum bladder capacity; OR time = duration of surgery;
RR = relative risk; volume = total volume infused or taken till spontaneous voiding/catheterization.
Table 3. Patients Who Reached POUR: Voiding Spontaneously

or Were Catheterized (n = 1,792)
Voiding less than threshold

Catheterized less than threshold
Reaching volume threshold (=POUR)
Spontaneous voiding
Catheterization*
Control
Group
n = 883
Index
Group
n = 909
505 (56.8)
2 (0.2)
376 (43)
273 (72.6)
103 (27.4)
611 (67)
10 (1)
288 (32)
223 (77.4)
65 (22.6)
Data are expressed as numbers and percentages.

*Nonsignificant (relative risk, 0.82; 95% CI, 0.63 to 1.08; P = 0.160).
Control group = threshold bladder volume 500ml; index group = threshold bladder volume maximum bladder capacity; POUR = postoperative
urinary retention.
Table 4. IPSS and QoL Scores Pre- and Postoperatively

(n = 1,762)
Control Group Index Group
n = 869
n = 893
Median (IQR) Median (IQR) P Value
IPSS score
Preoperative
1 day after surgery
1 week after surgery
1 month after surgery
QoL score
Preoperative
1 day after surgery
1 week after surgery
1 month after surgery
5.0 (2 to 9)
5.0 (2 to 9)
4.0 (1 to 8)
4.0 (1 to 8)
5.0 (2 to 9)
5.0 (1 to 9)
4.0 (1 to 7)
4.0 (1 to 8)
0.52
0.83
0.37
0.32
1.0 (0 to 2)
1.0 (0 to 2)
1.0 (0 to 2)
1.0 (0 to 2)
1.0 (0 to 2)
1.0 (0 to 2)
1.0 (0 to 2)
1.0 (0 to 2)
0.51
0.95
0.43
0.35
P values were based on MannWhitney U test.

IPSS = International Prostate Symptom Score; IQR = interquartile ranges;
QoL = Quality of Life.
hour until spontaneous voiding or bladder catheterization

had occurred.
We anticipated that the method of assessing maximum
voided bladder volume at home would be safe and reproducible and expected that it could be accomplished with minimal patient burden. From a subset of 822 patients who were
specifically asked about their experiences with the MBC
home measurement procedure, only three patients indicated
that they would have problems when asked to perform it
again before an operation, and eight patients found voiding
in a measuring bowl difficult. These results were irrespective
of age and sex. All patients were motivated to perform the
estimation of their MBC at home, if it could help to prevent
unnecessary bladder catheterization.
A limitation of our pragmatic study design is that several
other factors that potentially can influence the incidence of
POUR were not standardized. The selection of anesthetic
technique and type of local anesthetic used for spinal anesthesia, as well as the amount of fluids infused were left at
the discretion of the anesthesia team. Age, history of voiding
problems, and type of operation all can influence the incidence and frequency of catheterization.46 Still, as a result
Fig. 2. Ranges of maximum bladder capacities of all patients

(n = 1,840) calculated as maximum voided volume + residual
volume.
of randomization and a sufficiently large sample size, the

two study groups were well balanced with respect to demographic variables and clinical characteristics (table 1).
Of 4,500 patients asked to participate, only 1,840 were
included the study. For almost 1,900 patients the main reason not to participate was mention of the word bladder
catheterization and explanation about the aim of the study
could not change their mind. This suggests that an element
of denial may have been present. Other stated reasons for
refusal were that it would be too cumbersome or it would
cost too much time (fig.1).
By measuring MBC at home in a large sample of surgical patients and measuring IPSS en QoL scores pre- and
postoperatively, we were able to establish safe bladder volume ranges. None of the patients had significantly worsened
postoperative IPSS or QoL scores. Only three patients developed a urinary tract infection the first month after the operation. Taken together, this suggests that a regimen consisting
of serial postoperative ultrasound measurements can prevent
overdistension and bladder damage. The current study also
suggests that patients older than 60 yr will benefit less from
measuring their MBC at home (which may be a more cumbersome procedure for them).
For patients younger than 60 yr who are at risk for
bladder catheterization, measuring MBC preoperatively
will decrease the chance of being catheterized postoperatively (table 2).
In summary, this large randomized study shows that
measuring MBC preoperatively can decrease the incidence
of postoperative bladder catheterization. Asking elective surgical patients, who may be at risk for postoperative bladder catheterization, to determine their MBC at home is a
low cost, low-tech approach to reduce the potential need for
postoperative bladder catheterization. Taking into account
the very large number of surgical patients operated every
year millions of unnecessary too early catheterizations can
Table 5. Spinal Anesthesia: Articaine vs. Bupivacaine (n = 1,840)
Total
Articaine
Bupivacaine
Control Group
n = 287
Index Group
n = 295
Catheterization
Catheterization
70/287 (24.4)
24/213 (11.3)
46/74 (62.2)
48/295 (16.3)
19/234 (8.1)
27/61 (44.3)
RR (95% CI)
P Value
0.67 (0.48 to 0.93)

0.74 (0.39 to 1.40)
0.71 (0.42 to 0.95)
0.01*
0.35
0.03*
Numbers are expressed as numbers and percentages.

*Significant.
Control group = threshold bladder volume 500ml; index group = threshold bladder volume maximum bladder capacity; RR = relative risk.
safely be prevented when this strategy is implemented in the

relevant clinical practice guidelines.
Acknowledgments
The authors thank all patients for their participation, their
willingness to measure their maximum bladder volumes at
home, and answering the International Prostate Symptoms
Score and Quality of Life questionnaire through the telephone and e-mail. The authors also thank the nursing staff
at the Preoperative Anesthesiology Screening for their help
and support during the study; the research assistants for
their enthusiasm, good spirits, and support; all surgeons;
anesthesiologists; and operating room nursing staff for their
help and patience during the study. The authors thank Nic
J. G. M. Veeger, Ph.D. (Department MCL Academia, Medical
Center Leeuwarden The Netherlands), and Jaap de Vries,
M.D., Ph.D. (Heart Center, Medical Center Leeuwarden),
Leeuwarden, for their critical review and support during the
revision of the article. The authors had full access to all data
in the study and they take responsibility for the integrity
of the data and the accuracy of the data analysis. Verathon
Europe (IJsselstein, The Netherlands) supplied devices
for noninvasive bladder volume measurements during the
study period. They also supplied measuring bowls to allow
patients to measure bladder volumes at home.
Supported by the Department of Science, MCL Academia,
Medical Center, Leeuwarden, The Netherlands (10,000
Euro), and Verathon Europe, IJsselstein, The Netherlands
(25,000 Euro).
Competing Interests
Dr. Brouwer received an allowance for travel and accommodation from Verathon (IJsselstein, The Netherlands) for
lecturing on postoperative bladder retention for colleagues,
but no speaking or consultancy fees. All other authors have
no relationship with industry. The research assistants and
4th year medical students received travel allowances and
were paid for the number of hours worked. The other authors declare no competing interests.
Correspondence
Address correspondence to Dr. Brouwer: Department of
Anesthesiology, Medical Center Leeuwarden, Henri Dunantweg 2, P. O. Box888, 8901 BR Leeuwarden, The Netherlands. t.brouwer@znb.nl. This article may be accessed
for personal use at no charge through the Journal Web site,
www.anesthesiology.org.
References
1. Hinman F: Editorial: Postoperative over distension of the
bladder. Surg Gynaecol Obstet 1976; 142960:9012
2. Higgins PM, Karia SJ, Mehta K: The management of acute
urinary retention. Br J Urol 1981; 5:3448
3. Beadnell SW, Connaughton B, Farhood VW: Management of
postoperative urinary retention. J Oral Maxillofac Surg 1989;
47:130710
4. Brouwer TA: Rtention durines et anomalies de la miction
en postopratoire. [Postoperative urinary retention and voiding disturbances] Le Praticien en Anesthsie-Reanimation
2003:12935
5. Baldini G, Bagry H, Aprikian A, Carli F: Postoperative urinary retention: Anesthetic and perioperative considerations.
6. Darrah DM, Griebling TL, Silverstein JH: Postoperative urinary retention. Anesthesiol Clin 2009; 27:46584
7. Tammela T, Kontturi M, Lukkarinen O: Postoperative urinary
retention II: Micturition problems after the first catheterisation. Scand J Urol Nephrol 1986; 20:25760
8. Tammela T, Kontturi M, Lukkarinen O: Postoperative urinary
retention. I. Incidence and predisposing factors. Scand J Urol
Nephrol 1986; 20:197201
9. Tammela T: Postoperative urinary retention: Why the patient
cannot void. Scand J Urol Nephrol Suppl 1995; 175:757
10. Carpenter FG: Impairment and restoration of rat urinary

bladder responsiveness following distension. Am J Physiol
1983; 244:R10613
11. Gosling JA, Kung LS, Dixon JS, Horan P, Whitbeck C, Levin
RM: Correlation between the structure and function of the
rabbit bladder following partial outflow obstruction. J Urol
2000; 163:134956
12. Lau H, Lam B: Management of postoperative urinary retention: A randomized trial of in-out versus overnight catheterization. ANZ J Surg 2004; 74:65861
13. Slappendel R, Weber EWG: Non-invasive bladder volume measurement reduces the use of bladder catheters and urinary
tract infections in orthopaedic surgery. Eur J Anaesthesiol
1999; 16:5036
14. Lamonerie L, Marret E, Deleuze A, Lembert N, Dupont M,
Bonnet F: Prevalence of postoperative bladder distension
and urinary retention detected by ultrasound measurement.
Br J Anaesth 2004; 92:5446
15. Keita H, Diouf E, Tubach F, Brouwer T, Dahmani S, Mantz J,
Desmonts JM: Predictive factors of early postoperative urinary retention in the postanaesthesia care unit. Anesth Analg
2005; 101:5926
16. Joelsson-Alm E, Nyman CR, Lindholm C, Ulfvarson J, Svensen
C: Perioperative bladder distension: A prospective study.
Scand J Urol Nephrol 2008; 31:15
17. Pavlin DJ, Pavlin EG, Gunn HC, Taraday JK, Koerschgen ME:
Voiding in patients managed with or without ultrasound
monitoring of bladder volume after outpatient surgery.

Anesth Analg 1999; 89:907
18. Brouwer TA, Eindhoven BG, Epema AH, Henning RH: Validation
of an ultrasound scanner for determing urinary volumes in surgical patients and volunteers. J Clin Monit Comput 1999; 15:37985
19. Rosseland LA, Stubhaug A, Breivik H: Detecting postop
erative urinary retention with an ultrasound scanner. Acta
20. Pedersen LM, Mantoni T, Lynggaard MD, Schousboe BM,

Lauritzen JB, Pedersen BD, Jrgensen HL: [Postoperative urinary retention. Clinical assessment versus ultrasound examination]. Ugeskr Laeger 2007; 169:6058
21. Sarasin SM, Walton MJ, Singh HP, Clark DI: Can a urinary
tract symptom score predict the development of postoperative urinary retention in patients undergoing lower limb
arthroplasty under spinal anaesthesia? A prospective study.

Ann R Coll Surg Engl 2006; 88:3948
22. Elkhodair S, Parmar HV, Vanwaeyenbergh J: The role of the
IPSS (International Prostate Symptoms Score) in predicting
acute retention of urine in patients undergoing major joint
arthroplasty. Surgeon 2005; 3:635
23. Luger TJ, Garoscio I, Rehder P, Oberladsttter J, Voelckel W:
Management of arthroscopic knee surgery in low-dose spinal anaesthesia: Development of a simple algorithm. Arch
Orthop Trauma Surg 2008; 128:60712
24. Kamphuis ET, Ionescu TI, Kuipers PW, de Gier J, van

Venrooij GE, Boon TA: Recovery of storage and emptying
functions of the urinary bladder after spinal anesthesia with
lidocaine and with bupivacaine in men. Anesthesiology
1998; 88:3106
From Paris, France, in 1868, Furne, Jouvet and Company published the second volume of author
Louis Figuiers Les Merveilles de la Science ou Description Populaire des Inventions Modernes [The
Wonders of Science or Popular Description of Modern Inventions]. This volume focused on telegraphy,
electroplating, ballooning, and etherization. In Chapter 3, on page 645, figure 341 (left) depicts an
anguished medical studentturneddental patient seated next to the forlorn standing figure of dentist
Horace Wells (close up, right). The illustrations legend translates to: Horace Wels [sic] experience,
of the extraction of a tooth, after the inspiration of nitrogen protoxyd [nitrous oxide] made before the
students of a Boston hospital. This incomplete anesthetic would later be dubbed the Humbug
Affair by Wells detractors. (Copyright the American Society of Anesthesiologists, Inc.)
Ohio. UJYC@aol.com.
monitoring of bladder volume after outpatient surgery.

18. Brouwer TA, Eindhoven BG, Epema AH, Henning RH: Validation
of an ultrasound scanner for determing urinary volumes in surgical patients and volunteers. J Clin Monit Comput 1999; 15:37985
19. Rosseland LA, Stubhaug A, Breivik H: Detecting postop
erative urinary retention with an ultrasound scanner. Acta
20. Pedersen LM, Mantoni T, Lynggaard MD, Schousboe BM,

Lauritzen JB, Pedersen BD, Jrgensen HL: [Postoperative urinary retention. Clinical assessment versus ultrasound examination]. Ugeskr Laeger 2007; 169:6058
21. Sarasin SM, Walton MJ, Singh HP, Clark DI: Can a urinary
tract symptom score predict the development of postoperative urinary retention in patients undergoing lower limb
arthroplasty under spinal anaesthesia? A prospective study.

Ann R Coll Surg Engl 2006; 88:3948
22. Elkhodair S, Parmar HV, Vanwaeyenbergh J: The role of the
IPSS (International Prostate Symptoms Score) in predicting
acute retention of urine in patients undergoing major joint
arthroplasty. Surgeon 2005; 3:635
23. Luger TJ, Garoscio I, Rehder P, Oberladsttter J, Voelckel W:
Management of arthroscopic knee surgery in low-dose spinal anaesthesia: Development of a simple algorithm. Arch
Orthop Trauma Surg 2008; 128:60712
24. Kamphuis ET, Ionescu TI, Kuipers PW, de Gier J, van

Venrooij GE, Boon TA: Recovery of storage and emptying
functions of the urinary bladder after spinal anesthesia with
lidocaine and with bupivacaine in men. Anesthesiology
1998; 88:3106
From Paris, France, in 1868, Furne, Jouvet and Company published the second volume of author
Louis Figuiers Les Merveilles de la Science ou Description Populaire des Inventions Modernes [The
Wonders of Science or Popular Description of Modern Inventions]. This volume focused on telegraphy,
electroplating, ballooning, and etherization. In Chapter 3, on page 645, figure 341 (left) depicts an
anguished medical studentturneddental patient seated next to the forlorn standing figure of dentist
Horace Wells (close up, right). The illustrations legend translates to: Horace Wels [sic] experience,
of the extraction of a tooth, after the inspiration of nitrogen protoxyd [nitrous oxide] made before the
students of a Boston hospital. This incomplete anesthetic would later be dubbed the Humbug
Affair by Wells detractors. (Copyright the American Society of Anesthesiologists, Inc.)
Ohio. UJYC@aol.com.
Accuracy of Malignant Hyperthermia Diagnoses in

Hospital Discharge Records
Teeda Pinyavat, M.D., Henry Rosenberg, M.D., Barbara H. Lang, M.P.H., Cynthia A. Wong, M.D.,
Sheila Riazi, M.D., Joanne E. Brady, Ph.D., Lena S. Sun, M.D., Guohua Li, M.D., Dr.P.H.
ABSTRACT
Background: In 1997, the International Classification of Diseases (ICD), 9th Revision Clinical Modification (ICD-9) coding
system introduced the code for malignant hyperthermia (MH) (995.86). The aim of this study was to estimate the accuracy
of coding for MH in hospital discharge records.
Methods: An expert panel of anesthesiologists reviewed medical records for patients with a discharge diagnosis of MH based
on ICD-9 or ICD-10 codes from January 1, 2006 to December 31, 2008 at six tertiary care medical centers in North America.
All cases were categorized as possible, probable, or fulminant MH, history of MH (family or personal) or other.
Results: A total of 47 medical records with MH diagnoses were reviewed; 68.1% had a documented surgical procedure and
general anesthesia, and 23.4% (95% CI, 12.338.0%) had a possible, probable, or fulminant MH event. Dantrolene was
given in 81% of the MH events. All patients judged to have an incident MH event survived to discharge. Family and personal
history of MH accounted for 46.8% of cases. High fever without evidence of MH during admission accounted for 23.4%,
and the reason for MH coding was unclear in 6.4% of cases.
Conclusions: Approximately one quarter of ICD-9 or ICD-10 coded MH diagnoses in hospital discharge records refer to
incident MH episodes and an additional 47% to MH susceptibility (including personal history or family history). Information such as surgical procedure, anesthesia billing data, and dantrolene administration may aid in identifying incident MH
cases among those with an ICD-9 or ICD-10 coded MH diagnosis in their hospital discharge records. (Anesthesiology
2015; 122:55-63)
ALIGNANT hyperthermia (MH) is a rare pharmacogenetic disorder of skeletal muscle metabolism. When
a susceptible patient is exposed to a triggering agent (halogenated volatile anesthetic agent and/or succinylcholine) or event
(such as elevated environmental heat along with exercise), a
potentially fatal hypermetabolic reaction can occur.1 In 1997,
the International Classification of Diseases (ICD), 9th Revision
Clinical Modification coding system (ICD-9) added a code for
malignant hyperthermia: malignant hyperpyrexia due to anesthesia (995.86). The availability of a special ICD code for MH
has made it possible to conduct epidemiological research related
to MH, including use of the code to evaluate MH prevalence,
trends in MH occurrence, and coexisting diseases.24
However, the validity of the findings from studies based
on ICD-coded data has been a serious concern. ICD codes
are prone to error at several different levels including but
not limited to physician, coding, and sequencing errors.5,6
Physician errors occur when a diagnosis is omitted from
the record, made incorrectly, or improperly recorded using
the wrong terminology. Coding errors can then occur when
data abstracters, usually nonmedical personnel, misinterpret

The accuracy of International Classification of Diseases (ICD)
coding for the purpose of registry research has been questioned in several fields
An ICD code for malignant hyperthermia was created in the
9th edition, but accuracy in its use has not been assessed

In review by an expert panel of ICD coding for malignant hyperthermia over a 3-yr period, approximately 70% of coded cases
were considered to be malignant hyperthermia susceptible
The most common reason for inaccurate coding was high fever unrelated to anesthesia
or incompletely review the chart and make incorrect decisions about which diagnoses to code.6 Sequencing errors are
defined as improper assignment of a primary diagnosis versus a secondary diagnosis.5
Studies on the accuracy of ICD coded diagnoses vary
widely in their findings due to the fact that each diagnosis and procedure is prone to its own unique rate and type
Preliminary results were presented at the American Society of Anesthesiologists annual meeting, Washington, D.C., October 16, 2012, and
at the International Anesthesia Research Society annual meeting, Montreal, Quebec, Canada, May 1720, 2014.
Submitted for publication May 14, 2014. Accepted for publication August 13, 2014. From the Department of Anesthesiology, Columbia
University College of Physicians and Surgeons, New York, New York (T.P., B.H.L.); Department of Medical Education and Clinical Research,
Saint Barnabas Medical Center, Livingston, New Jersey (H.R.); Department of Anesthesiology, Northwestern University Feinberg School of
Medicine, Chicago, Illinois (C.A.W.); Department of Anesthesiology, Toronto General Hospital, Toronto, Ontario, Canada (S.R.); Departments
of Anesthesiology and Epidemiology, Columbia University College of Physicians and Surgeons and Mailman School of Public Health, New
York, New York ( J.E.B., G.L.); and Departments of Anesthesiology and Pediatrics, Columbia University College of Physicians and Surgeons,
New York, New York (L.S.S.).
January 2015
Accuracy of ICD Code for Malignant Hyperthermia
of errors. Factors influencing the accuracy of ICD-coded

diagnoses include disease type, current state of medical
knowledge and technology, clinical acumen, participation
of various practitioners from patient to administrative personnel, and implementation of coding practices. Codes are
most likely to be accurate when the disease is clearly defined
with observable signs and symptoms, physician experts
record the diagnosis in the chart, experienced coders with
access to all patient data assign the code, and the code is
not new.7 One study of 7,050 Medicare discharges from a
national sampling of U.S. hospitals found that while the
overall accuracy was 78.2%, it varied widely across different
conditions. The positive predictive value in this study was
highest for hip fractures (94%) and lowest for peripheral
vascular disease (50%).8
We conducted a critical review of a sample of hospitalizations containing the ICD code for MH at discharge to
estimate the accuracy of coding, identify common errors,
and develop strategies to improve the identification of incident MH episodes using ICD and other administrative data.
The medical records reviewed in this study were for the 3-yr
period from 2006 to 2008, the most recent years for which
electronic medical records were available at all the study sites
when the study was initiated in 2009.

This study meets the criteria for the Protection of Human
Subjects exemption 4 (research involving preexisting data) of
the U.S. Code of Federal Regulations (45 CFR 46.101). A
formal application was prepared and submitted to each of the
site institutional review boards (IRB) (Columbia University
Medical Center Institutional Review Board, New York, New
York; Saint Barnabas Medical Center Institutional Review
Board, Livingston, New Jersey; Northwestern University
Institutional Review Board, Chicago, Illinois; The Childrens
Hospital of Philadelphia Institutional Review Board, Philadelphia, Pennsylvania; New York University Langone Medical Center/School of Medicine Institutional Review Board,
New York, New York; University Health Network Research
Ethics Board, Toronto, Ontario, Canada), and the study was
either deemed exempt from full review by the IRB chair or
approved after review based on local IRB policies. Written
consent was waived by the IRB at all sites.
Design and Setting
For this multicenter study, hospital billing records with a
primary or secondary diagnosis of MH (ICD-9 995.86 or
ICD-10 T88.3) were identified at six academic medical
centers (Columbia University Medical Center and Morgan
Stanley Childrens Hospital, New York, New York; Saint
Barnabas Medical Center, Livingston, New Jersey; Northwestern Memorial Hospital, Chicago, Illinois; Childrens
Hospital of Philadelphia, Philadelphia, Pennsylvania; New
York University Medical Center, New York, New York; University Health Network, Toronto, Ontario, Canada).
Study Population and Data Collection

The electronic databases from the participating hospitals
were searched to identify patients with the discharge diagnosis of MH (ICD-9 995.86 or ICD-10 T88.3) for a 3-yr
period, between January 1, 2006 and December 31, 2008.
For each identified hospitalization, the medical record was
reviewed including all notes, labs, and all available clinical data from both paper and electronic sources. Other
hospitalizations for the same patient were not included in
the review.
The chart reviews were conducted by an expert panel
consisting of five anesthesiologists who are MH Hotline
consultants and have been trained in human subjects protection and the Health Insurance Portability and Accountability Act (Teeda Pinyavat, Henry Rosenberg, Cynthia A.
Wong, Sheila Riazi, Lena S. Sun). The records were anonymized by a nonpanelist coinvestigator or research coordinator trained in privacy policies and practices prior to review
by the expert panel. Two panelists independently reviewed
each record and completed a standard data abstraction form
individually (appendix).
If, in the judgment of the panelist, an adverse metabolic
reaction occurred, the MH clinical grading scale (CGS) was
calculated.9 Although the CGS does rely on an anesthesiologists judgment, it provides a validated means to quantify
and standardize the diagnosis. In addition to the CGS, the
likelihood of MH was classified as possible, probable or fulminant based on the expert panelists opinion. When there
was discrepancy between the two reviews, a third panelist
was assigned to reconcile the differences on site with the two
first panelists present to answer questions. When differences
did occur they were usually easily reconciled when information missed by one panelist was found by another. There
were no charts for which an agreement could not be made
on a final CGS score and assignment of MH probability by
three panelists.
Proportions and confidence intervals (95%) were calculated
using Stata Version 11.2 (StataCorp LP, College Station, TX).
Results
Demographics
A total of 47 patients were discharged with a diagnosis of
MH as indicated by ICD-9 code 995.86 or ICD-10 code
T88.3 from the six participating hospitals between January
1, 2006 and December 31, 2008. Medical records for all 47
patients were reviewed. Of these 47 patients, 10 (21%) were
children (age 17 yr or less) and 23 (49%) were male. The
mean age was 40 yr and the median age was 36 yr. Thirtytwo of the patients (68.1%) were admitted for or underwent a surgical procedure with general anesthesia during the
hospitalization.
Anesthesiology 2015; 122:55-63 56 Pinyavat et al.
MH Codes and Clinical MH Cases

Eleven of the 47 patients had an incident of possible, probable, or fulminant MH event triggered by anesthesia during
the index hospitalization (table1). Therefore, the positive
predictive value for incident MH when MH was coded as a
primary or secondary discharge diagnosis was 23.4% (95%
CI (12.338.0%). MH was coded as the primary diagnosis
in one case. This patient was transferred from an outside hospital for MH management.
The age range of the 11 patients with an incident of possible, probable, or fulminant MH was 477 yr. The mean
age was 33 yr and 66% were male. All patients had a surgical procedure during the admission except for one who
was transferred from an ambulatory surgical center for MH
management after liposuction. Five of the 11 cases were
emergent. Succinylcholine was administered in 6 of the
11 cases. In one patient, the sole triggering agent received
was succinylcholine. The remaining 10 patients received
volatile halogenated agents, most often sevoflurane or desflurane. The maximum temperature ranged from 36.3 to
40.4C with a mean of 38.7C (temperature information
was not available for two patients). All patients survived
to discharge.
Nine patients received dantrolene. One patient who did
not receive dantrolene was a 4-yr-old boy who had masseter
spasm after succinylcholine. Volatile anesthetics were discontinued and the patient was observed without further treatment. The second patient who did not received dantrolene
was a 5-yr-old boy with congenital ptosis who had signs of
hypermetabolism (temperature of 38.2C, heart rate of 165
beats/min, and an end tidal carbon dioxide of 60mm Hg)
during an anesthetic that resolved after discontinuation of
volatile anesthetic, cooling, and hyperventilation.
Table 1. Explanation for Malignant Hyperthermia Code for
Total Sample
MH incident
Possible
Probable
Fulminant
MH history
Patient
Family
Patient and family
Fever* unrelated to MH
Other
Total
Frequency (n)
11
4
4
3
22
12
6
4
11
3
47
23.4%
46.8%
23.4%
6.3%
*Fever defined as maximum temperature greater than 37.5 C. Other

includes cases of infected spinal hardware without fever, lumpectomy for
breast mass without complication, and lumbar decompression and fusion
complicated by postoperative respiratory arrest.
MH = malignant hyperthermia.
MH Codes and Personal or Family History

Twenty-two patients (46.8%) were coded due to a personal
(n = 12) or family (n = 6) history of MH, or both (n = 4).
In the cases of family history, all except one were in a firstdegree relative.
MH Codes and Non-MH Cases
Eleven records were miscoded as MH due to a high fever
unrelated to anesthesia (maximum temperature ranged from
40.5 to 42.2C with a mean of 41.5C). The mean age was
40 yr; 36% were male, and 72% had no surgical procedure
or anesthesia during admission. In-hospital all-cause mortality for these patients was 18%.
Dantrolene was administered in three cases determined
not to be related to MH and miscoded due to fever. One
patient was a 55-yr-old male with a history of traumatic brain
injury admitted from a rehabilitation facility with fever and
a diagnosis of urosepsis. Dantrolene was given empirically
for his fever of 41.1C and MH was listed on the differential
diagnosis in an emergency department note. Another patient
was a 91-yr-old female with heart failure, chronic obstructive
pulmonary disease, and renal insufficiency admitted with
pneumonia. She had received haloperidol, and dantrolene
was given due to suspected neuroleptic malignant syndrome
with a maximum temperature of 41.3C. The third patient
was a 2-yr-old female with severe pulmonary hypertension
who became unstable after receiving epoprostenol. Upon
arrival in the intensive care unit, dantrolene and antibiotics
were given for a fever of 41.1C.
The remaining three non-MH cases had no clear explanation for miscoding. Two were oncology patients. One
was a 75-yr-old male with hairy cell leukemia admitted for
osteomyelitis and possible infected spinal hardware with a
maximum temperature of 37.5C. The second patient was
a 69-yr-old female with breast cancer admitted for breast
lumpectomy and no recorded fever. The third patient was a
32-yr-old female with no recorded fever who had a respiratory arrest 2h after a lumbar decompression and fusion possibly due to opioid overdose.
MH Codes for Surgical Cases and Patients
Receiving Dantrolene
A total of 32 patients who underwent a surgical procedure
with general anesthesia, 11 were judged to have a clinical MH
event (34.3%) and 16 had personal or family history of MH
(50%). A total of 12 patients who were given dantrolene, 9
(75%) had a clinical MH event, none had a history or family
history of MH, and 3 (25%) had an event unrelated to MH.
In patients who had surgery and dantrolene, 9 out of 10 had
an incident MH event.
Discussion
Our results show that of 47 patients with ICD-9 or ICD-10
coded MH on hospital discharge, 23.4% had an incident
MH episode and 46.8% a personal and/or family history of
MH. Taken together, approximately 70% of cases were MH

susceptible. The most common reason for inaccurate coding
was high fever unrelated to anesthesia.
We were broad in our inclusion of all suspected cases,
from possible to fulminant, as true positives for the ICD
code. While this broad definition may have increased our
estimate of coding accuracy, we believe most researchers
using the ICD code to investigate MH are interested in any
possible cases rather than being limited to biopsy-proven
MH. Although we analyzed them separately, we did not
consider the personal and/or family history cases as coded
in error. In some ways, MH susceptibility can be viewed as
a chronic disease. Having a positive history or family history
has important clinical implications and the diagnosis should
be carried through the medical record on each admission.
However, because a different ICD diagnosis of MH susceptibility does not exist, the diagnosis of MH is assigned in
these cases.
Our study found that a personal or family history of
malignant hyperthermia is difficult to distinguish from an
incident diagnosis using discharge diagnoses alone. Some
databases such as Medicaid and Medicare databases, and
individual state hospital discharge databases, contain a present on admission flag to identify a preexisting condition.
In our previous study on MH prevalence in New York State
between 2001 and 2005, 52% (38 of 73) of cases with the
MH diagnosis were flagged as having the condition present
on admission.2 We found in the current study that 47% of
cases with the MH diagnosis were coded due to personal or
family history of MH. The results and conclusions of previous studies of MH prevalence using administrative databases
may be impacted by our finding that prevalence of MH
susceptibility is more likely to be accurately captured by
ICD code than MH incidence. The confusion in coding the
patient as MH susceptible versus having had an acute MH
event may also impact medical billing and reimbursement,
arguably the most important use of ICD codes today.
Including only surgical admissions increased the positive predictive value of the MH code for incident MH from
23.4% to 34.4%. Most of the patients with a suspected MH
event received dantrolene. Therefore, one way to improve
specificity of searches for incident MH events using administrative databases might be to include information on whether
a surgical procedure was performed and whether dantrolene
was administered.
Many ICD coding accuracy studies use physician diagnosis in the medical record or expert coders reabstraction of the
record as a gold standard.5,7,8,10 They address coding accuracy and assume physician accuracy.7 A definitive diagnosis
of MH is often problematic to make over the course of one
hospitalization. While muscle biopsy and caffeine-halothane
contracture testing can be performed at special testing centers to confirm MH susceptibility, there is no test that can
be applied acutely to distinguish MH from other causes of
hypermetabolism or hyperthermia. Moreover, contracture
tests are not reliable if performed in the first three months

after an acute event and therefore these results are not available during the same hospital admission, if at all. Many studies in the United States and Europe have used the CGS as
the clinical definition for MH.1,11,12 We chose the CGS and
expert opinion as our standard for true MH occurrence. In
choosing this external gold standard, we attempted to minimize the amount of physician error. For instance, in two of
the three non-MH cases in which dantrolene was given,
MH was perhaps on the treating physicians differential diagnosis but was deemed highly unlikely by our expert panel.
Validating the physicians notes in the instance of reported
previous personal or family history of MH, however, was
beyond the scope of the study.
Coding errors for MH accounted for approximately
32% of cases, most often involving patients with high
fever. Because the introduction of the MH code occurred
first in October 1997, familiarity with MH may be low
among medical coders.11 We also found that it was fairly
common for physicians to inappropriately use the term
malignant hyperthermiaperhaps reflecting the ambiguity of the terminology that currently existsto refer
to this hypermetabolic syndrome that is neither malignant nor always hyperthermic. Patients with fever who
were mistakenly coded with MH tended to have a higher
maximum temperature than patients who had a true MH
episode. They were sicker overall than patients who had a
true MH episode with most patients having an American
Society of Anesthesiologists Physical Status of 3 or greater
and a high mortality rate. They were also unlikely to have
had a surgical procedure or anesthesia or to have received
dantrolene. Excluding patients with diagnoses associated
with hyperthermia, such as sepsis and neuroleptic malignant syndrome, as Rosero et al.4 did in their study of MH
using the Nationwide Inpatient Sample, would eliminate
many of these coding errors.
Sequencing errors were studied widely when diagnosisrelated groups became the basis for hospital reimbursement.
The order of primary versus secondary diagnosis in these
studies was found be highly prone to error.5,10,13 The Institute of Medicine study found sequencing to be the number one reason for coding errors, leading to over 80% of
errors made in coding ischemic heart disease, for example.10
To minimize the impact of type of error and find as many
cases as possible for review, we included both primary and
secondary diagnoses of MH. The primary diagnosis is often
based on the admitting diagnosis and acute MH will usually
not appear as the admitting diagnosis, with the exception
of patient transfer for MH management. Including MH
as only a primary diagnosis would have missed all but one
admission in our study.
Limitations of our study include a small study sample
and a limited number of institutions. Our study was also
limited by the inability to identify cases in which the MH
code was omitted. Therefore, we are unable to report a code
specificity or sensitivity. Further studies of MH ICD code

accuracy might be of interest based on a larger sample of
hospitals, including both academic medical centers and
ambulatory surgical centers. Comparing MH ICD coding
accuracy in the United States and other countries would
also be of interest.
To our knowledge, this is the first study to examine the
ICD code for MH due to anesthesia. We estimate that the
code has a positive predictive value of approximately 23.4%
(95% CI, 12.338.0%) for incident MH, and 70.2% (95%
CI, 55.182.7%) for MH susceptibility. The code is more
likely to detect a history or family history of MH (47%)
than an incident MH event. The remainder of the cases were
miscoded, many due to high fever. Possible barriers unique
to accuracy of ICD coding for MH include variable clinical
presentation, lack of a pathognomonic test, rarity of the disease leading to inexperience among physicians and coders,
and relative newness of the code.
Using the vast amount of data available through administrative databases is important in expanding our knowledge
for rare diseases such as MH. It may be possible to improve
accuracy in finding MH events, if additional information
such as Current Procedural Terminology (CPT) billing codes
or Clinical Classification Software (CCS) procedure codes
are used to confirm that a surgical procedure took place, and
anesthesia was given during the admission; pharmacy databases may be used to confirm that dantrolene was administered. Further study is required to find the best algorithm of
codes to accurately capture incident MH cases in hospital
discharge records. Educating physicians and medical coders
about MH should also help increase coding accuracy.
Acknowledgments
The authors thank Anthony Isenalumhe, M.D., and Thomas Blanck, Ph.D., M.D., of the Department of Anesthesiology, New York University, New York, New York, and
Ronald Litman, D.O., and Theodora Goebel, R.N., V.S.N.,
at the Department of Anesthesiology, Childrens Hospital
of Philadelphia, Philadelphia, Pennsylvania, for their assistance with site chart reviews.
This study was supported in part by the Malignant Hyperthermia Association of the United States, Sherburne,
New York.
Competing Interests
Dr. Rosenberg received a one-time speaking fee from Eagle
Pharmaceuticals (Woodcliff Lake, New Jersey), a company
which manufactures Ryanodex, a concentrated formulation of dantrolene approved for the treatment of malignant
hyperthermia. The other authors declare no competing interests.
Correspondence
Address correspondence to Dr. Li: Departments of Anesthesiology and Epidemiology, Columbia University College
of Physicians and Surgeons and Mailman School of Public
Health, 622 West 168th St., PH5-505, New York, New York
10032. gl2240@cumc.columbia.edu. Information on purchasing reprints may be found at www.anesthesiology.org
or on the masthead page at the beginning of this issue.
Anesthesiologys articles are made freely accessible to all
readers, for personal use only, 6 months from the cover
date of the issue.
References
1. Larach MG, Gronert GA, Allen GC, Brandom BW, Lehman
EB: Clinical presentation, treatment, and complications of
malignant hyperthermia in North America from 1987 to
2006. Anesth Analg 2010; 110:498507
2. Brady JE, Sun LS, Rosenberg H, Li G: Prevalence of malignant
hyperthermia due to anesthesia in New York State, 20012005. Anesth Analg 2009; 109:11626
3. Li G, Brady JE, Rosenberg H, Sun LS: Excess comorbidities
associated with malignant hyperthermia diagnosis in pediatric
hospital discharge records. Paediatr Anaesth 2011; 21:95863
4. Rosero EB, Adesanya AO, Timaran CH, Joshi GP: Trends and
outcomes of malignant hyperthermia in the United States,
2000 to 2005. Anesthesiology 2009; 110:8994
5. Corn RF: The sensitivity of prospective hospital reimbursement to errors in patient data. Inquiry 1981; 18:35160
6. Lloyd SS, Rissing JP: Physician and coding errors in patient
records. JAMA 1985; 254:13306
7. OMalley KJ, Cook KF, Price MD, Wildes KR, Hurdle JF,
Ashton CM: Measuring diagnoses: ICD code accuracy. Health
Serv Res 2005; 40(5 Pt 2):162039
8. Fisher ES, Whaley FS, Krushat WM, Malenka DJ, Fleming
C, Baron JA, Hsia DC: The accuracy of Medicares hospital
claims data: Progress has been made, but problems remain.
Am J Public Health 1992; 82:2438
9. Larach MG, Localio AR, Allen GC, Denborough MA, Ellis FR,
Gronert GA, Kaplan RF, Muldoon SM, Nelson TE, Ording H:
A clinical grading scale to predict malignant hyperthermia
susceptibility. Anesthesiology 1994; 80:7719
10. Demlo LK, Campbell PM, Brown SS: Reliability of information abstracted from patients medical records. Med Care
1978; 16:9951005
11. Allen GC, Brubaker CL: Human malignant hyperthermia
associated with desflurane anesthesia. Anesth Analg 1998;
86:132831
12. Ording H, Brancadoro V, Cozzolino S, Ellis FR, Glauber V,
Gonano EF, Halsall PJ, Hartung E, Heffron JJ, Heytens L,
Kozak-Ribbens G, Kress H, Krivosic-Horber R, Lehmann-Horn
F, Mortier W, Nivoche Y, Ranklev-Twetman E, Sigurdsson S,
Snoeck M, Stieglitz P, Tegazzin V, Urwyler A, Wappler F: In
vitro contracture test for diagnosis of malignant hyperthermia
following the protocol of the European MH Group: Results
of testing patients surviving fulminant MH and unrelated
low-risk subjects. The European Malignant Hyperthermia
Group. Acta Anaesthesiol Scand 1997; 41:95566
13. Hsia DC, Krushat WM, Fagan AB, Tebbutt JA, Kusserow RP:
Accuracy of diagnostic coding for Medicare patients under the
prospective-payment system. N Engl J Med 1988; 318:3525
Appendix
Appendix. (Continued)
AE = adverse event; ALT = alanine transaminase; AMR = adverse metabolic reaction; Anesth. = anesthetic; Art/Ven = arterial/venous; ASC = ambulatory
surgery center; AST = aspartate transaminase; BE = base excess; BPM = beats per minute; CK = creatinine kinase; CPR = cardiopulmonary resuscitation;
CPT = Current Procedural Terminology; DIC = disseminated intravascular coagulopathy; ET = endotracheal; ETCO2 = end-tidal carbon dioxide; ETPCO2 =
end-tidal partial carbon dioxide concentration; F = female; FiO2 = fraction of inspired oxygen; GI = gastrointestinal; HR = heart rate; I = initial reaction; ICU
= intensive care unit; IRB = institutional review board; IV = intravenous; K = potassium; M = male; Max. = maximum; MH = malignant hyperthermia; Min. =
minimum; MM/YYYY = month/year; N = no; N/A = not applicable; N/D = not determined; PaCO2 = arterial carbon dioxide tension; pCO2 = partial pressure
of carbon dioxide; PetCO2 = end-tidal pressure of carbon dioxide; pO2 = partial pressure of oxygen; PPV = positive predictive value; pt = paties; R = recrudescence; Rx = reaction; SIDS = sudden infant death syndrome; Temp. = temperature; UNK = unknown; Y = yes; yr = years.
Relationship between Chronic Intermittent Hypoxia and

Intraoperative Mean Arterial Pressure in Obstructive Sleep
Apnea Patients Having Laparoscopic Bariatric Surgery
Alparslan Turan, M.D., Jing You, M.S., Cameron Egan, B.S., Alex Fu, M.D., Ignazia Gazmuri, M.D.,
Ashish Khanna, M.D., Yashar Eshraghi, M.D., Raktim Ghosh, M.D., Somnath Bose, M.D.,
Shahbaz Qavi, M.D., Lovkesh Arora, M.D., Daniel I. Sessler, M.D., Anthony G. Doufas, M.D., Ph.D.
ABSTRACT
Background: Recurrent nocturnal hypoxemia in obstructive sleep apnea enhances sympathetic function, decreases baroreceptor sensitivity, and weakens peripheral vascular responses to adrenergic signals. The authors hypothesized that the percentage
of total sleep time spent at oxyhemoglobin saturation (Sao2) less than 90% and minimum nocturnal Sao2 on preoperative
polysomnography are associated with decreased intraoperative mean arterial pressure.
Methods: The authors examined the records of all patients who had laparoscopic bariatric surgery at Cleveland Clinic between
2005 and 2009 and an available polysomnography study. The authors assessed the relationships between the percentage of
total sleep time spent at Sao2 less than 90% and minimum nocturnal Sao2, and the time-weighted average of mean arterial
pressure. The authors used multivariable regression models to adjust for prespecified clinical confounders.
Results: Two hundred eighty-one patients were included in the analysis. The average change in the time-weighted average of
mean arterial pressure was 0.02 (97.5% CI, 0.08, 0.04) mmHg for each 1% absolute increase in the percentage of sleep time
spent at Sao2 less than 90% (P = 0.50). The average change was 0.13 (97.5% CI, 0.27, 0.01) mmHg, for each 1% absolute
decrease in the minimum Sao2 (P = 0.04 > significance criterion of 0.025, Bonferroni correction). An unplanned analysis
estimated 1% absolute decrease in minimum Sao2 was associated with 0.22 (98.75% CI, 0.39, 0.04) mmHg, change in
mean arterial pressure (P = 0.002) in the time period between endotracheal intubation and trocar insertion.
Conclusion: Recurrent nocturnal hypoxemia in obstructive sleep apnea is not a risk marker for intraoperative hypotension.
BSTRUCTIVE sleep apnea (OSA) is common and

characterized by repeated partial or complete airway collapse during sleep, potentially leading to severe oxyhemoglobin
desaturation.1 Approximately 30% of the general population
suffers from OSA,2 while a similar fraction of surgical patients3
are at high risk for the disease, with most of them lacking a formal diagnosis.35 OSA has been linked to cardiovascular6,7 and
metabolic8 morbidity, while a diagnosis of moderate-to-severe
disease (15 apnea /hypopnea events per hour of sleep) has
been identified as an independent risk factor for all-cause and
cardiovascular mortality.9,10 As in other studies linking OSA to
cardiovascular morbidity,11,12 these investigations suggest that
nocturnal intermittent hypoxia, rather than sleep fragmentation, is the responsible risk-increasing element.
Patients suffering from OSA present with a chronic
enhancement in sympathetic adrenergic activity13,14 that is
considered one of the major mechanisms in the development
of cardiovascular morbidity in this population.15 This excess
in adrenergic signaling is associated with an almost twofold

Approximately 30% of the general population suffers from
obstructive sleep apnea. Patients suffering from obstructive
sleep apnea present with a chronic enhancement in sympathetic adrenergic activity that is considered one of the major
mechanisms in the development of cardiovascular morbidity
in this population. Thus, obstructive sleep apnea patients may
have increased risk for intraoperative and postoperative morbidity consequent to hemodynamic instability.
This study investigated whether nocturnal intermittent hypoxia consequent to obstructive sleep apnea, as quantified
by the percentage of total sleep time spent at Sao2 less than
90% and the minimum nocturnal Sao2, is associated with decreased intraoperative mean arterial pressure in patients undergoing laparoscopic bariatric surgery.

Recurrent nocturnal hypoxemia in obstructive sleep apnea is
not a risk marker for intraoperative hypotension in patients undergoing laparoscopic bariatric surgery.
increase in blood pressure variability,16 a significant attenuation of baroreflex sensitivity,17,18 and decreased peripheral
Submitted for publication February 24, 2014. Accepted for publication August 22, 2014. From the Department of Outcomes Research,
Cleveland Clinic, Cleveland, Ohio (A.T., C.E., A.F., I.G., Y.E., R.G., D.I.S.); Departments of Quantitative Health Sciences and Outcomes
Research, Cleveland Clinic, Cleveland, Ohio ( J.Y.); Anesthesiology Institute, Cleveland Clinic, Cleveland, Ohio (A.K., S.B., S.Q., L.A.); Department of Anesthesiology, Perioperative, and Pain Medicine, Stanford University School of Medicine, Stanford, California (A.G.D.); and Outcomes Research Consortium, Cleveland, Ohio (A.G.D.). Current affiliations: Boston University, Boston, Massachusetts (C.E.); Department of
Anesthesiology, Oregon Health and Science University, Portland, Oregon (A.F.); Department of Anesthesiology, MetroHealth Medical Center/
Case Western Reserve University, Cleveland, Ohio (Y.E.); and Internal Medicine, St. Vincent Charity Medical Center, Cleveland, Ohio (R.G.).
January 2015
vascular adrenergic responses,19 in OSA patients compared

to normal subjects. Both animal and human experiments
demonstrate that the magnitude of these alterations in cardiovascular homeostatic mechanisms is proportional to the
severity of the disease as measured by the number of apnea/
hypopnea events per hour of sleep and the associated extent
of recurrent hypoxemia.
The severity of nocturnal arterial desaturation was found
to be the strongest predictor of baroreceptor sensitivity
depression in OSA subjects.17 In a rat model of chronic intermittent hypoxia, the resulting sympathetic overactivity was
associated with a paradoxical decrease in the pressor response
to adrenergic signals20 suggesting peripheral adaptation to
chronic adrenergic overstimulation. Consistent with these
observations is a recent report that patients at high risk for
OSA (based on a validated risk prediction model) were more
likely to receive vasopressors to maintain hemodynamic
goals intraoperatively than low-risk patients.21
Even short periods of low intraoperative mean arterial
pressure (MAP) are strongly associated with postoperative
morbidity and associated organ damage.22 Impaired cardiovascular homeostasis may render OSA patients more vulnerable to hypotensive stimuli during a general anesthetic.23,24
Furthermore, defects in blood flow autoregulation mechanisms for vital target organs like the brain25,26 may decrease
the ability of subjects with OSA to cope with low perfusion insults. OSA patients may thus especially be at risk for
intraoperative and postoperative morbidity consequent to
hemodynamic instability including cardiovascular events
and transfers to intensive care settings.2729
We thus retrospectively examined the intraoperative
blood pressure in patients who had laparoscopic bariatric
surgery. Specifically, we tested the hypothesis that nocturnal
intermittent hypoxia consequent to OSA, as quantified by
the percentage of total sleep time spent at oxyhemoglobin
saturation (Sao2) less than 90% and the minimum nocturnal
Sao2, are associated with decreased intraoperative MAP.
Patients
Eligible were all patients who had laparoscopic bariatric procedures between June 2005 and December 2009 and had a
diagnosis of OSA with a polysomnography study performed
within two preoperative years. We excluded patients who did
not have a polysomnography report available and/or those
whose polysomnography reports were missing critical exposure variables such as the percentage of total sleep time spent
at Sao2 less than 90%, or minimum nocturnal Sao2. We also
excluded patients missing any of the prespecified potential
confounders. Patients with severe cardiopulmonary disease
associated with respiratory insufficiency and/or requiring
oxygen supplementation during daytime or sleep, were also
excluded.
Sleep studies were performed and evaluated, according to consensus and practice guidelines published by
the American Academy of Sleep Medicine.3033 Briefly,
14-channel polysomnography included monitoring the
electroencephalogram, electro-oculogram, electromyogram,
oronasal flow by thermocouples and nasal pressure, thoracic
and abdominal movement by inductance plethysmography,
and oxyhemoglobin saturation by pulse oximetry (Sao2).
The apnea/hypopnea and arousal indices were estimated by
dividing the number of all events occurring during sleep
with the total sleep time calculated as the sum of all sleep
stage periods.
Outcomes and Exposures

Time-weighted average intraoperative MAP was the main
outcome in our analysis. Intraoperative blood pressure data
were acquired from an electronic anesthesia record-keeping
system, which continuously records minute-by-minute data
from physiologic monitors during the intraoperative period.
When an arterial catheter was used, blood pressure was
recorded every minute. When noninvasive blood pressure
monitoring was employed, blood pressure was recorded at
1- to 5-min intervals. We removed blood pressure artifacts,
which were identified by out of range, abrupt change, and
moving average methods.
Our main exposure variables were the percentage of total
sleep time spent at Sao2 less than 90% and the minimum
nocturnal Sao2 listed in polysomnography reports, two
parameters indicating the nocturnal oxygenation status of
With approval from the Cleveland Clinic Institutional

Review Board (Cleveland, Ohio), we undertook a retrospective observational study using data from the Cleveland
Clinic Perioperative Health Documentation System. This
registry contains the complete anesthesia record for all noncardiac surgery patients, and is linked to the full electronic
medical record and about a dozen other data sources.
We used deidentified information from electronic records
and the paper charts of patients who had laparoscopic bariatric surgery at the Cleveland Clinic Main Campus between
June 2005 and December 2009, to retrospectively examine
the relationship between the severity of chronic intermittent
hypoxia, as measured by overnight polysomnography, and
intraoperative MAP in patients suffering from OSA.
Anesthetic Management
General anesthesia was induced with propofol, fentanyl, and
rocuronium or succinylcholine, and maintained with a volatile anesthetic in an oxygen/air mixture. Fentanyl, morphine,
alfentanil, remifentanil, or hydromorphone was titrated to
patients vital signs. Vasopressors such as ephedrine, phenylephrine, or epinephrine were used to maintain blood
pressure within 20% of preoperative level. Intraabdominal
pressure during the pneumoperitoneum was automatically
controlled at a level equal to, or less than 15mm H2O. Epidural analgesia was not used.
Anesthesiology 2015; 122:64-71 65 Turan et al.
Nocturnal Hypoxia and Intraoperative Hemodynamics
the patients with OSA. Investigators scrutinized the paper

charts of patients with a diagnosis of sleep-disordered breathing and recovered any existing polysomnography reports.
When polysomnography details were not included in the
medical charts, the investigators contacted the respective
sleep laboratories and recovered the full reports. From the
polysomnography reports, the total sleep time spent at Sao2
less than 90%, the minimum nocturnal Sao2, and the apnea/
hypopnea index (events per hour of sleep) were recorded.
Exposure variables that were of interest because of their
confounding potential were identified in our electronic registry and recorded. These included patients age, sex, and
race, as well as morphometric characteristics like body height
and weight that were used to estimate the body mass index
[= (weight in kg)/(height in m)2]. Important comorbid conditions including smoking status, hypertension, coronary
artery disease, and diabetes, as well as the use of antihypertensive medications and continuous positive airway pressure for management of OSA were also determined from
electronic or paper medical records. And finally, we also
recorded the type of bariatric surgery and various intraoperative variables including surgery duration, anesthetic drug
doses/gas concentrations, opioid use, as well as the amount
of fluids and the type and doses of vasopressors that were
given intraoperatively.
All patients with available information on the examined
exposure parameters and outcomes were included in the
analysis. We assessed the relationships between nocturnal
recurrent hypoxemia (severity measured by the percentage of total sleep time spent at Sao2 less than 90% and the
minimum nocturnal Sao2) and intraoperative time-weighted
average of MAP (a total of two analyses), each with a multivariable regression model. The time-weighted average of
MAP is equal to the sum of the portion of each time interval
in-between two adjacent MAP measurements multiplied by
the average of the corresponding two MAP measurements
and divided by the time interval between the first and the
last MAP measurements. We prespecified the following 11
potential confounders including age, sex, race, body mass
index, smoking status, diabetes mellitus, hypertension, coronary artery disease, preoperative use of antihypertensive
medications, continuous positive airway pressure therapy,
and type of bariatric surgery (i.e., laparoscopic gastroenterostomy vs. others). All the above potential confounders were
considered for inclusion in each model using a backward
selection procedure (alpha-to-stay = 0.20). A Bonferroni
correction was used to adjust for multiple testing; the significance criterion for each individual analysis was P value less
than 0.025 (i.e., 0.05/2).
Intraoperative opioids were converted to IV morphine
equivalents using published conversion tables.34 The total
amount of anesthetic gas (including isoflurane, desflurane,
and sevoflurane) was estimated by multiplying the end-tidal
gas concentration by the duration of volatile administration

in hours, and then dividing by the corresponding potency
(i.e., 1.17vol % for isoflurane, 1.8vol % for sevoflurane, and
6.6vol % for desflurane).
Secondary Analyses. We also evaluated the relationship
between percentage of total sleep time spent at Sao2 less than
90% and minimum nocturnal Sao2, and intraoperative use
of vasopressor (yes vs. no), including ephedrine, epinephrine,
and phenylephrine, each using a multivariable logistic regression. Among patients who received vasopressor, we further
assessed the relationship between percentage of total sleep
time spent at Sao2 less than 90% and minimum nocturnal
Sao2, and the total dose of vasopressor, using a multivariable
regression model. The dose of ephedrine was converted to
phenylephrine using a potency ratio between phenylephrine
(g) and ephedrine (mg) of 125, as suggested previously.35
Subsequently, the total dose of vasopressor (phenylephrine
equivalent) was calculated by adding of the dose of ephedrine and the dose of phenylephrine. Two patients given epinephrine were not included in this analysis. All the above
mentioned potential confounders were considered for inclusion in each model through the use of a backward selection
procedure (alpha-to-stay = 0.20).
We also conducted four unplanned exploratory analyses
focusing on the associations between the percentage of total
sleep time spent at Sao2 less than 90% and minimum nocturnal
Sao2, and the intraoperative time-weighted average of MAP,
separately for the time periods spanning from endotracheal
intubation to trocar insertion and from trocar insertion until
the end of case, using the same statistical methods as in the primary analysis. A Bonferroni correction was used to adjust for
multiple testing; the significance criterion for each individual
analysis was thus P value less than 0.0125 (i.e., 0.05/4).
The Statistical Analysis Software version 9.3 (SAS Institute, Cary, NC) was used for all analyses.
Results
One thousand six hundred forty-one adult patients had
laparoscopic bariatric surgery at the Cleveland Clinic Main
Campus between June 6, 2005 and December 30, 2009.
Among 335 who had an overnight polysomnography performed, 54 patients with missing minimum nocturnal Sao2,
percentage of total sleep time at Sao2 less than 90%, and/or
other covariates were excluded from the analyses, yielding
a final set of 281 patients (table1 and fig.1). More than
71% of our patients had a polysonography study performed
within 1 yr preoperatively, and 52% of the tests were done
within 6 months before surgery.
After removing artifacts (2% of all measurements), the
median of number of MAP measurements between endotracheal intubation and end of the case was 59 [Q1, Q3:
47, 91]. The observed mean (SD) of the intraoperative timeweighted average of MAP was 89 (11) mmHg. The observed
median of percentage of total sleep time spent at Sao2 less
than 90% was 8.1% [1.4, 26.9] and median of minimum
Table 1. Morphometrics, Polysomnography Variables, and

Clinical Characteristics (N = 281*)
Variable
Summary
Statistics
Age (yr)
4711
Sex (female), No. (%)
201 (72)
Race, No. (%)
Caucasian
216 (77)
African American
57 (20)
Others
8 (3)
Body mass index (kg/m2)
46 [42, 52]
Polysomnography parameters
8.1 [1.4, 26.9]
Time spent at Sao2 <90% (% of total sleep
time)
Minimum nocturnal Sao2 (%)
82 [74, 86]
AHI
AHI < 5
38 (14)
68 (24)

5 AHI < 15
64 (23)

15 AHI < 30
108 (39)

AHI 30
Smoking status (smokers), No. (%)
141 (50)
Diabetes, No. (%)
97 (35)
Hypertension, No. (%)
175 (70)
Coronary artery disease, No. (%)
26 (9)
Antihypertensive medication, No. (%)
67 (24)
Continuous positive airway pressure therapy,
178 (63)
No. (%)
Intraoperative
Type of gastric surgery, No. %

Gastroenterostomy
225 (80)
Gastric restrictive procedure
51 (18)

Gastroplasty
3 (1)
Removal of gastric restrictive device
1 (<1)
Duration of surgery (hours)
4.2 [3.7, 4.9]
From ET intubation to trocar insertion (hours)
0.50.1
From trocar insertion to end of case (hours)
3.51.1
TWA of MAP during surgery (mmHg)
8911
From ET intubation to trocar insertion (mmHg)
7912
From trocar insertion to end of case (mmHg)
9012
TWA of heart rate during surgery (beats/min)
7610
From ET intubation to trocar insertion
7613
(beats/min)
From trocar insertion to end of case (beats/min)
7610
Antihypertensive medications, No. (%)
76 (31)
Usage of vasopressors, No. (%)
153 (54)
Usage of ephedrine, No. (%)
97 (63)

Ephedrine (mg)
10 [10, 20]
Usage of epinephrine, No. (%)
2 (1)

Epinephrine (mg)
0.01 [0.01, 0.01]
Usage of phenylephrine, No. (%)
109 (71)
200 [100, 400]

Phenylephrine (g)
Amount of anesthetic gas (MAC hours)
2.8 [2.1, 3.4]
Propofol at induction (mg)
200 [200, 250]
Opioids (IV morphine equivalent) (mg)
30 [25, 40]
Crystalloids (l)
3.00.9
Colloids (l)
0.50.4
* Out of 335 patients who had nocturnal polysomnography, 54 patients
with missing total time Sao2 <90% and/or other covariates were excluded
from the analyses. One patient had missing amount of anesthetic gas;
three patients had missing AHI value. Statistics are reported as mean
SD, median [1st, 3rd quartile], or No. (%).
AHI = apnea/hypopnea index; ET = endotracheal; MAC = minimum alveolar
concentration; MAP = mean arterial pressure; Sao2 = oxyhemoglobin saturation; TWA = time-weighted average.
nocturnal Sao2 was 82% [74, 86]. The unadjusted associations are displayed in figure2.
We found that neither percentage of total sleep time spent
at Sao2 less than 90% (P = 0.50) nor minimum nocturnal
Sao2 (P = 0.04 > Bonferroni-corrected significance criterion
of 0.025) was associated with intraoperative time-weighted
average of MAP (table2). The estimated average change in
time-weighted average of MAP was 0.02 (97.5% CI, 0.08
to 0.04) mmHg for each 1% absolute increase in the percentage of total sleep time spent at Sao2 less than 90%, after
adjusting for smoking and coronary artery disease (retained
via the model selection procedure). The estimated average
change was 0.13 (97.5% CI, 0.27 to 0.01) mmHg, for
each 1% absolute decrease in the minimum nocturnal Sao2,
after adjusting for smoking and coronary artery disease. For
example, in a minimum nocturnal Sao2 reduction from 92
to 74%, the time-weighted average of MAP would decrease
2.3 mmHg (2.3 [97.5% CI, 4.8 to 0.1]).
Secondary Analyses
One hundred fifty-three (54%) patients were given vasopressor intraoperatively. Receiving vasopressor was not significantly associated with either percentage of total sleep time
spent at Sao2 less than 90% (P = 0.86) or minimum nocturnal Sao2 (P = 0.39). The estimated odds ratio of receiving
vasopressor was 1.00 (97.5% CI, 0.99 to 1.01) for each 1%
absolute increase in percentage of total sleep time spent at
Sao2 less than 90% and 1.01 (0.98 to 1.04) for each 1% absolute decrease in minimum nocturnal Sao2. Age, body mass
index, smoking, coronary artery disease, hypertension, and
type of bariatric surgery were adjusted for in both analyses.
Among the 154 patients given vasopressor, 97 received
ephedrine with median dose of 10 [10, 20] mg; two patients
were given epinephrine; and 109 patients received phenylephrine with a median dose of 200 [100, 400] g. The estimated total dose of vasopressor in phenylephrine equivalents
was 1,250 [400, 2,050] g. The estimated average change in
the total dose of vasopressor was 3 (97.5% CI, 14 to 8) g
for each 1% absolute increase in the percentage of total sleep
time spent at Sao2 less than 90% (P = 0.59), after adjusting
for age, sex, and preoperative use of antihypertensive medications. The estimated change in the dose was 9 (16, 34)
g for each 1% absolute decrease in the minimum nocturnal
Sao2 (P = 0.41), after adjusting for age, sex, smoking, diabetes, and preoperative use of antihypertensive medications.
Exploratory Analyses
The median number of MAP measurements between endotracheal intubation and trocar insertion was 9 [6, 13], while
between trocar insertion and end of case was 48 [37, 77].
Consistently with our primary analysis, we found that neither
the percentage of total sleep time spent at Sao2 less than 90%
nor the minimum nocturnal Sao2, was associated with timeweighted average of MAP, during the period spanned from
trocar insertion to end of case. However, the lower minimum
Fig. 1. Flow diagram for the selection of eligible patients. Sao2

= oxyhemoglobin saturation.
nocturnal Sao2 in the preoperative polysomnography was

significantly associated with lower time-weighted average
of MAP during the time period spanned from endotracheal
intubation to trocar insertion (P = 0.002; table 2).
Discussion
After adjusting for prespecified potential confounders, intermittent nocturnal hypoxia (measured by the percentage of
total sleep time spent at Sao2 less than 90% and the minimum nocturnal Sao2) was not significantly associated with
intraoperative MAP during laparoscopic bariatric surgery in
patients suffering from OSA.
Selection of chronic intermittent hypoxia indices as an

explanatory variable for intraoperative hypotension was based
on a priori information linking intermittent hypoxia to cardiovascular morbidity. The original clinical trials establishing the
relationship between OSA, sympathetic overactivity13,14,16,18
and cardiovascular morbidity6,15 were based on the number
of apnea/hypopnea events during sleep.6,13,14,16,18,19 However,
more recent research evidence has established recurrent nocturnal hypoxemia17,20 as the putative causative component.
Human17 and animal20 experiments show that intermittent
hypoxia, through an enhancement of sympathetic adrenergic mechanisms, may reduce the cardiovascular systems
ability to adequately compensate for acute hypotensive challenges. Both a depression of baroreceptor sensitivity17,18 and
the reduced vascular reactivity to adrenergic signals19,20 may
amount to the insufficiency of autonomic nervous control in
counteracting hypotension during anesthesia.
Overnight preoperative polysomnography showed that
approximately 60% of our bariatric patients had moderateto-severe OSA with an apnea/hypopnea index 15 events per
hour of sleep, which is consistent with the current epidemiological profile of the disease in the general population.2 A
median percentage of total sleep time spent at Sao2 less than
90% of 8.1% and a median nocturnal minimum Sao2 of 82%
indicate clinically important hypoxemia, which compares with
magnitudes of oxyhemoglobin desaturation in studies establishing OSA as an independent risk factor for stroke (median
% of total sleep time at Sao2 less than 90%: 0.4 vs. 0.1, with
and without ischemic stroke)36; cardiovascular disease (hypopneas with a 4% or more decrease in Sao2 were predictive vs.
those with less 4%)11; atrial fibrillation (each 1% absolute
decrease in the mean nocturnal Sao2 tripled the risk for atrial
fibrillation)37; insulin resistance (average nocturnal Sao2 less
than 94% and more than 2.2% of total sleep time at Sao2
less than 90% predicted insulin resistance)38; and sleep-related
pain (a decrease in the minimum nocturnal Sao2 from 92 to
75% approximately doubled the odds for reporting pain).39
Fig. 2. Scatter plots presenting the unadjusted associations between the time-weighted average (TWA) of mean arterial pressure
(MAP) during surgery, and the percentage of total sleep time spent at Sao2 <90% (A) and minimum nocturnal Sao2 (B), for the
281 patients included in the analysis. The solid lines are the fit regression lines; and the shaded regions are the 95% confidence
bands. Sao2 = oxyhemoglobin saturation.
Table 2. Association between Time-weighted Average of MAP and Percentage of Total Sleep Time Spent at Sao2 <90% and
Minimum Nocturnal Sao2 (N = 281)
Exposure Variable
Primary analysisintraoperative TWA of MAP
% of TST spent at Sao2 < 90% (for each 1% absolute increase)
Minimum nocturnal Sao2 (for each 1% absolute decrease)
Exploratory analysis 1: TWA of MAP from ET intubation to trocar insertion
Exploratory analysis 2: TWA of MAP from trocar insertion to end of case
Estimated of Slope* (97.5% CI)
P Value
0.02 (0.08 to 0.04)

0.13 (0.27 to 0.01)
0.50
0.04
Estimated of slope* (98.75% CI)
P Value
0.04 (0.12 to 0.03)

0.22 (0.39 to 0.04)
0.14
0.002
0.02 (0.09 to 0.06)

0.14 (0.30 to 0.03)
0.58
0.04
* All the prespecified potential confounders were considered for including in each model through the use of a backward selection procedure (alpha-tostay = 0.20), including age, sex, race, body mass index, smoking status, diabetes mellitus, hypertension, coronary artery disease, preoperative usage of
antihypertensive medications, continuous positive airway pressure therapy, and type of bariatric surgery (i.e., laparoscopic gastroenterostomy vs. others).
Smoking and coronary artery disease were retained in both primary analyses and two analyses for TWA of MAP from surgical incision to end of case. Age,
sex, race and type of surgery were adjusted for two analyses for TWA of MAP from endotracheal intubation to surgical incision. Bonferroni correction: The
significance criterion of 0.025 for the two primary analyses and the criterion of 0.0125 for the four exploratory analyses. Statistically significant.
ET = endotracheal; MAP = mean arterial pressure; Sao2 = oxyhemoglobin saturation; TST = total sleep time; TWA = time-weighted average.
Intraoperative hypotension, as traditionally defined, is common24,40 and as we22,41 and others4244 have shown, low intraoperative MAP may increase morbidity and mortality after noncardiac
surgery. A recent large-scale analysis of noncardiac surgery patients
found that intraoperative MAP less than 55 mmHg for 5min or less
was strongly associated with postoperative acute kidney injury (30%
increase) and myocardial injury (50% increase).22 If OSA patients
are prone to develop intraoperative hypotension, they might also be
at elevated risk for adverse postoperative outcomes related to their
decreased ability to cope with low perfusion insults;25,26 possibly
rendering this patient population vulnerable to postoperative central
nervous system45,46 and cardiovascular2729 morbidities.
Our exploratory analyses was based on the possibility
that pneumoperitoneum, with its profound cardiovascular
effects resulting in hyperdynamic circulation,47,48 may have
obscured the smaller potential effects of nocturnal hypoxia.
We observed that the minimum nocturnal Sao2 in preoperative polysomnography was significantly associated with timeweighted average MAP, during the period spanning from
endotracheal intubation to trocar insertion. A minimum nocturnal Sao2 of 74%, common occurrence among OSA populations, was associated with a reduction in average MAP of
approximately 4 mmHg (98.5% CI, 7.0 to 0.7). Although
the average effect of nocturnal hypoxemia on intraoperative
MAP appears to be small and was restricted to the period
between endotracheal intubation and trocar insertion, it is at
least consistent with the biological rationale underlying our
hypothesis. Large observational cohorts in noncardiac surgery
patients have shown that 50% of patients experience at least
a minute with MAP 60 mmHg or less, and that 30% experience at least 5min with MAP 60 mmHg or less.24 Such hypotensive episodes may thus be aggravated by nocturnal hypoxia,
and may place patients at risk.22,44 However, we urge cautious
interpretation of this finding since it was from an unplanned
post hoc analysis and only barely statistically significant.
As in any retrospective analysis, residual confounding is

likely and it is essentially impossible to estimate its magnitude and potential influence on our conclusions. Furthermore, certain factors, like fluid administration and the use of
drugs, might be mediating variables that transmit the effect
of chronic intermittent hypoxia on the mean intraoperative
arterial pressure. However, in our analysis, we did not investigate the possible mediator effects. Finally, although most
sleep studies were performed within one year preoperatively,
our ability to identify associations between nocturnal hypoxia
and intraoperative hypotension may have been degraded by
changes in the severity of OSA between the time nocturnal
hypoxemia was measured and the actual date of surgery.
In summary, chronic recurrent nocturnal hypoxemia
resulting from OSA is not associated with intraoperative hypotension. An exploratory post hoc analysis reveals a weak association and a small effect of nocturnal hypoxemia on MAP, only a
limited period intraoperatively. The significance and potential
clinical implications of this finding remain to be tested.
Acknowledgments
Support was provided solely from institutional and/or
departmental sources.
Competing Interests
Correspondence
Address correspondence to Dr. Doufas: Department of
Anesthesiology, Perioperative, and Pain Medicine, Stanford
University School of Medicine, 300 Pasteur Drive, Stanford,
California 94305, and Outcomes Research Consortium,
Cleveland, Ohio. agdoufas@stanford.edu. Information on
purchasing reprints may be found at www.anesthesiology.
org or on the masthead page at the beginning of this issue. Anesthesiologys articles are made freely accessible to
all readers, for personal use only, 6 months from the cover
date of the issue.
References
1. White DP, Younes MK: Obstructive sleep apnea. Compr
Physiol 2012; 2:254194
2. Peppard PE, Young T, Barnet JH, Palta M, Hagen EW, Hla KM:
Increased prevalence of sleep-disordered breathing in adults.
Am J Epidemiol 2013; 177:100614
3. Finkel KJ, Searleman AC, Tymkew H, Tanaka CY, Saager L,
Safer-Zadeh E, Bottros M, Selvidge JA, Jacobsohn E, Pulley D,
Duntley S, Becker C, Avidan MS: Prevalence of undiagnosed
obstructive sleep apnea among adult surgical patients in an
academic medical center. Sleep Med 2009; 10:7538
4. Young T, Evans L, Finn L, Palta M: Estimation of the clinically
diagnosed proportion of sleep apnea syndrome in middleaged men and women. Sleep 1997; 20:7056
5. Singh M, Liao P, Kobah S, Wijeysundera DN, Shapiro C,
Chung F: Proportion of surgical patients with undiagnosed
obstructive sleep apnoea. Br J Anaesth 2013; 110:62936
6. Peppard PE, Young T, Palta M, Skatrud J: Prospective study
of the association between sleep-disordered breathing and
hypertension. N Engl J Med 2000; 342:137884
7. Yaggi HK, Concato J, Kernan WN, Lichtman JH, Brass LM,
Mohsenin V: Obstructive sleep apnea as a risk factor for
stroke and death. N Engl J Med 2005; 353:203441
8. Punjabi NM, Beamer BA: Alterations in glucose disposal in
sleep-disordered breathing. Am J Respir Crit Care Med 2009;
179:23540
9. Young T, Finn L, Peppard PE, Szklo-Coxe M, Austin D, Nieto
FJ, Stubbs R, Hla KM: Sleep disordered breathing and mortality: Eighteen-year follow-up of the Wisconsin sleep cohort.
Sleep 2008; 31:10718
10. Punjabi NM, Caffo BS, Goodwin JL, Gottlieb DJ, Newman AB,
OConnor GT, Rapoport DM, Redline S, Resnick HE, Robbins
JA, Shahar E, Unruh ML, Samet JM: Sleep-disordered breathing and mortality: A prospective cohort study. PLoS Med
2009; 6:e1000132
11. Punjabi NM, Newman AB, Young TB, Resnick HE, Sanders
MH: Sleep-disordered breathing and cardiovascular disease:
An outcome-based definition of hypopneas. Am J Respir Crit
Care Med 2008; 177:11505
12. Gami AS, Pressman G, Caples SM, Kanagala R, Gard JJ,
Davison DE, Malouf JF, Ammash NM, Friedman PA, Somers
VK: Association of atrial fibrillation and obstructive sleep
apnea. Circulation 2004; 110:3647
13. Somers VK, Dyken ME, Clary MP, Abboud FM: Sympathetic
neural mechanisms in obstructive sleep apnea. J Clin Invest
1995; 96:1897904
14. Narkiewicz K, van de Borne PJ, Cooley RL, Dyken ME,
Somers VK: Sympathetic activity in obese subjects with and
without obstructive sleep apnea. Circulation 1998; 98:7726
15. Baguet JP, Barone-Rochette G, Tamisier R, Levy P, Ppin
JL: Mechanisms of cardiac dysfunction in obstructive sleep
apnea. Nat Rev Cardiol 2012; 9:67988
16. Narkiewicz K, Montano N, Cogliati C, van de Borne PJ,
Dyken ME, Somers VK: Altered cardiovascular variability in
obstructive sleep apnea. Circulation 1998; 98:10717
17. Ryan S, Ward S, Heneghan C, McNicholas WT: Predictors of
decreased spontaneous baroreflex sensitivity in obstructive
sleep apnea syndrome. Chest 2007; 131:11007
18. Carlson JT, Hedner JA, Sellgren J, Elam M, Wallin BG:
Depressed baroreflex sensitivity in patients with obstructive
sleep apnea. Am J Respir Crit Care Med 1996; 154:14906
19. Grote L, Kraiczi H, Hedner J: Reduced alpha- and beta(2)adrenergic vascular response in patients with obstructive
sleep apnea. Am J Respir Crit Care Med 2000; 162(4 Pt.
1):14807
20. Silva AQ, Schreihofer AM: Altered sympathetic reflexes and

vascular reactivity in rats after exposure to chronic intermittent hypoxia. J Physiol 2011; 589(Pt. 6):146376
21. Stierer TL, Wright C, George A, Thompson RE, Wu CL, Collop
N: Risk assessment of obstructive sleep apnea in a population of patients undergoing ambulatory surgery. J Clin Sleep
Med 2010; 6:46772
22. Walsh M, Devereaux PJ, Garg AX, Kurz A, Turan A, Rodseth
RN, Cywinski J, Thabane L, Sessler DI: Relationship between
intraoperative mean arterial pressure and clinical outcomes
after noncardiac surgery: Toward an empirical definition of
hypotension. Anesthesiology 2013; 119:50715
23. Stirt JA, Frantz RA, Gunz EF, Conolly ME: Anesthesia, catecholamines, and hemodynamics in autonomic dysfunction.
24. Bijker JB, van Klei WA, Kappen TH, van Wolfswinkel L,
Moons KG, Kalkman CJ: Incidence of intraoperative hypotension as a function of the chosen definition: Literature definitions applied to a retrospective cohort using automated
data collection. Anesthesiology 2007; 107:21320
25. Urbano F, Roux F, Schindler J, Mohsenin V: Impaired cerebral autoregulation in obstructive sleep apnea. J Appl Physiol
(1985) 2008; 105:18527
26. Foster GE, Brugniaux JV, Pialoux V, Duggan CT, Hanly PJ,
Ahmed SB, Poulin MJ: Cardiovascular and cerebrovascular
responses to acute hypoxia following exposure to intermittent hypoxia in healthy humans. J Physiol 2009; 587(Pt.
13):328799
27. Kaw R, Chung F, Pasupuleti V, Mehta J, Gay PC, Hernandez
AV: Meta-analysis of the association between obstructive
sleep apnoea and postoperative outcome. Br J Anaesth 2012;
109:897906
28. Mokhlesi B, Hovda MD, Vekhter B, Arora VM, Chung F,
Meltzer DO: Sleep-disordered breathing and postoperative
outcomes after elective surgery: Analysis of the nationwide
inpatient sample. Chest 2013; 144:90314
29. Memtsoudis SG, Stundner O, Rasul R, Chiu YL, Sun X,
Ramachandran SK, Kaw R, Fleischut P, Mazumdar M:
The impact of sleep apnea on postoperative utilization
of resources and adverse outcomes. Anesth Analg 2014;
118:40718
30. American Academy of Sleep Medicine Task Force: Sleeprelated breathing disorders in adults: Recommendations for
syndrome definition and measurement techniques in clinical
research. Sleep 1999; 22:66789
31. Iber C, Ancoli-Israel S, Chesson A, Quan S; for the American
Academy of Sleep Medicine: International classification of
sleep disorders: Diagnostic and coding manual, 2nd edition.
Westchester, IL, American Academy of Sleep Medicine, 2005
32. American Sleep Disorders Association: EEG arousals: Scoring
rules and examples: A preliminary report from the Sleep
Disorders Atlas Task Force of the American Sleep Disorders
Association. Sleep 1992; 15:17384
33. Epstein LJ, Kristo D, Strollo PJ Jr, Friedman N, Malhotra
A, Patil SP, Ramar K, Rogers R, Schwab RJ, Weaver EM,
Weinstein MD; Adult Obstructive Sleep Apnea Task Force
of the American Academy of Sleep Medicine: Clinical guideline for the evaluation, management and long-term care of
obstructive sleep apnea in adults. J Clin Sleep Med 2009;
5:26376
34. McPherson ML: Demystifying opioid conversion calculations:
A guide for effective dosing, 1st edition. Bethesda, American
Society of Health-System Pharmacists, 2009
35. Dyer RA, Reed AR, van Dyk D, Arcache MJ, Hodges O,
Lombard CJ, Greenwood J, James MF: Hemodynamic effects
of ephedrine, phenylephrine, and the coadministration
of phenylephrine with oxytocin during spinal anesthesia
for elective cesarean delivery. Anesthesiology 2009; 111:
75365
36. Redline S, Yenokyan G, Gottlieb DJ, Shahar E, OConnor GT,

Resnick HE, Diener-West M, Sanders MH, Wolf PA, Geraghty
EM, Ali T, Lebowitz M, Punjabi NM: Obstructive sleep apnea
hypopnea and incident stroke: The sleep heart health study.
Am J Respir Crit Care Med 2010; 182:26977
37. Gami AS, Hodge DO, Herges RM, Olson EJ, Nykodym J, Kara
T, Somers VK: Obstructive sleep apnea, obesity, and the risk of
incident atrial fibrillation. J Am Coll Cardiol 2007; 49:56571
38. Punjabi NM, Shahar E, Redline S, Gottlieb DJ, Givelber R,
Resnick HE; Sleep Heart Health Study Investigators: Sleepdisordered breathing, glucose intolerance, and insulin resistance: The Sleep Heart Health Study. Am J Epidemiol 2004;
160:52130
39. Doufas AG, Tian L, Davies MF, Warby SC: Nocturnal intermittent hypoxia is independently associated with pain in subjects
suffering from sleep-disordered breathing. Anesthesiology
2013; 119:114962
40. Moore LE, Sharifpour M, Shanks A, Kheterpal S, Tremper
KK, Mashour GA: Cerebral perfusion pressure below 60mm
Hg is common in the intraoperative setting. J Neurosurg
Anesthesiol 2012; 24:5862
41. Sessler DI, Sigl JC, Kelley SD, Chamoun NG, Manberg PJ, Saager
L, Kurz A, Greenwald S: Hospital stay and mortality are increased
in patients having a triple low of low blood pressure, low
bispectral index, and low minimum alveolar concentration of
volatile anesthesia. Anesthesiology 2012; 116:1195203
42. Monk TG, Saini V, Weldon BC, Sigl JC: Anesthetic management and one-year mortality after noncardiac surgery. Anesth
Analg 2005; 100:410
43. Kheterpal S, OReilly M, Englesbe MJ, Rosenberg AL,
Shanks AM, Zhang L, Rothman ED, Campbell DA, Tremper
KK: Preoperative and intraoperative predictors of cardiac
adverse events after general, vascular, and urological surgery.
44. Bijker JB, van Klei WA, Vergouwe Y, Eleveld DJ, van
Wolfswinkel L, Moons KG, Kalkman CJ: Intraoperative
hypotension and 1-year mortality after noncardiac surgery.
45. Flink BJ, Rivelli SK, Cox EA, White WD, Falcone G, Vail TP,
Young CC, Bolognesi MP, Krystal AD, Trzepacz PT, Moon
RE, Kwatra MM: Obstructive sleep apnea and incidence
of postoperative delirium after elective knee replacement
in the nondemented elderly. Anesthesiology 2012; 116:
78896
46. Bateman BT, Eikermann M: Obstructive sleep apnea predicts adverse perioperative outcome: Evidence for an association between obstructive sleep apnea and delirium.
47. Safran DB, Orlando R 3rd: Physiologic effects of pneumoperitoneum. Am J Surg 1994; 167:2816
48. Struthers AD, Cuschieri A: Cardiovascular consequences of
laparoscopic surgery. Lancet 1998; 352:56870
Propofol Attenuated Acute Kidney Injury after

Orthotopic Liver Transplantation via Inhibiting Gap
Junction Composed of Connexin 32
Chenfang Luo, M.D., Dongdong Yuan, M.D., Xiaoyun Li, M.D., Weifeng Yao, M.D.,
Gangjian Luo, M.D., Xinjin Chi, M.D., Haobo Li, M.Sc., Michael G. Irwin, M.D.,
Zhengyuan Xia, M.D., Ph.D., Ziqing Hei, M.D., Ph.D.
ABSTRACT
Background: Postliver transplantation acute kidney injury (AKI) severely affects patient survival, whereas the mechanism is
unclear and effective therapy is lacking. The authors postulated that reperfusion induced enhancement of connexin32 (Cx32)
gap junction plays a critical role in mediating postliver transplantation AKI and that pretreatment/precondition with the
anesthetic propofol, known to inhibit gap junction, can confer effective protection.
Methods: Male SpragueDawley rats underwent autologous orthotopic liver transplantation (AOLT) in the absence or presence of treatments with the selective Cx32 inhibitor, 2-aminoethoxydiphenyl borate or propofol (50mg/kg) (n = 8 per group).
Also, kidney tubular epithelial (NRK-52E) cells were subjected to hypoxiareoxygenation and the function of Cx32 was
manipulated by three distinct mechanisms: cell culture in different density; pretreatment with Cx32 inhibitors or enhancer;
Cx32 gene knock-down (n = 4 to 5).
Results: AOLT resulted in significant increases of renal Cx32 protein expression and gap junction, which were coincident
with increases in oxidative stress and impairment in renal function and tissue injury as compared to sham group. Similarly,
hypoxiareoxygenation resulted in significant cellular injury manifested as reduced cell growth and increased lactate dehydrogenase release, which was significantly attenuated by Cx32 gene knock-down but exacerbated by Cx32 enhancement. Propofol
inhibited Cx32 function and attenuated post-AOLT AKI. In NRK-52E cells, propofol reduced posthypoxic reactive oxygen
species production and attenuated cellular injury, and the cellular protective effects of propofol were reinforced by Cx32 inhibition but cancelled by Cx32 enhancement.
Conclusion: Cx32 plays a critical role in AOLT-induced AKI and that inhibition of Cx32 function may represent a new and
major mechanism whereby propofol reduces oxidative stress and subsequently attenuates post-AOLT AKI. (Anesthesiology
2015; 122:72-86)
OSTOPERATIVE acute kidney injury (AKI) is a

severe complication following liver transplantation,
which influences patient survival adversely.1 Reportedly,
approximately 30 to 50% of the patients undergoing liver
transplantation developed AKI, however, mechanisms
contribute to this complication is still unclear.2 Causes
of this complication are complicated and involve multi
factors, among which perioperative hypotension is considered to be one of the most important independent
risk factors. Perioperative hypotension always results in
renal ischemiareperfusion (I/R) injury and mechanism
of which is associated with oxidative stress.3,4 However,
effective therapy to combat postliver transplantation AKI
is lacking. Therefore, there is urgent need to explore the
underlying mechanisms of AKI and develop effective strategies for renal protection.

Acute kidney injury occurs after specific surgeries and in
creases patient morbidity and mortality

Anesthetized rats underwent autologous orthotopic liver
transplantation in the absence or presence of treatments with
a selective Cx32 inhibitor, 2-aminoethoxydiphenyl borate, or
propofol
Propofol inhibited Cx32 function and attenuated postautolo
gous orthotopic liver transplantation acute kidney injury
Connexins are a big family of transmembrane proteins

that express in all human organs and tissues. Approximately
21 isoforms have been found, and each of them forms channels with distinct regulation and permeability.5,6 Six of connexins compose a hemi-channel. Two hemi-channels in the
Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are available in both the
HTML and PDF versions of this article. Links to the digital files are provided in the HTML text of this article on the Journals Web site (www.
anesthesiology.org). The first three authors have contributed equally to this study. The last two authors share senior authorship.
Submitted for publication January 12, 2014. Accepted for publication August 6, 2014. From the Department of Anesthesiology, The Third
Affiliated Hospital of Sun Yat-Sen University, Guangzhou, Peoples Republic of China (C.L., D.Y., X.L., W.Y., G.L., X.C., Z.H.); and Department
of Anesthesiology, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong, Peoples Republic of China (H.L., M.G.I., Z.X.).
January 2015
neighboring cells dock together to form an integral gap junction (GJ), which manipulates the direct cell-to-cell signaling
transfer (molecule weight is less than 1kDa, such as calcium,
cyclic adenosine monophosphate, cyclic guanosine monophosphate, glutathione, etc.). This kind of signaling transfer
is not only essential for many physiological events, but also
relative with development of some disease.7,8 A most recent
study shows connexin32 (Cx32) expression always increases
when liver or brain damage become more and more serious and these damages could be blocked by Cx32 inhibitors
through decreasing oxidative stress and cell apoptosis.9,10
Cx32 expression is extremely abundant in kidney, especially
in renal proximal tubules, however, roles of Cx32 in the
development of AKI following liver transplantation is still
unknown.11 So far, effective therapeutic strategies for perioperative renal protection, especially targeting AKI after liver
transplantation are lacking.
As a commonly used anesthetic in clinical anesthesia and
intensive care unit sedation, propofol not only regulates the
balance between oxidants and antioxidants, but also confers
protective effects for different organs via different mechanisms, such as antioxidant,12,13 antiinflammatory effects,14
inhibition of inducible nitric oxide synthase enzyme and so
on.15 Xia et al.12 reported propofol reduced postischemic
oxidative stress and attenuated myocardial I/R injury in
patients, and Zhao et al.16 showed that propofol decreased
X-ray induced cellular toxicity through inhibiting Cx32.
Based on the above-mentioned findings illustrating that GJ
function inhibition reduced postischemic oxidative stress
and that propofol can inhibit Cx32 which is abundant in the
kidney, we hypothesized that GJ composed of Cx32 played
a critical role in the development of AKI after liver transplantation and propofol could protect against posttransplantation induced AKI through inhibiting Cx32, mechanism
of which might be relative with propofol reduced reactive
oxygen species (ROS) production and attenuated oxidative
stress in the setting of AKI through Cx32 GJ inhibition.
This viewpoint has never been reported until now. The new
mechanistic insight gained from the current study would
provide a basis for developing effective therapies to combat
liver transplantation-mediated AKI.

Animals and Treatment
Animal care followed National Institutes of Health criteria
for the care and use of laboratory animals in research. The
study was approved by the Laboratory Animal Care Committee of Sun Yat-Sen University (Guangzhou, Guangdong,
China). Male SpragueDawley rats (200 to 220g) were
obtained from Sun Yat-Sen University. In the initial intervention model establishment study, the animals were randomly
assigned into six parallel groups (n = 8 per group) using a
random number table taking into consideration of the
weight of the rats. These include sham operated, reperfusion
4-, 8-, 16-, 24-, and 48-h groups. Subsequent in vivo studies
were performed using the 8-h reperfusion model. The animals were then randomly assigned to four groups (n = 8 per
group) to explore the role of GJ function in postautologous
orthotopic liver transplantation [AOLT] renal injury, which
include sham, (2-aminoethoxydiphenyl borate [2-APB]) +
sham, AOLT, and (2-APB) + AOLT groups. Furthermore,
to explore the effectiveness and mechanisms of propofol protection, animals were randomly assigned to four groups (n =
8 per group) which include sham, propofol + sham, AOLT,
and propofol + AOLT. It is of notice that, with respect to our
success rate of approximately 80 to 90% for AOLT model
establishment,17,18 experiments were performed on a total of
129 animals.
Before operation, rats were treated intraperitoneally with
propofol (Sigma-Aldrich, St. Louis, MO) at doses 50mg/kg
for 3 days,19 and 2-APB (Sigma-Aldrich; a relative specific
inhibitor of Cx32) was given at 20mg/kg 4h20 before AOLT
or Scrape-and-load assay.20 The dose of propofol (i.e.,
50mg/kg, intraperitoneally) was chosen based on a preliminary experiment showing that propofol 50mg/kg, produced
a sedative response in rats, as determined by loss of reflex
responses to a painful stimulus while remaining sensitive to
skin incision.19 This dosage of propofol has been widely used
in in vivo studies in rats21,22 and produced beneficial effects.
Establishment of Rat AOLT Models
To minimize experimenter bias, the investigator who performed the operation was blinded to the treatment conditions. Rat AOLT models were established as previously
described.17,18 Initially, we used an open face guard to administer the inhalational ether anesthesia until the rats had no
response to a needle stimulus. After entering the abdominal
cavity, we resected and ligated the falciform ligament of the
liver, and severed the blood vessel along the esophagus. The
liver was revealed until the supra hepatic vena cava (SVC)
was liberated and then, the liver was placed back into its
original position. We prepared a bold line to guide the SVC
for blockage. When the upper region of the left renal vein
was completely liberated, the inferior vena cava (IVC) was
dissociated. We dissected the first hepatic portal and then
separated the portal vein (PV) from the convergence of the
inferior mesenteric and splenic veins. Both of the hepatic
artery and biliary were also liberated successively according to their anatomic relationship. Subsequently, the portal hepatics were ligated. Microvascular clamps were used
at the convergence of the inferior mesenteric, splenic veins,
hepatic artery, SVC, and IVC. The PV was punctured with
a 24-gauge needle in preparation for reperfusion and one
1-mm incision was made on the IVC wall as an outflow
tract. Precold 4C Ringer lactate solution was injected during reperfusion at 2.5ml/min until the liver color turned
yellow. Finally, after extracting the needle, the opening of
the PV and IVC was closed using 8-0 sutures. PV, SVC,
IVC, and hepatic artery were all unclamped. On average,
Anesthesiology 2015; 122:72-86 73 Luo et al.
Propofol Reduced Post-AOLT AKI via Inhibiting Cx32
the anhepatic phase lasted for 201min, which was concomitant with significantly lower mean arterial pressure in
this phase (vascular clamped) than in other phases as shown
in table1.
Mean Arterial Pressure Assay
The right femoral arteries were catheterized with a polyethylene catheter (outer diameter, 0.965mm; inner diameter,
0.58mm) for monitoring mean arterial pressure,23 which
was recorded before this operation and when PV, SVC, IVC,
and hepatic artery were clamped. Also, mean arterial pressure was recorded at different time points (1 min to 30 min),
when PV, SVC, IVC, and hepatic artery were unclamped.
Scrape-and-load Assay
Rats were treated with propofol (50mg/kg) or 2-APB
(20mg/kg) intraperitoneally. Four hours later, kidneys were
excised and freshly sliced. We placed a 27-gauge needle dipping into a solution containing 0.5% Lucifer Yellow (Invitrogen, Carlsbad, CA). The needle was used to mechanically
damage a small area of each slice to apply dyes. After 5min
incubation, kidney slices were rinsed in saline, fixed in 4%
paraformaldehyde for 30min, frozen in optimum cutting
temperature compound, cryosectioned into 10 m sections,
rinsed in saline again, mounted, and imaged by fluorescence
microscopy.20 Quantitative analysis of the distance of dye
spread was performed between the dye transfer front and
the scrape line.
Assessment of Kidney Damage
Creatinine and blood urea nitrogen were measured in blood
samples with an automatic biochemistry analyzer (Hitachi
7600- 020/7170A, Tokyo, Japan). Kidney specimens were
fixed in 10% buffered formalin, embedded in paraffin, and
processed for hematoxylineosin staining.
Measurement of Intracellular ROS Production, 15-F2tIsoprostane, and H2O2
Intracellular ROS production was estimated by using
2,7-dichlorofluorescein diacetate (Sigma-Aldrich). Formation of dichlorofluorescin was detected by fluorescence
spectrometer (HITACHI, Model No. F4500, Tokyo, Japan)
equipped with a fluorescein isothiocyanate filter at 488nm
and emission at 525
nm within 15
min. Dichlorofluorescin images were obtained with a fluorescence microscope
(Olympus IX71, Tokyo, Japan) at a magnification of 200
using identical acquisition settings for each section.24 15-F2tIsoprostane (15-F2t-Isop) and H2O2 levels are determined
using commercial assay kits (Cayman Chemical Company,
Ann Arbor, MI) as described.25
Cell Culture
NRK-52E cells (kidney tubular epithelial cells) were obtained
from American Type Culture Collection (Manassas, VA)
and cultured in Dulbeccos Modified Eagles Medium/F-12
supplemented with 10% fetal bovine serum. HeLa cells

expressing Cx32 under the control of a bidirectional tetracycline-inducible promoter was characterized previously26 and
cultured in Dulbeccos Modified Eagles Medium supplemented with 10% fetal bovine serum. Both kinds of cells
were grown at 37C in an atmosphere of 5% CO2 in air.
Cell Treatments and Survival Assay
NRK-52E cells were seeded at low density (25,000 cells/
cm2, no GJ formed) or high density (125,000 cells/cm2, GJ
formed) in 24-well plates to observe effects of GJ on hypoxia
reoxygenation (H/R) injury or other biochemical changes as
assessed below.26 The choosing of NRK-52E cell low density
at 25,000 cells/cm2 and high density at 125,000 cells/cm2 was
not only based on whether or not there was GJ formation,
but also based on our preliminary study findings showing that
the rates of cell proliferation peaked at 25,000 cells/cm2 and
maintained at a comparable level when the density was as high
as 125,000 cells/cm2 (data not shown) to exclude the possibility that NRK-52E cell growth is significantly down-regulated
due to reduced cell spreading when its density is lower than
20,000 cells/cm2 or higher than 200,000 cells/cm2.27
NRK-52E cells were pretreated with connexin channel
inhibitors 18--GA (GA), 10 M, for 1h (Sigma-Aldrich),
2-APB, 25 M, for 1h (Sigma-Aldrich); a Cx32 expression enhancer, retinoic acid (RA) 10 M, for 24h (SigmaAldrich) before inducing H/R injury and Parachute
dye-coupling assay was performed as described below.
In the preliminary dose-finding study, cells were pretreated with propofol (Sigma-Aldrich), respectively, at 1,
5, 15, 30, and 60 M for 1h before being processed for
Parachute dye-coupling assay, Cell Counting Kit-8 assay
(Dojindo Molecular Technologies, Tokyo, Japan), and
Western blotting analysis. And the concentration of 15
M, of propofol, was chosen for ensuring studies in NRK52E cells and these cells were pretreated for 1h with 15 M
propofol before being subjected to H/R injury. The choosing of propofol concentration at 15 M was based on our
dose-finding study results showing that this concentration
of propofol can profoundly decrease dye coupling and also
based on our previous study showing that propofol at 15
M can reduce the cytotoxicity of radiograph irrasiation
through inhibiting GJ activity.16 This concentration of propofol (15 M) is in the range of target plasma concentration of propofol 2 to 4 g/ml (i.e., 11 to 22 M) as used
clinically during major surgeries.28,29
H/R of NRK-52E Cells
Renal hypoperfusion is significant during liver transplantation, which is induced by hypotension. Renal hypoperfusion always results in renal I/R injury. Thus, we employed
NRK-52E cell H/R model as an in vitro study to mimic
renal cell I/R injury. Before the experiment, NRK-52E
cells were incubated in starving medium (serum-free and
glucose-free Dulbeccos Modified Eagles Medium/F-12)
Table 1. The Levels of Rat MAP during AOLT

Time Point
Before operation
Before vascular clamped
Vascular clamped for 1 min
Vascular clamped for 10 min
Vascular clamped for 20min (before
vascular unclamped)
Vascular unclamped for 1 min
Before operation finished
After operation for 1 min
After operation for 10 min
MAP (mmHg)
937.9
1019.1
448.5*
387.7*
406.3*
725.5
866.1
929.5
996.9
954.6
1036.8
Changes of rat MAP during AOLT. Data are mean SEM, n = 8 per group.
* P < 0.05 vs. before operation and before vascular clamping; P < 0.05 vs.
vascular clamping for 1, 10, and 20min, respectively.
AOLT = autologous orthotopic liver transplantation; MAP = mean artery
blood pressure.
for 12h and then exposed to hypoxia. In all the process of

H/R, cells were incubated in starving medium. NRK-52E
cells were cultured in a low-oxygen condition (95% N2 + 5%
CO2) for 24h in a humidified hypoxia incubator (Galaxy
48R; Eppendorf, Hamburg, Germany). The cells were then
exposed to normal oxygen condition (95% air + 5% CO2)
for reoxygenation for 4h. After the completion of the experiments, the supernatant and cells were collected separately for
further analysis.30 Control groups were cultured in normoxic
conditions for 28h, coincident with duration of H/R injury,
and the supernatant and cells were also harvested separately
for further analysis
Cell Counting Kit-8 Assay and Lactate Dehydrogenase
(LDH) Assay
Cells were seeded at low density (25,000 cells/cm2, no GJ
formed) or high density (125,000 cells/cm2, GJ formed)
in 24-well plates. At the end of different stimulation, Cell
Counting Kit-8 and LDH assays (Roche Diagnostics, Indianapolis, IN) were carried out according to the manufacturers
introduction as we described.25 According to the instruction
of the detection kit, to calculate percent cytotoxicity, three
kinds of controls (namely background control, low control
and high control) were included in each experiment. The
background control determines the LDH activity contained
in the assay medium. The absorbance value obtained from
this control was subtracted from all other absorbance values.
The low control determines the LDH activity released from
the untreated cells (spontaneous LDH release), whereas
the high control determines the maximum releasable LDH
activity in the cells (maximum LDH release). In the high
control, lysis reagent was added to the samples to get the
accurate estimate of maximum releasable LDH. To determine the percentage cytotoxicity, the average absorbance
values of the samples and controls were respectively calculated. And, the absorbance values of the background control
were subtracted from all other absorbance values. Thereafter,

the resulting extent of cytotoxicity (reflected as increased
LDH activity) was calculated using the following equation:
cytotoxicity (%) = (sample absorbance low control absorbance)/(high control absorbance low control absorbance)
100%. Thus, the value of LDH release we presented in
the current study is reflected as LDH release from treatment groups relative to high control (cells treated with lysis).
Then, we standardized the results of cytotoxicity (relative
LDH release) of the experimental control group as 1 and
the values of other groups were presented as relative values
compared to experimental control, with the purpose to easily
visualize the changes of cytotoxicity (relative LDH release)
resulted from changes of Cx32.
Parachute Dye-coupling Assay
GJ function was examined with Parachute dye-coupling
assay as described.5
Inhibition of Cx32 Expression by Small Interfering RNA
Transfection
Cells were transfected with two different kinds of small
interfering RNA (siRNA) targeting rat Cx32 gene (CACCAACAACACATAGAAA and GCATCTGCATTATCCTCAA, Cx32 siRNA1 and Cx32 siRNA2) or a nonspecific,
control siRNA (NC group). Transfection into NRK-52E
cells was carried out using Lipofectamine 2000 (Invitrogen)
according to the manufacturers instructions.
Western Blotting
Western blotting follows the standard procedures as
described.26 Anti-Cx32 (1:3,000; Sigma-Aldrich) and secondary antibody (1:2,000; Sigma-Aldrich) were used to
detect Cx32 expression. Anti--actin (Sigma-Aldrich) and
its corresponding secondary antibody were used at 1:4,000.
Quantitative data are presented as mean SEM. Statistical
analysis was performed by using SPSS 15.0 software (SPSS
Inc., Chicago, IL) and doseresponse curve fitting was made
by Sigmaplot 10.0 (Systat Software, Inc., Chicago, IL). And
the cure was formed by the graph properties, using the function of smoothed (spline) in the shape property. Multiple
comparisons among groups were analyzed using repeated
measures one-way ANOVA, followed by Tukey post hoc
comparisons. In our lab, the overall success rate of rat AOLT
model establishment is between 80 and 90%. According to
our previous studies,17 a sample size of n = 8 is needed and is
sufficient to achieve a statistical power in the study of intervention effectiveness. Therefore, when designing our studies,
we initially assigned 10 rats per groups while aiming for a
final n = 8 in each group, bearing in mind the rate of success
of the model. A two-tailed P value less than 0.05 was considered statistically significantly different.
Results
Remote Kidney Damage following AOLT In Vivo Studies at
Different Time Points
To explore effects of liver transplantation on kidneys, a well
established rat AOLT model31,32 was used in this study. In
this model, inferior vena cava interruption and significant
systemic hypotension occurred during the anhepatic phrase
which led to blood reflux disorder and caused renal ischemic
injury. In in vivo studies, we found that as reperfusion time
being extended, kidney pathological damage became more
and more serious gradually. Eight hours after reperfusion,
kidney damage became the most severe. Tissue injury recovered at 48h after reperfusion (fig.1A). Liver transplantation
is very complex. Hypotension is one of the most important
and obvious hemodynamic changes in the process, which
may result in remote organs injuries, including kidneys.
Thus, rat mean arterial pressure was recorded during AOLT.
Results showed that when PV, SVC, IVC, and hepatic artery
were clamped, mean arterial pressure decreased dramatically
while recovered gradually to the normal level as vascular
unclamped (table 1). Increase of creatinine and blood urea
nitrogen (fig. 1, B and C) (parameters of renal functional
impairment) as well as significant elevations in tissue H2O2
(fig. 1D) and 15-F2t-isoprostane (15-F2t-Isop) (fig. 1E) production, reflecting the level of oxidative stress,13 all peaked at
8h after post-AOLT, which mirrored the patterns of pathological injury of kidney.
Changes of Cx32 Expression after AOLT In Vivo Studies at
Different Time Point
Given that Cx32 expressed richly in kidney and that its overexpression has been shown to be related with organ damage,
so it was determined after AOLT. As shown in figure 1F, Cx32
protein was increased after AOLT and peaked at 8h after
reperfusion, which was coincident with the most severe kidney
pathological damage and functional impairment (fig. 1, AC).
Inhibition of Renal Cx32 GJ Attenuated Remote Kidney
Damage following AOLT with Concomitant Reduction of
Post-AOLT Renal Oxidative Stress
Results in figure 1 provided a clue that Cx32 might play an
important role in AOLT-mediated AKI. Therefore, 2-APB,
a relatively specific inhibitor of Cx32 channels20,33 was used
to explore roles of Cx32 in this pathology. Eight samples
are contained in every group. Scrape-and-load assay was
employed to detect GJ function composed of Cx32 in
vivo. Results indicated that dye spread was decreased obviously on kidney slices obtained from rats pretreated with
2-APB (fig.2A). Renal pathological damage, creatinine
and blood urea nitrogen were also reduced obviously (fig.
2, B and C) following the treatment with 2-APB. These
data indicate that inhibition of GJ function composed of
Cx32, could eliminate remote kidney damage after AOLT.
Of note, 2-APB treatment also significantly reduced ALOT
mediated renal productions of H2O2 and 15-F2t-Isop to a

level comparable to that in sham group (all P < 0.05, 2-APB
+ AOLT vs. AOLT; P > 0.05, 2-APB + AOLT vs. sham; fig.
2D). Dimethyl sulfoxide, the vehicle control of 2-APB had
no effect on the parameters (see fig. 1, Supplemental Digital
Content 1, http://links.lww.com/ALN/B94).
Cell Damage after H/R Was Cell Density Dependent
Results in figure 2 indicated that kidney damage after
AOLT could be attenuated by inhibiting Cx32 GJ function. To confirm this mechanism, we tested it on NRK52E cells, a kind of renal tubular epithelial cell line, which
mainly express Cx32. Figure3 illustrated effects of 24h
hypoxia and 4h reoxygenation (H24R4) on NRK-52E
cell survival and LDH release at low-density (25,000
cells/cm2, no GJ formed) and high-density cell culture
(125,000 cells/cm2, GJ formed) respectively, which is a
widely used method to observe effects of GJ.26 After being
subjected to H24R4, NRK-52E cell growth was reduced
at both low-density and high-density cell culture, by
approximately 43 and 71% respectively (P < 0.05 high
density vs. low density; fig. 3A). When cells were in contact with one another at high density, H24R4 damage was
substantially greater than that at low density, manifested
as greater extent of LDH release in the high-density condition than in the low-density condition (P < 0.02; fig.
3B) when the values as adjusted with their relative control.
These results indicated that posthypoxic cell damage was
density-dependent, being more severe in the high-density
condition where GJ was formed.
H24R4 Caused Densitydependent Cell Damage through GJ
Density-dependent cell damage is always modulated by GJ.34
Thus, we investigated roles of Cx32 gap function in H24R4
damage. Different methods were used to manipulate GJ function composed of Cx32 on NRK-52E cells. Dye coupling
was reduced by 18--GA (GA) and 2-APB, but increased by
the enhancer RA (fig.4A). At the same time, H24R4 damage was also reduced subsequent to pretreatment with GA
or 2-APB but exacerbated after RA treatment in NRK-52E
cells cultured in high-density (fig. 4B). However, cell growth
after H24R4 did not significantly differ at low-density cell
culture irrespective of treatments with either GA or 2-APB
or RA as compared with untreated control group (fig. 4B).
Changes in LDH release was just opposed the changes in cell
growth. Following preincubation with GA or 2-APB, LDH
release declined more obviously, treatment with RA resulted
in dramatic increase in LDH release at high-density cell culture (fig. 4C). Although, the same treatments were applied
on low-density cell culture, they did not cause significant
changes in posthypoxic LDH release (fig. 4C) and dimethyl
sulfoxide, the vehicle control of 2-APB, GA and RA had no
effects on the parameters in figure 4 (see fig. 2, Supplemental
Digital Content 1, http://links.lww.com/ALN/B94). These
Fig. 1. Remote kidney damage of rats following autologous orthotopic liver transplantation (AOLT) at different time points in vivo.
(A) Remote kidney damage of rats at different reperfusion time points after AOLT (hematoxylineosin staining; original magnification 200). (B and C) Levels of creatinine (Cr) and blood urea nitrogen (BUN) at different reperfusion time points after AOLT.
(D and E) Levels of H2O2 and 15-F2t-Isop of rat kidney at different reperfusion time points after AOLT. (F) Connexin32 (Cx32)
expression on rat kidney at different reperfusion time points after AOLT. Every group contains eight samples, n = 8. *P < 0.05
versus Sham group; #P < 0.05 versus 8h after reperfusion in BF.
findings suggested that Cx32 GJ plays a key role in posthypoxic cellular damage in NRK-52E cells.
Cx32 Gene Knock-down to Inhibit Cx32 GJ Function
Prevented H24R4-mediated Cell Damage in NRK-52E Cells
To confirm the impact of Cx32 function on H/R damage,
we synthesized two different Cx32 siRNAs (siRNA1 and
siRNA2) to specifically knock-down Cx32 expression
(fig.5A). Blocking Cx32 expression depressed dye coupling of NRK-52E cells (fig. 5, A and B). Although
Cx32 knock-down per se did not affect cell growth under
control condition, it significantly attenuated H24R4induced reduction of cell growth (fig. 5C) and reduced
posthypoxic LDH release (fig. 5D). The data are indicative that Cx32 GJ plays a key role in H/R-induced cell
damage.
Fig. 2. Remote kidney damage following autologous orthotopic liver transplantation (AOLT) was attenuated through inhibiting
gap junction (GJ) function composed of connexin32 (Cx32) in vivo. (A) Scrape-and-load assay was used to valuate functional
GJ in kidney tissue. Rats were treated with 2-aminoethoxydiphenyl borate (2-APB, 20mg/kg) or corresponding solvent for 3h.
Function of GJ is demonstrated by the spread of GJ-permeable Lucifer yellow. (B) Remote kidney damage of rat exposure to
2-APB (20mg/kg) or corresponding solvent for 3h before AOLT (hematoxylineosin staining; original magnification 200). (C) Serum creatinine (Cr) and blood urea nitrogen (BUN) levels of rats after reperfusion 8h, when rats were exposed to 2-APB (20mg/kg)
or corresponding solvent for 3h before AOLT. (D) H2O2 and 15-F2t-Isop levels of rat kidney after reperfusion 8h, when rats
were exposed to 2-APB (20mg/kg) or corresponding solvent for 3h before AOLT. Every group contains eight samples, n = 8.
*P < 0.05 versus Sham; #P < 0.05 versus AOLT group. Vehicle control of 2-APB is dimethyl sulfoxide, which has no effect on the
above-mentioned parameters (see fig. 1, Supplemental Digital Content 1, http://links.lww.com/ALN/B94).
Propofol Protected against H24R4-induced Cell Injury by

Inhibiting Cx32 GJ
Propofol had recently been shown to reduce kidney I/R
injury induced by hyperglycemia,35 a pathological condition
that is associated with increased oxidative stress. To assess
whether or not propofol can reduce H24R4-mediated renal

damage via inhibiting Cx32, we first examined the effects
of propofol on dye coupling of NRK-52E cells. Figure6,
AC, showed that propofol from 5 to 60 M decreased dye
coupling in a dose-dependent manner without cytotoxicity
production in low-density cells subjected to H24R4 but

completely cancelled H24R4-induced increases in LDH and
15-F2t-Isop that at high-density cells (fig. 7A), whereas lipid,
the vehicle control of propofol, had no effects on the results
in figures 6 and 7 (see fig. 2, Supplemental Digital Content
1, http://links.lww.com/ALN/B94). These results suggest
that inhibition of GJ function may be a major mechanism
whereby propofol attenuated oxidative stress and protected
against H24R4-induced cell damage.
Fig. 3. NRK-52E (cell line of normal rat kidney tubular epithelial

cells) cell damage of hypoxia for 24h and reoxygenation for 4h
(H24R4) was cell density dependent. (A) NRK-52E cell growth
at low- and high-density cell culture. (B) Relative lactate dehydrogenase (LDH) release of NRK-52E cells at low- and highdensity cell culture. Data were obtained from five independent
experiments each performed in quintuplicate, n = 5. *P < 0.05
versus control; #P < 0.05 versus H24R4 group at low-density
cell culture.
and had no significant effect on Cx32 expression in control

NRK-52E cells. By contrast, propofol at a concentration as
low as 15 M dramatically decreased dye coupling in Hela
cells that stably express Cx32 only, which means that the
effects of propofol on Cx32 expression and Cx32 GJ was
direct and specific (fig. 6D). This result suggests that propofol can profoundly inhibit Cx32 expression in situations
with Cx32 overexpression such as during AKI or acute renal
cell H/R injury. Also, the facts that GJ function inhibition
reduced H24R4-mediated cell damage (figs. 4 and 5) and
that propofol reduced dye coupling directly in Hela-Cx32
cells (fig. 6) collectively raised the possibility that propofol
could manipulate H24R4 damage via regulating Cx32 GJ
function. To test this hypothesis, NRK-52E cells seeded at
either low or high density were pretreated with propofol 15
M for 1h before being exposed to H24R4. Of note, propofol did not significantly enhance cell growth in low-density
cultured cells subjected to H24R4 but most profoundly
increased cell growth in high-density cultured cells subjected
to H24R4 (fig.7A). Propofol only slightly (despite statistically significantly) decreased LDH release and 15-F2t-Isop
Effects of Propofol on ROS Induced by H24R4 Damage

Mediated by GJ Composed of Cx32
The facts that organ-protective effects of propofol were
related to its effect in reducing ROS- mediated damage,36
together with our current findings that propofol could protect against H24R4-induced cell damage through inhibiting GJ prompted us to explore the relationship between
propofol-mediated GJ inhibition and ROS production. At
low-density cell culture, neither the GJ inhibitor 2-APB nor
the enhancer RA or propofol had effects on intracellular ROS
production in NRK-52E cells exposed to H24R4 (fig. 7B).
However, at high-density cell culture, ROS level fluctuated
as Cx32 function changed (fig. 7, B and C). As shown in
figure 7C, propofol similar to 2-APB profoundly inhibited
H24R4-induced increase of dichlorofluorescin florescence (a
reflection of H2O2 production) whereas RA further increased
dichlorofluorescin florescence in high-density cells, which
was quantified in figure 7B. These results demonstrated
that intracellular ROS was manipulated by GJ composed of
Cx32. Facts in this paragraph above suggested inhibition of
GJ is a major mechanism by which propofol reduced ROS,
at least at the current experimental settings. We had clarified
that propofol protected NRK-52E cells at low-density cell
culture, notwithstanding much less than that at high-density
cell culture (fig. 7A). The facts indicted propofol protected
cells through different mechanisms, one of which might be
relative with GJ. When NRK-52E cells (high-density cell culture) were pretreated with 2-APB or RA in combination with
propofol before H24R4, 2-APB magnified propofol protection but RA diminished it obviously (fig. 7, D and E). These
results strongly suggest that inhibition of Cx32 GJ may represent a major mechanism by which propofol confers cellular or
organ protection where Cx32 is overexpressed. This finding
might have clinical implications in guiding the optimal application of propofol for organ protection.
Propofol Decreased Remote Kidney Damage after AOLT
In Vivo via Inhibiting Cx32 GJ Function
Following our above demonstration in figures 6 and 7 that
propofol could protect NRK-52E cells against H/R damage through mediating Cx32 channel function, which
was related to inhibition of intracellular ROS, we wanted
to further confirm the protective effects and mechanism
of propofol in attenuating AOLT induced remote AKI in
vivo. With scrape-and-load assay, we demonstrated that
Fig. 4. NRK-52E cell damage of hypoxia for 24h and reoxygenation for 4h (H24R4) was regulated by gap junction (GJ) inhibitor
and enhancer. (A) Parachute dye-coupling assay was used to determine effects of inhibitors 18--GA (GA, 10 M) and 2-aminoethoxydiphenyl borate (2-APB, 25 M) decreasing GJ function but enhancer retinoic acid (RA, 10 M) increasing GJ function.
(B) Effects of GA (10 M), 2-APB (25 M), and RA (10 M) on NRK-52E cell growth at different cell density culture. (C) Effects of
GA (10 M), 2-APB (25 M) and RA (10 M) on relative lactate dehydrogenase (LDH) release of NRK-52E cells at different cell
density culture. Data were obtained from five independent experiments each performed in quintuplicate, n = 5. *P < 0.05 versus
control; #P < 0.05 versus H24R4 group at high-density cell culture. (B and C) Bars of control in low- or high-density cell culture
did not display in this figure, and bars of control were considered to be 1. Vehicle controls of 2-APB, GA (18--glycyrrhetinic
acid), and RA are all dimethyl sulfoxide, which have no significant effects on the above-mentioned parameters (see fig. 2,
Supplemental Digital Content 1, http://links.lww.com/ALN/B94).
propofol pretreatment decreased dye coupling on kidney

(fig.8A) that was associated with significant reduction in
renal Cx32 expression in rats subjected to AOLT but not
in sham-operated rats (fig. 8B). Furthermore, two different
methods, immunohistochemisty and immunofluorescence
were carried out to visualize the expression and localization
of Cx32. Both of the immunohistochemisty and immunofluorescence results (see figs. 3 and 4, Supplemental Digital
Content 1, http://links.lww.com/ALN/B94) showed that
in the areas of the renal tubule, Cx32 expression changed
obviously after AOLT, but by contrast, Cx32 expression in
the renal glomerulus was much less obvious than that in the
renal tubule. Cx32 expression increased in the AOLT group,
which could be inhibited by propofol, whereas lipid had
no effect on renal Cx32 expression and localization. These
results were coincident with the changes of Cx32 assessed
by Western blotting in figure 8B. Also, remote kidney damage was attenuated in the presence of propofol treatment
manifested as attenuated pathological damages (fig. 8C) and

reduced post-AOLT concentrations of creatinine and blood
urea nitrogen (fig. 8D). In line with its prevention of AOLT
induced increase in Cx32 expression and Cx32 GJ (fig. 8A),
propofol completely prevented AOLT induced increase in
H2O2 and 15-F2t-Isop production, but has no effect on
H2O2 and 15-F2t-Isop in sham-operated rats (fig. 8E). Findings as shown in figure 8 from in vivo studied confirmed
our in vitro study findings illustrated in figure 7, indicating
that inhibition of Cx32 GJ function is a key mechanism by
which propofol attenuated AOLT induced renal I/R injury.
The facts that propofol inhibited ROS production only in
high-density cells subjected to H24R4 (fig. 7B) where Cx32
GJ function was enhanced and that the enhancer RA cancelled propofol-mediated prevention of NRK-52E cells
against H/R injury (fig. 7, D and E) and cancelled propofolmediated attenuation of ROS production in NRK-52E cells
subjected to H/R (data not shown) indicates that inhibition
Fig. 5. Specific small interfering RNA (siRNA)modulated gap junction function composed of connexin32 (Cx32) and attenuated
NRK-52E cell damage of hypoxia for 24h and reoxygenation for 4h (H24R4). (A) Two different specific siRNAs decreased Cx32
expression. (B) Parachute dye-coupling assay was used to determine effects of these two siRNAs on gap junction function
composed of Cx32. (C) NRK-52E cell growth increased by these two different siRNAs when exposed to H24R4. (D) Relative
lactate dehydrogenase (LDH) release of NRK-52E cells decreased by these two different siRNAs when exposed to H24R4. Data
were obtained from four independent experiments each performed in quintuplicate, n = 4. *P < 0.05 versus control; #P < 0.05
versus H24R4 group.
of Cx32 GJ function is a major and novel mechanism of

propofol protection against AKI. This notion is confirmed in
in vivo models of AOLT where propofol prevented ALOTinduced Cx32 overexpression and increased in ROS production coincidently without affecting Cx32 expression and
ROS production on sham control rats (fig. 8).
Discussion
Liver transplantation is always considered to be the most
effective method to cure the final-stage liver disease. The
incidence of AKI is always at a high level and influences
patient survival severely, but the underlying mechanism
remains unclear.37 In our investigation, we chose a rat AOLT
model to explore Cx32 function in renal protection. Compared with pure hepatic I/R injury model, rat AOLT model
not only includes most procedures of liver transplantation,
but also includes blood reflux disorder. Compared with allogenic orthotopic liver transplantation, rat AOLT model has
its own advantages, such as avoiding complex condition of
recipients and reject reaction, keeping good repeatability and
high survival of rats. It is profitable for us to investigate the
role of Cx32 GJ in kidney damage following liver transplantation and the protective effects of propofol in this pathology.
Our in vitro studies, incorporating multiple manipulations,
provided strong evidence to show that increased Cx32 GJ
function is a major mechanism of renal H/R damage, and
inhibition of Cx32 GJ function could protect against renal
H/R damage. It should be noted that although 2-APB had
been used widely as a selective Cx32 inhibitor, which could
decrease function of Cx32 in vivo and in vitro,20 it may have
other nonspecific effects such as decreasing Ca2+ release or
inhibiting IP3 receptors38 when used at high doses. Our
Fig. 6. Propofol (pro) attenuated NRK-52E cell damage of hypoxia for 24h and reoxygenation for 4h (H24R4) through gap junction composed of connexin32 (Cx32). (AC) Effects of propofol (pro) on gap junction function of NRK-52E cells and Cx32 expression. (D) Propofol (pro, 15 M) decreased gap junction function on Hela cells just only expressing Cx32. Data were obtained from
four independent experiments each performed in quintuplicate, n = 4. *P < 0.05 versus control.
incorporation of the use of two different siRNAs specifically

to knock-down Cx32 expression confirmed the role of Cx32
in mediating posthypoxic cellular injury. Another important
novel finding was that propofol, as a common anesthetic
used in clinic, conferred protective effects against H/R damage of renal cells through inhibiting Cx32 channels and that
inhibition of Cx32 is a major mechanism whereby propofol
reduces ROS and oxidative stress under pathological conditions like AOLT-induced AKI.
Our joint in vitro and in vivo studies showed that GJ
composed of Cx32 plays a key role in kidney damage after
AOLT. GJ was always considered to be the direct cell-to-cell
transfer of electrical charge or small molecules. Intercellular
molecular signals transferred between neighboring cells contribute to cell growth, differentiation, normal physiology,
and response to trauma in all different organs.39 Molecular
signals enhancing cytotoxicity or inducing apoptosis were
called death signals, which were mainly manipulated by
GJ, especially in cancer cells.5,40
Propagation of death signals through GJ had been
widely explored. However, the intrinsic quality of them
had not been identified. These signals were usually considered to be calcium, cells metabolites or molecules triggering
activities of cellular death pathways.41 We investigated the
possibility of ROS as death signals, which was indispensable for I/R damage.42 ROS, including oxygen radicals and
nonradical compounds, exist at low levels in cells but play
an important role in regulating physiologic tubular functions and renal microcirculation. However, great amount
of ROS produced always disrupt the redox signaling pathway, resulting in cellular molecular structures damage. Both
oxygen radicals and nonradical compounds had much less
molecular mass than the upper limit of GJ permeable, which
mean that both of them might be transferred through GJ,
acting as death signals function. We showed that ROS
induced by H/R damage was manipulated by GJ composed
of Cx32. ROS generation under pathological conditions not
only damaged cells in which they were produced, but also
attacked neighboring cells through the transfer mediated by
GJ. In this transfer process, ROS level was raised in cells
connected with GJ, resulting in damage magnification. This
kind of mechanism was called bystander effect,43 which
was the reason why GJ could regulate kidney damage. We
firstly demonstrated that kidney damage could be protected
by inhibiting ROS transfer mediated by Cx32 channels. This
point of view has never been reported before.
Proper anesthetics choice used in perioperative or preoperative period benefitted patients who received organ
transplantation, because they contribute to organ protection
through different signal pathways.44 Propofol, as a commonly used anesthetic in clinic, had already been explored
for many years and considered to confer organ protection
through regulating different mechanisms, such as antioxidant, antiinflammatory, inhibiting apoptosis, and calcium
influx, but the in depth mechanism governing the above
mechanistic effects is still largely unclear.45
According to the reports, protective effects of propofol
were always considered to be relative with decreasing ROS
Fig. 7. Effects of propofol (pro) on reactive oxygen species (ROS) induced by hypoxia for 24h and reoxygenation for 4h (H24R4)
damage mediated by gap junction composed of connexin32 (Cx32). (A) Effects of propofol (pro, 15 M) on cell growth, relative
lactate dehydrogenase (LDH) release, and 15-F2t-Isop level of NRK-52E cells treated with H24R4 at different cell density culture.
(B) Effects of inhibitor 2-aminoethoxydiphenyl borate (2-APB, 25 M), enhancer retinoic acid (RA, 10 M), and propofol (pro, 15
M) on intracellular ROS at different cell density culture. (C) Dichlorofluorescin assay was used to test intracellular ROS when
cells were treated with different chemicals. Images were obtained with a fluorescence microscope, 200. (D) Effects of propofol
(pro, 15 M) cooperating with inhibitor 2-APB (25 M) or enhancer RA (10 M) on NRK-52E cell growth when exposed to H24R4.
(E) Effects of propofol (pro, 15 M) cooperating with inhibitor 2-APB (25 M) or enhancer RA (10 M) on relative LDH release of
NRK-52E cells when exposed to H24R4. (A and B) *P < 0.05 versus control; #P < 0.05 versus H24R4 group in low-density or
high-density cell culture; P < 0.05. (D and E) *P < 0.05 versus control; #P < 0.05 versus pro group. Data were obtained from
five independent experiments each performed in quintuplicate, n = 5. From A to B and D to E, bars of control in low- or highdensity cell culture did not display in this figure, and bars of control were considered to be 1. Vehicle controls of 2-APB and RA
are dimethyl sulfoxide, and Vehicle control of propofol is lipid. Neither dimethyl sulfoxide nor lipid has significant effects on the
above-mentioned parameters (see fig. 2, Supplemental Digital Content 1, http://links.lww.com/ALN/B94).
level. However, we found that at low-density cell culture

(no contact with each other and no GJ formed), ROS level
was not affected by propofol, but propofol decreased ROS
significantly at high-density cell culture (contact with each
other and GJ formed). The fact that propofol regulating
ROS through GJ may play a more important role against
H/R damage where GJ function may be enhanced, which
is totally distinct from classic signal pathways. We suppose
that propofol not only affected ROS distribution through
GJ directly, but also inhibited some signal transmission
resulting in enhanced ROS production. As a commonly

used anesthetic, propofol is not only just limited in operation room, but also is used postoperatively in intensive care
unit,46 especially in patients with severe conditions like those
after organ transplantation. As reported, propofol preconditioning with various time intervals had protective effects
against organ I/R injuries. Sufferers, who will undergo liver
transplantation, are always with anxiety and insomnia. Propofol has special curative effect for this kind of patients.46
Based on these reports, we pretreated rats with propofol
Fig. 8. Propofol (pro)-attenuated gap junction function of kidney tissue and kidney damage following autologous orthotopic liver
transplantation (AOLT). (A) Scrape-and-load assay was used to valuate functional gap junction in kidney tissue. Rats were
treated with propofol (pro, 50mg/kg) or corresponding solvent for 3 days intraperitoneally before Scrape-and-load assay.
(B) Effects of propofol (pro, 50mg/kg per day for 3 days, intraperitoneally) on expression of connexin32 (Cx32) after AOLT.
(C) Effects of propofol (pro, 50mg/kg, 3 days, intraperitoneally) on remote kidney damage of rats induced by AOLT (hematoxylin
eosin staining; original magnification 200). (D) Propofol (pro, 50mg/kg per day, for 3 days, intraperitoneally) attenuated serum
creatinine (Cr) and blood urea nitrogen (BUN) increase of rats induced by AOLT. (E) Propofol (pro, 50mg/kg day, for 3 days,
intraperitoneally) attenuated H2O2 and 15-F2t-Isop levels of rats induced by AOLT. Every group contains eight samples, n = 8.
*P < 0.05 versus sham; #P < 0.05 versus AOLT group. Vehicle control of propofol is lipid, which has no effects on the abovementioned parameters (see fig. 1, Supplemental Digital Content 1, http://links.lww.com/ALN/B94).
at its sedative doses for 3 days to observe its effects on GJ

composed of Cx32.47 Our in vivo study results showed that
pretreatment with propofol attenuated AOLT-induced AKI,
which provided a new strategy for organ protection.
Another protective effect of propofol was relative with
calcium increase. EC50 of propofol was 15 M (measured
as total concentration: protein bound and free).16,48 At this
clinically relevant anesthesia concentration, propofol could

increase calcium level in cells, which always affected cell
survival.49 But, our findings suggest that changes of calcium
were not the direct or major way for propofol to regulate
cell growth. Because, no effects of propofol were observed
at low-density cell culture (no GJ formed). A large number of reports documented that calcium contributed to the
closure of GJ, even acting on the GJ gating directly.50 The

results demonstrated that potential calcium increase induced
by propofol, if any, may have affected cell survival in a GJdependent manner, but not directly.
Liver transplantation was a serious attack for all the systems. During the operation, a large amount of harmful factors, such as proinflammatory cytokines and ROS and so on
were released into the blood induced by I/R, which initiated remote organs injury, leading to a high mortality rate
after the operation.51 In our studies, we investigated remote
kidney damage after AOLT and clarified that AKI could be
attenuated by inhibiting GJ function composed of Cx32,
which was relative with ROS reduction. Propofol reduced
postischemic or posthypoxic ROS production and oxidative
damage mainly via GJ-dependent mechanism and protected
against AOLT induced kidney damage. These findings might
not be universal to all the remote organs, but nonetheless
they may have translational applications in organ protection
of liver transplantation.
In conclusion, we have conducted a series of in vitro and
in vivo studies and demonstrated that Cx32 plays a critical role in AOLT-induced AKI and that inhibition of Cx32
GJ function may represent a new and major mechanism
whereby propofol reduces oxidative stress and subsequently
attenuates post-AOLT AKI.
Acknowledgments
This study is supported by the National Natural Science
Foundation of China (Beijing, China; grant nos. 81170449
and 81401628); key project of Natural Science Foundation
of Guangdong Province, China (Guangzhou, Guangdong
Province, China; grant no. S2011020002780); and 985 project (Beijing, China; grant no. 82000-1188190).
Competing Interests
Correspondence
Address correspondence to Dr. Hei (for access to raw data
and statistical analysis): Department of Anesthesiology, The
Third Affiliated Hospital of Sun Yat-sen University, Guangzhou, Peoples Republic of China. heiziqing@sina.com; and
Dr. Xia (regarding study design): Department of Anesthesiology, Li Ka Shing Faculty of Medicine, The University of
Hong Kong, Hong Kong, Peoples Republic of China. zyxia@
hku.hk. Information on purchasing reprints may be found at
www.anesthesiology.org or on the masthead page at the beginning of this issue. Anesthesiologys articles are made freely
accessible to all readers, for personal use only, 6 months
from the cover date of the issue.
References
1. Billings FT IV, Ball SK, Roberts LJ II, Pretorius M: Postoperative
acute kidney injury is associated with hemoglobinemia and
an enhanced oxidative stress response. Free Radic Biol Med
2011; 50:14807
2. Nadim MK, Genyk YS, Tokin C, Fieber J, Ananthapanyasut W,
Ye W, Selby R: Impact of the etiology of acute kidney injury on
outcomes following liver transplantation: Acute tubular necrosis versus hepatorenal syndrome. Liver Transpl 2012; 18:53948
3. Aronson S, Phillips-Bute B, Stafford-Smith M, Fontes M, Gaca
J, Mathew JP, Newman MF: The association of postcardiac
surgery acute kidney injury with intraoperative systolic
blood pressure hypotension. Anesthesiol Res Pract 2013;
2013:174091
4. Gang GT, Hwang JH, Kim YH, Noh JR, Kim KS, Jeong JY,
Choi DE, Lee KW, Jung JY, Shong M, Lee CH: Protection of
NAD(P)H: Quinone oxidoreductase 1 against renal ischemia/reperfusion injury in mice. Free Radic Biol Med 2014;
67:13949
5. Hong X, Wang Q, Yang Y, Zheng S, Tong X, Zhang S, Tao
L, Harris AL: Gap junctions propagate opposite effects in
normal and tumor testicular cells in response to cisplatin.
Cancer Lett 2012; 317:16571
6. Yano T, Ito F, Yamasaki H, Hagiwara K, Ozasa H, Nakazawa
H, Toma H: Epigenetic inactivation of connexin 32 in renal
cell carcinoma from hemodialytic patients. Kidney Int 2004;
65:1519
7. Anand RJ, Hackam DJ: The role of gap junctions in health
and disease. Crit Care Med 2005; 33(12 suppl):S5358
8. Asamoto M, Hokaiwado N, Murasaki T, Shirai T: Connexin
32 dominant-negative mutant transgenic rats are resistant to
hepatic damage by chemicals. Hepatology 2004; 40:20510
9. Zhang L, Li YM, Jing YH, Wang SY, Song YF, Yin J: Protective
effects of carbenoxolone are associated with attenuation of
oxidative stress in ischemic brain injury. Neurosci Bull 2013;
29:31120
10. Rash JE, Yasumura T, Davidson KG, Furman CS, Dudek FE,
Nagy JI: Identification of cells expressing Cx43, Cx30, Cx26,
Cx32 and Cx36 in gap junctions of rat brain and spinal cord.
Cell Commun Adhes 2001; 8:31520
11. Fujimoto E, Yano T, Ueno K: [Connexin32 as a tumor suppressor gene in renal cell carcinoma]. Nihon Yakurigaku
Zasshi 2007; 129:1059
12. Xia Z, Huang Z, Ansley DM: Large-dose propofol during
cardiopulmonary bypass decreases biochemical markers of
myocardial injury in coronary surgery patients: A comparison with isoflurane. Anesth Analg 2006; 103:52732
13. Xia Z, Godin DV, Chang TK, Ansley DM: Dose-dependent
protection of cardiac function by propofol during ischemia
and early reperfusion in rats: Effects on 15-F2t-isoprostane
formation. Can J Physiol Pharmacol 2003; 81:1421
14. Baki ED, Aldemir M, Kokulu S, Koca HB, Ela Y, Svac RG,
ztrk NK, Emmiler M, Adal F, Uzel H: Comparison of the
effects of desflurane and propofol anesthesia on the inflammatory response and s100 protein during coronary artery
bypass grafting. Inflammation 2013; 36:132733
15. Xia Z, Luo T, Liu HM, Wang F, Xia ZY, Irwin MG, Vanhoutte
PM: L-arginine enhances nitrative stress and exacerbates
tumor necrosis factor-alpha toxicity to human endothelial cells in culture: Prevention by propofol. J Cardiovasc
Pharmacol 2010; 55:35867
16. Zhao Y, Liu B, Wang Q, Yuan D, Yang Y, Hong X, Wang X,
Tao L: Propofol depresses the cytotoxicity of X-ray irradiation through inhibition of gap junctions. Anesth Analg 2011;
112:108895
17. Chi X, Zhang A, Luo G, Xia H, Zhu G, Hei Z, Liu X, Wei
J, Xia Z: Knockdown of myeloid differentiation protein-2
reduces acute lung injury following orthotopic autologous
liver transplantation in a rat model. Pulm Pharmacol Ther
2013; 26:3807
18. Yao W, Luo G, Zhu G, Chi X, Zhang A, Xia Z, Hei Z: Propofol
activation of the Nrf2 pathway is associated with amelioration of acute lung injury in a rat liver transplantation model.
Oxid Med Cell Longev 2014; 2014:258567
19. Inagawa G, Sato K, Kikuchi T, Nishihama M, Shioda M,
Koyama Y, Yamada Y, Andoh T: Chronic ethanol consumption
does not affect action of propofol on rat hippocampal acetylcholine release in vivo. Br J Anaesth 2004; 93:7379
20. Patel SJ, Milwid JM, King KR, Bohr S, Iracheta-Velle A, Li M,
Vitalo A, Parekkadan B, Jindal R, Yarmush ML: Gap junction
inhibition prevents drug-induced liver toxicity and fulminant
hepatic failure. Nat Biotechnol 2012; 30:17983
21. Liu KX, Rinne T, He W, Wang F, Xia Z: Propofol attenuates
intestinal mucosa injury induced by intestinal ischemiareperfusion in the rat. Can J Anaesth 2007; 54:36674
22. Liu KX, Chen SQ, Huang WQ, Li YS, Irwin MG, Xia Z: Propofol
pretreatment reduces ceramide production and attenuates
intestinal mucosal apoptosis induced by intestinal ischemia/
reperfusion in rats. Anesth Analg 2008; 107:188491
23. Liu L, Tian K, Zhu Y, Ding X, Li T: opioid receptor antagonist, ICI 174,864, is suitable for the early treatment of uncontrolled hemorrhagic shock in rats. Anesthesiology 2013;
119:37988
24. Bai X, Yan Y, Canfield S, Muravyeva MY, Kikuchi C, Zaja I,
Corbett JA, Bosnjak ZJ: Ketamine enhances human neural
stem cell proliferation and induces neuronal apoptosis via
reactive oxygen species-mediated mitochondrial pathway.
Anesth Analg 2013; 116:86980
25. Luo T, Xia Z: A small dose of hydrogen peroxide enhances
tumor necrosis factor-alpha toxicity in inducing human vascular endothelial cell apoptosis: Reversal with propofol.
26. He B, Tong X, Wang L, Wang Q, Ye H, Liu B, Hong X, Tao L,
Harris AL: Tramadol and flurbiprofen depress the cytotoxicity of cisplatin via their effects on gap junctions. Clin Cancer
Res 2009; 15:580310
27. Liu WF, Nelson CM, Pirone DM, Chen CS: E-cadherin engagement stimulates proliferation via Rac1. J Cell Biol 2006;
173:43141
28. De Hert SG, ten Broecke PW, Mertens E, Van Sommeren EW,
De Blier IG, Stockman BA, Rodrigus IE: Sevoflurane but not
propofol preserves myocardial function in coronary surgery
patients. Anesthesiology 2002; 97:429
29. De Hert SG, Cromheecke S, ten Broecke PW, Mertens E,
De Blier IG, Stockman BA, Rodrigus IE, Van der Linden PJ:
Effects of propofol, desflurane, and sevoflurane on recovery of myocardial function after coronary surgery in elderly
high-risk patients. Anesthesiology 2003; 99:31423
30. Wu T, Gan X, Zhou S, Ge M, Zhang Z, Hei Z: Histamine at
low concentrations aggravates rat liver BRL-3A cell injury
induced by hypoxia/reoxygenation through histamine H2
receptor in vitro. Toxicol In Vitro 2013; 27:37886
31. Jin C, Zhang PJ, Wu XM, Zhou B, Li Y, Liu XY, Feng M, Tao
LD: Impact of hypoxic preconditioning on apoptosis and its
possible mechanism in orthotopic liver autotransplantation
in rats. Hepatobiliary Pancreat Dis Int 2009; 8:405
32. Zhao HF, Zhang GW, Zhou J, Lin JH, Cui ZL, Li XH: Biliary
tract injury caused by different relative warm ischemia time
in liver transplantation in rats. Hepatobiliary Pancreat Dis Int
2009; 8:24754
33. Yang Y, Cao MH, Wang Q, Yuan DD, Li L, Tao L: The effects of
2-aminoethoxydiphenyl borate and diphenylboronic anhydride on gap junctions composed of Connexin43 in TM sertoli cells. Biol Pharm Bull 2011; 34:13907
34. Wang Q, You T, Yuan D, Han X, Hong X, He B, Wang L,
Tong X, Tao L, Harris AL: Cisplatin and oxaliplatin inhibit gap
junctional communication by direct action and by reduction
of connexin expression, thereby counteracting cytotoxic efficacy. J Pharmacol Exp Ther 2010; 333:90311
35. Carraretto AR, Vianna Filho PT, Castiglia YM, Golim Mde A,
Souza AV, Carvalho LR, Deffune E, Vianna PT: Do propofol
and isoflurane protect the kidney against ischemia/reperfusion injury during transient hyperglycemia? Acta Cir Bras
2013; 28:1616
36. Hsing CH, Chou W, Wang JJ, Chen HW, Yeh CH: Propofol
increases bone morphogenetic protein-7 and decreases oxidative stress in sepsis-induced acute kidney injury. Nephrol
Dial Transplant 2011; 26:116272
37. Randhawa PS, Shapiro R: Chronic renal failure after liver
transplantation. Am J Transplant 2005; 5:9678
38. Maruyama T, Kanaji T, Nakade S, Kanno T, Mikoshiba K:
2APB, 2-aminoethoxydiphenyl borate, a membrane-penetrable modulator of Ins(1,4,5)P3-induced Ca2+ release. J
Biochem 1997; 122:498505
39. Yeager M, Harris AL: Gap junction channel structure in the
early 21st century: Facts and fantasies. Curr Opin Cell Biol
2007; 19:5218
40. Krutovskikh VA, Piccoli C, Yamasaki H, Yamasaki H: Gap
junction intercellular communication propagates cell death
in cancerous cells. Oncogene 2002; 21:198999
41. Lahne M, Gale JE: Damage-induced activation of ERK1/2
in cochlear supporting cells is a hair cell death-promoting
signal that depends on extracellular ATP and calcium. J
Neurosci 2008; 28:491828
42. Gao S, Oh YB, Park BM, Park WH, Kim SH: Urotensin II protects ischemic reperfusion injury of hearts through ROS and
antioxidant pathway. Peptides 2012; 36:199205
43. Garcia-Rodrguez L, Prez-Torras S, Carri M, Cascante A,
Garca-Ribas I, Mazo A, Fillat C: Connexin-26 is a key factor mediating gemcitabine bystander effect. Mol Cancer Ther
2011; 10:50517
44. Head BP, Patel P: Anesthetics and brain protection. Curr
Opin Anaesthesiol 2007; 20:3959
45. Ito S, Sugiyama H, Kitahara S, Ikemoto Y, Yokoyama T:
Effects of propofol and pentobarbital on calcium concentration in presynaptic boutons on a rat hippocampal neuron. J
Anesth 2011; 25:72733
46. Xu Z, Jiang X, Li W, Gao D, Li X, Liu J: Propofol-induced
sleep: Efficacy and safety in patients with refractory chronic
primary insomnia. Cell Biochem Biophys 2011; 60:1616
47. Assad AR, Delou JM, Fonseca LM, Villela NR, Nascimento JH,
Verosa N, Lopes AG, Capella MA: The role of KATP channels on propofol preconditioning in a cellular model of renal
ischemia-reperfusion. Anesth Analg 2009; 109:148692
48. Milne SE, Troy A, Irwin MG, Kenny GN: Relationship
between bispectral index, auditory evoked potential index
and effect-site EC50 for propofol at two clinical end-points.
Br J Anaesth 2003; 90:12731
49. Ionta M, Ferreira RA, Pfister SC, Machado-Santelli GM:
Exogenous Cx43 expression decrease cell proliferation rate
in rat hepatocarcinoma cells independently of functional gap
junction. Cancer Cell Int 2009; 9:22
50. Bennett MV, Barrio LC, Bargiello TA, Spray DC, Hertzberg E,
Sez JC: Gap junctions: New tools, new answers, new questions. Neuron 1991; 6:30520
51. Ohkohchi N: Mechanisms of preservation and ischemic/
reperfusion injury in liver transplantation. Transplant Proc
2002; 34:26703
Long-term Effects of Single or Multiple Neonatal

Sevoflurane Exposures on Rat Hippocampal
Ultrastructure
Levana G. Amrock, B.S., Mathew L. Starner, B.A., Kathy L. Murphy, B.Vet.Med., Cert.V.A.,
Mark G. Baxter, Ph.D.
ABSTRACT
Background: Neonatal exposure to general anesthetics may pose significant neurocognitive risk. Human epidemiological
studies demonstrate higher rates of learning disability among children with multiple, but not single, exposures to anesthesia.
The authors employ a rat model to provide a histological correlate for these population-based observations. The authors examined long-term differences in hippocampal synaptic density, mitochondrial density, and dendritic spine morphology.
Methods: Twenty male rat pups (n = 5/condition) were exposed to 2.5% sevoflurane under one of four conditions: single 2-h
exposure on postnatal day 7 (P7); single 6-h exposure on P7; repeated 2-h exposures on P7, P10, and P13 for a cumulative 6h
of general anesthetics; or control exposure to 30% oxygen on P7, P10, and P13.
Results: Repeated exposure to general anesthetics resulted in greater synaptic loss relative to a single 2-h exposure (P < 0.001).
The magnitude of synaptic loss induced by three 2-h exposures (1.9770.040 m3 [mean SEM]) was more profound than
that of a single 6-h exposure (2.2800.045 m3, P = 0.022). Repeated exposures did not alter the distribution of postsynaptic
density length, indicating a uniform pattern of loss across spine types. In contrast, mitochondrial toxicity was best predicted
by the cumulative duration of exposure. Relative to control (0.5950.017), both repeated 2-h exposures (0.4790.015)
and a single 6-h exposure (0.4880.013) were associated with equivalent reductions in the fraction of presynaptic terminals
containing mitochondria (P < 0.001).
Conclusion: This suggests a threshold effect for general anestheticinduced neurotoxicity, whereby even brief exposures
induce long-lasting alterations in neuronal circuitry and sensitize surviving synapses to subsequent loss. (Anesthesiology
2015; 122:87-95)
VER the past 30 yr advances in medical and surgical

techniques have proven lifesaving for young children.
As a consequence, early and repeated exposure to general anesthetics (GA) has increased. Yet recent evidence suggests that
such practices pose significant neurocognitive risk. Human
epidemiological studies demonstrate that childhood exposure
to GA is associated with an increased risk for the later development of cognitive and behavioral impairment.1,2 This risk
is especially high among children with multiple anesthetic
exposures.35 However, the clinical implications and mechanism underlying these findings remain unclear. These studies
are retrospective in design and therefore limited in their ability to distinguish between the effects of anesthetic-induced
neurotoxicity and confounders such as coexistent medical
disease, surgical stress, and hospitalization.
In vivo animal studies demonstrate that GA is immediately
toxic to the developing brain, and even a single early exposure induces long-term cognitive and behavioral deficits.68
These deficits have been linked to disturbances in neuronal
circuitry, mitochondrial morphology, and dendritic spine

Human epidemiological studies suggest that children exposed
to multipleanesthetics, as opposed to a single anesthetic, might
be at greater risk for development of cognitive dysfunction
The underlying mechanism by which injury is increased with
multiple exposures is not known
The effect of a single exposure or multiple exposures with
equivalent total duration of exposure on brain ultrastructure
was evaluated by electron microscopy in rodents

Repeated exposure to sevoflurane led to a greater loss of synapses in comparison to a single exposure
Anesthetic exposure led to a reduction in the number of synaptic terminals with mitochondria
Interestingly, this reduction was correlated to total anesthetic
exposure rather than frequency of exposure
These data suggest that a brief anesthetic exposure might
sensitize the brain to subsequent anesthetic induced injury
development.912 Our previous work demonstrates that relative to a single neonatal exposure, rats repeatedly exposed to
This work has been submitted to the International Anesthesia Research Society annual meeting, May 17, 2014, Montreal, Quebec, Canada;
American Society of Anesthesiologists annual meeting, October 11, 2014, New Orleans, Louisiana; and to the Society for Neuroscience annual
meeting, November 15, 2014, Washington, D.C.
Submitted for publication June 2, 2014. Accepted for publication September 4, 2014. From the Department of Anesthesiology (L.G.A.), Department of Neuroscience (M.L.S.), Departments of Neuroscience, Anesthesiology, and Geriatrics and Palliative Medicine (M.G.B.), the Icahn School of
Medicine at Mount Sinai, New York, New York; and the Department of Biomedical Services, University of Oxford, Oxford, United Kingdom (K.L.M.).
January 2015
Effect of Multiple Neonatal Anesthetic Exposures
isoflurane had greater deficits in spatial working memory.13

We therefore hypothesize that, as compared to a single exposure, brief but repeated neonatal exposures would produce
greater neurohistologic damage.
Characterizing the dose-related effects of neonatal anesthetic exposure is critical for the development of safe clinical
practices and informed preoperative counseling. Here, we
use a rat model to investigate the long-term ultrastructural
effects of single or multiple neonatal exposures to GA. Specifically, we examine differences in synaptic density, mitochondrial density, and dendritic spine head morphology.
Our study demonstrates that relative to a single exposure,
repeated neonatal exposure to GA is highly neurotoxic.
Although alterations in presynaptic mitochondrial localization appear to be dependent on the cumulative duration of
exposure, synaptic losses are best predicted by the total number of anesthetic exposures.

Animals
All experiments were carried out in strict accordance with
the recommendations of the Guide for the Care and of Use
of Laboratory Animals of the National Institutes of Health
(Bethesda, MD). The protocol (LA10-00071) was approved
by the Icahn School of Medicine Animal Use and Care
Committee (New York, New York).
Male Long-Evans rat pups from six natural litters were
randomly divided into four balanced experimental groups.
Rats were then exposed to 2.5% sevoflurane in 30% oxygen
under one of four conditions: (12h) a single 2-h exposure on postnatal day 7 (P7), followed by control exposure
to 30% oxygen on P10 and P13; (16h) a single 6-h anesthetic exposure on P7, followed by control exposure to 30%
oxygen on P10 and P13; (32h) repeated 2-h sevoflurane
exposures on P7, P10, and P13 for a cumulative 6h of GA;
or a control exposure to 30% oxygen on P7, P10, and P13.
At P91, rats underwent behavioral testing using a spatial
memory task, followed by euthanasia on P105-112. Behavioral results are not shown, as testing was interrupted by an
unforeseen circumstance and as a result the rats displayed
atypical learning curves (an increase in errors associated with
the interruption in testing). Given this disrupted testing, we
cannot draw any conclusions from the behavioral data.
Anesthesia
Sevoflurane was administered in a monitored anesthesia
chamber via a calibrated flowmeter and agent-specific vaporizer. Chamber concentrations of carbon dioxide, oxygen,
and sevoflurane were monitored throughout (VitalStore,
Vetronic Services Ltd., Abbotskerswell, United Kingdom).
Anesthesia was induced with 6% sevoflurane in 30% oxygen
until loss of the righting reflex and response to toe and tail
pinch, at which point exposure time commenced. Anesthesia
was maintained by 2.5% sevoflurane in 30% oxygen. Rectal
temperature was continuously monitored, and rats were kept

normothermic at 36.50.5C throughout the experiment
(PowerLab ADInstruments Ltd., Oxford, United Kingdom). Pulse oximetry was continuously monitored (VitalStore, Vetronic Services Ltd.), and oxygen saturation was
maintained about 95% in all but two pups (see mortality
data below). The respiratory rate and skin color were visually
inspected every 20min. Pups in the 6-h sevoflurane condition were removed from the chamber every 2h and massaged
to provide stimulation. Separate pups were used for transcardial arterial blood gas analysis (Radiometer ABL80, Cleveland, OH), which determined that all conditions produced
mild hypercapnia. After anesthesia, the chamber was flushed
with 30% oxygen until the pups recovered and displayed
righting reflexes. The pups were then rubbed with bedding
material from their home cage, and returned to their natural
litter where they remained until weaning at P21. Mortality
within group 16h was 20% (2 of 10 pups). There was no
mortality across the remaining experimental groups.
Tissue Processing and Perfusion
Twenty male rats (n = 5/condition) were euthanized at ~3.5
months of age (P105-112). Rats were deeply anesthetized
with sodium pentobarbital 100mg/kg and transcardially
perfused with 1% paraformaldehyde in 0.1M phosphatebuffered saline (PBS, pH 7.4) for 1min, followed by 4%
paraformaldehyde in a 0.1M PBS for 11min. The brains
were removed, postfixed at 4C in 4% paraformaldehyde
and 0.125% gluteraldehyde in PBS overnight, and coronally
sectioned on a Vibratome (Leica VT1000S, Bannockburn,
IL). Three blocks containing hippocampal CA1 were then
manually selected from each animal and placed a 0.1M
PBS solution. The blocks then underwent cryo-protection
in increasing concentrations of glycerol (from 10 to 30%
in PBS), followed by a freeze-plunge.14 In brief, blocks
were rapidly immersed in liquid propane cooled to 190C
in a Universal Cryofixation System KF80 (Reichert-Jung,
Vienna, Austria). The blocks were then immersed in 1.5%
uranyl acetate and anhydrous methanol for 24h at 90C
in an Automatic Freeze-Substitution System (Lecia, Vienna,
Austria). The block temperature was then raised by 4C/h to a
temperature of 45C at which point the blocks were washed
with anhydrous methanol and infiltrated with Lowicryl resin
(Electron Microscopy Sciences, Hatfield, PA). The resin was
polymerized under ultraviolet light (360nm) first for 48h at
45C, followed by 24h at 0C. The block faces were then
trimmed to the stratum pyramidale and radiatum of CA1.
Fifteen or more consecutive 90nm ultrathin sections were
cut with a diamond knife (Diatome, Bienne, Switzerland),
and mounted on formvar/carbon-coated nickel slot grids
(Electron Microscopy Sciences).
Morphologic Analyses
All imaging, synaptic, and mitochondrial analyses were performed by an experimenter blind to condition. Samples were
Anesthesiology 2015; 122:87-95 88 Amrock et al.
randomly recoded, and the code remained hidden until all

morphologic analyses were complete. Serial section micrographs were taken with a Hitachi H-7000 transmission electron microscope (Hitachi High Technologies America, Inc.,
Pleasanton, CA) using a systematic-random approach. Sample
size was guided by previous publications using the disector
method for synaptic and mitochondrial analyses.1517 Five sets
of 15 serial images were captured at 17,000 using an AMT
advantage CCD camera (Advanced Microscopy Techniques,
Danvers, MA). To ensure the correct subregion was captured,
all imaging was performed 95 to 105 m from the pyramidal
cell layer. Images were imported into Adobe Photoshop CS5
version 12.0.4 (Adobe Systems Inc., San Jose, CA).
Synaptic Analysis
Previous studies including rats from a similar study population have demonstrated that neonatal exposure to anesthesia
induces marked hippocampal neurodegeneration and reductions in subicular synaptic volumetric density that persist
into adulthood.6,11,18 On this basis, synaptic density was
selected as a measure of interest. Disector analysis was used to
obtain a stereologically unbiased population of synapses for
quantitative analysis. This methodology has been previously
described.15,16,1921 In brief, all axospinous synapses were
identified in the first two layers of each 15-section serial set.
The postsynaptic density (PSD) was selected as the counting
unit. PSD was defined by the presence of synaptic vesicles in
the axon terminal and a distinct density in the postsynaptic
dendritic spine.16 Synapses were counted only if they were
contained in the reference layer (section 1), but not in the
corresponding look-up layer (section 2). This procedure
was repeated for the middle two (sections 8 and 9) and last
two (sections 14 and 15) images of each serial set. To increase
sampling efficiency, the reference and look-up layers were
reversed and the images were reanalyzed (fig.1, A and B).
The disector area was 49 m2 and the height of the disector was 180nm. Axospinous synapse density was calculated
as the total number of counted synapses from both images,
divided by the total volume of the disector.
Postsynaptic Density Analysis
Postsynaptic density length was used as a surrogate measure
for dendritic spine head size.22 For each rat, 15 sets of three
serial sections were analyzed, and all axospinous synapses in
the second, middle section, were identified. Each synapse
was followed through the three-image series, and the longest
PSD was measured (fig.1, CE). For perforated synapses,
the total length was reported as the sum of each PSD segment.16 On average 200 PSDs were measured per rat.
Presynaptic Mitochondrial Analysis
For each 15-section series, the first five serial images were
analyzed using the disector analysis described above. All
presynaptic mitochondria were labeled and identified by
the following criteria as: (1) electron dense and (2) enclosed
within a bouton containing greater than 3 synaptic vesicles.
Mitochondrial density (m3) was calculated as the total

number of mitochondria present in the first reference layer
and not the second look-up layer, divided by the total volume of the disector.
To control for a treatment effect on the raw number of
boutons present, we analyzed the percentage of presynaptic terminals containing mitochondria. For each series, the
eighth or middle section was used as a reference, and all
boutons containing greater than 3 synaptic vesicles were
identified and numbered. Each bouton was then followed
throughout the series and the number of mitochondria
recorded (fig.2).17 On average, 250 boutons were analyzed
per rat. Boutons extending beyond 15 sections were included
in the analyses so as not to preferentially exclude large boutons. The fraction of boutons containing mitochondria was
calculated using Microsoft Excel (Microsoft Corporation,
Redmond, WA).
Linear fixed-effects analyses (e.g., one-way ANOVA) cannot
account for the dependence between multiple observations
from the same animal. We therefore adopted a nested, multilevel approach.23 Multilevel linear regression models were
constructed to determine the main effect of treatment group
on synaptic density, mitochondrial density, and the fraction
of presynaptic terminals containing mitochondria. Random
intercepts accounted for baseline variation and clustering
between dams, animals, and series. To better evaluate for
differences between experimental groups, post hoc pairwise
comparisons with Bonferroni correction were performed.
Additionally, the correlation between mean synaptic density
and mean presynaptic mitochondrial density was evaluated
using Pearsons correlation coefficient. To confirm that the
observed relationship was not due to common group effects,
data was centered on the grand mean and reanalyzed. Statistical significance of the above multilevel analyses and post
hoc comparisons was defined as a P value of less than 0.05.
Finally, differences in the cumulative distribution of PSD
length were assessed in a pairwise manner via a two-sample
KolmogorovSmirnov test. To reduce family-wise error, statistical significance was defined as a P value less than 0.0083
(or 0.05/6). All analyses were preformed using Stata version
11.2 (StataCorp, College Station, TX).
Results
Effects of Anesthesia on Synaptic Density
Neonatal exposure to anesthesia was associated with longlasting reductions in mean CA1 synaptic density (fig.3).
Multilevel linear regression models controlling for baseline
variation at the dam, animal, and series level determined that
synaptic density differed based on the duration and number
of sevoflurane exposures (table1). As compared to control, a
single 2-h exposure on P7 was not associated with alterations
in synaptic density ( = 0.078, P = 0.605), while a single
Fig. 1. (A and B) Electron micrographs demonstrating the disector method for synaptic density. A is considered the reference
layer, B is considered the look-up. Axospinous synapses present only in the reference layer, and not in the look-up are shaded
in blue. Synapses present in both layers are shaded in green. Scale bar, 0.5 m. (CE) Electron micrographs demonstrating
postsynaptic density (PSD) length analysis. All axospinous synapses in the middle layer (D) were identified. For each synapse,
the longest PSD in any of the three serial sections was identified and measured. Scale bar, 0.5 m.
Fig. 2. Electron micrographs demonstrating the analysis of the fraction of boutons containing mitochondria. For each 15-layer
series, the 8th or middle section (*) was used as the reference and all boutons identified. Each bouton (green) was then followed
throughout the series, and the presence of mitochondria (purple) recorded. Scale bar, 500 m.
6-h exposure yielded substantial synaptic losses ( = 0.336,

P = 0.022). Furthermore, repeated 2-h exposures on P7, 10,
and 13 were associated with a marked decrease in synaptic
density ( = 0.681, P < 0.001).
In order to further evaluate for dose-related differences,
post hoc pairwise tests using the Bonferroni correction were
performed (table2). As compared to a single 2-h exposure,
both a single 6-h exposure on P7 and multiple 2-h exposures
on P7, 10 and 13 resulted in significant decreases in mean
synaptic density (P = 0.042 and P <0.001, respectively).
Notably, multiple 2-h anesthetic exposures (of cumulative
6-h duration) resulted in greater synaptic loss as compared
to a single 6-h exposure (P = 0.022).
Effects of Anesthesia on Presynaptic Density Length
Early exposure to sevoflurane was associated with a shift
in the cumulative distribution of PSD length, a surrogate
measure of dendritic spine size and morphology (fig.4).22
Pairwise two-sample KolmogorovSmirnov tests demonstrated that a single 2-h exposure resulted in smaller PSD
lengths relative to a single 6-h exposure and multiple 2-h
exposures (D = 0.086, P = 0.003 and D = 0.085, P = 0.006,
respectively). However, no differences in PSD length were
observed between a cumulative 6-h anesthetic exposure
and control (multiple 2-h exposures D = 0.037, P = 0.615
and one 6-h exposure D = 0.053, P = 0.163, respectively,
table3). This suggests that while a single brief neonatal exposure induces a selective decrease in larger spines,
repeated and prolonged exposures do not alter the distribution of spines.
Effects of Anesthesia on Presynaptic Mitochondrial

Density
Disector analysis revealed a decrease in mean presynaptic
mitochondrial density associated with longer cumulative
durations of sevoflurane exposure (table4). Relative to
control, multilevel analysis revealed a slight, but nonsignificant decrease in mitochondrial density after a single
2-h exposure on P7 ( = 0.041, P = 0.864). Prolonged
and repeated neonatal anesthetic exposure was associated
with significant reductions in mean mitochondrial density. As compared to control, both a single 6-h exposure
and multiple 2-h exposures were associated with a marked
decrease in presynaptic mitochondrial density ( = 0.543,
P = 0.023 and = 0.748, P = 0.002, respectively). However, in contrast to the above synaptic data, repeated 2-h
exposures did not result in a greater magnitude of mitochondrial loss relative to a single prolonged exposure (P =
0.156, table2).
To determine if the observed reduction in mean presynaptic mitochondrial density was secondary to bouton loss, we
calculated the fraction of boutons containing mitochondria.
Multilevel linear regression confirmed that the anesthesia
associated reductions in presynaptic mitochondrial density
was independent of bouton-loss (table4). As compared to
control, a single 6-h and multiple 2-h exposures to sevoflurane resulted in a decreased fraction of boutons containing
mitochondria ( = 0.107, P < 0.001 and = 0.116, P =
0.001, respectively). Post hoc analysis did not reveal a significant difference between these two cumulative 6-h exposures
(P = 0.770).
Fig. 3. Mean synaptic and mitochondrial density by treatment group: multiple and prolonged exposure to anesthesia was associated with greater reductions in both synaptic and mitochondrial density. Synaptic and mitochondrial density are expressed
as mean SEM. Treatment group (n = 5/group): 12h = single 2-h exposure postnatal day 7 (P7); 16h = single 6-h exposure (P7); 32h = three 2-h exposures (P7, P10, and P13). a, b, c indicates groups with differing superscripts (a, b, or c)
are significantly different from one another (statistical significance defined as P < 0.05). For example, for synaptic density: the
control and 12-h groups do not differ from each other, but both differ from the 16-h and 32-h groups, which also differ
from one another.
Table 1. Multilevel Linear Regression for Synaptic Density

Variable
Treatment group
12 h
16 h
32 h
Control
Coefficient
95% CI
P Value*
0.078
0.336
0.681
2.630
0.371 to 0.216
0.623 to 0.049
0.972 to 0.390
2.415 to 2.845
0.605
0.022
<0.001
Level 1 (dam) n = 6; level 2 (animal) n = 20; and level 3 (series) n = 100.

*Statistical significance as defined by P < 0.05. Treatment group: 12h =
single 2-h exposure postnatal day 7 (P7); 16h = single 6-h exposure (P7);
and 32h = three 2-h exposures (P7, 10, and 13).
Correlation between Synaptic Loss and Mitochondrial

Density
Analysis with Pearsons correlation coefficient demonstrated a
statistically significant relationship between the mean synaptic density and the mean presynaptic mitochondrial density
(r = 0.750, P < 0.001; fig.5). To control for common group
effects, the correlation between synaptic and mitochondrial
density was reevaluated using mean-centered data from each
group. Results remained significant (r = 0.500, P = 0.025),
and suggest a possible relationship between the alterations
in mitochondrial localization and observed synaptic losses.
Overall, these findings indicate that relative to a single
prolonged exposure, repeated neonatal exposure to anesthesia is highly neurotoxic, and results in greater ultrastructural
damage. Synapses appear to be particularly vulnerable, and
brief but repeated exposure results in marked and long-lasting synaptic loss. In contrast, mitochondrial toxicity is best
predicted by the cumulative duration of exposure, not the
total number of exposures.
Discussion
In this study, we demonstrate that neonatal sevoflurane
exposure produces long-lasting and dose-related ultrastructural damage. This damage is characterized by synaptic loss,
decreased presynaptic mitochondrial localization, and a shift
in the distribution of PSD length. Of note, the magnitude
of synaptic loss induced by three 2-h exposures was greater
than that of a single 6-h exposure of equivalent total length.
Conversely, animals exposed to a cumulative 6h of anesthesia had marked alterations in mitochondrial localization
that were independent of the total number of exposures.
Fig. 4. Cumulative frequency distribution: measurements

revealed a dose-related shift in postsynaptic density (PSD)
length. A single 2-h exposure on postal day 7 (P7) was associated with a decrease in larger head-bearing spines. The
detail is magnified and shown in the inset.
These findings provide a potential histological correlate for

human population-based studies demonstrating that children with repeated exposures to anesthesia are at increased
risk of attention-deficit/hyperactivity disorder and cognitive
impairment.1,35 One study by Flick et al.3 demonstrated a
2.12-fold increased risk of learning disability associated with
multiple, but not single, early exposures. The risk of learning disability has also been shown to increase with longer
cumulative durations of anesthetic exposures.1 Nonetheless,
these retrospective studies are limited in their ability to control for surgical stress, perioperative inflammation, hypoxia,
and comorbidity. The present study therefore provides an
important and experimentally controlled model to further
investigate these observations.
This is the first study of which we are aware to characterize
the long-term ultrastructural effects associated with multiple
early anesthetic exposures. We previously demonstrated that
relative to a single exposure, adult rats with multiple early
isoflurane exposures had greater impairment on a spatial
working memory task.13 Our work builds on prior literature
showing that even a single early exposure to N-methyl-d-aspartate antagonists or gamma aminobutyric acid mimetics
is immediately neurotoxic and triggers widespread apoptotic neurodegeneration.24,25 Of note, a 6-h neonatal exposure to GA has been shown to induce a 21-fold increase in
Table 2. Post Hoc Pairwise Comparisons with Bonferroni Correction

Synaptic Density
Treatment Groups*
12h vs.
16h vs.
16 h
23 h
23 h
2
3.03
16.98
5.31
P Value
0.042
<0.001
0.022
Mitochondrial Density
2
4.41
8.76
0.74
P Value
0.036
0.003
0.156
Fraction of Boutons with

Mitochondria
2
11.50
13.56
0.09
P Value
<0.001
<0.001
0.770
*Treatment groups: 12h = single 2-h exposure postnatal day 7 (P7); 16h = single 6-h exposure (P7); and 32h = three 2-h exposures (P7, 10, and 13).
Statistical significance as defined by P < 0.05.
Table 3. KolmogorovSmirnov Tests

Treatment Groups*
Control vs.
12h vs.
16h vs.
12 h
16 h
32 h
16 h
32 h
32 h
P Value
0.062
0.053
0.037
0.086
0.085
0.036
0.056
0.163
0.615
0.003
0.006
0.694
*Treatment groups: 12h = single 2-h exposure postnatal day 7 (P7); 16h
= single 6-h exposure (P7); and 32h = three 2-h exposures (P7, 10, and
13). Statistical significance as defined by P < 0.0083.
hippocampal CA1 neuronal degeneration.6 Lunardi et al.

subsequently demonstrated that rats with a single 6-h exposure to a cocktail of midazolam, nitrous oxide, and isoflurane
on P7, had long-lasting reductions in synaptic density and
the number of multiple synaptic boutons. Here we further
establish that relative to a single exposure, brief but repeated
exposures are associated with even greater long-term synaptic losses.
Dendritic spine morphology is thought to be closely
related to synaptic function.26 PSD length and spine head
volume are directly correlated with synaptic strength, the
number of receptors, and the availability of presynaptic
docked vesicles.16,22,27,28 Briner et al.9 previously demonstrated that rats exposed to propofol on P10 had lower numbers of mature prefrontal cortex spines with a head diameter
between 0.3 and 0.4 m. Similarly, we found that a single
2-h exposure to sevoflurane on P7 is associated with a selective decrease in larger head-bearing spines. However, longer
6-h exposures did not shift the cumulative distribution of
spine types. In the setting of a decreased spine density, this
suggests that brief neonatal anesthetic-exposures decrease
the stability of large spines, while prolonged exposure
causes a more uniform pattern of neurodegeneration, and
is unrelated to spine-type or maturity. This is of note, as it
demonstrates that even short exposures to GA may induce
long-term alterations in neuronal circuitry.
Impaired mitochondrial function and morphogenesis has
been suggested as a mechanism for anesthetic neurotoxicity and long-term cognitive impairment. Mitochondria are
dynamic organelles, critical to synaptogenesis, plasticity,
Fig. 5. Scatter plot of mean presynaptic mitochondrial density

by mean synaptic density.
calcium buffering, and vesicle recruitment.2931 Neonatal exposure to general anesthetics has been shown to alter
mitochondrial fission/fusion and activate the intrinsic,
mitochondrial-dependent, apoptotic cascade.12,32,33 A recent
study by Sanchez et al.10 showed persistent alterations in
mitochondrial morphology, function, and presynaptic localization after a single 6-h neonatal exposure to anesthesia.
Our results similarly demonstrate long-lasting reductions
in presynaptic mitochondrial density associated with longer cumulative exposures to sevoflurane. In the context of
our synaptic data, this suggests a threshold effect for GAinduced neurotoxicity. It is possible that early alterations in
mitochondrial trafficking or fission sensitize surviving synapses to subsequent exposures, and may in part explain the
clinical effects observed with multiple exposures.
The neurotoxicity associated with general anesthesia
is age dependent, and peak toxicity occurs during periods
of rapid neurodevelopment and synaptogenesis.25,34 In
humans, this period extends from the third trimester to several years postnatally, and up until P14 in rats. In the current
study, rat pups were exposed to GA between P7 and P13, a
timeframe thought to approximate the brain growth spurt
occurring at birth in humans.35 More conservative neuroinformatics models estimate that P7-10 is the neurodevelopmental equivalent of premature human infants born in
Table 4. Multilevel Linear Regression for Presynaptic Mitochondrial Density and the Fraction of Boutons Containing Mitochondria
Mitochondrial Density
Variable
Treatment group
12 h
16 h
32 h
Control
Boutons with Mitochondria
Coefficient
95% CI
P Value*
Coefficient
0.041
0.543
0.748
2.972
0.509 to 0.428
1.011 to 0.074
1.217 to 0.280
2.641 to 3.303
0.864
0.023
0.002
<0.001
0.000
0.107
0.116
0.595
95% CI
0.062 to 0.062
0.168 to 0.045
0.178 to 0.054
0.551 to 0.638
P Value*
0.999
0.001
<0.001
<0.001
Level 1 (dam) n = 6; level 2 (animal) n = 20; and level 3 (series) n = 100.

*Statistical significance as defined by P < 0.05. Treatment group: 12h = single 2-h exposure postnatal day 7 (P7); 16h = single 6-h exposure (P7); and
32h = three 2-h exposures (P7, 10, and 13).
the late-third trimester.3638 Nevertheless, these infants are at

high risk of morbidity and mortality, and often require early
and repeated exposure to anesthesia. Clarifying the neurohistologic impact of multiple anesthetic exposures is therefore
essential for clinical decision-making and the development
of pharmacological intervention.
Our research has several limitations. First, the number of
rats included in each experimental group was low (n = 5/
group). Although such a sample size is comparatively large for
stereologic electron microscopy studies, we may still be underpowered to detect more subtle differences or shifts in PSD
length.1517 Second, we cannot address whether there are any
sex-related differences, as all of the animals were male. This is
significant, as prior studies have reported that adult females
with single postnatal anesthetic-exposures have greater deficits in spatial working memory, while a more recent study
demonstrated long-term cognitive dysfunction in male, but
not female rats.13,39,40 Finally, our study employs a nonsurgical experimental model. Noxious stimulation itself causes
marked neurodegeneration, and prior research suggests that
sedation with ketamine may actually protect neonates from
pain-related neurotoxicity.41,42 It will be important in future
work to utilize models that incorporate a surgical stimulus
and examine effects in both males and females.
In conclusion, repeated neonatal exposure to GA is associated with greater long-term reductions in both synaptic density and presynaptic mitochondrial localization. Furthermore,
a single 2-h anesthetic exposure shifts the distribution of PSD
length, and alters dendritic spine morphology. This suggests
a threshold effect for GA-induced neurotoxicity, such that
even though brief exposures induce long-term alterations in
neuronal circuitry, longer exposures affect mitochondrial localization and sensitize surviving synapses to subsequent loss.
Acknowledgments
The authors thank John Morrison, Ph.D., from the
Departments of Neuroscience and Geriatrics and Palliative
Medicine; Yuko Hara, Ph.D., from the Department of Neuroscience; William Janssen, B.S., from the Department of
Neuroscience; Dara Dickstein, Ph.D., from the Departments
of Neuroscience and Geriatrics and Palliative Medicine; and
Allison Sowa, B.S., from the Department of Neuroscience,
the Icahn School of Medicine at Mount Sinai, New York,
New York, for their technical assistance with electron microscopy and analysis. The authors thank Paula Croxson,
Ph.D., from the Departments of Neuroscience and Psychiatry, the Icahn School of Medicine at Mount Sinai, for her
assistance with anesthesia.
Support was provided solely from institutional and/or
departmental sources.
Competing Interests
Correspondence
Address correspondence to Ms. Amrock: Icahn School of
Medicine at Mount Sinai, 1470 Madison Ave, Room 10116,
New York, New York 10029. levana.amrock@mssm.edu.

Information on purchasing reprints may be found at www.
anesthesiology.org or on the masthead page at the beginning of this issue. Anesthesiologys articles are made freely
References
1. Wilder RT, Flick RP, Sprung J, Katusic SK, Barbaresi WJ,
Mickelson C, Gleich SJ, Schroeder DR, Weaver AL, Warner
DO: Early exposure to anesthesia and learning disabilities
in a population-based birth cohort. Anesthesiology 2009;
110:796804
2. Yu CK, Yuen VM, Wong GT, Irwin MG: The effects of anaesthesia on the developing brain: A summary of the clinical
evidence. F1000Res 2013; 2:166
3. Flick RP, Katusic SK, Colligan RC, Wilder RT, Voigt RG,
Olson MD, Sprung J, Weaver AL, Schroeder DR, Warner DO:
Cognitive and behavioral outcomes after early exposure to
anesthesia and surgery. Pediatrics 2011; 128:e105361
4. Sprung J, Flick RP, Katusic SK, Colligan RC, Barbaresi WJ,
Bojani K, Welch TL, Olson MD, Hanson AC, Schroeder DR,
Wilder RT, Warner DO: Attention-deficit/hyperactivity disorder after early exposure to procedures requiring general
anesthesia. Mayo Clin Proc 2012; 87:1209
5. DiMaggio C, Sun LS, Li G: Early childhood exposure to anesthesia and risk of developmental and behavioral disorders in
a sibling birth cohort. Anesth Analg 2011; 113:114351
6. Jevtovic-Todorovic V, Hartman RE, Izumi Y, Benshoff ND,
Dikranian K, Zorumski CF, Olney JW, Wozniak DF: Early
exposure to common anesthetic agents causes widespread
neurodegeneration in the developing rat brain and persistent
learning deficits. J Neurosci 2003; 23:87682
7. Yon JH, Daniel-Johnson J, Carter LB, Jevtovic-Todorovic V:
Anesthesia induces neuronal cell death in the developing
rat brain via the intrinsic and extrinsic apoptotic pathways.
Neuroscience 2005; 135:81527
8. Satomoto M, Satoh Y, Terui K, Miyao H, Takishima K, Ito M,
Imaki J: Neonatal exposure to sevoflurane induces abnormal
social behaviors and deficits in fear conditioning in mice.
9. Briner A, Nikonenko I, De Roo M, Dayer A, Muller D, Vutskits
L: Developmental stage-dependent persistent impact of propofol anesthesia on dendritic spines in the rat medial prefrontal cortex. Anesthesiology 2011; 115:28293
10. Sanchez V, Feinstein SD, Lunardi N, Joksovic PM, Boscolo
A, Todorovic SM, Jevtovic-Todorovic V: General anesthesia
causes long-term impairment of mitochondrial morphogenesis and synaptic transmission in developing rat brain.
11. Lunardi N, Ori C, Erisir A, Jevtovic-Todorovic V: General anesthesia causes long-lasting disturbances in the ultrastructural
properties of developing synapses in young rats. Neurotox
Res 2010; 17:17988
12. Boscolo A, Milanovic D, Starr JA, Sanchez V, Oklopcic A, Moy
L, Ori C C, Erisir A, Jevtovic-Todorovic V: Early exposure to
general anesthesia disturbs mitochondrial fission and fusion
in the developing rat brain. Anesthesiology 2013; 118:108697
13. Murphy KL, Baxter MG: Long-term effects of neonatal single
or multiple isoflurane exposures on spatial memory in rats.
Front Neurol 2013; 4:87
14. Janssen WG, Vissavajjhala P, Andrews G, Moran T, Hof PR,
Morrison JH: Cellular and synaptic distribution of NR2A and
NR2B in macaque monkey and rat hippocampus as visualized with subunit-specific monoclonal antibodies. Exp
Neurol 2005; 191(suppl 1):S2844
15. Bloss EB, Puri R, Yuk F, Punsoni M, Hara Y, Janssen WG,
McEwen BS, Morrison JH: Morphological and molecular
changes in aging rat prelimbic prefrontal cortical synapses.

Neurobiol Aging 2013; 34:20010
16. Hara Y, Park CS, Janssen WG, Roberts MT, Morrison JH, Rapp
PR: Synaptic correlates of memory and menopause in the
hippocampal dentate gyrus in rhesus monkeys. Neurobiol
Aging 2012; 33:421.e1728
17. Hara Y, Yuk F, Puri R, Janssen WG, Rapp PR, Morrison JH:
Presynaptic mitochondrial morphology in monkey prefrontal
cortex correlates with working memory and is improved with
estrogen treatment. Proc Natl Acad Sci U S A 2014; 111:48691
18. Nikizad H, Yon JH, Carter LB, Jevtovic-Todorovic V: Early exposure to general anesthesia causes significant neuronal deletion
in the developing rat brain. Ann N Y Acad Sci 2007; 1122:6982
19. Tigges J, Herndon JG, Rosene DL: Preservation into old age
of synaptic number and size in the supragranular layer of
the dentate gyrus in rhesus monkeys. Acta Anat (Basel) 1996;
157:6372
20. de Groot DM, Bierman EP: A critical evaluation of methods
for estimating the numerical density of synapses. J Neurosci
Methods 1986; 18:79101
21. Sterio DC: The unbiased estimation of number and sizes of
arbitrary particles using the disector. J Microsc 1984; 134(Pt
2):12736
22. Harris KM, Stevens JK: Dendritic spines of CA 1 pyramidal
cells in the rat hippocampus: serial electron microscopy
with reference to their biophysical characteristics. J Neurosci
1989; 9:298297
23. Aarts E, Verhage M, Veenvliet JV, Dolan CV, van der Sluis S: A
solution to dependency: Using multilevel analysis to accommodate nested data. Nat Neurosci 2014; 17:4916
24. Johnson SA, Young C, Olney JW: Isoflurane-induced neuroapoptosis in the developing brain of nonhypoglycemic mice.
J Neurosurg Anesthesiol 2008; 20:218
25. Ikonomidou C, Bosch F, Miksa M, Bittigau P, Vckler J,
Dikranian K, Tenkova TI, Stefovska V, Turski L, Olney JW:
Blockade of NMDA receptors and apoptotic neurodegeneration in the developing brain. Science 1999; 283:704
26. Arellano JI, Benavides-Piccione R, Defelipe J, Yuste R:
Ultrastructure of dendritic spines: Correlation between synaptic and spine morphologies. Front Neurosci 2007; 1:13143
27. Nusser Z, Lujan R, Laube G, Roberts JD, Molnar E, Somogyi
P: Cell type and pathway dependence of synaptic AMPA
receptor number and variability in the hippocampus. Neuron
1998; 21:54559
28. Schikorski T, Stevens CF: Morphological correlates of functionally defined synaptic vesicle populations. Nat Neurosci
2001; 4:3915
29. Billups B, Forsythe ID: Presynaptic mitochondrial calcium

sequestration influences transmission at mammalian central
synapses. J Neurosci 2002; 22:58407
30. Chang DT, Honick AS, Reynolds IJ: Mitochondrial trafficking
to synapses in cultured primary cortical neurons. J Neurosci
2006; 26:703545
31. Zucker RS: Calcium- and activity-dependent synaptic plasticity. Curr Opin Neurobiol 1999; 9:30513
32. Boscolo A, Starr JA, Sanchez V, Lunardi N, DiGruccio MR, Ori
C, Erisir A, Trimmer P, Bennett J, Jevtovic-Todorovic V: The
abolishment of anesthesia-induced cognitive impairment
by timely protection of mitochondria in the developing rat
brain: The importance of free oxygen radicals and mitochondrial integrity. Neurobiol Dis 2012; 45:103141
33. Yon JH, Carter LB, Reiter RJ, Jevtovic-Todorovic V: Melatonin
reduces the severity of anesthesia-induced apoptotic neurodegeneration in the developing rat brain. Neurobiol Dis
2006; 21:52230
34. Vutskits L: General anesthesia: A gateway to modulate synapse formation and neural plasticity? Anesth Analg 2012;
115:117482
35. Dobbing J, Sands J: Comparative aspects of the brain growth
spurt. Early Hum Dev 1979; 3:7983
36. Rice D, Barone S Jr: Critical periods of vulnerability for the
developing nervous system: evidence from humans and
animal models. Environ Health Perspect 2000; 108(suppl
3):51133
37. Clancy B, Finlay BL, Darlington RB, Anand KJ: Extrapolating
brain development from experimental species to humans.
Neurotoxicology 2007; 28:9317
38. Clancy B, Darlington RB, Finlay BL: Translating develop
mental time across mammalian species. Neuroscience 2001;
105:717
39. Lee BH, Chan JT, Kraeva E, Peterson K, Sall JW: Isoflurane
exposure in newborn rats induces long-term cognitive dysfunction in males but not females. Neuropharmacology 2014;
83:917
40. Boscolo A, Ori C, Bennett J, Wiltgen B, Jevtovic-Todorovic V:
Mitochondrial protectant pramipexole prevents sex-specific
long-term cognitive impairment from early anaesthesia exposure in rats. Br J Anaesth 2013; 110(suppl 1):i4752
41. Istaphanous GK, Loepke AW: General anesthetics and the
developing brain. Curr Opin Anaesthesiol 2009; 22:36873
42. Anand KJ, Garg S, Rovnaghi CR, Narsinghani U, Bhutta
AT, Hall RW: Ketamine reduces the cell death following
inflammatory pain in newborn rat brain. Pediatr Res 2007;
62:28390
Assessment of Neutrophil Gelatinase-associated

Lipocalin in the Brain-dead Organ Donor to Predict
Immediate Graft Function in Kidney Recipients
A Prospective, Multicenter Study
Laurent Muller, M.D., Ph.D., Armelle Nicolas-Robin, M.D., Ph.D., Sophie Bastide, M.D.,
Orianne Martinez, M.D., Guillaume Louart, M.D., Jean-Christian Colavolpe, M.D., Florence Vachiery, M.D.,
Sandrine Alonso, M.Sc., Jean-Yves Lefrant, M.D., Ph.D., Bruno Riou, M.D., Ph.D.; for AzuRea Group*
ABSTRACT
Background: Delayed graft function is a major determinant of long-term renal allograft survival. Despite considerable efforts
to improve donor selection and matching, incidence of delayed graft function remains close to 25%. As neutrophil gelatinaseassociated lipocalin (NGAL) has been shown to predict acute renal failure, the authors tested the hypothesis that NGAL
measurement in brain-dead donors predicts delayed graft function in kidney recipients.
Methods: In a prospective, multicenter, observational study, serum NGAL was measured in donors at the time of transfer to
operating room. The primary endpoint was the delayed graft function, defined as the need for renal replacement therapy during the first week posttransplantation.
Results: Among 159 included brain-dead donors, 146 were analyzable leading to 243 renal transplantations. Of these, 56
(23%) needed renal replacement therapy. Donors NGAL values were similar in case of both delayed and normal graft function in recipients. The area under the receiver-operating curve for NGAL to predict the need for renal replacement therapy
before day 8 was 0.50 (95% CI, 0.42 to 0.59). The area under curve for NGAL to predict failure to return to a normal graft
function at day 8 was 0.51 (95% CI, 0.44 to 0.59). Using multivariate analysis, NGAL was not associated to the need for renal
replacement therapy (odds ratio, 0.99; 95% CI, 0.98 to1.00) or failure to return to a normal graft function at day 8 (odds
ratio, 1.00; 95% CI, 0.99 to 1.00).
Conclusion: NGAL measurement in brain-dead donors at the time of recovery failed to predict delayed or normal graft function in kidney recipients. (Anesthesiology 2015; 122:96-105)
ENAL transplantation is considered as the standard

therapy for end-stage chronic kidney disease1 because
it allows a better quality of life with a lower cost than dialysis.1,2 Although living donation has been shown as a valuable
and safe response to shortage of organs for transplantation,
a large majority of kidney transplantation is obtained from
organ donation after brain death (BD).3,4 Quality of donor
care is a major determinant of graft success and increases
the number of eligible donors.5,6 Despite considerable efforts
to improve quality of care for transplanted patients, 20 to
30% of delayed graft function (DGF, defined as a need for
renal replacement therapy [RRT] within 8 days after transplantation) are usually reported after renal transplantation
from deceased donors.79 As early recovery of graft function

Delayed graft function in renal transplant recipients still remains
at 25%

Despite the ability to predict acute renal failure earlier than serum
creatinine rises in critically ill patients, neutrophil gelatinase-associated lipocalin measurements in blood samples obtained from
brain-dead donors before kidney graft harvesting failed to predict either delayed or normal graft function in kidney recipients
determines long-term graft survival, DGF is a fundamental

prognosis variable.9,10 Expanded criteria for donors (ECD),
especially age greater than 60 yr, could be a major explanation
Submitted for publication April 7, 2014. Accepted for publication August 14, 2014. From the Department of Anesthesiology and Critical Care (L.M., G.L., J.-Y.L.) and Department of Biostatistics and Clinical Epidemiology (S.B., S.A.), Nmes University Hospital, Place du Pr
Robert Debr, Nmes, France; Department of Anesthesiology and Critical Care, CHU Piti-Salptrire, Paris, France, and UMRS INSERM 956,
Universit Pierre et Marie Curie, Paris, France (A.N.-R., B.R.); Department of Anesthesiology and Critical Care, CHU Lapeyronie, Montpellier,
France (O.M.); Department of Anesthesiology and Critical Care, CHU Timone, Marseille, France ( J.-C.C.); and Department of Anesthesiology
and Critical Care, CHU Saint Eloi, Montpellier, France (F.V.).
*Members of the AzuRea Group are listed in the appendix.
January 2015
for the high rate of DGF.11,12 Nevertheless, if ECD influence

graft success, some studies surprisingly report only a slight
difference of graft prognosis obtained from standard criteria
donors versus ECD.7,11 These observations reflect the influence of other detrimental factors such as cold ischemic time,
Human Leukocyte Antigen match, female sex, Caucasian
origin,11 and the global quality of donor management.13 It
has been demonstrated that cold ischemic time and DGF
are two major factors of long-term renal allograft survival.10
Because DGF rate remains high, obtaining new criteria to
predict DGF could be of particular interest to detect occult
renal failure in brain-dead14 donors to improve donor
recipient matching. Recently, it was hypothesized that new
biomarkers could better predict acute kidney injury (AKI)
than classical markers such as serum creatinine or blood urea
nitrogen.15 The neutrophil gelatinase-associated lipocaline
(NGAL) was identified as a 25kDa polypeptide chain of 178
amino acids that belongs to the lipocalin family of proteins
and is produced by polymorphonuclear neutrophil cells and
predominantly by renal tubule epithelial cells.16 In intensive care unit (ICU), NGAL predicts AKI 48h earlier than
serum creatinine and/or Risk, Injury, Failure, Loss, Endstage kidney disease classification17 criteria.18 The NGAL
measurement could be also useful for assessing graft function after renal transplantation.1921 In contrast, NGAL has
been poorly studied to assess occult renal dysfunction in BD
donors. Only one previous study reported that both urinary
NGAL (U-NGAL) and serum NGAL (S-NGAL) failed to
predict DGF in 99 BD donors.22 However, this previous
study has several limitations: it was a monocentric study,
with a potential center effect; measurement of S-NGAL was
performed before the occurrence of BD in a significant proportion of patients (36%); and all patients were given glucocorticoids, which does not reflect the practice of all centers.
These limitations justify a new multicenter study to clarify
the usefulness of NGAL as biomarker in BD patients for predicting DGF in kidney recipients. Therefore, in the current
study, we test the hypothesis that S-NGAL measurement in
BD organ donors predicts DGF in kidney recipients.

Considering that the current study was noninterventional
and that such investigation did not need significant additional blood sample (volume <0.5ml), the institutional
review board approved this study (Institutional Review
Board, Nmes University Hospital, France, file number
2011.01.01) and stated that informed consent of next-of-kin
was not required as organ donors are not considered human
subjects. However, the next-of-kin was systematically orally
informed and could refuse the participation. According to
the French law, the National Committee on Health Research
Data Analysis (CCTIRS, Comit Consultatif sur le Traitement
de lInformation en matire de Recherche dans le domaine de la
Sant, Nmes France, file number: 11.140) and the National
Committee on Information Technology (Commission Nationale de lInformatique et des Liberts, Paris, France, file number EGY/FLR/AR114292) also approved this study. This
prospective, multicenter, observational study was conducted
by a French anesthesia and intensive care research network
(AzuRea group). Patients were enrolled in 19 medical or surgical ICU of eight French teaching hospitals more than an
11-month period (February to December 2011). Data were
anonymously collected using electronic case report form.
Brain-dead Organ Donors
All BD patients in whom organ retrieval was performed
were eligible. Clinical diagnosis of BD was performed with
a strict adherence to the standard international criteria23
and confirmed by either computed tomography angiography or electroencephalography, according to the French law.
An interview of relatives and/or closest was systematically
performed to confirm the absence of known refusal of the
BD patients to organ donation; we also checked the French
national registry of refusal of organ donation. Standard medical contraindications to organ donation were also checked.
Patients in whom a contraindication to kidney donation was
identified were excluded. In BD donors, the following variables were collected: age, sex, weight, height, medical history
of cardiac failure (defined as New York Heart Association 3
or 4 classes), coronary artery disease, arterial hypertension,
peripheral obliterating arterial disease, tobacco consumption (more than one pack per day during 20 yr), diabetes
mellitus, hyperlipidemia (defined as chronic treatment for
hypercholesterolemia or hypertriglyceridemia). The ECD
for kidney were defined by the following elements: age older
than 60 yr, age between 50 and 60 yr and one of the following risk criteria as arterial hypertension, stroke as a cause of
BD, and at least one serum creatinine dosage greater than
130 mol/l. The sepsis-related organ failure assessment
score24 was recorded at the time of NGAL sampling.
We recorded mean arterial blood pressure (mmHg), heart
rate (beats/min), at the time of transfer to the operating
room and the worst value of mean arterial pressure during
the 24h before, urine output during the 6 and 12h before
transfer, the total amount and the type of fluids infused
to donor during the 24h before transfer, and the dose of
norepinephrine administered at the time of transfer to the
operating room. Serum creatinine and blood urea nitrogen,
serum lactate, arterial blood gases, proteinuria, at the time of
transfer to the operating room, and their worst values in the
previous 48h were also recorded.
Sample Collection and NGAL Detection
To assess the whole period at risk for AKI before kidney harvesting, a blood sample (0.25ml) was withdrawn just before
entering into the operating room. Each center was equipped
with a portable biological diagnostic device allowing NGAL
assay in few minutes (Triage MeterPro Meter; Biosite Inc.,
San Diego, CA). The NGAL measurements were performed
Anesthesiology 2015; 122:96-105 97 Muller et al.
Donor NGAL and Delayed Graft Function in Recipient
by fluorescence immunoassay specific kits (Triage meter

NGAL test, Alere, Jouy en Josas, France). The range of
quantification was 60 to 1,300ng/ml. For this technique,
the reported variation coefficient was 10 to 15%.
Kidney Recipients
Assessment of renal function in kidney recipients was based
on data obtained from the different transplantation units
during the first postgraft week. DGF was the primary endpoint and was defined as any RRT before day 8 (D8) posttransplantation.25 The secondary endpoint was the failure to
return to a normal graft function. A normal graft function
was defined as urine output greater than 1,000ml/day and
serum creatinine less than 167 mol/l during 2 consecutive
days at D8 posttransplantation.12,25 We also recorded cold
ischemic time defined as the time from aortic clamping
in donor and graft declamping in recipient for each transplanted kidney.
All statistical analyses were performed using SAS (SAS
Institute, Cary, NC) version 9.3. Quantitative variables
were expressed as median with 25th and 75th percentiles
and compared using MannWhitney tests. Qualitative variables were expressed as proportions and compared using
chi-square tests (or Fisher exact tests when appropriate).
Receiver operator characteristic curves were generated to
evaluate the capacity of NGAL to predict DGF and failure
to return to a normal graft function. A logistic mixed model
with three hierarchical levels was used with a center random
effect and a donor random effect within each center to take
into account the link between the two transplanted kidneys
from the same donor. Univariate and multivariate analyses
were successively performed to assess the capacity of NGAL
to predict DGF by adjusting on other expected predictors
such as cold ischemic time, ECD, serum creatinine.
The sample size calculation was based on a previous study
conducted by our group in which we found a proportion of
DFG of 30%.26 We tested the hypothesis that NGAL is a significant predictor of dialysis before D8 posttransplantation,
that is, that it could predict RRT with a value of area under
the receiver operator characteristic curve of 0.82 and accuracy
precision of 0.07 (keeping the lower value of the CI >0.75).27
We calculated that 132 patients were required, to be conservative we decided thus to enroll at least 150 patients.
All P values were two tailed and a P value less than 0.05
was considered significant.
Results
During the study period, 159 BD patients with renal donation were included. Characteristics of BD patients and their
management are shown in table1. Figure1 shows study flow
chart. Dosage of S-NGAL was not performed for 13 patients
leading to 146 analyzable BD donors that could have theoretically led to 292 renal transplantations (fig. 1). Finally,
29 transplants were not grafted and 20 recipients were not

assessed for renal function at D8, for a total of 243 transplants assessable for primary and secondary endpoints. The
mean cold ischemic time in recipient population was 14.5
(12 to 19) h.
Among 243 recipients (115 and 128 right and left transplants, respectively), 56 (23%) needed RRT before D8 after
graft. The median S-NGAL value was similar in case of RRT
requirement or not (103 [60 to 300] vs. 102 [60 to 260]
ng/ml, P = 0.94). The distribution of S-NGAL values was
strictly comparable in case of DGF as compared to those
without DGF (fig.2A). The AUC of S-NGAL for predicting RRT in recipients was 0.50 (95% CI, 0.42 to 0.59) (fig.
2B). The AUC of serum creatinine measured at the same
time was 0.60 (95% CI, 0.53 to 0.67). At D8, multivariate
analysis showed that S-NGAL was not significantly associated to DGF, whereas a 1-h longer cold ischemic time and
a 10 mol/l increase in creatinine plasma value were both
significantly associated to DGF (table2).
Among the 243 recipients, 112 (46%) failed to return to
a normal graft function at D8. The S-NGAL dosage were
similar in case of normal graft function or not (102 [60 to
226] vs. 106.5 [60 to 317.5] ng/ml, P = 0.69). The distribution of S-NGAL values between recipients with normal
graft function or not was similar (fig.3A). The AUC the
receiver operator characteristic curve of NGAL for predicting the failure to return to a normal graft function at D8 was
0.51 (95% CI, 0.44 to 0.59) (fig. 3B). In the logistic mixed
model, S-NGAL was not predictive of failure to return to a
normal graft function, whereas a longer cold ischemic time
and an increase in creatinine plasma value were both significantly predictive of failure to return to a normal graft function at D8 (table3).
For the need of RRT at day 8, the logistic mixed model
showed no significant impact center effect (P = 0.17) or no
significant impact of the link between the two transplanted
kidneys from the same donor (P = 0.29). For the return to
normal renal function at day 8, the logistic mixed model
showed no significant impact center effect (P = 0.26) or no
significant impact of the link between the two transplanted
kidneys from the same donor (P = 0.06).
Discussion
In the current prospective multicenter study, NGAL measurement in BD organ donors failed to predict DGF or
normal graft function in kidney recipients. Cold ischemic
time and an increase in serum creatinine in BD donor were
significantly associated to DGF.
NGAL is one of the most promising and studied AKI
biomarkers.15,28 In healthy subjects, proximal and distal
tubular epithelial cells synthesize and secrete low levels of
NGAL that are detectable in plasma. An extra-renal production of NGAL (gut, bone marrow, and lungs) also participates to NGAL plasma level. In the glomerulus, plasma
NGAL is rapidly filtered. The filtered NGAL is largely
Table 1. Characteristics of Brain-dead Donors (n = 159)

Variables
Values
Women
Men
Age (yr)
Weight (kg)
Height (cm)
Medical history
Chronic cardiac failure or coronary artery disease (missing data = 3)
Hypertension
Peripheral vascular occlusive arteriopathy (missing data = 3)
Diabetes mellitus
Dyslipidemia
Smoker
Cause of brain death
Stroke
Traumatic brain injury
Cardiac arrest
Others
SOFA score
Maximal creatinine in the last 48h before operating room (mol/l)
Maximal urea in the last 48h before operating room (mmol/l)
Urine output in the last 12h before operating room (ml kg1 h1)
Urine output in the last 6h before operating room (ml kg1 h1)
Vasoactive drugs (missing data = 11)
Norepinephrine
Dobutamine
Dopamine
Terlipressine
Time from ICU admission to brain death diagnosis (days)
Time from brain death to aorta clamping (h)
Blood lactates at time of operating room (mmol/l)
Heart rate (beats/min)
Minimal mean arterial pressure in the last 24h before operating room (mmHg)
Mean arterial pressure just before the transfer to operating room (mmHg)
Norepinephrine posology before the transfer to operating room (mg/h)
Arterial pH before the transfer to operating room
Arterial Po2 before the transfer to operating room (mmHg)
Arterial Pco2 before the transfer to operating room (mmHg)
Proteinuria before the transfer to operating room (g/l)
Cristalloids in the last 24h before operating room
Colloids in the last 24h before operating room
NGAL (ng/ml) (13 missing data)
59 (37%)
100 (63%)
53 (3862)
72 (6381)
172 (165179)
11 (7%)
37 (23%)
2 (1%)
13 (8%)
15 (9%)
46 (29%)
84 (53%)
57 (36%)
12 (8%)
6 (4%)
10 (912)
90 (68115)
6.0 (4.78.4)
1,700 (1,1002,500)
775 (4851,200)
145 (98%)
15 (10%)
7 (5%)
2 (1%)
2 (13)
14 (1017)
1.6 (1.12.2)
94 (82112)
66 (6074)
85 (7696)
1.78 (1.003.6)
7.40 (7.307.44)
309 (177398)
37 (3244)
0.20 (0.100.52)
3,000 (2,0004,000)
500 (01,000)
108.5 (60.0285.0)
Results are expressed as n (%) or median (25th percentile75th percentile).

ICU = intensive care unit; NGAL = neutrophil gelatinase-associated lipocalin; SOFA = Sequential Organ Failure Assessment.
reabsorbed by proximal tubule, via receptor binding (Megalin) and endocytosis. Therefore, in physiological conditions,
NGAL is barely detectable in urine or plasma. In case of
AKI with acute tubular injury, plasma and urine levels of
NGAL rapidly and markedly increase. This major and early
response (within few hours) is of particular interest to early
detect and prevent AKI. In children scheduled for cardiac
surgery, the AUC for NGAL at 2 and 4h postbypass were
0.99 and 1.0, respectively.29 In a study involving 88 general
ICU patients, Constantin et al. showed that plasma NGAL
increases 48h before Risk, Injury, Failure, Loss, End-stage
kidney disease criteria. In this study, a cutoff of 155 nmol/l
showed a sensitivity and specificity to predict AKI of 82 and

97%, respectively, with an AUC of 0.92 (95% CI, 0.85 to
0.97). These results were weakened by less conclusive findings in adult cardiac surgery, ICU, and emergency patients,
in whom AUC for S-NGAL or U-NGAL varied from 0.54
to 0.79, which can be considered as moderately informative.3036 The main concern of NGAL interpretation in these
large studies is that there is a wide overlap of NGAL values between AKI and non-AKI patients. Although a very
low value (<200 nmol/l) is usually associated with no AKI,
a higher value can be recorded both in AKI and non-AKI
patients (false positive), especially in patients with a high
159 brain-dead donors

Paris Piti Salptrire
Montpellier Lapeyronie
Montpellier Gui de Chauliac
Clichy Beaujon
Nice Saint Roch
Marseille Timone
Saint Etienne
Nmes
16
17
35
7
43
13
13
15
with missing S-NGAL dosage

No graft for technical reasons
Left kidney, n = 13
Right kidney, n = 16
No available DGF at day 8
Left kidney, n = 5
Right kidney, n = 15
243 transplanted kidneys

Left = 128
Right = 115
53 (23%) recipients with RRT before day 8

131 (54%) recipients with normal graft function at day 8
Fig. 1. Study flow chart. DGF = delayed graft function; RRT =

renal replacement therapy; S-NGAL = serum neutrophil gelatinase-associated lipocalin.
systemic inflammatory response syndrome, which is usually

observed during BD.37 In the frame of renal transplantation,
NGAL measurement after transplantation was suggested to
be useful to follow transplanted patients and to predict graft
recovery or failure.20 Therefore, the usefulness of NGAL in
the setting of kidney transplantation from BD donor is challenged. In the frame of BD donor, only one study tested
the ability of NGAL to predict DGF.22 In this study, neither
S-NGAL nor U-NGAL in donor could accurately predict
DGF in recipients. This previous study had several limitations: monocentric study, measurement of S-NGAL performed before the occurrence of BD in 36% of cases and
systematic use of glucocorticoids. The design of the current
study ruled out this methodological limitations and showed
similar results.
Several hypotheses can be raised to explain our negative
results concerning NGAL as biomarker predictive for DGF
that markedly differ to other clinical situations in critically ill
patients. First, we decided to perform NGAL measurement
at the time of surgery to detect an occult renal dysfunction
after the occurrence of BD and hemodynamic stabilization.
However, Hollmen et al.22 performed NGAL measurement
earlier and they also failed to observe positive results. Thus,
the timing does not appear as a valuable explanation for our
negative results. Second, it has been widely demonstrated

that an aggressive management of BD donor increases
the number and the quality of grafts.6 Therefore, it can be
hypothesized that a high quality of ICU care may limit AKI
in BD donors. In our study, the median S-NGAL was very
low: 108.5ng/ml. Such low NGAL levels can be due to the
efficacy of an aggressive donor resuscitation leading to a low
proportion of significant AKI in our donors. Also, low levels
of serum creatinine at the same time (table 1) confirm this
hypothesis. It should also be noted that the S-NGAL values
reported by Hollmen et al.22 in a similar setting were also
relatively low (median = 212ng/ml). This reinforces the idea
that AKI is nowadays well controlled by aggressive donor
resuscitation. As the timing of S-NGAL measurement was
performed earlier in the Hollmen et al.22 study and because
the median S-NGAL value was twice the value observed in the
current study (212 vs. 108.5ng/ml), it can be hypothesized
that S-NGAL value decreases over time as a consequence of
efficient donor management. It could also be hypothesized
that serial measurement of NGAL (at admission and before
surgery) could have been more informative, but the current
result do not allow supporting this hypothesis. It could be
interesting to test the ability of NGAL variations over time
in donors to predict DGF in recipients. Third, the major
factors leading to DGF might be postrecovery factors rather
than factors related to pregraft donor conditions. In the current study, the multivariate analysis shows that cold ischemic
time has the highest OR (tables 2 and 3), suggesting that
postrecovery conditions are major determinant of graft function. Recently, Malinoski et al.38 showed that age, cold ischemic time, and creatinine are independent factors of DGF.
Finally, the weak prediction of DGF by S-NGAL observed
in the current study may be related to the type of measurement itself. It could be argued that S-NGAL is less specific
of renal damages than U-NGAL since U-NGAL detected
in urine mainly derives from renal epithelial cells as a consequence of AKI.39 However, in the study from Hollmen
et al.,22 U-NGAL was not superior to S-NGAL to predict
DGF. S-NGAL can be produced through nonrenal (bone
marrow, gut, and lungs) synthesis NGAL.16 Another way of
extra-renal NGAL production could be the systemic inflammatory response syndrome induced by BD, as NGAL is an
acute phase protein, which depends on cytokine release.16
As S-NGAL is filtered through the glomerulus, one cannot
exclude that high U-NGAL concentration may also reflect
extra-renal overproduction of NGAL. However, because we
observed low values of NGAL in BD patients, the hypothesis
that extra-renal S-NGAL production may interfere with its
diagnostic accuracy is not likely.
The following limitations have to be considered to assess
the clinical relevance of our results. First, in this multicenter
trial, there was no recommended donor algorithm management, especially indications for RRT. Thus, we cannot exclude
management heterogeneity between the different units leading to a less powerful interest of NGAL measurement. The
Fig. 2. (A) Distribution of serum neutrophil gelatinase-associated lipocalin (S-NGAL) values obtained from donors with no subsequent renal replacement therapy (RRT) in recipients at day 8 (red dashed line) and from donors with subsequent RRT in recipients
at day 8 (blue solid line). The density function represents the number of patients with a given value of NGAL. (B) Receiver operator characteristic (ROC) curve for S-NGAL to predict the need for renal replacement therapy at day 8 postgraft. The area under
the ROC curve was 0.50 (95% CI, 0.42 to 0.59). No cutoff value could be identified.
Table 2. Donors and Transplantation Factors that Influence the Requirement for Renal Replacement Therapy in Recipients in Uniand Multivariate Logistic Mixed Model Analyses
Univariate Analysis
Factor
NGAL increase >1ng/ml
Cold ischemic time (per increase = 1h)
Serum creatinine increase in donor >10 mol/l
Multivariate Analysis
OR (95% CI)
P Value
OR (95% CI)
P Value
1.001 (0.9991.002)
1.115 (1.0571.177)
1.064 (1.0061.126)
0.48
<0.001
0.03
0.999 (0.9971.002)
1.116 (1.0571.178)
1.072 (1.0041.145)
0.57
<0.001
0.04
NGAL = neutrophil gelatinase-associated lipocalin; OR = odds ratio.
Fig. 3. (A) Distribution of serum neutrophil gelatinase-associated lipocalin (S-NGAL) values obtained from donors with subsequent abnormal graft function in recipients at day 8 (red dashed line) and from donors with subsequent normal graft function in recipients at day 8 (blue solid line). The density function reflects the number of patients with a given value of S-NGAL.
(B) Receiver operator characteristic (ROC) curve for S-NGAL to predict the failure to return to a normal renal graft function at day
8 postgraft. A normal graft function at day 8 was defined as urine output greater than 1,000ml/day and creatinine less than 167
mol/l during two consecutive days at day 8 posttransplantation.12,25 The area under the ROC curve was 0.51 (95% CI, 0.44 to
0.59). No cutoff value could be identified.
current data show a trend for a center effect for slow graft
function then we cannot reject the hypothesis that this study
is underpowered to show a significant center effect. Second,
we did not analyze recipient characteristics, especially age,
warm ischemia time during transplantation, and severity
scores. This lack of recipients data theoretically limits the
introduction of confounding variables in the model. Therefore, the current results need to be interpreted with caution.
Despite these limitations, DGF percentage is comparable

to literature and usual predicting factors for DGF were also
found. Then, the current population probably reflects a representative donor and recipient population. Moreover, because
the NGAL values reported in the current study are very low
(table 1), we think that the lack of link between NGAL and
DGF is partly due to the lack of renal failure in donor and not
because of confounding variables. Third, we did not analyze
Table 3. Donors and Transplantation Factors that Influence Failure to Return to a Normal Graft Function in Recipients in Uni- and
Multivariate Logistic Mixed Model Analyses
Univariate Analysis
Factor
NGAL increase >1ng/ml
Extended criteria donor
Cold ischemic time increase >1 h
Serum creatinine increase in donor >10 mol/ml
Multivariate Analysis
OR (95% CI)
P Value
OR (95% CI)
P Value
1.001 (0.9991.003)
1.573 (0.8203.019)
1.073 (1.0221.127)
1.089 (1.0211.161)
0.26
0.17
0.005
0.01
1.000 (0.9981.002)
1.642 (0.8483.180)
1.073 (1.0211.128)
1.091 (1.0161.172
0.87
0.14
0.006
0.02
NGAL = neutrophil gelatinase-associated lipocalin; OR = odds ratio.
other relevant biomarker like cystatin C or cycle cell arrest

biomarker or association of several markers that have recently
been shown as promising.40 Finally, It could be objected that
only terminal creatinine value could have been studied, rather
than creatinine variation. The current result suggests that creatinine variation over time is a better predictor than NGAL for
DGF. This finding reinforces the use of scores as Risk, Injury,
Failure, Loss, End-stage kidney disease or surrogates; those
are mainly based on creatinine variation, and widely recommended in ICU patients. It has been demonstrated that donor
terminal creatinine is strongly associated with DGF.41,42 In the
current study, we did not specifically studied final (terminal)
creatinine value, but we recorded the variation of creatinine
during the last 48h of donor resuscitation. This variation was
significantly associated with DGF. As we recorded creatinine
value at the time of surgical procedure, this creatinine variation takes into account terminal creatinine value. Then, it can
be postulated that the 10% creatinine variation reported in
the current study indirectly reflect terminal creatinine value,
but we did not specifically tested this hypothesis.
Conclusion
Unlike what has been previously demonstrated in ICU and
emergency patients, S-NGAL measurement in BD kidney
donor at the time of surgery does not predict delayed or normal graft function after transplantation.
Acknowledgments
The authors thank intensive care and organ donation teams
of Clermont-Ferrand, Grenoble, Lyon, Marseille, Montpellier, Nice, Nmes, and Paris Piti-Salptrire. The authors
thank Audrey Ayral, Franoise Casano, Loubna Elotmani,
and Sophie Lloret (Department of Anesthesiology and Critical Care, Nmes University Hospital, Place du Pr Robert Debr, Nmes, France) for their substantial contribution to the
current study. The authors also thank Mariella Lomma (Department of Biostatistics and Clinical Epidemiology, Nmes
University Hospital, Place du Pr Robert Debr) for reviewing
English writing.
To support this project, the major part of grants (30,000)
was obtained from Nmes Teaching Hospital (Place du Pr
Robert Debr, Nmes, France). A private grant, obtained from
ALERE society (ALERE France, Jouy en Josas, France), was
also used to complete the NGAL dosing funding (3,000).
This society also provided dosing portable machines distributed in each participating unit.
Competing Interests
Correspondence
Address correspondence to Dr. Lefrant: Department of Anesthesiology and Critical Care, Nmes University Hospital,
Place du Pr Robert Debr, 30029 Nmes 09, France. jean.
yves.lefrant@chu-nimes.fr. Information on purchasing reprints may be found at www.anesthesiology.org or on the
masthead page at the beginning of this issue. Anesthesiologys articles are made freely accessible to all readers, for personal use only, 6 months from the cover date of the issue.
References
1. Shapiro R: End-stage renal disease in 2010: Innovative
approaches to improve outcomes in transplantation. Nat Rev
Nephrol 2011; 7:6870
2. Page TF, Woodward RS: Cost-effectiveness of Medicares coverage of immunosuppression medications for kidney transplant recipients. Expert Rev Pharmacoecon Outcomes Res
2009; 9:43544
3. Ibrahim HN, Foley R, Tan L, Rogers T, Bailey RF, Guo H,
Gross CR, Matas AJ: Long-term consequences of kidney
donation. N Engl J Med 2009; 360:45969
4. Cecka JM: Kidney transplantation in the United States.
Clinical Transplants 2008; pp 118
5. Salim A, Martin M, Brown C, Rhee P, Demetriades D, Belzberg
H: The effect of a protocol of aggressive donor management:
Implications for the national organ donor shortage. J Trauma
2006; 61:42933; discussion 4335
6. Malinoski DJ, Patel MS, Daly MC, Oley-Graybill C, Salim A;
UNOS Region 5 DMG Workgroup: The impact of meeting
donor management goals on the number of organs transplanted per donor: Results from the United Network for
Organ Sharing Region 5 prospective donor management
goals study. Crit Care Med 2012; 40:277380
7. Hassanain M, Tchervenkov J, Cantarovich M, Metrakos P,
Paraskevas S, Keith D, Baran D, Fernandez M, Mangel R,
Chaudhury P: Delayed graft function has an equally bad
impact on deceased donor renal graft survival in both standard criteria donors and expanded criteria donors. Transplant
Proc 2009; 41:1334
8. Hassanain M, Tchervenkov JI, Cantarovich M, Metrakos P,
Paraskevas S, Keith D, Baran D, Fernandez M, Mangel R,
Chaudhury P: Recovery of graft function early posttransplant
determines long-term graft survival in deceased donor renal
transplants. Transplant Proc 2009; 41:1246
9. Sinz MM, Toro JC, Poblete HB, Perez LF, Nicovani VH,
Carrera MG: Incidence and factors associated with delayed
graft function in renal transplantation at Carlos Van Buren
Hospital, January 2000 to June 2008. Transplant Proc 2009;
41:26558
10. Quiroga I, McShane P, Koo DD, Gray D, Friend PJ, Fuggle

S, Darby C: Major effects of delayed graft function and cold
ischaemia time on renal allograft survival. Nephrol Dial
Transplant 2006; 21:168996
11. Lebranchu Y, Halimi JM, Bock A, Chapman J, Dussol B,
Fritsche L, Kliem V, Oppenheimer F, Pohanka E, Salvadori
M, Soergel M, Tufveson G; MOST International Study
Group: Delayed graft function: Risk factors, consequences
and parameters affecting outcome-results from MOST, A
Multinational Observational Study. Transplant Proc 2005;
37:3457
12. Halloran PF, Hunsicker LG: Delayed graft function: State of
the art, November 1011, 2000. Summit meeting, Scottsdale,
Arizona, USA. Am J Transplant 2001; 1:11520
13. Salim A, Martin M, Brown C, Belzberg H, Rhee P, Demetriades
D: Complications of brain death: Frequency and impact on
organ retrieval. Am Surg 2006; 72:37781
14. Saleh DA, Shebl FM, El-Kamary SS, Magder LS, Allam A,
Abdel-Hamid M, Mikhail N, Hashem M, Sharaf S, Stoszek
SK, Strickland GT: Incidence and risk factors for communityacquired hepatitis C infection from birth to 5 years of age
in rural Egyptian children. Trans R Soc Trop Med Hyg 2010;
104:35763
15. McIlroy DR, Wagener G, Lee HT: Biomarkers of acute kidney injury: An evolving domain. Anesthesiology 2010;
112:9981004
16. Mishra J, Ma Q, Prada A, Mitsnefes M, Zahedi K, Yang J,
Barasch J, Devarajan P: Identification of neutrophil gelatinase-associated lipocalin as a novel early urinary biomarker
for ischemic renal injury. J Am Soc Nephrol 2003; 14:253443
17. Singbartl K, Kellum JA: AKI in the ICU: Definition, epidemiology, risk stratification, and outcomes. Kidney Int 2012;
81:81925
18. Constantin JM, Futier E, Perbet S, Roszyk L, Lautrette A,
Gillart T, Guerin R, Jabaudon M, Souweine B, Bazin JE,
Sapin V: Plasma neutrophil gelatinase-associated lipocalin is
an early marker of acute kidney injury in adult critically ill
patients: A prospective study. J Crit Care 2010; 25:176.e16
19. Malyszko J, Malyszko JS, Bachorzewska-Gajewska H,
Poniatowski B, Dobrzycki S, Mysliwiec M: Neutrophil gelatinase-associated lipocalin is a new and sensitive marker of
kidney function in chronic kidney disease patients and renal
allograft recipients. Transplant Proc 2009; 41:15861
20. Hollmen ME, Kyllnen LE, Inkinen KA, Lalla ML, Salmela KT:
Urine neutrophil gelatinase-associated lipocalin is a marker
of graft recovery after kidney transplantation. Kidney Int
2011; 79:8998
21. Rostami Z, Nikpoor M, Einollahi B: Urinary neutrophil gelatinase associated lipocalin (NGAL) for early diagnosis of acute
kidney injury in renal transplant recipients. Nephrourol Mon
2013; 5:74552
22. Hollmen ME, Kyllnen LE, Inkinen KA, Lalla ML, Merenmies
J, Salmela KT: Deceased donor neutrophil gelatinase-associated lipocalin and delayed graft function after kidney transplantation: A prospective study. Crit Care 2011; 15:R121
23. Wijdicks EF, Varelas PN, Gronseth GS, Greer DM; American
Academy of Neurology: Evidence-based guideline update:
Determining brain death in adults: Report of the Quality
Standards Subcommittee of the American Academy of
Neurology. Neurology 2010; 74:19118
24. Vincent JL, Moreno R, Takala J, Willatts S, De Mendona
A, Bruining H, Reinhart CK, Suter PM, Thijs LG: The SOFA
(Sepsis-related Organ Failure Assessment) score to describe
organ dysfunction/failure. On behalf of the Working Group
on Sepsis-Related Problems of the European Society of
Intensive Care Medicine. Intensive Care Med 1996; 22:70710
25. Perico N, Cattaneo D, Sayegh MH, Remuzzi G: Delayed graft
function in kidney transplantation. Lancet 2004; 364:181427
26. Boutin C, Vachiry-Lahaye F, Alonso S, Louart G, Bouju A,
Lazarovici S, Perrigault PF, Capdevila X, Jaber S, Colson P,
Jonquet O, Ripart J, Lefrant JY, Muller L; pour le groupe

AzuRa: [Anaesthetic management of brain-dead for organ
donation: Impact on delayed graft function after kidney
transplantation]. Ann Fr Anesth Reanim 2012; 31:42736
27. Ray P, Le Manach Y, Riou B, Houle TT: Statistical evaluation
of a biomarker. Anesthesiology 2010; 112:102340
28. Vanmassenhove J, Vanholder R, Nagler E, Van Biesen W:
Urinary and serum biomarkers for the diagnosis of acute
kidney injury: An in-depth review of the literature. Nephrol
29. Mishra J, Dent C, Tarabishi R, Mitsnefes MM, Ma Q, Kelly
C, Ruff SM, Zahedi K, Shao M, Bean J, Mori K, Barasch J,
Devarajan P: Neutrophil gelatinase-associated lipocalin
(NGAL) as a biomarker for acute renal injury after cardiac
surgery. Lancet 2005; 365:12318
30. Cruz DN, de Cal M, Garzotto F, Perazella MA, Lentini P,
Corradi V, Piccinni P, Ronco C: Plasma neutrophil gelatinaseassociated lipocalin is an early biomarker for acute kidney
injury in an adult ICU population. Intensive Care Med 2010;
36:44451
31. de Geus HR, Bakker J, Lesaffre EM, le Noble JL: Neutrophil
gelatinase-associated lipocalin at ICU admission predicts for
acute kidney injury in adult patients. Am J Respir Crit Care
Med 2011; 183:90714
32. Di Somma S, Magrini L, De Berardinis B, Marino R, Ferri E,
Moscatelli P, Ballarino P, Carpinteri G, Noto P, Gliozzo B,
Paladino L, Di Stasio E: Additive value of blood neutrophil
gelatinase-associated lipocalin to clinical judgement in acute
kidney injury diagnosis and mortality prediction in patients
hospitalized from the emergency department. Crit Care 2013;
17:R29
33. Wagener G, Gubitosa G, Wang S, Borregaard N, Kim M, Lee
HT: Urinary neutrophil gelatinase-associated lipocalin and
acute kidney injury after cardiac surgery. Am J Kidney Dis
2008; 52:42533
34. Tuladhar SM, Pntmann VO, Soni M, Punjabi PP, Bogle RG:
Rapid detection of acute kidney injury by plasma and urinary neutrophil gelatinase-associated lipocalin after cardiopulmonary bypass. J Cardiovasc Pharmacol 2009; 53:2616
35. Koyner JL, Bennett MR, Worcester EM, Ma Q, Raman J,
Jeevanandam V, Kasza KE, OConnor MF, Konczal DJ, Trevino
S, Devarajan P, Murray PT: Urinary cystatin C as an early biomarker of acute kidney injury following adult cardiothoracic
surgery. Kidney Int 2008; 74:105969
36. Haase-Fielitz A, Bellomo R, Devarajan P, Story D, Matalanis
G, Dragun D, Haase M: Novel and conventional serum biomarkers predicting acute kidney injury in adult cardiac
surgeryA prospective cohort study. Crit Care Med 2009;
37:55360
37. Murugan R, Venkataraman R, Wahed AS, Elder M, Carter
M, Madden NJ, Kellum JA; HIDonOR Study Investigators:
Preload responsiveness is associated with increased interleukin-6 and lower organ yield from brain-dead donors. Crit
Care Med 2009; 37:238793
38. Malinoski DJ, Patel MS, Ahmed O, Daly MC, Mooney S,
Graybill CO, Foster CE, Salim A; United Network for Organ
Sharing (UNOS) Region 5 Donor Management Goals (DMG)
Workgroup: The impact of meeting donor management goals
on the development of delayed graft function in kidney
transplant recipients. Am J Transplant 2013; 13:9931000
39. Yang J, Mori K, Li JY, Barasch J: Iron, lipocalin, and kidney
epithelia. Am J Physiol Renal Physiol 2003; 285:F918
40. Kashani K, Al-Khafaji A, Ardiles T, Artigas A, Bagshaw
SM, Bell M, Bihorac A, Birkhahn R, Cely CM, Chawla LS,
Davison DL, Feldkamp T, Forni LG, Gong MN, Gunnerson KJ,
Haase M, Hackett J, Honore PM, Hoste EA, Joannes-Boyau
O, Joannidis M, Kim P, Koyner JL, Laskowitz DT, Lissauer
ME, Marx G, McCullough PA, Mullaney S, Ostermann M,
Rimmel T, Shapiro NI, Shaw AD, Shi J, Sprague AM, Vincent
JL, Vinsonneau C, Wagner L, Walker MG, Wilkerson RG,
Zacharowski K, Kellum JA: Discovery and validation of cell

cycle arrest biomarkers in human acute kidney injury. Crit
Care 2013; 17:R25
41. Moore J, Tan K, Cockwell P, Krishnan H, McPake D, Ready A,
Mellor S, Hamsho A, Ball S, Lipkin G, Borrows R: Predicting
early renal allograft function using clinical variables. Nephrol
42. Irish WD, Ilsley JN, Schnitzler MA, Feng S, Brennan DC:
A risk prediction model for delayed graft function in the
current era of deceased donor renal transplantation. Am J
Transplant 2010; 10:227986
Appendix. AzuRea Group: www.azurea.org

The AzuRea group is a research network in perioperative medicine.
In addition to authors who are affiliated to Azurea group, the following persons are members of this group and have participated to
the current study:
-Bernard Allaouchiche, M.D., Ph.D., Department of Anesthesiology and Critical Care, Edouard Herriot Hospital,
Hospices Civils de Lyon, Lyon, France.
-Xavier Capdevila, M.D., Ph.D., Department of Anesthesiology and Critical Care, CHU Lapeyronie, Montpellier, France.
-
Jean Michel Constantin, M.D., Ph.D., Department of
Anesthesiology and Critical Care, University Hospital of
Estaing, Clermont Ferrand, France.
-Olivier Fourcade, M.D., Ph.D., Department of Anesthesiology and Critical Care, University Hospital of Toulouse,
Toulouse, France.
-Carole Icha, M.D., Ph.D., Department of Anesthesiology and Critical Care, University Hospital of Nice, Nice,
France.
-Samir Jaber, M.D., Ph.D., Department of Anesthesiology
and Critical Care, CHU Saint Eloi, Montpellier, France.
-Marc Leone, M.D., Ph.D., Department of Anesthesiology
and Critical Care, Hopital Nord, University Hospital of
Marseille, Marseille, France.
-Serge Molliex, M.D., Ph.D., Department of Anesthesiology
and Critical Care, University Hospital of Saint Etienne, Saint
Etienne, France.
-Jerome Morel, M.D., Ph.D., Department of Anesthesiology and Critical Care, University Hospital of Saint Etienne,
Saint Etienne, France.
-Jean Franois Payen, M.D., Ph.D., Department of Anesthesiology and Critical Care, University Hospital of Grenoble,
Grenoble, France.
-Vincent Piriou, M.D., Ph.D., Department of Anesthesiology and Critical Care, University Hospital of Lyon Sud,
Hospices Civils de Lyon, Lyon, France.
Modulation of Stress versus Time Product during

Mechanical Ventilation Influences Inflammation as Well
as Alveolar Epithelial and Endothelial Response in Rats
Peter M. Spieth, M.D., Pedro L. Silva, Ph.D., Cristiane S. N. B. Garcia, Ph.D., Debora S. Ornellas, Ph.D.,
Cynthia S. Samary, Ph.D., Lillian Moraes, M.Sc., Maira Bentes, R.R.T., Marcelo M. Morales, M.D., Ph.D.,
Michael Kasper, Ph.D., Andreas Gldner, M.D., Robert Huhle, M.Sc., Thea Koch, M.D., Ph.D.,
Paolo Pelosi, M.D., Marcelo Gama de Abreu, M.D., Ph.D., Patricia R. M. Rocco, M.D., Ph.D.
ABSTRACT
Background: Mechanical ventilation can lead to lung biotrauma when mechanical stress exceeds safety thresholds. The
authors investigated whether the duration of mechanical stress, that is, the impact of a stress versus time product (STP),
influences biotrauma. The authors hypothesized that higher STP levels are associated with increased inflammation and with
alveolar epithelial and endothelial cell injury.
Methods: In 46 rats, Escherichia coli lipopolysaccharide (acute lung inflammation) or saline (control) was administered intratracheally. Both groups were protectively ventilated with inspiratory-to-expiratory ratios 1:2, 1:1, or 2:1 (n = 12 each), corresponding to low, middle, and high STP levels (STPlow, STPmid, and STPhigh, respectively). The remaining 10 animals were
not mechanically ventilated.
Results: In animals with mild acute lung inflammation, but not in controls: (1) messenger RNA expression of interleukin-6
was higher in STPhigh (28.113.6; mean SD) and STPlow (28.916.0) versus STPmid (7.47.5) (P < 0.05); (2) expression of
the receptor for advanced glycation end-products was increased in STPhigh (3.61.6) versus STPlow (2.31.1) (P < 0.05); (3)
alveolar edema was decreased in STPmid (0 [0 to 0]; median, Q1 to Q3) compared with STPhigh (0.8 [0.6 to 1]) (P < 0.05); and
(4) expressions of vascular cell adhesion molecule-1 and intercellular adhesion molecule-1 were higher in STPlow (3.01.8)
versus STPhigh (1.20.5) and STPmid (1.40.7) (P < 0.05), respectively.
Conclusions: In the mild acute lung inflammation model used herein, mechanical ventilation with inspiratory-to-expiratory
of 1:1 (STPmid) minimized lung damage, whereas STPhigh increased the gene expression of biological markers associated with
inflammation and alveolar epithelial cell injury and STPlow increased markers of endothelial cell damage. (Anesthesiology
2015; 122:106-16)
ENERAL anesthesia usually requires mechanical ventilation to maintain adequate gas exchange. However,
mechanical ventilation may impose stress on the lung parenchyma, triggering a proinflammatory response that damages
the lung, a phenomenon known as ventilator-induced lung
injury (VILI).1 In patients with the acute respiratory distress
syndrome, the use of protective mechanical ventilation with
tidal volumes (VT) of 4 to 8ml/kg and positive end-expiratory pressure (PEEP) has been shown to reduce lung inflammation and mortality.2
Most surgical patients undergoing general anesthesia differ from intensive care patients in terms of the absence of
lung injury. However, because inflammatory mediators can
be released into the circulation during surgery, making lungs

Mechanical ventilation can lead to lung biotrauma when the
mechanical stress caused by the ventilation exceeds safety
thresholds

In a mild acute lung inflammation model in rats, using
mechanical ventilation with an inspiratory-to-expiratory ratio
of 1:1 minimized lung damage, whereas an inspiratory-
to-expiratory ratio of 2:1 led to increased gene expression of inflammatory mediators and markers of alveolar epithelial cell injury
vulnerable to VILI,3 the use of protective mechanical ventilation has been recommended during general anesthesia.4
anesthesiology.org). The first two authors contributed equally to this work.
Submitted for publication December 18, 2013. Accepted for publication July 8, 2014. From the Pulmonary Engineering Group, Department of Anesthesiology and Intensive Care Therapy, University Hospital Carl Gustav Carus, Dresden, Technische Universitt Dresden, Dresden, Germany (P.M.S.,
A.G., R.H., T.K., M.G.d.A.); Laboratory of Pulmonary Investigation (P.L.S., C.S.N.B.G., D.S.O., C.S.S., L.M., M.B., P.R.M.R.), and Laboratory of Cellular and
Molecular Biology (M.M.M.), Carlos Chagas Filho Biophysics Institute, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil; Rio de Janeiro Federal Institute of Education, Science and Technology, Rio de Janeiro, Brazil (C.S.N.B.G.); Institute of Anatomy, Technical University Dresden, Dresden,
Germany (M.K.); and IRCCS AOU San Martino-IST, Department of Surgical Sciences and Integrated Diagnostics, University of Genoa, Genoa, Italy (P.P.).
January 2015
Mechanical stress of the lung parenchyma is a major factor determining VILI.1 In lungs, stress is proportional to
strain, that is, proportional to the change in lung volume
at end-inspiration in relation to the resting volume at barometric pressure and to the specific lung elastance.5 Some
investigators claim that as long as stress is kept below a certain threshold, lung injury might be avoided.5 It has been
shown that not only the magnitude but also the duration of
stress triggers an inflammatory genomic response in cultured
primary alveolar epithelial cells.6 This suggests that even
below what is considered a safety threshold for magnitude of
stress, proinflammatory and profibrotic responses may still
be triggered in lungs if stress is applied for longer periods of
timeas a result of a phenomenon that can be called stress
versus time product (STP). The use of the term STP may be
advantageous over pressure versus time product, because it
applies to both tensile and compressive forces that result in a
biological response. Furthermore, the term STP can be used
beyond the setting of mechanical ventilationfor example,
in reference to cell preparations subjected to stretching. To
the best of our knowledge, however, the biological impact of
STP on lung tissue in vivo has not yet been characterized.
In the current study, we investigated the effects of STP
on gas exchange, respiratory system mechanics, biological markers of inflammation, fibrosis, apoptosis, as well as
alveolar epithelial and endothelial damage in mechanically
ventilated rats with intratracheal lipopolysaccharide (acute
lung inflammation) or saline (control) administration. Modulation of STP was accomplished by using different inspiratory-to-expiratory (I:E) ratios, with I:E = 1:2, 1:1, and 2:1
corresponding to low, middle, and high STP levels (STPlow,
STPmid, and STPhigh, respectively). In the mild acute lung
inflammation model used, but not in controls, we expected
that STP might modulate the alveolar epithelial and endothelial biological responses without major impact on lung
function. Accordingly, our primary hypothesis, which was
used for sample size calculation, was that STPhigh would
lead to increased gene expression of interleukin-6 (IL-6) in
lung tissue compared with STPmid and STPlow. Our secondary hypothesis was that STPhigh, compared with STPmid and
STPlow, would increase the gene expression of markers of
fibrosis, apoptosis, and mechanical stress, as well as lung tissue damage.

This study was approved by the Health Sciences Center Ethics Committee (CEUA-CCS, 019) at the Federal University
of Rio de Janeiro, Rio de Janeiro, Brazil. Animals received
humane care in compliance with the Principles of Laboratory Animal Care formulated by the National Society for
Medical Research and the Guide for the Care and Use of
Laboratory Animals prepared by the U.S. National Academy of Sciences.
Animal Preparation and Experimental Protocol

Forty-six male Wistar rats (300 to 360g) were used for this
study. Animals were anesthetized with sevoflurane (2.5 Vol%)
and subjected to intratracheal instillation of Escherichia coli
lipopolysaccharide (Serotype 055:B5; Sigma Aldrich, St.
Louis, MO; 0.55 to 0.66mg/kg suspended in 0.9% saline
with total volume equal to 20 l, n = 18) or saline (20 l,
n = 18). After recovering from anesthesia, animals remained
under close observation.
Twenty-four hours after induction of acute lung inflammation, animals were anesthetized by intraperitoneal injection of ketamine (100mg/kg) and midazolam (5mg/kg).
An intravenous line (24 gauge) was placed in the tail vein
and anesthesia was maintained intravenously with ketamine
(50mg kg1 h1) and midazolam (2.5mg kg1 h1). Animals
were placed and kept in supine position throughout the
whole experiment. After median neck incision, a polyethylene catheter (PE 50) was introduced into the right internal
carotid artery for blood sampling and mean arterial blood
pressure measurement. Heart rate, mean arterial blood pressure, and rectal temperature were continuously recorded
(Networked Multiparameter Veterinary Monitor LifeWindow 6000V; Digicare Animal Health, Boynton Beach, FL).
Body temperature was maintained at 37.5 1C using a
heating blanket. Administration of lactated ringer solution
(8ml kg1 h1) was performed via the tail vein to keep fluid
homeostasis and mean arterial blood pressure higher than 70
mmHg. A 14-gauge cannula was used for tracheostomy and
a fluid-filled tube was inserted in the esophagus for esophageal pressure measurements. Muscle paralysis was achieved
with continuous infusion of pancuronium, 1mg kg1 h1.
Lungs were mechanically ventilated (Servo-I; MAQUET,
Solna, Sweden) in pressure-controlled ventilation mode with
inspiratory driving pressure to reach tidal volumes (VT) of
6ml/kg, fraction of inspired oxygen (Fio2) of 0.4, and PEEP
of 5cm H2O. Respiratory rate was set at 70 breaths/min and
adjusted to arterial pH of 7.35 to 7.45. The I:E time ratio
was set to 1:1. After the onset of mechanical ventilation, a
recruitment maneuver (30cm H2O constant inflation for 15
s) was performed in all animals to achieve similar baseline
conditions. Thirty-six animals were then randomly assigned
to low, middle, and high STP levels (STPlow, STPmid, and
STPhigh) for 2h (n = 12 each). Modulation of STP was
accomplished by changing I:E ratios to 1:2, 1:1, and 2:1,
corresponding to STPlow, STPmid, and STPhigh, respectively.
At the end of this experimental protocol, animals were sacrificed by exsanguination and the lungs were removed for
further analysis. The remaining 10 animals (n = 5 each in the
acute lung inflammation and control groups) did not receive
mechanical ventilation and were used for measurement of
wet-to-dry ratio and molecular biology analysis only. Blood
(300 l) was drawn into a heparinized syringe to determine
arterial partial pressure of oxygen, arterial partial pressure of
carbon dioxide, and arterial pH (i-STAT; Abbott Laboratories, Chicago, IL).
Anesthesiology 2015; 122:106-16 107 Spieth et al.
Stress vs. Time Product and Biotrauma
Data Acquisition and Processing

Airflow, tracheal, and esophageal pressures were measured.
Airway pressure (Paw) was measured with a SCIREQ differential pressure transducer (UT-PDP-300; SCIREQ, Montreal, Canada). Changes in esophageal pressure (Pes), which
reflect changes in chest wall pressure, were measured with a
30-cm long water-filled catheter (PE205) with side holes at
the tip connected to a SCIREQ differential pressure transducer (UT-PL-400; SCIREQ). The catheter was passed into
the stomach and then slowly returned into the esophagus.
Its proper positioning was assessed using the occlusion test.
In brief, this method consists of comparing the variation
between Pes and Paw during spontaneous inspiratory efforts
made against a closed airway. The presence of comparable
changes in Pes and Paw (difference lower than 5%, phase
angle close to nil) indicates that the changes in Pes accurately
reflect changes in pleural pressure. Airflow and tracheal and
esophageal pressures were continuously recorded using LabView-based software (National Instruments, Austin, TX).7
Transpulmonary pressures were calculated during inspiration
and expiration as the difference between airway and esophageal pressures. All signals were filtered (100 Hz), amplified
in a four-channel conditioner (SC-24; SCIREQ), and sampled at 200 Hz with a 12-bit analogue to digital converter
(NI-DAQmx 8.7.1; National Instruments). Peak and mean
airway pressures, as well as peak and mean transpulmonary
pressures (Paw,p, Paw,m, PL,p, and PL,m, respectively), were computed. The work of breathing performed by the ventilator
was calculated using the area enclosed by the inspiratory and
expiratory part of the airway pressure volume loop. Also, we
calculated the pressuretime product using airway pressure
(PTPaw) and transpulmonary pressure (PTPL). The elastance
(Ers) and resistance (Rrs) of the respiratory system were calculated using the equation of motion.7 All respiratory variables
were computed from continuous recordings (5min) of Paw,
Pes, and airflow by routines written in MATLAB (Version
7.14; The Mathworks, Natick, MA). All functional measurements were obtained at baseline and at 1 and 2h.
Postmortem Analysis
A laparotomy was performed immediately after blood sampling at the end of experiments and heparin (1,000 IU) was
injected intravenously in the caval vein. The trachea was
clamped at end-expiration at a continuous airway pressure
of 5cm H2O, and the abdominal aorta and vena cava were
sectioned to quickly sacrifice the animal by exsanguination.
End-expiratory Lung Volume
Lungs were removed en bloc and the end-expiratory lung
volume was measured as described elsewhere.8 In brief, a jar
containing sufficient 0.9% saline with a submerged surplus
weight was placed on a common laboratory scale, which was
subsequently set to zero. The lungs were fixed to a laboratory
stand by means of a thread with the surplus weight and completely submerged in saline 0.9%. The liquid displaced by
the submerged lungs corresponds to the weight on the scale.

Because the specific gravity of saline 0.9% differs no more
than 2 to 3% from 1g/cm3, the volume of the organ may be
expressed directly by the weight gain registered on the scale.
Lung Histology
The right lower lung lobe was fixed in 4% buffered formaldehyde solution, paraffin embedded, cut in slices of 4-m
thickness, and stained with hematoxylineosin. Photomicrographs at magnifications of 25, 100, and 400 were
obtained from four nonoverlapping fields of view per section
using a light microscope. Diffuse alveolar damage (DAD)
was quantified using a weighted scoring system by an expert
in lung pathology (M.K.) blinded to the experimental protocol, as described elsewhere.9 In brief, values from 0 to 4 were
used to represent the severity of alveolar edema, hemorrhage,
inflammatory infiltration, and alveolar overdistension, with
0 standing for no effect and 4 for maximum severity. In addition, the extent of each score characteristic per field of view
was determined with values from 0 to 4, with 0 standing for
no appearance and 4 for complete involvement. Scores were
calculated as the product of severity and extent of each feature, being situated in the range 0 to 16. Cumulated DAD
score was calculated as sum of single score characteristics
yielding score values from 0 to 64.
Lung Wet-to-dry Weight Ratio
The wet-to-dry ratio was determined in the right middle
lobe as described elsewhere.10 In brief, the right middle lobe
was separated, weighed (wet weight), and then dried in a
microwave at low power (200W) for 5min. The drying process was repeated until the difference between two consecutive lung weight measurements was less than 0.002g. The
last weight measured represented the dry weight.
Markers of Inflammatory, Alveolar Epithelial and
Endothelial Cell Injury, and Fibrogenesis
Quantitative real-time reverse-transcription polymerase
chain reaction was performed to measure the messenger
RNA (mRNA) expression of IL-6, procaspase-3, receptor for advanced glycation end-products (RAGE), surfactant protein-B, type III procollagen, intercellular adhesion
molecule (ICAM)-1, and vascular cell adhesion molecule
(VCAM)-1. Glycerylaldehyde-3-phosphate dehydrogenase
was used as a housekeeping gene. Central slices of the left
lung were cut, collected in cryotubes, quick-frozen by immersion in liquid nitrogen, and stored at 80C. Total RNA was
extracted using the spin or vacuum total RNA isolation
system (Promega, Fitchburg, WI). RNA concentration was
measured by spectrophotometry in Nanodrop ND-1000
(Wilmington, DE). First-strand complementary DNA was
synthesized from total RNA using M-MLV Reverse Transcriptase Kit (Invitrogen, Carlsbad, CA). The primers used are
shown in the Supplemental Digital Content 1, http://links.
lww.com/ALN/B83. Relative mRNA levels were measured
with a SYBR green detection system using ABI 7500 realtime polymerase chain reaction (Applied Biosystems, Foster City, CA). Samples were measured in triplicate. Relative
gene expression was calculated as a ratio of the average gene
expression levels compared with the reference gene (glycerylaldehyde-3-phosphate dehydrogenase) and expressed as fold
change relative to nonmechanically ventilated animals.
Statistical Analyses
The sample size calculation for testing the primary hypothesis (the gene expression of IL-6 in lung tissue is increased
with STPhigh compared with STPlow and STPmid in acute
lung inflammation) was based on effects estimates obtained
from pilot studies as well as on previous measurements made
by us (mean value and dispersion, respectively). Accordingly,
we expected that a sample size of six animals per level of STP
would provide the appropriate power (1- = 0.8) to identify significant ( = 0.05) differences in IL-6 gene expression, considering an effect size d = 2.2, two-sided test, and
multiple comparisons (n = 3) (* = 0.0167, * Bonferroni
adjusted).
All data are expressed as mean and SD. The primary and
secondary hypotheses were tested with one-way and twoway ANOVAs, as appropriate (Prism for Mac, Version 5.0a;
GraphPad Software, La Jolla, CA). DAD scores are shown as
median and interquartile range. Statistics were calculated by
KruskallWallis tests. Adjustments for multiple comparisons
were performed according to Bonferroni. Stepwise curve-fit
regression analyses using linear, logarithmic, quadratic, and
exponential functions with PTPL as the nondependent and
morphological and molecular biology data as the dependent
variables were performed using IBM SPSS Statistics 20.0
(IBM Corp., Armonk, NY). The global significance level for
all tests was P value less than 0.05.
achieving the highest value in STPhigh and the lowest value

in STPlow.
Histological evaluation of DAD score characteristics
revealed increased alveolar edema in STPhigh compared with
STPmid (fig.3). Other features as well as cumulated DAD
score did not differ statistically among STP levels.
Gene expression of IL-6 was higher with STPlow and
STPhigh than STPmid, whereas mRNA expression of procaspase-3 was comparable among STP levels (fig.4). The
mRNA expression of RAGE was higher in STPhigh than
STPmid, whereas mRNA levels of surfactant protein-B did
not differ significantly among STP levels. The analysis of
gene expression of type III procollagen showed no statistically significant differences among STP settings. STPlow
yielded increased mRNA expression of VCAM-1 compared
with STPhigh, and of ICAM-1 compared with STPmid.
PTPL was significantly associated with cumulated DAD
score, inflammatory infiltration, and mRNA expression of
VCAM-1 (table3).
Data obtained in control animals are shown in Supplemental Digital Content 2, http://links.lww.com/ALN/B84, and
Supplemental Digital Content 3, http://links.lww.com/ALN/
B85. PTPaw was higher in STPhigh compared with other STPs
(see table 1, Supplemental Digital Content 2, http://links.
lww.com/ALN/B84). No significant differences were observed
among levels of STP in all other functional (see figs. 13,
Supplemental Digital Content 3, http://links.lww.com/ALN/
B85), morphological (see table 2, Supplemental Digital Content 2, http://links.lww.com/ALN/B84), and biological variables (see fig. 4, Supplemental Digital Content 3, http://links.
lww.com/ALN/B85). Additional data showing the significant
regressions between PTPL and morphological and molecular
biology in controls and in controls pooled with lipopolysaccharide-treated animals are reported in figure 5, Supplemental
Results
Discussion
All animals survived the intratracheal administration of lipopolysaccharide and saline, as well as the 24-h period that
followed.
In animals subjected to mild acute lung inflammation,
heart rate, mean arterial blood pressure, and the amount
of fluid given did not differ significantly among groups
(table1). The same was true for gas exchange, end-expiratory
lung volume, and lung wet-to-dry ratio (fig.1).
Respiratory variables are shown in figure2. Paw,m was
higher in STPhigh compared with other STPs, whereas Paw,p
did not differ among STP levels. PL,p was also similar among
the groups, but PL,m was higher with STPhigh than STPlow.
Intrinsic PEEP, peak flow values, volume-independent elastance (%E1), or lobe Ers did not differ among STP levels.
As depicted in table2, the work of breathing performed
by the ventilator was comparable among STP levels. The
PTPaw was higher in STPhigh compared with STPlow and
STPmid. The PTPL differed significantly among STP levels,
The main finding of this study was that in animals with

mild acute lung inflammation, but not in controls, (1) gene
expression of IL-6 was higher in STPhigh and STPlow versus
STPmid; (2) expression of RAGEs was increased in STPhigh
versus STPlow; (3) alveolar edema was decreased in STPlow
compared with STPhigh; and (4) expressions of VCAM-1
and ICAM-1 were higher in STPlow, compared with STPhigh
and STPmid, respectively. Therefore, our data suggest that
mechanical ventilation with I:E = 1:1, compared with
I:E = 1:2 and 2:1, minimized VILI in the model of mild
acute lung inflammation used in this investigation.
To assess the effects of STP on acute lung inflammation,
we chose a model with intratracheal instillation of E. coli
lipopolysaccharide with lung mechanical and histological
impairment.11,12 Animals were ventilated with a protective
ventilatory strategy to rule out the possible effects of higher
VT, whereas the I:E ratio was modulated to accomplish different levels of STP. During mechanical ventilation, stress/
Table 1. Hemodynamics and Fluids Infused in Acute Lung Inflammation Animals
MAP, mmHg
HR, min1
Fluids infused, ml
STPlow
STPmid
STPhigh
STPlow
STPmid
STPhigh
STPlow
STPmid
STPhigh
Baseline
1h MV
2h MV
8516
977
9714
38846
42828
38825
8626
844
7916
39251
41134
36343
738
827
8515
39346
40734
38137
5.71.9
5.91.4
5.20.9
Data are presented as mean SD. Comparisons among groups were performed by means of two-way ANOVA (HR and MAP) and one-way ANOVA (fluids
infused). Fluids infused represent the total volume of fluids infused during the experiment. There were no significant differences among groups.
HR = heart rate; I:E = inspiratory-to-expiratory; MAP = mean arterial pressure; MV = mechanical ventilation; STPhigh = high level of stress vs. time product
(I:E 2:1); STPlow = low level of stress vs. time product (I:E 1:2); STPmid = middle level of stress vs. time product (I:E 1:1).
Fig. 1. Gas exchange, lung volume, and wet-to-dry ratio in acute lung inflammation animals. Data are presented as mean SD.
Comparisons among groups were performed using two-way ANOVA (A, fraction of arterial partial pressure of oxygen and inspired oxygen fraction [Pao2/Fio2]; B, arterial partial pressure of carbon dioxide [Paco2]) and one-way ANOVA (C, end-expiratory
lung volume [EELV]; D, lung wet-to-dry ratio). I:E = inspiratory-to-expiratory; MV= mechanical ventilation; NV = nonventilated;
STPhigh = high levels of stress versus time product (I:E 2:1); STPlow = low level of stress versus time product (I:E 1:2); STPmid =
middle level of stress versus time product (I:E 1:1).
Fig. 2. Respiratory mechanics in acute lung inflammation animals. Data are presented as mean SD. Comparisons among
groups were performed using two-way ANOVA. (A) Peak airway pressure (Paw,p); (B) mean airway pressure (Paw,m); (C) peak
transpulmonary pressure (PL,p); (D) mean transpulmonary pressure (PL,m); (E) elastance of the respiratory system (Ers); (F) resistance of the respiratory system (Rrs). I:E = inspiratory-to-expiratory; MV = mechanical ventilation; STPhigh = high levels of stress
versus time product (I:E 2:1); STPlow = low level of stress versus time product (I:E 1:2); STPmid = middle level of stress versus time
product (I:E 1:1).
strain of lung tissue occurs at all phases of the respiratory

cycle. However, inspiration represents the phase of the respiratory cycle that exposes the lung tissue to higher stress. We
chose to modulate STP by using I:E = 1:2, 1:1, and 2:1 in
a pressure-controlled ventilation mode for the following
reasons. First, the airflow magnitude and profile during
pressure-controlled ventilation is similar across different I:E
settings and affects VILI.13 Second, we aimed at maintaining
the expiratory strain in the lungsthat is, end-expiratory
lung volumecomparable across the groups, whereas modulating only the total amount of inspiratory stress.
Interleukin-6 and procaspase-3 mRNA expressions were

analyzed because of their role as mediators of inflammation and/
or apoptosis, respectively, in the pathogenesis of VILI.14 The
lung expression of type III procollagen was evaluated because
it is the first collagen to be remodeled in the development of
fibrogenesis, and also an early marker of lung parenchyma
remodeling.15 RAGE and surfactant protein-B mRNA were
chosen because they closely reflect alveolar types I and II cell
injury respectively.16,17 In the pulmonary endothelium, the gene
expression of the adhesion molecules VCAM-1 and ICAM-1
is increased during stress induced by mechanical ventilation.18
Table 2. WOB Performed by the Ventilator and PressureTime Products in Acute Lung Inflammation Animals
WOB, cm H2O ml
PTPaw, cm H2Os
PTPL, cm H2Os
Baseline
1h MV
2h MV
Group Effects
STPlow
STPmid
STPhigh
STPlow
STPmid
STPhigh
STPlow
7.21.2
6.51.8
6.51.5
7.41.3
6.81.7
6.92.0
2.51.0
7.41.7
6.21.1
7.10.9
6.81.5
6.91.1
10.11.8
1.80.8
8.31.1
6.41.3
7.41.8
5.91.7
6.91.4
9.51.8
2.10.6
ns
STPmid
STPhigh
2.10.4
1.90.5
2.30.5
3.21.1
2.60.6
3.70.8
P < 0.001 STPlow vs. STPhigh

P < 0.001 STPmid vs. STPhigh
P < 0.05 STPlow vs. STPmid
P <0 .001 STPlow vs. STPhigh
P < 0.001 STPmid vs. STPhigh
Data are given as mean SD. Comparisons among groups were performed by means of two-way ANOVA on ranks, to account for unequal variances
between groups, with P value adjustment according to the BonferroniHolm method. Mechanical ventilation at baseline was performed with an I:E ratio of
1:1 in all groups.
I:E = inspiratory-to-expiratory; MV = mechanical ventilation; ns = not significant; PTPaw = airway pressure vs. time product; PTPL = transpulmonary pressure
vs. time product; STPhigh = high level of stress vs. time product (I:E = 2:1); STPlow = low level of stress vs. time product (I:E = 1:2); STPmid = middle level of
stress vs. time product (I:E = 1:1); WOB = work of breathing performed by the ventilator.
In both control animals and in those subjected to mild

acute lung inflammation, there was an association between
STP and Paw, as reflected by higher Paw,m at STPhigh and
STPmid compared with STPlow. In mild acute lung inflammation, PTPL was highest during STPhigh and lowest during
STPlow, suggesting that STP could be effectively modulated,
whereas the work of breathing performed by the ventilator
was roughly constant. However, those findings were not
accompanied by changes in oxygenation, Ers and end-expiratory lung volume, suggesting that higher Paw,m did not yield
lung recruitment. Furthermore, Paw,p was comparable among
groups. Taken together, those data seem to indicate that the
strain achieved during inspiration did not differ significantly
at the different investigated STP levels. Also, the peak airflow and the inspiratory airflow profile itself were comparable among groups, suggesting that alveolar epithelial and
endothelial responses were determined mainly by STP.
Previous studies indicate that IL-6, a proinflammatory
cytokine, contributes to VILI.19,20 In the current study, in
mild acute lung inflammation, the increase in IL-6 expression with STPhigh and STPlow may be related to mechanotransduction in lung tissue by increased distortion of the
alveolar-capillary barrier during inspiration and expiration
respectively.21
STPhigh increased the gene expression of the marker of
alveolar type I cell injury (RAGE) compared with STPmid.
It has been demonstrated that increased mechanical stress,
as achieved by proportionally high strain during inspiration,
may result in VILI.22 The current work adds to the previous
knowledge that also the duration of the intratidal stress may
play a role in the biological impact of mechanical ventilation. In line with our findings, Broccard et al.,23 using an
isolated perfused rabbit lung model, showed that cumulative
stress, as measured by airway plateau pressure, contributed
more to VILI than did the magnitude of dynamic stress per
se, represented by PL,p.
In mild acute lung inflammation, the markers of
mechanical stress in the alveolar endothelium were increased
with STPlow as compared with both STPmid (ICAM-1) and

STPhigh (VCAM-1). There are different possible explanations
for this apparent discrepancy: (1) although the straining of
the pulmonary endothelium correlates with the straining of
the alveolar epithelium, the magnitude is not identical.24
Because the endothelial layer is surrounded by alveoli, the
expansion of the epithelial layer should be more pronounced
than that of the endothelium. In fact, excessive alveolar
expansion can even decrease the radius of pulmonary capillaries, therefore decreasing the straining of the endothelium during inspiration; and (2) because the duration of
expiration was longer in the STPlow group than in the other
groups, longer time periods may have changed fluid shear
stress, increasing the stimuli to endothelial cells.25
Besides fluid shear stress, the intensity of pulmonary perfusion has been identified as an important determinant of
VILI in isolated rabbit lungs.26 We previously demonstrated
that redistribution of perfusion from dorsal to ventral lung
zones is accompanied by decreased damage in a saline lavage
model of acute respiratory distress syndrome.27,28 Also, we
showed that hypervolemia worsens lung damage in a model
of sepsis-induced acute respiratory distress syndrome.21
It could be suggested that modulation of I:E may yield
changes in pulmonary blood flow, thus altering lung injury
heterogeneity. However, if changes in pulmonary blood flow
occurred, then we would expect differences in gas exchange
and hemodynamics, which were not observed. Furthermore,
the DAD was comparable among groups, whereas the gene
expression of markers of inflammation, as well as alveolar
epithelial and endothelial cell damage did differ according to
I:E, suggesting that lung injury heterogeneity is not the main
mechanism explaining our findings. In fact, the new experiments in controls, with intratracheal instillation of saline,
showed that modulation of I:E did not affect gas exchange
and DAD. Taking those facts into account, the biological
impact of I:E modulation is more likely determined by STP.
The concept of stress and strain in the lungs has been
recently reviewed.29 There are numerous factors contributing
Fig. 3. Diffuse alveolar damage (DAD) score in acute lung inflammation animals. Data are presented as median (interquartile range)
as well as minima and maxima, respectively. Statistical analysis was performed using KruskallWallis tests. DAD score ranged
from 016 points. Score values were determined by multiplying characteristic extent by characteristic severity. The cumulated DAD
score was calculated by adding the scores obtained for each individual characteristic and ranged from 0 to 64. I:E = inspiratory-toexpiratory; lipopolysaccharide-NV = nonventilated animals with acute lung inflammation; STPhigh = high levels of stress versus time
product (I:E 2:1); STPlow = low level of stress versus time product (I:E 1:2); STPmid = middle level of stress versus time product (I:E 1:1).
to the development of VILI, namely increased plateau pressures, tidal volumes, patientventilator asynchrony, high
respiratory rate, or airway flow.30 In a porcine model of
VILI, Protti et al.5 reported that VILI developed only when
reaching a strain 1.5 to 2. Recently, the same group showed
that in addition to global strain, static strain and strain components also contributed to VILI.31 Another key factor to
understand VILI development is that even if global lung
mechanics is not indicative of lung stress and strain, the
regional distribution of stress and strain may easily overcome
potentially injurious levels in the presence of lung inhomogeneity.32 This is likely a reason why we did not observe major
differences among groups regarding Ers or plateau pressures,
while having clear evidence of differences in mRNA expression of cytokines favoring STPmid.
Possible Clinical Implications
The current study was designed to develop a model of
mild acute lung inflammation (first hit) with preserved
lung function. Therefore, our data are important for better
understanding the impact of STP on lung morphological

and biological responses. Even though an I:E ratio equal to
1:2 is commonly used during protective mechanical ventilation, our data suggest that, in the presence of mild acute
lung inflammation, an I:E of 1:1 may be appropriate. Conversely, in healthy lungs, when a first inflammatory hit is not
present, the modulation of I:E does not seem to play a major
role in terms of lung protection.
Limitations
This study has several limitations that should be addressed:
(1) we were not able to measure strain, that is, the ratio
between tidal volume and aerated lung volume at endexpiration. However, the analysis of aerated area may be
regarded as a surrogate of aerated lung volume; (2) although
global lung stress was not directly measured, transpulmonary
pressure represents the global stress exerted on the lungs,29
and PTPL was probably a valid surrogate of STP; (3) a
model of mild acute lung inflammation induced by intratracheal endotoxin instillation was used. Theoretically, the
Fig. 4. Gene expression in acute lung inflammation animals. Data are presented as mean SD. Comparisons among groups
were performed using one-way ANOVA. Messenger RNA (mRNA) expressions of genes are normalized to the respective housekeeping gene (glycerylaldehyde-3-phosphate dehydrogenase). (A) interleukin-6 (IL-6); (B) caspase 3; (C) receptor of advanced
glycation end-products (RAGE); (D) surfactant protein B (SP-B); (E) type III procollagen (PC III); (F) vascular cell adhesion molecule-1 (VCAM-1); (G) intercellular cell adhesion molecule-1 (ICAM-1). CT = threshold cycle; I:E = inspiratory-to-expiratory;
lipopolysaccharide-NV = nonventilated animals with acute lung inflammation; MV = mechanical ventilation; STPhigh = high levels
of stress versus time product (I:E 2:1); STPlow = low level of stress versus time product (I:E 1:2); STPmid = middle level of stress
versus time product (I:E 1:1).
Table 3. Association Analysis between Transpulmonary PressureTime Product and Postmortem Analysis in Lipopolysaccharidetreated Animals
Cumulated DAD score

Inflammatory infiltration
VCAM-1
Curve Fit
R2
P Value
Quadratic
Quadratic
Logarithmic
0.631
0.596
0.502
0.398
0.355
0.252
0.022
0.037
0.034
Significant results of the curve-fit regression analysis conducted with transpulmonary pressuretime product as nondependent variable and all postmortem
measurements as dependent variable in animals with lipopolysaccharide-induced acute lung inflammation. Curve fits were calculated for linear, logarithmic,
quadratic, and exponential functions.
DAD = diffuse alveolar damage; VCAM-1 = vascular cell adhesion molecule-1.
intravenous administration of lipopolysaccharide could have

also been used, but, in our experience, this leads to hemodynamic impairment, which may introduce confounding factors in the molecular biology analysis. Thus, our results may
not be extrapolated to other experimental models; (4) a fixed
PEEP level was applied and thus we cannot rule out that
different results may be obtained at higher PEEP levels; and
(5) the observational time was restricted to 2h, and thus the
expression of mediators was quantified using real-time polymerase chain reaction instead of enzyme-linked immunosorbent assay.11 Furthermore, larger observation times may lead
to increased mechanical stretch and therefore greater biological impact on the epithelium and endothelium.
In conclusion, in the mild acute lung inflammation model
tested in this study, mechanical ventilation with I:E = 1:1
(STPmid) minimized lung damage, whereas STPhigh increased
the gene expression of biological markers associated with
inflammation and alveolar epithelial cell injury and STPlow
increased markers of endothelial cell damage.
Acknowledgments
The authors thank Andre da Silva, B.S. (Laboratory of Pulmonary Investigation, Carlos Chagas Filho Biophysics Institute, Federal University of Rio de Janeiro, Rio de Janeiro,
Brazil), for animal care, Ana Lucia da Silva, B.S. (Laboratory of Pulmonary Investigation, Carlos Chagas Filho Biophysics Institute, Federal University of Rio de Janeiro), for
her help with microscopy, and Claudia Buchweitz, Ph.D.,
Filippe Vasconcellos, B.A., and Moira Schttler, B.A. (Rio
de Janeiro, Brazil), for their assistance in editing the article.
The authors are indebted to Hannes Krause, M.S. (Pulmonary Engineering Group, Department of Anesthesiology and
Intensive Care Medicine, University Hospital Dresden, Dresden University of Technology, Dresden, Germany), for his
assistance with tables and figures.
Supported by the Centers of Excellence Program (PRONEXFAPERJ; Rio de Janeiro, Brazil), Brazilian Council for Scientific
and Technological Development (CNPq; Brasilia, Brazil), Rio
de Janeiro State Research Supporting Foundation (FAPERJ; Rio
de Janeiro, Brazil), So Paulo State Research Supporting Foundation (FAPESP; So Paulo, Brazil), National Institute of Science and Technology of Drugs and Medicine (INCT-INOFAR;
Brasilia, Brazil), Coordination for the Improvement of Higher
Level Personnel (CAPES; Brasilia, Brazil), German Academic
Exchange Service (DAAD; Bonn, Germany), and departmental
funds. MAQUET (Solna, Sweden) provided technical support.
Competing Interests
Correspondence
Address correspondence to Dr. Rocco: Laboratory of Pulmonary Investigation, Carlos Chagas Filho Biophysics Institute,
Federal University of Rio de Janeiro, Rio de Janeiro, Brazil.
prmrocco@gmail.com. Information on purchasing reprints
may be found at www.anesthesiology.org or on the masthead page at the beginning of this issue. Anesthesiologys articles are made freely accessible to all readers, for personal
use only, 6 months from the cover date of the issue.
References
1. Tremblay LN, Slutsky AS: Ventilator-induced lung injury:
From the bench to the bedside. Intensive Care Med 2006;
32:2433
2. The Acute Respiratory Distress Syndrome Network:
Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000; 342:13018
3. Wolthuis EK, Choi G, Dessing MC, Bresser P, Lutter R, Dzoljic
M, van der Poll T, Vroom MB, Hollmann M, Schultz MJ:
Mechanical ventilation with lower tidal volumes and positive
end-expiratory pressure prevents pulmonary inflammation
in patients without preexisting lung injury. Anesthesiology
2008; 108:4654
4. Determann RM, Royakkers A, Wolthuis EK, Vlaar AP, Choi
G, Paulus F, Hofstra JJ, de Graaff MJ, Korevaar JC, Schultz
MJ: Ventilation with lower tidal volumes as compared with
conventional tidal volumes for patients without acute lung
injury: A preventive randomized controlled trial. Crit Care
2010; 14:R1
5. Protti A, Cressoni M, Santini A, Langer T, Mietto C, Febres
D, Chierichetti M, Coppola S, Conte G, Gatti S, Leopardi O,
Masson S, Lombardi L, Lazzerini M, Rampoldi E, Cadringher
P, Gattinoni L: Lung stress and strain during mechanical ventilation: Any safe threshold? Am J Respir Crit Care Med 2011;
183:135462
6. Yerrapureddy A, Tobias J, Margulies SS: Cyclic stretch magnitude and duration affect rat alveolar epithelial gene expression. Cell Physiol Biochem 2010; 25:11322
7. Gama de Abreu M, Quelhas AD, Spieth P, Brauer G, Knels
L, Kasper M, Pino AV, Bleyl JU, Hubler M, Bozza F, Salluh J,
Kuhlisch E, Giannella-Neto A, Koch T: Comparative effects
of vaporized perfluorohexane and partial liquid ventilation in oleic acid-induced lung injury. Anesthesiology 2006;
104:27889
8. Scherle W: A simple method for volumetry of organs in quantitative stereology. Mikroskopie 1970; 26:5760
9. Spieth PM, Knels L, Kasper M, Domingues Quelhas A,
Wiedemann B, Lupp A, Hbler M, Neto AG, Gianella Neto
A, Koch T, Gama de Abreu M: Effects of vaporized perfluorohexane and partial liquid ventilation on regional distribution
of alveolar damage in experimental lung injury. Intensive
Care Med 2007; 33:30814
10. Peterson BT, Collins ML, Gray LD, Azghani AO: Aerosolized
Pseudomonas elastase and lung fluid balance in anesthetized
sheep. J Appl Physiol (1985) 1992; 72:192733
11. Santos CL, Moraes L, Santos RS, Oliveira MG, Silva JD, MaronGutierrez T, Ornellas DS, Morales MM, Capelozzi VL, Jamel
N, Pelosi P, Rocco PR, Garcia CS: Effects of different tidal
volumes in pulmonary and extrapulmonary lung injury with
or without intraabdominal hypertension. Intensive Care Med
2012; 38:499508
12. Riva DR, Oliveira MB, Rzezinski AF, Rangel G, Capelozzi
VL, Zin WA, Morales MM, Pelosi P, Rocco PR: Recruitment
maneuver in pulmonary and extrapulmonary experimental
acute lung injury. Crit Care Med 2008; 36:19008
13. Garcia CS, Abreu SC, Soares RM, Prota LF, Figueira RC,
Morales MM, Capelozzi VL, Zin WA, Rocco PR: Pulmonary
morphofunctional effects of mechanical ventilation with
high inspiratory air flow. Crit Care Med 2008; 36:2329
14. Fanelli V, Mascia L, Puntorieri V, Assenzio B, Elia V, Fornaro
G, Martin EL, Bosco M, Delsedime L, Fiore T, Grasso S,
Ranieri VM: Pulmonary atelectasis during low stretch ventilation: Open lung versus lung rest strategy. Crit Care Med
2009; 37:104653
15. Krebs J, Pelosi P, Tsagogiorgas C, Haas J, Yard B, Rocco
PR, Luecke T: Time course of lung inflammatory and fibrogenic responses during protective mechanical ventilation in
healthy rats. Respir Physiol Neurobiol 2011; 178:3238
16. Sugahara K, Iyama K, Sano K, Kuroki Y, Akino T, Matsumoto

M: Overexpression of surfactant protein SP-A, SP-B, and SP-C
mRNA in rat lungs with lipopolysaccharide-induced injury.
Lab Invest 1996; 74:20920
17. Yamakawa N, Uchida T, Matthay MA, Makita K: Proteolytic
release of the receptor for advanced glycation end products from in vitro and in situ alveolar epithelial cells. Am J
Physiol Lung Cell Mol Physiol 2011; 300:L51625
18. Miyao N, Suzuki Y, Takeshita K, Kudo H, Ishii M, Hiraoka
R, Nishio K, Tamatani T, Sakamoto S, Suematsu M, Tsumura
H, Ishizaka A, Yamaguchi K: Various adhesion molecules
impair microvascular leukocyte kinetics in ventilator-induced
lung injury. Am J Physiol Lung Cell Mol Physiol 2006;
290:L105968
19. Yoshikawa S, King JA, Lausch RN, Penton AM, Eyal FG,
Parker JC: Acute ventilator-induced vascular permeability
and cytokine responses in isolated and in situ mouse lungs.
J Appl Physiol (1985) 2004; 97:21909
20. Wilson MR, Choudhury S, Goddard ME, ODea KP, Nicholson
AG, Takata M: High tidal volume upregulates intrapulmonary
cytokines in an in vivo mouse model of ventilator-induced
lung injury. J Appl Physiol (1985) 2003; 95:138593
21. Silva PL, Cruz FF, Fujisaki LC, Oliveira GP, Samary CS,
Ornellas DS, Maron-Gutierrez T, Rocha NN, Goldenberg R,
Garcia CS, Morales MM, Capelozzi VL, Gama de Abreu M,
Pelosi P, Rocco PR: Hypervolemia induces and potentiates
lung damage after recruitment maneuver in a model of sepsis-induced acute lung injury. Crit Care 2010; 14:R114
22. Plataki M, Hubmayr RD: The physical basis of ventilatorinduced lung injury. Expert Rev Respir Med 2010; 4:37385
23. Broccard AF, Hotchkiss JR, Suzuki S, Olson D, Marini JJ:
Effects of mean airway pressure and tidal excursion on
lung injury induced by mechanical ventilation in an isolated perfused rabbit lung model. Crit Care Med 1999;
27:153341
24. Roan E, Waters CM: What do we know about mechanical

strain in lung alveoli? Am J Physiol Lung Cell Mol Physiol
2011; 301:L62535
25. Cannon D, Kalso EJ, Sykes MK: Effects of the pattern of ventilation and of an increase in cardiac output on the distribution of
blood flow to a hypoxic lung lobe. Br J Anaesth 1988; 60:8190
26. Broccard AF, Hotchkiss JR, Kuwayama N, Olson DA, Jamal
S, Wangensteen DO, Marini JJ: Consequences of vascular
flow on lung injury induced by mechanical ventilation. Am J
Respir Crit Care Med 1998; 157(6 Pt 1):193542
27. Carvalho AR, Spieth PM, Gldner A, Cuevas M, Carvalho
NC, Beda A, Spieth S, Stroczynski C, Wiedemann B, Koch T,
Pelosi P, de Abreu MG: Distribution of regional lung aeration
and perfusion during conventional and noisy pressure support ventilation in experimental lung injury. J Appl Physiol
(1985) 2011; 110:108392
28. Spieth PM, Carvalho AR, Gldner A, Kasper M, Schubert R,
Carvalho NC, Beda A, Dassow C, Uhlig S, Koch T, Pelosi P,
Gama de Abreu M: Pressure support improves oxygenation
and lung protection compared to pressure-controlled ventilation and is further improved by random variation of pressure
support. Crit Care Med 2011; 39:74655
29. Gattinoni L, Carlesso E, Caironi P: Stress and strain within
the lung. Curr Opin Crit Care 2012; 18:427
30. Rocco PR, Dos Santos C, Pelosi P: Pathophysiology of ventilator-associated lung injury. Curr Opin Anaesthesiol 2012;
25:12330
31. Protti A, Andreis DT, Monti M, Santini A, Sparacino CC,
Langer T, Votta E, Gatti S, Lombardi L, Leopardi O, Masson
S, Cressoni M, Gattinoni L: Lung stress and strain during
mechanical ventilation: Any difference between statics and
dynamics? Crit Care Med 2013; 41:104655
32. Mead J, Takishima T, Leith D: Stress distribution in lungs:
A model of pulmonary elasticity. J Appl Physiol 1970;
28:596608
Volatile Organic Compounds during Inflammation and

Sepsis in Rats
A Potential Breath Test Using Ion-mobility Spectrometry
Tobias Fink, M.D., Alexander Wolf, M.D., Felix Maurer, D.Sc., Frederic W. Albrecht, M.D.,
Nathalie Heim, Beate Wolf, Anne C. Hauschild, M.Sc., Bertram Bdeker, Jrg I. Baumbach, Ph.D.,
Thomas Volk, M.D., Ph.D., Daniel I. Sessler, M.D., Ph.D., Sascha Kreuer, M.D., Ph.D.
ABSTRACT
Background: Multicapillary column ion-mobility spectrometry (MCC-IMS) may identify volatile components in exhaled
gas. The authors therefore used MCC-IMS to evaluate exhaled gas in a rat model of sepsis, inflammation, and hemorrhagic
shock.
Methods: Male SpragueDawley rats were anesthetized and ventilated via tracheostomy for 10h or until death. Sepsis was
induced by cecal ligation and incision in 10 rats; a sham operation was performed in 10 others. In 10 other rats, endotoxemia
was induced by intravenous administration of 10mg/kg lipopolysaccharide. In a final 10 rats, hemorrhagic shock was induced
to a mean arterial pressure of 355 mmHg. Exhaled gas was analyzed with MCC-IMS, and volatile compounds were identified using the BS-MCC/IMS-analytes database (Version 1209; B&S Analytik, Dortmund, Germany).
Results: All sham animals survived the observation period, whereas mean survival time was 7.9h in the septic animals, 9.1h
in endotoxemic animals, and 2.5 h in hemorrhagic shock. Volatile compounds showed statistically significant differences
in septic and endotoxemic rats compared with sham rats for 3-pentanone and acetone. Endotoxic rats differed significantly
from sham for 1-propanol, butanal, acetophenone, 1,2-butandiol, and 2-hexanone. Statistically significant differences were
observed between septic and endotoxemic rats for butanal, 3-pentanone, and 2-hexanone. 2-Hexanone differed from all other
groups in the rats with shock.
Conclusions: Breath analysis of expired organic compounds differed significantly in septic, inflammation, and sham rats.
MCC-IMS of exhaled breath deserves additional study as a noninvasive approach for distinguishing sepsis from inflammation.
HOUSANDS of volatile organic compounds (VOCs)

are expelled with each breath. The composition of these
VOCs can change with the state of health.110 Breath analysis
has consequently become a promising field of research. One
suitable tool for breath analysis is multicapillary column
ion-mobility spectrometry (MCC-IMS). The MCC-IMS
is a noninvasive and rapid method for detection of VOCs
in exhaled breath. The principles of MCC-IMS have been
described previously.11,12 MCC-IMS is already used extensively for detection of drugs, chemicals, and explosives.13,14
But this technique has already been used for evaluation of
pulmonary diseases, that is, chronic obstructive pulmonary
disease,13 acute respiratory distress syndrome,4 lung cancer,
and airway infections.1,6,15 Furthermore, it also appears possible to detect volatile metabolites arising from various bacterial species.710
Based on these results using MCC-IMS for diagnosis of
pulmonary diseases and inflammation, the question rises if
systemic inflammation and sepsis are also appropriated targets

Septic patients are not always readily identified by caregivers,
and there is a short window for optimum antibiotic treatment

Exhaled gas from rats given endotoxin compared with the gas
from rats with bacterial sepsis was found to be significantly different and different from rats who were in hemorrhagic shock
Breath analysis appears to be able to distinguish inflammation
from infection
for this novel noninvasive device. To our knowledge, only one

study from Guamn et al.15 used gas chromatographymass
spectrometry to evaluate the systemic inflammatory response
to intraperitoneal lipopolysaccharide in rats.
But whether MCC-IMS can be used on exhaled breath
as a measure of sepsis and systemic inflammation remains
unknown. We therefore compared patterns of exhaled
organic compounds in septic, endotoxemic shock (ES),
Submitted for publication November 25, 2013. Accepted for publication July 24, 2014. From the Department of Anesthesiology, Intensive
Care, and Pain Therapy, Saarland University Medical Center, Homburg (Saar), Germany (T.F., A.W., F.M., F.W.A., N.H., B.W., T.V., S.K.); Computational Systems Biology Group, Max Planck Institute for Informatics, Saarbrcken, Germany (A.C.H.); B&S Analytik, BioMedicalCenter,
Dortmund, Germany (B.B., J.I.B.); Faculty of Applied Chemistry, Reutlingen University, Reutlingen, Germany ( J.I.B.); and Department of
Outcomes Research, Cleveland Clinic, Cleveland, Ohio (D.I.S.).
January 2015
Volatile Organic Compounds during Inflammation and Sepsis
and sham-control rats during an observation period of

10h. A further group of pressure-controlled hypotension
was used to distinguish between infection, inflammation,
and noninfectious shock in exhaled metabolites. Noninfectious shock was included to distinguish changes in
VOCs resulting from infection/inflammation versus hypotension alone.
Material and Methods

Animals
All experiments were conducted with approval from our
Animal Care and Use Committee (Landesamt fr Soziales, Gesundheit und Verbraucherschutz; Saarbrcken; Germany) and in accordance with the German Animal Welfare
Act. Male SpragueDawley rats (200 to 300g body weight)
were obtained from Charles River (Sulzfeld, Germany) and
kept in the institutional animal facility under controlled
conditions (temperature 20 2C and 505% humidity).
Animals had free access to water; standard pellet food was
withheld for 12h before the experiment.
Chemicals
Pure lipopolysaccharide of Escherichia coli serotype O26:B6
was obtained from Sigma-Aldrich (Munich, Germany).
Surgical Procedures
Rats were anesthetized with pentobarbital 60mg/kg intraperitoneally and positioned on a warming plate; an open
tracheotomy was performed to facilitate breathing. The right
external jugular vein was catheterized to allow for infusion,
and the left carotid artery was catheterized for continuous
measurement of mean arterial pressure (MAP) and heart rate
(Philips IntelliVue MP20 Junior, Boeblingen, Germany).
Afterward the rats were connected to the respirator
(KTR-5 small animal ventilator; Hugo Sachs ElektronikHarvard Apparatus, March-Hugstetten, Germany) and
ventilated with highly purified synthetic air (Air Liquid,
Ludwigshafen, Germany). The MCC-IMS (type Breath
discovery; B&S Analytik, Dortmund, Germany) was connected to the same tank of synthetic air.
Tidal volume was adapted to a physiological lift of the
thorax wall and respiratory rate set to 65min1 at an inspiration ratio of 45% and a positive end-expiratory pressure of 1
to 2cm H2O. General anesthesia was maintained throughout the study with pentobarbital intravenously as needed.
The rats were given intravenous isotonic solution at a rate
of 10ml kg1 h1 (Sterofundin ISO; B. Braun, Melsungen,
Germany).
Experimental Protocol
We evaluated four groups of 10 rats each, randomly assigned to
sepsis (CLI), ES, hemorrhagic shock (HS), and sham control:
CLI: Sepsis was induced using a modified model of cecal
ligation and incision as previously described.16,17 The
cecum was exteriorized through a midline laparotomy,

and the cecum and paracecal blood vessels were ligated
below the ileocecal valve. The ligated cecum was
incised 1.5cm on the antimesenteric side and replaced
in the abdomen, and the abdominal wall was closed.
Sham: Sham-operated rats were treated likewise but
without ligation and incision.
ES: ES was induced by administration of 10mg/kg lipopolysaccharide E. coli serotype O26:B6 intravenously.
HS: A modified pressure-regulated shock model was used
as described preciously,18,19 but without resuscitation
and retransfusion. HS was induced by rapid arterial
blood withdrawal by way of the carotid artery with a
MAP of 355 mmHg.
Induction of sepsis or sham operation was carried out by
the same investigator and subsequently blinded during the
observation period. ES and HS groups were also carried out
by the same investigator, but blinding was not realizable.
When measurements were finished, all surviving rats were
killed with an overdose of pentobarbital.
Arterial blood (0.2ml) was sampled every 2.5h for analysis of pH, base excess, lactate, hemoglobin, glucose, partial
pressures of oxygen, and carbon dioxide (Radiometer ABL
800 Basic, Willich, Germany). After insertion of the arterial catheter, and 5h after induction of CLI, blood samples
(0.5ml) were taken for enzyme-linked immunosorbent assay
(ELISA) analysis of pro- and antiinflammatory cytokines.
MCC-IMS Measurements
We used a Bioscout 2011 MCC-IMS based on a 550 MBq
63Ni ionization source with an electrical field strength of
300V/cm and equipped with MCC using a general-purpose, slightly polar standard capillary column phase OV-5
(Multichrom, Novosibirsk, Russia).
Before connecting the rats to the MCC-IMS, we measured background concentrations of organic volatile
compounds from the respirator for 1h. Throughout the
experiment, the rats were ventilated with synthetic air as
described previously.20 To avoid any extraneous signals,
polytetrafluoroethylene (Bohlender, Grnsfeld, Germany)
was used for all external connections, and the MCC-IMS
sampling tube was directly connected to the exhalatory
line of the respirator. Breath humidity and cluster reactions
between ions and water lead to significant changes of the
IMS spectra; therefore we used a MCC coupled with an ionmobility spectrometer to avoid this error.
Exhaled gas (10ml per measurement) was collected in the
sample loop before entering the multicapillary column of the
ion-mobility spectrometer. This arrangement allowed three
measurements per hour. The retention time in relation to the
preseparation of the MCC, drift time of the ions within the
IMS, and the intensity of the analytes were delivered by the
MCC-IMS and subsequently identified using the BS-MCC/
IMS-analytes database (Version 1209; B&S Analytik).
Anesthesiology 2015; 122:117-26 118 Fink et al.
Peaks representing various VOCs were identified using

Visual Now 3.1 (B&S Analytik). All signals above the threshold of 1 mV to 5V (peak range: lower limit is defined as five
times over the background noise; the upper limit depends
on the saturation of the drift tube) were defined as a peak
and characterized by their position with drift time, retention
time, and concentration related to the peak intensity.
Cytokine Assay
Cytokine concentrations were measured by ELISA. Positive controls of each cytokine were measured routinely with
each assay (ELISA Antibodies BD OptEIA; BD Biosciences
Pharmingen, San Diego, CA).
Statistical Analysis and Signal Processing
Mathematical and statistical evaluation was carried out
using SigmaPlot (version 12.5; Systat Software, Erkrath,
Germany).
Based on our previous experience, we confirmed a number of 10 rats per group as sufficient to see clear differences.
Animals were randomized in four groups using a random
number generator (Excel 2010; Microsoft, Redmond, WA).
Survival time was analyzed using a log-rank model
according to KaplanMeier, followed by an all-pairwise
comparison using Tukey test.
Data were tested for distribution normality (KolmogorovSmirnov test) and analyzed using repeated-measures ANOVA, followed by post hoc multiple comparisons
with HolmSidak method. When appropriate, repeatedmeasures ANOVA on ranks were used. Based on the survival
rate, statistical analysis for each volatile compound was performed from 0 to 6h in the sham, sepsis, and endotoxemic
group.
Hemorrhagic shock group was merely analyzed after 2h
using one-way ANOVA followed by post hoc multiple comparisons with the HolmSidak method. A two-tailed P value
of less than 0.05 was considered statistically significant.
Baseline concentrations of tumor necrosis factor-, interleukin-6, and interleukin-10 were similar at the beginning
of the experiment. However, concentrations of all three
increased significantly in the septic and endotoxemic rat
(table2).
Hemodynamic Parameters
Baseline MAP was similar in all groups and remained
stable throughout in the sham-operated rats (fig.2).
Induction of HS resulted in a significant decline of MAP.
Injection of lipopolysaccharide resulted in an early reduction of MAP from 1 to 2h and further at 5h and thereafter. MAP gradually declined in the CLI rats, becoming
significantly lower than in the sham group after 5h. The
heart rate remained within normal ranges during the
entire observation period, without significant differences
between the groups.
Multicapillary Column Ion-mobility Spectrometry
During the experiment, about a hundred different signals
between a threshold of 1 mV and 5V (peak definition) were
detected in exhaled air. Seven signals presented with significant differences between groups during the first 6h of the
observation period. Those seven were considered potentially
useful VOCs for distinguishing among treatments. Figure3
illustrates the typical IMS chromatogram of 3-pentanone
and a three-dimensional plot for this setting.
Results
Survival Time, Blood Gas Analysis, and Cytokines
All sham rats survived the observation period, whereas others died earlier: HS: mean, 2.50.3h; 95% CI, 1.9 to 3.1h;
CLI: mean, 7.90.4h; CI, 7.2 to 8.7h; and, ES: mean,
9.10.4h; and CI, 8.3 to 9.9h. All P values less than 0.001
versus sham (fig.1).
All blood gas values remained stable and normal in shamoperated rats. In contrast, blood lactate concentrations
increased significantly over time in CLI, ES, and HS rats.
Arterial base excess and pH were significantly lower in the
CLI and ES groups than sham rats (table1). Respiratory
parameters (partial pressure of oxygen, partial pressure of
carbon dioxide, and oxygenation index), hemoglobin, and
blood glucose remained within normal ranges throughout
the experiment in both groups.
Fig. 1. Survival rate. After hemorrhagic shock (HS), endotoxemic shock (ES), and sepsis (CLI), mean survival time was significantly decreased compared with sham control. *P < 0.001
for sham versus HS, ES, and CLI.
Table 1. Arterial Blood Gas Analysis

0h
Survival rate (N = )
Sham
10
CLI
10
ES
10
HS
10
pH
Sham
7.39 (7.347.49)
CLI
7.37 (7.317.44)
ES
7.41 (7.327.52)
HS
7.40 (7.307.50)
Base excess
Sham
0.5 (6.0 to 5.7)
CLI
1.1 (8.4 to 5.4)
ES
2.3 (2.4 to 6.1)
HS
2.5 (0.2 to 5.8)
Lactate
Sham
0.9 (0.51.6)
CLI
0.9 (0.51.8)
ES
0.7 (0.51.0)
HS
0.6 (0.41.0)
Partial pressure of oxygen
Sham
82.5 (59.1119)
CLI
73.1 (59.8121)
ES
83.9 (51.5127)
HS
72.1 (35.6100)
Partial pressure of carbon dioxide
Sham
35.1 (24.040.9)
CLI
37.6 (30.444.2)
ES
36.6 (23.344.9)
HS
37.7 (30.747.8)
Oxygen index
Sham
393 (281567)
CLI
348 (285576)
ES
400 (245480)
HS
343 (257585)
2.5 h
5h
7.5 h
10 h
10
10
10
5
10
10
10
10
7
8
10
2
5
7.42 (7.387.45)
7.35 (7.207.44)
7.36 (7.287.46)
7.41 (7.307.54)
7.41 (7.357.48)
7.26* (7.16739)
7.36 (7.147.52)
7.42 (7.367.54)
7.23* (7.137.40)
7.30 (7.117.46)
7.39 (7.357.42)
7.23 (7.237.26)
7.29 (7.257.30)
3.6 (6.1 to 1.8)

5.6 (12.3 to 0.2)
5.1 (9.4 to 0.0)
5.7 (12.3 to 0.6)
3.9 (5.4 to 2.6)

10.1* (17.9 to 2.1)
6.0* (14.4 to 0.7)
0.6 (0.50.8)
1.1 (0.61.8)
1.0 (0.51.4)
3.2 (2.26.0)
0.9 (0.61.4)
2.4 (1.04.7)
2.6* (1.45.3)
0.7 (0.41.1)
4.1* (1.37.2)
4.9* (3.09.4)
0.6 (0.40.9)
6.2 (5.27.2)
4.5 (3.95.4)
75.8 (63.787.2)
79.4 (48.899.7)
75.2 (58.5105)
73.0 (50.0123)
82.1 (41.5131)
82.7 (47.895.2)
91.9 (60.7166)
77.8 (50.7154)
73.0 (68.187.0)
87.7 (85.1127)
79.1 (61.2123)
77.6 (46.293.4)
80.6 (68.9126)
31.3 (25.337.9)
32.4 (23.939.4)
33.2 (28.238.9)
28.1 (19.431.0)
30.9 (22.537.1)
32.9 (25.743.7)
34.6 (20.444.5)
29.4 (18.241.6)
24.5 (17.735.6)
29.0 (17.036.4)
32.0 (26.040.0)
29.7 (28.431.0)
24.2 (21.628.0)
361 (303415)
378 (246661)
358 (301537)
348 (469619)
391 (197623)
394 (232475)
438 (279500)
370 (279480)
315 (227405)
419 (289790)
382 (241633)
371 (324414)
384 (343504)
4.0 (7.2 to 1.5)

3.8 (6.9 to 3.2)
14.8* (25.9 to 7.1) 15.0 (18.0 to 12.0)
11.6* (23.8 to 1.7) 10.9 (13.6 to 80)
Data are given as means 95% CIs. Arterial blood gas analysis revealed normal baseline values for pH, base excess, lactate, partial pressure of oxygen,
and carbon dioxide. Lactate values increase significantly, and pH and base excess decrease significantly over time, when compared with sham control.
Respiratory parameters: partial pressure of oxygen and carbon dioxide as oxygenation index (partial pressure of oxygen/oxygen fraction) remained within
normal ranges throughout the experiment. Statistical analysis was performed from 0 to 7.5h.
* P < 0.05 vs. sham group. P < 0.05 vs. corresponding baseline.
CLI = sepsis; ES = endotoxemic shock; HS = hemorrhagic shock.
With the BS-MCC/IMS-analytes database (Version

1209), all seven potential compounds could be identified
according to their retention time and drift time. All these
VOCs are shown in table3.
As a result of the measured survival time (fig. 1), the statistical analysis of each VOC was performed at 0, 2, 4, and 6h
in the sham, CLI, and ES (each n = 10 per group at these time
points). Sham-operated rats presented with approximately
stable concentrations of all VOCs throughout the observation
period. During the experiment, seven VOCs decreased significantly in lipopolysaccharide-treated rats, when compared with
sham. At 2h and thereafter 1-propanol (P < 0.04), butanal
(P < 0.001), acetophenone (P < 0.022), 1,2-butandiol (P <
0.001), 3-pentanone (P < 0.001), acetone (P < 0.001), and
2-hexanone (P < 0.001), all decreased (table 3 and figs.46).
In the septic rats, though, only two compounds decreased

significantly. 3-Pentanone (P < 0.001) and acetone (P = 0.02)
decreased at 6h (table 3 and figs. 4 and 5).
Overall, there were significant differences between ES and
CLI in the concentrations of three VOCs: butanal (P < 0.001
at 2h and thereafter), 3-pentanone (P < 0.001 at 2h and
thereafter), and 2-hexanone (P < 0.001 at 2h and thereafter
(table 3 and figs. 4 and 6).
Noninfectious shock was included to distinguish changes
in VOCs resulting from infection or inflammation versus
hypotension alone. Due to the high mortality in this group,
a single statistical evaluation was performed at 2h after
induction of HS:
There was significant increase in 2-hexanone (P < 0.001)
when compared with all other groups, and for butanal
Table 2. Cytokine Response

0h
(N = 10 per Group)
Tumor necrosis factor-, pg/ml
Sham
0.0 (0.00.0)
CLI
0.1 (0.00.3)
ES
0.2 (0.02.0)
Interleukin-10, pg/ml
Sham
0.0 (0.00.0)
CLI
5.1 (0.016.5)
ES
2.3 (0.07.5)
Interleukin-6, pg/ml
Sham
374 (279677)
CLI
375 (346404)
ES
315 (160412)
5h
(N = 10 per Group)
0.0 (0.00.0)
17.9 (0.037.1)*
27.10 (2.362.1)*
0.0 (0.00.0)
312 (84.4784)*
184 (10.0365)*
377 (267505)
11,223 (1,22331,790)*
17,663 (1,24641,598)*
CLI and ES significantly increased tumor necrosis factor-, interleukin-10,

and interleukin-6 compared with sham-operated rats. Data are given as
means 95% CI (n = 10 per group at 0 and 5h).
* P < 0.001 vs. baseline and corresponding sham group.
(P < 0.006), 3-pentanone (P < 0.001), and acetone (P < 0.001)

when compared with endotoxemia. However, none of the
VOCs decreased significantly compared with sham (table 3).
Table4 gives an overview of changes in volatile compounds with significant differences after 6h.
Discussion
Currently, diagnosis of sepsis and septic shock relies on nonspecific physiological signs, and early appropriate treatment
remains of prime importance. Common diagnostic tools,
such as blood culture techniques, procalcitonin, ELISA, or
polymerase chain reaction, harbor a series of drawbacks and
limitations.2126 MCC-IMS may prove to be an innovative
tool to complement conventional diagnostic techniques.15
Major advantages of the technique lie in its noninvasive
character and near-real-time bedside analysis.8
We are aware of only a single previous study in which expired
gas was used to detect a systemic inflammatory response.15 In
that study, IMS and gas chromatographymass spectrometry
were used to evaluate the systemic inflammatory response to
intraperitoneal lipopolysaccharide in rats. Our use of MCC-IMS
technology for VOC assessment of sepsis appears to be novel.
We observed a total of 100 signals in exhaled air, between
a threshold of 1 mV and 5V, which was defined as peak
range. Seven VOCs presented with significant difference
between the investigated groups during the first 6h of the
experiment and have to be chosen as potential VOCs to distinguish between research groups.
All detected compounds could be identified; the most
were ketones, hydrocarbons, alcohols, and ethers. Previous
investigators, using different methodologies, found similar
compounds in the breath of humans,27,28 rats,15 and mice29
and also in the headspace samples of bacteria cultures.7,8,10,30
Fig. 2. Mean arterial pressure was significantly lower in the

hemorrhagic shock (HS) than in any of the other groups. In
endotoxemic shock (ES) and sepsis (CLI) groups, mean arterial pressure decreased over time throughout the experiment
compared with sham control. Statistical analysis for sham,
CLI, and ES was performed from 0 to 6 h: *P < 0.05 for sham
versus CLI. **P < 0.05 for sham versus ES. #P < 0.05 for CLI
versus corresponding baseline. ##P < 0.05 for ES versus
corresponding baseline. Statistical analysis for HS was performed at 2 h: &P < 0.05 for HS versus all other groups. Data
are given as means SEMs (n = 10 per group at the beginning
of the experiment).
Our focus, though, was not on peak identity, but in differences between sepsis, inflammation, and shock.
Our experimental model of fulminant four-quadrant
peritonitis presumably better represents clinical sepsis than
lipopolysaccharide endotoxemia, but also poses several problems. During CLI sepsis, rats presented with three clinical
entities: infection, inflammation, and shock. To distinguish
among them, we used two further models: one with a systemic inflammatory response induced by lipopolysaccharide
and another with a pressure-controlled HS.
In comparison with untreated animals, most volatile
compounds detected in sepsis and inflammation animals
declined over time, as did hemodynamic compensation.
However, none of these compounds similarly declined during HS; in contrast, HS was the only condition associated
with an increase in VOCs. Also surprisingly, the sterile
inflammation after injection of lipopolysaccharide seemed to
have more altered the most VOCs than the complex model
of polymicrobial sepsis. Thus, the observed alterations in volatile compounds during sepsis are unlikely to be the result of
hemodynamic decompensation, but seem more likely due to
factors specific to sepsis, inflammation, or consequent pulmonary impairments. Furthermore, all animals presented
with a normal oxygenation index during the experiment, so
the distinction between groups is unlikely influenced by the
degree of lung injury.
Fig. 3. Ion-mobility spectrometry chromatogram of 3-pentanone. (A) Shows a typical ion-mobility spectrometer chromatogram
for, respectively, individual rats in the sham, sepsis (CLI), endotoxemic shock (ES), and hemorrhagic shock (HS) groups after 4h.
(B) A three-dimensional plot of the same rats. Signals above a threshold of 1 mV defined a peak. The peaks were characterized
by their position in the ion-mobility spectrometry chromatogram, using the inverse reduced mobility (axis of abscissa; 1/K0valuewhich is proportional to the drift time of the ions) and the retention time (axis of ordinates) as well as the peak expansion.
Subsequently, peaks were identified according to database BS-multicapillary column/ion-mobility spectrometry-analytes (Version 1209; B&S Analytik, Dortmund, Germany). For visualization, 3-pentanone is labeled as representative peak in this setting.
Comparing the 3-pentanone peak, the drift time of 0.551 Vs/cm2, and a retention time of 8.2 s are equal in all chromatograms.
The peak intensity after 4h was 685 mV for sham, 370 mV for CLI, 60 mV for ES, and 585 mV for HS for this peak and is given
as difference in color in A and size in B.
Table 3. Identification and Changes in MCC-IMS Peaks

0h
2h
Survival rate (N =)
Sham
10
10
CLI
10
10
ES
10
10
HS
10
6
1-Propanol (CAS 71-23-8)
Sham
29.3 (13.460.9)
31.9 (9.872.1)
CLI
31.2 (16.070.2)
30.1 (10.962.8)
ES
24.0 (6.351.3)
15.8* (5.438.8)
HS
31.8 (24.447.0)
32.1 (26.451.5)
Butanal (CAS 123-72-8)
Sham
11.8 (6.115.8)
13.4 (6.318.1)
CLI
12.1 (8.115.3)
12.8 (6.916.9)
ES
8.3 (5.815.4)
6.9* (4.512.0)
HS
11.3 (7.018.3)
13.9 (6.524.1)
Acetophenone (CAS 98-86-2)
Sham
3.8 (1.76.1)
4.5 (1.68.3)
CLI
4.0 (1.86.1)
3.6 (1.75.2)
ES
2.7 (1.67.1)
2.0* (1.12.9)
HS
3.4 (2.26.6)
3.1 (1.96.5)
1,2-Butandiol (CAS 584-03-2)
Sham
4.3 (1.59.0)
4.7 (2.18.8)
CLI
4.1 (2.55.8)
4.0 (1.96.1)
ES
2.9 (1.65.1)
2.4* (1.33.6)
HS
2.8 (1.94.4)
2.7 (1.84.4)
3-Pentanone (CAS 96-22-0) shown in figure 4
Sham
481 (309658)
541 (215753)
CLI
495 (319736)
489 (232655)
ES
404 (238929)
153* (131432)
HS
553 (204700)
654 (371697)
Acetone (CAS 67-64-1) shown in figure 5
Sham
118 (54.2210)
163 (31.4303)
CLI
126 (53.6286)
125 (44.9218)
ES
106 (31.2209)
58.8* (21.0149)
HS
167 (23.9358)
217 (130319)
2-Hexanone (CAS 591-78-6) shown in figure 6
Sham
21.3 (13.329.7)
22.2 (10.430.9)
CLI
22.9 (18.631.3)
20.8 (14.426.9)
ES
18.3 (10.533.7)
12.6* (7.118.2)
HS
21.0 (17.727.0)
27.6 (21.033.1)
4h
6h
8h
10 h
10
10
10
2
10
10
10
10
3
7
10
2
5
30.6 (8.571.7)
22.7 (7.342.9)
10.9* (4.321.7)
35.7 (26.947.2)
28.3 (7.461.2)
14.6 (6.922.1)
10.6* (6.118.1)
26.7 (7.355.2)
16.7 (10.429.1)
12.5 (9.817.9)
26.2 (7.246.5)
12.7 (9.316.0)
14.3 (10.523.3)
13.6 (5.919.8)
11.8 (9.713.5)
6.2* (4.29.6)
21.3 (8.430.7)
12.7 (4.719.0)
10.0 (6.012.4)
6.3* (4.69.3)
12.8 (4.520.6)
10.1 (8.312.1)
5.8 (4.78.9)
13.2 (4.720.8)
9.7 (8.810.6)
5.2 (4.06.9)
5.1 (1.713.0)
2.8 (1.04.1)
1.9* (1.52.4)
2.8 (2.82.8)
4.6 (1.910.5)
2.6 (1.74.6)
1.9* (1.12.7)
5.1 (1.614.1)
3.3 (2.14.4)
2.1 (1.52.6)
5.0 (1.711.0)
3.4 (3.03.9)
2.0 (1.72.2)
4.7 (1.89.3)
3.6 (1.85.5)
2.3* (1.72.8)
2.3 (1.92.6)
4.4 (1.87.5)
3.5 (1.95.0)
2.5* (1.63.1)
4.7 (2.210.3)
3.0 (2.04.3)
2.4 (1.23.0)
4.8 (2.08.2)
2.3 (2.02.5)
2.6 (1.83.0)
528 (173785)
353 (94.2553)
77.8* (28.3304)
576 (558585)
499 (107767)
213* (80.8293)
75.8* (33.3240)
485 (157779)
189 (129224)
69 (32181)
482 (141769)
179 (113244)
67 (43135)
172 (27.8361)
79.4* (22.8143)
33.5* (21.162.2)
158 (140167)
154 (25.9325)
46.3* (22.564.3)
28.6* (21.748.7)
155 (21.9338)
44 (36.249.5)
29 (22.238.8)
153 (19.8320)
43 (33.752.2)
38 (24.479.5)
20.7 (9.029.1)
18.2 (9.622.7)
10.5* (6.015.0)
33.8 (29.043.2)
19.5 (9.328.8)
16.0 (8.020.0)
11.3* (7.914.5)
19.3 (8.630.7)
19.7 (13.025.9)
12.1 (9.514.6)
19.0 (7.229.7)
18.0 (13.222.7)
13.7 (9.920.6)
Data are given as means 95% CI. Seven volatile organic compounds with significant changes for peak intensities among the groups were found during
the experiment (presented in table 3). Three of them were illustrated in figures 46. Statistical analysis was performed from 0 to 6h.
* P < 0.05 vs. sham group. P < 0.05 vs. corresponding baseline. P < 0.05 vs. LPS. P < 0.05 vs. all other groups.
CAS = chemical abstract service registry number; CLI = sepsis; ES = endotoxemic shock; HS = hemorrhagic shock; MCC-IMS = multicapillary column
ion-mobility spectrometry.
A detailed peak examination revealed an early and strongly

reduced release of ketones, acetone, and 3-pentanone during
sepsis and inflammation. Our findings are consistent with previous reports showing that blood ketone bodies and the ketogenic
capacity of the liver are inhibited during sepsis in rats.31,32 These
findings seem to be independent of nutritional conditions because
fasting before and during sepsis did not result in hypoglycemia.32
Instead, increased tumor necrosis factor- and interleukin-6
plasma concentrations may mediate ketogenic impairment.33
Volatile acetone decreased over time in sepsis, inflammation, and HS, suggesting a hemodynamic dependence. But
interestingly, other studies reported an increased acetone

in exhaled breath of patients with heart failure, associated
with an increase in blood brain natriatic peptide concentration.34,35 It thus seems unlikely that reduced acetone concentrations result from cardiac decompensation.
Guamn et al.15 selected acetone and a further 14 volatile
compounds as possible compounds that can be potentially used
to separate healthy rats from endotoxemic rats 24h after lipopolysaccharide treatment using solid phase microextraction gas
chromatographymass spectrometry Schubert et al.35 reported
that n-pentane concentrations are higher and that acetone
Fig. 4. Peak course of 3-pentanone. For 3-pentanone,

sham and hemorrhagic shock (HS) rats exhibited stable signal strength during the experiment. In endotoxemic shock
(ES) and sepsis (CLI), 3-pentanone concentrations declined
during the observation period. Statistical analysis was
performed from 0 to 6 h: *P < 0.05 for sham versus CLI.
**P < 0.05 for sham versus ES. P < 0.05 for CLI versus
ES. #P < 0.05 for CLI versus corresponding baseline.
##P < 0.05 for ES versus corresponding baseline. Data are
given as means SEMs.
Fig. 6. Peak course of 2-hexanone. 2-Hexanone increased

significantly more after induction of hemorrhagic shock (HS)
than other treatments. 2-Hexanone decreased in endotoxemic shock (ES) compared with sham control and sepsis
(CLI). Statistical analysis was performed from 0 to 6 h: **P <
0.05 for sham versus ES. P < 0.05 for CLI versus ES. #P <
0.05 for CLI versus corresponding baseline. ##P < 0.05 for
ES versus corresponding baseline. Statistical analysis for
HS was performed at 2 h: &P < 0.05 for HS versus all other
groups. Data are given as means SEMs.
Table 4. Overview of Changes in Volatile Compound Profiles
1-Propanol
2h
Sham
CLI
ES
HS
6h
3-Pentanone
2h
6h
Acetone
2h
6h
2-Hexanone
2h
6h
Four volatile organic compounds displayed significant differences for CLI

or ES after 6h, when compared with corresponding baseline values. Two
hours after HS, 2-hexanone was associated with a significant increase in
intensity compared with baseline. Significant changes in volatile compound
profiles are given as directional arrows: stable, decline, increase.
Fig. 5. Peak course of acetone. Examples of an acetone peak

declining during hemorrhagic shock (HS), endotoxemic shock
(ES), and sepsis (CLI). Statistical analysis was performed from
0 to 6 h: *P < 0.05 for sham versus CLI. **P < 0.05 for sham
versus ES. #P < 0.05 for CLI versus corresponding baseline.
##P < 0.05 for ES versus corresponding baseline. Data are
given as means SEMs.
concentrations are lower in the blood of mechanically ventilated septic patients than in healthy controls. However, GC/
MS revealed no differences in the concentration of either substance in the breath of the patient groups. In contrast, previous
studies showed increased concentrations of acetone, dimethyl
sulfide, 2-butanone, and 2-pentanone in the breath of patients
with liver cirrhosis and in the headspace samples of different
bacteria.8,29,36 The extent to which any of these VOCs might
serve as useful diagnostic criteria for sepsis thus remains unclear.
Our study is limited in so far as rodent models only partially correlate with human conditions.37 However, the sepsis
and the inflammation model we chose are standard ones that

have been used in many previous studies. A more important
limitation is that we do not know whether expired breath
markers of inflammation and sepsis can detect the conditions
in humans before they are otherwise clinically apparent.
In summary, MCC-IMS analysis of expired gas may help
identify sepsis or inflammation. Future studies should evaluate plausible mechanisms for changes in VOC quantities and
which breath profiles best indicate a septic or inflammatory
state. Thereafter, breath analysis for sepsis and inflammation
needs to be confirmed in humans, and its sensitivity and
specificity determined.
Acknowledgments
The authors thank Jan Baumbach, Ph.D. (Department of
Mathematics and Computer Science, University of Southern
Denmark, Odense, Denmark), for his encouragement and
Stefan Wagenpfeil, Ph.D. (Department of Biometry, Epidemiology, and Medical Informatics, University Medical Center, Homburg (Saar), Germany), for his statistical support.
This study contains data taken from the thesis presented
by Nathalie Heim (Department of Anesthesiology, Intensive
Care, and Pain Therapy, University Medical Center) as part
of the requirements for the obtention of the degree Doctor of Medicine at Saarland University Medical Center and
Saarland University Faculty of Medicine.
The data analysis work performed in this article (Dr.
Baumbach) was supported by Deutsche Forschungsgemeinschaft (Bonn, Germany) within the Collaborative Research
Center (Sonderforschungsbereich) SFB 876 Providing Information by Resource-Constrained Analysis, project TB1
Resource-Constrained Analysis of Spectrometry Data. All
other funding was provided from department sources.
Competing Interests
Correspondence
Address correspondence to Dr. Fink: Department of Anesthesiology, Intensive Care, and Pain Therapy, Saarland
University Medical Center, Kirrbergerstr., D-66421 Homburg
(Saar), Germany. tobias.fink@uks.eu. Information on purchasing reprints may be found at www.anesthesiology.org
or on the masthead page at the beginning of this
issue.
Anesthesiologys articles are made freely accessible to all
readers, for personal use only, 6 months from the cover date
of the issue.
References
1. Westhoff M, Litterst P, Maddula S, Bdeker B, Rahmann S,
Davies AN, Baumbach JI: Differentiation of chronic obstructive pulmonary disease (COPD) including lung cancer from
healthy control group by breath analysis using ion mobility
spectrometry. Int J Ion Mobil Spec 2010; 13:1319
2. Westhoff M, Litterst P, Madulla S, Bdeker B, Baumbach
JI: Statistical and bioinformatical methods to differentiate
chronic obstructive pulmonary disease (COPD) including
lung cancer from healthy control by breath analysis using ion
mobility spectrometry. Int J Ion Mobil Spec 2011; 14:13949
3. Bessa V, Darwiche K, Teschler H, Sommerwerck U, Rabis
T, Baumbach JI, Freitag L: Detection of volatile organic
compounds (VOCs) in exhaled breath of patients with

chronic obstructive pulmonary disease (COPD) by ion mobility spectrometry. In J Mobil Spec 2011; 14:713
4. Schubert JK, Mller WP, Benzing A, Geiger K: Application
of a new method for analysis of exhaled gas in critically ill
patients. Intensive Care Med 1998; 24:41521
5. Baumbach JI, Westhoff M: Ion mobility spectrometry to detect
lung cancer and airway infection. Spectroscopy Europe 2006;
18:227
6. Westhoff M, Litterst P, Freitag L, Urfer W, Bader S, Baumbach
JI: Ion mobility spectrometry for the detection of volatile
organic compounds in exhaled breath of patients with lung
cancer: Results of a pilot study. Thorax 2009; 64:7448
7. Dwivedi P, Puzon G, Tam M, Langlais D, Jackson S, Kaplan K,
Siems WF, Schultz AJ, Xun L, Woods A, Hill HH Jr: Metabolic
profiling of Escherichia coli by ion mobility-mass spectrometry with MALDI ion source. J Mass Spectrom 2010;
45:138393
8. Maddula S, Blank LM, Schmid A, Baumbach JI: Detection of
volatile metabolites of Escherichia coli by multi capillary column coupled ion mobility spectrometry. Anal Bioanal Chem
2009; 394:791800
9. Purkhart R, Khler H, Liebler-Tenorio E, Meyer M, Becher
G, Kikowatz A, Reinhold P: Chronic intestinal Mycobacteria
infection: Discrimination via VOC analysis in exhaled breath
and headspace of feces using differential ion mobility spectrometry. J Breath Res 2011; 5:027103
10. Jnger M, Vautz W, Kuhns M, Hofmann L, Ulbricht S,
Baumbach JI, Quintel M, Perl T: Ion mobility spectrometry
for microbial volatile organic compounds: A new identification tool for human pathogenic bacteria. Appl Microbiol
Biotechnol 2012; 93:260314
11. Baumbach JI: Ion mobility spectrometry coupled with multicapillary columns for metabolic profiling of human breath. J
Breath Res 2009; 3:034001
12. Baumbach JI, Eiceman GA: Ion mobility spectrometry:
Arriving on site and moving beyond a low profile. Appl
Spectrosc 1999; 53:338A55A
13. Borsdorf H, Eiceman GA: Ion mobility spectrometry principles and applications. Appl Spectrosc Rev 2006; 41:32375
14. Eiceman GA, Stone JA: Ion mobility spectrometers in national
defence. Anal Chem 2004; 76:390A7A
15. Guamn AV, Carreras A, Calvo D, Agudo I, Navajas D, Pardo
A, Marco S, Farr R: Rapid detection of sepsis in rats through
volatile organic compounds in breath. J Chromatogr B Analyt
Technol Biomed Life Sci 2012; 881882:7682
16. Fink T, Heymann P, Taha-Melitz S, Taha A, Wolf B, Rensing
H, Volk T, Mathes AM: Dobutamine pretreatment improves
survival, liver function, and hepatic microcirculation after
polymicrobial sepsis in rat. Shock 2013; 40:12935
17. Scheiermann P, Hoegl S, Revermann M, Ahluwalia D, Zander
J, Boost KA, Nguyen T, Zwissler B, Muhl H, Hofstetter C:
Cecal ligation and incision: An acute onset model of severe
sepsis in rats. J Surg Res 2009; 151:1327
18. Rensing H, Jaeschke H, Bauer I, Ptau C, Datene V, Pannen
BH, Bauer M: Differential activation pattern of redox-sensitive transcription factors and stress-inducible dilator systems heme oxygenase-1 and inducible nitric oxide synthase
in hemorrhagic and endotoxic shock. Crit Care Med 2001;
29:196271
19. Jellestad L, Fink T, Pradarutti S, Kubulus D, Wolf B, Bauer I,
Thiemermann C, Rensing H: Inhibition of glycogen synthase
kinase (GSK)-3- improves liver microcirculation and hepatocellular function after hemorrhagic shock. Eur J Pharmacol
2014; 724:17584
20. Wolf A, Baumbach JI, Kleber A, Maurer F, Maddula S, Favrod
P, Jang M, Fink T, Volk T, Kreuer S: Multi-capillary columnion mobility spectrometer (MCC-IMS) breath analysis in
ventilated rats: A model with the feasibility of long-term measurements. J Breath Res 2014; 8:016006
21. Mylotte JM, Tayara A: Blood cultures: Clinical aspects and
controversies. Eur J Clin Microbiol Infect Dis 2000; 19:15763
22. Sankar V, Webster NR: Clinical application of sepsis biomarkers. J Anesth 2013; 27:26983
23. Schaub N, Frei R, Mller C: Addressing unmet clinical needs in
the early diagnosis of sepsis. Swiss Med Wkly 2011; 141:w13244
24. Vincent JL, Donadello K, Schmit X: Biomarkers in the critically
ill patient: C-reactive protein. Crit Care Clin 2011; 27:24151
25. Shafazand S, Weinacker AB: Blood cultures in the critical care
unit: Improving utilization and yield. Chest 2002; 122:172736
26. Casserly B, Read R, Levy MM: Multimarker panels in sepsis.
Crit Care Clin 2011; 27:391405
27. Ligor T, Ligor M, Amann A, Ager C, Bachler M, Dzien A,
Buszewski B: The analysis of healthy volunteers exhaled
breath by the use of solid-phase microextraction and GC-MS.
J Breath Res 2008; 2:046006
28. Phillips M, Herrera J, Krishnan S, Zain M, Greenberg J,
Cataneo RN: Variation in volatile organic compounds in the
breath of normal humans. J Chromatogr B Biomed Sci Appl
1999; 729:7588
29. Vautz W, Nolte J, Bufe A, Baumbach JI, Peters M: Analyses of
mouse breath with ion mobility spectrometry: A feasibility
study. J Appl Physiol (1985) 2010; 108:697704
30. Filipiak W, Sponring A, Baur MM, Filipiak A, Ager C,

Wiesenhofer H, Nagl M, Troppmair J, Amann A: Molecular
analysis of volatile metabolites released specifically by
Staphylococcus aureus and Pseudomonas aeruginosa. BMC
Microbiol 2012; 12:113
31. Lanza-Jacoby S, Rosato E, Braccia G, Tabares A: Altered
ketone body metabolism during gram-negative sepsis in the
rat. Metabolism 1990; 39:11517
32. Ohtoshi M, Jikko A, Asano M, Uchida K, Ozawa K, Tobe

T: Ketogenesis during sepsis in relation to hepatic energy
metabolism. Res Exp Med (Berl) 1984; 184:20919
33. Pailla K, El-Mir MY, Cynober L, Blonde-Cynober F: Cytokinemediated inhibition of ketogenesis is unrelated to nitric
oxide or protein synthesis. Clin Nutr 2001; 20:3137
34. Marcondes-Braga FG, Gutz IG, Batista GL, Saldiva PH, AyubFerreira SM, Issa VS, Mangini S, Bocchi EA, Bacal F: Exhaled
acetone as a new biomaker of heart failure severity. Chest
2012; 142:45766
35. Schubert JK, Miekisch W, Birken T, Geiger K, NldgeSchomburg GF: Impact of inspired substance concentrations
on the results of breath analysis in mechanically ventilated
patients. Biomarkers 2005; 10:13852
36. Van den Velde S, Nevens F, Van Hee P, van Steenberghe D,
Quirynen M: GC-MS analysis of breath odor compounds in
liver patients. J Chromatogr B Analyt Technol Biomed Life Sci
2008; 875:3448
37. Seok J, Warren HS, Cuenca AG, Mindrinos MN, Baker HV,
Xu W, Richards DR, McDonald-Smith GP, Gao H, Hennessy
L, Finnerty CC, Lpez CM, Honari S, Moore EE, Minei JP,
Cuschieri J, Bankey PE, Johnson JL, Sperry J, Nathens AB,
Billiar TR, West MA, Jeschke MG, Klein MB, Gamelli RL,
Gibran NS, Brownstein BH, Miller-Graziano C, Calvano
SE, Mason PH, Cobb JP, Rahme LG, Lowry SF, Maier RV,
Moldawer LL, Herndon DN, Davis RW, Xiao W, Tompkins
RG; Inflammation and Host Response to Injury, Large Scale
Collaborative Research Program: Genomic responses in
mouse models poorly mimic human inflammatory diseases.
Proc Natl Acad Sci U S A 2013; 110:350712
Extracellular Histones Play an Inflammatory Role in Acid

Aspiration-induced Acute Respiratory Distress Syndrome
Yanlin Zhang, Ph.D., M.D., Zongmei Wen, Ph.D., M.D., Li Guan, Ph.D., Ping Jiang, M.D.,
Tao Gu, M.D., Jinyuan Zhao, Ph.D., M.D., Xin Lv, Ph.D., M.D., Tao Wen, Ph.D.
ABSTRACT
Background: Systemic inflammation is a key feature in acid aspiration-induced acute respiratory distress syndrome (ARDS),
but the factors that trigger inflammation are unclear. The authors hypothesize that extracellular histones, a newly identified
inflammatory mediator, play important roles in the pathogenesis of ARDS.
Methods: The authors used a hydrochloric acid aspiration-induced ARDS model to investigate whether extracellular histones are
pathogenic and whether targeting histones are protective. Exogenous histones and antihistone antibody were administered to mice.
Heparin can bind to histones, so the authors studied whether heparin could protect from ARDS using cell and mouse models. Furthermore, the authors analyzed whether extracellular histones are clinically involved in ARDS patients caused by gastric aspiration.
Results: Extracellular histones in bronchoalveolar lavage fluid of acid-treated mice were significantly higher (1.8320.698)
at 3h after injury than in sham-treated group (0.630.153; P = 0.0252, n = 5 per group). Elevated histones may originate
from damaged lung cells and neutrophil infiltration. Exogenous histones aggravated lung injury, whereas antihistone antibody
markedly attenuated the intensity of ARDS. Notably, heparin provided a similar protective effect against ARDS. Analysis of
plasma from ARDS patients (n = 21) showed elevated histones were significantly correlated with the degree of ARDS and were
higher in nonsurvivors (2.7230.2933, n = 7) than in survivors (1.7250.1787, P = 0.006, n = 14).
Conclusion: Extracellular histones may play a contributory role toward ARDS by promoting tissue damage and systemic
inflammation and may become a novel marker reflecting disease activity. Targeting histones by neutralizing antibody or
heparin shows potent protective effects, suggesting a potentially therapeutic strategy. (Anesthesiology 2015; 122:127-39)
CUTE lung injury is a serious clinical condition characterized by acute diffuse, inflammatory lung injury
leading to enhanced alveolar-capillary permeability, edema,
hypoxemia, or hemorrhage. Acute respiratory distress
syndrome (ARDS) is the most severe form of acute lung
injury.1,2 Notably, a recent report3the Berlin Definition
recommends use of three categories of ARDS, based on the
degree of hypoxemia, to replace the definition of acute lung
injury. Currently, most treatment for ARDS is just supportive care. Of these, lung-protective ventilation and fluidconservative management have proven to be able to reduce
mortality and morbidity, respectively.4,5 However, despite
the advances in basic and clinical research, ARDS still represents a life-threatening problem among intensive care unit
patients, with an in-hospital mortality of 40% or more.1,2
The most common causes of ARDS are lung infection,
aspiration of gastric contents, sepsis, multiple trauma, and
other insults, all of which are thought to initiate a dysregulated inflammation and inappropriate accumulation and
activation of leukocytes within the lungs that lead to alveolar

Acid aspiration is known to be a cause of acute lung injury.

Extracellular histones were significantly elevated in the bronchoalveolar lavage from mice with acid-induced lung injury
versus sham mice and in human patients who died from acute
lung injury compared to survivors with acute lung injury. Extracellular histones may be causal, and targeting histones may
be a reasonable therapeutic strategy.
barrier disruption, and subsequently a severe condition such

as respiration failure.2,6 It is noticeable that systemic inflammation is a common pathological feature shared by different
etiologies-caused ARDS as demonstrated by a rapid influx
of leukocytes and releases of proinflammatory cytokines.2,7
However, the factors or mechanisms that trigger systemic
inflammation in ARDS are largely unclear. Identification of
key mediators that may modulate ARDS-associated inflammation is highly desirable.
anesthesiology.org). Drs. Zhang, Z. Wen, and Lv contributed equally to this work.
Submitted for publication January 28, 2014. Accepted for publication July 31, 2014. From the Research Center of Occupational Medicine,
Peking University Third Hospital, Beijing, Peoples Republic of China (Y.Z., L.G., J.Z.); Department of Anesthesiology, Shanghai Pulmonary
Hospital, Tongji University School of Medicine, Shanghai, Peoples Republic of China (Z.W., X.L.); Department of Emergency, Shanghai Pulmonary Hospital, Tongji University School of Medicine, Shanghai, Peoples Republic of China (P.J.); Department of Oncology, First Hospital
of Qinhuangdao, Qinhuangdao, Hebei Province, Peoples Republic of China (T.G.); and Beijing Institute of Hepatology, Beijing Youan Hospital affiliated with Capital Medical University, Beijing, Peoples Republic of China (T.W.).
January 2015
Extracellular Histones Promote ARDS
More recently, extracellular histones are recognized as a

pivotal mediator in systemic inflammatory diseases, both
infectious and noninfectious, including sepsis, acute ischemiareperfusion injuries of the kidney or liver, and traumaassociated lung injury.810 It reveals that extracellular histones
have many functions including induction of endothelial damage, coagulation activation, platelet aggregation, and cytokine production,1114 all of which are likely involved in the
pathogenesis of inflammatory organ injuries such as ARDS.
We speculate that extracellular histones play a contributory
role in the initiation and progression of ARDS and modulation of histones may have some therapeutic potential. To test
this hypothesis, we prepared a mouse model of acid aspirationinduced ARDS, which is clinically relevant model of acute
gastric aspiration injury6,15 to investigate whether extracellular histones released after lung injury have pathogenic roles.
Besides, targeting histones by specific antihistone antibody or
activated protein C has been shown to be protective.8,16 We
question whether neutralization of histones by other chemicals such as heparin may have similar beneficial effects, which
could yield new strategies for management of ARDS. Heparin is commonly used as a potent anticoagulant, but it is also
found to have antiinflammatory effects.17,18 However, the
molecular mechanism of heparin-mediated antiinflammation
remains to be defined. It reveals that histones have high affinity toward heparin19 and some histone-induced effects such
as platelet aggregation could be inhibited in the presence of
heparin.11 Therefore, we sought to examine whether heparin
has the potential to attenuate acid aspiration-induced lung
injury through binding to histones. Furthermore, we recruited
21patients with ARDS caused by aspiration of gastric contents to analyze the clinical relevance of extracellular histones.
Induction of Lung Injury with Acid Aspiration in Mice

Before acid aspiration, mice fasted overnight but were
allowed water ad libitum. For the induction of acid-induced
lung injury, mice were anesthetized using sodium pentobarbital (50mg/kg), and injured by intratracheal instillation of
hydrochloric acid (HCl, 0.1N, pH 1.5) into the lung via a
tracheal catheter. The sham mice underwent the same procedure but received sterile saline instead.

Reagents
Calf thymus histones from Sigma-Aldrich (St Louis, MO)
and human recombinant histone H4 from New England
BioLabs (Ipswich, MA) were obtained. Heparin-sodium
was obtained from Sigma-Aldrich. Mouse antihistone H4
mAb was prepared following the previous protocol involving
autoimmune mice.16,20 Histone H4 ELISA (enzyme-linked
immunosorbent assay) kit was obtained from USCN Life
Science, Inc. (Wuhan, China).
Preparation of Mouse Samples

In another group of mice, the lungs were flushed with 1ml
phosphate-buffered saline to obtain bronchoalveolar lavage
fluid (BALF). BALF was centrifuged at 1,000g for 10min
at 4C and supernatants were stored at 80C until further
analysis. The lung lobes were rapidly excised, flash frozen in
liquid nitrogen, and stored at 80C. Mouse blood was collected by retro-orbital bleeding in a tube containing sodium
citrate as anticoagulant and centrifuged to separate plasma,
and then kept at 80C.
Animals
Eight to ten week-old male C57BL/6 mice, weighing 25
30g, were purchased from the Experimental Animal Center of Peking University (Peking, China). Mice were housed
in an air-conditioned room at 25C with a 12h darklight
cycle and allowed to acclimate upon arrival for 3 days before
experimentation. All experimental protocols of this study
were approved by the Institutional Animal Care and Use
Committee of Health Sciences Center, Peking University,
Beijing, Peoples Republic of China.
Measurement of Extracellular Histones in BALF and

Plasma of Mice
As nucleosomes are complexed with histones and DNA,
assaying the concentrations of nucleosomes allows for the
relative quantification of histones.16,21 We first measured the
concentrations of nucleosome in BALF and plasma of mice
using a Cell Death Detection ELISA kit (Roche Applied
Science, Mannheim, Germany). We further characterized
extracellular histones, especially histone H4 by ELISA and
Western blotting because histone H4 is believed to play a
Assessment of ARDS in Mice

The extent of ARDS was assessed by blood gas analysis,
pulmonary edema, and lung histology. An abdominal aorta
catheter with sodium citrate was inserted to obtain arterial
blood and arterial partial oxygen tension (PaO2) was analyzed with a gas analyzer (Ciba Corning-170 blood gas analyzer, Ciba Corning, Etobicoke, ON, Canada) at different
time points after acid aspiration.
The ratio of wet to dry lung weights was calculated to
assess pulmonary edema. Lung tissues were rapidly excised
and rinsed in phosphate-buffered saline (PBS) to remove contaminating blood. After removal of excessive phosphate-buffered saline by careful drying on tissue paper, lung tissues were
weighed (wet weight). Then the tissues were dried in an oven
at 60 for 72h, followed by a second weighing (dry weight).
For lung histology, part of lung tissues were fixed with
10% buffered formalin, embedded in paraffin, and 5 m
sections were obtained and stained with hematoxylin and
eosin. The stained sections were evaluated and scored by
pathologists who were blinded to the experimental protocol. The severity of lung injury was scored according to the
following parameters: hemorrhage, alveolar edema, alveolar
exudates, necrosis, and leukocyte infiltration.
Anesthesiology 2015; 122:127-39 128 Zhang et al.
central role in mediating cytotoxicity in contrast to other

individual histones including H1, H2A, H2B, and H38,9.
Quantification of Lactate Dehydrogenase Activity and
Neutrophil Activation
It has been suggested that extracellular histones can be released
from severely injured tissues or inflammatory leukocytes, especially neutrophils. Lactate dehydrogenase (LDH) is a cytoplasmic enzyme and its activity reflects the degree of tissue damage.
We measured LDH activity in BALF and plasma of mice
using an Automated Multi-parameteric Analyzer (AU 5400,
Olympus, Tokyo, Japan) according to an automated procedure. For determination of neutrophil activation, we assayed
myeloperoxidase activity, an index of neutrophil, monocyte/
macrophage infiltration,22,23 in BALF and plasma of mice
using a commercial kit (BioVision, Milpitas, CA) according
to the manufacturers recommended protocol. In addition, to
strengthen the specificity, we stained paraffin-embedded lung
sections with anti-Ly6G (Abcam, Cambridge, United Kingdom), which is commonly used as a surface marker for neutrophils24 to visualize neutrophil infiltration.
Treatment of Exogenous Histones, Antihistone Antibody,
and Heparin
In a separate set of experiment, we sought to examine whether
histones released after acid-caused lung injury play a major
role in the pathogenesis of ARDS. To this end, a mixture of
all individual histones (20mg/kg) isolated from calf thymus
or human recombinant histone H4 (5mg/kg) was given by
pulmonary instillation to mice shortly after HCl challenge.
Meanwhile, another group of mice received antihistone H4
mAb (20mg/kg) intravenously or heparin (250 IU/kg) by
subcutaneous injection immediately after acid aspiration,
with an aim of further confirming the pathogenic role of
histones as well as exploring possible treatment strategies.
The dosage of histones, antihistone H4 mAb, or heparin was
adopted on the basis of our previous experiments.
Measurement of Cytokines
Frozen lung tissues of mice were homogenized in lysis buffer containing 50mM Tris-HCl, 150mM NaCl, 0.1% SDS,
0.5% sodium deoxycholate and cocktail protease inhibitors.
Lung homogenates were centrifuged at 10,000g at 4C for
15min, and supernatants were stored at 80C until assays
were performed. We quantified interleukin (IL)-1, IL-6,
IL-10 and tumor necrosis factor (TNF)- levels in BALF,
plasma and lung homogenates using the ProcartaPlex Multiplex Immunoassay from eBioscience (Vienna, Austria),
according to the manufacturers protocol.
Assay for Histone Cytotoxicity In Vitro
Human lung epithelial cells (BEAS-2B) were obtained from
American Type Culture Collection (ATCC: CRL9609).
They were cultured in Dulbeccos modified eagle medium
supplemented with 10% fetal bovine serum, and 1%
penicillin-streptomycin. After the cells grew to 7080%

confluence, they were stimulated with calf thymus histones
(50 g/ml) or recombinant histone H4 (20 g/ml) for 1h at
37. In another set of experiment, calf thymus histones or
recombinant histone H4 were preincubated with antihistone
H4 antibody (20 g/ml) or heparin (200/ml) for 1h at
room temperature before administration to cells. Cells were
harvested and stained with propidium iodide to analyze cell
damage via flow cytometry. In addition, cell culture supernatants were analyzed for IL-1, IL-6, IL-10, and TNF-
levels using the ProcartaPlex Multiplex Immunoassay.
Measurement of Extracellular Histones and Inflammatory
Markers in Patients with ARDS
A total of 21 ARDS patients caused by aspiration of gastric
contents who were admitted to the emergency ward were
recruited from Shanghai Pulmonary Hospital, Tongji University School of Medicine, Shanghai, Peoples Republic of
China. These patients met the criteria for ARDS defined by
the Berlin Definition.3 Inclusion criteria were as follows: (1)
acute onset of lung injury, (2) arterial oxygen tension/fraction of inspired oxygen (PaO2/FiO2) less than or equal to
300 mmHg, (3) infiltrative shadow seen on chest radiography. All the patients were observed early in the course of
ARDS. Patients who had sepsis or ARDS for more than 3
days were not included. All subjects or their relatives gave
their informed consent for inclusion before they participated
in the study. The study was approved by the Ethics Committee and Institutional Review Board of Shanghai Pulmonary
Hospital, Shanghai, Peoples Republic of China, which followed the recommendations of the Declaration of Helsinki
for biomedical research involving human subjects. Fifteen
age- and sex-matched healthy volunteers were included as
controls. The blood samples were collected from patients at
admission, plasma was separated, aliquoted and stored at
80C for future analysis. Given that the incidence of gastric
aspiration-induced ARDS is relatively low, and it was just a
preliminary evaluation instead of a randomized, controlled
study for human patients, so the sample-size calculation of
patients was not conducted.
For animal survival analysis, the Log-rank (Mantel-Cox)
test was applied, with all data censored by 24h. All values
were expressed as mean SD. Data were analyzed using
unpaired Student t test (for two groups), one-way analysis
of variance followed by Tukey post-tests (for more than two
groups). In addition, data that were obtained at multiple
time points throughout the experiment were analyzed using
a two-way analysis of variance for PaO2, lung dry/wet ratio,
and nucleosomes/histones, LDH, and myeloperoxidase,
respectively. Multiple comparisons were corrected by Bonferroni post-tests, in which group allocation (acid aspiration
vs. sham) and time were treated as two factors. Correlations
between variables were assessed using Pearson correlation
analysis. The number of mice in each group was selected

based on our preliminary experiments. The samples sizes (n)
are indicated in the figures. Results were considered statistically significant when P < 0.05. All statistical analyses were
calculated using GraphPad Prism v5 (GraphPad Software,
Inc., San Diego, CA).
Results
Extracellular Histones Are Dramatically Released in Mice
with Acid Aspiration-induced Lung Injury
It showed that administration of HCl to mice produced significant lung injury, as demonstrated by the decreased PaO2
and a remarkable increase of lung wet/dry weight ratio, both
of which were viewed as the hallmark of ARDS2,4 (fig.1, A
and B). Lung histological examination at 6h after HCl challenge showed multifocal alveolar hemorrhage, disruption of
alveolar wall, and massive infiltration of inflammatory cells
in mice (fig. 1C). After lung injury, the levels of extracellular
nucleosome that consist of histones and DNA were found to
be increased in BALF of mice, with parallel results obtained
from mice plasma (fig. 1, D and E). The release of nucleosome into BALF and plasma occurred in a time-dependent
manner, which correlated with the severity of ARDS.
Further analysis of histones revealed that individual histone H2A, H2B, H3, and H4 were simultaneously detected
in plasma of mice with ARDS (data not shown and fig. 1F).
Considering that histone H4 is a primary factor exerting
toxic effects,8,16 we further measured the concentrations of
histone H4 by ELISA in BALF and plasma of mice. Parallel
BALF levels and plasma levels of histone H4 were observed
to be increased markedly in mice with ARDS (fig. 1, G and
H), which was in line with the changes of nucleosomes.
Both assays for nucleosome and histone H4 were comparable (BALF and plasma nucleosome vs. BALF and plasma
histone H4: Pearson correlation coefficient r = 0.7117,
P < 0.0001; r = 0.6368, P= 0.0002), indicating that nucleosome or histone H4 can be a shared marker for ARDS.
Collectively, these data suggest that large quantities of
nucleosomes are released into circulation during the course
of severe lung injury, which can be presented in the form of
individual histones including H4.
Massive Injured Tissue Cells and Neutrophil Activation
may Cause the Released Extracellular Histones
Extracellular histones can be derived from dying tissue cells
or from the degradation of inflammatory cells.9,25 Neutrophil
extracellular traps formed in response to systemic inflammation may be an important source of released histones
found in the circulation.23,26,27 We first assayed LDH activity that reflects the degree of tissue damage, and as expected
we observed a time-dependent increase of LDH activity in
BALF and plasma of acid aspiration-treated mice (fig.2, A
and B), suggesting an occurrence of drastic cell death. Then,
we measured neutrophil activation, which is considered to
be a major mechanism for inflammation associated with
lung injury27,28 by measuring myeloperoxidase activity in

BALF and plasma of mice. After acid aspiration, parallel
BALF and plasma myeloperoxidase activities were found to
be increased markedly in a time-dependent manner (fig. 2, C
and D). Moreover, immunohistochemical analysis of Ly6G
staining confirmed a significant neutrophil infiltration at 6h
in the lung sections of acid-treated mice (fig. 2E). There was
a strong correlation of BALF nucleosome levels with BALF
LDH (r = 0.7406, P < 0.0001), and with BALF myeloperoxidase activity (r = 0.7080, P < 0.0001) (Supplemental Digital
Content 1, fig. 1, http://links.lww.com/ALN/B88). Taken
together, it suggested that the increased levels of extracellular
histones were likely derived from dying lung tissue cells as
well as neutrophil infiltration in response to HCl challenge
in mice.
Exogenous Histones Aggravate Tissue Damage and
Systemic Inflammation in Mice with ARDS
In further experiments, we aimed to ascertain whether extracellular histones released during the course of ARDS contributed to inflammation and lung injury in this model.
We showed that pulmonary instillation of low doses of
calf thymus histones (20mg/kg) or recombinant histone
H4 (5mg/kg) alone resulted in mild lung damage in mice,
as evidenced by an increase of BALF LDH levels at 6h
(431.247.88 or 366.327.02) in contrast to sham-treated
mice(213.422.59; P = 0.007, P = 0.004) (fig.3A). Notably,
administration of exogenous histone mixture or recombinant
histone H4 in conjunction with HCl significantly increased
the death rate of mice, as compared with only HCl-treated
group (fig. 3, B and C). Histone mixture or recombinant histone H4 remarkably aggravated acid-induced lung injury and
systemic inflammation in mice, as evidenced by enhanced
lung injury scores (fig. 3D) and the increased levels of inflammatory cytokines such as TNF-, IL-1, IL-6 and IL-10 in
BALF and lung homogenates (fig. 3, EH and Supplemental Digital Content 1, fig. 2, http://links.lww.com/ALN/
B88). Several of these cytokines are known to be associated
with development and progression of ARDS in humans.29,30
Therefore, our data confirmed that extracellular histones
released from lung cell death or neutrophil infiltration are
a major contributor to ARDS in mice after acid aspiration.
Neutralization of Histones Is Protective in Mice with Lung
Injury
Targeting extracellular histones by specific neutralizing antibody (e.g., antihistone H4 antibody) or activated protein
C was previously shown to be protective in several inflammatory conditions.8,9 In this study, we sought to examine
whether heparin, which is capable of binding to histones19,31
(Supplemental Digital Content 1, fig. 3, http://links.lww.
com/ALN/B88) has similarly protective potentials. Antihistone H4 antibody was administered separately as a positive
control. It showed that either heparin or antihistone antibody significantly reduced lung injury in acid-treated mice,
Fig. 1. HCl aspiration caused ARDS as well as elevated histone (nucleosome) levels in BALF and plasma of mice. (A) 0.1 N HCl
was instilled intratracheally in mice and caused a significant decrease in PaO2 in a time-dependent manner (mean SD, n =
56 mice/group).*P < 0.01 vs. the sham group. (B) HCl aspiration caused a significant increase in lung wetdry ratio of mice
(mean SD, n = 56 mice per group).*P < 0.01 versus the sham group. (C) Hematoxylin and eosinstained sections of lung
damage at 6h after HCl administration to mice. (a) sham-treated mice; (b) HCl-treated mice. Obvious pathological changes
were observed in the lungs of HCl-treated mice, such as alveolar hemorrhage, infiltration of neutrophils. Scale bars: 100 m.
(D) BALF nucleosomes and (E) plasma nucleosomes were measured using ELISA in mice. (F) An example of histone H4 presence was detected by Western blot in the plasma of mice after HCl aspiration. The blots are representative of at least three
independent experiments. (G) BALF histone H4 and (H) plasma histone H4 were measured using ELISA in mice. These levels
were increased significantly after HCl aspiration in a time-dependent manner (mean SD, n = 56 mice per group). *P < 0.01
versus the shams. ARDS = acute respiratory distress syndrome; BALF = bronchoalveolar lavage fluid; ELISA = enzyme-linked
immunosorbent assay; HCl = hydrochloric acid; PaO2 = partial oxygen tension.
demonstrated by improved lung histology (fig.4, A and B).

Parallel BALF and tissue levels of TNF-, IL-1, IL-6, and
IL-10 levels were found to be reduced remarkably with blockade of extracellular histones (fig. 4, CF and Supplemental
Digital Content 1, fig. 4, http://links.lww.com/ALN/B88).
Collectively, this confirms that extracellular histones play a
detrimental role in the progression of ARDS and targeting
histones by neutralizing antibody or heparin is protective.
In Vitro Studies
We further elucidated whether heparin mediated-protection is
dependent on binding to extracellular histones. To address this
issue, we first stimulated human lung epithelial cells (BEAS2B) with calf thymus histones or recombinant histone H4 and
observed that administration of exogenous histones caused

enhanced rate of cell death, confirming that extracellular histones possess direct cytotoxic effects. Preincubation of exogenous
histones or recombinant histone H4 with heparin lowered cytotoxicity, as demonstrated by significantly improved cell viability
rate and reduced inflammatory cytokines levels in the supernatant of culture medium (fig.5). Antihistone antibody was also
administered as a positive control. These results confirmed that
heparin mediates cytoprotection via targeting histones.
Extracellular Histones Indicate Disease Severity in
Patients with Gastric Aspiration-induced ARDS
Having shown that extracellular histones were released
notably in mice with acid-induced ARDS and aggravated
Fig. 2. Extracellular histones were likely derived from dying lung cells and infiltrated neutrophils. Quantification of (A) BALF
and (B) plasma LDH activity, a marker reflecting tissue damage, was determined using an Automated Multiparameteric Analyzer. Quantification of (C) BALF and (D) plasma myeloperoxidase activity, a marker of granules in neutrophils and monocytes,
was determined by ELISA. Both LDH and myeloperoxidase activity were increased remarkably in mice after HCl aspiration.
*P < 0.01 vs. the sham-treated mice. Immunohistochemical analysis of Ly6G staining at 6h in the lung sections of (E ) shamtreated mice and (F) HCl-treated mice. Large number of Ly6G-stained cells was observed in the lungs of HCl-treated mice,
suggesting massive neutrophil infiltration. Arrowheads indicate positive cells. Scale bars: 50 m. There was a strong correlation
of BALF nucleosome levels with BALF, LDH, and BALF myeloperoxidase, suggesting that cell death and neutrophils, monocytes, or both contributed to the release of nulceosomes/histones in mice (Supplemental Digital Content fig. 1, http://links.lww.
com/ALN/B88). BALF = bronchoalveolar lavage fluid; ELISA = enzyme-linked immunosorbent assay; HCl = hydrochloric acid;
LDH = lactate dehydrogenase.
systemic inflammation, we checked whether this is similar

to gastric aspiration-induced ARDS in patients. Totally, 21
ARDS patients (15 men and 6 women with a mean age of
588.3 years) caused by gastric aspiration were enrolled in
the study. These patients were evidenced to have had gastric aspiration events, and diagnosed by identification of an
infiltrative shadow on a chest x-ray, and the relevant aspiration signs and symptoms including wheezing, shortness of
breath, cyanosis, hypotension, and hypoxia. Baseline characteristics, demographic details, and biochemical parameters
of each patient were routinely recorded and summarized
in Supplemental Digital Content 1, table 1, http://links.
lww.com/ALN/B88. Based on the Berlin definition,3 these

patients were categorized into three groups according to
degree of hypoxemia: mild (200 mmHg < PaO2/FiO2
300 mmHg, n = 6), moderate (100 mmHg < PaO2/FiO2
200 mmHg, n = 10), and severe (PaO2/FiO2 100 mmHg,
n = 5). Among these patients, seven (five men and two
women) died within 24h of injury.
We measured levels of extracellular nucleosomes/histone
H4, LDH, myeloperoxidase, and inflammatory cytokines
in the plasma of these patients. It showed that mean extracellular nucleosome or histone H4 levels were significantly
higher in ARDS patients (2.1090.179 or 14.781.134)
Fig. 3. Exogenous histones aggravated lung damage and inflammation in mice with ARDS. (A) Pulmonary instillation of exogenous
histone mixture (20mg/kg) or recombinant histone H4 (5mg/kg) caused mild lung damage in mice, evidenced by an increase of
BALF LDH levels at 6h in contrast to sham-treated mice. (B) Histone mixture or (C) histone H4 plus HCl significantly decreased
the survival rate of HCl-treated mice (*P < 0.05). (D) Lung histological scores were higher in histone mixture or histone H4 plus
HCl groups than in HCl alone group (mean SD). BALF (E) TNF- (F) IL-1 (G) IL-6 (H) IL-10 levels were all notably increased
in histones plus HCl groups as compared with HCl-treated mice (mean SD). ARDS = acute respiratory distress syndrome;
BALF = bronchoalveolar lavage fluid; HCl = hydrochloric acid; IL-1 = interleukin-1; IL-6 = interleukin 6; IL-10 = interleukin-10;
LDH = lactate dehydrogenase; TNF- = tumor necrosis factor-.
as compared with healthy controls (0.526
0.513 or
0.4660.0765; both P < 0.0001). Moreover, their levels
correlated well with the degree of ARDS (all P < 0.05; fig.6,
A and B). Notably, nucleosome or histone H4 levels in nonsurviving patients (2.7230.2933 or 17.581.555) were
significantly higher than those in survivors (1.7250.1787
or 11.811.261, P = 0.006 or P = 0.012; fig. 6, C and
D). We also observed that LDH, myeloperoxidase, and
inflammatory cytokines including TNF-, IL-1, IL-6,
and IL-10 were all increased markedly in patients with
ARDS (Supplemental Digital Content 1, fig. 4 and fig. 5,

http://links.lww.com/ALN/B88). However, their levels did
not correlate with the degree of ARDS, except for myeloperoxidase activity between moderate and severe ARDS
(589.4150.4 vs. 1427321.2, P=0.017; Supplemental
Digital Content 1, fig. 5C, http://links.lww.com/ALN/
B88). Moreover, there were significant differences in the
levels of myeloperoxidase (551.8135.3 vs.1249355.4,
P = 0.0378), TNF- (19.724.21 vs.56.139.42, P =
0.0006), and IL-6 (27.253.17 vs. 46.610.94, P=0.041)
Fig. 4. Antihistone antibody or heparin provides similar protection against ARDS. (A) Administration of antihistone H4 antibody
or heparin significantly reduced lung damage of HCl-treated mice. The representative of H&E-stained sections indicated a significant improvement in the lung of antihistone antibody or heparin-treated mice at 6h. (a) sham-treated mice, (b) HCl-treated
mice, (c) antihistone antibody plus HCl-treated mice, (d) heparin plus HCl-treated mice. Scale bars: 100 m. (B) Lung histological
scores were lower in antihistone antibody or heparin plus HCl groups than in HCl alone group (*P < 0.01). BALF (C) TNF- (D) IL1 and (E) IL-6 were notably decreased in antihistone antibody- or heparin-treated groups as compared with HCl-treated group
(*P < 0.05). (F) There was no difference in IL-10 levels between antihistone antibody or heparin-treated groups and HCl-treated
group. ARDS = acute respiratory distress syndrome; HCl = hydrochloric acid; H&E = hematoxylin and eosin; IL-1 = interleukin1; IL-6 = interleukin-6; IL-10 = interleukin-10; TNF- = tumor necrosis factor-.
between survivors and nonsurviving ARDS patients (Supplemental Digital Content 1, figs. 5 and 6, http://links.
lww.com/ALN/B88). There was a significant correlation
of plasma nucleosomes with plasma myeloperoxidase (r =
0.4905, P =0.025), and IL-6 (r = 0.6158, P = 0.002) levels
(table1), suggesting an inflammatory role of extracellular
nucleosomes/histones in ARDS. We concluded that extracellular histones could reflect disease severity of ARDS in
clinical situations.
Discussion
So far, ARDS is still the leading cause for respiratory failure in critically ill patients with high morbidity and mortality.1,2 Aspiration of gastric contents is one of the major
causes for ARDS.6,21,32 There is a broad range of conditions that predispose to gastric aspiration-induced ARDS
(e.g., general anesthesia, alcohol and narcotic abuse, and
neurologic disorders). Despite therapy, 3050% of gastric
aspiration-induced ARDS patients die as a result of respiratory failure.6 However, unlike sepsis-associated ARDS, the
underlying mechanism for acid aspiration-induced ARDS is
less well-studied.5
Accumulating evidence reveals that uncontrolled and

persistent inflammation is deeply implicated in the progress
of ARDS, characterized by massive leukocyte activation,
free radical production, and cytokine release.29 Moreover,
damage-associated molecular pattern molecules such as
extracellular histones, high mobility group box-1 protein,
mitochondrial DNA, and formyl peptides have been suggested as a primary culprit of uncontrolled inflammation in
addition to pathogen.3335 Among these, extracellular histones were particularly identified as a novel damage-associated
molecular pattern molecule involved in a series of inflammatory events.9,25 Histones are a group of nuclear proteins that
form heterooctamers to wind up the double-stranded DNA
to form nucleosomes.34 Histones play important roles in the
regulation of DNA repair, gene transcription, and chromatin condensation.16 In some cases, histones in the form of
nucleosome fragments can be detected in the cytoplasm or
even in the extracellular milieu such as peripheral circulation. Increased concentrations of histones are observed
in the circulation of patients with myocardial infarction,
stroke, infections, trauma, cancer and autoimmune diseases.26,34,3638 However, the roles of extracellular histones are
not well studied, and whether extracellular histones merely
Fig. 5. Heparin can inhibit histone-mediated cell death in vitro. (A) PI staining of cell death via flow cytometry indicated that preincubation of exogenous histones or histone H4 with antihistone antibody or heparin could significantly decrease cell death in
vitro. (B) Administration of histone mixture or recombinant histone H4 increased the levels of TNF-, IL-1, IL-6, and IL-10 in the
supernatant of culture medium, whereas antihistone antibody or heparin preincubation remarkably decreased these cytokines
(*P < 0.01). IL-1 = interleukin-1; IL-6 = interleukin-6; IL-10 = interleukin-10; PI = propidium iodide; TNF- = tumor necrosis factor-.
act as bystanders or are active mediators of disease remains

unclear. Recently Xu et al. 8 for the first time reported that
extracellular histones are key mediators of cell damage and
organ dysfunction during the hyperinflammatory reactions
such as sepsis. Extracellular histones were increased markedly in response to sepsis and induced endothelial cytotoxicity and triggered an inflammatory and thrombotic response
that eventually led to multiple organ dysfunction syndrome
and death.8 Subsequently Huang et al. 10 studied the role
of extracellular histones in sterile inflammatory liver injury
and showed that endogenous histones derived from necrotic
hepatocytes following hepatic ischemiareperfusion injury
serve as a crucial link between initial tissue damage and
activation of inflammation. Moreover Allam et al. 9 revealed
that circulating histones from dying renal cells aggravated
acute kidney injury in mice and neutralization of histones
using a mAb significantly protected against injury. Fuchs
et al. found that circulating histones could directly activate

platelets and lead to thrombocytopenia in vivo, which may
be another important mechanism of organ damage.11 All of
these findings indicate a strikingly novel role of extracellular
histones in systemic inflammation and organ dysfunction.
However, the role of extracellular histones in the pathogenesis of ARDS caused by acid aspiration is still not clear.
Here, we first prepared a murine model of HCl aspiration-induced ARDS to study the role of extracellular histones, as aspiration of gastric acid is clinically relevant and is
associated with high mortality rates of ARDS.15,29,30 HCl has
been reported to initially damage pulmonary airway epithelia, which in turn triggers an inflammatory response followed
by edema formation and disruption of the alveolar capillary membrane.5,29,30 We demonstrated that HCl aspiration
caused severe lung injury in mice, together with the significantly increased levels of extracellular histones in BALF and
Fig. 6. Extracellular histones were observed to increase in patients with ARDS caused by gastric aspiration. Plasma (A) nucleosome (B) histone H4 levels were significantly higher in ARDS patients than in healthy controls (*P < 0.001). There were also
statistical differences in the levels of nucleosome or histone H4 between mild, moderate and severe ARDS (*P < 0.05). Plasma
(C) nucleosomes (D) histone H4 levels were significantly higher in nonsurviving patients than in survivors (*P < 0.05). ARDS =
acute respiratory distress syndrome.
Table 1. Correlation of Plasma Nucleosomes with Various
Variables in ARDS Patients
ARDS (n = 21)
LDH
Myeloperoxidase
TNF-
IL-1
IL-6
IL-10
P Value
0.2078
0.4905
0.3376
0.3104
0.6158
0.2108
0.3987
0.025*
0.0719
0.1142
0.002*
0.4528
* P < 0.05 was considered to be statistically significant.

ARDS = acute respiratory distress syndrome; IL = interleukin; LDH= lactate
degydrogenase; TNF- = tumor necrosis factor-.
plasma of mice, which correlated with the severity of ARDS.

We next determined the source of extracellular histones by
assaying LDH activity and neutrophil infiltration. LDH is a
classic marker representing the extent of tissue damage. The
significantly increased levels of LDH in BALF and plasma
suggested an occurrence of prominent damage of the lungs
during the course of acid-induced ARDS. Myeloperoxidase is a heme protein stored in granules of neutrophils and
monocytes, and thus is a classic marker for neutrophil or
macrophage infiltration.22 We showed that parallel BALF
and plasma myeloperoxidase activities were elevated markedly after lung injury, indicating a significant activation of
neutrophils or macrophages. Besides, immunohistochemical
staining of Ly6G, a commonly used marker for neutrophils,

indicated a prominent infiltration of neutrophils, further
supporting these findings. Collectively, we concluded that
the elevated levels of extracellular histones might originate
from dying lung cells or neutrophil infiltration in this model.
Large quantities of histones in the extracellular milieu
have pathological importance. It is reported that histones
have direct cytotoxicity by binding phospholipids, disrupting cell membranes, and causing calcium influx.25,27 In addition, histones can serve as a damage-associated molecular
pattern molecule to promote inflammation by mediating
downstream inflammatory responses leading to cytokine
production.9,25 Consistent with these studies, we found that
a low dose of exogenous histones or recombinant histone
H4 aggravated HCl-induced ARDS in mice, whereas blockade of histones by either a neutralizing antibody or heparin provided significant protection against ARDS. It thus
confirmed an important pathological and targetable role of
extracellular histones in acid-induced ARDS. Based on our
results, we propose a hypothetical role of extracellular histones in mediating inflammation and lung injury after acid
challenge (fig.7). In brief, dying lung tissue cells as well as
neutrophil infiltration in response to HCl challenge release
large quantities of histones, which are directly cytotoxic or
act as an endogenous damage-associated molecular pattern
molecule to promote innate immunity and systemic inflammation. Furthermore, histones may in turn attract more
innate cells to inflammation site and amplify inflammation
Fig. 7. The hypothesized model of extracellular histones in mediating inflammation and lung injury after acid challenge. It is proposed that extracellular histones may originate from cell damage or infiltrated neutrophils as a result of acid aspiration, causing
direct cytotoxicity and activation of innate immunity, which in turn attracts more inflammatory cells and amplify inflammation
by promoting cytokine production that eventually contribute to the pathogenesis of ARDS. ARDS = acute respiratory distress
syndrome; DAMPs = damage associated molecular patterns.
by promoting more cytokine production, all of which eventually add an important factor contributing to ARDS (fig.
7).
Notably, heparin provides a similar protective effect as a
neutralizing histone H4 antibody in vivo and in vitro experiments. Heparin is a glycosaminoglycan well known for its
anticoagulant properties.17 Heparin also possesses antiinflammatory effects and has been successfully used for the
treatment of inflammatory conditions such as ischemia
reperfusion injury.18,39,40 However, the mechanisms responsible for the antiinflammatory effects of heparin are not well
understood. It has been reported that heparin binds to many
cellular proteins including histones through electrostatic
interactions of high affinity.19 This occurs because heparin
is highly sulfated and has strong negative charges, whereas
histones are positively charged. However, no one attributed
the binding of heparin to histones to its antiinflammation
property before histones were revealed to be an inflammatory mediator. In this study, we reported that heparin-treated
mice with ARDS had a significantly improved survival rate
and reduced lung inflammation compared to control mice,
and heparin directly interfered with histone-induced cell
death in human lung epithelial cells. We thus raise the possibility that the binding of heparin to histones is responsible
for the antiinflammatory effects of heparin, which adds a

novel explanation for the underlying mechanisms related
to heparin-mediated antiinflammation. It also suggests that
drugs targeting histones may be a therapeutic option in
patients with ARDS. However, a concern that heparin treatment could increase bleeding risk in patients should not be
ignored. To address such an issue, we intend to explore the
protective effect of nonanticoagulant N-Acetyl heparin in
the future, which excludes the anticoagulant property while
retaining its antiinflammation ability.41
ARDS remains a serious clinical concern around the
world. In some cases, patients after gastric aspiration events
are at high risk of pulmonary dysfunction and the development of severe ARDS.1,7,30 Over the years, no therapeutic agents have demonstrated a clear benefit during ARDS
treatment. Here, we questioned whether there was a clinical
relevance of extracellular histones in patients with ARDS
caused by acid aspiration. We found that extracellular histones were low or even undetectable in healthy subjects but
increased drastically in ARDS patients and the levels of histones correlated well with the degree of ARDS. Moreover,
we observed the statistical difference in plasma histone levels
among nonsurvivors or survivors. Thus, we conclude that
extracellular histones may serve as an important marker for
clinically monitoring the severity of lung injury in humans.

Animal studies suggest that extracellular histones could
become a potential therapeutic target. So it is conceivable
that therapeutic strategies targeting histones will provide
novel pharmacological approaches to treat ARDS in humans.
Antihistone therapies such as administration of neutralizing
antibody, activated protein C or heparin may complement
the available strategies for management of ARDS. But currently, no relevant data is available regarding whether ARDS
patients receiving heparin or activated protein C intervention
are protected. Future epidemiological studies are needed to
address this possibility. Likewise, the development of specific
neutralizing antihistone antibody is warranted.
In summary, our data clearly show that extracellular histones play an inflammatory role in acid aspiration-induced
ARDS by stimulating systemic inflammation and contributing to lung damage. The levels of extracellular histones may
act as a novel marker indicating disease activity in mice and
humans with ARDS. Blockade of histones by specific neutralizing antibody or heparin may be a potentially therapeutic strategy in the treatment of ARDS.
Acknowledgments
The authors thank Jun Xu, Ph.D. (Cardiovascular Biology
Research Program, Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma), for technical support. The
authors also thank Xiaochen Shu, Ph.D. (Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden), for his assistance in statistical analysis.
This work was supported by grants from the National
Natural Science Foundation of China (grant nos. 81272142,
81372094, 81400052), Beijing, Peoples Republic of China;
and by Beijing Municipal Health System High-level Talent
Training Project (to Dr. Tao Wen), Beijing, Peoples Republic
of China. This work was also funded by Shanghai Municipal
Committee of Science and Technology (14ZR1434400, to Dr.
Zongmei Wen), Shanghai, Peoples Republic of China, and
by Pujiang Talent Program (14PJD030, to Dr. Zongmei Wen),
Shanghai, Peoples Republic of China.
Competing Interests
Correspondence
Address correspondence to Dr. Tao Wen: Beijing Institute of
Hepatology, Beijing Youan Hospital affiliated with Capital
Medical University, No. 8 Xitouiao, Youan Men Wai, Beijing
100069, Peoples Republic of China. wentao528@gmail.com.
References
1. Matthay MA, Zimmerman GA: Acute lung injury and the
acute respiratory distress syndrome: Four decades of inquiry
into pathogenesis and rational management. Am J Respir Cell
Mol Biol 2005; 33:31927
2. Matthay MA: The acute respiratory distress syndrome.

Pathogenesis and treatment. Annu Rev Pathol 2011; 6:14763
3. Ranieri VM, Rubenfeld GD, Thompson BT, Ferguson ND,
Caldwell E, Fan E, Camporota L, Slutsky AS: Acute respiratory distress syndrome. The Berlin Definition. JAMA 2012;
307:252633
4. Matthay MA, Ware LB, Zimmerman GA: The acute respiratory
distress syndrome. J Clin Invest 2012; 122:273140
5. Ambrosio AM, Luo R, Fantoni DT, Gutierres C, Lu Q, Gu
WJ, Otsuki DA, Malbouisson LM, Auler JO, Jr, Rouby JJ;
Experimental ARDS Study Group: Effects of positive end-expiratory pressure titration and recruitment maneuver on lung
inflammation and hyperinflation in experimental acid aspiration-induced lung injury. Anesthesiology 2012; 117:132234
6. Raghavendran K, Nemzek J, Napolitano LM, Knight PR:
Aspiration-induced lung injury. Crit Care Med 2011; 39:81826
7. Matt U, Warszawska JM, Bauer M, Dietl W, Mesteri I,
Doninger B, Haslinger I, Schabbauer G, Perkmann T, Binder
CJ, Reingruber S, Petzelbauer P, Knapp S: Bbeta(15-42) protects against acid-induced acute lung injury and secondary
pseudomonas pneumonia in vivo. Am J Respir Crit Care Med
2009; 180:120817
8. Xu J, Zhang X, Pelayo R, Monestier M, Ammollo CT, Semeraro
F, Taylor FB, Esmon NL, Lupu F, Esmon CT: Extracellular histones are major mediators of death in sepsis. Nat Med 2009;
15:131821
9. Allam R, Scherbaum CR, Darisipudi MN, Mulay SR, Hgele
H, Lichtnekert J, Hagemann JH, Rupanagudi KV, Ryu M,
Schwarzenberger C, Hohenstein B, Hugo C, Uhl B, Reichel
CA, Krombach F, Monestier M, Liapis H, Moreth K, Schaefer L,
Anders HJ: Histones from dying renal cells aggravate kidney
injury via TLR2 and TLR4. J Am Soc Nephrol 2012; 23:137588
10. Huang H, Evankovich J, Yan W, Nace G, Zhang L, Ross M,
Liao X, Billiar T, Xu J, Esmon CT, Tsung A: Endogenous
histones function as alarmins in sterile inflammatory liver
injury through Toll-like receptor 9 in mice. Hepatology 2011;
54:9991008
11. Fuchs TA, Bhandari AA, Wagner DD: Histones induce rapid and
profound thrombocytopenia in mice. Blood 2011; 118:370814
12. Semeraro F, Ammollo CT, Morrissey JH, Dale GL, Friese

P, Esmon NL, Esmon CT: Extracellular histones promote
thrombin generation through platelet-dependent mechanisms: Involvement of platelet TLR2 and TLR4. Blood 2011;
118:195261
13. Yamamichi S, Fujiwara Y, Kikuchi T, Nishitani M, Matsushita
Y, Hasumi K: Extracellular histone induces plasma hyaluronan-binding protein (factor VII activating protease) activation in vivo. Biochem Biophys Res Commun 2011; 409:4838
14. Ammollo CT, Semeraro F, Xu J, Esmon NL, Esmon CT:
Extracellular histones increase plasma thrombin generation
by impairing thrombomodulin-dependent protein C activation. J Thromb Haemost 2011; 9:1795803
15. Kardara M, Hatziantoniou S, Sfika A, Vassiliou AG,
Mourelatou E, Muagkou C, Armaganidis A, Roussos C,
Orfanos SE, Kotanidou A, Maniatis NA: Caveolar uptake and
endothelial-protective effects of nanostructured lipid carriers
in acid aspiration murine acute lung injury. Pharm Res 2013;
30:183647
16. Wen Z, Liu Y, Li F, Ren F, Chen D, Li X, Wen T: Circulating
histones exacerbate inflammation in mice with acute liver
failure. J Cell Biochem 2013; 114:238491
17. Zhao D, Ding R, Mao Y, Wang L, Zhang Z, Ma X: Heparin rescues sepsis-associated acute lung injury and lethality through
the suppression of inflammatory responses. Inflammation
2012; 35:182532
18. Young E: The anti-inflammatory effects of heparin and
related compounds. Thromb Res 2008; 122:74352
19. Alcantara FF, Iglehart DJ, Ochs RL: Heparin in plasma samples causes nonspecific binding to histones on Western blots.
J Immunol Methods 1999; 226:118
20. Monestier M, Fasy TM, Losman MJ, Novick KE, Muller S:
Structure and binding properties of monoclonal antibodies
to core histones from autoimmune mice. Mol Immunol 1993;
30:106975
21. Bosmann M, Grailer JJ, Ruemmler R, Russkamp NF, Zetoune
FS, Sarma JV, Standiford TJ, Ward PA: Extracellular histones
are essential effectors of C5aR- and C5L2-mediated tissue
damage and inflammation in acute lung injury. FASEB J 2013;
27:501021
22. Fuchs TA, Kremer Hovinga JA, Schatzberg D, Wagner DD,
Lmmle B: Circulating DNA and myeloperoxidase indicate
disease activity in patients with thrombotic microangiopathies. Blood 2012; 120:115764
23. Papayannopoulos V, Metzler KD, Hakkim A, Zychlinsky A:
Neutrophil elastase and myeloperoxidase regulate the formation of neutrophil extracellular traps. J Cell Biol 2010;
191:67791
24. Jaeschke H, Hasegawa T: Role of neutrophils in acute inflammatory liver injury. Liver Int 2006; 26:9129
25. Xu J, Zhang X, Monestier M, Esmon NL, Esmon CT:
Extracellular histones are mediators of death through TLR2
and TLR4 in mouse fatal liver injury. J Immunol 2011;
187:262631
26. Holdenrieder S, Stieber P: Clinical use of circulating nucleosomes. Crit Rev Clin Lab Sci 2009; 46:124
27. Abrams ST, Zhang N, Manson J, Liu T, Dart C, Baluwa F, Wang
SS, Brohi K, Kipar A, Yu W, Wang G, Toh CH: Circulating
histones are mediators of trauma-associated lung injury. Am
J Respir Crit Care Med 2013; 187:1609
28. Taooka Y, Ohe M, Chen L, Sutani A, Higashi Y, Isobe T:
Increased expression levels of Integrin 91 and CD11b on
circulating neutrophils and elevated serum IL-17A in elderly
aspiration pneumonia. Respiration 2013; 86:36775
29. Cornlio FD, Martins TM, de Andrade LE, de Freitas AC, Lazo
Chica JE, Artrio SC, Faccioli LH, Paula RA: Anti-inflammatory
effects of ellagic acid on acute lung injury induced by acid in
mice. Mediators Inflamm 2013; 2013:164202
30. Davidson BA, Vethanayagam RR, Grimm MJ, Mullan BA,

Raghavendran K, Blackwell TS, Freeman ML, Ayyasamy
V, Singh KK, Sporn MB, Itagaki K, Hauser CJ, Knight PR,
Segal BH: NADPH oxidase and Nrf2 regulate gastric
aspiration-induced inflammation and acute lung injury. J

Immunol 2013; 190:171424
31. Wildhagen KC, Garca de Frutos P, Reutelingsperger CP,

Schrijver R, Arest C, Ortega-Gmez A, Deckers NM, Hemker
HC, Soehnlein O, Nicolaes GA: Nonanticoagulant heparin
prevents histone-mediated cytotoxicity in vitro and improves
survival in sepsis. Blood 2014; 123:1098101
32. Lockey DJ, Coats T, Parr MJ: Aspiration in severe trauma: A
prospective study. Anaesthesia 1999; 54:10978
33. Chen GY, Nuez G: Sterile inflammation: Sensing and reacting to damage. Nat Rev Immunol 2010; 10:82637
34. De Meyer SF, Suidan GL, Fuchs TA, Monestier M, Wagner DD:
Extracellular chromatin is an important mediator of ischemic
stroke in mice. Arterioscler Thromb Vasc Biol 2012; 32:188491
35. Henson PM: Dampening inflammation. Nat Immunol 2005;
6:117981
36. Holdenrieder S, Stieber P, Bodenmller H, Busch M, Von

Pawel J, Schalhorn A, Nagel D, Seidel D: Circulating nucleosomes in serum. Ann N Y Acad Sci 2001; 945:93102
37. Zeerleder S, Zwart B, Wuillemin WA, Aarden LA, Groeneveld
AB, Caliezi C, van Nieuwenhuijze AE, van Mierlo GJ,
Eerenberg AJ, Lmmle B, Hack CE: Elevated nucleosome levels in systemic inflammation and sepsis. Crit Care Med 2003;
31:194751
38. Roth GA, Lubsczyk BA, Pilz J, Faybik P, Hetz H, Krenn CG:
Nucleosome serum levels in acute hepatic failure and MARS
treatment. Transplant Proc 2009; 41:420710
39. Taha MO, Gonalves PF, Vidigal RO, Sacchi GP, Pinheiro

FG, Brando LS, Feguri S, Moraes-Filho AM, Ghadie MM,
Magalhes KG, Simes MJ, Tersariol IL, Monteiro HP,
Oliveira-Jnior IS, Oliveira I, Jr, Jurkiewicz A, Caricati-Neto A:
Protective effects of heparin on hepatic ischemia and reperfusion lesions in rabbits. Transplant Proc 2009; 41:8125
40. Nakamura T, Vollmar B, Winning J, Ueda M, Menger MD,
Schfers HJ: Heparin and the nonanticoagulant N-acetyl
heparin attenuate capillary no-reflow after normothermic
ischemia of the lung. Ann Thorac Surg 2001; 72:11838; discussion 11889
41. Lakshmi TS, Shanmugasundaram N, Shanmuganathan S,
Babu M: Efficacy of desulfated heparin mitigating inflammation in rat burn wound model. J Biomed Mater Res B Appl
Biomater 2011; 97:21523
PAIN MEDICINE
Disruption of Cortical Connectivity during Remifentanil

Administration Is Associated with Cognitive Impairment
but Not with Analgesia
Ahmad Khodayari-Rostamabad, M.Sc., Ph.D., Sren S. Olesen, M.D., Ph.D.,
Carina Graversen, M.Sc., Ph.D., Lasse P. Malver, M.D., Geana P. Kurita, M.Sc.,
Per Sjgren, M.D., Ph.D., D.M.Sc., Lona L. Christrup, M.Sc., Ph.D.,
Asbjrn M. Drewes, M.D., Ph.D., D.M.Sc.
ABSTRACT
Background: The authors investigated the effect of remifentanil administration on resting electroencephalography functional
connectivity and its relationship to cognitive function and analgesia in healthy volunteers.
Methods: Twenty-one healthy male adult subjects were enrolled in this placebo-controlled double-blind cross-over study.
For each subject, 2.5min of multichannel electroencephalography recording, a cognitive test of sustained attention (continuous reaction time), and experimental pain scores to bone-pressure and heat stimuli were collected before and after infusion
of remifentanil or placebo. A coherence matrix was calculated from the electroencephalogram, and three graph-theoretical
measures (characteristic path-length, mean clustering coefficient, and relative small-worldness) were extracted to characterize
the overall cortical network properties.
Results: Compared to placebo, most graph-theoretical measures were significantly altered by remifentanil at the alpha and low
beta range (8 to 18 Hz; all P < 0.001). Taken together, these alterations were characterized by an increase in the characteristic
path-length (alpha 17% and low beta range 24%) and corresponding decrements in mean clustering coefficient (low beta
range 25%) and relative small-worldness (alpha 17% and low beta range 42%). Changes in characteristic path-lengths
after remifentanil infusion were correlated to the continuous reaction time index (r = 0.57; P = 0.009), while no significant
correlations between graph-theoretical measures and experimental pain tests were seen.
Conclusions: Remifentanil disrupts the functional connectivity network properties of the electroencephalogram. The findings
give new insight into how opioids interfere with the normal brain functions and have the potential to be biomarkers for the
sedative effects of opioids in different clinical settings. (Anesthesiology 2015; 122:140-9)
EMIFENTANIL is a potent central acting -receptor

opioid agonist used for balanced anesthesia and analgosedation.1,2 Due to its remarkable titrability it has proven
to be useful for scientific research and provides an ideal drug
to study the mechanism of action underlying the effects of
opioids.
During the last decade, electroencephalography has
been increasingly used to study and monitor central acting drug effects. The advantage of electroencephalography
is that it provides a reliable and direct measure of brain
activity and has relatively low cost and ease of administration compared to other neuroimaging methods. Electroencephalography has previously been used to study
central effects of remifentanil, and clear relationships
between electroencephalographic alterations and remifentanil pharmacokinetics have been demonstrated.36 Most
of these studies, however, focused on the classical power
spectral density of the electroencephalogram. Thus, the
complex information captured by electroencephalography
was reduced to quantitative measures of brain rhythmicity. Consequently, these studies did not assess the brain

During the last decade, electroencephalography has been increasingly used to study and monitor central-acting drug effects due to the reliable and direct measure of brain activity
without the need of an advanced and expensive setup.
A clear association between remifentanil pharmacokinetics
and the power density spectrum of the electroencephalography has previously been demonstrated. However, the effect of
remifentanil on functional network connectivity, as evaluated
by electroencephalography analysis, has not been studied.

Remifentanil altered graph-theoretical measures of the electroencephalography, characterized by an increase in path
length in the alpha and low beta frequency ranges.
Changes in path length were correlated to continuous reaction
time, a measure of sedation. However, a correlation between
electroencephalography measures and pain perception was
not apparent.
Remifentanil alters functional network connectivity in the brain,
and the changes in the electroencephalography have the potential to serve as markers of remifentanil induced sedation
but not analgesia.
This article is featured in This Month in Anesthesiology, page 1A. Corresponding article on page 8.
January 2015
PAIN MEDICINE
networks that can be inferred from multichannel electroencephalography data in terms of functional connectivity.7 To characterize the complex networks in the brain,
the characteristics obtained by functional connectivity
may be subjected to graph-theoretical measures.8,9 Hence,
graph-theory is build on a network that is reconstructed
through functional connectivity analysis and decomposes
the multiple connectivity features extracted from the
multichannel electroencephalogram into a few composite
measures of the overall brain network performance and
functionalityfor further details see figure1. Analysis of
functional cortical connectivity followed by graph-theory
(and other advanced statistical methods) has provided
valuable information on disease mechanisms underlying
neurological and psychiatric disorders including autism,10
schizophrenia,11 bipolar disorder,12 and temporal lobe epilepsy.13 Also, in a recent study based on electroencephalographic alterations and analysis of cortical connectivity,
disruption of frontoparietal communication was shown
to be a possible biomarker of anaesthesia.14
In the current study, we investigated the effect of remifentanil and placebo infusion on graph-theoretical measures
of functional cortical connectivity using multichannel electroencephalography. We hypothesized that in comparison to
placebo; remifentanil infusion in healthy volunteers would
induce significant changes in network properties of functional connectivity in the resting electroencephalogram. The
aims of the study were as follows: (1) validate reproducibility of the graph-theoretical measures of cortical connectivity obtained from resting state electroencephalography, (2)
investigate changes in graph-theoretical measures of cortical
functional connectivity after remifentanil infusion compared
to placebo, and (3) correlate putative changes in graph-theoretical measures to the sedative and analgesic effects of remifentanil as assessed by measurement of continues reaction
time (CRT) and experimental pain stimulation.

Study Subjects
Twenty-four healthy males were invited to participate in
this study, which was approved by the Ethics Committee for the Region of Northern Jutland, Aalborg, Denmark
(N-20110014) and the Danish Health and Medicines
Authority (EudraCT No. 2009013465). All subjects provided written informed consent. The study was conducted
according to the Declaration of Helsinki and the rules of
Good Clinical Practice. The Good Clinical Practice Unit at
Aarhus University Hospital monitored the study, which was a
locally approved extension of the main study registered with

www.clinicaltrials.gov (NCT01375348). Data from the current study were collected as an explorative supplement to gain
insight into mechanisms of action for remifentanil.
Before enrollment, all subjects underwent a clinical
examination to ensure all were in good health with no
history of chronic pain, psychiatric disorders, or drug
abuse. Subjects were asked to fast for 6h before experimental investigations and not use any analgesics for 14
days before study start. Immediately before the test session, participants were required to pass a drug screen and
an alcohol test.
Study Protocol and Medication
The study was an investigator-initiated, double-blind, placebo-controlled, cross-over study of remifentanil in healthy
volunteers. A diagram of the study protocol is illustrated in
figure2. Each session started with a pretreatment recording
of resting state electroencephalography followed by an assessment of cognitive performance by CRT and subjective pain
scores using quantitative sensory testing (QST) for bone and
heat pain. After the pretreatment assessment, the subject was
randomly assigned to either remifentanil or placebo infusions. Steady state was assumed after 25min of infusion and
independent recordings of electroencephalography, CRT,
and QST were obtained again. After a 2-h drug washout
period, a new pretreatment assessment of resting state electroencephalography, CRT, and QST was obtained and the
second infusion was given (either placebo or remifentanil
depending on the initial infusion). The second infusion was
followed by reassessment of electroencephalography, CRT,
and QST as described for the first session.
Blinding of treatment allocation was ensured by two identical 50ml infusions, prepared by a pharmacist with no other
involvement in the study according to a randomization plan
generated at www.randomization.com. The remifentanil
infusion contained 2mg remifentanil (Ultiva; GlaxoSmithKline Pharma A/S, Brndby, Denmark) dissolved in 33.3ml
isotonic saline to a concentration of 60 g/ml, while the placebo infusion contained an equal amount of isotonic saline.
All infusions were administered by an infusion syringe pump
(Codan Green Stream SY-P; Codan Medizinische Gerte
GmbH & Co KG, Lensahn, Germany) ensuring a remifentanil dose of 0.1 g kg1 min1.
Continuous Reaction Time
The CRT is a test of sustained attention and vigilance and
has been well-validated in the Danish population.15 CRT
was tested in terms of the response to auditory stimuli using
Submitted for publication April 5, 2014. Accepted for publication August 29, 2014. From the Mech-Sense, Department of Gastroenterology and Hepatology, Aalborg University Hospital, Aalborg, Denmark (A.K.-R. S.S.O., C.G., L.P.M., A.M.D.); The Multidisciplinary Pain Centre,
Rigshospitalet, University of Copenhagen, Copenhagen, Denmark (G.P.K.); Section of Palliative Medicine, Department of Oncology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark (G.P.K., P.S.); Department of Drug Design and Pharmacology, Faculty
of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark (L.L.C.); Department of Electrical and Computer Engineering, McMaster University, Hamilton, Ontario, Canada (A.K.-R.); and Center for Sensory-Motor Interactions (SMI), Department of Health
Science and Technology, Aalborg University, Aalborg, Denmark (A.M.D.).
Anesthesiology 2015; 122:140-9 141 Khodayari-Rostamabad et al.
Remifentanil Alters Brain Connectivity Networks
Fig. 1. Functional brain connectivity networks assessed in terms of graph-theoretical analysis. The multichannel electroencephalography recordings (A) are transformed into coherence (COH) measures (B). The symmetrical coherence matrix to the left
contains all pair-wise couplings, which is represented on a circle to the right. A coherence value of 0 represents no coherence,
while a value of 1 represents maximum coherence. The coherence values are mapped into appropriate weights of functional distance between electrodes (C), with the constraint that no coherence represents maximum distance, while maximum coherence
is associated with zero distance. The coherence values and corresponding weights of functional distance are then subjected to
graph-theoretical analysis (D). In this study, three graph measures were used for analysis. The mean clustering coefficient (i) is
a measure of the level of clustered connectivity around electroencephalography electrodes. In the graph to the left, c is strongly
connected to electrodes b and d, which are also connected. This forms a functional triangle, while no triangles are present in the
figure to the right. Hence, in the graph to the left, groups of neurons forms a functional unit and have higher functional connectivity. The characteristic path-length (ii) is a measure of functional distance between any two electroencephalogram electrode pairs.
In the graph to the left, the distance from b to e is longer than in the graph to the right, and hence information is transferred more
rapidly in the latter graph due to the shorter path-length. Taken together, quantification of small-world networks (iii) provides a
score of the balance between segregation and integration of the network. In the graph to the left, the network is perfectly ordered
in contrary to the graph to the right, which is completely random. The small-network graph in the middle has a low path-length
and high-clustering coefficient, which is known to be associated with normal brain function.
a response unit and a push-button system controlled by commercially available software (EKHO; Bitmatic, Viby J, Denmark). One hundred auditory beep signals (500 Hz, 90 dB)
were delivered through headphones to the subject at random
intervals of 2 to 6 s. The subject was instructed to press a button as soon as he heard a beep signal. The reaction time, that
is, the time from emission of the sound signal to activation
of the button was measured in milliseconds. The CRT index
PAIN MEDICINE
Fig. 2. Outline of the study protocol. On the same day, each

subject received both remifentanil (REM) and placebo (PLA)
infusions in random order with a 2-hr washout period in between. For each infusion, pre- as well as post-treatment data
were collected. Data collection included resting state electroencephalography (EEG), continuous reaction time (CRT) as
well as subjective pain scores by quantitative sensory testing
(QST) for experimental bone and heat pain.
calculation was based on the fractiles of the distribution of

the measurements. The 10, 50 (median), and 90% fractiles
were determined and the reaction time index was calculated
as: 50 percentile/(90 percentile to 10 percentile).16
Quantitative Sensory Testing
The QST measures were recorded using a standard numerical rating scale where: 0 = no pain, 5 = moderate pain,
and 10 = worst imaginable pain.17 Bone pain was assessed
by pressure stimulation on the dominant leg 15cm below
the patella using a handheld algometer (Type 2; Somedic
production AB, Hrby, Sweden) with a probe diameter
of 2mm. The pressure increase rate was 225 mmHg/sec
adjusted to a probe diameter of 11mm2 according to recommendations by Andresen et al.18 Subjects were instructed
to press a button when they experienced moderate pain
(numerical rating scale = 5), which terminated the stimulation. Heat pain was assessed by thermal stimulation (TSAII,
NeuroSensory analyzer; Medoc, Ramat Yishai, Israel) on the
right volar forearm midway between the wrist and cubital
fossa. The stimulation intensity was increased by 1C/s from
a baseline temperature of 32C to a maximum temperature
of 52C. Subjects were instructed to press a button when
they experienced moderate pain (numerical rating scale =
5), which caused the stimulator immediately to return to
the baseline temperature of 32C. Three successive stimulations were performed, and the average temperature to inflict
moderate pain was used for the analysis of analgesic effect
to heat pain.
Electroencephalography Recordings and Data
Preprocessing
Electroencephalography data was collected by 62 electrodes
mounted according to the extended 1020 system, using
a Neuroscan SynAmp2 system (Neuroscan 4.3.1, El Paso,
TX) and a standard electroencephalography cap (QuickCap, International, Neuroscan). The data were recorded
using a reference electrode located between Cz and FCz,
a ground electrode in the occipital area of the scalp and a
sampling frequency of 1,000 Hz. Electrode impedance was
kept below 5 k. During recordings, data were bandpass
filtered (0.5 to 200 Hz) using the Neuroscan software. Subjects were instructed to rest with open eyes and gaze fixed
on a remote point. Although the open eyes design may
introduce varying visual inputs, this design was chosen to
avoid sleep patterns in the electroencephalogram after remifentanil infusion due to its sedative effect. Furthermore, all
subjects were positioned likewise to minimize variation in
visual input. For each subject, a total of 2.5min of resting
state electroencephalography data were recorded for each of
the four conditions (i.e., before and after remifentanil and
placebo infusions).
The preprocessing procedure was as follows. In the first
phase, the 50 Hz power supply noise was suppressed by
a notch filter (501 Hz), then a bandpass filter (1 to 70
Hz) was applied to attenuate high-frequency artifacts that
may otherwise interfere with the visual inspection process
in later steps. To validate data quality and to remove excessive muscle and movement artifacts, an electroencephalography expert visually inspected the linked ear rereferenced
electroencephalography data. Furthermore, electrodes with
low signal-to-noise ratio were interpolated by the immediate neighboring electrodes. All these steps were performed
in Neuroscan EDIT 4.3 software (Neuroscan). Then, using
Matlab R2011b software (Mathworks Inc., Natick, MA) the
following analyses were performed: First, data were bandpass filtered (0.7 to 37 Hz). Second, rejection of artifact
intervals with high amplitude was done (i.e., amplitudes 5.5
times the corresponding SD) in either the linked ear or the
Cz referenced version of the data. Third, since the connectivity analysis depends on the electroencephalography reference
system used, the electroencephalography data was rereferenced into the average reference to make results comparable
to previous literature.10,19 Finally, to make a fair comparison
between recordings, all recordings were reduced to equal
length (initial 88 s of deartifacted data) before performing
connectivity analysis.
Electroencephalogram Functional Connectivity
The magnitude squared coherence (COH) between all electrode pairs was extracted from the multichannel electroencephalogram using the Welchs modified periodogram
averaging method with nonoverlapping windows. The window length was 2 s, resulting in 0.5 Hz frequency resolution.
To partly reduce low-frequency eye-movement and eye-blink
artifacts, we used the 2 to 30 Hz range for data analysis. The
electroencephalogram coherence values were calculated separately for the following frequency bands: delta (2 to 3.5 Hz),
theta (4 to 7.5 Hz), alpha (8 to 12 Hz), beta1 (12.5 to 18
Hz), and beta2 (18.5 to 30 Hz).
Graph-theoretical Analysis
Graph-theoretical analysis was applied to functional connectivity data to characterize the cortical network as illustrated in figure1. In general, graph-theoretical analysis
is used to identify overall network properties and can be
Table 1. Comparison of Continuous Reaction Time and Experimental Heat and Bone Pain between Remifentanil and Placebo
Infusion
Remifentanil
Pretreatment
Continuous reaction time index
Heat pain (C)
Bone pressure pain (mmHg/cm2)
2.40.64
45.52.4
40,6307,943
Placebo
Posttreatment
1.80 .79
47.93
57,12014,050
Pretreatment
2.40.53
45.72.9
42,9107,880
Posttreatment
2.30.56
46.22.2
40,99810,330
Remifentanil vs.
Placebo
F = 14.1; P < 0.001*
F = 23.1; P < 0.001*
F = 39.8; P < 0.001*
* Significant result after adjusting for multiple comparisons.
constructed in several ways.9 First, the input to the graph

should reflect the connection strength (functional distance) between electrodes and should therefore be considered carefully. As coherence measures are characterized by
values on a continuous scale from 0 to 1 (0 = no coherence
and 1 = maximum coherence), but with no information
with respect to direction of information flux, a weighted
undirected network design was applied.8 Second, the output measures should be selected appropriately to reflect
relevant characteristics of the network. In this study, we
extracted the characteristic path-length, mean clustering
coefficient, and relative small-worldness to quantify complimentary properties of the network, and taken together
they can be used to characterize the overall network
performance.8,20
The characteristic path-length is a measure of the average
functional distance between electrode pairs. It is a measure
of how rapidly information from different specialized brain
regions can be combined and thus provides a measure of
functional cerebral integration. The mean clustering coefficient is the average level of clustered connectivity around
electrodes. Hence, a high clustering coefficient indicates
the presence of local groups of neurons making specialized
functional units (e.g., sensory or motor areas). In contrast to
the average path-length, the clustering coefficient thus provides a measure of functional cortical segregation. Networks
that are both integrated and segregated are commonly called
small-world networks, which reflect the balance of local segregation and global integration.
In this study, electroencephalogram electrodes were
used in the sensor space as nodes and functional connectivity measures as edges, and the graph analysis was performed using the brain connectivity toolbox.20 The mean
clustering coefficient was computed based on the definition in Onnela et al.,21 while relative small-worldness was
computed as mean clustering coefficient/characteristic
path-length, which is a simplified version of the definition in Humphries and Gurney (2008).22 When calculating the characteristic path-length, the coherence values
were mapped into appropriate measures of functional distance between electrodes by applying the inverse function
f(COH) = log(COH). By this mapping, a coherence value
of zero (no connection) was associated with infinite distance, while a coherence value of 1 (maximum connection)
was associated with zero distance.
This was an exploratory study and therefore no formal power
calculation was performed. The primary aim of our study
was to investigate changes in graph-theoretical measures of
cortical functional connectivity after remifentanil infusion.
Previous studies using graph-theoretical measures on resting
state electroencephalography after administration of analgesics and sedatives have typically enrolled between 10 and 30
subjects.14 Consequently, we considered a sample size of 20
subjects to provide sufficient statistical power. In order to
allow for possible dropouts 24 subjects were enrolled.
Descriptive statistics are reported as mean SD unless
otherwise indicated. Analyses of CRT, QST, and electroencephalogram coherence and graph-theoretical measures
followed a two-step procedure: First, pretreatment conditions were compared using paired t tests. Second, differences
between groups (i.e., remifentanil vs. placebo) were compared
using analysis of covariance with the pretreatment value of
the analyzed parameter as a co-variable. Significant changes
in graph-theoretical measures were correlated to changes
in CRT and QST using Pearson correlation coefficient. A
P value of less than 0.05 was considered as an indication
of statistical significance. In case of multiple comparisons,
the Bonferroni correction was used. The software package
STATA version 11.2 (StataCorp LP, College Station, TX)
was used for the statistical analysis.
Results
Twenty-four healthy Caucasian male volunteers were
screened for inclusion in the study. Two subjects had a positive urine drug screen and were excluded, while another
subject was excluded due to a history of psychiatric disorder. Hence, 21 subjects were enrolled in the study. The average age was 23.52.1 yr and the average body weight was
78.710.4kg.
CRT and QST
The CRT scores and QST parameters are reported in table1.
All assessments were comparable between the two pretreatment conditions (all P 0.2). After placebo infusion, no
significant changes were seen in CRT scores (P = 0.18) or
QST parameters (P = 0.07 for heat pain, and P = 0.15 for
bone pressure pain). In contrast, all CRT and QST parameters were changed significantly by remifentanil infusion (all
P < 0.001).
PAIN MEDICINE
seen (P < 0.001) as well as a decrease in the relative smallworldness (P < 0.001). Furthermore, a decrease in the mean
clustering coefficient was evident (P = 0.02), although this
was not significant after adjustment for multiple comparisons (table 2). Taken together, the alterations were characteristic for disruption of cortical networks during remifentanil
treatment.
Fig. 3. Grand mean electroencephalogram coherence across

all 62 electrodes for the four different conditions (i.e., the two
pretreatment conditions, postplacebo and postremifentanil
infusions). Data are provided for each frequency band. Error
bars indicate SD. The mean coherence were different at alpha
(8 to 12 Hz; P = 0.007) and beta1 (12.5 to 18 Hz; P < 0.001)
bands when comparing pretreatment-adjusted remifentanil
and placebo effects.
Electroencephalogram Functional Connectivity

The average coherence across all pairs of electrodes for the
four different conditions (i.e., the pre- and post-treatments
to remifentanil and placebo) is illustrated in figure3. Mean
coherences at pretreatment conditions were comparable at
all frequency bands (all P > 0.1). After remifentanil infusion,
mean coherence was reduced in the alpha (P = 0.007) and
beta1 bands (P < 0.001), while no differences were seen after
placebo infusion (all P > 0.1).
The grand mean functional connectivity measures from
all subjects for the four conditions are illustrated for the
beta1 band in figure4. Based on visual inspection, the overall functional connectivity was decreased after remifentanil
infusion, with less long-range connections and less synchronized cortical activity primarily in the frontal and parietal
brain areas. In contrast, no major differences were seen after
placebo infusion.
Reproducibility of Pretreatment Graph-theoretical
Measurements
Pretreatment graph measures are provided in table2. All
measures were comparable between pretreatment conditions
for all frequency bands (all P 0.09).
Remifentanil versus Placebo Effects on Graph-theoretical
Measurements
When comparing the changes in graph-theoretical measurements of cortical network properties between remifentanil
and placebo conditions in the beta1 band, a significant
increase in characteristic path-length was seen, whereas the
mean clustering coefficient and relative small-worldness
decreased (all P < 0.001; table 2). For the alpha band a significant increase in the characteristic path-length was also
Correlation Analysis
Putative correlations between the CRT index scores, QST
pain parameters (bone and heat pain scores), and graph-theoretical measures influenced by remifentanil infusion (i.e.,
electroencephalogram frequency range alpha and beta1)
were explored. A negative correlation between the characteristic path-length and CRT index was seen in the alpha band
(r = 0.57; P = 0.009; fig.5). No other significant correlations between the graph-theoretical measures and the CRT
or QST parameters were evident after adjustment for multiple comparisons (all P 0.04).
Discussion
The current study investigated how remifentanil and placebo affects graph-theoretical measures obtained from
functional connectivity network measures in resting-state
electroencephalography data from healthy volunteers. After
remifentanil infusion a reduction in overall efficiency of
cortical networks was seen in the alpha and beta1 frequency
ranges (8 to 18 Hz) of the electroencephalogram, whereas
no significant changes were seen after placebo treatment.
Hence, remifentanil disrupted the complex cortical network subserving normal brain function. These alterations
were associated with loss of stability of sustained attention,
while no associations were found regarding the analgesic
effect.
Remifentanil Effects on Cortical Network Properties
To our knowledge, the effect of remifentanil analgesia (i.e.,
subanesthetic concentrations) on electroencephalogram
graph-theoretical measures has not been investigated previously. We found an increase in the characteristic path-length
of the cortical networks with corresponding decrements in
mean clustering coefficient and relative small-worldness. In
experimental studies, neuronal networks optimized for complexity show small-world characteristics, and graph measures
of these networks are comparable to real cortical networks.9
The retrieved values do not directly reflect any specific
underlying neurobiological processes, but taken together,
the findings translate to a reduction in overall efficiency of
the networks between different neuronal centers, and an
alteration of the balance between the integration and segregation of the networks.23 This is in contrast to most previous
studies where small-worldness and scale-free properties were
maintained during general anesthesia.2428 However, previous studies have not used opioids for sedation, but typically
used propofol or other anesthetics with an entirely different
Fig. 4. Comparison of grand mean cortical networks at beta1 band among recordings pre- and post-treatment to remifentanil
and placebo. The four circles illustrate and compare connectivity with coherence values greater than 0.25. The locations of the
62 electroencephalogram electrodes used in the study are shown at the top.
receptor profile. Furthermore, most previous studies looking

at graph-theoretical measures (including small-worldness)
were based on functional resonance imaging. Hence, they
did not measure the direct electrical activity in the brain,
but rather indirect measures of activity based on changes
in blood flow. Also, many anesthetics have effects on the
cardiovascular system, which confounds the use of imaging methods.29 For these reasons the findings in the current
paper cannot be directly compared to previous studies and
more research is needed to study the unique effects of opioid
induced sedation on cortical connectivity.
While multichannel electroencephalography and graphtheoretical analysis has not previously been used after
remifentanil administration, simpler methods to address
cerebral connectivity during remifentanil anesthesia based
on pair-wise connectivity or associations have been reported.
Hayashi et al. 6 investigated resting state electroencephalography before and after induction of anesthesia using a
combination of sevoflurane and remifentanil. They found

that with the induction of anesthesia, frontal alpha activity
measured by bicoherence was increased while occipital alpha
activity was decreased. Decreased alpha coherence reflecting
less synchronized cortical activity in frontal electrodes was
also seen in our study and although the methods are different, they stress the importance of frontal changes during
remifentanil administration.
The changes in functional connectivity after remifentanil
were confined to the alpha and beta1 frequency ranges (8
to 18 Hz). This is in line with a previous study, where Kortelainen et al. 5 studied the effect of adding remifentanil to
propofol anaesthesia. A decrease in beta frequency (>14 Hz)
activity was seen in both light and deep anesthesia, while
increased alpha band (7 to 14 Hz) activity was seen during
deep anesthesia only. Although we did not investigate deep
anesthesia in the current study, the importance of alpha and
beta1 bands modifications after remifentanil infusion was
PAIN MEDICINE
Table 2. Comparison of Functional Connectivity Measures between Remifentanil and Placebo Infusion
Remifentanil
Frequency
Band
Delta (23.5 Hz)
Theta (47.5 Hz)
Alpha (812 Hz)
Beta1 (12.518 Hz)
Beta2 (18.530 Hz)
Placebo
Network
Measure
Pretreatment
Posttreatment
Pretreatment
Posttreatment
Remifentanil vs. Placebo
L
C
S
L
C
S
L
C
S
L
C
S
L
C
S
1.160.26
0.190.09
0.220.30
1.130.17
0.190.04
0.180.08
1.020.16
0.220.04
0.220.08
1.410.26
0.160.04
0.120.06
1.680.27
0.130.03
0.080.04
1.050.34
0.260.12
0.330.30
1.170.20
0.200.05
0.180.08
1.190.22
0.200.04
0.180.06
1.750.26
0.120.03
0.070.03
1.840.25
0.120.04
0.070.04
1.210.28
0.180.04
0.160.07
1.160.20
0.170.02
0.160.04
1.080.19
0.210.04
0.200.07
1.440.24
0.150.03
0.110.04
1.660.23
0.130.02
0.080.03
1.150.31
0.200.07
0.210.17
1.140.19
0.180.03
0.170.05
1.040.16
0.210.04
0.220.07
1.400.20
0.150.02
0.110.03
1.660.19
0.130.02
0.080.03
F = 0.48; P = 0.49
F = 2.46; P = 0.12
F = 1.67; P = 0.20
F = 3.25; P = 0.08
F = 0.43; P = 0.51
F < 0.01; P = 0.98
F = 28.6; P < 0.001*
F = 5.76; P = 0.02
F = 17.9; P < 0.001*
F = 46.1; P < 0.001*
F = 42.8; P < 0.001*
F = 39.9; P < 0.001*
F= 7.83; P = 0.008
F = 2.03; P = 0.16
F = 1.77; P = 0.19
* Significant result after adjustment for multiple comparisons.

C = mean clustering coefficient; L = characteristic path-length; S = relative small-worldness.
Fig. 5. Correlation between the characteristic path-length and

continuous reaction time (CRT) index for the alpha band after
remifentanil infusion.
supported by this study. Finally, in a recent study Lee et al.24

used network analysis to explore the effect of propofol. They
found disruption of hub structure in the same frequency
bands as in our study. Hubs facilitate functional integration
of the network and in their study they are interpreted to
reflect the coordination of information flow in the brain.30
Although propofol is an anesthetic without major analgesic
properties, the findings support those in the current study.
From a functional perspective, the retrieved changes in
connectivity graph measures were associated with impaired
cognitive function as indicated by the correlation to changes
in CRT index. CRT is primarily a simple neuropsychological test of vigilance and sustained attention.31 Other studies
involving mentally demanding tests of sustained attention
and vigilance have demonstrated that functional network
organization can change over relatively short time scales.32
Thus, increased path-length and an asymmetrical pattern
of connectivity (right > left) in frontoparietal regions
were associated with decreased attention.31 In particular,
our study showed that the increase in characteristic pathlength was associated with changes in CRT. This further
validates the findings since the characteristic path-length is
a measure of functional cortical integration and cross-talk
between distinct brain areas. Along this line, abnormal cortical connectivity seems to be a common trait for various
neurological and psychiatric disorders including Alzheimers
disease, schizophrenia, and autism.9 These are characterized
by varying degree of cognitive dysfunction related to abnormal integration of large-scale network (i.e., disconnection
syndromes).33
In contrast, no association to analgesia was seen. We have
recently found changes in brain networks between insular
and cingulate sources during morphine treatment in a study
where pain specific evoked brain potentials to phasic stimuli
were used.34 However, the current study explored the general
properties of the brain in contrast to studies on evoked brain
potentials. To our knowledge, there are no graph-theoretical
studies on brain networks in patients with chronic pain, but
our findings are supported by a previous study where withdrawal of opioids increased functional connectivity in the
alpha and beta bands.35 Finally, it is well known that remifentanil may cause hyperalgesia rather than analgesia in many
subjects.36 Although speculative the balance between analgesia
and hyperalgesia (together with sedative effects on pain assessment) may therefore confound the results. In the experimental
settings, phasic pain stimuli are also less sensitive to opioid
treatment.37 Hence, it could be recommended in future studies to study the effects of remifentanil on long-lasting pain
stimuli such as ischemia or other types of tonic pain.
Methodological Considerations
Remifentanil is a fast acting opioid with a half-life of less
than 20min.1 Consequently a 2-hr washout period (equal
to more than six half-lives) was considered sufficient time

ensuring the brain to return to baseline condition before
the second infusion was initiated. To validate the stability
of the functional connectivity graph-theoretical measures of
the electroencephalogram, the test re-test reproducibility of
the two pretreatment recordings was evaluated. All measures
were reproducible and thus confirm the stability of cortical connectivity during pretreatment conditions. Along this
line, no change in functional connectivity and the associated graph-theoretical measures were observed after placebo
infusion (analysis not shown), which further strengthen the
validity of the changes seen in functional connectivity after
remifentanil.
Some methodological limitations are present in this
study. In order to control for confounding factors, enrolled
subjects comprised a very homogenous group (i.e., opioid
naive healthy males). Therefore, an extension of this study
is needed to assess the effects of remifentanil on functional
cortical connectivity in women and in the clinical conditions
such as in chronic pain patients taking opioids on a regular
basis. Additionally, it should be noted that we have used the
sensor space for functional connectivity analysis rather than
applying source localization first to use neuroanatomical reference points. This design was chosen to enable comparison
to previous studies on remifentanil also reporting findings
obtained from surface electrodes. Finally, it should be noted
that electroencephalography and electromyography have an
overlap in some frequency bands. To minimize the effect of
the electromyography component, we applied both manual
and automatic deartifacting to the data, and reduced the
analyzed frequency range to 2 to 30 Hz.
Conclusions
Remifentanil administration is associated with significant
changes in functional cortical connectivity including a
reduction in overall efficiency of the cortical network. These
modifications seem to disrupt the complex cortical network
subserving normal brain function and are associated with
cognitive performance but not with analgesia.
Acknowledgments
This study was supported by The Danish Cancer Society
(Krftens Bekmpelse), Copenhagen, Denmark, and The
Danish Agency for Science, Technology, and Innovation
(Det Strategiske Forskningsrd), Copenhagen, Denmark
(grant no. 10-092786).
Competing Interests
Correspondence
Address correspondence to Dr. Drewes: Mech-Sense, Department of Gastroenterology and Hepatology, Aalborg University Hospital, Mlleparkvej 4, 9000 Aalborg, Denmark.
amd@rn.dk. Information on purchasing reprints may be
found at www.anesthesiology.org or on the masthead page
at the beginning of this issue. Anesthesiologys articles are

made freely accessible to all readers, for personal use only,
6 months from the cover date of the issue.
References
1. Brkle H, Dunbar S, Van Aken H: Remifentanil: A novel,
short-acting, mu-opioid. Anesth Analg 1996; 83:64651
2. Servin FS, Billard V: Remifentanil and other opioids. Handb
Exp Pharmacol 2008; 182:283311
3. Egan TD, Minto CF, Hermann DJ, Barr J, Muir KT, Shafer SL:
Remifentanil versus alfentanil: Comparative pharmacokinetics and pharmacodynamics in healthy adult male volunteers.
4. Noh GJ, Kim KM, Jeong YB, Jeong SW, Yoon HS, Jeong
SM, Kang SH, Linares O, Kern SE: Electroencephalographic
approximate entropy changes in healthy volunteers during
remifentanil infusion. Anesthesiology 2006; 104:92132
5. Kortelainen J, Koskinen M, Mustola S, Seppnen T: Effects
of remifentanil on the spectrum and quantitative parameters of electroencephalogram in propofol anesthesia.
6. Hayashi K, Mukai N, Sawa T: Simultaneous bicoherence
analysis of occipital and frontal electroencephalograms in
awake and anesthetized subjects. Clin Neurophysiol 2014;
125:194201
7. Sleigh JW: The study of consciousness comes of age.
8. Bullmore E, Sporns O: Complex brain networks: Graph theoretical analysis of structural and functional systems. Nat Rev
Neurosci 2009; 10:18698
9. Stam CJ, Reijneveld JC: Graph theoretical analysis of complex
networks in the brain. Nonlinear Biomed Phys 2007; 1:3
10. Peters JM, Taquet M, Vega C, Jeste SS, Fernndez IS, Tan J,
Nelson CA III, Sahin M, Warfield SK: Brain functional networks in syndromic and non-syndromic autism: A graph
theoretical study of EEG connectivity. BMC Med 2013; 11:54
11. Rubinov M, Knock SA, Stam CJ, Micheloyannis S, Harris AW,
Williams LM, Breakspear M: Small-world properties of nonlinear brain activity in schizophrenia. Hum Brain Mapp 2009;
30:40316
12. Kim DJ, Bolbecker AR, Howell J, Rass O, Sporns O, Hetrick
WP, Breier A, ODonnell BF: Disturbed resting state EEG synchronization in bipolar disorder: A graph-theoretic analysis.
Neuroimage Clin 2013; 2:41423
13. Antony AR, Alexopoulos AV, Gonzlez-Martnez JA, Mosher JC,
Jehi L, Burgess RC, So NK, Galn RF: Functional connectivity
estimated from intracranial EEG predicts surgical outcome in
intractable temporal lobe epilepsy. PLoS One 2013; 8:e77916
14. Lee U, Ku S, Noh G, Baek S, Choi B, Mashour GA: Disruption
of frontal-parietal communication by ketamine, propofol,
and sevoflurane. Anesthesiology 2013; 118:126475
15. Elsass P: Continuous reaction times in cerebral dysfunction.
Acta Neurol Scand 1986; 73:22546
16. Lauridsen MM, Thiele M, Kimer N, Vilstrup H: The continuous reaction times method for diagnosing, grading, and monitoring minimal/covert hepatic encephalopathy. Metab Brain
Dis 2013; 28:2314
17. Breivik H, Borchgrevink PC, Allen SM, Rosseland LA,
Romundstad L, Hals EK, Kvarstein G, Stubhaug A: Assessment
of pain. Br J Anaesth 2008; 101:1724
18. Andresen T, Pfeiffer-Jensen M, Brock C, Drewes AM, ArendtNielsen L: A human experimental bone pain model. Basic
Clin Pharmacol Toxicol 2013; 112:11623
19. Quraan MA, McCormick C, Cohn M, Valiante TA, McAndrews
MP: Altered resting state brain dynamics in temporal lobe epilepsy can be observed in spectral power, functional connectivity and graph theory metrics. PLoS One 2013; 8:e68609
PAIN MEDICINE

52:105969
21. Onnela JP, Saramki J, Kertsz J, Kaski K: Intensity and
coherence of motifs in weighted complex networks. Phys
Rev E Stat Nonlin Soft Matter Phys 2005; 71(6 Pt 2):065103
quantitative method for determining canonical network
equivalence. PLoS One 2008; 3:e0002051
networks maintain a scale-free organization across consciousness, anesthesia, and recovery: evidence for adaptive
reconfiguration. Anesthesiology 2010; 113:108191
27. Liu X, Ward BD, Binder JR, Li SJ, Hudetz AG: Scale-free functional connectivity of the brain is maintained in anesthetized
healthy participants but not in patients with unresponsive
wakefulness syndrome. PLoS One 2014; 9:e92182
28. Schrter MS, Spoormaker VI, Schorer A, Wohlschlger A,
Czisch M, Kochs EF, Zimmer C, Hemmer B, Schneider G,
Jordan D, Ilg R: Spatiotemporal reconfiguration of large-scale
brain functional networks during propofol-induced loss of
consciousness. J Neurosci 2012; 32:1283240
29. Shushruth S: Exploring the neural basis of consciousness

through anesthesia. J Neurosci 2013; 33:17578
30. Sporns O: Network attributes for segregation and integration
in the human brain. Curr Opin Neurobiol 2013; 23:16271
31. Sun Y, Lim J, Kwok K, Bezerianos A: Functional cortical connectivity analysis of mental fatigue unmasks hemispheric
asymmetry and changes in small-world networks. Brain
Cogn 2014; 85:22030
32. Piantoni G, Cheung BL, Van Veen BD, Romeijn N, Riedner
BA, Tononi G, Van Der Werf YD, Van Someren EJ: Disrupted
directed connectivity along the cingulate cortex determines
vigilance after sleep deprivation. Neuroimage 2013; 79:21322
33. Catani M, Ffytche DH: The rises and falls of disconnection
syndromes. Brain 2005; 128(Pt 10):222439
34. Lelic D, Olesen AE, Gregersen H, Dahan A, Kolesnikov
Y, Drewes AM: Morphine modifies the cingulate-operculum network underlying painful rectal evoked potentials.
Neuropharmacology 2014; 77:4227
35. Fingelkurts AA, Fingelkurts AA, Kivisaari R, Autti T, Borisov
S, Puuskari V, Jokela O, Khknen S: Opioid withdrawal
results in an increased local and remote functional connectivity at EEG alpha and beta frequency bands. Neurosci Res
2007; 58:409
36. Sivak EL, Davis PJ: Review of the efficacy and safety of remifentanil for the prevention and treatment of pain during and
after procedures and surgery. Local Reg Anesth 2010; 3:3543
37. Olesen AE, Andresen T, Staahl C, Drewes AM: Human experimental pain models for assessing the therapeutic efficacy of
analgesic drugs. Pharmacol Rev 2012; 64:72279
Immediate Rescue Designs in Pediatric Analgesic Trials

A Systematic Review and Meta-analysis
Joe Kossowsky, Ph.D., Carolina Donado, M.D., Charles B. Berde, M.D., Ph.D.
ABSTRACT
Background: Designing analgesic clinical trials in pediatrics requires a balance between scientific, ethical, and practical concerns. A previous consensus group recommended immediate rescue designs using opioid sparing as a surrogate measure of
analgesic efficacy. The authors summarize the performance of rescue analgesic designs in pediatric trials of four commonly
used classes of analgesics: opioids, nonsteroidal antiinflammatory drugs, acetaminophen, and local anesthetics.
Methods: MEDLINE, Embase, CINAHL, The Cochrane Library, and Web of science were searched in April 2013. The 85
studies selected were randomized or controlled clinical trials using immediate rescue paradigms in postoperative pain settings.
A random-effects meta-analysis was used to synthesize predefined outcomes using Hedges g. Difference between the means of
the treatment arms were also expressed as a percentage of the corresponding value in the placebo group (placebo-treatment/
placebo). Distributions of pain scores in study and control groups and relationships between opioid sparing and pain scores
were examined.
Results: For each of the four study drug classes, significant opioid sparing was demonstrated in a majority of studies by one
or more of the following endpoints: (1) total dose (milligram per kilogram per hour), (2) percentage of children requiring
rescue medication, and (3) time to first rescue medication (minutes). Pain scores averaged 2.4/10 in study groups, 3.4/10 in
control groups.
Conclusions: Opioid sparing is a feasible pragmatic endpoint for pediatric pain analgesic trials. This review serves to guide
future research in pediatric analgesia trials, which could test whether some specific design features may improve assay sensitivity while minimizing the risk of unrelieved pain. (Anesthesiology 2015; 122:150-71)
LINICAL trials to evaluate efficacy or effectiveness

of analgesics pose ethical and scientific challenges
for all ages, but especially for children. For sound scientific reasons, the standard approach to adult acute pain trials involves enrollment of patients with moderate to severe
pain, randomization between active drug and placebo, and
comparison of pain scores over time between active and placebo subjects as the primary measure of analgesic efficacy
(fig.1A).1 The ethical basis of this approach rests on adults
making informed decisions to bear the risk of assignment
to a placebo group and potentially to experience continued
pain during the study period. Comparative effectiveness trials involving no placebo group are relevant for guiding clinical decision-making, but they pose statistical problems for
establishing drug efficacy.2
For children, as vulnerable subjects, there is greater ethical concern about a significant risk of unrelieved pain in
clinical trials, particularly where there are existing effective
treatments. For ethical and practical reasons, pediatric analgesic trials have proceeded slowly. The U.S. Congress, the
Food and Drug Administration, and the European Union

Traditional approaches to pediatric analgesic trials may leave
patients with insufficiently treated pain
Whether rescue treatments are generally effective in pediatric
analgesia trials remains unclear

The investigators performed a meta-analysis of pediatric trials
with four classes of analgesics, using rescue/opioid sparing
designs
Average pain scores were low and similar in control and experimental analgesic groups, confirming the ethical basis of
opioid-sparing rescue designs
Opioid-sparing designs also showed good assay sensitivity
have generated incentives for pharmaceutical companies to

test analgesics in children, and some funding mechanisms
via the Food and Drug Administration and the National
Institutes of Health have supported a small number of investigator-initiated pediatric trials for off-patent analgesics that
lack commercial incentives.3 Despite these initiatives, enrollment rates in pediatric analgesic trials have been very low.4
anesthesiology.org). The first two authors contributed equally to this article.
Submitted for publication April 16, 2014. Accepted for publication August 14, 2014. From the Department of Anesthesiology, Perioperative and Pain Medicine, Boston Childrens Hospital, Harvard Medical School, Boston, Massachusetts ( J.K., C.D., C.B.B.); and Department of
Clinical Psychology and Psychotherapy, University of Basel, Basel, Switzerland ( J.K.).
January 2015
Pain Medicine
hypotheses were (1) immediate opioid rescue designs provide

reasonable assay sensitivity as pragmatic surrogate measures of
analgesia efficacy, (2) surrogate efficacy effect sizes vary with
drug class and study design, choice of opioid-sparing endpoints, method of rescue analgesic administration, and type
of surgery, and (3) subjects randomized to either study drug
or control/placebo groups with these designs had acceptably
low mean pain scores (additional technical aspects and details
relevant to researchers in the field can be found in Supplemental Digital Content 1, http://links.lww.com/ALN/B91).

The search strategy is described in detail in Section A1 of
Supplemental Digital Content 1, http://links.lww.com/
ALN/B91.
Fig. 1. (A) Typical time course of pain scores for a doubleblind, parallel-group, placebo-controlled, active comparator
analgesic trial. Note that in general, requirement for rescue
analgesia results in termination of pain scoring for that subject. (B) Idealized time course of hourly rescue dosing of a
short-acting opioid. (C) Idealized time course of the pain.
Note that, depending on the dosing schedule for rescue analgesics, in some trials of this design, pain scores remain lower
in the active group than in the placebo group. VAS = visual
analog scale. Reproduced, with permission, from Berde et al.
Pediatrics 2012; 129:35464.1
In November 2010, an expert consensus group was convened by the Anesthetic, Analgesic and Addiction Drugs
section of the Food and Drug Administration to address
pediatric analgesic trial design issues. Recommendations
from this group were published in Pediatrics in 2012.1 One
prominent recommendation was to regard rescue-analgesic
sparing as a pragmatic surrogate primary endpoint in pediatric analgesic trials. In order to maintain the scientific advantages of the traditional adult design, subjects could still be
blinded and randomized between study drug and placebo.
By providing immediate access to incremental rescue analgesia, especially via patient-controlled analgesia (PCA) or
nurse-controlled analgesia (NCA) pumps, we expected that
these designs would reduce the odds that subjects, especially
those randomized to placebo, would experience unrelieved
severe pain (fig. 1, B and C). This recommendation was
based on committee members opinions and informal review
of a small number of successful trials, but not on any systematic review or quantitative analysis.
With this background, we now attempt a systemic review
and quantitative analysis of rescue analgesic designs in pediatric trials of four commonly used classes of analgesics for
acute pain: opioids, nonsteroidal antiinflammatory drugs
(NSAIDs), acetaminophen, and local anesthetics. Our
* Available at: www.globalrph.com/narcoticonv.htm. Accessed June
27, 2014.
Available at: www.meta-analysis.com. Accessed June 27, 2014.
Selection Criteria
Studies were included if they met the following criteria: (1)
randomized or controlled clinical trial; (2) children and adolescents aged 18 yr; (3) use of immediate rescue paradigms;
and (4) assessed rescue medication and/or pain scores in
postoperative pain setting.
For the purpose of this study, we included articles only
if they (1) included placebo or control groups; (2) used IV
opioids as rescue medication; and (3) used opioids, NSAIDs,
acetaminophen, or local anesthetics as the study drug, which
was tested for efficacy against a placebo or other control.
We chose to evaluate the following three analgesic sparing
outcomes: (1) rescue opioid usage (milligram per kilogram
per hour), (2) percentage of subjects requiring rescue medication, or (3) time to first rescue medication (minutes). No
language restrictions were applied. In calculating rescue opioid usage, opioids were converted to morphine equivalents.*
In assessing the hourly usage, we chose the shortest reported
timeframe over an hour and divided the morphine equivalent by this timeframe. For articles reporting pain scores, we
standardized all the pain scores to a 0 to 10 scale.5,6
Data Synthesis and Statistical Analysis
Data were extracted independently by two reviewers (J.K.
and C.D.). Inconsistencies were resolved in consensus meetings. Additional study variables were extracted and tabulated, including age range of patients; sample sizes; type(s)
of surgery; dose regimes for study drugs and rescue opioid,
method of pain assessment; criteria for opioid administration; and duration of follow-up. Each included study was
graded for quality and scored using the Jadad criteria.7
Two methods of assessing opioid sparing were used. First,
we used a meta-analytic approach. We estimated Hedges g
and the 95% CI between the study drug and control groups.8
Data management and calculations were performed using
Comprehensive Meta-Analysis, version 2.0. Since considerable heterogeneity was expected, all analyses were performed
with a random-effects model.9 We assessed the presence of
publication bias using the fail-safe N method.10 To assess
Anesthesiology 2015; 122:150-71 151 Kossowsky et al.
Rescue Designs for Pediatric Analgesic Trials
Fig. 2. Flow chart of literature search with summary of excluded and included studies. IM = intramuscular; LA = local anesthetic;
NSAIDs = nonsteroidal antiinflammatory drugs.
heterogeneity between studies, Q-statistics were calculated. A

statistically significant Q indicates a heterogeneous distribution of odds ratios between studies, meaning that systematic
differences, possibly influencing the results, are present.11 In
addition, the degree of inconsistency was quantified by the
I2 statistic, which measures the percentage of variation across
studies that is due to heterogeneity rather than chance.12 A
0% value means no heterogeneity, and higher values represent an increase in heterogeneity. Generally, heterogeneity is
categorized at 25% (low), 50% (moderate), and 75% (high).
Given the large number of possible factors influencing the
standardized effect size, we chose not to conduct any sensitivity analyses to explain the heterogeneity, except for subgroup
analyses with regard to type of surgery to assess its influence
on assay sensitivity and choice of outcome.
Second, mean opioid sparing by the study drug was
expressed as percent of maximum possible effect (placebo
study/placebo),13 as in pharmacologic studies.14
A third approach to data synthesis explored the distribution of pain scores in subjects randomized to study drug or
control conditions. A perfect rescue paradigm should in
theory result in low and nearly equal pain scores in both
active and placebo groups (fig. 1C).1 We wanted to evaluate
the degree to which this aim is achieved in practice and to
depict this balance between assay sensitivity and potential
for unrelieved pain in control subjects. To do this we plotted each of these opioid-sparing endpoints against the mean
pain scores for subjects in the control/placebo group. We
refer to the relationship between these opioid-sparing endpoints and the pain in the controls as the efficacyburden
relationship for each study. This relationship is meant to
assess both how well a study performed in terms of effect size
(opioid sparing between test drug and control groups) and
how severe the pain was in the group that was randomized
not to receive the test drug.
Pain Medicine
Results
Study Selection
The study selection procedure is summarized in figure2.
The summary of characteristics of the studies is shown in
table1.1598 All included studies were controlled blinded trials with quality scores of 3 to 5 on the Jadad scale.7 Selected
results of publication bias and tests of heterogeneity, as well
as efficacy measures of the individual studies and the graphical representation of the efficacyburden relationship are
presented in tables 2 and 3 in Section A3 of Supplemental
Opioids as the Study Drug
Details of the 33 included articles (48 active study drug
arms) can be seen in Section A2 of Supplemental Digital
Content 1, http://links.lww.com/ALN/B91. Two special
cases are noteworthy. Trials using intrathecal or epidural opioids as the study drug gave the largest effect sizes and highest efficacy to burden relationships. Intrathecal and epidural
opioids, especially morphine and hydromorphone, have
markedly higher potency and duration of action compared
to the same opioids given systemically. Conversely, when the
ultrashort-acting opioid remifentanil31 was given intraoperatively as study drug, it produced a significant negative postoperative opioid-sparing effect, that is, greater opioid use in
patients receiving study drug versus placebo, consistent with
previous animal and adult human studies showing remifentanils potential to induce hyperalgesia and acute tolerance.99
Total Opioid Usage (Milligram per Kilogram per Hour). This
endpoint was recorded for 32 study drug arms among 22
studies including a wide range of surgeries.
Significant standardized mean differences (reductions in opioid use) were found for 16 of 28 active study
drug arms (Hedges g = 0.84; 95% CI, 1.22 to 0.47,
P < 0.001) (fig.3). Subgroup analyses found a highest effect
in scoliosis surgeries (Hedges g = 3.23; 95% CI, 5.01
to 1.45, P < 0.001), intermediate effect sizes in thoracic
and cardiac surgeries, (Hedges g = 0.90; 95% CI, 1.25
to 0.55, P < 0.001), and very modest or insignificant
effects in adenotonsillectomies (Hedges g = 0.47; 95%
CI, 1.16 to 0.22, P = 0.18), and urological/abdominal
procedures (Hedges g = 0.17; 95% CI 0.43 to 0.09,
P = 0.20). Significant percent reductions in opioid use
were found in 17 of 32 study drug arms. Intrathecal morphine25 after spinal fusion showed the highest sparing
effect and efficacyburden relationship.
Percentage Requiring Rescue Medication. This endpoint
was recorded for 21 study drug arms among 16 studies. It
was used commonly for ambulatory/short-stay surgeries,
including urologic surgeries and adenotonsillectomies, not
at all for scoliosis surgery or thoracic surgery.
Percentage requiring rescue medication showed significant
standardized mean difference from control in 11 of 21 treatment arms (Hedges g = 0.83; 95% CI, 1.15 to 0.52; P <
0.001). Subgroup analyses showed that the effect magnitude

is higher for the adenotonsillectomies (Hedges g = 0.99;
95% CI, 1.53 to 0.44; P < 0.001) (driven by one article)15
and smaller for urological procedures (Hedges g = 0.42;
95% CI, 0.89 to 0.05; P = 0.08). Ten of 21 study drug
arms demonstrated significant percent opioid-sparing effect.
Time to First Rescue Medication (Minutes). This endpoint
was recorded for 20 study drug arms among 15 studies. It was
used commonly for ambulatory/short-stay surgeries, as well
as in one open-heart surgery and one spinal fusion surgery.
Time to first rescue medication shows a strong mean difference in favor of the study drug versus control in 11 of 19
treatment arms (Hedges g = 1.64; 95% CI, 2.43 to 0.84;
P < 0.001). Subanalyses by the type of surgery found small
differences in the magnitude of the effect for adenotonsillectomies (Hedges g = 1.44; 95% CI, 2.80 to 0.09;
P = 0.04), and urological procedures (Hedges g = 1.86;
95% CI, 3.96 to 0.25; P = 0.08). Thirteen out of 20
treatment arms demonstrated significant percentage opioid-sparing effect. Overall, time to first rescue seems to be
an outcome with high assay sensitivity in single-dose opioid studies.
NSAID as the Study Drug
Content 1, http://links.lww.com/ALN/B91. The largest
number of studies involved adenotonsillectomies (13 articles). Overall the degree of opioid-sparing effect of NSAIDs
varied considerably.
studies. It was used commonly for adenotonsillectomies but
also for one study involving idiopathic scoliosis surgery and
one involving other types of orthopedic surgery. Significant
standardized mean differences (reductions in opioid use)
were found for 10 of 15 active study drug arms (Hedges g
= 0.92; 95% CI, 1.32 to 0.52; P < 0.001) (fig.4). The
effect magnitude was found to be slightly higher in tonsillectomy surgeries than the general mean (Hedges g = 1.15;
95% CI, 1.92 to 0.38; P = 0.003). No other type of surgery could be evaluated separately due to the small number
of studies. Significant percent reductions in opioid use were
found in 11 of 21 study drug arms.
Percentage Needing Rescue Medication. This endpoint was
recorded for 26 study drug arms among 18 studies. It was
used commonly for adenotonsillectomies. Percentage requiring rescue medication showed significant standardized mean
difference from control for 8 of 18 treatment arms (Hedges g
= 0.52; 95% CI, 0.66 to 0.38; P < 0.001). Ten out of 17
articles demonstrated significant percent opioid sparing effect.
was recorded for 15 study drug arms among 11 studies. It was
used commonly for ambulatory/short-stay surgeries, mainly
for adenotonsillectomies. Time to first rescue medication
Table 1. Selected Characteristics of Included Studies

Intervention Drug
Intervention Group
Source
Surgery
Name
Dose
Route
Mean
Age (yr)
SD or
Range
Ali (2008)15
Tonsillectomy
Tramadol
1mg/kg
IV
30
7.53
1.88
Antila (2006)16
Tonsillectomy
Dextromethorphan
Tramadol
1mg/kg
1mg/kg
Oral
IV
30
15
7.46
11.9
1.85
2.40
Ayatollahi (2012)17
Batra (2008)18
Tonsillectomy
Urological
Tramadol
Fentanyl
2mg/kg
0.25 g/kg
Infil
IT
42
14
7.06
0.62
2.21
0.24
Bean-Lijewski
(1996)19
Campbell (1992)20
Other surgeries
Fentanyl
Fentanyl
Meperidine
0.5 g/kg
1 g/kg
1mg/kg
IT
IT
IM
13
15
25
0.57
0.63
4.10
0.08
0.29
2.60
Urological
Fentanyl
1 g/kg
Epidur
17
5.30
3.30
Dawson (2001)21
Doyle (1993a)22
Tonsillectomy
Appendectomy
Dextromethorphan
Morphine
1mg/kg
20 g kg1 h1
Oral
IV
19
20
7.02
10.2
1.98
(612)
Doyle (1993b)23
Appendectomy
Eschertzhuber
(2008)24
Scoliosis
Morphine
Morphine
Morphine
4 g kg1 h1
10 g kg1 h1
5 g/kg
IV
IV
IT
15
15
14
10.4
10.3
15.0
(6.512.9)
(7.212.4)
2.00
Gall (2001)25
Spinal fusion
Morphine
Morphine
15 g/kg
2 g/kg
IT
IT
14
10
15.0
17.0
2.00
3.00
Ganesh (2008)26
Noncardiac
thoracic
Open Heart
Tympanomastoid
Morphine
Fentanyl
5 g/kg
2 g/ml
IT
Epidur
10
16
15.0
0.18
2.00
0.07
Morphine
Dextromethorphan
7 g/kg
1mg/kg
Epidur
Oral
20
19
2.37
12.2
1.46
3.40
Tonsillectomy
Urological
Tramadol
Fentanyl
2mg/kg
1 g/kg
Infil
Epidur
20
17
15.3
4.00
2.20
(36.92)
Urological
Remifentanil
0.3 g kg1 min1
IV
15
2.48
1.18
Krane (1987)32
Other surgeries
Remifentanil
Remifentanil
Morphine
0.6 g kg1 min1

0.9 g kg1 min1
0.1mg/kg
IV
IV
Epidur
15
15
15
2.69
2.72
7.70
1.02
1.31
NR
Lawhorn (1994)33
Lawhorn (1997)34
Mane (2011)35
Other surgeries
Urological
Cleft palate repair
McDonnell (2008)36
40 g/kg
30 g/kg
0.25 g/kg
0.25 g/kg
100 g/kg
Epidur
Epidur
Block
Block
IV
10
100
15
15
18
8.50
4.09
16.4
14.9
14.8
5.30
2.57
NR
NR
1.70
Ozcengiz (2001)37
Ideopathic
Scoliosis
Urological
Butorphanol
Butorphanol
Fentanyl
Meperidine
Morphine
Rosen (1989)38
Rose (1999)39
Open Heart
Tonsillectomy
Tramadol
Morphine
Morphine
Dextromethorphan
2mg/kg
0.03mg/kg
0.075mg/kg
0.5mg/kg
Epidur
Epidur
Epidur
Oral
38
40
16
19
6.85
6.97
2 to 12
7.80
1.80
1.76
NR
1.70
Sharma (2011)40
Other surgeries
Dextromethorphan
Methadone
Methadone
Methadone
1mg/kg
0.1mg/kg
0.2mg/kg
0.3mg/kg
Oral
IV
IV
IV
19
10
10
11
7.90
14.0
13.0
14.0
1.60
2.00
2.00
2.00
Opioids
Hammer (2005)27
Hasan (2004)28
Heiba (2012)29
Kawaraguchi
(2006)30
Kim (2013)31
(Continued)
Pain Medicine
Control Group
Outcome Variables
Type of
Control
Mean
Age (yr)
SD or
Range
Rescue
Opioid
Total
Dose
Placebo control
30
7.61
1.93
Meperidine
Placebo control
15
12.5
1.90
Fentanyl
Placebo control
Add on
42
14
7.40
0.56
1.38
0.29
Fentanyl
Fentanyl
Placebo control
28
4.20
3.00
Meperidine
Not clearly stated
CHEOPS
Add on
17
4.60
3.80
Morphine
N/A
Placebo control
Placebo control
21
20
7.75
9.60
2.70
(612)
Morphine
Morphine
X
X
Rescue medication%
that need rescue
Not clearly stated
Not clearly stated
Placebo control
15
10.5
(8.7012.1)
Morphine
Not clearly stated
N/A
4 points
self-report
mCHEOPS
Control
14
15.0
1.00
Piritamide
Not clearly stated
VAS
Placebo control
10
15.0
4.00
Morphine
Rescue medication
total dose
VAS
Add on
16
0.18
0.14
Nalbuphine
FACES
Control
Placebo control
20
19
2.36
11.5
1.56
3.70
Fentanyl
Morphine
X
X
Placebo control
Add on
20
18
15.2
4.30
3.20
(3.007.08)
Meperidine
Pentazocine
Placebo control
15
2.74
1.19
Remifentanil
Rescue medication
total dose
Pain score
Rescue medication
total dose
Not clearly stated
Rescue medication
Time to first rescue
Rescue medication
total dose
Control
15
7.80
NR
Morphine
Add on
Add on
Add on
10
100
15
8.60
3.97
14.0
4.70
2.37
NR
Morphine
Morphine
Fentanyl
X
X
X
Placebo control
19
14.5
1.90
Morphine
Control
38
6.76
1.76
Morphine
Placebo control
Placebo control
16
19
2 to 12
7.90
NR
1.60
Morphine
Morphine
X
X
Control
30
15.0
2.00
Morphine
Min
X
X
X
X
X
X
X
X
Primary
Outcome
Pain Scale
Rescue medication
number of requests
FACES
Rescue medication
number of rescue
Not clearly stated
Rescue medication
VAS
mCHEOPS
VAS
FACES
VAS
VAS
CHEOPS
mCHEOPS
Rescue medication
Not clearly stated
Not clearly stated
Not clearly stated
N/A
Rescue medication
total dose
Not clearly stated
N/A
Not clearly stated

Rescue medication
total dose
VAS
VAS
Pharmacokinetics
Nurses pain
scale
N/A
N/A
N/A
N/A
(Continued)
Table 1. (Continued)
Intervention Drug
Intervention Group
Source
Surgery
Name
Dose
Route
Mean
Age (yr)
SD or
Range
Suominen (2004)41
Open Heart
Morphine
20 g/kg
IT
35
1.19
(0.0116.7)
Suski (2010)42
Dextromethorphan
3045 mg
Oral
30
15.9
2.40
Tarkkila (2003)43
Ugur (2008)44
Ideopathic
Scoliosis
Tonsillectomy
Tonsillectomy
Umuroglu (2004)45
Tonsillectomy
Viitanen (2001)46
Tonsillectomy
Remifentanil
Tramadol
Tramadol
Morphine
Tramadol
Tramadol
1 g/kg
2mg/kg
2mg/kg
0.1mg/kg
1.5mg/kg
2mg/kg
IV
IM
Infil
IV
IV
IV
25
15
15
15
15
40
3.83
8.20
8.40
7.13
6.06
1 to 3
2.08
1.70
1.60
2.51
2.08
NR
Watcha (1992)47
Other surgeries
Morphine
0.1mg/kg
IV
31
8.50
3.70
Adarsh (2012)48
Cleft palate repair
Diclofenac
1mg/kg
Rectal
30
2.80
1.73
Antila (2006)16
Tonsillectomy
Ketoprofen
2mg/kg
IV
15
12.5
2.30
Bean-Lijewski
(1996)19
Bridge (2000)49
Dawson (1996)50
Other surgeries
Ketorolac
0.75mg/kg
IM
29
4.00
2.20
Strabismus
Cleft palate repair
Ketorolac
Ketorolac
Kokki (1994)51
Elective surgery
Ibuprofen
Kokki (1998)52
Tonsillectomy
Ketoprofen
0.3mg/kg
Kokki (1999a)53
Kokki (1999b)54
Strabismus
Major surgery
Ketoprofen
Ketoprofen
Ketoprofen
Ketoprofen
Kokki (2000)55
Tonsillectomy
Kokki (2001)56
Tonsillectomy
(phase I)
Tonsillectomy
(phase III)
Nonsteroidal
antiinflammatory
drugs
Kokki (2002)57
Tonsillectomy
3mg per eye

Ophthalm
1mg/kg + 0.5mg/
IV
kg QID
40mg kg1 day1
Rectal
17
18
5.25 (4.259.75)
7 to 20
NR
40
2.42
1.11
IV
55
3.33
(1.56.17)
1mg/kg
3mg/kg
1+1mg/kg
1+4mg/kg
IV
IV
IV
IV
55
55
30
24
2.67
2.67
6.92
6.58
(1.257.08)
(1.256.00)
(4.179.33)
(1.5814.8)
Ketoprofen
25mg/kg
IV
42
3.67
(1.178.08)
Ketoprofen
Ketoprofen
25mg/kg
1+1mg/kg
Rectal
IV
42
54
2.75
3.25
(1.422.25)
1.92
Ketoprofen
0.3mg/kg
IV
33
3.50
1.75
Ketoprofen
Ketoprofen
Ketoprofen
1mg/kg
3mg/kg
0.5+3mg/kg preqx
IV
IV
IV
29
32
47
3.25
3.08
10.0
2.17
1.83
1.00
Ketoprofen
IV
42
12.0
3.00
IV
27
7.92
2.92
Kokki (2004)58
Strabismus
Ketoprofen
0.5+3mg/kg
postqx
1+1mg/kg
Korpela (2007)59
Tonsillectomy
Naproxen
10mg/kg
Oral
30
1.80
(0.706.20)
Morton (1999)60
Appendectomy
Diclofenac
1mg/kg TID
Rectal
20
10.6
(6.0013.0)
Munro (2002)61
Ketorolac
0.5mg/kg
IV
20
14.1
1.20
Nikanne (1997)62
Ideopathic
scoliosis
Tonsillectomy
Ketoprofen
1+1mg/kg
IV
80
3.17
(1.009.25)
Oztekin (2002)63
Tonsillectomy
Diclofenac
1mg/kg
Rectal
20
8.40
0.53
(Continued)
Pain Medicine
Control Group
Outcome Variables
Type of
Control
Mean
Age (yr)
SD or
Range
Rescue
Opioid
Total
Dose
Control
36
1.19
(0.0312.7)
Morphine
Placebo control
30
16.5
2.70
Morphine
Placebo control
Placebo control
25
15
2.92
8.50
1.83
2.10
Oxycodone
Meperidine
Placebo control
15
6.96
2.08
Meperidine
Placebo control
40
1 to 3
NR
Meperidine
Placebo control
32
10.0
3.60
Morphine
Control
30
2.26
1.43
Fentanyl
Placebo control
15
12.50
1.90
Fentanyl
Placebo control
28
4.20
3.00
Meperidine
Placebo control
Control
13
16
6.67
7 to 20
(4.1712.4)
NR
Morphine
Morphine
Placebo control
41
2.78
1.32
Morphine
Placebo control
55
3.42
(1.256.58)
Fentanyl
Placebo control
Placebo control
29
23
5.33
7.08
(3.837.83)
(1.2514.4)
Fentanyl
Sulfentanyl
Placebo control
39
3.75
(1.176.08)
Placebo control
45
3.33
Placebo control
35
placebo control
Min
Pain Scale
Rescue medication%
that need rescue
Not clearly stated
N/A
NRS
X
X
Discharge time
Pain score
N/A
VAS
Not clearly stated
CHEOPS
Rescue medication%
that need rescue
Not clearly stated
N/A
X
X
Fentanyl
1.92
Fentanyl
4.00
2.25
Fentanyl
20
11.0
1.00
Oxycodone
Placebo control
29
7.17
3.00
Fentanyl
Placebo control
29
1.60
(0.805.90)
Fentanyl
Control
20
10.15
(513)
Morphine
Placebo control
15
13.9
1.30
Morphine
Placebo control
84
3.33
(0.837.92)
Fentanyl
Control
20
8.90
0.45
Morphine
X
X
Primary
Outcome
VAS
Rescue medication
total dose
Rescue medication
number of doses
Not clearly stated
Hannallah
Pain score
Not clearly stated
CHEOPS
N/A
VAS
CHEOPS
Rescue medication Maunuksela

total dose
Rescue medication% Maunuksela
that need rescue
Vomiting
Rescue medication
total dose
Rescue medication
total dose
Maunuksela
Maunuksela
Lenght of hospital stay
VAS
Rescue medication%
that need rescue
VAS
Rescue medication
number of doses
Rescue medication%
that need rescue
Rescue medication
total dose
Pain score
N/A
Maunuksela
OPS
N/A
VAS

that need rescue
Not clearly stated
VAS
(Continued)
Intervention Drug
Intervention Group
Source
Surgery
Name
Dose
Route
Mean
Age (yr)
SD or
Range
Ryhanen (1994)64
Rugyte (2007)65
Urological
Pectus
Diclofenac
Ketoprofen
1mg/kg
1mg/kg
IM
IV
70
14
3.80
14.00
1.70
(13.015.0)
Sheeran (2004)66
Tonsillectomy
Rofecoxib
0.5mg/kg
Oral
23
7.20
1.80
Sims (1994)67
Sutters (1995)68
Sutters(1999)69
Indomethacin
Ketorolac
Ketorolac
13
45
36
10.1
7.06
12.7
1.80
2.41
3.51
Ketoprofen
2mg/kg
1mg/kg
1mg/kg + 0.5mg/
kg QID
1mg/kg
Rectal
IM
IV
Tuomilehto (2000)70
Abdominal
Tonsillectomy
Orthopedic
Surgery
Adenoidectomy
IV
40
2.67
(1.336.83)
Tuomilehto (2002)71
Tonsillectomy
Ketoprofen
Ketoprofen
1mg/kg
2mg/kg
Oral
IM
40
40
4.17
3.50
(1.678.42)
(1.008.33)
Vetter (1994)72
Viitanen (2003)73
Orthopedics
Tonsillectomy
Ketoprofen
Ketorolac
Ibuprofen
2mg/kg
0.8mg/kg
15mg/kg
IV
IV
Rectal
40
25
41
2.50
13.0
3.20
(1.256.25)
2.00
(1.006.90)
Watcha (1992)47
Other surgeries
Ketorolac
0.9mg/kg
IV
32
8.3
3.80
Bremerich (2001)74
Cleft palate
Dashti (2009)75
Gandhi (2012)76
Tonsillectomy
Ophthalmic
Spine
10mg/kg
20mg/kg
40mg/kg
40mg/kg
40mg/kg
20mg/kg
30mg/kg
Rectal
Rectal
Rectal
Rectal
Rectal
Hiller (2012)77
Acetaminophen
Acetaminophen
Acetaminophen
Acetaminophen
Acetaminophen
Acetaminophen
Acetaminophen
IV
20
20
20
53
48
47
18
0.97
1.01
0.79
10.2
5.60
7.40
15.1
0.73
0.87
0.75
2.84
3.40
3.40
2.00
Kocum (2013)78
Tonsillectomy
Acetaminophen
15mg/kg
IV
40
4.70
1.00
Korpela (1999)79
Elective surgery
Korpela (2007)59
Tonsillectomy
Acetaminophen
Acetaminophen
Acetaminophen
Acetaminophen
20mg/kg
40mg/kg
60mg/kg
10mg/kg
Rectal
Rectal
Rectal
Oral
30
30
30
30
3.10
3.80
4.20
1.30
1.90
2.20
2.30
(0.805.60)
Mercan (2007)80
Inguinal
Morton (1999)60
Appendectomy
Acetaminophen
Acetaminophen
Acetaminophen
Rectal
Rectal
Rectal
65
78
20
4.00
3.92
9.90
2.97
2.93
(5.0012.0)
Van der Marel

(2007)81
Viitanen (2003)73
Abdominal
Acetaminophen
Rectal
29
0.00
(0.000.25)
Tonsillectomy
Acetaminophen
2025mg/kg p3
2025mg/kg p4
20mg/kg + 15mg/
kg QID
3040mg/kg +
20mg/kg QID
40mg/kg
Rectal
40
2.70
(1.006.40)
Carney (2010)82
Appendectomy
Ropivacaine 0.75%
2.5mg/kg
Block
19
NR
(4.0016.0)
Chaudhary (2012)83
Coban (2008)84
Edwards (2011)85
Cardiac
Cleft palate
Appendectomy
Ropicacine 0.5%
Ropivacaine
Bupivacaine 0.25%
0.050.06mg/kg
0.2mg/kg
0.5ml/kg
Block
Infil
Infil
14
10
29
5.50
1.90
11.8
1.82
1.00
(10.812.9)
Giannoni (2001)86
Heiba (2012)29
Hermansson
(2013)87
Inanoglu (2009)88
Tonsillectomy
Tonsillectomy
Abdominal
21
20
17
7.00
14.9
3.40
2.90
2.50
(0.5012.6)
30
6.00
1.30
Acetaminophen
Local anesthetics
Tonsillectomy
Ropivacaine 0.01%
0.15ml/kg
Infil
Lidocaine
2mg/kg
Infil
Bupivacaine
0.20.4mg kg1 h1 SC infusion
Bupivacaine 0.25%
NR
Infil
(Continued)
Pain Medicine
Control Group
Outcome Variables
Type of
Control
Mean
Age (yr)
SD or
Range
Rescue
Opioid
Total
Dose
Control
Placebo control
73
17
3.50
13.0
1.80
(1015)
Meperidine
Morphine
Placebo control
22
7.60
2.20
Morphine
Placebo control
Placebo control
Placebo control
15
42
32
10.7
7.08
12.67
2.10
2.22
4.22
Morphine
Fentanyl
Morphine
Placebo control
20
3.83
(1.258.42)
Fentanyl
Placebo control
40
2.75
(1.337.08)
Fentanyl
Control
Placebo control
25
38
13.0
2.60
2.30
(1.006.00)
Morphine
Meperidine
X
X
Placebo control
32
10.0
3.60
Morphine
Placebo control
20
1.04
0.97
Piritramide
Control
Control
51
30
9.45
6.60
2.22
4.15
Meperidine
Fentanyl
Placebo control
18
14.4
1.90
Oxycodone
Placebo control
40
4.30
1.00
Meperidine
Placebo control
30
4.40
2.10
Morphine
Placebo control
29
1.60
(0.805.90)
Fentanyl
Control
59
3.97
2.92
Meperidine
Control
20
10.15
(5.0013.0)
Morphine
Placebo control
25
0.00
(0.000.08)
Morphine
Placebo control
38
2.60
(1.006.00)
Meperidine
Rescue medication%
that need rescue
N/A
Placebo control
21
NR
(516)
Morphine
CHIPPS
Placebo control
Control
Placebo control
Control
Placebo control
Placebo control
Placebo control
13
10
29
30
21
20
15
5.70
1.80
12.3
11.9
7.40
15.2
2.80
1.58
0.80
(11.313.3)
(10.813.1)
3.70
3.20
(0.65.8)
Fentanyl
Morphine
Morphine
Rescue medication
total dose
Pain score
Not clearly stated
Pain score
Fentanyl
Meperidine
Morphine
Placebo control
30
6.20
1.60
Fentanyl
Min
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Primary
Outcome
Not clearly stated
Rescue medication
total dose
Rescue medication
total dose
Not clearly stated
Rescue medication
Not clearly stated
Pain Scale
N/A
VAS
CHEOPS
VAS
CHEOPS
FACES

that need rescue
Rescue medication%
that need rescue
OPS
Not clearly stated

Rescue medication%
that need rescue
Not clearly stated
VAS
N/A
Not clearly stated
CHIPPS
Not clearly stated

Pain score
VAS
OPS
Rescue medication
total dose
Rescue medication
total dose
Not clearly stated
N/A
VAS
CHEOPS
VAS
Rescue medication%
that need rescue
Not clearly stated
OPS
Rescue medication
total dose
Rescue medication
N/A
Pain score
Not clearly stated
Rescue medication
number of doses
Pain score
VAS
VAS
mOPS
CHIPPS
FACES
VAS
VAS
N/A
CHEOPS
(Continued)
Intervention Drug
Intervention Group
Source
Surgery
Name
Dose
Route
Mean
Age (yr)
SD or
Range
Jagannathan
(2009)89
Klamt (2003)90
Krane (1987)32
Urological
Bupivacaine 0.25%
0.1ml/kg
Block
25
3.62
1.87
Abdominal
Other surgeries
Ropivacaine 0.1%
Bupivacaine 0.25%
0.2ml kg1 h\1

0.1ml/kg
Epidur
Epidur
17
13
2.82
6.20
2.67
NR
Kundra (2006)91
Inguinal hernia
Bupivacaine 0.25%
0.25ml/kg
ILIH-I
34
6.10
3.60
Meara (2010)92
Cleft palate
Bupivacaine 0.25%
Bupivacaine 0.25%
Bupivacaine 0.25%
0.25ml/kg
0.25ml/kg
1ml/h
34
34
32
5.80
5.80
6 to 9
3.10
3.70
NR
Muthukumar
(2012)93
Cleft palate
Lidocaine
7mg/kg
ILIH-M
ILIH-S
SC
infusion
Infil
25
2.10
1.70
Ohara (2004)94
Spinal
Park (2004)95
Ryhanen (1994)64
Tonsillectomy
Urological
7mg/kg
4ml/h
4ml/h
30 mg
1ml/kg
1ml/kg
Infil
Epidur
Epidur
Infil
Epidur
Epidur
25
10
12
66
57
50
2.60
13 to 21
13 to 21
7.00
3.90
3.90
1.90
NR
NR
2.00
1.80
1.80
Splinter (2010)96
Appendectomy
Lidocaine
Bupivacaine 0.1%
Bupivacaine 0.065%
Ropivacaine 0.5%
Bupivacaine 0.25%
Bupivacaine 0.25% +
Epinephrine
Ropivacaine 0.2%
0.250.5ml/kg
Block
18
10.2
3.00
Tirotta (2009)97
Cardiac
0.55ml/h
4.47
(0.2514.7)
Urological
5%
SC
infusion
Topical
35
Usmani (2009)98
Bupivacaine 0.25% /
Levobupivacaine
EMLA cream
30
6.00
2.00
Lidocaine 1%
0.5ml/kg
IV
30
7.00
2.00
APDS = All India Institute of Medical Sciences pain discomfort scale; CHEOPs = Childrens Hospital of Eastern Ontario Pain Scale; CHIPPS = Children and
Infants Postoperative Pain Scale; Epidur = epidural; Faces = The Wong-Baker FACES Pain Rating Scale; FLACC = The Face, Legs, Activity, Cry, Consolability
scale; ILIH-I = inferomedial approach to ilioinguinaliliohypogastric nerve block; ILIH-M = medial approach to ilioinguinaliliohypogastric nerve block; ILIH-S =
superomedial approach to ilioinguinaliliohypogastric nerve block; IM = intramuscular; Infil = infiltration; IT = intrathecal; IV = intravenous; mCHEOPs = modified CHEOPs; mOPS = modified OPS; N/A = not applicable; NRS = numeric rate scale; Ophthalm = ophthalmological drops; OPS = Observational Pain Scale;
SC infusion = subcutaneous infusion; VAS = visual analog scale.
shows a moderate mean difference in favor of the study drug

versus control in one of four treatment arms (Hedges g =
0.32; 95% CI, 0.53 to 0.10; P = 0.004). Since there were
only four studies that could be included, no subgroup analyses
could be conducted. Only 2 of 11 treatment arms demonstrated significant percent opioid-sparing effect and one article
had a slightly negative, yet nonsignificant sparing effect.53
Overall, total opioid dose in milligram per kilogram per
hour seems to be the opioid-sparing outcome with the highest assay sensitivity for NSAID trials.
Acetaminophen as the Study Drug
Content 1, http://links.lww.com/ALN/B91. General conclusions were (1) rectal acetaminophen showed greatest
effect size in all three measures at doses of 40 or 60mg/kg,
and variable effects at lower doses; and (2) only one oral and
two IV acetaminophen studies met full inclusion criteria for
this review; effect size using IV acetaminophen was strongly

influenced by study methodology.
Total Opioid Usage (Milligram per Kilogram per Hour).
This endpoint was recorded for 13 study drug arms among
9 studies. It was used commonly for ambulatory/short-stay
surgeries. Significant standardized mean differences (reductions in opioid use) were found for four of eight active study
drug arms (Hedges g = 2.12; 95% CI, 3.50 to 0.75;
P = 0.002) (fig.5), though this was heavily influenced by
the effectiveness of 40 and 60mg/kg in the study by Korpela
et al.79 Significant percent reductions in opioid use were
found in 4 of 13 study drug arms.
Percentage Needing Rescue Medication. This endpoint
was used commonly to evaluate rectal acetaminophen in
ambulatory/short-stay surgeries. Percentage requiring rescue medication showed significant standardized mean difference from control in 5 of 10 treatment arms (Hedges
g = 0.82; 95% CI, 1.20 to 0.44; P < 0.001). Noteworthy, the effects sizes found in the study reported by
Pain Medicine
Control Group
Outcome Variables
Type of
Control
Mean
Age (yr)
SD or
Range
Rescue
Opioid
Add on
23
4.14
1.89
Morphine
Add on
Control
18
15
3.53
7.80
2.83
NR
Tramadol
Morphine
Placebo control
34
5.20
3.30
Fentanyl
Placebo control
33
6 to 9
NR
Morphine
Pain score
FACES
Placebo control
25
2.70
2.10
Fentanyl
Cardiovascular response
FLACC
Placebo control
25
9
2.90
13 to 21
2.10
NR
Morphine
Not clearly stated
VAS
Placebo control
Control
64
73
7.00
3.50
3.00
1.80
Fentanyl
Meperidine
Pain score
Not clearly stated
OPS
N/A
Placebo control
18
10.6
2.90
Morphine
N/A
Placebo control
37
3.51
(0.2516.7)
Morphine
Placebo control
30
7.00
3.00
Fentanyl
Rescue medication
total dose
Rescue medication
total dose
Rescue medication
number of doses
Korpela et al.79 were much smaller compared to the previous outcome and similar to the group mean. Four of six
study drug arms demonstrated significant percent opioidsparing effect.
was recorded for five study drug arms among four studies.
Time to first rescue medication shows no effect in favor of
the study drug versus control in any arms (Hedges g = 0.07;
95% CI, 0.32 to 0.19; P = 0.60). Since there were only
four studies that could be included, no subgroup analyses
could be conducted. No study found any percent differences
between the study drug and control, including even those
finding significant differences for 40mg/kg rectal acetaminophen when assessing total rescue opioid usage.59,73
Overall, for acetaminophen trials, total opioid dose in
milligram per kilogram per hour seems to be the outcome
with the highest assay sensitivity, yet due to low number of
studies, these results should be considered preliminary.
Total
Dose
Min
Primary
Outcome
Pain Scale
Pain score
CHIPPS
X
X
Not clearly stated

Rescue medication
Rescue medication
total dose
N/A
N/A
X
X
X
APDS
N/A
N/A
Local Anesthetics as the Study Drug

Details of the general characteristic of the 20 included articles (27 active study drug arms) can be seen in Section A2
of Supplemental Digital Content 1, http://links.lww.com/
ALN/B91.
studies. Half of the articles administered the local anesthetics
via infiltration and the other half via peripheral block or epidural. Significant standardized mean differences (reductions
in opioid use) were found for 6 of 14 active study drug arms
(Hedges g = 0.72; 95% CI, 1.18 to 0.27; P = 0.002)
(fig.6). Abdominal surgeries (including inguinal hernia procedures) showed a nonsignificant effect (Hedges g = 0.40;
95% CI, 1.01 to 0.20; P = 0.19). No other type of surgery
could be evaluated due to the low number of studies. Singledose studies showed significant standardized mean differences
(Hedges g = 1.16 95% CI, 1.85 to 0.47; P = 0.001),
whereas studies evaluating continuous infusion showed
none (Hedges g = 0.16; 95% CI, 0.55 to 0.22; P = 0.41).
Fig. 3. Forest plot for opioids as study drug. Expressed as Hedges g score and 95% CIs. Negative scores favor study drug over
control. *Study included more than one active treatment arm.
Significant percent reductions in opioid use were found in

10 of 18 study drug arms.
Percentage Needing Rescue Medication. This endpoint
was used most commonly for abdominal/urological surgeries. Percentage requiring rescue medication showed significant standardized mean difference from control for 11 of
15 treatment arms (Hedges g = 1.19; 95% CI, 1.56 to
0.82; P < 0.001). Abdominal surgeries (including inguinal hernia procedures) were very close to the mean value
(Hedges g = 1.21; 95% CI, 1.72 to 0.70; P < 0.001).
No other type of surgery could be evaluated due to the small

number of studies. Eight of 15 study drug arms demonstrated significant percent opioid-sparing effect.
was used most commonly for abdominal/urological surgeries. Time to first rescue medication shows a strong difference in favor of the study drug versus control in 9 of 10
treatment arms (Hedges g = 1.55; 95% CI, 2.11 to 0.99;
P < 0.001). In subgroup analyses, the urological and abdominal procedures show a smaller magnitude of the effect (Hedges
Pain Medicine
Fig. 4. Forest plot for nonsteroidal antiinflammatory drugs as study drug. Expressed as Hedges g score and 95% CIs. Negative
scores favor study drug over control. *Study included more than one active treatment arm.
Fig. 5. Forest plot for acetaminophen as study drug. Expressed as Hedges g score and 95% CIs. Negative scores favor study
drug over control. *Study included more than one active treatment arm.
Fig. 6. Forest plot for local anesthetics as study drug. Expressed as Hedges g score and 95% CIs. Negative scores favor study
drug over control. *Study included more than one active treatment arm.
g = 1.05; 95% CI, 1.45 to 0.65; P < 0.001). Nine of 12

treatment arms demonstrated significant percent opioid-sparing effect.
Local anesthetic trials were highly variable. They
involved wound infiltration (9 of 20 articles, two of those
with continuous subcutaneous infusion), peripheral nerve
blocks (5 of 20 articles), and epidural blocks (4 of 20
articles, only one of those with continuous infusion).
Some involved single injection, others involved continuous infusions.
Pain Scores
Among the entire group of 85 clinical trials, 62 reported pain
scores. Mean pain scores were 2.31.5 in the study drug arms
and 3.41.2 in the control arms (P < 0.001). Eighteen trials
used PCA or NCA, whereas 44 administered rescue analgesia by nurse-administered boluses. No statistical differences
were found between PCA/NCA and nurse-administered rescue trials on pain scores in the study drug arms (2.51.3
vs. 2.21.6; P = 0.43) and in the control arms (3.41.5 vs.
3.32.1; P = 0.84). Studies using observational measures did
not differ from those using self-report measures (3.61.7 vs.
3.12.1; P = 0.25). No trial using PCA/NCA had a mean
pain score greater than five in either study arm or control
arm. Similar results for the pain score measures were found
when using parametric and nonparametric methods.
A linear regression found no associations between pain
scores in the control group and type of surgery, type of pain
scale (objective vs. self-report), and PCA/NCA versus nurseadministered boluses.
EfficacyBurden Relationships
When opioids were used as the study drug, a positive relationship was found between pain in the control group and
time to first rescue medication ( = 0.37, R2 = 0.44), a
small relationship between pain in the control group and
total opioid use ( = 0.16, R2 = 0.16), but no relationship
with the percentage requiring recue medications ( = 0.06,
R2 = 0.01) (figs. 1 and 2 in Supplemental Digital Content
1, http://links.lww.com/ALN/B91, which depict the efficacy
burden relationship for opioids as the study drug). Although
time to rescue showed good assay sensitivity for opioid sparing, trials using this endpoint had higher pain scores in control
groups compared with trials using the other primary opioidsparing endpoints.
When NSAIDs were used as the study drug, a positive
strong relationship was found between pain in the control
group and time to first rescue medication ( = 0.70, R2 = 0.25),
Pain Medicine
Discussion
Our systematic review and quantitative analysis examined
rescue analgesic designs in pediatric trials of four commonly
used classes of analgesics for acute pain. We considered these
designs from the standpoint of usefulness as a surrogate measure of analgesic efficacy and from the standpoint of burden of
unrelieved pain in the subjects, particularly in control groups.
Fig. 7. Efficacyburden relationship for nonsteroidal antiinflammatory drugs as the study drug. (A) Rescue opioid usage
(milligram per kilogram per hour) as the outcome. (B) Percentage requiring rescue medication as the outcome. (C) Time to
first rescue medication (minutes) as the outcome. Percent
MPE = percent maximum possible effect: placebostudy/
placebo.
a small relationship between pain in the control group and

percentage requiring recue medications ( = 0.18, R2 = 0.025),
but no relationship with the total opioid use ( = 0.12, R2 = 0.01)
(figs.7 and 8).
When acetaminophen was used as the study drug, no
relationship was found between pain in the control group
and percentage requiring recue medications ( = 0.007,
R2 = 0), or total opioid use ( = 0.06, R2 = 0.0001). Since
only one study provided pain scores when looking at time
to first rescue medication, no association could be assessed
(figs. 3 and 4 in Supplemental Digital Content 1, http://
links.lww.com/ALN/B91, which depict the efficacyvurden relationship for acetaminophen as the study drug).
When local anesthetics were used as the study drug, a
positive relationship was found between pain in the control group and time to first rescue medication ( = 0.53,
R2 = 0.53), percentage requiring recue medications ( =
0.29, R2 = 0.40), and total opioid use ( = 0.29, R2 =
0.40) (figs.9 and 10).
Opioid Sparing as a Surrogate Measure of Analgesic

Efficacy
Although opioid sparing could be demonstrated in a high percentage of trials in this systematic review, the magnitude of rescue opioid sparing varied greatly. Some sources of variability
in these trials appears due to: (1) the test drug (dose, bioavailability, intrinsic efficacy, time course of action relative to the
timing of measurements), (2) type of opioid-sparing endpoint
(milligram per kilogram per hour, time to rescue, percent of
subjects needing rescue), (3) method of analysis (Hedges g
vs. percent sparing), (4) type of surgery, and (5) a range of
additional demographic variables. Despite an initial survey
of almost 6,000 abstracts from pediatric analgesic trials using
nested search terms, ultimately only 85 trials fit our inclusion
criteria for quantitative analysis. Analysis indicated great heterogeneity in each of the five sources of variability. Based on
this, recommendations can be only somewhat provisional.
We evaluated the three most common reported rescueanalgesicsparing outcomes: (1) total dose (milligram per
kilogram per hour), (2) percentage of children requiring rescue medication, and (3) time to first rescue medication (minutes). Variations in the design methodologies of the analgesic
clinical trial influence the sensitivity for detecting differences
in each of these outcomes, making it difficult to designate
one of these outcomes as an accepted standard for all studies.
Total dose seems to be the outcome most often chosen. This
parameter can be used in single-dose or multiple-doses trials.
However, this outcome is susceptible to the nonlinear interactions between the study drug and the rescue medication. In
addition, for a single-dose study drug with strong efficacy but
short duration of action, effect size magnitudes will depend
on the time period chosen for recording between-group differences, and use of a long-acting rescue opioid might wash
out between-group differences. One recent trial published
after we completed the systematic review, used PCA via sufentanil as the rescue analgesic to evaluate thoracic paravertebral blockade for the Nuss operation for pectus excavatum.
This study showed excellent opioid sparing, but also had high
pain scores in the control group.100 In future studies evaluating rescue opioids with relatively short context-sensitive
half-times, we believe that it is important to permit escalation of dosing parameters in the setting of unrelieved pain.
Time to first rescue medication appears useful in singledose studies (e.g., nerve blocks, infiltrations) and allows us
to have a more clear view of the primary effect of the study
drug. However, this parameter appears less useful for very
short duration study drugs or multiple dose trials. Finally,
Fig. 8. Nonsteroidal antiinflammatory drugs as the study drug. (A) Rescue opioid usage (milligram per kilogram per hour) as
the outcome. (B) Percentage requiring rescue medication as the outcome. (C) Time to first rescue medication (minutes) as the
outcome. Colorless circles indicate studies without pain scores.
percentage requiring rescue medication can be useful in small

procedures, when the pain scores and the time and amount
of rescue medication needed are low.
Pain Scores, Burden on Control Subjects, and Ethical
Considerations
From an ethical standpoint, it was noteworthy that, among
the 83 studies included in this review, the mean pain scores
in the control arm were mild or moderate (averaging 3.4
out of 10 in a standardized scale), and only very rarely
severe. PCA/NCA paradigms seem particularly effective for
preventing high pain scores among children randomized to

control groups.
Previous research indicates that, in routine clinical use,
patients and nurses do not dose PCA or NCA, respectively,
based solely on pain intensity, but dosing is based also on a
range of other factors, including anticipation of future pain
and on side-effects. In PCA/NCA trials, dosing was not titrated
to give equal scores in the study drug and control groups, but
rather mean pain scores were lower in the study drug arms (P
= 0.013). Based on this, there is a potential for opioid sparing alone to underestimate the analgesic effect of a study drug.
Pain Medicine
endpoints. It is clear that variability in study design, type of

surgery, method of opioid delivery, duration of study drug
administration, and reported outcome measures all impact
on the likelihood and degree of positive findings and thus
our quantitative results should be interpreted as exploratory.
Conclusions and Recommendations for

Future Trials
Fig. 9. Efficacyburden relationship for local anesthetics

as the study drug. (A) Rescue opioid usage (milligram per
kilogram per hour) as the outcome. (B) Percentage requiring
rescue medication as the outcome. (C) Time to first rescue
medication (minutes) as the outcome. Percent MPE = percent
maximum possible effect: placebostudy/placebo.
Future studies may consider the potential utility of composite

efficacy measures based on both pain scores and opioid sparing.
Limitations
As with all quantitative systematic reviews, meta-analyses are
only as good as the data that are reported and the description
of methods in each study. Data of many studies could not be
included due to lack of necessary information. In our metaanalysis, we included multiple treatment arms from a single
study. We are aware that duplicating the number of patients
in a control group between two comparisons may generate a
unit-of-analysis error. This could have been avoided by either
splitting the shared group resulting in a smaller sample size
and including two or more comparisons, by combining
groups to create pairwise comparisons, or by undertaking a
multiple treatment analysis. However, the goal of our study
was not to test a specific drugs efficacy in either the prevention or treatment of pain but to see if immediate rescue
designs are feasible and usable outcomes in both these kinds
of studies, over various types of surgeries and intervention
drugs, and to consider the utility of different opioid-sparing
Immediate rescue analgesic trials show reasonable assay sensitivity and tolerably low burden (low-moderate pain scores)
for children after surgery. From a clinicians standpoint, opioid sparing matters only if it is associated with meaningful
improvements in clinical outcome measures, including pain
scores, reductions in side-effects such as nausea, vomiting,
itching, bowel dysfunction, etc., as well as in the time course
of recovery, rehabilitation, and postoperative behavior.
Acetaminophen, NSAIDs, spinal morphine, and a range of
types of regional anesthesia are indeed effective at reducing
systemic opioid use, but the degree of opioid sparing varies
both with drug, dose, technique, and type of surgery.
Patient-controlled analgesia/nurse-controlled analgesia
paradigms seem ideal for more extensive surgeries, repeated
dosing of the study drug, or trials involving long-acting study
drugs. Surrogate efficacy measures are important for guiding
analgesic prescribing in infants and children, but they are not
the only outcome measures that are essential components of
pediatric analgesic trials. As outlined by previous consensus
groups, including the Initiative on Methods, Measurement, and
Pain Assessment in Clinical Trials,101 measures of safety, sideeffects, hospital stay, complications, behavioral measures, and
functional recovery parameters are essential as well. The combination of these elements will allow for ethical and feasible
study designs in future pediatric analgesic trials.
Acknowledgments
The authors thank Paul A. Bain, Ph.D. (Countway Library of
Medicine, Harvard University, Boston, Massachusetts), and
Alison Clapp, M.L.I.S. (Boston Childrens Hospital, Harvard
Medical School, Boston, Massachusetts), for their assistance
in searching the literature, and Ted Kaptchuk, B.A. (Beth
Israel Deaconess Medical Center, Harvard Medical School,
Boston, Massachusetts), for providing mentoring, which
was supported by National Institutes of Health grant no.
2K24AT004095 (Bethesda, Maryland).
This study was supported by a grant project (P2BSP1_148628) awarded to Dr. Kossowsky by the Swiss National Science Foundation (Bern, Switzerland) and by the
Sara Page Mayo Endowment for Pediatric Pain Research, Education and Treatment (Boston, Massachusetts) to Dr. Berde.
Competing Interests
Correspondence
Address correspondence to Dr. Kossowsky: Department of
Anesthesiology, Perioperative and Pain Medicine, Boston
Fig. 10. Local anesthetics as the study drug. (A) Rescue opioid usage (milligram per kilogram per hour) as the outcome.
(B) Percentage requiring rescue medication as the outcome. (C) Time to first rescue medication (minutes) as the outcome.
Colorless circles indicate studies without pain scores.
Childrens Hospital, 333 Longwood Avenue, Boston, Massachusetts 02446. joe.kossowsky@childrens.harvard.edu.

References
1. Berde CB, Walco GA, Krane EJ, Anand KJ, Aranda JV, Craig
KD, Dampier CD, Finkel JC, Grabois M, Johnston C, Lantos J,
Lebel A, Maxwell LG, McGrath P, Oberlander TF, Schanberg
LE, Stevens B, Taddio A, von Baeyer CL, Yaster M, Zempsky
WT: Pediatric analgesic clinical trial designs, measures,
and extrapolation: Report of an FDA scientific workshop.

Pediatrics 2012; 129:35464
2. Tramr MR, Reynolds DJ, Moore RA, McQuay HJ: When placebo controlled trials are essential and equivalence trials are
inadequate. BMJ 1998; 317:87580
3. Permanand G, Mossialos E, McKee M: The EUs new paediatric medicines legislation: Serving childrens needs? Arch Dis
Child 2007; 92:80811
4. Finkel JC, Finley A, Greco C, Weisman SJ, Zeltzer L:
Transdermal fentanyl in the management of children with
chronic severe pain: Results from an international study.
Cancer 2005; 104:284757
5. Tomlinson D, von Baeyer CL, Stinson JN, Sung L: A systematic review of faces scales for the self-report of pain intensity
in children. Pediatrics 2010; 126:e116898
Pain Medicine
6. von Baeyer CL, Spagrud LJ: Systematic review of observational (behavioral) measures of pain for children and adolescents aged 3 to 18 years. Pain 2007; 127:14050
7. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ,
Gavaghan DJ, McQuay HJ: Assessing the quality of reports of
randomized clinical trials: Is blinding necessary? Control Clin
Trials 1996; 17:112
8. Hedges LV, Olkin I: Statistical Methods for Meta-analysis.
New York, Academic Press, 1985
9. Brockwell SE, Gordon IR: A comparison of statistical methods for meta-analysis. Stat Med 2001; 20:82540
10. Rosenthal R: Meta-analytic Procedures for Social Research.
London, United Kingdom, Sage Publications, 1991
11. Cochran WG: The combination of estimates from different
experiments. Biometrics 1954; 10:10129
12. Higgins JP, Thompson SG, Deeks JJ, Altman DG: Measuring
inconsistency in meta-analyses. BMJ 2003; 327:55760
13. Wong I, St John-Green C, Walker SM: Opioid-sparing effects of
perioperative paracetamol and nonsteroidal anti-inflammatory
drugs (NSAIDs) in children. Paediatr Anaesth 2013; 23:47595
14. Dewey WL, Harris LS, Howes JF, Nuite JA: The effect of various neurohumoral modulators on the activity of morphine
and the narcotic antagonists in the tail-flick and phenylquinone tests. J Pharmacol Exp Ther 1970; 175:43542
15. Ali SM, Shahrbano S, Ulhaq TS: Tramadol for pain relief in
children undergoing adenotonsillectomy: A comparison with
dextromethorphan. Laryngoscope 2008; 118:15479
16. Antila H, Manner T, Kuurila K, Salanter S, Kujala R, Aantaa
R: Ketoprofen and tramadol for analgesia during early recovery after tonsillectomy in children. Paediatr Anaesth 2006;
16:54853
17. Ayatollahi V, Behdad S, Hatami M, Moshtaghiun H,

Baghianimoghadam B: Comparison of peritonsillar infiltration effects of ketamine and tramadol on post tonsillectomy
pain: A double-blinded randomized placebo-controlled clinical trial. Croat Med J 2012; 53:15561
18. Batra YK, Lokesh VC, Panda NB, Rajeev S, Rao KL: Doseresponse study of intrathecal fentanyl added to bupivacaine
in infants undergoing lower abdominal and urologic surgery.
Paediatr Anaesth 2008; 18:6139
19. Bean-Lijewski JD, Hunt RD: Effect of ketorolac on bleeding
time and postoperative pain in children: A double-blind, placebo-controlled comparison with meperidine. J Clin Anesth
1996; 8:2530
20. Campbell FA, Yentis SM, Fear DW, Bissonnette B: Analgesic
efficacy and safety of a caudal bupivacaine-fentanyl mixture
in children. Can J Anaesth 1992; 39:6614
21. Dawson GS, Seidman P, Ramadan HH: Improved postoperative pain control in pediatric adenotonsillectomy with dextromethorphan. Laryngoscope 2001; 111:12236
22. Doyle E, Robinson D, Morton NS: Comparison of patientcontrolled analgesia with and without a background infusion
after lower abdominal surgery in children. Br J Anaesth 1993;
71:6703
23. Doyle E, Harper I, Morton NS: Patient-controlled analgesia
with low dose background infusions after lower abdominal
surgery in children. Br J Anaesth 1993; 71:81822
24. Eschertzhuber S, Hohlrieder M, Keller C, Oswald E,

Kuehbacher G, Innerhofer P: Comparison of high- and lowdose intrathecal morphine for spinal fusion in children. Br J
Anaesth 2008; 100:53843
25. Gall O, Aubineau JV, Bernire J, Desjeux L, Murat I: Analgesic
effect of low-dose intrathecal morphine after spinal fusion in
children. Anesthesiology 2001; 94:44752
26. Ganesh A, Adzick NS, Foster T, Cucchiaro G: Efficacy of addition of fentanyl to epidural bupivacaine on postoperative
analgesia after thoracotomy for lung resection in infants.
27. Hammer GB, Ramamoorthy C, Cao H, Williams GD, Boltz

MG, Kamra K, Drover DR: Postoperative analgesia after spinal blockade in infants and children undergoing cardiac surgery. Anesth Analg 2005; 100:12838
28. Hasan RA, Kartush JM, Thomas JD, Sigler DL: Oral dextromethorphan reduces perioperative analgesic administration
in children undergoing tympanomastoid surgery. Otolaryngol
Head Neck Surg 2004; 131:7116
29. Heiba MH, Atef A, Mosleh M, Mohamed R, El-Hamamsy M:
Comparison of peritonsillar infiltration of tramadol and lidocaine for the relief of post-tonsillectomy pain. J Laryngol
Otol 2012; 126:113841
30. Kawaraguchi Y, Otomo T, Ota C, Uchida N, Taniguchi A,
Inoue S: A prospective, double-blind, randomized trial
of caudal block using ropivacaine 0.2% with or without
fentanyl 1 microg kg-1 in children. Br J Anaesth 2006;
97:85861
31. Kim SH, Lee MH, Seo H, Lee IG, Hong JY, Hwang JH:

Intraoperative infusion of 0.6-0.9 gkg(-1)min(-1) remifentanil induces acute tolerance in young children after
laparoscopic ureteroneocystostomy. Anesthesiology 2013;
118:33743
32. Krane EJ, Jacobson LE, Lynn AM, Parrot C, Tyler DC: Caudal
morphine for postoperative analgesia in children: A comparison with caudal bupivacaine and intravenous morphine.
33. Lawhorn CD, Brown RE Jr: Epidural morphine with butorphanol in pediatric patients. J Clin Anesth 1994; 6:914
34. Lawhorn CD, Stoner JM, Schmitz ML, Brown RE Jr, Stewart
FW, Volpe P, Shirey R: Caudal epidural butorphanol plus
bupivacaine versus bupivacaine in pediatric outpatient genitourinary procedures. J Clin Anesth 1997; 9:1038
35. Mane RS, Sanikop CS, Dhulkhed VK, Gupta T: Comparison
of bupivacaine alone and in combination with fentanyl or
pethidine for bilateral infraorbital nerve block for postoperative analgesia in paediatric patients for cleft lip repair: A
prospective randomized double blind study. J Anaesthesiol
Clin Pharmacol 2011; 27:236
36. McDonnell C, Zaarour C, Hull R, Thalayasingam P, Pehora
C, Ahier J, Crawford MW: Pre-treatment with morphine does
not prevent the development of remifentanil-induced hyperalgesia. Can J Anaesth 2008; 55:8138
37. Ozcengiz D, Gunduz M, Ozbek H, Isik G: Comparison of
caudal morphine and tramadol for postoperative pain control in children undergoing inguinal herniorrhaphy. Paediatr
Anaesth 2001; 11:45964
38. Rosen KR, Rosen DA: Caudal epidural morphine for con
trol of pain following open heart surgery in children.
39. Rose JB, Cuy R, Cohen DE, Schreiner MS: Preoperative oral
dextromethorphan does not reduce pain or analgesic consumption in children after adenotonsillectomy. Anesth Analg
1999; 88:74953
40. Sharma A, Tallchief D, Blood J, Kim T, London A, Kharasch
ED: Perioperative pharmacokinetics of methadone in adolescents. Anesthesiology 2011; 115:115361
41. Suominen PK, Ragg PG, McKinley DF, Frawley G, But WW,
Eyres RL: Intrathecal morphine provides effective and safe
analgesia in children after cardiac surgery. Acta Anaesthesiol
Scand 2004; 48:87582
42. Suski M, Bujak-Gizycka B, Madej J, Kacka K, Dobrogowski J,
Woron J, Olszanecki R, Korbut R: Co-administration of dextromethorphan and morphine: Reduction of post-operative
pain and lack of influence on morphine metabolism. Basic
Clin Pharmacol Toxicol 2010; 107:6804
43. Tarkkila P, Viitanen H, Mennander S, Annila P: Comparison
of remifentanil versus ketamine for paediatric day case adenoidectomy. Acta Anaesthesiol Belg 2003; 54:21722
44. Ugur MB, Yilmaz M, Altunkaya H, Cinar F, Ozer Y, Beder L:

Effects of intramuscular and peritonsillar injection of tramadol before tonsillectomy: A double blind, randomized, placebo-controlled clinical trial. Int J Pediatr Otorhinolaryngol
2008; 72:2418
45. Umurolu T, Eti Z, Cifti H, Yilmaz G F: Analgesia for
adenotonsillectomy in children: A comparison of morphine,
ketamine and tramadol. Paediatr Anaesth 2004; 14:56873
46. Viitanen H, Annila P: Analgesic efficacy of tramadol 2mg

kg(-1) for paediatric day-case adenoidectomy. Br J Anaesth
2001; 86:5725
47. Watcha MF, Jones MB, Lagueruela RG, Schweiger C, White
PF: Comparison of ketorolac and morphine as adjuvants during pediatric surgery. Anesthesiology 1992; 76:36872
48. Adarsh E, Mane R, Sanikop C, Sagar S: Effect of pre-operative
rectal diclofenac suppository on post-operative analgesic
requirement in cleft palate repair: A randomised clinical trial.
Indian J Anaesth 2012; 56:2659
49. Bridge HS, Montgomery CJ, Kennedy RA, Merrick PM:

Analgesic efficacy of ketorolac 0.5% ophthalmic solution
(Accular) in paediatric strabismus surgery. Paediatr Anaesth
2000; 10:5216
50. Dawson KH, Egbert MA, Myall RW: Pain following iliac crest
bone grafting of alveolar clefts. J Craniomaxillofac Surg 1996;
24:1514
51. Kokki H, Hendolin H, Maunuksela EL, Vainio J, Nuutinen
L: Ibuprofen in the treatment of postoperative pain in
small children. A randomized double-blind-placebo controlled parallel group study. Acta Anaesthesiol Scand 1994;
38:46772
52. Kokki H, Nikanne E, Tuovinen K: I.v. intraoperative ketoprofen in small children during adenoidectomy: A dose-finding
study. Br J Anaesth 1998; 81:8704
53. Kokki H, Homan E, Tuovinen K, Purhonen S: Perioperative
treatment with i.v. ketoprofen reduces pain and vomiting in
children after strabismus surgery. Acta Anaesthesiol Scand
1999; 43:138
54. Kokki H, Tuovinen K, Hendolin H: The effect of intravenous
ketoprofen on postoperative epidural sufentanil analgesia in
children. Anesth Analg 1999; 88:103641
55. Kokki H, Tuomilehto H, Tuovinen K: Pain management after
adenoidectomy with ketoprofen: Comparison of rectal and
intravenous routes. Br J Anaesth 2000; 85:83640
56. Kokki H, Salonen A, Nikanne E: Perioperative intravenous
ketoprofen neither prolongs operation time nor delays discharge after adenoidectomy in children. Paediatr Anaesth
2001; 11:5964
57. Kokki H, Salonen A: Comparison of pre- and postoperative
administration of ketoprofen for analgesia after tonsillectomy in children. Paediatr Anaesth 2002; 12:1627
58. Kokki H, Purhonen S, Tersvirta M, Ylnen P: Ketoprofen
for add-on pain treatment to paracetamol after strabismus
surgery in children. Clin Drug Investig 2004; 24:23744
59. Korpela R, Silvola J, Laakso E, Meretoja OA: Oral naproxen
but not oral paracetamol reduces the need for rescue analgesic after adenoidectomy in children. Acta Anaesthesiol Scand
2007; 51:72630
60. Morton NS, OBrien K: Analgesic efficacy of paracetamol and
diclofenac in children receiving PCA morphine. Br J Anaesth
1999; 82:7157
61. Munro HM, Walton SR, Malviya S, Merkel S, Voepel-Lewis T,
Loder RT, Farley FA: Low-dose ketorolac improves analgesia
and reduces morphine requirements following posterior spinal fusion in adolescents. Can J Anaesth 2002; 49:4616
62. Nikanne E, Kokki H, Tuovinen K: Comparison of perioperative ketoprofen 2.0mg kg-1 with 0.5mg kg-1 i.v. in
small children during adenoidectomy. Br J Anaesth 1997;
79:6068
63. Oztekin S, Hepaular H, Kar AA, Ozzeybek D, Artikaslan

O, Elar Z: Preemptive diclofenac reduces morphine use after
remifentanil-based anaesthesia for tonsillectomy. Paediatr
Anaesth 2002; 12:6949
64. Ryhnen P, Adamski J, Puhakka K, Leppluoto J, Vuolteenaho
O, Ryhnen J: Postoperative pain relief in children. A comparison between caudal bupivacaine and intramuscular
diclofenac sodium. Anaesthesia 1994; 49:5761
65. Rugyte D, Kokki H: Intravenous ketoprofen as an adjunct
to patient-controlled analgesia morphine in adolescents with
thoracic surgery: A placebo controlled double-blinded study.
Eur J Pain 2007; 11:6949
66. Sheeran PW, Rose JB, Fazi LM, Chiavacci R, McCormick L:
Rofecoxib administration to paediatric patients undergoing
adenotonsillectomy. Paediatr Anaesth 2004; 14:57983
67. Sims C, Johnson CM, Bergesio R, Delfos SJ, Avraamides EA:
Rectal indomethacin for analgesia after appendicectomy in
children. Anaesth Intensive Care 1994; 22:2725
68. Sutters KA, Levine JD, Dibble S, Savedra M, Miaskowski C:
Analgesic efficacy and safety of single-dose intramuscular
ketorolac for postoperative pain management in children
following tonsillectomy. Pain 1995; 61:14553
69. Sutters KA, Shaw BA, Gerardi JA, Hebert D: Comparison

of morphine patient-controlled analgesia with and without
ketorolac for postoperative analgesia in pediatric orthopedic
surgery. Am J Orthop (Belle Mead NJ) 1999; 28:3518
70. Tuomilehto H, Kokki H, Tuovinen K: Comparison of intravenous and oral ketoprofen for postoperative pain after adenoidectomy in children. Br J Anaesth 2000; 85:2247
71. Tuomilehto H, Kokki H: Parenteral ketoprofen for pain management after adenoidectomy: Comparison of intravenous
and intramuscular routes of administration. Acta Anaesthesiol
Scand 2002; 46:1849
72. Vetter TR, Heiner EJ: Intravenous ketorolac as an adjuvant to
pediatric patient-controlled analgesia with morphine. J Clin
Anesth 1994; 6:1103
73. Viitanen H, Tuominen N, Vrniemi H, Nikanne E, Annila
P: Analgesic efficacy of rectal acetaminophen and ibuprofen
alone or in combination for paediatric day-case adenoidectomy. Br J Anaesth 2003; 91:3637
74. Bremerich DH, Neidhart G, Heimann K, Kessler P, Behne
M: Prophylactically-administered rectal acetaminophen does
not reduce postoperative opioid requirements in infants and
small children undergoing elective cleft palate repair. Anesth
Analg 2001; 92:90712
75. Dashti GA, Amini S, Zanguee E: The prophylactic effect of
rectal acetaminophen on postoperative pain and opioid
requirements after adenotonsillectomy in children. Middle
East J Anesthesiol 2009; 20:2459
76. Gandhi R, Sunder R: Postoperative analgesic efficacy of single high dose and low dose rectal acetaminophen in pediatric ophthalmic surgery. J Anaesthesiol Clin Pharmacol 2012;
28:4604
77. Hiller A, Helenius I, Nurmi E, Neuvonen PJ, Kaukonen
M, Hartikainen T, Korpela R, Taivainen T, Meretoja OA:
Acetaminophen improves analgesia but does not reduce opioid requirement after major spine surgery in children and
adolescents. Spine (Phila Pa 1976) 2012; 37:E122531
78. Kocum AI, Sener M, Caliskan E, Bozdogan N, Micozkadioglu
D, Yilmaz I, Aribogan A: Intravenous paracetamol and dipyrone for postoperative analgesia after day-case tonsillectomy
in children: A prospective, randomized, double blind, placebo
controlled study. Braz J Otorhinolaryngol 2013; 79:8994
79. Korpela R, Korvenoja P, Meretoja OA: Morphine-sparing

effect of acetaminophen in pediatric day-case surgery.
80. Mercan A, Sayin MM, Saydam S, Ozmert S, Tiryaki T: When to
add supplemental rectal paracetamol for postoperative analgesia with caudal bupivacaine in children? A prospective,
Pain Medicine
double-blind, randomized study. Paediatr Anaesth 2007;

17:54751
81. van der Marel CD, Peters JW, Bouwmeester NJ, Jacqz-Aigrain
E, van den Anker JN, Tibboel D: Rectal acetaminophen does
not reduce morphine consumption after major surgery in
young infants. Br J Anaesth 2007; 98:3729
82. Carney J, Finnerty O, Rauf J, Curley G, McDonnell JG, Laffey
JG: Ipsilateral transversus abdominis plane block provides
effective analgesia after appendectomy in children: A randomized controlled trial. Anesth Analg 2010; 111:9981003
83. Chaudhary V, Chauhan S, Choudhury M, Kiran U, Vasdev S,
Talwar S: Parasternal intercostal block with ropivacaine for
postoperative analgesia in pediatric patients undergoing cardiac surgery: A double-blind, randomized, controlled study. J
Cardiothorac Vasc Anesth 2012; 26:43942
84. Coban YK, Senoglu N, Oksuz H: Effects of preoperative
local ropivacaine infiltration on postoperative pain scores in
infants and small children undergoing elective cleft palate
repair. J Craniofac Surg 2008; 19:12214
85. Edwards TJ, Carty SJ, Carr AS, Lambert AW: Local anaesthetic
wound infiltration following paediatric appendicectomy: A
randomised controlled trial: Time to stop using local anaesthetic wound infiltration following paediatric appendicectomy? Int J Surg 2011; 9:3147
86. Giannoni C, White S, Enneking FK, Morey T: Ropivacaine
with or without clonidine improves pediatric tonsillectomy
pain. Arch Otolaryngol Head Neck Surg 2001; 127:126570
87. Hermansson O, George M, Wester T, Christofferson R: Local
delivery of bupivacaine in the wound reduces opioid requirements after intraabdominal surgery in children. Pediatr Surg
Int 2013; 29:4514
88. Inanoglu K, Ozbakis Akkurt BC, Turhanoglu S, Okuyucu S,
Akoglu E: Intravenous ketamine and local bupivacaine infiltration are effective as part of a multimodal regime for reducing
post-tonsillectomy pain. Med Sci Monit 2009; 15:CR539543
89. Jagannathan N, Sohn L, Sawardekar A, Ambrosy A, Hagerty
J, Chin A, Barsness K, Suresh S: Unilateral groin surgery in
children: Will the addition of an ultrasound-guided ilioinguinal nerve block enhance the duration of analgesia of a
single-shot caudal block? Paediatr Anaesth 2009; 19:8928
90. Klamt JG, Garcia LV, Stocche RM, Meinberg AC: Epidural

infusion of clonidine or clonidine plus ropivacaine for postoperative analgesia in children undergoing major abdominal
surgery. J Clin Anesth 2003; 15:5104
91. Kundra P, Sivashanmugam T, Ravishankar M: Effect of needle
insertion site on ilioinguinal-iliohypogastric nerve block in
children. Acta Anaesthesiol Scand 2006; 50:6226
92. Meara DJ, Livingston NR, Sittitavornwong S, Ness T, Boyce
J, Wang D, Waite PD: Continuous infusion of bupivacaine
for pain control after anterior iliac crest bone grafting for
alveolar cleft repair in children. Cleft Palate Craniofac J 2011;

48:6904
93.
Muthukumar M, Arya VK, Mathew PJ, Sharma RK:
Comparison of haemodynamic responses following different concentrations of adrenaline with and without lignocaine for surgical field infiltration during cleft lip and cleft
palate surgery in children. Anaesth Intensive Care 2012;
40:1149
94. OHara JF Jr, Cywinski JB, Tetzlaff JE, Xu M, Gurd AR,
Andrish JT: The effect of epidural vs intravenous analgesia
for posterior spinal fusion surgery. Paediatr Anaesth 2004;
14:100915
95. Park AH, Pappas AL, Fluder E, Creech S, Lugo RA, Hotaling
A: Effect of perioperative administration of ropivacaine
with epinephrine on postoperative pediatric adenotonsillectomy recovery. Arch Otolaryngol Head Neck Surg 2004;
130:45964
96. Splinter WM, Thomson ME: Somatic paravertebral block
decreases opioid requirements in children undergoing
appendectomy. Can J Anaesth 2010; 57:20610
97. Tirotta CF, Munro HM, Salvaggio J, Madril D, Felix DE,
Rusinowski L, Tyler C, Decampli W, Hannan RL, Burke RP:
Continuous incisional infusion of local anesthetic in pediatric patients following open heart surgery. Paediatr Anaesth
2009; 19:5716
98. Usmani H, Pal Singh S, Quadir A, Chana RS: A comparison
between EMLA cream application versus lidocaine infiltration for postoperative analgesia after inguinal herniotomy in
children. Reg Anesth Pain Med 2009; 34:1069
99. Joly V, Richebe P, Guignard B, Fletcher D, Maurette P,
Sessler DI, Chauvin M: Remifentanil-induced postoperative
hyperalgesia and its prevention with small-dose ketamine.
100. Qi J, Du B, Gurnaney H, Lu P, Zuo Y: A prospective randomized observer-blinded study to assess postoperative analgesia provided by an ultrasound-guided bilateral thoracic
paravertebral block for children undergoing the Nuss procedure. Reg Anesth Pain Med 2014; 39:20813
101. McGrath PJ, Walco GA, Turk DC, Dworkin RH, Brown MT,
Davidson K, Eccleston C, Finley GA, Goldschneider K,
Haverkos L, Hertz SH, Ljungman G, Palermo T, Rappaport
BA, Rhodes T, Schechter N, Scott J, Sethna N, Svensson OK,
Stinson J, von Baeyer CL, Walker L, Weisman S, White RE,
Zajicek A, Zeltzer L; PedIMMPACT: Core outcome domains
and measures for pediatric acute and chronic/recurrent pain
clinical trials: PedIMMPACT recommendations. J Pain 2008;
9:77183
A Randomized Control Trial of Bupivacaine and

Fentanyl versus Fentanyl-only for Epidural Analgesia
during the Second Stage of Labor
Margaret G. Craig, M.D., Erica N. Grant, M.D., M.Sc., Weike Tao, M.D., Donald D. McIntire, Ph.D.,
Kenneth J. Leveno, M.D.
ABSTRACT
Background: The purpose of this prospective, double-blinded, parallel-arm, randomized trial was to examine the effects of
epidural bupivacaine on the length of the second stage of labor in nulliparous women.
Methods: The authors assessed length of second-stage labor, degree of motor blockade, mode of delivery, and visual analog
scores in 310 nulliparous women with labor epidurals randomized to receive either: (1) 0.125% bupivacaine and fentanyl 2
g/ml or (2) fentanyl 10 g/ml alone via epidural using double blinding.
Results: The median duration of the second stage was 75min (41, 128) in the bupivacaine/fentanyl group versus 73min (42,
120) in the fentanyl-only group (P = 0.17) with a median difference of 6.0 (95% CI, 6.0 to 18.0). Furthermore, there was
no difference in degree of motor blockade, incidence of operative delivery, visual analog scores, or neonatal outcomes between
the two groups. No adverse events were reported.
Conclusions: Use of epidural bupivacaine/fentanyl or a fentanyl-only infusion during the second stage of labor did not
affect the duration of the second stage of labor, degree of motor blockade, mode of delivery, pain relief, and maternal or
neonatal outcomes. However, in the fentanyl-only infusion group, there was a fivefold increase in opioid exposure to the
fetus with unknown effects on neurobehavior, an outcome not assessed beyond the immediate postnatal period in this study.
NALGESIA administered via epidural catheters is the

most popular method of pain relief during childbirth
with approximately two thirds of American women receiving such analgesia.1 There has been longstanding concern that
epidural analgesia interferes with labor which has prompted
several studies to be performed at Parkland Hospital (Dallas, Texas) to determine the effects of epidural analgesia on
labor.28 These studies in the aggregate included randomization of 4,465 women, 2,703 of which were nulliparous.
Results from these studies have shown that epidural analgesia
did not lead to increased cesarean deliveries, but did lengthen
both the first and second stages of labor. However, these differences were small, with the first stage being lengthened by an
average of 36min and the second stage by 15min.8 Although
these increased labor intervals seem small, they were of clinical
importance since operative vaginal delivery rates were increased
with epidural analgesia. This increased operative delivery rate
was attributed to increased lower extremity motor blockade
due to local anesthetic agents delivered via epidural analgesia.
Our study was designed to assess whether the degree of
lower extremity motor blockade during the second stage of
labor could adversely affect the birthing process by interfering with maternal expulsive efforts during delivery. We
hypothesized that the group receiving epidural fentanyl in

Epidural analgesia through the second stage of labor is associated in some studies with prolonged second-stage duration
and increased instrumental deliveries
Whether providing epidural analgesia without local anesthetics
alters these outcomes is unknown

In 310 nulliparous women with epidural analgesia randomized
at the onset of second stage to receive epidural fentanyl alone
or with bupivacaine, there was no difference in duration of
second stage, degree of motor block, or instrumental delivery
To achieve similar degrees of analgesia, women receiving epidural fentanyl without bupivacaine required a fivefold
increased dose of fentanyl
comparison with epidural bupivacaine and fentanyl would

have decreased lower extremity motor blockade and thus a
shortened second stage of labor.

Study Design
This prospective, double-blinded, parallel-arm, randomized
trial was approved by The University of Texas Southwestern
This article is featured in This Month in Anesthesiology, page 1A.

Submitted for publication November 6, 2013. Accepted for publication August 22, 2014. From the Department of Anesthesiology and Pain
Management (M.G.C., E.N.G., W.T.) and Department of Obstetrics and Gynecology (D.D.M., K.J.L.), The University of Texas Southwestern
Medical Center, Dallas, Texas.
January 2015
PAIN MEDICINE
Institutional Review Board, Dallas, Texas, registered with

ClinicalTrials.gov (NCT01621230), and designed to assess
the superiority of epidural fentanyl-only versus bupivacaine
and fentanyl on the length of second stage of labor. This
study was monitored by a Safety and Data Monitoring
Board.
All women presenting for labor induction at Parkland
Hospital, Monday through Friday, were screened for inclusion in the study. The labor and delivery unit is a closed staff
unit; therefore, protocols derived from labor and delivery,
described in Williams Obstetrics Section 4, were used by
the obstetrical service reducing potential bias in management.9 During labor, eligible women who had an existing
epidural were approached for participation and research staff
obtained informed written consent. Three hundred ten nulliparous women at term (37 weeks), in active labor and
with an existing, functional labor epidural in place were
enrolled into the study. The primary outcome was the length
of the second stage of labor. Secondary outcomes of interest
were motor strength score,10 mode of delivery, intravenous
meperidine bolus requirements, satisfaction with pain relief,
and indices of maternal and neonatal well-being.
Methods
During the study period, labor and delivery at Parkland
Hospital was staffed with a protocol-driven dedicated
obstetrical anesthesiology service to ensure minimal variation in practice. Epidural analgesia was achieved via placement of an epidural catheter, subsequent 3ml of 1.5%
lidocaine with epinephrine 1:200,000 as test dose solution
for possible intravascular or intrathecal placement, followed
by 0.25% bupivacaine and 100 g of fentanyl as the initial
bolus using our standard continuous epidural kit (B. Braun
Medical Inc., Bethlehem, PA). The initial bolus was then
followed with the standard epidural solution used at our
institution, 0.125% bupivacaine with 2 g/ml fentanyl,
infused at 10ml/h. Anesthesia providers were notified by
the nurse of patients reporting a visual analog score (VAS)
of greater than 2 for assessment and possible intervention.
If needed, additional local anesthetic epidural boluses consisting of either bupivacaine 0.25% (typically), bupivacaine
0.5%, and/or lidocaine 2% with epinephrine 1:200,000
were given per the anesthesia provider. If these epidural
boluses were determined to be inadequate for pain relief,
all participants could receive 25mg of intravenous meperidine every hour. These boluses were in addition to patientcontrolled epidural analgesia demand boluses that were
programmed to deliver 5ml of the study drug every 30min
once enrolled into the study.
Routine cervical examinations were performed approximately every 2h and randomization occurred at any point
between 8 and 10cm cervical dilation, with subsequent
infusion of either our standard epidural solution of 0.125%
bupivacaine with 2 g/ml of fentanyl or epidural fentanylonly (10 g/ml concentration). In both groups, infusion
rates were started at 10ml/h. Infusions were prepared by

Parklands Investigational Drug Services Pharmacy in identical, unlabeled cartridges as part of the blinding effort.
Data Collection
Research personnel performed motor strength assessments
(table1), as well as VAS at randomization, 10cm cervical
dilation, and every 30min thereafter until delivery. According to the motor strength scale, a score of 5 indicates no
weakness in hip flexion. To assess this variable, subjects were
asked to flex their knees and lift buttocks off of the bed without assistance. VAS was measured on a linear scale from 0 to
10 (0 = no pain; 10 = worst possible pain). Epidural analgesia was provided per study protocol until either spontaneous delivery of the neonate or operative vaginal delivery
or cesarean birth was planned. At this point, the study was
terminated and routine anesthesia care resumed.
Women were asked within approximately 1h after delivery if they were satisfied with their analgesia and if they
would choose to have the same form of analgesia with subsequent deliveries. In addition, breastfeeding status of the
parturient was also assessed upon discharge. All data were
abstracted by research staff using study-specific forms.
The sample size for this study was determined using historical
data from Parkland Hospital. Randomization was performed
in permuted blocks with computer-generated 1:1 allocation
sequence. Epidural infusion cassettes were sequentially numbered. The baseline rate for the primary outcome and the
length of second stage were estimated for women with singleton, term gestation pregnancies, requiring labor induction,
and without labor epidural analgesia. The median (first and
third quartile) lengths of the second stage of labor were estimated at 28min (15, 58) in women without epidural analgesia. Assuming a one-third increase in the length of the second
stage to 37min due to epidural bupivacaine, a sample size of
155 per arm (310 total) was required for 80% power to detect
such a difference using a two-sided Wilcoxon rank sum test.
Length of second stage was analyzed using Wilcoxon rank
Table 1. Motor Strength Score
Score
1
2
3
4
5
6
Criteria
Unable to move feet or knees (complete block)
Able to move feet only
Able to move knees
Detectable weakness of hip flexion while supine
Ability to flex knees and lift buttock off bed
(detection of hip flexion)
Able to perform partial knee bend
Adapted from modified Bromage description of degree of motor blockade

(Reproduced, with permission, from Breen TW et al. Epidural anesthesia
for labor in an ambulatory patient. Anesth Analg 1993; 77:91924).10 Adaptations are themselves works protected by copyright. So in order to publish
this adaptation, authorization must be obtained both from the owner of
the copyright in the original work and from the owner of copyright in the
translation or adaptation.
Anesthesiology 2015; 122:172-7 173 Craig et al.
Epidural Analgesia Effects on Second-stage Labor
Table 2. Maternal Demographic Characteristics and Antepartum

Complications in Women Randomized to Bupivacaine vs. No
Bupivacaine during the Second Stage of Labor
Characteristic
Fig. 1. Participant flow diagram.
sum test. Motor strength and VAS were analyzed using a

mixed-effect model to estimate the mean (SEM) under the
construct of repeated observations per subject. No assumptions were made restricting the estimation of the covariance
structure. Treatment assignment was assumed to be a fixed
effect and time, a random effect, accounting for the incompleteness of observations over time. The reported P values
were TukeyKramer adjusted for multiple testing of these
pairwise comparisons. The remaining outcomes are categorical and were analyzed using Pearson chi-square test. Analysis
of results was performed using intent-to-treat in all subjects.
Statistical significance was accepted with a P value of less than
0.05 (SAS version 9.2; SAS Institute, Inc., Cary, NC).
Results
Between September 7, 2009 and July 16, 2012, a total of
1,013 women met inclusion criteria for this study, 481
(48%) were consented and 310 (64%) were randomized
(fig.1). A total of 171 women were not randomized due to
cesarean birth before reaching 8 to 10cm cervical dilation
or precipitous delivery. One hundred fifty-four women were
allocated to standard epidural infusion of 0.125% bupivacaine with 2 g/ml fentanyl and 156 women were allocated
to infusion of fentanyl-only (10 g/ml). As shown in table2,
there were no statistically significant differences in maternal
demographics or antepartum complications between the two
study groups. The time from randomization to delivery was
similar as well, 169min in the bupivacaine/fentanyl arm and
149min in the fentanyl-only arm (P = 0.29).
Bupivacaine
(N = 154)
Age (yr)
9 (6)
16
7 (5)
35
Mean SD
22.65.4
Race/ethnicity
Black
14 (9)
Hispanic
133 (86)
White
6 (4)
Other
1 (1)
BMI
Mean SD
31.67.2
Hypertension
47 (31)
Postterm
9 (6)
Ruptured
51 (33)
membranes,
not in labor
Diabetes
9 (6)
Randomization to delivery (min)
Median
169 (93, 243)
(interquartile
range)
Mean SD
198155
No Bupivacaine
(N = 156)
P Value
7 (4)
8 (5)
22.95.0
0.59
0.81
0.64
0.40
16 (10)
133 (85)
3 (2)
4 (3)
32.16.0
36 (23)
11 (7)
46 (29)
0.51
0.13
0.67
0.49
11 (7)
0.67
149 (100, 221)
0.29
180143
0.27
All data are shown as frequency (percent) unless otherwise annotated in

the table.
BMI = body mass index.
Duration of the Second Stage of Labor

The median duration of the second stage was 75min (41,
128) in the bupivacaine/fentanyl group versus 73min (42,
120) in the fentanyl-only group (P = 0.17) with a median
difference of 6.0min (95% CI, 6.0 to 18.0) (fig.2).
Motor Blockade
Motor strength scores were not different between the two
study groups with a narrow range suggesting that neither
group had a dense motor block. Figure3 illustrates the distribution of scores over the measured time points.
Mode of Delivery
There was no difference in the rates of spontaneous delivery
(P = 0.09), forceps-assisted delivery (P = 0.17), or cesarean
delivery (P = 0.38) between the two study groups (table3).
A subset analysis did reveal a possible association between
a low motor strength score and forceps delivery (P = 0.02).
This seems plausible because in patients with a dense motor
block, the second stage was artificially shortened by instrumental vaginal delivery; however, being a secondary analysis,
this finding is hypothesis generating and requires further
investigation.
Relief of Labor Pain
Visual analog scores were similar between the two groups
with scores ranging from 0 to 7 with a median score between
PAIN MEDICINE
Table 3. Mode of Delivery in Women Randomized to

Bupivacaine vs. No Bupivacaine during the Second Stage of
Labor
Mode of Delivery
Bupivacaine
No
Bupivacaine
Spontaneous
Forceps-assisted
Cesarean
112 (73)
19 (12)
23 (15)
126 (81)
12 (8)
18 (12)
P Value
0.09
0.17
0.38
All data are shown as frequency (percent).
Fig. 2. Duration of the second stage of labor in respective

groups.
1 and 2 throughout the second stage. Figure4 illustrates the

distribution of scores over measured time points.
There was a small but significant difference in the number
of intravenous meperidine boluses given during the second
stage. Specifically, the bupivacaine/fentanyl group received
an average of 1.2 boluses versus 1.9 in the fentanyl-only group
(P = 0.004). However, the amount of meperidine did not differ significantly between the two groups, 0.321.20mg and
0.351.12mg, respectively (P = 0.45). Notably, the total
amount of opioid (in morphine equivalents) administered in
the second stage was statistically significant, 7.626.10mg
in the bupivacaine/fentanyl group and 35.9525.22mg in
the fentanyl-only group (P 0.001).
Maternal and Neonatal Well-being
Similarly, the rates of chorioamnionitis (defined as maternal fever 100.4F [38C] in labor) did not differ between
the study groups (P = 0.48). As shown in table4, there
were no differences in neonatal outcomes with regards to
1- and 5-min Apgar scores (P = 0.55 and 0.57, respectively),
umbilical artery pH (P = 0.12), administration of naloxone
(P = 0.31), admission to intensive care unit (P = 0.66), or
breastfeeding at discharge (P = 0.65).
Fig. 3. Distribution of motor strength scores over measured

time points in respective groups.
Discussion
Historically, when high concentrations of local anesthetic
were used in labor epidural infusions, motor blockade was
significant during the second stage of labor and often resulted
in epidural infusions being turned down or completely off
to improve maternal expulsive efforts.11 We highlight three
clinical observations from our study of paramount interest
and importance that address the aforementioned practices.
First, lower-dose epidural bupivacaine infusions such as
that used in this trial moderately impeded lower extremity
motor function during childbirth. In fact, all women were
able to lift their legs against gravity (motor strength score: 3
or higher), further indication that no woman had an extensive motor block. This counters a previously described theory of epidural local anesthetics resulting in poor maternal
expulsive efforts, thought to be a cause of increased operative
deliveries.12
Second, several investigators have reported that a major
disadvantage to lowering or discontinuing the epidural infusion during the second stage of labor is the potential for
inadequate analgesia.1316 In two separate studies, Chestnut
et al.13,14 randomized women to either bupivacaine-containing solutions or placebo infusion of saline during the second stage. Women randomized to the saline groups reported
higher pain scores. Alternatively, Lindow et al.16 infused
epidural opioid in lieu of saline in comparison with epidural bupivacaine/fentanyl. Again, results were similar with
women reporting higher pain scores and a greater need for
Fig. 4. Distribution of visual analog scores over measured

time points in respective groups.
Epidural Analgesia Effects on Second-stage Labor
Table 4. Indices of Maternal and Neonatal Well-being in

Women Randomized to Bupivacaine vs. No Bupivacaine during
the Second Stage of Labor
Outcome
Chorioamnionitis
Breastfeeding at
discharge
Apgar 3 at 1 min
Apgar 3 at 5 min
Umbilical artery
pH 7.10
Naloxone
administered
Admission to
intensive care unit
Bupivacaine
(N = 154)
No
Bupivacaine
(N = 156)
P Value
44 (29)
113 (73)
39 (25)
118 (76)
0.48
0.65
11 (7)
1 (1)
2 (1)
14 (9)
2 (1)
7 (5)
0.55
0.57
0.12
1 (1)
0.31
2 (1)
3 (2)
0.66
All data are shown as frequency (percent).
rescue analgesia of nitrous oxide and/or perineal infiltration of lidocaine in the fentanyl-only group. Of note, these
studies had sample sizes of less than 100 women.
A meta-analysis was performed analyzing the possible
consequences of discontinuing epidural analgesia late in
labor.17 The only significant finding was an increase in pain.
In contrast, our study found that pain scores were equally
satisfactory among both groups. This is likely due to the
higher concentration of fentanyl (10 g/ml) used in our
study versus saline and a fentanyl concentration of 1.6 g/ml
used in the study by Lindow et al.16
Last, substituting epidural bupivacaine with fentanyl-only
(10 g/ml) during the second stage of labor was not associated with deleterious maternal or neonatal effects. Although
there was a negligible amount of intravenous meperidine
given in both study arms, epidural fentanyl administration was five times greater in the fentanyl-only group. This
increase in fetal opioid exposure is because epidural fentanyl
is rapidly absorbed systemically. Although the bioavailability is unknown in laboring women, transplacental transfer is
approximately 90%.18
We report three limitations to our study. First, to estimate
sample size, from our database, we included women of parity
0 and 1, which would explain the shorter median duration of
the second stage at 28min (15, 58) than what was observed
in this study. Our study consisted of all nulliparous women.
Second, having a third arm, with subjects that had a placebo
infusion started after randomization, would have allowed us
to evaluate the effects of discontinuing the epidural entirely
on the outcomes of interest. But such a design would have
raised ethical concerns regarding the withholding of effective
analgesia. Last, parental opioids are known to induce neurobehavioral depression. The literature regarding the effects
of epidural opioids remains unclear.19 However, we did not
assess the newborn beyond the routine 1- and 5-min Apgar
scores, so, there exists the possibility that this exposure could
have resulted in neurobehavioral changes, even subtle ones.
In summary, we found that the use of epidural bupivacaine/fentanyl versus fentanyl-only neither did lengthen
the second stage of labor nor did affect the degree of motor
blockade, mode of delivery, satisfaction with pain analgesia, and maternal or neonatal outcomes. However, the use
of epidural fentanyl-only resulted in a fivefold increase in
fetal opioid exposure. Although not assessed, this exposure
could result in neonatal neurobehavioral depression, both
short and long term, and should be an outcome of interest in future studies. Therefore, anesthetic management
should be tailored to the individual needs of the obstetrical
patient, balancing the risks and benefits to the mother and
her newborn.
Acknowledgments
Dr. Grants contribution to this work was conducted with
support from the Center for Translational Medicine, The
University of Texas Southwestern Medical Center, Dallas,
Texas, and the National Institutes of Health/National Center
for Advancing Translational Sciences, Bethesda, Maryland
(grant no. KL2TR000453). The University of Texas Southwestern IRB number and approval date: 052009-049 and
November 12, 2009.
Competing Interests
Correspondence
Address correspondence to Dr. Grant: Department of Anesthesiology and Pain Management, The University of Texas
Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas 75390. erica.grant@utsouthwestern.edu.
This article may be accessed for personal use at no charge
through the Journal Web site, www.anesthesiology.org.
References
1. Osterman MJ, Martin JA: Epidural and spinal anesthesia use
during labor: 27-state reporting area, 2008. Natl Vital Stat Rep
2011; 59:113, 16
2. Gambling DR, Sharma SK, Ramin SM, Lucas MJ, Leveno KJ,
Wiley J, Sidawi JE: A randomized study of combined spinal-epidural analgesia versus intravenous meperidine during labor: Impact on cesarean delivery rate. Anesthesiology
1998; 89:133644
3. Ramin SM, Gambling DR, Lucas MJ, Sharma SK, Sidawi JE,
Leveno KJ: Randomized trial of epidural versus intravenous
analgesia during labor. Obstet Gynecol 1995; 86:7839
4. Alexander JM, Sharma SK, McIntire DD, Leveno KJ: Epidural
analgesia lengthens the Friedman active phase of labor.
Obstet Gynecol 2002; 100:4650
5. Alexander JM, Sharma SK, McIntire DD, Wiley J, Leveno KJ:
Intensity of labor pain and cesarean delivery. Anesth Analg
2001; 92:15248
6. Sharma SK, Alexander JM, Messick G, Bloom SL, McIntire
DD, Wiley J, Leveno KJ: Cesarean delivery: A randomized trial
of epidural analgesia versus intravenous meperidine analgesia during labor in nulliparous women. Anesthesiology 2002;
96:54651
7. Sharma SK, Sidawi JE, Ramin SM, Lucas MJ, Leveno KJ,
Cunningham FG: Cesarean delivery: A randomized trial of
PAIN MEDICINE
epidural versus patient-controlled meperidine analgesia during labor. Anesthesiology 1997; 87:48794
8. Sharma SK, McIntire DD, Wiley J, Leveno KJ: Labor analgesia and
cesarean delivery: An individual patient meta-analysis of nulliparous women. Anesthesiology 2004; 100:1428; discussion 6A
9. Cunningham F, Leveno K, Bloom S, Hauth J, Rouse D, Spong
C: Williams Obstetrics, 23rd edition. New York, McGraw-Hill,
2010, pp 374587
10. Breen TW, Shapiro T, Glass B, Foster-Payne D, Oriol NE:
Epidural anesthesia for labor in an ambulatory patient.
11. Comparative Obstetric Mobile Epidural Trial (COMET) Study
Group UK: Effect of low-dose mobile versus traditional epidural techniques on mode of delivery: A randomised controlled trial. Lancet 2001; 358:1923
12. Anim-Somuah M, Smyth R, Jones L: Epidural versus non-epidural or no analgesia in labour (Review). Cochrane Database
Syst Rev 2011; CD000331
13. Chestnut DH, Vandewalker GE, Owen CL, Bates JN, Choi
WW: The influence of continuous epidural bupivacaine analgesia on the second stage of labor and method of delivery in
nulliparous women. Anesthesiology 1987; 66:77480
14. Chestnut DH, Laszewski LJ, Pollack KL, Bates JN, Manago
NK, Choi WW: Continuous epidural infusion of 0.0625%
bupivacaine-0.0002% fentanyl during the second stage of
labor. Anesthesiology 1990; 72:6138
15. Abenhaim HA, Fraser WD: Impact of pain level on secondstage delivery outcomes among women with epidural analgesia: Results from the PEOPLE study. Am J Obstet Gynecol
2008; 199:500.e16
16. Lindow SW, Dhillon AR, Husaini SW, Russell IF: A randomised
double-blind comparison of epidural fentanyl versus fentanyl
and bupivacaine for pain relief in the second stage of labour.
BJOG 2004; 111:107580
17. Torvaldsen S, Roberts CL, Bell JC, Raynes-Greenow CH:
Discontinuation of epidural analgesia late in labour for reducing the adverse delivery outcomes associated with epidural
analgesia. Cochrane Database Syst Rev 2004; CD004457
18. Moiss EC, de Barros Duarte L, de Carvalho Cavalli R,
Lanchote VL, Duarte G, da Cunha SP: Pharmacokinetics and
transplacental distribution of fentanyl in epidural anesthesia for normal pregnant women. Eur J Clin Pharmacol 2005;
61:51722
19. Lieberman E, Odonoghue C: Unintended effects of epidural
analgesia during labor: A systematic review. Am J Obstet
Gynecol 2002; 186(5 suppl Nature):S3168
Brain Serotonin Content Regulates the Manifestation of

Tramadol-induced Seizures in Rats
Disparity between Tramadol-induced Seizure and
Serotonin Syndrome
Yohei Fujimoto, M.D., Tomoharu Funao, M.D., Ph.D., Koichi Suehiro, M.D., Ph.D.,
Ryota Takahashi, M.D., Ph.D., Takashi Mori, M.D., Ph.D., Kiyonobu Nishikawa, M.D., Ph.D.
ABSTRACT
Background: Tramadol-induced seizures might be pathologically associated with serotonin syndrome. Here, the authors
investigated the relationship between serotonin and the seizure-inducing potential of tramadol.
Methods: Two groups of rats received pretreatment to modulate brain levels of serotonin and one group was treated as a sham
control (n=6 per group). Serotonin modulation groups received either para-chlorophenylalanine or benserazide + 5-hydroxytryptophan. Serotonin, dopamine, and histamine levels in the posterior hypothalamus were then measured by microdialysis,
while simultaneously infusing tramadol until seizure onset. In another experiment, seizure threshold with tramadol was investigated in rats intracerebroventricularly administered with either a serotonin receptor antagonist (methysergide) or saline (n=6).
Results: Pretreatment significantly affected seizure threshold and serotonin fluctuations. The threshold was lowered in parachlorophenylalanine group and raised in benserazide + 5-hydroxytryptophan group (The mean SEM amount of tramadol
needed to induce seizures; sham: 43.14.2mg/kg, para-chlorophenylalanine: 23.22.8mg/kg, benserazide + 5-hydroxytryptophan: 59.416.5mg/kg). Levels of serotonin at baseline, and their augmentation with tramadol infusion, were less in the
para-chlorophenylalanine group and greater in the benserazide + 5-hydroxytryptophan group. Furthermore, seizure thresholds
were negatively correlated with serotonin levels (correlation coefficient; 0.71, P < 0.01), while intracerebroventricular methysergide lowered the seizure threshold (P < 0.05 vs. saline).
Conclusions: The authors determined that serotonin-reduced rats were predisposed to tramadol-induced seizures, and
that serotonin concentrations were negatively associated with seizure thresholds. Moreover, serotonin receptor antagonism precipitated seizure manifestation, indicating that tramadol-induced seizures are distinct from serotonin syndrome.
RAMADOL, one of the most frequently prescribed

analgesics worldwide, has a multitude of pharmacological properties, acting as an agonist at opioid receptors,
while inhibiting reuptake of serotonin (5-hydroxytryptamine
[5-HT]) and noradrenaline.14 However, along with its therapeutic actions, tramadol has been identified as a major cause
of drug-induced seizures; about 8% of new cases over the
last decade were due to tramadol exposure.5 Although recent
data are not available, a retrospective study revealed that
602 patients out of 557,004 cases reported to the California
Poison Control System during a 2.5-yr period had suffered
from tramadol-induced seizures.6 In correspondence with the
increasing sale of tramadol, the number of patients experiencing tramadol-induced seizures will likely rise.
Elucidating the pathogenesis of tramadol-induced seizures will help reduce the occurrence of such events. However, multiple factors underlie the potentiation of seizures,
and several of tramadols pharmacological features could
individually influence seizure potency.7,8 In general, alteration of brain monoamines and activation of opioid receptors

Tramadol, a commonly prescribed analgesic, is a significant
cause of medication-induced seizures. The mechanism by
which seizure threshold is reduced by tramadol is not known.
In addition to -receptor agonism, tramadol inhibits uptake of
serotonin. Excessive serotonin levels lead to the serotonin syndrome that is characterized by seizures. However, the role of
serotonergic neurotransmission in tramadol-induced seizures is
not known.
The effect of augmenting and antagonizing serotonergic neurotransmission on tramadol-induced seizures was evaluated.

Tramadol-induced seizure thresholds were reduced by serotonin depletion and increased by serotonin augmentation. Serotonin antagonists also reduced seizure threshold.
The results suggest that tramadol-induced seizures are not
related to serotonin uptake inhibition and that these seizures
are distinct from the serotonin syndrome.
are known to modulate the development of epileptic episodes.9,10 Further, Spiller et al.11 suggested that inhibiting the
Submitted for publication March 3, 2014. Accepted for publication August 12, 2014. From the Department of Anesthesiology, Osaka City
University Graduate School of Medicine, Osaka, Japan.
January 2015
PAIN MEDICINE
reuptake of monoamines could be what is responsible for

tramadol intoxication rather than its agonistic actions on the
opioid receptor, since administration of an opioid receptor
antagonist seems to aggravate seizures.11,12
5-Hydroxytryptamine is generally known to protect
against seizures,13 and selective serotonin reuptake inhibitors
(SSRIs) are presumed to be preventative in various animal
models of seizure.1417 To our knowledge, there is no direct
evidence determining the role of 5-HT in the generation of
tramadol-induced seizures; specifically, whether this therapeutic agent potentiates or protects against seizure activity.
Serotonin syndrome (SS), caused by excessive amounts of
5-HT, is another adverse side effect of tramadol, with neuromuscular hyperactivity i.e., clonus, myoclonus, and even
seizures regarded as the major symptoms characterizing the
disorder.18,19 Furthermore, concomitant use of tramadol with
drugs that affect 5-HT, such as SSRIs and tricyclic antidepressants (TCAs), is associated with a higher risk of developing
seizure and/or SS.18,2024 These observations have led some
to believe that tramadol-induced seizures may be one of the
symptoms of SS. Thus, clarifying the role of 5-HT could aid
in differentiating between tramadol-induced seizures and SS.
Interestingly, some antihistamines have been shown to
relieve symptoms of SS.25 This is surprising since antihistaminergic activity has generally been suggested to potentiate
seizures5,26; however, antihistamines and histamine release
inhibitors have been reported to attenuate the potential of
tramadol to induce seizures.27 Although no direct evidence
has been proposed regarding histamine neuromodulation
with tramadol, there may be connections between tramadolinduced seizure and histamine.
The aim of the current study was to investigate the role of
5-HT and histamine in the generation of tramadol-induced
seizures. To accomplish this, we used intracerebral in vivo
microdialysis to characterize the variation of brain 5-HT and
histamine concentrations before, during, and after tramadolinduced seizures while simultaneously evaluating hemodynamics during all experiments. We also focused on the effect
of 5-HT receptor antagonism on seizure threshold.

Animals
Upon approval from the Institutional Animal Care and Use
Committee of our institution (Osaka City University Graduate School of Medicine, Osaka, Japan), 7- to 9-week-old male
SpragueDawley rats weighing 240 to 300g (Clea Inc., Osaka,
Japan) were acquired for this study. Rats had free access to
food and water and were housed in plastic cages under a 12h
lightdark cycle. After probe implantation and intraarterial
and venous cannulation, rats were housed individually in a
cage. Experiments were carried out during the light cycle and
after experiment completion, rats were euthanized with an
intravenous overdose of pentobarbital (50mg/kg).
Surgical Preparation
Before surgery, rats were administered an intramuscular
injection of cefazolin 25mg/kg into their triceps muscle.
Rats were then anesthetized with 3 to 5% sevoflurane
(Abbvie Japan, Tokyo, Japan) in oxygen via a nose cone.
For microdialysis, rats were implanted with an intracerebral
guide cannula (7mm in length and 0.5mm in OD, AG-7;
Eicom, Kyoto, Japan) 7 days before the experiment by using
a stereotaxic frame (Narishige, Tokyo, Japan). Guide cannulas were fixed with dental cement and aimed at the posterior hypothalamus according to the following coordinates:
1.0mm lateral and 4.0mm posterior to bregma, and 7.0mm
ventral to dura.28 Stainless-steel stylets were inserted into
the guide cannulas to prevent obstruction. Rats were singlehoused postoperatively, and allowed one week to recover.
Rats were then anesthetized 2 days before microdialysis, and
an intravenous catheter was inserted into their right jugular
vein, while an arterial catheter was inserted into their right
carotid artery. Catheters were filled with heparinized saline
and routed subcutaneously to exit at the back of the neck.
Seven days before the start of the antagonist study, rats
were implanted with guide cannulas (4mm in length and
0.5mm in OD, AG-4; Eicom) aimed at their lateral ventricle according to the coordinates: 2.0mm lateral and 1.0mm
posterior to bregma, and 3.5mm ventral to dura.28 Intravenous catheters were inserted into their right jugular vein 2
days before the experiment.
Drugs
The following drugs were used: Tramadol HCl (4mg kg1 min1
intravenous injection; Sigma Aldrich Japan, Tokyo, Japan),
4-chloro-dl-phenylalanine methyl ester HCl (para-chlorophenylalanine [PCPA]; 160mg/kg, intraperitoneal injection;
Sigma Aldrich Japan), benserazide HCl (30mg/kg intraperitoneal injection; Sigma Aldrich Japan), 5-hydroxytryptophan
(5-HTP; 30mg/kg intraperitoneal injection; Sigma Aldrich
Japan), methysergide maleate (40 g/animal intracerebroventricular injection; Sigma Aldrich Japan). All compounds
were dissolved in 0.9% NaCl. Volumes of intraperitoneally
administered drugs were 1.0ml/kg of body weight except for
5-HTP, which was 5.0ml/kg because of its solubility to 0.9%
NaCl. Volumes of intracerebroventricularly administered
drugs were 10 l/animal. The dosage of tramadol was determined from our preliminary study.
Experiment 1: In Vivo Microdialysis and Hemodynamic
Evaluation during Seizure Generation with Tramadol
Group Allocation and Drug Pretreatment for 5-HT
Modulation. To deplete brain 5-HT, we used PCPA, an
inhibitor of tryptophan hydroxylase, which is the rate-limiting enzyme in 5-HT synthesis.29 Alternatively, we concomitantly administrated 5-HTP along with benserazide,
immediate precursors of 5-HT, and a peripheral aromatic
amino acid decarboxylase inhibitor to increase 5-HT concentration in the central nervous system.30
Anesthesiology 2015; 122:178-89 179 Fujimoto et al.
Influence of Serotonin on Tramadol-induced Seizures
Using the envelope method, rats were randomly allocated

to each experimental group, which consisted of six total subjects. For the PCPA group, rats received PCPA (160mg/kg)
72, 48, and 24h before microdialysis. Dose and timing of
PCPA administration has been previously described by Prinssen et al.31 and was expected to decrease extracellular 5-HT
levels by about 90% from baseline. For the benserazide +
5-HTP group, rats received benserazide (30mg/kg) 120min
before and 5-HTP (30mg/kg) 90min before microdialysis.
Details have been described by Baumann et al.30 who successfully demonstrated that the combined administration of
these two drugs resulted in the elevation of 5-HT without
altering other parameters. Finally, the sham group of rats
received no pharmacological pretreatment. An experimenter
who was not involved in monitoring the seizure experiment
performed the pretreatments.
In Vivo Microdialysis. On the morning of microdialysis testing, rats were brought to the laboratory and a microdialysis probe, which was confirmed to have a recovery rate of
51% before the experiment, was carefully inserted into the
guide cannulas. The probe was connected to extension cannulas and was perfused using a microdialysis pump (ESP-32;
Eicom) with Ringers solution containing 147.0mM NaCl,
4.0mM KCl, and 2.3mM CaCl2 at a constant flow rate of
2 l/min. Each rat was placed in a plastic cage for 90min
before drug intravenous injection so that they could acclimate to the surrounding environment. After constant perfusion for 90min, the perfusate was collected every 5min to
determine basal dopamine, 5-HT, and histamine concentrations. Infusion of drugs via the venous catheter was concurrent with collection of the dialysate, which was diluted and
directly measured using the following procedures.
Analytical Procedure. Dopamine and 5-HT concentrations
were measured in dialysates using high-performance liquid
chromatography (Eicompak PP-ODS II, 30mm 4.6mm
I.D.; Eicom) with electrochemical detection (HTEC-500;
Eicom), as previously described by Ohmura et al.32 The
mobile phase, which consisted of 2.1mM sodium 1-decansulfonate, 0.1mM EDTA-2Na/0.1M phosphate buffer
(pH 6.0), and 2% (v/v) methanol, was pumped at a rate
of 250 l/min. Concentrations of 5-HT and dopamine in
the dialysate were measured using a pure graphite working electrode (WE-PG; Eicom) and a salt bridge AgAgCl
reference electrode. The working potential was set at 450
mV, while the signal from the current-potential converter
(the integrator output) was filtered with a low-pass in-line
noise filter and integrated by a computerized data acquisition system using chromatography data software (PowerChrom; AD-Instruments Pty Ltd., Castle Hill, New South
Wales, Australia).
Histamine was analyzed by the high-performance liquid chromatography method coupled with a post-column
ortho-phthalaldehyde derivatization of the target amines.33
The high-performance liquid chromatography system consisted of an EP-700 pump system, a M505 auto-injector,
an EHA-500 reaction pump system (Eicom), and an RF10AXL fluorescence detector (Shimadzu, Kyoto, Japan). The
mobile phase consisted of methanol-0.1M NaH2PO4 (1:9,
v/v) containing 170mg/l sodium 1-octanesulfonate (flow
rate, 0.5ml/min). Histamine was separated on an analytical reverse-phase column (Eicompak SC-5ODS; 150mm
3.0mm I.D.; Eicom), and the eluent line was connected by a
T piece with a reagent line, which mixes 0.05% ortho-phthalaldehyde containing 2.5% 2-mercaptoethanol solution and
0.5M K2CO3 in a short reaction coil (9 m 0.5mm I.D.)
at a flow rate of 0.2ml/min (pump 1, 0.1ml/min; pump
2, 0.1ml/min). The eluent and basic ortho-phthalaldehyde
solution were mixed in a reaction coil (3 m 0.5mm I.D.),
while the separation and reaction procedures were performed
at 40.0 0.1C in a CTO-10A column oven (Shimadzu).
The fluorescence intensity was subsequently measured at
450nm with excitation at 340nm in a fluorescence detector.
The chromatography data were analyzed in the same way as
the 5-HT and dopamine data.
Drug Administration and Evaluation of Seizure Activity.
Seizures induced with tramadol infusion were monitored by
two researchers who were blinded to the group allocation
at the same time as microdialysis perfusion. Seizure activity was defined as the fifth stage of the modified Racine
score (tonicclonic seizures),34 since tramadol-induced seizures typically manifest as generalized tonicclonic.3537
We avoided evaluating electroencephalograms because this
methodology may not accurately reflect seizure severity of
tramadol-induced seizures.38 Drug infusion was discontinued with seizure onset, and the latency to seizure was
recorded to determine the amount of tramadol needed to
induce the epileptic activity.
Blood Pressure and Heart Rate Determination. An extension of tubing from the arterial catheter was attached to a
pressure transducer to record blood pressure on a polygraph
(UB-104U; Unique Medical Co., Ltd., Tokyo, Japan) when
rats were placed in plastic cages. Along with microdialysis
and seizure activity experiments, we measured arterial blood
pressure and heart rate (HR), which was calculated from
the arterial pressure waveform using an on-line computer
system (Unique Acquisition; Unique Medical Co., Ltd.).
Mean arterial blood pressure and HR were obtained before
the infusion (baseline value), and again 20 s before and 60
s after seizure manifestation (preconvulsion value/postconvulsion value, respectively). To validate the volume effect of
infusion, 0.9% NaCl was infused at the same rate to the
six rats making up the control group, which did not receive
microdialysis. Since rats in the control group did not exhibit
seizures, values at 9 and 11min after infusion were substituted for the pre- and postconvulsion values, respectively.
Experiment 2: Antagonist Study
To demonstrate the direct involvement of 5-HT in seizure generation, 12 rats were randomly allocated to two
groups; one group of rats (n=6) was given 0.9% NaCl
PAIN MEDICINE
intracerebroventricularly (saline group), while the other

group (n=6) was given 40 g of methysergide maleate, a
nonselective 5-HT 1/2 receptor antagonist, intracerebroventricularly in a volume of 10 l39 (methysergide group). On
the day of the experiment, rats were brought to the laboratory and an intracerebroventricular port (AMI-4; Eicom)
was attached to the guide cannula. Two hours after the procedure, test drugs and tramadol were administered through
the intracerebroventricular port and the intravenous catheter, respectively. As in Experiment 1, 4mg kg1 min1 of
tramadol was infused until rats exhibited generalized tonic
clonic seizures. The amount of tramadol needed to induce
these seizures was then compared between saline and methysergide groups.
Data Calculations and Statistical Analysis for
Both Experiments
The number of rats in each group was based on the results of
our preliminary experiment (n=6 for each group); the difference of the mean amount of tramadol needed to induce
seizures was about 20mg/kg and the expected SD was 10,
with an of 0.05 and a of 0.80.
Amounts of 5-HT, dopamine, and histamine were calculated by comparing the peak heights of samples and standards.
The minimal dose of tramadol (mg/kg) needed to induce a
seizure was calculated from the latency and was used as an
index of seizure threshold. Comparisons of these data between
groups were performed with an unpaired t test and one-way
ANOVA followed by the HolmSidak post hoc test. Extracellular 5-HT, dopamine, and histamine concentrations are
expressed as pg/l of sample. These data were analyzed by a
two-way ANOVA with time as a repeated measure and with
group as the independent factor. When there were significant
interactions between these two variables, a simple effects test
was performed (respective F values along with degrees of freedom are indicated). Comparisons between the variables at
each time point within the same group were made by oneway repeated-measures ANOVA, and comparisons between
pretreatment groups at the respective time points of measurement were made using one-way ANOVA. When there was a
significant difference, we used the HolmSidak test as a post
hoc analysis. We also utilized Spearman product-moment correlations. All analyses were performed using Sigma Plot ver.
11 (Systat Software Inc., San Jose, CA), and statistical significance for all analysis was set at P value less than 0.05. Data are
presented as mean and SEM.
Results
Seizure Activity and Tramadol Threshold
Every rat manifested tonicclonic seizures on infusion of tramadol and no infusion resulted in lethality. Furthermore, seizure threshold was significantly affected by pretreatment with
PCPA or benserazide + 5-HTP. The mean SEM latency to
seizure was 647.263.3 s in the Sham group, 347.717.4 s
in the PCPA group, and 891.7101.3 s in the benserazide +
5-HTP group. Figure1 shows that PCPA pretreatment significantly reduced the threshold of tramadol needed to induce
seizures, while pretreatment with benserazide + 5-HTP
increased the tramadol threshold for seizure induction.
Extracellular Monoamine Concentrations in
the Posterior Hypothalamus
Continuous infusion of tramadol resulted in marked increases
in both 5-HT and dopamine concentrations (figs.2 and 3).
A two-way ANOVA demonstrated significant intergroup
differences of these variables (5-HT: F[2,120]=51.9,
P <0.001, dopamine: F[2,120]=9.3, P <0.01), as well
as differences between time points (5-HT: F[8, 120]=11.2,
P<0.001, dopamine: F[8,120]=24.2, P<0.001). Since
significant interactions were detected between time and
group factors (5-HT: F[16,120]=6.2, P <0.01, dopamine: F[16,120]=2.4, P < 0.01), we performed individual
simple effects. We determined that 5-HT concentrations
were significantly elevated from baseline values during the
infusion of tramadol in all groups, except for values at 0
to 5min in the benserazide + 5-HTP group (Sham group:
F[8,40]=18.3, P < 0.001, PCPA group: F[8,40]=18.3,
P < 0.001, benserazide + 5-HTP group: F[8,40]=7.6,
P<0.001) (fig. 2). Both of the pretreatments significantly
affected 5-HT levels at baseline as well as during tramadol
infusion. A one-way ANOVA with a post hoc HolmSidak
test revealed significant intergroup differences on respective
time-points from baseline to 10 to 15min, the period at
which tramadol was infused (P < 0.05, respectively, F values
are indicated in fig. 2). Regarding extracellular dopamine, a
one-way ANOVA revealed that concentrations were comparable among groups at baseline and were similarly elevated
during the period which tramadol infusion had been conducted (0 to 15min). Within group difference at different
time periods (Sham group: F[8,40]=12.0, P<0.001, PCPA
group: F[8,40]=12.8, P < 0.001, benserazide + 5-HTP
group: F[8,40]=7.0, P < 0.001) (intergroup differences
with F values at each time-point; described in fig. 3) Baseline
extracellular histamine concentrations were similar among
groups and were not affected by tramadol infusions (fig.4).
Individual 5-HT concentrations at the time of seizure manifestation and the amount of tramadol used to induce seizures
in each subject are plotted in figure5, with a significant positive correlation between factors (correlation coefficient 0.71,
P < 0.01, Spearman correlation).
Hemodynamics during the Experiment
Hemodynamic changes during the experiment are presented
in table1. Similar hemodynamic alterations were observed in
each group to which tramadol was administered. With tramadol infusion, HR gradually decreased to 60 to 80% of the
baseline value of each rat at the time before the seizure, and
recovered to almost 100% after the seizure. Mean blood pressure values were similar between groups before and after seizures. At the onset of seizure, almost all rats seemed to express
Fig. 1. Total amount of tramadol needed to induce seizures in each group. Each group consisted of six rats. Data are shown as
means and bars indicate SEM. *P < 0.05, ** P < 0.01 compared with the value of the Sham group. A one-way ANOVA was used to
detect statistical significance between groups and a HolmSidak post hoc test was performed. 5-HTP = 5-hydroxytryptophan;
PCPA = para-chlorophenylalanine.
marked hypertension and tachycardia; however, we could not

obtain proper arterial waveforms because they were markedly
affected by seizure activity. Finally, the control group (infused
with 0.9% NaCl) did not exhibit any change in mean blood
pressure or HR during the infusion period.
Effects of Serotonin Receptor Antagonism on Tramadolinduced Seizures
An intracerebroventricular administration of methysergide
markedly potentiated seizure propagation compared to saline
administration in rats (P < 0.05, Unpaired t test) (fig.6). The
mean SEM amount of tramadol used to induce seizures
was 80.85.4mg/kg in saline groups, while that used in the
methysergide group was 60.14.2mg/kg.
Discussion
This study showed the protective role of 5-HT on tramadolinduced seizures. We have revealed a correlation between
brain 5-HT concentration and the threshold of tramadol
needed to induce a seizure. We also showed that extracellular
5-HT augmented by tramadol actually protected against seizure. Additionally, the administration of a serotonin receptor antagonist, methysergide, potentiated seizure generation.
These findings strongly suggest that tramadol-induced
seizures are a distinctly different concept from SS. Interestingly, tramadol did not influence extracellular concentrations of histamine; consequently, histamine is not presumed
to play a role in tramadol-induced epileptic activity.
Our study is unique in that we only used tramadol as a
drug to induce seizure activity. Furthermore, its continuous
infusion enabled proper time-course validation of the monoamine contribution to seizure development. Most of the previously reported experiments on tramadol-related seizures often
used various epileptic models that could have affected the
potency of tramadol to induce seizures, such as pentylenetetrazole-27,40,41 and maximal electroshock-induced seizures7,8,41;
therefore, these previous studies may not be appropriate in
qualifying the seizure potential of tramadol. Indeed, Bankstahl
et al.41 reported a discrepancy between seizure models in that
tramadol reduced the threshold of pentylenetetrazole to induce
seizures but increased that of maximal electroshock-induced
seizures. Generally, 5-HT exerts a preventative role in seizure
onset under various experimental conditions.1315,17 Moreover, administration of SSRIs, including fluoxetine, is related
to seizure attenuation in some animal models of epilepsy.1417
For example, Yan et al.15 found that the anticonvulsant effect
of fluoxetine selectively correlated with an enhanced synaptic
availability of 5-HT. Interestingly, the pharmacological action
of tramadol on 5-HTergic neurotransmission resembles the
PAIN MEDICINE
Fig. 2. Extracellular 5-hydroxytryptamine (5-HT) concentrations in the posterior hypothalamus of each group. Each group consisted of six rats. Data are shown as means and vertical bars indicate SEM. F values with degrees of freedom (numbers in the parenthesis) are indicated in some cases. Horizontal bars indicate the mean duration of tramadol infusion in each group. A two-way
repeated-measures ANOVA demonstrated significant main effects in the difference among pretreatment groups (F[2,120] = 51.9, P
< 0.001) and between time-points (F[8,120] = 11.2, P < 0.001). In addition, a significant interaction between groups and time factors was been detected (F[16,120] = 6.2, P < 0.001). Significant simple effects were detected regarding both group and time factors. Expresses P < 0.05 compared with the baseline values of each group. (One-way repeated-measures ANOVA followed by
a HolmSidak post hoc analysis.) The table under the figure indicates the intergroup differences at each time-point. Expresses
P < 0.01 difference between groups at each time-point detected by a one-way ANOVA, and F values with degrees of freedom are also
indicated below. Post hoc analyses were demonstrated using the HolmSidak test, which indicated some significant intergroup differences between Sham and para-chlorophenylalanine (PCPA) groups, Sham and benserazide + 5-hydroxytryptophan (5-HTP) groups,
and PCPA and benserazide + 5-HTP groups, respectively. *Significant difference in mean 5-HT concentrations between the two groups.
actions of SSRIs.4 Furthermore, lower baseline 5-HT levels

are correlated with subsequent increases in generalized seizure
duration and frequency.42 Our primary finding was consistent
with these previous reports and we assume that 5-HT was
responsible for the seizure-resilient properties that we observed
in this study.
In contrast to our findings and the discussed literature,
Raffa et al.10 reported that the seizure potency of tramadol
was not affected by reserpine pretreatment, which nonspecifically depletes 5-HT. Along with its deleterious effects
on 5-HT, reserpine induces depletion of various brain catecholamines that themselves may affect seizure potency,
and induce hypotension,43 which may have a negative effect
on seizures. To omit the possible influence of our various
pretreatment regimens on hemodynamics, we measured
hemodynamic changes during the experiment and found
that alterations in hemodynamic values were similar among
groups. Mean blood pressure was unaffected by the infusion

of tramadol, and the reduction in HR that we observed during tramadol administration was observed in each group.
More specifically, HR gradually dropped to about 70 to 80%
of its baseline value before seizure onset, however, this was
not a consequence of volume load because the same infusion rate of 0.9% NaCl did not result in any hemodynamic
change. At seizure onset, on the other hand, tachycardia was
observed in all of the tramadol-infused rats. Tachycardia is
regarded as one of the risk factors associated with tramadolinduced seizures.6 However, as the authors of the previous
report mentioned, it was not clear whether tachycardia was
present before the seizures they induced or if it occurred
subsequent to seizures. The results of our experiment indicate that the latter incident may be the case. Moreover, this
suggests that seizures induced by tramadol are not a consequence of hemodynamic changes.
Fig. 3. Extracellular dopamine concentrations in the posterior hypothalamus of each group. Each group consisted of six rats.
Data are shown as means and vertical bars indicate SEM. F values with degrees of freedom (numbers in the parenthesis) are
indicated in some cases. Horizontal bars indicate the mean duration of tramadol infusion in each group. A two-way repeatedmeasures ANOVA demonstrated significant main effects in the difference among pretreatment groups (F[2,120] = 9.3, P < 0.01)
and between time-points (F[8,120] = 24.2, P < 0.001). In addition, a significant interaction between groups and time factors was
detected (F[16,120] = 2.4, P < 0.01). Significant simple effects were detected regarding both group and time factors. Expresses
P < 0.05 compared with the baseline values of each group. (One-way repeated-measures ANOVA followed by a HolmSidak
post hoc analysis.) The table under the figure indicates the intergroup differences at each time-point. ,Express P < 0.05 and
P < 0.01 differences between groups at each time-point detected by a one-way ANOVA, and F values with degrees of freedom
are also indicated below. Post hoc analyses were demonstrated with the HolmSidak test, which indicated some significant
intergroup differences between Sham and para-chlorophenylalanine (PCPA) groups, Sham and benserazide + 5-hydroxytryptophan (5-HTP) groups, and PCPA and benserazide + 5-HTP groups, respectively. *Significant difference in mean dopamine
concentrations between two groups.
In the clinical setting, it is sometimes difficult to distinguish tramadol-induced seizures from SS, since neuromuscular symptoms of SS include clonus, myoclonus, and even
seizures.18,19 We confirmed that 5-HT elevation resulted in
an anticonvulsive effect and that 5-HT depleted rats were
more susceptible to seizures. Moreover, a nonselective 5-HT
antagonist methysergide, which may be a potent SS reliever,44
promoted seizure generation. These findings indicate that
tramadol-induced seizures are distinctly different from SS,
and that neuromuscular symptoms of SS are qualitatively
different from tramadol-induced seizures. This is clinically
of importance that, when we face seizure-like symptoms
with tramadol user, therapeutics for one syndrome probably
exacerbate the other.
SS tends to arise in patients concomitantly taking TCAs
or SSRIs with tramadol compared to those with tramadol
monotherapy.24,45,46 There can be no doubt that TCAs or

SSRIs in addition to tramadol, synergistically augment neuronal 5-HT, which accounts for the occurrence of SS; however, as we have already discussed, the same trend has been
reported with cases of tramadol-induced seizures.2024 This
seems to be contradictory to our current findings on the
anticonvulsive effect of 5-HT, though this discrepancy
could be explained by the influence of drug metabolism
and characteristics of the tramadol user. Drugdrug interactions are speculated to be associated with the risk of developing tramadol-related seizures,47 as is much more evident
in cases of SS.24,45,46 Interestingly, tramadol, TCAs, and
SSRIs are all commonly metabolized by cytochrome P450
2D6 (CYP2D6),48 therefore, there may be some competition for enzymatic reaction when these agents are taken
concomitantly. Furthermore, TCAs and SSRIs are known
PAIN MEDICINE
Fig. 4. Extracellular histamine concentrations in the posterior hypothalamus of each group. Each group consisted of six rats.
Data are shown as means and vertical bars indicate SEM. Horizontal bars indicate the mean duration of tramadol infusion in
each group. A two-way repeated-measures ANOVA did not detect a significant effect of time and intergroup group difference in
histamine concentration. 5-HTP = 5-hydroxytryptophan; PCPA = para-chlorophenylalanine.
to inhibit the activity of CYP2D6.49 These interactions may

result in the insufficient metabolism of these drugs, yielding extremely high levels of tramadol, overwhelming the
anticonvulsive property of 5-HT, and thereby provoking
seizures. This is consistent with observations that the occurrence of tramadol-induced seizures is dose-independent and
that the dose that precipitates seizures differs greatly among
individuals.6,20,36,37 Gardner et al.22 suggested that a small
subset of the population is sensitive to seizures induced by
tramadol because of a genetic polymorphism of CYP2D6,
resulting in individuals who are poor metabolizers of tramadol. Aside from this small population, it remains that patients
who are prescribed tramadol along with antidepressants are
at higher risk of developing tramadol-related seizures. Moreover, it has been revealed that tramadol users are more likely
to have been exposed to drugs that interact with tramadol,
such as CYP2D6 substrates and inhibitors.22 Furthermore,
patients with major depression have been proven to be sixfold more likely to have an unprovoked seizure than the
general population.50 Interestingly, most documented cases
of tramadol overdose are in relation to suicide attempts or
patients with mental disorders, a population of individuals

that may already be susceptible to tramadol intoxication.6,12
Eventually, our results indicated that SSRI can theoretically
relieve tramadol induced seizure by 5-HT augmentation,
however, its administration seems to affect drug metabolism
and sometimes militated against seizure threshold. In addition, it should be noticed that a concomitant administration
of SSRI with tramadol will probably cause SS, so that we
do not recommend SSRI as a reliever of tramadol-induced
seizure.
Based on the observation that antihistamines such as
cyproheptadine can relieve symptoms of SS,25 and that histamine plays a possible role in tramadol-induced seizure,27
we measured both histamine as well as 5-HT in the current study. We selected the posterior hypothalamus as the
region of monoamine measurement since this area is rich in
histaminergic neurons,51,52 and is thus an appropriate area
to evaluate the possible effect of tramadol on histamine.53
Furthermore, this was also a reasonable area for confirming
5-HT modulations using pretreatments, because 5-HTergic
innervations from the raphe nucleus distribute over the entire
Fig. 5. Correlation between seizure threshold and 5-hydroxytryptamine (5-HT) concentration at the time of seizure manifestation.
Seizure threshold was defined as the minimum amount of tramadol needed to induce seizure (mg/kg). An overall n = 18 was
used in this analysis. Spearman correlation revealed a significant correlation between these factors (correlation coefficient: 0.71,
P<0.01). 5-HTP = 5-hydroxytryptophan; PCPA = para-chlorophenylalanine.
Table 1. Hemodynamics before and after Seizures in Each Group
Baseline Value
Preconvulsion Value
Postconvulsion Value
Mean Blood Pressure/HR
Tramadol infusion
Sham group
PCPA group
Benserazide + 5-HTP group
125 (5)/397.5 (23.1)

122.2 (2.4)/362 (15.8)
113.8 (1.9)/381.3 (9.7)
Control group (0.9% NaCl infusion)
120.7 (1.3)/377 (6.5)
117.5 (7.2)/287 (5.9)

123 (4.8)/267.2 (11.6)
120.8 (2.5)/310.5 (11.4)
115.2 (1.7)/390 (8.5)
(9min after infusion)
126 (5.1)/387.2 (18.4)

133.8 (4.8)/343 (17.2)
124.2 (3.4)/364.7 (20)
116.3 (1.6)/392 (10.9)
(11min after infusion)
Each group consisted of six rats. Data are shown as mean (SEM). Unit for blood pressure was expressed as mmHg and HR as beats/minutes. A two-way
repeated-measures ANOVA was used. There was no intergroup difference in either blood pressure or HR between tramadol-infused groups.
5-HTP = 5-hydroxytryptophan; HR = heart rate; PCPA = para-chlorophenylalanine.
forebrain (including the hypothalamus).5456 To date, there

has been no reported pharmacological effect of tramadol
on brain histamine, although Rehni et al.27 showed that the
administration of antihistamines can attenuate the pentylenetetrazole-induced seizure potentiating effect of tramadol.
More specifically, this group suggested that tramadol exerted
a seizurogenic effect on pentylenetetrazole-treated mice, possibly through an opioid receptor-dependent release of histamine. However, this finding was contrary to the widely
accepted observation that antihistaminergic agents produce
seizure activity in human and animals.5,26 Moreover, the histamine-releasing potency of tramadol has proven to be much
smaller than that of morphine57; hence, it is not likely that

opioid-induced histamine release would have a major effect
on seizure development with tramadol. In the current study,
we demonstrated that hypothalamic histamine concentrations remained unchanged throughout the measurement
period. Furthermore, it should be noted that we avoided
using antihistamines since they have been shown to potentiate seizures5,26 and could likely enhance the seizure potency
of tramadol.22,24 Our findings indicate that tramadol has
no pharmacological action on extracellular histamine and,
despite the previous report discussed above,27 brain histamine
does not play a major role in tramadol-induced seizures.
PAIN MEDICINE
Fig. 6. Effect of intracerebroventricular administration of either methysergide or saline on seizure threshold in normal rats. Both
groups consisted of six rats. Data are shown as means and bars indicate SEM. *P < 0.05 compared with saline group values. An
unpaired t test was used to detect statistical significance between groups.
Finally, we measured dopamine concentration for fear

that it might be affected by pretreatment.58 Dopamine may
influence seizure potency through its activation of various
dopamine receptors.9,59 Tramadol inhibits dopamine uptake,
however, our understanding of how dopamine contributes to
the pharmacological effect of tramadol remains limited.2,60
In the current study, dopamine was similarly elevated
among all three groups during tramadol infusion, suggesting that dopamine was not responsible for the difference in
the seizure threshold observed between groups. However, it
remains possible that elevations of neuronal dopamine can
affect the seizure-inducing potential of tramadol, and further study involving specific dopamine modulation like this
experiment may help in our understanding of this issue.
In conclusion, to our knowledge, this is the first study to
provide evidence of the relationship between extracellular
5-HT and the seizure-inducing potential of tramadol. We
found that depletion of 5-HT resulted in a lowered threshold for seizure induction, whereas elevated 5-HT raised this
threshold. Furthermore, serotonergic receptor antagonism
lowered the threshold. Accordingly, SS, which is caused by
excessive 5-HT, is undoubtedly different from tramadolinduced seizures. Thus, in the clinical setting, we should be
careful when a tramadol user manifests seizure-like symptoms,
since the therapeutics for SS will aggravate tramadol-induced
seizures. We also showed that tramadol does not elevate brain
histamine content, and thus, histamine is likely not involved
in the development of tramadol-induced seizures.
Acknowledgments
This study was supported by a Grant-in-Aid for scientific research in Japan ( Japan Society for the Promotion of Science,
Tokyo, Japan; grant no. J122640011).
Competing Interests
Correspondence
Address correspondence to Dr. Funao: Department of Anesthesiology, Osaka City University Graduate School of
Medicine, 1-5-7, Asahimachi, Abenoku, Osaka 545-8586,
Japan. funaot@iris.eonet.ne.jp. Information on purchasing
reprints may be found at www.anesthesiology.org or on
the masthead page at the beginning of this issue. Anesthesiologys articles are made freely accessible to all readers,
for personal use only, 6 months from the cover date of the
issue.
References
1. Bamigbade TA, Davidson C, Langford RM, Stamford JA:
Actions of tramadol, its enantiomers and principal metabolite, O-desmethyltramadol, on serotonin (5-HT) efflux and
uptake in the rat dorsal raphe nucleus. Br J Anaesth 1997;
79:3526
2. Driessen B, Reimann W, Giertz H: Effects of the central analgesic tramadol on the uptake and release of noradrenaline
and dopamine in vitro. Br J Pharmacol 1993; 108:80611
3. Sevcik J, Nieber K, Driessen B, Illes P: Effects of the
central analgesic tramadol and its main metabolite,
O-desmethyltramadol, on rat locus coeruleus neurones. Br J
Pharmacol 1993; 110:16976
4. Bloms-Funke P, Dremencov E, Cremers TI, Tzschentke TM:

Tramadol increases extracellular levels of serotonin and noradrenaline as measured by in vivo microdialysis in the ventral hippocampus of freely-moving rats. Neurosci Lett 2011;
490:1915
5. Thundiyil JG, Kearney TE, Olson KR: Evolving epidemiology
of drug-induced seizures reported to a Poison Control Center
System. J Med Toxicol 2007; 3:159
6. Marquardt KA, Alsop JA, Albertson TE: Tramadol exposures reported to statewide poison control system. Ann
Pharmacother 2005; 39:103944
7. Manocha A, Sharma KK, Mediratta PK: Tramadol, a centrally
acting opioid: Anticonvulsant effect against maximal electroshock seizure in mice. Indian J Physiol Pharmacol 1998;
42:40711
8. Manocha A, Sharma KK, Mediratta PK: On the mechanism
of anticonvulsant effect of tramadol in mice. Pharmacol
Biochem Behav 2005; 82:7481
9. Freitas RM, Vasconcelos SM, Sousa FC, Viana GS, Fonteles
MM: Pharmacological studies of the opioids, mood stabilizer
and dopaminergic drugs on pilocarpine-induced seizures
and status epilepticus. Neurosci Lett 2006; 408:848
10. Raffa RB, Stone DJ Jr: Unexceptional seizure potential of tramadol or its enantiomers or metabolites in mice. J Pharmacol
Exp Ther 2008; 325:5006
11. Spiller HA, Gorman SE, Villalobos D, Benson BE, Ruskosky
DR, Stancavage MM, Anderson DL: Prospective multicenter
evaluation of tramadol exposure. J Toxicol Clin Toxicol 1997;
35:3614
12. Shadnia S, Soltaninejad K, Heydari K, Sasanian G, Abdollahi
M: Tramadol intoxication: A review of 114 cases. Hum Exp
Toxicol 2008; 27:2015
13. Trindade-Filho EM, de Castro-Neto EF, de A Carvalho R,
Lima E, Scorza FA, Amado D, Naffah-Mazzacoratti Mda G,
Cavalheiro EA: Serotonin depletion effects on the pilocarpine
model of epilepsy. Epilepsy Res 2008; 82:1949
14. Yan QS, Jobe PC, Dailey JW: Evidence that a serotonergic
mechanism is involved in the anticonvulsant effect of fluoxetine in genetically epilepsy-prone rats. Eur J Pharmacol
1994; 252:10512
15. Yan QS, Jobe PC, Cheong JH, Ko KH, Dailey JW: Role of serotonin in the anticonvulsant effect of fluoxetine in genetically
epilepsy-prone rats. Naunyn Schmiedebergs Arch Pharmacol
1994; 350:14952
16. Hernandez EJ, Williams PA, Dudek FE: Effects of fluoxetine
and TFMPP on spontaneous seizures in rats with pilocarpine-induced epilepsy. Epilepsia 2002; 43:133745
17. Browning RA, Wood AV, Merrill MA, Dailey JW, Jobe PC:

Enhancement of the anticonvulsant effect of fluoxetine following blockade of 5-HT1A receptors. Eur J Pharmacol 1997;
336:16
18. Iqbal MM, Basil MJ, Kaplan J, Iqbal MT: Overview of serotonin syndrome. Ann Clin Psychiatry 2012; 24:3108
19. Dunkley EJ, Isbister GK, Sibbritt D, Dawson AH, Whyte IM:
The hunter serotonin toxicity criteria: Simple and accurate
diagnostic decision rules for serotonin toxicity. QJM 2003;
96:63542
20. Kahn LH, Alderfer RJ, Graham DJ: Seizures reported with
tramadol. JAMA 1997; 278:1661
21. Jick H, Derby LE, Vasilakis C, Fife D: The risk of seizures
associated with tramadol. Pharmacotherapy 1998; 18:60711
22. Gardner JS, Blough D, Drinkard CR, Shatin D, Anderson G,
Graham D, Alderfer R: Tramadol and seizures: A surveillance
study in a managed care population. Pharmacotherapy 2000;
20:142331
23. Ripple MG, Pestaner JP, Levine BS, Smialek JE: Lethal combination of tramadol and multiple drugs affecting serotonin.
Am J Forensic Med Pathol 2000; 21:3704
24. Sansone RA, Sansone LA: Tramadol: Seizures, serotonin

syndrome, and coadministered antidepressants. Psychiatry
(Edgmont) 2009; 6:1721
25. Simons FE: Advances in H1-antihistamines. N Engl J Med
2004; 351:220317
26. Yamada K, Takizawa F, Tamura T, Kanda T: The effect of
antihistamines on seizures induced by increasing-current
electroshocks: Ketotifen, but not olopatadine, promotes the
seizures in infant rats. Biol Pharm Bull 2012; 35:6937
27. Rehni AK, Singh TG, Singh N, Arora S: Tramadol-induced
seizurogenic effect: A possible role of opioid-dependent histamine H1 receptor activation-linked mechanism. Naunyn
Schmiedebergs Arch Pharmacol 2010; 381:119
28. Paxinos G, Watson C: The Rat Brain in Stereotaxic

Coordinates, Compact 6th edition. San Diego, Academic
Press, 2009, pp 40-2 and 65-7
29. Koe BK, Weissman A: p-Chlorophenylalanine: A specific
depletor of brain serotonin. J Pharmacol Exp Ther 1966;
154:499516
30. Baumann MH, Williams Z, Zolkowska D, Rothman RB:

Serotonin (5-HT) precursor loading with 5-hydroxy-l-tryptophan (5-HTP) reduces locomotor activation produced by
(+)-amphetamine in the rat. Drug Alcohol Depend 2011;
114:14752
31. Prinssen EP, Assi MB, Koek W, Kleven MS: Depletion of
5-HT disrupts prepulse inhibition in rats: Dependence on
the magnitude of depletion, and reversal by a 5-HT precursor. Neuropsychopharmacology 2002; 26:3407
32. Ohmura Y, Izumi T, Yamaguchi T, Tsutsui-Kimura I, Yoshida T,
Yoshioka M: The serotonergic projection from the median raphe
nucleus to the ventral hippocampus is involved in the retrieval
of fear memory through the corticotropin-releasing factor type
2 receptor. Neuropsychopharmacology 2010; 35:12718
33. Miyamoto Y, Yoshimoto R, Yumoto M, Ishihara A, Takahashi
K, Kotani H, Kanatani A, Tokita S: Simultaneous fluorometric
measurement of histamine and tele-methylhistamine levels
in rodent brain by high-performance liquid chromatography.
Anal Biochem 2004; 334:8996
34. Racine RJ: Modification of seizure activity by electrical
stimulation. II. Motor seizure. Electroencephalogr Clin
Neurophysiol 1972; 32:28194
35. Jovanovi-Cupi V, Martinovi Z, Nesi N: Seizures associated
with intoxication and abuse of tramadol. Clin Toxicol (Phila)
2006; 44:1436
36. Talaie H, Panahandeh R, Fayaznouri M, Asadi Z, Abdollahi
M: Dose-independent occurrence of seizure with tramadol. J
Med Toxicol 2009; 5:637
37. Boostani R, Derakhshan S: Tramadol induced seizure: A

3-year study. Caspian J Intern Med 2012; 3:4847
38. Jang HS, Jang IS, Lee MG: The effects of tramadol on electroencephalographic spectral parameters and analgesia in rats.
Korean J Physiol Pharmacol 2010; 14:1918
39. Yu XM, Hua M, Mense S: The effects of intracerebroventricular injection of naloxone, phentolamine and methysergide on
the transmission of nociceptive signals in rat dorsal horn neurons with convergent cutaneous-deep input. Neuroscience
1991; 44:71523
40. Lesani A, Javadi-Paydar M, Khodadad TK, Asghari-Roodsari
A, Shirkhodaei M, Norouzi A, Dehpour AR: Involvement
of the nitric oxide pathway in the anticonvulsant effect of
tramadol on pentylenetetrazole-induced seizures in mice.
Epilepsy Behav 2010; 19:2905
41. Bankstahl M, Bankstahl JP, Bloms-Funke P, Lscher W:

Striking differences in proconvulsant-induced alterations of
seizure threshold in two rat models. Neurotoxicology 2012;
33:12737
42. Shouse MN, Staba RJ, Ko PY, Saquib SF, Farber PR:
Monoamines and seizures: Microdialysis findings in locus
PAIN MEDICINE
ceruleus and amygdala before and during amygdala kindling. Brain Res 2001; 892:17692
43. Shore P, Giachetti A: Reserpine: Basic and clinical pharmacology, Handbook of Psychopharmacology. Edited by Iversen L,
Iversen S, Snyder S. New York, Springer, 1978, pp 197219
44. Yamada J, Sugimoto Y, Horisaka K: The evidence for the
involvement of the 5-HT1A receptor in 5-HT syndrome
induced in mice by tryptamine. Jpn J Pharmacol 1989;
51:4214
45. Mahlberg R, Kunz D, Sasse J, Kirchheiner J: Serotonin syndrome with tramadol and citalopram. Am J Psychiatry 2004;
161:1129
46. Fox MA, Jensen CL, Murphy DL: Tramadol and another

atypical opioid meperidine have exaggerated serotonin syndrome behavioural effects, but decreased analgesic effects,
in genetically deficient serotonin transporter (SERT) mice.
Int J Neuropsychopharmacol 2009; 12:105565
47. Reus VI, Rawitscher L: Possible interaction of tramadol and
antidepressants. Am J Psychiatry 2000; 157:839
48. Subrahmanyam V, Renwick AB, Walters DG, Young PJ, Price
RJ, Tonelli AP, Lake BG: Identification of cytochrome P-450
isoforms responsible for cis-tramadol metabolism in human
liver microsomes. Drug Metab Dispos 2001; 29:114655
49. Ereshefsky L, Riesenman C, Lam YW: Antidepressant drug
interactions and the cytochrome P450 system. The role of
cytochrome P450 2D6. Clin Pharmacokinet 1995; 29(suppl
1):108; discussion 189
50. Hesdorffer DC, Hauser WA, Annegers JF, Cascino G: Major
depression is a risk factor for seizures in older adults. Ann
Neurol 2000; 47:2469
51. Haas HL, Sergeeva OA, Selbach O: Histamine in the nervous
system. Physiol Rev 2008; 88:1183241
52. Panula P, Yang HY, Costa E: Histamine-containing neurons

in the rat hypothalamus. Proc Natl Acad Sci U S A 1984;
81:25726
53. Yoshitake T, Yamaguchi M, Nohta H, Ichinose F, Yoshida H,
Yoshitake S, Fuxe K, Kehr J: Determination of histamine in
microdialysis samples from rat brain by microbore column
liquid chromatography following intramolecular excimerforming derivatization with pyrene-labeling reagent. J
Neurosci Methods 2003; 127:117
54. Saavedra JM, Palkovits M, Brownstein MJ, Axelrod J:

Serotonin distribution in the nuclei of the rat hypothalamus
and preoptic region. Brain Res 1974; 77:15765
55. Jacobs BL, Azmitia EC: Structure and function of the brain
serotonin system. Physiol Rev 1992; 72:165229
56. Rubenstein JL: Development of serotonergic neurons and

their projections. Biol Psychiatry 1998; 44:14550
57. Baldo BA, Pham NH: Histamine-releasing and allergenic
properties of opioid analgesic drugs: Resolving the two.
Anaesth Intensive Care 2012; 40:21635
58. OLeary OF, Bechtholt AJ, Crowley JJ, Hill TE, Page ME, Lucki
I: Depletion of serotonin and catecholamines block the acute
behavioral response to different classes of antidepressant
drugs in the mouse tail suspension test. Psychopharmacology
(Berl) 2007; 192:35771
59. Gangarossa G, Di Benedetto M, OSullivan GJ, Dunleavy

M, Alcacer C, Bonito-Oliva A, Henshall DC, Waddington JL,
Valjent E, Fisone G: Convulsant doses of a dopamine D1
receptor agonist result in Erk-dependent increases in Zif268
and Arc/Arg3.1 expression in mouse dentate gyrus. PLoS
One 2011; 6:e19415
60. Sprague JE, Leifheit M, Selken J, Milks MM, Kinder DH,
Nichols DE: In vivo microdialysis and conditioned place preference studies in rats are consistent with abuse potential of
tramadol. Synapse 2002; 43:11821
EDUCATION
Charles D. Collard, M.D., Editor
Alan Jay Schwartz, M.D., M.S. Ed., Associate Editor
Images in Anesthesiology: Inversion of the Right

Hemidiaphragm due to Massive Hemothorax after
Central Line Placement
Allan F. Simpao, M.D., Jorge A. Galvez, M.D., Alan Jay Schwartz, M.D., M.S.Ed.,
Mohamed A. Rehman, M.D.
9-month-old, 6.6-kg, male child

with biliary atresia presented for an
orthotopic liver transplant. Anesthetic induction, endotracheal intubation, and arterial
catheter insertion proceeded uneventfully. Two
attempts were required to place a right internal
jugular catheter by using ultrasound guidance;
although the first attempt appeared technically
correct, that attempt was aborted due to scant
blood return from the finder needle.
During surgery, the patient experienced a 4 to 5cm H2O increase in peak inspiratory pressures as well as sustained hypotension and anemia despite administration of blood products and vasoactive medications. Blood loss in the operative field appeared
limited and inconsistent with the observed hypotension. The figure and video (see Supplemental Digital Content 1, http://links.
lww.com/ALN/B17) illustrate an inverted diaphragm with the native liver retracted (fig. A, arrow); the diaphragm bulged in an
abnormal caudal direction during inspiration.1 An inverted diaphragm is abnormal, as shown by the drawing (fig. B) indicating normal and abnormal positions during respiration, with abnormal being the case when a supradiaphragmatic mass effect is
present. The surgeons evacuated 280ml of blood from the thorax and repaired the right subclavian artery after the transplant
procedure. The patient recovered fully and without long-term sequelae noted at a 3-month follow-up visit.
Inversion of a hemidiaphragm should raise suspicion of a supradiaphragmatic mass effect, such as pleural effusion, tension
pneumothorax, or massive hemothorax.2 In this case, vascular injury occurred likely during central catheter insertion. Unexplained hypotension with no apparent blood loss in the face of the abnormal bulge of the diaphragm suggests hematoma as
the cause of the supradiaphragmatic mass effect. Early (occurring within 24h after insertion) central catheter complications
include pneumothorax, hemothorax, cardiac tamponade, catheter leak, dysrhythmia, hydrothorax, and death.3
Competing Interests
Correspondence
Address correspondence to Dr. Simpao: simpaoa@email.chop.edu
References
1. Kolar P, Neuwirth J, Sanda J, Suchanek V, Svata Z, Volejnik J, Pivec M: Analysis of diaphragm movement during tidal breathing and
during its activation while breath holding using MRI synchronized with spirometry. Physiol Res 2009; 58:38392
2. Lowe SH, Cosgrove DO, Joseph AE: Inversion of the right hemidiaphragm shown on ultrasound examination. Br J Radiol 1981; 54:7547
3. Lovell M, Baines D: Fatal complication from central venous cannulation in a paediatric liver transplant patient. Paediatr Anaesth
2000; 10:6614
anesthesiology.org).
From the Perelman School of Medicine at the University of Pennsylvania and the Childrens Hospital of Philadelphia, Philadelphia,
Pennsylvania.
Copyright 2013, the American Society of Anesthesiologists, Inc. Lippincott Williams & Wilkins. Anesthesiology 2015; 122:190
January 2015
Clinical Concepts and Commentary
Jerrold H. Levy, M.D., F.A.H.A., F.C.C.M., Editor
Pretransfusion Testing and Transfusion of

Uncrossmatched Erythrocytes
Michael L. Boisen, M.D., Ryan A. Collins, M.D., Mark H. Yazer, M.D., Jonathan H. Waters, M.D.
LOOD transfusion is the most frequent procedure

performed during hospital admissions1 and many
transfusions are administered in the perioperative period,
often on a time-sensitive basis. In this clinical commentary,
key points related to pretransfusion testing are reviewed with
an emphasis on the electronic crossmatch, as well as the use
of uncrossmatched erythrocytes in situations where crossmatch-compatible units are not yet available for transfusion.
Preoperative Blood Orders

For the typical surgical patient, a determination is made preoperatively as to the level of pretransfusion testing ordered
(none, type and screen only, or number of units to crossmatch). Such determinations may be based on individual
physician judgment of the expected/typical amount of surgical blood loss for a given procedure, the patients preoperative
hemoglobin concentration, or on an institutions maximum
surgical blood order schedule (MSBOS). The MSBOS lists
the recommended extent of pretransfusion testing for common surgical procedures and is intended to optimize the
amount of pretransfusion testing performed on each patient
thereby reducing costs and unnecessary testing.2
Limitations of the traditional MSBOS include that it
may not be based on local data and may be updated infrequently. More recently, advances in medical informatics have
allowed the MSBOS to be updated based on institution- and
procedure-specific median transfusion rates, an approach
with the potential to significantly reduce unnecessary testing and crossmatching.3 The accuracy of the MSBOS in
predicting transfusion requirements may also be improved
by accounting for patient-specific variables such as preoperative hematocrit and the lowest tolerable hematocrit.4 An
advantage of using a data-driven MSBOS is that once the

mechanism to collect data is in place, data-driven revisions
can easily be made on an annual basis or whenever there is a
significant change in practice (such as using tranexamic acid
in joint replacement surgeries). In any case, the MSBOS represents a guideline to direct pretransfusion testing; the number of units ordered for any given patient may need to be
modified in the light of the patients preoperative condition,
coagulation status, and clinically significant antibodies that
may make finding compatible erythrocytes difficult. In other
words, the recommendations made in the MSBOS should
be interpreted for each individual patient and their underlying disease and antibody status.
Pretransfusion Testing
Pretransfusion testing is a multistep process aimed at avoiding potentially fatal hemolytic transfusion reactions. The
process begins on the clinical ward with identification of the
intended recipient and collection of a properly labeled blood
sample. When the sample and requisition are received in
the transfusion laboratory, blood bank personnel review the
recipients transfusion history, perform the necessary testing,
and if ordered, crossmatch erythrocytes.
Historical Review
The patients electronic blood bank record is reviewed for
previous ABO and RhD type results and the presence of
anti-erythrocyte antibodies. Previous ABO and RhD type
results are important because discrepancies between previous and current results can signal a wrong-blood-in-tube
miscollection and prompt recollection of a patient sample.
Historical antibodies are important because, although their
This article is featured in This Month in Anesthesiology, page 1A.

Submitted for publication March 17, 2014. Accepted for publication July 8, 2014. From the Department of Anesthesiology (M.L.B.) and
Department of Pathology (R.A.C.), University of Pittsburgh, Pittsburgh, Pennsylvania; Department of Pathology, University of Pittsburgh and
Institute for Transfusion Medicine, Pittsburgh, Pennsylvania (M.H.Y.); and Departments of Anesthesiology and Bioengineering and McGowan
Institute for Regenerative Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania ( J.H.W.).
January 2015
Pretransfusion Testing and Uncrossmatched Erythrocytes
titers may have decreased to below the threshold of detectability in the current specimen, they may result in a clinically
significant hemolytic reaction if the recipient is transfused
with erythrocytes bearing the corresponding antigen (an
anamnestic immune response). Thus, the historical review
is a first pass look at the patients transfusion history and
provides some information on the likelihood that the patient
will present with clinically significant antibodies.
Blood Type (or Group) Determination
Determination of the recipients ABO type is performed using
both forward and reverse testing phases; these two phases of
testing produce complementary information that serves to
confirm each others result. Forward typing is performed by
mixing the recipients erythrocytes with commercially available anti-A and anti-B sera and observing for agglutination
(clumping together of cells indicating antibody has bound to
its target on the erythrocytes). The reverse type is performed
using the recipients serum and commercially available group
A and B erythrocytes. Agglutination patterns of the forward
and reverse types and the compatible erythrocytes for transfusion are listed in table1.
Typing for the RhD antigen is performed in a similar
manner as the forward type, with commercially available
anti-D sera reacting with RhD antigen expressed on recipient erythrocytes. Unlike the ABO blood group, antibodies directed against antigens in the Rh blood group do not
occur naturally and are only made in response to a sensitizing exposure such as previous transfusion or pregnancy. Such
antibodies are detected by the antibody screen.5
Antibody Screen
The antibody screen is an antibody detection test in which
the recipients serum is added to a reference panel of commercially available erythrocytes with a known pattern of
antigen expression that, between them, include all clinically
significant non-ABO antigens known to cause clinically significant hemolysis.
A positive antibody screen signifies the presence of at
least one antibody directed against red cell surface antigens.
Development of red cell antibodies, known as alloimmunization, occurs as a result of exposure to erythrocyte antigens during pregnancy or a previous transfusion.6 When an
antibody is detected in the antibody screen, the blood bank

must perform additional testing to identify the specificity
of the antibody. If the antibody is clinically significant, that
is, if it can cause the premature destruction of transfused
erythrocytes, antigen-negative erythrocyte units must be
located. This search for compatible (i.e., antigen-negative)
erythrocytes can take several hours or even longer (e.g., days
and weeks) depending on the number and nature of the antibodies and can result in significant surgical delays (median
delay of 12h in one report).5 To reduce surgical delays and
cancellations related to unexpected antibodies, pretransfusion testing can be completed up to 30 to 45 days in advance
provided the patient has not been pregnant or transfused
in the preceding 90 days. If a patient has been pregnant or
transfused in the preceding 90 days, a type and screen is valid
for up to 72h.5 The patient must also commit to maintaining their identification from the blood bank (usually a bracelet applied at the time of sample collection) that links the
patient to their blood bank testing.5
Serologic Crossmatch
Serologic crossmatching involves physically mixing donor
erythrocytes with recipients plasma and it can be performed
in two ways. An immediate spin crossmatch was the traditional method of confirming ABO compatibility between
a potential donor unit and the recipients plasma. A more
extensive crossmatch involving antihuman globulin (sometimes referred to as Coombs reagent) is used to ensure
compatibility between antigen-negative erythrocytes and the
serum of a recipient with a current or historical non-ABO
antibody; incompatibility in either of these crossmatches
is indicated by the presence of erythrocyte agglutination
or hemolysis. Serologic crossmatching adds approximately
10min to pretransfusion testing using the immediate spin
method, and approximately 45min to perform an antihuman globulin crossmatch (table2).
Computer or Electronic Crossmatch
With the improved sensitivity of modern antibody screening methods for detecting nearly all clinically significant
erythrocyte antibodies, it was recognized that a patient who
has a negative antibody screen and no historical antibodies
can safely be issued any ABO and RhD typespecific unit
Table 1. Forward and Reverse Type Agglutination Patterns of Recipient Blood Types in the ABO Antigen System
Forward
Recipient Type
A
B
AB
O
Reverse
Anti-A
Anti-B
A cells
B cells
Compatible Donors
+
+
A, O
B, O
AB, A, B, O
O
+ indicates the presence of agglutination (a positive finding indicating an antibodyantigen interaction), indicates no agglutination.
Group O is the universal erythrocyte donor, whereas group AB recipients can receive erythrocytes from any donor blood type.
Anesthesiology 2015; 122:191-5 192 Boisen et al.
EDUCATION
without performing an immediate spin crossmatch. Starting in the 1980s, blood banks began selectively replacing
conventional immediate spin serologic crossmatching with
computerized systems involving bar codes and laser wands to
identify and issue ABO-compatible units.7 The blood banks
computer has logic that recognizes when an incompatible
unit has been selected for transfusion and will not permit
that unit to be issued. To be eligible for erythrocyte issuing using the electronic crossmatch, the recipient must have
had their ABO and RhD group determined twice, a negative
antibody screen, and no prior record of non-ABO antierythrocyte antibodies.7 As non-ABO antierythrocyte antibodies
are found in only a small percentage of all recipients (i.e., in
approximately 5% of hospitalized surgical patients who do
not have sickle-cell disease), most patients qualify for this
crossmatch system. Electronic crossmatch technology allows
compatible units to be issued in less than 5min because
physically mixing the recipients plasma with the donors
erythrocytes is not required. If any of the above-mentioned
requirements are not met, the electronic crossmatch cannot
be used and serologic crossmatching is necessary.8
The main advantage of the electronic crossmatch is that
erythrocytes can be issued in mere minutes. This can lead
to reductions in the costs associated with laboratory testing,
in the number of units ordered but not transfused, and in
improved blood inventory management.
A logical extension of electronic crossmatch systems,
remote electronic blood issue systems have been implemented with substantial improvements in both blood
availability and system efficiency.9 In effect, these systems
are vending machines containing erythrocytes that can be
accessed if and when a patient requires a transfusion; these
systems incorporate a mechanism to capture essential information such as the member of the healthcare team who
acquires the units, the time that the units were removed,
the number of units and specific units removed as well as
a mechanism for the main blood bank to be electronically
contacted (signaled) when the system is accessed and/or
units are removed. They can be installed in locations that are
remote from the blood bank itself, such as near an operating room or intensive care unit, thereby reducing the time
required to transport the units to the patient. One such
electronically controlled blood refrigerator system installed
in a cardiac surgical operating theater reduced the median

time to delivery of urgently required erythrocyte units from
24min to 59 s and brought about significant reductions in
unnecessary requests for erythrocyte units.10
Emergency-release Uncrossmatched Erythrocytes
Uncrossmatched group O (universal donor) erythrocytes are
issued in emergencies when transfusion is required before
compatibility testing is complete. Because group O erythrocytes do not express A or B antigens, ABO-incompatibility
hemolytic transfusion reactions are avoided. Blood banks
maintain a supply of group O erythrocytes that can be issued
immediately or even stored in a refrigerator at the pointof-care, such as in the emergency room or operating room
environments. The use of uncrossmatched erythrocytes
is indicated when the recipients bleeding and/or anemia
is so severe that even the typically short delay necessary to
complete pretransfusion testing and provide crossmatched
erythrocytes would jeopardize the patients survival. Because
the blood banks inventory of group O erythrocytes (and
perhaps to a greater extent, group AB plasma) is finite, it is
of critical importance to send a bleeding patients sample to
the laboratory as soon as possible so that type-specific crossmatched erythrocytes can be issued.
The main risk associated with transfusion of uncrossmatched erythrocytes (aside from the risks common to all
red cell transfusions) is the risk of a hemolytic transfusion
reaction in a recipient with a preexisting red cell antibody
who is transfused with erythrocytes expressing the corresponding antigen. Because uncrossmatched erythrocytes
are issued before the antibody screen is complete, the
uncrossmatched erythrocytes could be incompatible with a
recipients antibody leading to hemolysis of the donor unit
with an associated risk of organ injury if the antibody fixes
complement, or if the hemolysis is brisk. The prevalence of
these clinically significant red cell antibodies was recently
found to be 1.9% in a tertiary hospital emergency department population, and only 0.5% in patients younger than
30 yr, although the incidence can be higher in other populations.11 However, estimates of the actual risk of hemolysis
after the transfusion of uncrossmatched erythrocytes are
lower still (typically <1%) as summarized in table3. Again,
although acute hemolysis due to ABO incompatibility
Table 2. Time Required for Various Pretransfusion Testing Methods

Test
ABO type
Antibody screen*
Serologic crossmatch (assuming compatible units are in inventory)
Electronic crossmatch
Emergency-release uncrossmatched erythrocytes
Approximate Time Required

10 min
4560min depending on the technique used
Immediate spin: 10 min antihuman globulin: 45 min
5 min
<5 min
* As the ABO type part of a type and screen is usually performed concomitantly with the antibody screen, the time required to complete
a type and screen is usually determined by the speed at which the antibody screen is performed. Note that crossmatched erythrocytes
cannot be issued until both parts of a type and screen are completed.
Pretransfusion Testing and Uncrossmatched Erythrocytes
Table 3. Summary of Uncrossmatched Transfusion Studies
Study
Mulay, 201217
Radkay, 20126
Miraflor, 201115
Goodell, 201018
Ball, 200919
Dutton, 200514
Unkle, 199120
Lefebre, 198721
Schwab, 198622
Gervin, 198423
Blumberg, 197824
Total
Number of Recipients
Number of
Uncrossmatched
Erythrocyte Units
Issued
Rate of
Hemolysis
Rate of New
Antibody Formation
1,407
218
132
262
153
161
135
133
99
160
46
2,906
4,144
1,065
1,570
1,002
511
581
Not reported
537
410
875
221
10,916
1/1,407 (0.02%)
1/218 (0.5%)
1/132 (0.8%)
1/262 (0.4%)
0
0
0
0
0
0
0
4/2,906 (0.1%)
7/232* (3%)
4/218 (1.8%)
1/132
Not reported
Not reported
1/161 (0.6%)
3/135 (2.2%)
Not reported
Not reported
Not reported
Not reported
16/878 (1.8%)
Incidence of hemolysis and alloimmunization after emergency-release uncrossmatched blood transfusion in civilian centers.
* Denominator includes only patients with a subsequent antibody screen available.
will not occur with group O erythrocytes, there is a small

chance that the recipient will have an antibody against
another erythrocyte antigen. Although hemolytic reactions are infrequent and those mediated by the majority of
non-ABO antibodies tend to be less severe than hemolytic
reactions caused by an ABO-incompatible transfusion (or
other antibodies that fix complement), close clinical and
laboratory monitoring for hemolysis and indices of organ
injury is advised for patients who turn out to have received
antigen-incompatible units.
It should be noted that for patients with negative antibody
screens, erythrocytes are not routinely matched for antigens
other than ABO and RhD. Thus, uncrossmatched erythrocytes would not be expected to confer any increased risk of
causing the recipient to become alloimmunized compared
with crossmatched erythrocytes. Studies of uncrossmatched
erythrocyte transfusions have detected new antierythrocyte
antibodies after 1.8 to 3% of uncrossmatched transfusion
episodes (table 3), comparable with observations with crossmatched erythrocytes.12 Confounding of these estimates is
likely due to massively transfused patients often receiving
both uncrossmatched and crossmatched units during their
hospital course.
Group O RhD-negative blood is typically used in
emergencies when the recipients RhD status is unknown.
Because RhD-negative blood is a scarce resource, priority
for its use is given to RhD-negative females with childbearing potential to prevent alloimmunization against D
antigen and subsequent risk of hemolytic disease of the
fetus and newborn. RhD-positive erythrocytes can be
transfused to RhD-negative patients who have not made
anti-D antibodies as a result of prior exposure to RhD
antigen because D-negative individuals do not constitutively produce anti-D antibodies. In experimental studies,
anti-D antibodies are detected in approximately 80% of

healthy RhD-negative volunteers immunized with RhDpositive blood,13 but the alloimmunization rate observed
in clinical studies of hospitalized patients who were transfused at least one unit of RhD-positive blood ranges from
only 10 to 33%.1416
In summary, preoperative blood orders should be based
on an MSBOS that is informed whenever possible by
institution-specific data. The pretransfusion testing process
demands accuracy at each step, and for maximum efficiency
should be completed before the day of surgery. For most
surgeries, a type and screen is sufficient, particularly given
the rapidity with which blood can be issued by blood banks
using the computer crossmatch. When crossmatched blood
is not yet available, the published experience supports the
safety of transfusing uncrossmatched erythrocytes; clinicians should not hesitate to use them when clinically indicated in the management of patients with life-threatening
hemorrhage.
Acknowledgments
Competing Interests
Correspondence
Address correspondence to Dr. Waters: Departments of Anesthesiology and Bioengineering and McGowan Institute for
Regenerative Medicine, University of Pittsburgh, 300 Halket
Street, Pittsburgh, Pennsylvania 15213. watejh@upmc.edu.
anesthesiology.org or on the masthead page at the begin-
EDUCATION
ning of this issue. Anesthesiologys articles are made freely

References
1. Wier LM, Pfuntner A, Maeda J, Stranges E, Ryan K, Jagadish
P, Collins Sharp B, Elixhauser A: HCUP Facts and Figures:
Statistics on Hospital-based Care in the United States, 2009.
Rockville, Agency for Healthcare Research and Quality, 2011
2. Friedman BA, Oberman HA, Chadwick AR, Kingdon KI: The
maximum surgical blood order schedule and surgical blood
use in the United States. Transfusion 1976; 16:3807
3. Frank SM, Rothschild JA, Masear CG, Rivers RJ, Merritt WT,
Savage WJ, Ness PM: Optimizing preoperative blood ordering with data acquired from an anesthesia information management system. Anesthesiology 2013; 118:128697
4. Palmer T, Wahr JA, OReilly M, Greenfield ML: Reducing
unnecessary cross-matching: A patient-specific blood ordering system is more accurate in predicting who will receive a
blood transfusion than the maximum blood ordering system.
5. McWilliams B, Yazer MH, Cramer J, Triulzi DJ, Waters JH:
Incomplete pretransfusion testing leads to surgical delays.
Transfusion 2012; 52:213944; quiz 2145
6. Radkay L, Triulzi DJ, Yazer MH: Low risk of hemolysis after transfusion of uncrossmatched red blood cells.
Immunohematology 2012; 28:3944
7. Butch SH, Judd WJ, Steiner EA, Stoe M, Oberman HA:
Electronic verification of donor-recipient compatibility: The
computer crossmatch. Transfusion 1994; 34:1059
8. Sfwenberg J, Hgman CF, Cassemar B: Computerized delivery controlA useful and safe complement to the type and
screen compatibility testing. Vox Sang 1997; 72:1628
9. Cox C, Enno A, Deveridge S, Seldon M, Richards R, Martens
V, Woodford P: Remote electronic blood release system.
10. Staves J, Davies A, Kay J, Pearson O, Johnson T, Murphy MF:
Electronic remote blood issue: A combination of remote
blood issue with a system for end-to-end electronic control
of transfusion to provide a total solution for a safe and
timely hospital blood transfusion service. Transfusion 2008;
48:41524
11. Saverimuttu J, Greenfield T, Rotenko I, Crozier J, Jalaludin
B, Harvey M: Implications for urgent transfusion of uncrossmatched blood in the emergency department: The prevalence
of clinically significant red cell antibodies within different

patient groups. Emerg Med (Fremantle) 2003; 15:23943
12. Heddle NM, OHoski P, Singer J, McBride JA, Ali MA, Kelton
JG: A prospective study to determine the safety of omitting
the antiglobulin crossmatch from pretransfusion testing. Br J
Haematol 1992; 81:57984
13. Gunson HH, Stratton F, Cooper DG, Rawlinson VI: Primary
immunization of Rh-negative volunteers. Br Med J 1970;
1:5935
14. Dutton RP, Shih D, Edelman BB, Hess J, Scalea TM: Safety
of uncrossmatched type-O red cells for resuscitation from
hemorrhagic shock. J Trauma 2005; 59:14459
15. Miraflor E, Yeung L, Strumwasser A, Liu TH, Victorino GP:
Emergency uncrossmatched transfusion effect on blood type
alloantibodies. J Trauma Acute Care Surg 2012; 72:4852; discussion 523
16. Yazer MH, Triulzi DJ: Detection of anti-D in D- recipients
transfused with D+ red blood cells. Transfusion 2007;
47:2197201
17. Mulay SB, Jaben EA, Johnson P, Badjie K, Stubbs JR: Risks
and adverse outcomes associated with emergency-release
red blood cell transfusion. Transfusion 2013; 53:141620
18. Goodell PP, Uhl L, Mohammed M, Powers AA: Risk of hemolytic transfusion reactions following emergency-release RBC
transfusion. Am J Clin Pathol 2010; 134:2026
19. Ball CG, Salomone JP, Shaz B, Dente CJ, Tallah C, Anderson
K, Rozycki GS, Feliciano DV: Uncrossmatched blood transfusions for trauma patients in the emergency department:
Incidence, outcomes and recommendations. Can J Surg 2011;
54:1115
20. Unkle D, Smejkal R, Snyder R, Lessig M, Ross SE: Blood antibodies and uncrossmatched type O blood. Heart Lung 1991;
20:2846
21. Lefebre J, McLellan BA, Coovadia AS: Seven years experi
ence with group O unmatched packed red blood cells in a
regional trauma unit. Ann Emerg Med 1987; 16:13449
22. Schwab CW, Civil I, Shayne JP: Saline-expanded group O

uncrossmatched packed red blood cells as an initial resuscitation fluid in severe shock. Ann Emerg Med 1986; 15:12827
23. Gervin AS, Fischer RP: Resuscitation of trauma patients

with type-specific uncrossmatched blood. J Trauma 1984;
24:32731
24. Blumberg N, Bove JR: Un-cross-matched blood for emergency transfusion. One years experience in a civilian setting.
JAMA 1978; 240:20579
Review Article
David S. Warner, M.D., Editor
Regulation of Cerebral Autoregulation by Carbon

Dioxide
Lingzhong Meng, M.D., Adrian W. Gelb, M.B.Ch.B.
ABSTRACT
Cerebral autoregulation describes a mechanism that maintains cerebral blood flow stable despite fluctuating perfusion pressure. Multiple nonperfusion pressure processes also regulate cerebral perfusion. These mechanisms are integrated. The effect of
the interplay between carbon dioxide and perfusion pressure on cerebral circulation has not been specifically reviewed. On the
basis of the published data and speculation on the aspects that are without supportive data, the authors offer a conceptualization delineating the regulation of cerebral autoregulation by carbon dioxide. The authors conclude that hypercapnia causes
the plateau to progressively ascend, a rightward shift of the lower limit, and a leftward shift of the upper limit. Conversely,
hypocapnia results in the plateau shifting to lower cerebral blood flows, unremarkable change of the lower limit, and unclear
change of the upper limit. It is emphasized that a sound understanding of both the limitations and the dynamic and integrated
nature of cerebral autoregulation fosters a safer clinical practice. (Anesthesiology 2015; 122:196-205)
erebral blood flow (CBF) is tightly controlled to

meet the disproportionately high metabolic rate of the
brain and to wash out the large amount of metabolic wastes
thus produced. Multiple physiological processes are engaged
in the regulation of CBF.1 Cerebral autoregulation is a mechanism that maintains a stable CBF for a given magnitude of
cerebral metabolic rate in spite of fluctuation of cerebral perfusion pressure (CPP).2 The original conceptualization was
proposed by Lassen3 who, in 1959, insightfully took the data
from 11 groups of subjects reported in 7 different studies and
drew the first plot. Lassens approach was carefully reviewed
and critiqued.4,5 Nonetheless, cerebral autoregulation is regularly referenced in clinical practice to guide arterial blood pressure management in both neurological and non-neurological
patients, with or without increased intracranial pressure.
Cerebral autoregulation is visualized as a correlation plot
of CBF (axis of ordinate) against CPP (axis of abscissas) (fig.1).
The three key elements of the autoregulation curve are (1) the
lower limit, (2) the upper limit, and (3) the plateau. The lower
and upper limits are the two sharp inflection points indicating
the boundary of pressure-independent flow (the plateau) and
the start of pressure-passive flow. The most quoted numbers are the lower limit (CPP)=60 mmHg, the upper limit
(CPP)=150 mmHg, and the plateau (CBF)=50ml/min per
100g.2 It needs to be pointed out that these numbers are the
means of various groups of subjects in the studies, without any
note of the range of distribution or SD.2,3 For an individual

patient, these means may either underestimate or overestimate
the true values. Indeed, it was cautioned by Drummond4 that
there are enormous interindividual and study-to-study variations in the lower limit. The very wide range of the distribution of the lower limit, 40 to 110 mmHg, can be recognized
not only in young and healthy volunteers6 but also in cardiac
patients undergoing cardiopulmonary bypass with -stat acid
base management.7 The position of the lower limit also depends
on the mechanism of hypotension. For example, Fitch etal.8
showed in baboons that during hemorrhagic hypotension,
the lower limit resided at a mean arterial pressure (MAP) that
was 65% of the baseline value, whereas during drug-induced
hypotension (halothane alone, halothane plus trimetaphan, and
halothane plus nitroprusside), the lower limit shifted to a lower
MAP that was 35 to 40% of the baseline value.
The execution of cerebral autoregulation relies on the robust
cerebrovascular reactivity that engenders dilation to a decrease
in CPP and constriction to an increase in CPP (fig. 1). However, cerebrovascular reactivity is not exclusively linked to CPP.
Changes in other physiological processes, notably, carbon dioxide, can also alter cerebral vasomotor tone and thus regulate
CBF. Intuitively, perfusion pressure and nonperfusion pressure
CBF-regulating processes interact and integrate at the point of
cerebrovascular resistance regulation; thus, the effect of the interplay of distinct processes on CBF may differ to a stand-alone
This article is featured in This Month in Anesthesiology, page 1A. The figures were prepared by Annemarie B. Johnson, C.M.I., Medical
Illustrator, Vivo Visuals, Winston-Salem, North Carolina.
Submitted for publication April 4, 2014. Accepted for publication July 22, 2014. From the Department of Anesthesia and Perioperative
Care, University of California San Francisco, San Francisco, California.
January 2015
EDUCATION
Fig. 1. Cerebral autoregulation is visualized as a correlation plot between cerebral blood flow (CBF) and cerebral perfusion pressure
(CPP). CBF remains stable between the lower limit (LL) and the upper limit (UL) (portion B, plateau). CBF is pressure passive at the CPP
range below the lower limit (portion A) and above the upper limit (portion C). This illustration uses a CPP of 60 mmHg as the lower limit,
a CPP of 150 mmHg as the upper limit, and a CBF of 50ml/min per 100g as the plateau. However, these regularly quoted numbers
are not fixed; rather, they vary interindividually and intraindividually depending on a variety of factors. Therefore, we take a note of SD
to emphasize that these parameters have a wide range of distribution. The cerebrovascular reactivity is also illustrated.
process. Therefore, cerebral autoregulation should be regarded

as a mobile instead of a fixed plot when nonperfusion pressure
processes are also involved. Ignorance of this dynamic and integrative nature of CBF regulation can be risky in clinical practice
because the position and shape of the autoregulation curve may
have shifted away from what we normally assume. Although the
above cited numbers may apply in some young, healthy, normotensive, nonanesthetized, and resting subjects and when perfusion pressure is the only process engaged in CBF regulation,
they may not be suitable in other circumstances. We emphasize
that the autoregulation curve conceived by Lassen is not a onesize-fits-all phenomenon; rather, its position and shape may
change following changes in pertinent physical, medical, neurological, or physiological conditions. Therefore, proper arterial
blood pressure management in the effort of CBF optimization
can only be fulfilled when the influences of nonperfusion pressure processes on cerebral autoregulation are appreciated.
Carbon dioxide is a known powerful modulator of cerebral vasomotor tone,911 and change in arterial blood carbon
dioxide partial pressure (Paco2) is frequently encountered in
clinical care.12,13 The influence of carbon dioxide on cerebral
autoregulation has not been specifically reviewed. A sound
understanding of this topic facilitates clinical decision making
and promotes a culture of safe practice. On the basis of the
published large animal and human data and some speculation
on the components that are without direct data support, we
present a detailed discussion of the integrated effect of carbon
dioxide and perfusion pressure on the cerebral circulation.
Effect of Hypercapnia on Cerebral

Autoregulation
Hypercapnia increases CBF by cerebral vasodilation. Thus,
there are two pertinent queries when considering its effect on
Anesthesiology 2015; 122:196-205 197
cerebral autoregulation. The first is how it affects the lower limit.

The combined vasodilatory effects imposed by hypotension and
hypercapnia could shift the lower limit rightward. The other
query is how it affects the upper limit. The dilation induced
by hypercapnia could adversely affect the hypertension-induced
constriction, rendering a leftward shift of the upper limit.
Harper14 in 1966 was among the earliest to show in dogs
that the lower limit was lost and the pressureflow relationship became linear during severe hypercapnia (Paco2=70 to
90 mmHg) compared with normocapnia (Paco2=30 to 40
mmHg). Also in dogs, Hggendal and Johansson15 showed
that the lower limit occurred at an MAP of 50 to 60 mmHg
under normocapnia (Paco2=30 to 50 mmHg); however,
during hypercapnia (Paco2=70 to 95 mmHg), the lower
limit shifted to a much higher MAP of 80 to 100 mmHg.
Later, Raichle and Stone16 demonstrated in monkeys that
the lower limit was shifted upward and rightward and ultimately abolished by acutely increasing Paco2 in a stepwise
manner (Paco2=33, 48, 57, and 70 mmHg, respectively).
Regarding the upper limit, Ekstrm-Jodal etal.17 showed
in dogs that autoregulation was conserved until 225 mmHg
(MAP) during normocapnia; however, during moderate
hypercapnia (Paco2=40 to 60 mmHg), the autoregulatory pressure range was shorter, with an upper limit of 150
mmHg, and during more marked hypercapnia (Paco2 >60
mmHg), the upper limit was as low as 125 mmHg.
In human subjects, McCulloch etal.18 found that the
threshold at which hypercapnia significantly impaired
autoregulation averaged 56 mmHg (Paco2) during sevoflurane anesthesia and 61 mmHg during propofol anesthesia.
In patients undergoing cardiopulmonary bypass, Murkin
etal.19 showed that -stat acidbase management preserved,
whereas pH-stat (carbon dioxide supplementation) impaired
Carbon Dioxide and Cerebral Autoregulation
autoregulation. In ventilated very low-birth-weight infants,

Kaiser etal.20 found that autoregulation was progressively
impaired and cerebral perfusion became progressively pressure
passive with escalating hypercapnia. However, none of these
human studies examined in detail how the plateau, the lower
limit, and the upper limit are affected by hypercapnia.1820
In light of these studies, we propose the following construct as illustrated in figure2 to explain the effect of hypercapnia on cerebral autoregulation. The ensuing discussion in
this section is based on this figure unless specified otherwise.
The plateau shifts upward during hypercapnia due to the
cerebral vasodilation being induced.
When CPP is decreasing, cerebral resistance vessels dilate
until the lower limit is reached. However, with hypercapniainduced dilation, maximal dilation and therefore the lower
limit is reached at a higher CPP during hypercapnia than
normocapnia, that is, the lower limit is shifted rightward. For
example, at point E on the CPP axis at normocapnia, cerebral resistance vessels are not maximally dilated yet because the
CPP is higher than the lower limit; however, for the same CPP
at hypercapnia, cerebral resistance vessels are already maximally
dilated due to the additional dilation imposed by hypercapnia.
We overlap the portions of the autoregulation curve below the
lower limit based on the premise that the calibers of the maximally dilated cerebral resistance vessels at hypercapnia and normocapnia are the same. This premise is plausible if there is truly
a fixed size limit to which cerebral resistance vessels can dilate.
Otherwise, a different construct would ensue.
When CPP is increasing, cerebral resistance vessels constrict

until the upper limit is reached. However, at hypercapnia, the
upper limit is reached at a lower CPP than normocapnia, that
is, the upper limit is shifted leftward due to the antagonism of
the hypertension-induced vasoconstriction by the hypercapniainduced vasodilation. For example, at point F on the CPP axis
at normocapnia, cerebral resistance vessels are not maximally
constricted yet because the CPP is lower than the upper limit;
however, for the same CPP at hypercapnia, they are already
maximally constricted due to the case that further constriction
is negated by hypercapnia. Because different degrees of hypercapnia exert differing strengths of dilation, the calibers of the
maximally constricted cerebral resistance vessels are different at
differing severity of hypercapnia. As a consequence, the portion
of the autoregulation curve above the upper limit at hypercapnia does not overlap with normocapnia and has a steeper slope.
In summary, during hypercapnia, the plateau of the autoregulation curve is shifted upward and shortened, the lower limit
is shifted rightward, and the upper limit is shifted leftward. The
extent of these changes depends on the severity of hypercapnia.
At severe hypercapnia when cerebral resistance vessels are maximally dilated, the plateau is lost and the pressureflow relationship is linear.
Effect of Hypocapnia on Cerebral

Autoregulation
Hypocapnia decreases CBF by cerebral vasoconstriction. As
such, there are two pertinent queries relating to its effect on
Fig. 2. Effect of hypercapnia on cerebral autoregulation. Autoregulation curves are in black at normocapnia and red at hypercapnia.
Cerebral resistance vessels are illustrated in red/pink. The bold solid blue arrows indicate the dynamic shift of the maximally dilated
and constricted cerebral resistance vessels at hypercapnia. The dashed black and blue lines/arrows indicate the lower and upper
limits at normocapnia and hypercapnia, respectively. A=the curve below the lower limit; B=the plateau at normocapnia (B0), mild
hypercapnia (B1), and severe hypercapnia (B2); C=the curve above the upper limit at normocapnia (C0), mild hypercapnia (C1), and
severe hypercapnia (C2); CBF=cerebral blood flow; CPP=cerebral perfusion pressure; LL=the lower limit at normocapnia (LL0), mild
hypercapnia (LL1), and severe hypercapnia (LL2); R=calibers of cerebral resistance vessels at normocapnia (R0), mild hypercapnia
(R1), and severe hypercapnia (R2); UL=the upper limit at normocapnia (UL0), mild hypercapnia (UL1), and severe hypercapnia (UL2).
Anesthesiology 2015; 122:196-205 198
EDUCATION
cerebral autoregulation. The first is how the lower limit moves

because the effects of hypotension and hypocapnia on cerebral
resistance vessels are opposite; one dilates and the other constricts. The other query is how the upper limit shifts because both
hypertension and hypocapnia induce cerebral vasoconstriction.
In 1965, Harper and Glass21 showed by bleeding dogs
that when MAP was reduced to 100 mmHg from a baseline
of 150 mmHg, the CBF response to hypocapnia persisted
but was much reduced, and when MAP was further reduced
to 50 mmHg, the CBF response to hypocapnia was lost. They
proposed an over-ride mechanism to explain their finding
by theorizing that responding to tissue ischemia and hypoxia
takes precedence over the maintenance of tissue carbon dioxide homeostasis. Work conducted by Whitelaw etal.22 in
newborn piglets subjected to bleeding showed that hypocapnia failed to further decrease CBF when MAP was reduced to
38 mmHg or less even though it could produce a substantial
decrease in CBF at an MAP of 45 mmHg or above.
In the context of drug-induced hypotension, serial works
by Artru etal.2325 showed that the cerebrovascular response
to hypocapnia was abolished in dogs during hypotension
to an MAP of 50 mmHg induced with sodium nitroprusside,23,24 trimethaphan,23 and nitroglycerin.25 In contrast,
Matta etal.26 showed that the cerebrovascular response
to hypocapnia was attenuated, but not abolished, during
nitroprusside-induced hypotension (MAP=60 mmHg) in
patients anesthetized with isoflurane. Similarly, Endoh etal.27
showed that nicardipine, nitroglycerin, and prostaglandin E1induced hypotension (MAP=55 to 60 mmHg) attenuated,
but did not abolish, the cerebrovascular hypocapnia response
in patients anesthetized with propofol and fentanyl.
With regard to the hypotension induced by anesthetic
agents, Okuda etal.28 showed that the cerebrovascular hypocapnia response was abolished at an MAP of 45 mmHg
induced by halothane in baboons. Artru etal.24 showed in dogs
that the cerebrovascular response to hypocapnia was partially
preserved during isoflurane-induced hypotension to an MAP
of 50 mmHg. In patients, Matta etal.26 found that the cerebrovascular carbon dioxide reactivity was attenuated during
isoflurane-induced hypotension as low as 60 mmHg (MAP).
Collectively, there is ample evidence demonstrating that
the cerebrovascular reactivity to hypocapnia is significantly
attenuated or abolished during hemorrhage-, drug-, or anesthesia-induced hypotension.2128 However, these studies do
not directly tackle the question of how hypocapnia affects
cerebral autoregulation. The effect of hypotension on cerebrovascular hypocapnia reactivity and effect of hypocapnia
on cerebrovascular pressure reactivity, that is, autoregulation,
are related but distinctive issues. The former is typically studied via investigating the change in CBF following a change
in Paco2 under hypotensive condition, whereas the latter via
investigating the change in CBF following a change in CPP
at hypocapnia. Surprisingly, there is scant evidence directly
addressing the latter issue. A study conducted by Artru etal.
examined the effect of hypocapnia on the lower limit while
Anesthesiology 2015; 122:196-205 199
the CPP was gradually decreased via hemorrhage in dogs.

They found that hypocapnia did not cause a substantial
shift of the lower limit which was at 61% of baseline CPP
at hypocapnia and 59% of baseline CPP at normocapnia,
and that the slopes of the autoregulation curve below the
lower limit did not significantly differ between hypocapnia
and normocapnia.29
In light of these considerations, we propose the constructs
illustrated in figures3 and 4 to describe the effect of hypocapnia on cerebral autoregulation. The constructs in figures
3 and 4 are based on the distinct speculation on the effect
of hypocapnia on the upper limit. The ensuing discussion
in this section is based on these two figures unless specified
otherwise. The plateau descends to a lower CBF with hypocapnia due to cerebral vasoconstriction.
To the best of our knowledge, the study by Artru etal.29 is
the only one so far that directly examined the effect of hypocapnia on the lower limit. In accordance with the results in
the study by Artru etal., we have kept the position of the
lower limit at hypocapnia the same as normocapnia and the
slope of the autoregulation curve below the lower limit at
hypocapnia not significantly different from normocapnia.
Nonetheless, the abundant evidence that the cerebrovascular reactivity to hypocapnia is significantly weakened or lost
during hypotension makes the drawing of the autoregulation
curve in the proximity of the lower limit during hypocapnia complex.2128 One could argue that the plateau should
swing upward when the CPP is critically decreased due to the
attenuation of the cerebrovascular hypocapnia response by
hypotension. This seems rational when considering that the
over-ride of the hypocapnia vasoconstrictive effect could
have restored the decreased CBF. Indeed, this consideration
seems supported by the rabbit data presented by Czosnyka
etal.30 where a smooth upswing of the plateau in the proximity of the lower limit was demonstrated even though it was
not clear whether the curve was generated during hypocapnia.
However, such speculations suggest that during hypocapnia
the CBF at hypotension could be higher than normotension.
To date, the available data show that the CBF during combined hypocapnia and hypotension is not higher than that
when hypocapnia and normotension are combined.22,29 In
addition, over-ride is a mechanism dealing specifically with
the effect of hypotension on cerebrovascular carbon dioxide
reactivity that is different to the effect of hypocapnia on cerebrovascular pressure reactivity or autoregulation.
How the upper limit is affected by hypocapnia is not
clear due to the lack of data. There are two lines of speculation. If we hypothesize that the calibers of the maximally constricted cerebral resistance vessels are the same
between hypocapnia and normocapnia, the construct in
figure 3 applies. In this case, the upper limit shifts leftward due to the background constriction induced by
hypocapnia. For example, at point D on the CPP axis
at normocapnia, cerebral resistance vessels are not maximally constricted yet. However, for the same CPP at
Fig. 3. Effect of hypocapnia on cerebral autoregulation. We speculate that the calibers of the maximally constricted cerebral
resistance vessels at normocapnia and hypocapnia are the same. Autoregulation curves are in black at normocapnia and blue
at hypocapnia. Cerebral resistance vessels are illustrated in red/pink. The bold solid blue arrow indicates the dynamic shift of
the maximally constricted cerebral resistance vessels at hypocapnia. The dashed black and blue lines/arrows indicate the lower
and upper limits at normocapnia and hypocapnia, respectively. A=the curve below the lower limit at normocapnia (A0), mild
hypocapnia (A1), and severe hypocapnia (A2); B=the plateau at normocapnia (B0), mild hypocapnia (B1), and severe hypocapnia
(B2); C=the curve above the upper limit at normocapnia (C0), mild hypocapnia (C1), and severe hypocapnia (C2); CBF=cerebral
blood flow; CPP=cerebral perfusion pressure; LL=the lower limit at normocapnia (LL0), mild hypocapnia (LL1), and severe
hypocapnia (LL2); R=calibers of cerebral resistance vessels at normocapnia (R0), mild hypocapnia (R1), and severe hypocapnia
(R2); UL=the upper limit at normocapnia (UL0), mild hypocapnia (UL1), and severe hypocapnia (UL2).
Fig. 4. Effect of hypocapnia on cerebral autoregulation. We speculate that the caliber of the maximally constricted cerebral
resistance vessels at hypocapnia is smaller than normocapnia and assume that hypocapnia causes a rightward shift of the upper limit. Autoregulation curves are in black at normocapnia and blue at hypocapnia. Cerebral resistance vessels are illustrated
in red/pink. The bold solid blue arrow indicates the dynamic shift of the maximally constricted cerebral resistance vessels at
hypocapnia. The dashed black and blue lines/arrows indicate the lower and upper limits at normocapnia and hypocapnia,
respectively. A=the curve below the lower limit at normocapnia (A0), mild hypocapnia (A1), and severe hypocapnia (A2); B=the
plateau at normocapnia (B0), mild hypocapnia (B1), and severe hypocapnia (B2); C=the curve above the upper limit at normocapnia (C0), mild hypocapnia (C1), and severe hypocapnia (C2); CBF=cerebral blood flow; CPP=cerebral perfusion pressure;
LL=the lower limit at normocapnia (LL0), mild hypocapnia (LL1), and severe hypocapnia (LL2); R=calibers of cerebral resistance
vessels at normocapnia (R0), mild hypocapnia (R1), and severe hypocapnia (R2); UL=the upper limit at normocapnia (UL0), mild
hypocapnia (UL1), and severe hypocapnia (UL2).
hypocapnia, they are maximally constricted as a result of

the additional constriction imposed by hypocapnia. We
overlap the portions of the autoregulation curve above the
Anesthesiology 2015; 122:196-205 200
upper limit at hypocapnia and normocapnia. The overlap

is plausible in considering that, for a given CPP higher
than the upper limit, the CBF at hypocapnia should be
EDUCATION
the same as normocapnia because the flow resistances

determined by the calibers of cerebral resistance vessels
are the same.
The other line of speculation is to hypothesize that the
caliber of the maximally constricted cerebral resistance vessels at hypocapnia is smaller than normocapnia due to the
extra constriction imposed by hypocapnia. In this case, the
construct in figure 4 may apply. The upper limit may or
may not shift rightward even though we opt for a rightward
shift in this discussion. This line of speculation is shared by
Paulson etal.31 who believed that, at hypocapnia, the upper
limit shifted rightward and as a consequence, the plateau
was lengthened. However, neither Paulson etal. nor the two
studies cited by Paulson etal.14,15 specifically studied the
effect of hypocapnia on the upper limit. Future research is
warranted to address this unknown aspect.
In summary, at hypocapnia, the plateau of the autoregulation curve shifts downward; any change in the lower limit is
unremarkable; however, how the upper limit moves is not clear.
Methodological Considerations
We have discussed the integrated effect of carbon dioxide
and perfusion pressure on the cerebral circulation based on
the published large animal and human experimental data
and some speculation on the aspects that are without data.
Effort has been made to ensure that the available data and
the proposed conceptualization do not conflict. Nonetheless, challenges exist due to the differences in the methods
used by previous investigators that include study subjects,
CPP manipulation, Paco2 manipulation, CBF measurement,
and anesthesia choice (table1). Data do not automatically
deliver concepts or theories, and it is especially unlikely that
a single data set could do so. To establish a complete theory,
careful data analysis and logical extrapolation are needed in
addition to some speculation on the components that are
without direct data support.
Studies on the relationship between distinct physiological processes can be confounded by unrecognized or
unmeasured processes associated with the processes being
studied. Sympathetic nervous activity can be a confounder
when studying the effect of hypercapnia on cerebral circulation. Busija and Heistad32 showed that the hypercapniainduced increase in CBF was further increased by bilateral
sympathectomy and attenuated by bilateral electrical stimulation of the superior cervical ganglion in anesthetized cats
and unanesthetized rabbits. Cassaglia etal.33 also reported
that the CBF was augmented by sympathetic withdrawal
via bilateral superior cervical ganglionectomy during acute
hypercapnia in sleeping lambs. Zhang etal.34 showed that
the sensitivity of cerebral vasoreactivity to hypercapnia was
attenuated by augmented sympathetic activity induced via
lower body negative pressure in healthy young subjects.
Moreover, it is known that hypercapnia can increase sympathetic activity alone35 and with hypoxia.36 Collectively, this
evidence suggests that the hypercapnia-induced increase
Anesthesiology 2015; 122:196-205 201
in CBF is buffered by sympathetic nervous activity. During anesthesia, sympathetic activity is altered, ranging from
inhibition to stimulation depending on the anesthetic agent
being used,37 implying that sympathetic activity could confound a study on the effect of hypercapnia on cerebral circulation in anesthetized subjects.
Hypocapnia is achieved via hyperventilation in all studies we referenced. However, hyperventilation itself can be a
confounder. Alexander etal.38 showed that hyperventilation
caused a consistent increase in blood pressure and decrease
in cardiac output in patients anesthetized with propofol and
remifentanil. The associated changes in blood pressure and
cardiac output can confound studies on the effect of hypocapnia on cerebral circulation due to the influence of systemic circulation on cerebral perfusion.
A variety of anesthetic agents were used in previous studies (table 1). Type of anesthesia can also be a confounder. It is
known that volatile agents cause an increase in CBF due to
the intrinsic cerebral vasodilatory property,39 whereas propofol
representing intravenous agents exerts the opposite effect.40 As
a result, volatile agents impair cerebral autoregulation while
propofol preserves it,41 somewhat resembling hypercapnia and
hypocapnia, respectively. Therefore, the effect of carbon dioxide on cerebral autoregulation should also be considered in the
context of the anesthetic agent being used.
We focused our discussion on the integrated effect of
carbon dioxide and perfusion pressure on the cerebral circulation. Although it is important to understand the physiology comprehensively, we did not include every aspect that
engages in CBF regulation. Oxygen is one of these aspects
that deserve emphasis. In 1940s, Kety and Schmidt42 showed
in volunteers that an inspired oxygen fraction of 0.85 to 1.0
was associated with a 13% reduction in CBF, whereas an
inspired oxygen fraction of 0.1 produced a 35% increase in
CBF. In 1960s, Hggendal and Johansson15 showed that a
decrease in arterial blood oxygen saturation from 90% to 20
to 30% led to an increase in CBF of 100%. Gupta etal.43
found that the threshold for hypoxic cerebral vasodilation
was at a peripheral oxygen saturation of 90% in healthy
volunteers, much higher than previously reported. Brown
etal.44 argued that it is the arterial blood oxygen content
that is fundamentally important in the regulation of CBF.
On the mechanism, adenosine, nitric oxide, cyclic nucleotides, and adenosine triphosphatesensitive K+ channels are
all implicated as being responsible for the hypoxia-induced
cerebral vasodilation.45 On the interplay among oxygen, carbon dioxide, and perfusion pressure on the cerebral circulation, it was found that the cerebrovascular carbon dioxide
reactivity was attenuated during acute hypoxia46 and the
impairment of dynamic cerebral autoregulation during isocapnic hypoxia could be prevented with hypocapnia.47 Conversely, brain tissue oxygen tension is regulated by carbon
dioxide and perfusion pressure, resembling the well-known
CBF regulation by carbon dioxide and CPP, respectively.48
This observation reflects that one of the fundamental goals
Anesthesiology 2015; 122:196-205 202

Phencyclidine
Details not disclosed
Propofol infusion or sevoflurane, in
addition to remifentanil infusion
Fentanyl, diazepam
Thiopentone, nitrous oxide
Chloralose, urethane
Halothane, nitrous oxide
Isoflurane, nitrous oxide
Isoflurane, fentanyl infusion
Propofol and fentanyl infusion
Nitrous oxide, halothane
Thiopental, nitrous oxide
Dog
Dog
Monkey
Dog
Human
Human (cardiopulmonary
bypass)
Very low-birth-weight infant
Dog
Newborn piglet
Dog
Dog
Dog
Human
Human
Baboon
Dog
Human
Harper14
Hggendal15
Raichle16
Ekstrm-Jodal17
McCulloch18
Murkin19
Kaiser20
Harper21
Whitelaw22
Artru23
Artru24
Artru25
Matta26
Endoh27
Okuda28
Artru29
Paulson31
References are tabulated in the sequence they are referenced in the text.
Phencyclidine, thiopentone,
nitrous oxide
Thiopentone, nitrous oxide
Pentobarbital
Anesthesia
Baboon
Subject
Fitch8
First Author
Table 1. Details of the Methods Used in the Studies Cited
Xe washout, electromagnetic flowmeter

Diversion of sagittal sinus
blood flow
133
Xe washout
133
Doppler ultrasonography
KetySchmidt (krypton 85)
blood flow
blood flow
blood flow
Xe washout
133

Xe washout
133
Cerebral Blood Flow

Measurement
Angiotensin
Bleeding
Nitroprusside, isoflurane
Nicardipine, nitroglycerin,
prostaglandin E1
Halothane
Nitroglycerin
Isoflurane
Tracheal suction
Bleeding
Bleeding
Nitroprusside, trimethaphan
Bleeding and m
etaraminol
infusion
Bleeding, thoracic aorta
clamping
Phenylephrine infusion
Bleeding, halothane,
trimetaphan, nitroprusside
Bleeding
Bleeding
Cerebral Perfusion Pressure Manipulation
Hyperventilation
Hyperventilation, breathing
carbon dioxide
Ventilation adjustment
Hyperventilation
Hyperventilation
Breathing carbon dioxide
Hyperventilation
Hyperventilation
carbon dioxide
Hypoventilation
Hypoventilation
carbon dioxide
Breathing carbon dioxide
Constant
Carbon Dioxide
Manipulation
EDUCATION
of cerebral perfusion is oxygen delivery. In summary, oxygen

regulates CBF both alone and via an integrated mechanism
that involves interplay with carbon dioxide, perfusion pressure, and maybe other physiological processes.
It needs to be noted that the flat plateau (zero tilt) of the
autoregulation curve is likely an idealized drawing. In reality,
cerebral autoregulation may execute on a (slightly) tilted plateau that is different to pressure-passive flow.49 Moreover, the
sharp inflection points at both the lower and the upper limits should probably be drawn as a round shoulder rather
than a sharp elbow because the former conforms to normal
physiology, whereas the latter is derived as a result of statistical processing.4
Clinical Implications
Cerebral autoregulation is an important mechanism in protecting the brain from ischemia and overperfusion in the
face of fluctuating perfusion pressure. As such, it is regularly
referenced in clinical practice when taking care of patients
with or without neurologic pathophysiologies. However,
ignorance of both the limitations and the dynamic/integrative nature of this concept can do more harm than good. The
practice of applying a fixed number learned from textbooks
or other resources in an individual patient is risky for the
following reasons. First, it can either underestimate or overestimate the true value of the lower limit, the upper limit, or
the plateau of an individual patient because the commonly
quoted numbers are the means of the populations studied
without noting the SD or range of distribution. Second, the
functional status of cerebral autoregulation is not routinely
monitored in clinical care. It can be impaired in a variety
of situations such as traumatic brain injury50 and anesthesia
with volatile agents.41 If so, CBF becomes pressure passive
and a different conceptual framework is needed. Finally,
nonperfusion pressure conditions or processes, such as carbon dioxide as discussed in this article, can alter the position
and shape of the autoregulation curve via their modulating
effect on cerebral vasomotor tone. Therefore, cerebral autoregulation is a dynamic process that is regulated by nonperfusion pressure but CBF-regulating aspects.
The aim of this review is to reanalyze the conceptualization
of cerebral autoregulation and not to deal specifically with
the relationship between decreased (or increased) blood pressure and neurological outcome. This was recently reviewed.51
Nonetheless, a pertinent clinical question is what the practical
strategy of arterial blood pressure management is when realtime cerebral perfusion and autoregulation are not monitored.
Unfortunately, there is no single or simple answer. It depends
on the patients neurologic pathophysiology including cerebral
metabolic need, adequacy of perfusion, intracranial pressure,
and integrity of cerebral autoregulation, in addition to the
presence of cardiac disease, pulmonary disease, anemia, and
so on, as well the largely unknown effects of vasoactive drugs.
Clinical care should balance the needs of different organ systems. The complexity of this philosophy is illustrated, for
Anesthesiology 2015; 122:196-205 203
example, with the triple H (hypertension, hypervolemia,

and hemodilution) therapy that benefits the brain but may
harm the heart52 and the perioperative -blockade therapy
that helps the heart but may hurt the brain.53 Although the
causeeffect role of blood pressure in these dilemmas is hard to
define, it is clear that blood pressure management is a decision
characterized by priority and balance. The ultimate vindication of any intervention should be based on randomized and
controlled trials demonstrating an overall beneficial outcome
and this applies equally to the care of the systemic and cerebral
circulations. Thus far, large meaningful trials are lacking.
The clinical implications of the effect of carbon dioxide on cerebral autoregulation are summarized in figure5.
An increase in CBF due to hypercapnia renders the match
between CBF and cerebral metabolic rate tilted toward
more CBF than needed if cerebral metabolic rate remains
unchanged (fig. 2). This may be seen as a safer situation in
terms of the maintenance of the supply of cerebral metabolic substrates. However, it needs to be noted that an acute
increase in Paco2 not only shifts the plateau up but also
shortens it. A shrunken plateau increases the chance of CBF
fluctuation as CPP fluctuates. Under anesthesia, a Paco2 of
greater than 55 mmHg should be regarded as having eliminated autoregulation.18 Therefore, a tighter CPP control is
needed to avoid CBF fluctuation although some protection may come from the higher-than-needed CBF (assuming a stable metabolic demand) that would allow a greater
decrease in perfusion pressure before ischemia.
Fig. 5. Clinical implications of the effects of hypercapnia

and hypocapnia on cerebral autoregulation. CBF = cerebral blood flow; = increase or upward shift; = decrease
or downward shift; = rightward shift; = leftward shift;
=insignificant shift.
If cerebral metabolic rate remains unchanged, a decrease

in CBF due to hypocapnia renders the brain at risk of cerebral ischemia (figs. 3 and 4). It is true that the cerebrovascular response to hypocapnia is attenuated during hypotension
secondary to hemorrhage, drug, or anesthesia.2128 However,
this over-ride mechanism deals specifically with the effect of
hypotension on the cerebrovascular carbon dioxide reactivity,
not with the effect of hypocapnia on cerebrovascular pressure
reactivity (autoregulation). Studies showed that the CBF was
significantly reduced during combined hypocapnia and hypotension.24,29 There is no evidence showing that the CBF during combined hypocapnia and hypotension is increased above
baseline. Therefore, our cautious recommendation is to avoid
hypotension during hypocapnia to decrease the ischemic risk.
The decision to implement hypocapnia in clinical care should
be weighed against the inherent ischemic risk it incurs. The
deleterious effect of hypocapnia in patients with head trauma
was reviewed.54 Hypocapnia was also associated with unfavorable functional outcomes at 90 days after acute stroke.55
Summary
Cerebral autoregulation is a mechanism that maintains CBF
stable despite the fluctuation of CPP. As such, it is regularly
referenced in clinical care; however, ignorance of its dynamic
nature and limitations can do more harm than good. Nonperfusion pressure but CBF-regulating processes such as carbon dioxide affect the efficiency of pressure autoregulation
because they intercept at the same targetthe cerebrovascular reactivity. The integrated effect of carbon dioxide and
perfusion pressure on cerebral circulation is discussed based
on the published large animal and human data and some
speculation on the aspects that are without data support.
We showed that during hypercapnia, the plateau ascends
and shortens, the lower limit shifts rightward, and the upper
limit leftward. Conversely, during hypocapnia, the plateau
descends and the lower limit remains unchanged. How the
upper limit is affected by hypocapnia is not clear; nonetheless, we provided two lines of speculation: one line assuming
the same calibers of the maximally constricted cerebral resistance vessels at hypocapnia and normocapnia and the other
assuming a smaller caliber at hypocapnia than normocapnia.
Acknowledgments
Competing Interests
Correspondence
Address correspondence to Dr. Meng: Department of Anesthesia and Perioperative Care, University of California San
Francisco, 521 Parnassus Avenue, Suite C450, San Francisco,
California 94143. mengl@anesthesia.ucsf.edu. Information
on purchasing reprints may be found at www.anesthesiolAnesthesiology 2015; 122:196-205 204
ogy.org or on the masthead page at the beginning of this

issue. Anesthesiologys articles are made freely accessible to
all readers, for personal use only, 6 months from the cover
date of the issue.
References
1. Lassen NA, Christensen MS: Physiology of cerebral blood
flow. Br J Anaesth 1976; 48:71934
2. Paulson OB, Strandgaard S, Edvinsson L: Cerebral autoregulation. Cerebrovasc Brain Metab Rev 1990; 2:16192
3. Lassen NA: Cerebral blood flow and oxygen consumption in
man. Physiol Rev 1959; 39:18338
4. Drummond JC: The lower limit of autoregulation: Time to
revise our thinking? Anesthesiology 1997; 86:14313
5. Tzeng YC, Ainslie PN: Blood pressure regulation IX: Cerebral
autoregulation under blood pressure challenges. Eur J Appl
Physiol 2014; 114:54559
6. Larsen FS, Olsen KS, Hansen BA, Paulson OB, Knudsen GM:
Transcranial Doppler is valid for determination of the lower
limit of cerebral blood flow autoregulation. Stroke 1994;
25:19858
7. Joshi B, Ono M, Brown C, Brady K, Easley RB, Yenokyan
G, Gottesman RF, Hogue CW: Predicting the limits of cerebral autoregulation during cardiopulmonary bypass. Anesth
Analg 2012; 114:50310
8. Fitch W, Ferguson GG, Sengupta D, Garibi J, Harper AM:
Autoregulation of cerebral blood flow during controlled
hypotension in baboons. J Neurol Neurosurg Psychiatry
1976; 39:101422
9. Gibbs FA, Gibbs EL, Lennox WG: Changes in human cerebral
blood flow consequent on alterations in blood gases. Am J
Physiol 1935; 111:55763
10. Meng L, Gelb AW, Alexander BS, Cerussi AE, Tromberg BJ,
Yu Z, Mantulin WW: Impact of phenylephrine administration
on cerebral tissue oxygen saturation and blood volume is
modulated by carbon dioxide in anaesthetized patients. Br J
Anaesth 2012; 108:81522
11. Kety SS, Schmidt CF: The effects of active and passive hyperventilation on cerebral blood flow, cerebral oxygen consumption, cardiac output, and blood pressure of normal
young men. J Clin Invest 1946; 25:10719
12. Gelb AW, Craen RA, Rao GS, Reddy KR, Megyesi J, Mohanty
B, Dash HH, Choi KC, Chan MT: Does hyperventilation
improve operating condition during supratentorial craniotomy? A multicenter randomized crossover trial. Anesth Analg
2008; 106:58594
13. Ni Chonghaile M, Higgins B, Laffey JG: Permissive hypercapnia: Role in protective lung ventilatory strategies. Curr Opin
Crit Care 2005; 11:5662
14. Harper AM: Autoregulation of cerebral blood flow: Influence of
the arterial blood pressure on the blood flow through the cerebral cortex. J Neurol Neurosurg Psychiatry 1966; 29:398403
15. Hggendal E, Johansson B: Effects of arterial carbon dioxide
tension and oxygen saturation on cerebral blood flow autoregulation in dogs. Acta Physiol Scand Suppl 1965; 258:2753
16. Raichle ME, Stone HL: Cerebral blood flow autoregulation
and graded hypercapnia. Eur Neurol 1971; 6:15
17. Ekstrm-Jodal B, Hggendal E, Linder LE, Nilsson NJ:
Cerebral blood flow autoregulation at high arterial pressures
and different levels of carbon dioxide tension in dogs. Eur
Neurol 1971; 6:610
18. McCulloch TJ, Visco E, Lam AM: Graded hypercapnia and
cerebral autoregulation during sevoflurane or propofol anesthesia. Anesthesiology 2000; 93:12059
19. Murkin JM, Farrar JK, Tweed WA, McKenzie FN, Guiraudon
G: Cerebral autoregulation and flow/metabolism coupling
during cardiopulmonary bypass: The influence of Paco2.
EDUCATION
20. Kaiser JR, Gauss CH, Williams DK: The effects of hypercapnia on cerebral autoregulation in ventilated very low birth
weight infants. Pediatr Res 2005; 58:9315
21. Harper AM, Glass HI: Effect of alterations in the arterial carbon dioxide tension on the blood flow through the cerebral
cortex at normal and low arterial blood pressures. J Neurol
Neurosurg Psychiatry 1965; 28:44952
22. Whitelaw A, Karlsson BR, Haaland K, Dahlin I, Steen PA,
Thoresen M: Hypocapnia and cerebral ischaemia in hypotensive newborn piglets. Arch Dis Child 1991; 66(10 Spec
No):11104
23. Artru AA, Colley PS: Cerebral blood flow responses to hypocapnia during hypotension. Stroke 1984; 15:87883
24. Artru AA: Partial preservation of cerebral vascular responsiveness to hypocapnia during isoflurane-induced hypotension in dogs. Anesth Analg 1986; 65:6606
25. Artru AA: Cerebral vascular responses to hypocapnia during nitroglycerin-induced hypotension. Neurosurgery 1985;
16:46872
26. Matta BF, Lam AM, Mayberg TS, Eng CC, Strebel S:
Cerebrovascular response to carbon dioxide during sodium
nitroprusside- and isoflurane-induced hypotension. Br J
Anaesth 1995; 74:296300
27. Endoh H, Honda T, Komura N, Shibue C, Watanabe I, Shimoji
K: Effects of nicardipine-, nitroglycerin-, and prostaglandin
E1-induced hypotension on human cerebrovascular carbon
dioxide reactivity during propofol-fentanyl anesthesia. J Clin
Anesth 1999; 11:5459
28. Okuda Y, McDowall DG, Ali MM, Lane JR: Changes in CO2
responsiveness and in autoregulation of the cerebral circulation during and after halothane-induced hypotension.
JNeurol Neurosurg Psychiatry 1976; 39:22130
29. Artru AA, Katz RA, Colley PS: Autoregulation of cerebral
blood flow during normocapnia and hypocapnia in dogs.
30. Czosnyka M, Brady K, Reinhard M, Smielewski P, Steiner LA:
Monitoring of cerebrovascular autoregulation: Facts, myths,
and missing links. Neurocrit Care 2009; 10:37386
31. Paulson OB, Olesen J, Christensen MS: Restoration of autoregulation of cerebral blood flow by hypocapnia. Neurology
1972; 22:28693
32. Busija DW, Heistad DD: Effects of activation of sympathetic
nerves on cerebral blood flow during hypercapnia in cats
and rabbits. J Physiol 1984; 347:3545
33. Cassaglia PA, Griffiths RI, Walker AM: Sympathetic withdrawal augments cerebral blood flow during acute hypercapnia in sleeping lambs. Sleep 2008; 31:172934
34. Zhang P, Huang G, Shi X: Cerebral vasoreactivity during
hypercapnia is reset by augmented sympathetic influence.
JAppl Physiol (1985) 2011; 110:3528
35. Somers VK, Mark AL, Zavala DC, Abboud FM: Contrasting
effects of hypoxia and hypercapnia on ventilation and sympathetic activity in humans. J Appl Physiol (1985) 1989; 67:21016
36. Morgan BJ, Crabtree DC, Palta M, Skatrud JB: Combined
hypoxia and hypercapnia evokes long-lasting sympathetic
activation in humans. J Appl Physiol (1985) 1995; 79:20513
37. Maignan E, Dong WX, Legrand M, Safar M, Cuche JL:
Sympathetic activity in the rat: Effects of anaesthesia on noradrenaline kinetics. J Auton Nerv Syst 2000; 80:4651
38. Alexander BS, Gelb AW, Mantulin WW, Cerussi AE, Tromberg
BJ, Yu Z, Lee C, Meng L: Impact of stepwise hyperventilation on cerebral tissue oxygen saturation in anesthetized
patients: A mechanistic study. Acta Anaesthesiol Scand 2013;
57:60412
Anesthesiology 2015; 122:196-205 205
39. Matta BF, Mayberg TS, Lam AM: Direct cerebrovasodilatory

effects of halothane, isoflurane, and desflurane during propofol-induced isoelectric electroencephalogram in humans.
Anesthesiology 1995; 83:9805; discussion 27A
40. Eng C, Lam AM, Mayberg TS, Lee C, Mathisen T: The influence of propofol with and without nitrous oxide on cerebral blood flow velocity and CO2 reactivity in humans.
41. Strebel S, Lam AM, Matta B, Mayberg TS, Aaslid R, Newell
DW: Dynamic and static cerebral autoregulation during isoflurane, desflurane, and propofol anesthesia. Anesthesiology
1995; 83:6676
42. Kety SS, Schmidt CF: The effects of altered arterial tensions
of carbon dioxide and oxygen on cerebral blood flow and
cerebral oxygen consumption of normal young men. J Clin
Invest 1948; 27:48492
43. Gupta AK, Menon DK, Czosnyka M, Smielewski P, Jones JG:
Thresholds for hypoxic cerebral vasodilation in volunteers.
44. Brown MM, Wade JP, Marshall J: Fundamental importance of
arterial oxygen content in the regulation of cerebral blood
flow in man. Brain 1985; 108 (Pt 1):8193
45. Armstead WM: Role of nitric oxide, cyclic nucleotides, and
the activation of ATP-sensitive K+ channels in the contribution of adenosine to hypoxia-induced pial artery dilation.
JCereb Blood Flow Metab 1997; 17:1008
46. Ogoh S, Nakahara H, Ueda S, Okazaki K, Shibasaki M,
Subudhi AW, Miyamoto T: Effects of acute hypoxia on cerebrovascular responses to carbon dioxide. Exp Physiol 2014;
99:84958
47. Querido JS, Ainslie PN, Foster GE, Henderson WR, Halliwill
JR, Ayas NT, Sheel AW: Dynamic cerebral autoregulation during and following acute hypoxia: Role of carbon dioxide. J
Appl Physiol (1985) 2013; 114:118390
48. Hemphill JC III, Knudson MM, Derugin N, Morabito D,
Manley GT: Carbon dioxide reactivity and pressure autoregulation of brain tissue oxygen. Neurosurgery 2001; 48:37783;
discussion 3834
49. Tan CO: Defining the characteristic relationship between
arterial pressure and cerebral flow. J Appl Physiol (1985)
2012; 113:1194200
50. Engelborghs K, Haseldonckx M, Van Reempts J, Van Rossem
K, Wouters L, Borgers M, Verlooy J: Impaired autoregulation
of cerebral blood flow in an experimental model of traumatic
brain injury. J Neurotrauma 2000; 17:66777
51. Bijker JB, Gelb AW: Review article: The role of hypotension
in perioperative stroke. Can J Anaesth 2013; 60:15967
52. Sen J, Belli A, Albon H, Morgan L, Petzold A, Kitchen N:
Triple-H therapy in the management of aneurysmal subarachnoid haemorrhage. Lancet Neurol 2003; 2:61421
53. Devereaux PJ, Yang H, Yusuf S, Guyatt G, Leslie K, Villar JC,
Xavier D, Chrolavicius S, Greenspan L, Pogue J, Pais P, Liu
L, Xu S, Mlaga G, Avezum A, Chan M, Montori VM, Jacka
M, Choi P; POISE Study Group: Effects of extended-release
metoprolol succinate in patients undergoing non-cardiac
surgery (POISE trial): A randomised controlled trial. Lancet
2008; 371:183947
54. Curley G, Kavanagh BP, Laffey JG: Hypocapnia and the
injured brain: More harm than benefit. Crit Care Med 2010;
38:134859
55. Takahashi CE, Brambrink AM, Aziz MF, Macri E, Raines J,
Multani-Kohol A, Hinson HE, Lutsep HL, Clark WM, Fields
JD: Association of intraprocedural blood pressure and end
tidal carbon dioxide with outcome after acute stroke intervention. Neurocrit Care 2014; 20:2028
Carol Wiley Cassella, M.D., Editor
Electrocardiogram
Audrey Shafer, M.D.
The repolarization of a t wave

an ocean roller over an invisible sand bar, a lull
until a p wave begins its small rise above the isoelectric
the chop of look-at-me qrss
not called complexes for nothing
peaks, troughs, weddings, divorces
births transforming more than bodies into mothers
and deaths
underneath the swells
salt churns cold and warm currents
a low t wave stretches into a u
waves in succession
storms of qrss surging into ts
wide and lopsided as sinking freighters
From the Stanford University School of Medicine and Veterans Affairs Palo Alto Health Care System, Palo Alto, California. ashafer@
stanford.edu
Accepted for publication May 30, 2014.
Anesthesiology,
122
No 1
Anesthesiology,
V 122 VNo
1 206
206
January 2015
EDUCATION
till a pause between series

the ocean a windless lake
no more ps, no more qrss, no more ts
this is my autobiography:
the story of a wave
and then the next
Mind To Mind
CORRESPONDENCE
Is the Standard Supplied by the

Association for the Advancement
of Medical Instrumentation the
Measure of All Things for Noninvasive
Continuous Hemodynamic Devices?
To the Editor:
The systematic review and meta-analysis by Kim et al.1 raise
important questions regarding the correct way to technically
evaluate the new generation of continuous noninvasive arterial pressure monitors (CNBP).
Continuous noninvasive arterial pressure monitors
devices have been designed to extend the physiological information available regarding a patients hemodynamic status.
As rightly mentioned by De Hert,2 they cannot completely
replace invasive monitoring, because blood gas sampling
may also be required. However they have the potential to
bridge the gap between noninvasive but intermittent and continuous but invasive arterial pressure measurements.1
Continuous noninvasive arterial pressure monitors are
designed to measure blood pressure (BP) trending data,
beat-to-beat rhythms, and waveforms. The BP signal can
then be further processed to provide additional, and important insights into dynamic fluid responsiveness parameters,
stroke volume, and cardiac output changes. If accurate and
clinically useful, these monitors may potentially have quite
profound impact on patient care. It is important to build
on our understanding of the performance and limitations of
these newly available devices.
There is currently no accepted standard way to evaluate the
clinical performance of this new generation of monitors. A
standard from the Association for the Advancement of Medical Instrumentation (AAMI) is available3 for intermittent
noninvasive cuff sphygmomanometers (NBP) by utilizing
an upper-arm cuff. In the absence of consistency in reporting
accuracy results by the papers reviewed in the meta-analysis,
the authors used this AAMI standard benchmark for accuracy and precision (58 mmHg). In doing so, the authors
conclusion was that the precision and accuracy results found
for CNBP are outside those deemed acceptable by the AAMI
standard. However, we believe that there are some important
technical and methodological limitations associated with
applying this standard that lead us to suggest using caution
with the interpretation and extrapolation of these results.
Firstly, we fully agree with Kim et al. that most studies
evaluating these devices did not report bias and error the way
the AAMI recommended, and following the recommendations
in that standard would have led to significantly lower error
values. Indeed, it is accepted by Kim et al. that AAMI recommendations3,4 have not been implemented correctly in
Copyright 2014, the American Society of Anesthesiologists, Inc. Lippincott
Williams & Wilkins. Anesthesiology 2015; 122:208-21
any of the studies reviewed in their meta-analysis. In fact,

the AAMI criteria define, in a very detailed manner, the
way comparisons of intermittent oscillometric sphygmomanometers with intraarterial BP (IBP) measurements should
be conducted. The critical analytical method missing from
these studies is the use of the zero-zone proposed within
the AAMI guidelines. The zero-zone is calculated using
IBP-values of a minimum of 30 s: Beat-to-beat IBP-values
are averaged and their standard deviations (SDs) are calculated. This results in the zero-zones (mean 1 SD), to
which the device-under-test is then compared.
Figure1 illustrates the process to determine differences
to the device-under-test derived using the AAMI zero-zone
definition. The range of the reference diastolic BP (dBP) is
66 to 74 mmHg, which defines the zero-zone. If the dBP
value determined by the sphygmomanometer is 67 mmHg
the error for this determination is 0 mmHg! The range of the
reference systolic BP (sBP) is 123 to 136 mmHg. If the sBP
determined by the sphygmomanometer is 121 mmHg the
error for this determination is 2 mmHg.
Using the zero-zone approach, the AAMI benchmark of
acceptable accuracy and precision of 58 mmHg are wider
than suggested in the analysis used by Kim et al. This change
in allowable bias is mathematically equivalent to increasing
the AAMI benchmark by half of the zero-zone range (the
SD of IBP). Adding the maximum allowable ranges of the
Fig. 1. The process of determining the error of the device-under-test toward the reference intraarterial blood pressure (IBP)
signal using the Association for the Advancement of Medical
Instrumentation zero-zone: The beat-to-beat diastolic and systolic blood pressure (sBP) values of the reference IBP signal are
determined for the shaded 30-s intervals (zero-zone calculation zones I and II) and their ranges, averages, and SDs are calculated. The NBP device-under-test derives its measurements
from 0 to 30 s (dashed triangle). In this example, the range of
the reference beat-to-beat diastolic is 66 to 74 mmHg, which
defines the diastolic zero-zone (black rectangle). If the beatto-beat diastolic value determined by NBP is 67 mmHg, the
zero-zone difference is 0 whereas the mean difference (i.e., the
difference to the average IBP value, horizontal dotted line) is
3. The range of the reference sBP is 123 to 136 mmHg. If the
sBP determined by NBP is 121 mmHg, the zero-zone difference is 2 mmHg whereas the mean difference is 9.
January 2015
CORRESPONDENCE
zero-zones as defined by the AAMI (20 mmHg for sBP and 12

mmHg for dBP)3 results in the benchmarks being extended
by 10 to 1518 mmHg for sBP and by 6 to 1114 mmHg
for dBP. Would that still be a clinically acceptable bias?
None of the studies cited by Kim et al. applied this zerozone, but simply calculated accuracy as the average (more or
less beat-to-beat) difference between CNBP and IBP (for this
example 9 mmHg for sBP and 3 mmHg for dBP). It is obvious that this omission has a profound negative impact on the
reported accuracy and precision. In our view, the data should be,
where possible, reanalyzed in the manner outlined in the AAMI
standard. Comparison of the data with the AAMI benchmark
standard without such zero-zone analysis is not appropriate.
Secondly, the original AAMI approach requires patients to
be hemodynamically stable, so that the sphygmomanometer is
able to deliver plausible values. Maximum allowable ranges of
the zero-zones are exactly defined by AAMI and all data from
a subject shall be excluded if the reference zero-zone is greater
than the maximum ranges.3 Establishing the accuracy of
CNBP in hemodynamically unstable situations like induction
of anesthesia5,6 or even transfemoral aortic valve implantation
procedures7 provokes additional and challenging questions as
to what is the appropriate methodology and what statistics to
use? Including studies in the meta-analysis with data gained
from unstable patients or patients episodes may help explain
the high variation in accuracy and precision results reported.
As stated by Kim et al., their inclusion criteria may limit the
collective comparison of all studies in a single meta-analysis.
The need to continue to explore methodologies for the
testing and evaluation of all BP monitoring devices under
real-life clinical conditions is emphasized by the results
reported by Wax et al.8 They analyzed the difference between
sphygmomanometer and IBP readings from 24,225 patients
in daily clinical routine. The real-world average bias (SD)
was 1 mmHg (16 mmHg) for sBP and 5 mmHg (11
mmHg) for dBPwithout applying a zero-zone.
In conclusion, there is a growing clinical requirement for
the noninvasive monitoring of continuous BP, fluid, and stroke
volume. Such devices are now available to clinicians and are
being used in increasing numbers. It is therefore of major
importance to further discuss acceptable and reasonable continuous BP evaluation standards. A new evaluation standard
should be defined taking into consideration the inherent beatto-beat nature and trending capability of these new devices.
Competing Interests
CNSystems Medizintechnik AG (Graz, Austria) develops,
manufactures, and markets the continuous noninvasive
arterial pressure technology. LiDCO Ltd. (London, United
Kingdom) has integrated the continuous noninvasive arterial
pressure technology into their LiDCOrapid products. The
authors are inventors and named on one or more patents of
continuous noninvasive technology.
Drs. Fortin and OBrien are CEOs and founders of their
companies, receive salary, and have equity interests. Ms.Lerche
and Dr. Flotzinger are employees of CNSystems AG.
Jrgen Fortin, Ph.D., Katja Lerche, M.Sc., Doris Flotzinger, Ph.D., Terence OBrien, Ph.D. CNSystems Medizintechnik AG, Graz, Austria (J.F.). juergen.fortin@cnsystems.com
References
1. Kim SH, Lilot M, Sidhu KS, Rinehart J, Yu Z, Canales C,

Cannesson M: Accuracy and precision of continuous noninvasive arterial pressure monitoring compared with invasive arterial pressure: A systematic review and meta-analysis.
2. De Hert S: Noninvasive hemodynamic monitoring devices.
3. Non-invasive
SphygmomanometersPart
2:
Clinical
Investigation of Automated Measurement Type, ANSI/AAMI/
ISO 810602:2013. Arlington, Association for the Advancement
of Medical Instrumentation, 2013, pp 122
4. Non-invasive
2:
Clinical
Validation of Automated Measurement Type, ANSI/AAMI/ISO
810602:2009. Arlington, Association for the Advancement of
Medical Instrumentation, 2009, pp 121
5. Ilies C, Bauer M, Berg P, Rosenberg J, Hedderich J, Bein B, Hinz
J, Hanss R: Investigation of the agreement of a continuous noninvasive arterial pressure device in comparison with invasive
radial artery measurement. Br J Anaesth 2012; 108:20210
6. Gayat E, Mongardon N, Tuil O, Sievert K, Chazot T, Liu N,
Fischler M: CNAP() does not reliably detect minimal or
maximal arterial blood pressures during induction of anaesthesia and tracheal intubation. Acta Anaesthesiol Scand 2013;
57:46873
7. Schramm C, Baat L, Plaschke K: Continuous noninvasive
arterial pressure: Assessment in older and high-risk patients
under analgesic sedation. Blood Press Monit 2011; 16:
2706
8. Wax DB, Lin HM, Leibowitz AB: Invasive and concomitant noninvasive intraoperative blood pressure monitoring:
Observed differences in measurements and associated therapeutic interventions. Anesthesiology 2011; 115:9738
(Accepted for publication September 3, 2014.)
In Reply:
We thank Fortin et al. for the letter to the editor related to
our recently published article.1 We fully agree with their statements and we would like to thank them for seconding our
main message: continuous noninvasive blood pressure monitoring systems are generating a lot of interest in our community and there is currently no accepted standard way to evaluate
the clinical performance of these systems. We, in our metaanalysis, used the Association for the Advancement of Medical Instrumentation standards2 only because it was cited by 15
of the 28 articles included. We fully acknowledged that this
is probably a misleading way to analyze these systems but we
wanted to stress this point to generate discussions and provoke
some changes/improvements. Even more interesting to us is
the fact that studies citing the Association for the Advancement
of Medical Instrumentation standards did not subsequently
follow them. Here again, we agree with Fortin et al.
We believe that the next actions should be undertaken to
move the field forward:
Anesthesiology 2015; 122:208-21 209 Correspondence
CORRESPONDENCE
zero-zones as defined by the AAMI (20 mmHg for sBP and 12

mmHg for dBP)3 results in the benchmarks being extended
by 10 to 1518 mmHg for sBP and by 6 to 1114 mmHg
for dBP. Would that still be a clinically acceptable bias?
None of the studies cited by Kim et al. applied this zerozone, but simply calculated accuracy as the average (more or
less beat-to-beat) difference between CNBP and IBP (for this
example 9 mmHg for sBP and 3 mmHg for dBP). It is obvious that this omission has a profound negative impact on the
reported accuracy and precision. In our view, the data should be,
where possible, reanalyzed in the manner outlined in the AAMI
standard. Comparison of the data with the AAMI benchmark
standard without such zero-zone analysis is not appropriate.
Secondly, the original AAMI approach requires patients to
be hemodynamically stable, so that the sphygmomanometer is
able to deliver plausible values. Maximum allowable ranges of
the zero-zones are exactly defined by AAMI and all data from
a subject shall be excluded if the reference zero-zone is greater
than the maximum ranges.3 Establishing the accuracy of
CNBP in hemodynamically unstable situations like induction
of anesthesia5,6 or even transfemoral aortic valve implantation
procedures7 provokes additional and challenging questions as
to what is the appropriate methodology and what statistics to
use? Including studies in the meta-analysis with data gained
from unstable patients or patients episodes may help explain
the high variation in accuracy and precision results reported.
As stated by Kim et al., their inclusion criteria may limit the
collective comparison of all studies in a single meta-analysis.
The need to continue to explore methodologies for the
testing and evaluation of all BP monitoring devices under
real-life clinical conditions is emphasized by the results
reported by Wax et al.8 They analyzed the difference between
sphygmomanometer and IBP readings from 24,225 patients
in daily clinical routine. The real-world average bias (SD)
was 1 mmHg (16 mmHg) for sBP and 5 mmHg (11
mmHg) for dBPwithout applying a zero-zone.
In conclusion, there is a growing clinical requirement for
the noninvasive monitoring of continuous BP, fluid, and stroke
volume. Such devices are now available to clinicians and are
being used in increasing numbers. It is therefore of major
importance to further discuss acceptable and reasonable continuous BP evaluation standards. A new evaluation standard
should be defined taking into consideration the inherent beatto-beat nature and trending capability of these new devices.
Competing Interests
CNSystems Medizintechnik AG (Graz, Austria) develops,
manufactures, and markets the continuous noninvasive
arterial pressure technology. LiDCO Ltd. (London, United
Kingdom) has integrated the continuous noninvasive arterial
pressure technology into their LiDCOrapid products. The
authors are inventors and named on one or more patents of
continuous noninvasive technology.
Drs. Fortin and OBrien are CEOs and founders of their
companies, receive salary, and have equity interests. Ms.Lerche
and Dr. Flotzinger are employees of CNSystems AG.
Jrgen Fortin, Ph.D., Katja Lerche, M.Sc., Doris Flotzinger, Ph.D., Terence OBrien, Ph.D. CNSystems Medizintechnik AG, Graz, Austria (J.F.). juergen.fortin@cnsystems.com
References

2. De Hert S: Noninvasive hemodynamic monitoring devices.
3. Non-invasive

2:
Clinical
Investigation of Automated Measurement Type, ANSI/AAMI/
ISO 810602:2013. Arlington, Association for the Advancement
of Medical Instrumentation, 2013, pp 122
4. Non-invasive

2:
Clinical
Validation of Automated Measurement Type, ANSI/AAMI/ISO
5. Ilies C, Bauer M, Berg P, Rosenberg J, Hedderich J, Bein B, Hinz
J, Hanss R: Investigation of the agreement of a continuous noninvasive arterial pressure device in comparison with invasive
radial artery measurement. Br J Anaesth 2012; 108:20210
6. Gayat E, Mongardon N, Tuil O, Sievert K, Chazot T, Liu N,
Fischler M: CNAP() does not reliably detect minimal or
maximal arterial blood pressures during induction of anaesthesia and tracheal intubation. Acta Anaesthesiol Scand 2013;
57:46873
7. Schramm C, Baat L, Plaschke K: Continuous noninvasive

arterial pressure: Assessment in older and high-risk patients
under analgesic sedation. Blood Press Monit 2011; 16:
2706
8. Wax DB, Lin HM, Leibowitz AB: Invasive and concomi
tant noninvasive intraoperative blood pressure monitoring:
Observed differences in measurements and associated therapeutic interventions. Anesthesiology 2011; 115:9738
In Reply:
We thank Fortin et al. for the letter to the editor related to
our recently published article.1 We fully agree with their statements and we would like to thank them for seconding our
main message: continuous noninvasive blood pressure monitoring systems are generating a lot of interest in our community and there is currently no accepted standard way to evaluate
the clinical performance of these systems. We, in our metaanalysis, used the Association for the Advancement of Medical Instrumentation standards2 only because it was cited by 15
of the 28 articles included. We fully acknowledged that this
is probably a misleading way to analyze these systems but we
wanted to stress this point to generate discussions and provoke
some changes/improvements. Even more interesting to us is
the fact that studies citing the Association for the Advancement
of Medical Instrumentation standards did not subsequently
follow them. Here again, we agree with Fortin et al.
We believe that the next actions should be undertaken to
move the field forward:
Correspondence
1.Standards are needed for the evaluation of continuous

noninvasive blood pressure monitoring systems.
2.These standards should probably define separate
benchmarks for systolic, diastolic, and mean arterial pressure. Because systolic, diastolic, and mean
arterial pressure are inherently different by nature,
different acceptability threshold should be applied.
3.These standards should probably develop a methodology for assessing the trending ability of these systems. Because these systems will be used as continuous
monitors in the clinical setting, their trending ability
is as important as their instantaneous accuracy.
4.When such standards exist, clinician scientists should
follow them carefully when conducting clinical studies testing these systems.
In conclusion, we would like to thank Fortin et al. for echoing our main message and reinforcing it. We truly believe that
the development and validation of this next generation of blood
pressure monitoring systems is promising, but will require close
collaboration between industry, clinician scientist, and regulatory agencies to make them beneficial to our patients.
Competing Interests
Maxime Cannesson is a consultant for Edwards Lifesciences
(Irvine, California), Covidien (Boulder, Colorado), Masimo
Corp. (Irvine, California), ConMed (Irvine, California), and
Philips Medical System (Suresnes, France). A Nexfin monitor (BMEYE B.V., Amsterdam, The Netherlands) and a CNAP
monitor (CNSystems, Graz, Austria) were loaned to Maxime
Cannesson and his research team in 2010. Maxime Cannesson publicly endorsed the Nexfin technology in a BMEYE
newsletter. The other authors declare no competing interests.
Maxime Cannesson, M.D., Ph.D., Joseph Rinehart,
M.D., Sang-Hyun Kim, M.D., Ph.D. University of California
Irvine Health, Irvine, California (M.C.). mcanness@uci.edu
References

2. Non-invasive sphygmomanometersPart 2: Clinical vali
dation of automated measurement type, ANSI/AAMI/ISO
Inotrope Use in Cardiac Surgery:

ACause of Worse Outcomes, or Just
a Marker of Patients Who Are at Risk?
To the Editor:
Although Dr. Nielsen et al.1 are to be commended for their
efforts to investigate the potential detrimental effects of
inotropic therapy in cardiac surgery, we believe that methodological problems significantly limit the validity of their
conclusions.
1. Insufficient information about the result of the matching
process is provided, but enough to indicate what appears
to be a significant flaw in methodology. First, their matching algorithm discarded a large number of both treatment
and control patients (n = 6,005 patients were identified to
be included in analysis; after propensity matching, only
n = 2,340 [39%] remained). This implies a considerable
lack of common support (overlap between the propensity
score distributions of the two cohorts), which, even in the
presence of a good match, increases risk of bias through
unmeasured confounders and makes the estimate of the
treatment effect unreliable.2 Second, the authors cite
Donald Rubin (coinventor of propensity score matching),
but use only one of the three metrics he recommends to
judge the quality of a match: absolute standardized difference. Neither the variance ratios of the propensity scores
between groups, nor the ratio of variance of the residuals of each covariate is reported.3,4 These are important,
because the match is vulnerable to systematic differences
in how the propensity scores were assigned. Finally, greedy
matching depends on the order of patients, so it should
be preceded by randomizing the order of patients in the
dataset, which the authors do not report.
2. These design decisions in the propensity matching algorithm leave the study open to the possibility that these
unmeasured confoundersand not the effect of inotropic
therapyare responsible for the observed outcome difference. Some variables were treated as overly simplistic
dichotomous variables, which fail to capture important
differences between patients, such that inotropic support
may continue to act as nothing more than a marker for
sicker patients with less well-functioning ventricles. Left
ventricular ejection fraction was treated as a binary variable: less than or equal to 30% or greater than 30%. Therefore, their propensity matching would not differentiate
between a patient with a baseline left ventricular ejection
fraction of 35% and one with a baseline left ventricular
ejection fraction of 65%. Duration of myocardial insult
was captured only by cardiopulmonary bypass (CPB)
time, again treated as a dichotomous variable (>120min
or 120min). There are two problems with this decision:
first, the need for inotropic support is more closely related
to the duration of myocardial ischemia, i.e., the aortic
cross-clamp time, than the time on bypass. Although
CPB time is correlated with cross-clamp time, different surgeons may adopt different temporal approaches
to weaning from bypass such that two surgeons with
the same cross-clamp time will have very different CPB
times. Indeed, the use of CPB time without cross-clamp
time would prevent differentiation of a patient who had
no aortic cross-clamp and no myocardial ischemia (e.g., a
Correspondence
1.Standards are needed for the evaluation of continuous

noninvasive blood pressure monitoring systems.
2.These standards should probably define separate
benchmarks for systolic, diastolic, and mean arterial pressure. Because systolic, diastolic, and mean
arterial pressure are inherently different by nature,
different acceptability threshold should be applied.
3.These standards should probably develop a methodology for assessing the trending ability of these systems. Because these systems will be used as continuous
monitors in the clinical setting, their trending ability
is as important as their instantaneous accuracy.
4.When such standards exist, clinician scientists should
follow them carefully when conducting clinical studies testing these systems.
In conclusion, we would like to thank Fortin et al. for echoing our main message and reinforcing it. We truly believe that
the development and validation of this next generation of blood
pressure monitoring systems is promising, but will require close
collaboration between industry, clinician scientist, and regulatory agencies to make them beneficial to our patients.
Competing Interests
Maxime Cannesson is a consultant for Edwards Lifesciences
(Irvine, California), Covidien (Boulder, Colorado), Masimo
Corp. (Irvine, California), ConMed (Irvine, California), and
Philips Medical System (Suresnes, France). A Nexfin monitor (BMEYE B.V., Amsterdam, The Netherlands) and a CNAP
monitor (CNSystems, Graz, Austria) were loaned to Maxime
Cannesson and his research team in 2010. Maxime Cannesson publicly endorsed the Nexfin technology in a BMEYE
newsletter. The other authors declare no competing interests.
Maxime Cannesson, M.D., Ph.D., Joseph Rinehart,
M.D., Sang-Hyun Kim, M.D., Ph.D. University of California
Irvine Health, Irvine, California (M.C.). mcanness@uci.edu
References

2. Non-invasive sphygmomanometersPart 2: Clinical vali
dation of automated measurement type, ANSI/AAMI/ISO
Inotrope Use in Cardiac Surgery:

ACause of Worse Outcomes, or Just
a Marker of Patients Who Are at Risk?
To the Editor:
Although Dr. Nielsen et al.1 are to be commended for their
efforts to investigate the potential detrimental effects of
inotropic therapy in cardiac surgery, we believe that methodological problems significantly limit the validity of their
conclusions.
1. Insufficient information about the result of the matching
process is provided, but enough to indicate what appears
to be a significant flaw in methodology. First, their matching algorithm discarded a large number of both treatment
and control patients (n = 6,005 patients were identified to
be included in analysis; after propensity matching, only
n = 2,340 [39%] remained). This implies a considerable
lack of common support (overlap between the propensity
score distributions of the two cohorts), which, even in the
presence of a good match, increases risk of bias through
unmeasured confounders and makes the estimate of the
treatment effect unreliable.2 Second, the authors cite
Donald Rubin (coinventor of propensity score matching),
but use only one of the three metrics he recommends to
judge the quality of a match: absolute standardized difference. Neither the variance ratios of the propensity scores
between groups, nor the ratio of variance of the residuals of each covariate is reported.3,4 These are important,
because the match is vulnerable to systematic differences
in how the propensity scores were assigned. Finally, greedy
matching depends on the order of patients, so it should
be preceded by randomizing the order of patients in the
dataset, which the authors do not report.
2. These design decisions in the propensity matching algorithm leave the study open to the possibility that these
unmeasured confoundersand not the effect of inotropic
therapyare responsible for the observed outcome difference. Some variables were treated as overly simplistic
dichotomous variables, which fail to capture important
differences between patients, such that inotropic support
may continue to act as nothing more than a marker for
sicker patients with less well-functioning ventricles. Left
ventricular ejection fraction was treated as a binary variable: less than or equal to 30% or greater than 30%. Therefore, their propensity matching would not differentiate
between a patient with a baseline left ventricular ejection
fraction of 35% and one with a baseline left ventricular
ejection fraction of 65%. Duration of myocardial insult
was captured only by cardiopulmonary bypass (CPB)
time, again treated as a dichotomous variable (>120min
or 120min). There are two problems with this decision:
first, the need for inotropic support is more closely related
to the duration of myocardial ischemia, i.e., the aortic
cross-clamp time, than the time on bypass. Although
CPB time is correlated with cross-clamp time, different surgeons may adopt different temporal approaches
to weaning from bypass such that two surgeons with
the same cross-clamp time will have very different CPB
times. Indeed, the use of CPB time without cross-clamp
time would prevent differentiation of a patient who had
no aortic cross-clamp and no myocardial ischemia (e.g., a
CORRESPONDENCE
right-sided procedure done with the heart perfused and

beating continuously) from one with the same duration of
bypass but with a cross-clamp and cardioplegia. Second,
the doseresponse pattern of a need for greater inotropic
support with longer periods of cross-clamping is unlikely
to be a simple threshold effect at 120min. The dichotomous treatment of CPB time is unable to distinguish, for
instance, between patients on bypass for 125min versus
325min. Linear or at least multilevel treatment of left
ventricular ejection fraction and cross-clamp time would
have improved the ability to adjust for the potential confounding effect of differences in baseline function and of
longer periods of myocardial ischemia.
3. The two cohorts are separated only by the presence or
absence of inotrope use; there is no ability to study the
doseresponse of low- versus high-dose inotropes, multiple inotropes, and so on. A subsequent logistic regression
in the inotrope group could have assessed the relationship
between inotrope dose (for instance, using a scoring system such as Vasoactive-Inotropic Score)5 and mortality.
Also, why was the use of norepinephrine excluded (resulting in 967 patients excluded from analysis)? It has positive
inotropic qualities in addition to being a vasopressor (as
the authors state), and in many institutions is the first-line
agent in heart surgery.
4. The larger design flaw in this study is that the retrospective
approach fails to compensate for the differences between
patients that would lead anesthesiologists and surgeons to
make the decision to use an inotrope in the first place. Propensity matching (even with better variable selection) probably cannot ever capture important variables that affect this
decision: the quality of cardioplegia and myocardial protection, the presence of air emboli to the coronary arteries
during or immediately after weaning from CPB, and most
importantly, the appearance (by gross visualization and
echocardiography) of ventricular function before weaning
from CPB. Although it is probably true that at least some
surgeons and anesthesiologists use inotropes routinely even
in patients who have no objective evidence that they need
them, the retrospective design does not isolate this subgroup.
The concomitant use of a vasodilator (to control blood pressure in the setting of hyperdynamic and/or hypertensive
physiology when an inotrope is added to an already wellfunctioning ventricle) might be a better marker for patients
who do not need the inotrope. That comparisoninotrope
plus vasodilator versus neitherwould be a more interesting guide for clinicians, as it could answer a more important question: does raising cardiac output (independent of
changing blood pressure) improve or worsen outcomes?
As is, the Nielsen study mostly can be said to demonstrate
that inotrope use is a marker for poor cardiac function after
bypass and hence worse outcomes (hardly surprising), and it
risks broadly discouraging the use of an important therapy
that is lifesaving in selected patients.
Competing Interests
Bryan G. Maxwell, M.D., M.P.H., Jack O. Wasey, B.M.,
B.Ch., Eugenie S. Heitmiller, M.D. Johns Hopkins
University School of Medicine, Baltimore, Maryland (B.G.M.).
bmaxwell@jhu.edu
References
1. Nielsen DV, Hansen MK, Johnsen SP, Hansen M, Hindsholm
K, Jakobsen CJ: Health outcomes with and without use of
inotropic therapy in cardiac surgery: Results of a propensity
score-matched analysis. Anesthesiology 2014; 120:1098108
2. Stuart EA: Matching methods for causal inference: A review
and a look forward. Stat Sci 2010; 25:121
3. Rubin DB: Using propensity scores to help design observational studies: Application to the tobacco litigation. Heal Serv
Outcomes Res Methodol 2001; 2:16988
4. Rubin DB: The design versus the analysis of observational
studies for causal effects: Parallels with the design of randomized trials. Stat Med 2007; 26:2036
5. Gaies MG, Gurney JG, Yen AH, Napoli ML, Gajarski RJ, Ohye
RG, Charpie JR, Hirsch JC: Vasoactive-inotropic score as a
predictor of morbidity and mortality in infants after cardiopulmonary bypass. Pediatr Crit Care Med 2010; 11:2348
In Reply:
We thank Dr. Maxwell et al. for their interest and comments
on our study,1 and we will try to address some of their concerns regarding the methodology that was used in the study
and our interpretation of the findings.
Dr. Maxwell et al. express concern about the propensity
scorebased matching process. We were able to match 56%
of the patients treated with inotropic therapy with a nontreated patient. The fact that it was not possible to match
100% of the patients indicates that the distribution of the
propensity score among the treated and nontreated patients
did not fully overlap. However, we do not agree that discarding patients in the matching process from the original
cohort per se implies that the internal validity of the study
is affected. It may, however, imply that concern should be
taken before extrapolating the findings to patients that differ
from the characteristics of our matched population. When
matching our patients, we assessed the balance using both
absolute standardized differences and variance ratios.
The variance ratios of the individual covariates in the
matched population ranged from a low of 0.86 (critical preoperative state) to high of 1.18 (off pump surgery; table1).
Along with the standardized differences of less than 10%, we
find strong indications of a well-balanced matching. Notably, the covariate postinfarct septal rupture had a very low
variance ration of 0.6, which is probably due to the very
few patients characterized by this covariate (five patient in
treated group and three with no inotropic therapy).
CORRESPONDENCE
right-sided procedure done with the heart perfused and

beating continuously) from one with the same duration of
bypass but with a cross-clamp and cardioplegia. Second,
the doseresponse pattern of a need for greater inotropic
support with longer periods of cross-clamping is unlikely
to be a simple threshold effect at 120min. The dichotomous treatment of CPB time is unable to distinguish, for
instance, between patients on bypass for 125min versus
325min. Linear or at least multilevel treatment of left
ventricular ejection fraction and cross-clamp time would
have improved the ability to adjust for the potential confounding effect of differences in baseline function and of
longer periods of myocardial ischemia.
3. The two cohorts are separated only by the presence or
absence of inotrope use; there is no ability to study the
doseresponse of low- versus high-dose inotropes, multiple inotropes, and so on. A subsequent logistic regression
in the inotrope group could have assessed the relationship
between inotrope dose (for instance, using a scoring system such as Vasoactive-Inotropic Score)5 and mortality.
Also, why was the use of norepinephrine excluded (resulting in 967 patients excluded from analysis)? It has positive
inotropic qualities in addition to being a vasopressor (as
the authors state), and in many institutions is the first-line
agent in heart surgery.
4. The larger design flaw in this study is that the retrospective
approach fails to compensate for the differences between
patients that would lead anesthesiologists and surgeons to
make the decision to use an inotrope in the first place. Propensity matching (even with better variable selection) probably cannot ever capture important variables that affect this
decision: the quality of cardioplegia and myocardial protection, the presence of air emboli to the coronary arteries
during or immediately after weaning from CPB, and most
importantly, the appearance (by gross visualization and
echocardiography) of ventricular function before weaning
from CPB. Although it is probably true that at least some
surgeons and anesthesiologists use inotropes routinely even
in patients who have no objective evidence that they need
them, the retrospective design does not isolate this subgroup.
The concomitant use of a vasodilator (to control blood pressure in the setting of hyperdynamic and/or hypertensive
physiology when an inotrope is added to an already wellfunctioning ventricle) might be a better marker for patients
who do not need the inotrope. That comparisoninotrope
plus vasodilator versus neitherwould be a more interesting guide for clinicians, as it could answer a more important question: does raising cardiac output (independent of
changing blood pressure) improve or worsen outcomes?
As is, the Nielsen study mostly can be said to demonstrate
that inotrope use is a marker for poor cardiac function after
bypass and hence worse outcomes (hardly surprising), and it
risks broadly discouraging the use of an important therapy
that is lifesaving in selected patients.
Competing Interests
Bryan G. Maxwell, M.D., M.P.H., Jack O. Wasey, B.M.,
B.Ch., Eugenie S. Heitmiller, M.D. Johns Hopkins
University School of Medicine, Baltimore, Maryland (B.G.M.).
bmaxwell@jhu.edu
References
2. Stuart EA: Matching methods for causal inference: A review
and a look forward. Stat Sci 2010; 25:121
3. Rubin DB: Using propensity scores to help design observational studies: Application to the tobacco litigation. Heal Serv
Outcomes Res Methodol 2001; 2:16988
4. Rubin DB: The design versus the analysis of observational
studies for causal effects: Parallels with the design of randomized trials. Stat Med 2007; 26:2036
5. Gaies MG, Gurney JG, Yen AH, Napoli ML, Gajarski RJ, Ohye
RG, Charpie JR, Hirsch JC: Vasoactive-inotropic score as a
predictor of morbidity and mortality in infants after cardiopulmonary bypass. Pediatr Crit Care Med 2010; 11:2348
In Reply:
We thank Dr. Maxwell et al. for their interest and comments
on our study,1 and we will try to address some of their concerns regarding the methodology that was used in the study
and our interpretation of the findings.
Dr. Maxwell et al. express concern about the propensity
scorebased matching process. We were able to match 56%
of the patients treated with inotropic therapy with a nontreated patient. The fact that it was not possible to match
100% of the patients indicates that the distribution of the
propensity score among the treated and nontreated patients
did not fully overlap. However, we do not agree that discarding patients in the matching process from the original
cohort per se implies that the internal validity of the study
is affected. It may, however, imply that concern should be
taken before extrapolating the findings to patients that differ
from the characteristics of our matched population. When
matching our patients, we assessed the balance using both
absolute standardized differences and variance ratios.
The variance ratios of the individual covariates in the
matched population ranged from a low of 0.86 (critical preoperative state) to high of 1.18 (off pump surgery; table1).
Along with the standardized differences of less than 10%, we
find strong indications of a well-balanced matching. Notably, the covariate postinfarct septal rupture had a very low
variance ration of 0.6, which is probably due to the very
few patients characterized by this covariate (five patient in
treated group and three with no inotropic therapy).
Correspondence
Table 1. Variance Ratios in Propensity Matched Sample

Covariates
Variance Ratio
Demographics
Age (yr)
Females
EuroSCORE
Patient-related EuroSCORE variables
Chronic pulmonary disease
Extra cardiac arteriopathy
Neurologic dysfunction disease
Previous cardiac surgery
Serum creatinine >200 mol/l
Active endocarditis
Critical preoperative state
Other patient-related variables
Preoperative arrhythmia
Preoperative RRT
Cardiac-related EuroSCORE variables
Unstable angina
Recent myocardial infarction
Pulmonary hypertension
LVEF 30%
Procedure-related EuroSCORE variables
Emergency surgery
CABG only
Thoracic aortic surgery
Postinfarct septal rupture
Other procedure-related variables
Intravenous anesthesia
Epidural supplement
CPB time >120 min
Off-pump surgery
Cardiac center
Center A
Center B
Center C
1.01
1.01
1.00
0.98
0.99
0.93
0.91
1.05
0.87
0.86
1.00
0.87
0.86
1.05
1.00
0.91
0.89
0.98
0.99
0.60
0.99
1.09
1.01
1.18
1.00
1.00
1.03
CABG = coronary artery bypass grafting; CPB = cardiopulmonary bypass;

LVEF = left ventricular ejection fraction; RRT = renal replacement therapy.
Table 2. One-year Mortality by Left Ventricular Function and

Treatment Group
EuroSCORE
EF Definition
Control
Inotropes
LVEF 50
LVEF 3149
LVEF 30
P value
27 (3.76%)
19 (4.85%)
3 (5.00%)
0.6538
80 (11.10%)
45 (11.90%)
8 (11.27%)
0.9223
EF = ejection fraction; LVEF = left ventricular ejection fraction.
Furthermore, the order of the patients was randomized

before matching although this was not specified in the article.
As we already discussed in our article, we were not able
to control for residual confounding relating to intraoperative
events not accounted for by procedure scoring.
We agree that extracorporeal circulation (ECC) time and
perhaps especially cross-clamp (CC) time have some impact
on whether patients are treated with inotropes and with which

dose. We also agree that there might be a correlation between
ECC/CC times and mortality. Thus, it is possible that the
dichotomization with on/off bypass and then ECC time
greater than/less than 120 min was too tight. However, we
could not demonstrate any statistically significant difference
in the mortality when the treated and nontreated was handled
individually according to ECC time in either 30- or 60-min
intervals. Including CC time in a conditional linear regression
analysis on the matched cohort reduced the risk estimate of
30-day mortality from adjusted hazard ratio 3.71 (2.116.53,
95%CI) to a hazard ratio 2.39 (1.573.63, 95% CI).
The authors likewise express concern about our dichotomization of ventricular function into less than or equal to
30% and greater than 30%. It could be argued that a more
detailed categorization of patients with normal left ventricular function would be relevant.
However as table2 indicates, there was no significant difference in mortality between different states of left ventricular function but a constant difference between treated and
nontreated patients.
As we wanted to primarily investigate the effect of inotropic therapy, we excluded the patients who had exclusively
received vasopressors and we think it is well established that
norepinephrine has only minimal inotropic effect, as it is primarily an -receptor agonist and only have very little, if any,
direct -effect.
Dr. Maxwell and colleagues suggest studying the doseresponse of inotropic therapy by a vasoactive-active scoring
system. To our knowledge the scoring system in concern has
been validated only on a neonatal and infant population,
and we do not find it appropriate to apply on an adult population. Our data were not granular enough to distinguish
whether inotropes were given in parallel or sequential therapy, but we have future studies on a larger population size
that will focus on the association between specific inotropes,
dosages, and outcome.
The problem of capturing the anesthesiologists gut
feeling will exist in both propensity score matching studies and in a randomized clinical trial, unless a very fixed
protocol. We know that initiation of inotropes is highly
dependent on the name of the anesthetist,2 thus we tried
to adjust for the provider effect in the conditional regression analysis.
Although an observational study as the present always
will be challenged with the risk of confounding by indication, we believe that the design and statistical analyses of
our study are enough robust to interpret data as to raise
concern about a possible harmful effect of inotropic therapy in cardiac surgery. The concern of a possible risk that
could exceed the beneficial effect of inotropic therapy to
some patients is not novel. There is a growing body of
literature, especially from large-scale randomized trials of
nonsurgical heart failure patients that indicate that inotropic therapy may cause more harm than good. In the light
CORRESPONDENCE
of the lack of randomized clinical trials demonstrating any

improvement of clinical outcomes including mortality
from perioperative inotropic therapy,35 we are restraint to
ignore or explain away the underlying signal our dataraise.
Competing Interests
Dorthe Viemose Nielsen, M.D., Sren Paaske Johnsen,
M.D., Ph.D., Malene Krslund Hansen, M.B.B.S., CarlJohan Jakobsen, M.D. Aarhus University Hospital, Aarhus,
Denmark (D.V.N.). dorthe.viemose.nielsen@skejby.rm.dk
References
2. Nielsen DV, Johnsen SP, Madsen M, Jakobsen CJ: Variation in use
of peroperative inotropic support therapy in cardiac surgery:
Time for reflection? Acta Anaesthesiol Scand 2011; 55:3528
3. Cuffe MS, Califf RM, Adams KF Jr, Benza R, Bourge R,

Colucci WS, Massie BM, OConnor CM, Pina I, Quigg R,
Silver MA, Gheorghiade M; Outcomes of a Prospective Trial
of Intravenous Milrinone for Exacerbations of Chronic Heart
Failure (OPTIME-CHF) Investigators: Short-term intravenous
milrinone for acute exacerbation of chronic heart failure: A
4. Felker GM, Benza RL, Chandler AB, Leimberger JD, Cuffe
MS, Califf RM, Gheorghiade M, OConnor CM; OPTIMECHF Investigators: Heart failure etiology and response to
milrinone in decompensated heart failure: Results from the
OPTIME-CHF study. J Am Coll Cardiol 2003; 41:9971003
5. Amsallem E: Phosphodiesterase III inhibitors for heart failure. Cochrane Database Syst Rev 2005:CD002230
Lung Ultrasonography for the

Detection of Anesthesia-induced Lung
Atelectasis
To the Editor:
We have read with great interest the article by Acosta etal.1
exploring the use of lung ultrasound as a mean to detect
intraoperative atelectasis. However, despite their statement
to the contrary, the occurrence of B lines in the setting of
atelectasis has already been described by others. Although B
lines were initially thought to originate from the interaction
of the ultrasound beam with thickened subpleural interlobular septa found in alveolar-interstitial pathologies,2 recent
work has challenged this hypothesis. In an elegant series of
experiments, Soldati et al.35 have shown that B lines are
observed when the ultrasound beam interacts at the pleural surface with lung tissue of a specific density. Although
this can occur with the replacement of subpleural air by an
ultrasound-conductive substance (e.g., water, pus, blood,

and fibrous tissue), the withdrawal of air (e.g., resorption
atelectasis) will also lead to the genesis of B lines. Demonstrating this last point, in an ex vivo animal model of graded
atelectasis, B lines were observed in increasing numbers with
increasing atelectasis. Pathologic examination of the excised
lungs showed diffusely compressed alveoli mixed with sporadic areas of normally expanded distal air spaces.5
The study by Acosta et al. comes at an interesting moment.
The recent publication of two randomized controlled trials6,7
exploring the impact of intraoperative mechanical ventilation
parameters on postoperative pulmonary complications has
generated much interest.8 Although some have linked the negative results of the PROVHILO (PROtective Ventilation using
HIgh versus LOw positive end-expiratory pressure) trial to a
lack of regular recruitment maneuvers and a positive end-expiratory pressure set too high,9 others blame the use of inappropriately high tidal volumes in the control group for the positive
results in the IMPROVE (Intraoperative PROtective VEntilation) trial.10 In the absence of imagery supporting claims of
atelectasis or overdistention, the culprits usually blamed for
the development of postoperative pulmonary complications,
it is unlikely this question will be resolved before more data
becomes available. Interestingly, recent anesthesiology literature has demonstrated the advantage of hemodynamic optimization11 championing the concept that individualization is
preferable to a one size fits all approach. Likewise, individualization of mechanical ventilation parameters might be an interesting avenue to explore if we wish to decrease the occurrence
of postoperative respiratory complications. This hypothesis is
supported by spiral computed tomography studies reporting
significant interpatient variability in the amount of atelectasis
induced by general anesthesia.12,13 Therefore, we believe that
bedside monitoring to detect lung atelectasis or overdistention
is needed. Although magnetic resonance imaging and computed tomography fulfill this requirement, they cannot be used
in an intraoperative setting except in specially designed operating rooms and could not realistically be repeated throughout a
procedure. Because lung ultrasonography can be performed at
the bedside and is devoid of any ionizing radiation, the present
study by Acosta et al., although interesting in and of itself, is
an important milestone toward establishing lung ultrasonography as a tool to optimize intraoperative mechanical ventilation
parameters. Other investigators have described loss of aeration
scales that have allowed the study of the therapeutic effects of
antibiotics in ventilator-associated pneumonia,14 the effect of
different levels of positive end-expiratory pressure in patients
with acute respiratory distress syndrome on lung reexpansion15
and the detection of patients likely to fail extubation after a
successful spontaneous breathing trial.16 Although not developed specifically for the diagnosis and monitoring of anesthesia-induced atelectasis, the use of these scales would have been
an interesting addition to the study by Acosta etal. Whether
to optimize intraoperative mechanical ventilation parameters
or to assist anesthesiologists in the care of hypoxemic patients,
CORRESPONDENCE
of the lack of randomized clinical trials demonstrating any

improvement of clinical outcomes including mortality
from perioperative inotropic therapy,35 we are restraint to
ignore or explain away the underlying signal our dataraise.
Competing Interests
Dorthe Viemose Nielsen, M.D., Sren Paaske Johnsen,
M.D., Ph.D., Malene Krslund Hansen, M.B.B.S., CarlJohan Jakobsen, M.D. Aarhus University Hospital, Aarhus,
Denmark (D.V.N.). dorthe.viemose.nielsen@skejby.rm.dk
References
2. Nielsen DV, Johnsen SP, Madsen M, Jakobsen CJ: Variation in use
of peroperative inotropic support therapy in cardiac surgery:
Time for reflection? Acta Anaesthesiol Scand 2011; 55:3528
3. Cuffe MS, Califf RM, Adams KF Jr, Benza R, Bourge R,

Colucci WS, Massie BM, OConnor CM, Pina I, Quigg R,
Silver MA, Gheorghiade M; Outcomes of a Prospective Trial
of Intravenous Milrinone for Exacerbations of Chronic Heart
Failure (OPTIME-CHF) Investigators: Short-term intravenous
milrinone for acute exacerbation of chronic heart failure: A
4. Felker GM, Benza RL, Chandler AB, Leimberger JD, Cuffe
MS, Califf RM, Gheorghiade M, OConnor CM; OPTIMECHF Investigators: Heart failure etiology and response to
milrinone in decompensated heart failure: Results from the
OPTIME-CHF study. J Am Coll Cardiol 2003; 41:9971003
5. Amsallem E: Phosphodiesterase III inhibitors for heart failure. Cochrane Database Syst Rev 2005:CD002230
Lung Ultrasonography for the

Detection of Anesthesia-induced Lung
Atelectasis
To the Editor:
We have read with great interest the article by Acosta etal.1
exploring the use of lung ultrasound as a mean to detect
intraoperative atelectasis. However, despite their statement
to the contrary, the occurrence of B lines in the setting of
atelectasis has already been described by others. Although B
lines were initially thought to originate from the interaction
of the ultrasound beam with thickened subpleural interlobular septa found in alveolar-interstitial pathologies,2 recent
work has challenged this hypothesis. In an elegant series of
experiments, Soldati et al.35 have shown that B lines are
observed when the ultrasound beam interacts at the pleural surface with lung tissue of a specific density. Although
this can occur with the replacement of subpleural air by an
ultrasound-conductive substance (e.g., water, pus, blood,

and fibrous tissue), the withdrawal of air (e.g., resorption
atelectasis) will also lead to the genesis of B lines. Demonstrating this last point, in an ex vivo animal model of graded
atelectasis, B lines were observed in increasing numbers with
increasing atelectasis. Pathologic examination of the excised
lungs showed diffusely compressed alveoli mixed with sporadic areas of normally expanded distal air spaces.5
The study by Acosta et al. comes at an interesting moment.
The recent publication of two randomized controlled trials6,7
exploring the impact of intraoperative mechanical ventilation
parameters on postoperative pulmonary complications has
generated much interest.8 Although some have linked the negative results of the PROVHILO (PROtective Ventilation using
HIgh versus LOw positive end-expiratory pressure) trial to a
lack of regular recruitment maneuvers and a positive end-expiratory pressure set too high,9 others blame the use of inappropriately high tidal volumes in the control group for the positive
results in the IMPROVE (Intraoperative PROtective VEntilation) trial.10 In the absence of imagery supporting claims of
atelectasis or overdistention, the culprits usually blamed for
the development of postoperative pulmonary complications,
it is unlikely this question will be resolved before more data
becomes available. Interestingly, recent anesthesiology literature has demonstrated the advantage of hemodynamic optimization11 championing the concept that individualization is
preferable to a one size fits all approach. Likewise, individualization of mechanical ventilation parameters might be an interesting avenue to explore if we wish to decrease the occurrence
of postoperative respiratory complications. This hypothesis is
supported by spiral computed tomography studies reporting
significant interpatient variability in the amount of atelectasis
induced by general anesthesia.12,13 Therefore, we believe that
bedside monitoring to detect lung atelectasis or overdistention
is needed. Although magnetic resonance imaging and computed tomography fulfill this requirement, they cannot be used
in an intraoperative setting except in specially designed operating rooms and could not realistically be repeated throughout a
procedure. Because lung ultrasonography can be performed at
the bedside and is devoid of any ionizing radiation, the present
study by Acosta et al., although interesting in and of itself, is
an important milestone toward establishing lung ultrasonography as a tool to optimize intraoperative mechanical ventilation
parameters. Other investigators have described loss of aeration
scales that have allowed the study of the therapeutic effects of
antibiotics in ventilator-associated pneumonia,14 the effect of
different levels of positive end-expiratory pressure in patients
with acute respiratory distress syndrome on lung reexpansion15
and the detection of patients likely to fail extubation after a
successful spontaneous breathing trial.16 Although not developed specifically for the diagnosis and monitoring of anesthesia-induced atelectasis, the use of these scales would have been
an interesting addition to the study by Acosta etal. Whether
to optimize intraoperative mechanical ventilation parameters
or to assist anesthesiologists in the care of hypoxemic patients,
Correspondence
lung ultrasonography is likely to have a bright future in our

operating rooms.
Acknowledgments
Supported by Grant of the Fondation 2013, Fondation danesthsiologie et ranimation du Qubec; Grant
of the Fonds de dveloppement 2014, Dpartement
danesthsiologie, Universit de Montral.
Competing Interests
Martin Girard, M.D., Vincent Gnreux, M.D., Audrey
Monastesse, M.D. Centre Hospitalier de lUniversit de
Montral, Montral, Qubec, Canada (M.G.). martin.girard@
umontreal.ca
References
1. Acosta CM, Maidana GA, Jacovitti D, Belaunzarn A, Cereceda
S, Rae E, Molina A, Gonorazky S, Bohm SH, Tusman G:
Accuracy of transthoracic lung ultrasound for diagnosing
anesthesia-induced atelectasis in children. Anesthesiology
2014; 120:13709
2. Lichtenstein D, Mzire G, Biderman P, Gepner A, Barr O:
The comet-tail artifact. An ultrasound sign of alveolar-interstitial syndrome. Am J Respir Crit Care Med 1997; 156:16406
3. Soldati G, Copetti R, Sher S: Sonographic interstitial syndrome:
The sound of lung water. J Ultrasound Med 2009; 28:16374
4. Soldati G, Giunta V, Sher S, Melosi F, Dini C: Synthetic comets: A new look at lung sonography. Ultrasound Med Biol
2011; 37:176270
5. Soldati G, Inchingolo R, Smargiassi A, Sher S, Nenna

R, Inchingolo CD, Valente S: Ex vivo lung sonography:
Morphologic-ultrasound relationship. Ultrasound Med Biol
2012; 38:116979
6. Futier E, Constantin JM, Paugam-Burtz C, Pascal J, Eurin M,
Neuschwander A, Marret E, Beaussier M, Gutton C, Lefrant
JY, Allaouchiche B, Verzilli D, Leone M, De Jong A, Bazin JE,
Pereira B, Jaber S; IMPROVE Study Group: A trial of intraoperative low-tidal-volume ventilation in abdominal surgery. N
Engl J Med 2013; 369:42837
7. The PROVE Network Investigators, for the Clinical Trial Network
of the European Society of Anaesthesiology: High versus low
positive end-expiratory pressure during general anaesthesia for
open abdominal surgery (PROVHILO trial): A multicentre randomised controlled trial. Lancet 2014; 384:495503
8. Goldenberg NM, Steinberg BE, Lee WL, Wijeysundera DN,
Kavanagh BP: Lung-protective ventilation in the operating
room: Time to implement? Anesthesiology 2014; 121:1848
9. Futier E: Positive end-expiratory pressure in surgery: Good
or bad? Lancet 2014; 384:4724
10. Wanderer JP, Blum JM, Ehrenfeld JM: Intraoperative

low-tidal-volume ventilation. N Engl J Med 2013; 369:1861
11. Corcoran T, Rhodes JE, Clarke S, Myles PS, Ho KM:

Perioperative fluid management strategies in major surgery:
A stratified meta-analysis. Anesth Analg 2012; 114:64051
12. Andersson LE, Bth M, Thrne A, Aspelin P, Odeberg
Wernerman S: Effect of carbon dioxide pneumoperitoneum
on development of atelectasis during anesthesia, examined by
spiral computed tomography. Anesthesiology 2005; 102:2939
13. Reinius H, Jonsson L, Gustafsson S, Sundbom M, Duvernoy
O, Pelosi P, Hedenstierna G, Fredn F: Prevention of atelectasis in morbidly obese patients during general anesthesia and
paralysis: A computerized tomography study. Anesthesiology
2009; 111:97987
14. Bouhemad B, Liu ZH, Arbelot C, Zhang M, Ferarri F, Le-Guen

M, Girard M, Lu Q, Rouby JJ: Ultrasound assessment of antibiotic-induced pulmonary reaeration in ventilator-associated
pneumonia. Crit Care Med 2010; 38:8492
15. Bouhemad B, Brisson H, Le-Guen M, Arbelot C, Lu Q, Rouby
JJ: Bedside ultrasound assessment of positive end-expiratory
pressure-induced lung recruitment. Am J Respir Crit Care
Med 2011; 183:3417
16. Soummer A, Perbet S, Brisson H, Arbelot C, Constantin JM,
Lu Q, Rouby JJ; Lung Ultrasound Study Group: Ultrasound
assessment of lung aeration loss during a successful weaning
trial predicts postextubation distress*. Crit Care Med 2012;
40:206472
In Reply:
Thank you very much for giving us the opportunity to reply
the letter by Dr. Girard et al. about our recent paper in Anes1
thesiology. In their interesting letter, Dr. Girard etal. describe
different theories about the genesis of B lines in the setting
of atelectasis and commented that this very lung sonography (LUS) sign has already been described. Their argument
is based on publications related to patients with pre-existing
pulmonary diseases and on data derived from ex-vivo animal
and laboratory models.25 As practicing anesthesiologists who
simply apply LUS as a diagnostic tool, we focused our literature search primarily on clinical studies employing LUS and
thereby may have missed important evidence coming from
primary ultrasound research, however, we do not only agree
with the criticism but are grateful to the authors for having
raised our awareness for the complexity of LUS.
To our knowledge, the occurrence of anesthesia-induced
atelectasis in children has never before been studied by LUS in
detail. For this reason, we cannot infer with certainty that the
LUS signsincluding B linesfound in adults and in atelectasis of different origins are similar to or even identical with the
ones we saw in anesthesia-induced atelectasis by compressive
mechanism in our children. This lack of reliable information
made us define anesthesia-induced atelectasis a posteriori and
analyze the prevalence of LUS signs associated with such atelectasis (please see table 11). This is the reason why we presented
our resultsincluding those related to B linesas novel contributions to the clinical understanding and diagnosis of atelectasis in children undergoing general anesthesia.
In the second part of their letter, Dr. Girard et al. highlight the role anesthesia-induced atelectasis may play in
creating local inflammatory responses within the lungs and
in causing postoperative pulmonary complications.6 Such
lung inflammation appears any time cyclic ventilation is
applied to a partially collapsed lung, the root cause being
tidal recruitment (the opening and closing of an atelectatic
area during the breathing cycle) and tidal overdistension
(the excess volume or pressure that normally aerated areas
Correspondence
lung ultrasonography is likely to have a bright future in our

operating rooms.
Acknowledgments
Supported by Grant of the Fondation 2013, Fondation danesthsiologie et ranimation du Qubec; Grant
of the Fonds de dveloppement 2014, Dpartement
danesthsiologie, Universit de Montral.
Competing Interests
Martin Girard, M.D., Vincent Gnreux, M.D., Audrey
Monastesse, M.D. Centre Hospitalier de lUniversit de
Montral, Montral, Qubec, Canada (M.G.). martin.girard@
umontreal.ca
References
2014; 120:13709
3. Soldati G, Copetti R, Sher S: Sonographic interstitial syndrome:
The sound of lung water. J Ultrasound Med 2009; 28:16374
2011; 37:176270

2012; 38:116979
Engl J Med 2013; 369:42837
7. The PROVE Network Investigators, for the Clinical Trial Network
of the European Society of Anaesthesiology: High versus low
positive end-expiratory pressure during general anaesthesia for
open abdominal surgery (PROVHILO trial): A multicentre randomised controlled trial. Lancet 2014; 384:495503
8. Goldenberg NM, Steinberg BE, Lee WL, Wijeysundera DN,
Kavanagh BP: Lung-protective ventilation in the operating
room: Time to implement? Anesthesiology 2014; 121:1848
9. Futier E: Positive end-expiratory pressure in surgery: Good
or bad? Lancet 2014; 384:4724
10. Wanderer JP, Blum JM, Ehrenfeld JM: Intraoperative

low-tidal-volume ventilation. N Engl J Med 2013; 369:1861
11. Corcoran T, Rhodes JE, Clarke S, Myles PS, Ho KM:

Perioperative fluid management strategies in major surgery:
A stratified meta-analysis. Anesth Analg 2012; 114:64051
12. Andersson LE, Bth M, Thrne A, Aspelin P, Odeberg
Wernerman S: Effect of carbon dioxide pneumoperitoneum
on development of atelectasis during anesthesia, examined by
spiral computed tomography. Anesthesiology 2005; 102:2939
13. Reinius H, Jonsson L, Gustafsson S, Sundbom M, Duvernoy
O, Pelosi P, Hedenstierna G, Fredn F: Prevention of atelectasis in morbidly obese patients during general anesthesia and
paralysis: A computerized tomography study. Anesthesiology
2009; 111:97987
14. Bouhemad B, Liu ZH, Arbelot C, Zhang M, Ferarri F, Le-Guen

M, Girard M, Lu Q, Rouby JJ: Ultrasound assessment of antibiotic-induced pulmonary reaeration in ventilator-associated
pneumonia. Crit Care Med 2010; 38:8492
15. Bouhemad B, Brisson H, Le-Guen M, Arbelot C, Lu Q, Rouby
JJ: Bedside ultrasound assessment of positive end-expiratory
pressure-induced lung recruitment. Am J Respir Crit Care
Med 2011; 183:3417
16. Soummer A, Perbet S, Brisson H, Arbelot C, Constantin JM,
Lu Q, Rouby JJ; Lung Ultrasound Study Group: Ultrasound
assessment of lung aeration loss during a successful weaning
trial predicts postextubation distress*. Crit Care Med 2012;
40:206472
In Reply:
Thank you very much for giving us the opportunity to reply
the letter by Dr. Girard et al. about our recent paper in Anes1
thesiology. In their interesting letter, Dr. Girard etal. describe
different theories about the genesis of B lines in the setting
of atelectasis and commented that this very lung sonography (LUS) sign has already been described. Their argument
is based on publications related to patients with pre-existing
pulmonary diseases and on data derived from ex-vivo animal
and laboratory models.25 As practicing anesthesiologists who
simply apply LUS as a diagnostic tool, we focused our literature search primarily on clinical studies employing LUS and
thereby may have missed important evidence coming from
primary ultrasound research, however, we do not only agree
with the criticism but are grateful to the authors for having
raised our awareness for the complexity of LUS.
To our knowledge, the occurrence of anesthesia-induced
atelectasis in children has never before been studied by LUS in
detail. For this reason, we cannot infer with certainty that the
LUS signsincluding B linesfound in adults and in atelectasis of different origins are similar to or even identical with the
ones we saw in anesthesia-induced atelectasis by compressive
mechanism in our children. This lack of reliable information
made us define anesthesia-induced atelectasis a posteriori and
analyze the prevalence of LUS signs associated with such atelectasis (please see table 11). This is the reason why we presented
our resultsincluding those related to B linesas novel contributions to the clinical understanding and diagnosis of atelectasis in children undergoing general anesthesia.
In the second part of their letter, Dr. Girard et al. highlight the role anesthesia-induced atelectasis may play in
creating local inflammatory responses within the lungs and
in causing postoperative pulmonary complications.6 Such
lung inflammation appears any time cyclic ventilation is
applied to a partially collapsed lung, the root cause being
tidal recruitment (the opening and closing of an atelectatic
area during the breathing cycle) and tidal overdistension
(the excess volume or pressure that normally aerated areas
CORRESPONDENCE
receive during inspiration).69 This sequence of events calls

for the use of protective ventilator settings also during anesthesia if such kind of lung injury was to be prevented.1014
However, even in the light of recent studies, our knowledge
on how to best implement protective ventilation strategies is
scarce.15,16 These latest attempts to provide convincing scientific support for the hypothesis that particular ventilator
settings (a combination of high vs. low tidal volume or high
vs. low positive end-expiratory pressure, with and without
recruitment maneuvers) would have proven effects on both,
the mechanisms of lung injury during anesthesia and on
postoperative pulmonary complications failed.
Furthermore, Dr. Girard et al. suggested that the link
between atelectasis and postoperative pulmonary complications should be established by an imaging tool capable of
detecting atelectasis. We totally agree with this statement.
Although the presence of atelectasis might be suspected
when respiratory mechanics and gas exchange are suboptimal, in vivo it can only be diagnosed with 100% certainty by imaging means. Therefore, the cited studies fail
to demonstrate a causative relationship between particular
ventilator settings and atelectasis, let alone postoperative
pulmonary complications.
Therefore, before we could address such clinically relevant hypotheses, we first had to validate LUS by magnetic
resonance imaging as a noninvasive reliable tool to detect
atelectasisat least in children. Now that LUS has demonstrated its high sensitivity and specificity for diagnosing
atelectasis, we have a good chance to reveal tidal recruitment as the main mechanism of ventilator-associated lung
injury in partially collapsed lungs. However, due to methodological reasons, LUS will fail as a diagnostic tool when the
lungs become overdistended. To detect this other important
injurious condition the monitoring of dead space by way of
volumetric capnography is a valid option.17,18 Thus, today
at least two noninvasive bedside methods are available to
detect the main mechanisms of ventilator-induced lung
injury. We believe it is about time to start using them in our
daily practice for the benefit of our patients.
Competing Interests
Gerardo Tusman, M.D., Cecilia M. Acosta, M.D.,
Stephan H Bohm, M.D. Hospital Privado de Comunidad,
Mar del Plata, Argentina (G.T.). gtusman@hotmail.com
References
2014; 120:13709
3. Soldati G, Copetti R, Sher S: Sonographic interstitial syn

drome: The sound of lung water. J Ultrasound Med 2009;
28:16374
2011; 37:176270

2012; 38:116979
6. Tusman G, Bhm SH, Warner DO, Sprung J: Atelectasis and
perioperative pulmonary complications in high-risk patients.
Curr Opin Anaesthesiol 2012; 25:110
7. Shennib H, Mulder DS, Chiu RC: The effects of pulmo
nary atelectasis and reexpansion on lung cellular immune
defenses. Arch Surg 1984; 119:2747
8. Carney D, DiRocco J, Nieman G: Dynamic alveolar mechanics and ventilator-induced lung injury. Crit Care Med 2005;
33(suppl 3):S1228
9. Zupancich E, Paparella D, Turani F, Munch C, Rossi A,

Massaccesi S, Ranieri VM: Mechanical ventilation affects
inflammatory mediators in patients undergoing cardiopulmonary bypass for cardiac surgery: A randomized clinical
trial. J Thorac Cardiovasc Surg 2005; 130:37883
10. Michelet P, DJourno XB, Roch A, Doddoli C, Marin V,

Papazian L, Decamps I, Bregeon F, Thomas P, Auffray
JP: Protective ventilation influences systemic inflammation after esophagectomy: A randomized controlled study.
11. Wolthuis EK, Choi G, Dessing MC, Bresser P, Lutter R, Dzoljic
M, van der Poll T, Vroom MB, Hollmann M, Schultz MJ:
Mechanical ventilation with lower tidal volumes and positive
end-expiratory pressure prevents pulmonary inflammation
in patients without preexisting lung injury. Anesthesiology
2008; 108:4654
12. Licker M, Diaper J, Villiger Y, Spiliopoulos A, Licker V, Robert
J, Tschopp JM: Impact of intraoperative lung-protective interventions in patients undergoing lung cancer surgery. Crit
Care 2009; 13:R41
13. Hemmes SN, Serpa Neto A, Schultz MJ: Intraoperative ventilatory strategies to prevent postoperative pulmonary complications: A meta-analysis. Curr Opin Anaesthesiol 2013;
26:12633
1 4. Goldenberg NM, Steinberg BE, Lee WL, Wijeysundera DN,
Kavanagh BP: Lung-protective ventilation in the operating room: Time to implement? Anesthesiology 2014;
121:1848
Engl J Med 2013; 369:42837
16. The PROVE Network Investigators, for the Clinical Trial

Network of the European Society of Anaesthesiology: High
versus low positive end-expiratory pressure during general
anaesthesia for open abdominal surgery (PROVHILO trial):
A multicentre randomised controlled trial. Lancet 2014;
384:495503
17. Tusman G, Sipmann FS, Borges JB, Hedenstierna G, Bohm
SH: Validation of Bohr dead space measured by volumetric
capnography. Intensive Care Med 2011; 37:8704
18. Tusman G, Sipmann FS, Bohm SH: Rationale of dead space
measurement by volumetric capnography. Anesth Analg
2012; 114:86674
Correspondence
Early Childhood Anesthetic

Neurotoxicity and Unmeasured
Covariates: Theres the RUB
To the Editor:
In retrospective investigations that use cognitive and intellectual test data as endpoints, the potential for a wide variety
of medical and sociologic covariates to influence the results
is well recognized. Investigations of the effect of anesthetic
exposure during infancy have invariably attempted to take
these potential covariates (including birth weight, prematurity, parental presence and education, gender, the disease
process that led to the anesthetic, and coincident disease
processes) into account. The thoroughness of that process is critical to any interpretation or extrapolation of the
conclusions.
In that light, I write to request additional information
about the 112-child subset of the Raine cohort who were the
subject of the recent report by Ing et al.1 Their investigation
suggested that even very brief anesthetic exposure in the first
3 yr of life results in adverse effects on the development of
language skills.
First, I request that the authors provide a table of the surgical procedures, analogous to that provided in their original
article about the Raine cohort.2 Many of the children in that
first cohort underwent upper airway procedures (mygingotomy and tubes, tonsillectomy and adenoidectomy, cleft lip
and palate repair, tracheostomy) that might well have some
relevance to the acquisition of language skills. At a minimum, chronic middle ear effusions are known to be associated with impairments of receptive language and verbal
aspects of cognition.3 Furthermore, prospective study of
children with middle ear effusions has failed to demonstrate
subsequent differences in language skills between those who
do and do not undergo mygingotomy and tubes, suggesting that the adverse effect of middle ear effusions cannot be
assumed to have been rectified by their mygingotomy and
tubes procedures.4 What do the authors and the editorialists5 think of the potential for a contribution by those disease
states to the deficits that were identified?
Second, were any subjects excluded on the basis that the
disease states necessitating the procedures or the treatment
thereof might have an impact on learning and development?
In the very careful retrospective study by Block et al.,6 the
apparently appropriate exclusions entailed fully 56% of
anesthetized children who were otherwise eligible on the
basis of complete records. In addition to the airway-related
procedures mentioned above, the original Raine included
procedures requiring cardiopulmonary bypass, operations
involving the orbit or retina, cranioplasty, tracheostomy,
dialysis access. I am concerned that many of these might
reasonably be expected to influence cognitive development,
This letter was sent to the author of the referenced Editorial
View, who declined to reply.
either directly or indirectly. Not incidentally, these issues

may have influenced other studies that have reported an association between anesthetic exposure and cognitive development. Twenty-six percent of the children in the two Olmsted
county cohort studies underwent unspecified Ear, nose, and
throat procedures and 11.5% underwent neurosurgical or
ophthalmologic procedures.7,8
In connection with the same concern about the influence of coincident disease processes, Ing et al. calculated a
Resource Utilization Band (RUB) score to estimate, based
on the number of International Classification of Diseases,
Ninth Revision, diagnostic codes, the potential impact of
comorbidities on neuropsychological test scores. Figure 2
of their study provides data for the original Raine cohort
of 2,868 children, comparing RUB-corrected and RUBuncorrected neuropsychological and academic achievement
scores.1 For every metric, that correction moved the CI bar
closer to the line of no effect, reminding us of the potential for the intrusion of comorbidities. However, those same
RUB correction data, as described in the Methods section,
are not provided for the restricted cohort described in the
present article (their fig. 1).1 What did the RUB correction
reveal? Incidentally, those many of us who are unfamiliar
with RUB scores will wonder why the RUB-driven percentage change in the various scores (fig. 2) is not the same for
every cognitive metric. The editorialists did not comment on
this correction process.5 I would like to know whether they
have scrutinized this correction method and whether they
accept its validity, both qualitatively and quantitatively.
Finally, a correction for gender was applied. This is necessary and appropriate because there was a male preponderance in the exposed subset of children and males are known
to perform less well on standardized testing in general and
on language-oriented tests in particular.3,9,10 The methodology was not specified. Again, were the reviewers and the editorialists privy to information about that correction and can
we the readers of Anesthesiology be confident that it is qualitatively and quantitatively valid? To give us further insight
into the impact of the gender variable, can the authors perform a statistically meaningful comparison of exposed boys
(n = 73) versus unexposed boys (n = 333).
This is an all but terrifying issue for parents. Accordingly,
I think that investigations of this topic should be tempered
by the most rigorous consideration of the possibility that
apparent neurotoxic effects of anesthetics might represent
the intrusion of medical and sociologic covariates. I hope
that the authors1 and the editorialists5 will provide the readership with information and informed opinion as to the
potential impact of comorbidity-related influences on the
association between anesthesia during the first 3 yr of life
and impairment of language skills.
Competing Interests
The author declares no competing interests.
CORRESPONDENCE
John C. Drummond, M.D., F.R.C.P.C., The University of

California, and Veterans Affairs Medical Center, San Diego,
California. jdrummond@ucsd.edu
References
1. Ing CH, DiMaggio CJ, Malacova E, Whitehouse AJ, Hegarty
MK, Feng T, Brady JE, von Ungern-Sternberg BS, Davidson
AJ, Wall MM, Wood AJ, Li G, Sun LS: Comparative analysis of outcome measures used in examining neurodevelopmental effects of early childhood anesthesia exposure.
2. Ing C, DiMaggio C, Whitehouse A, Hegarty MK, Brady J, von
Ungern-Sternberg BS, Davidson A, Wood AJ, Li G, Sun LS: Longterm differences in language and cognitive function after childhood exposure to anesthesia. Pediatrics 2012; 130:e47685
3. Paradise JL, Dollaghan CA, Campbell TF, Feldman HM,

Bernard BS, Colborn DK, Rockette HE, Janosky JE, Pitcairn
DL, Sabo DL, Kurs-Lasky M, Smith CG: Language, speech
sound production, and cognition in three-year-old children
in relation to otitis media in their first three years of life.
Pediatrics 2000; 105:111930
4. Paradise JL, Campbell TF, Dollaghan CA, Feldman HM,

DL, Kurs-Lasky M, Sabo DL, Smith CG: Developmental outcomes after early or delayed insertion of tympanostomy
tubes. N Engl J Med 2005; 353:57686
5. Flick RP, Nemergut ME, Christensen K, Hansen TG: Anestheticrelated neurotoxicity in the young and outcome measures:
The devil is in the details. Anesthesiology 2014; 120:13035
6. Block RI, Thomas JJ, Bayman EO, Choi JY, Kimble KK, Todd
MM: Are anesthesia and surgery during infancy associated with altered academic performance during childhood?

110:796804

9. Hansen TG, Pedersen JK, Henneberg SW, Pedersen DA,

Murray JC, Morton NS, Christensen K: Academic performance
in adolescence after inguinal hernia repair in infancy: A
nationwide cohort study. Anesthesiology 2011; 114:107685
10. Campbell TF, Dollaghan CA, Rockette HE, Paradise JL,

Feldman HM, Shriberg LD, Sabo DL, Kurs-Lasky M: Risk factors for speech delay of unknown origin in 3-year-old children. Child Dev 2003; 74:34657
In Reply:
We thank Dr. Drummond for his interest in our work.1
In response to his request, we have provided a table of
surgical procedures performed in the restricted cohort of
children who had data on all outcomes and covariates of
interest (table1). We also have applied the resource utilization band (RUB) comorbidity correction to our restricted
cohort as requested and found that the risks of deficit in
the restricted cohort were consistent with those reported
Table 1. Procedures Performed on Children Exposed

to Anesthesia in the Restricted Cohort (n = 781)
Procedure
Myringotomy
Inguinal and umbilical hernia
Dental Procedure
Minor skin and nail procedure
Orchiopexy, hydrocele, and varicocele
Tonsillectomy and adenoidectomy
Circumcision
Procedure on extraocular muscles
Hypo/epispadias repair and chordee release
Finger and hand surgery
Procedures on month/tongue and
cleft lip and palate repair
Nasolacrimal duct probe
Computed tomography scan
Foot and knee surgery
Lymph node excision
Minor rectal/anal procedure
Nasal airway procedure
Procedure on orbit, lens, or retina
Tracheostomy and removal
Bone marrow biopsy
Crainiectomy
Gastric and bowel repair and resection
Laparotomy and laparoscopy
Magnetic resonance imaging
PDA ligation
Total
n (%)
38 (24.5)
16 (10.3)
14 (9.0)
10 (6.5)
10 (6.5)
10 (6.5)
8 (5.2)
7 (4.5)
6 (3.9)
5 (3.2)
5 (3.2)
4 (2.6)
3 (1.9)
3 (1.9)
2 (1.3)
2 (1.3)
2 (1.3)
2 (1.3)
2 (1.3)
1 (0.6)
1 (0.6)
1 (0.6)
1 (0.6)
1 (0.6)
1 (0.6)
155 (100)
Due to patients with multiple exposures, the number of procedures

exceeds the number of exposed patients.
PDA = patent ductus arteriosus.
in the full cohort. However, because of the smaller sample

size of the restricted cohort, two of the outcomes could
not be modeled after adding RUB, and the remaining outcomes had wider 95% CIs than those reported in the full
cohort. We would emphasize again that the primary purpose of the restricted cohort analysis was not to quantify
the relative risk of cognitive deficits associated with the
exposure to anesthesia, but rather to assess whether certain outcome measures were more sensitive than others in
measuring differences between the exposed and unexposed
children.
Comorbid illnesses including otitis media are potential
confounders and may play a role in the cognitive outcome
differences between children exposed and unexposed to
anesthesia. Although 33% of the procedures in the full Raine
cohort were otolaryngological in nature, they were unlikely
to sufficiently explain the observed excess risk of cognitive
deficits in the exposed children because similar results have
been reported in children who underwent inguinal hernia
surgery only.2 In addition, the association between otitis media and developmental outcomes is disputed, with a
meta-analysis of prospective studies finding the association
to be negligible.3
CORRESPONDENCE
John C. Drummond, M.D., F.R.C.P.C., The University of

California, and Veterans Affairs Medical Center, San Diego,
California. jdrummond@ucsd.edu
References
2. Ing C, DiMaggio C, Whitehouse A, Hegarty MK, Brady J, von
Ungern-Sternberg BS, Davidson A, Wood AJ, Li G, Sun LS: Longterm differences in language and cognitive function after childhood exposure to anesthesia. Pediatrics 2012; 130:e47685
3. Paradise JL, Dollaghan CA, Campbell TF, Feldman HM,

DL, Sabo DL, Kurs-Lasky M, Smith CG: Language, speech
sound production, and cognition in three-year-old children
in relation to otitis media in their first three years of life.
Pediatrics 2000; 105:111930
4. Paradise JL, Campbell TF, Dollaghan CA, Feldman HM,

DL, Kurs-Lasky M, Sabo DL, Smith CG: Developmental outcomes after early or delayed insertion of tympanostomy
tubes. N Engl J Med 2005; 353:57686
5. Flick RP, Nemergut ME, Christensen K, Hansen TG: Anestheticrelated neurotoxicity in the young and outcome measures:
The devil is in the details. Anesthesiology 2014; 120:13035
6. Block RI, Thomas JJ, Bayman EO, Choi JY, Kimble KK, Todd
MM: Are anesthesia and surgery during infancy associated with altered academic performance during childhood?

110:796804

9. Hansen TG, Pedersen JK, Henneberg SW, Pedersen DA,

Murray JC, Morton NS, Christensen K: Academic performance
in adolescence after inguinal hernia repair in infancy: A
nationwide cohort study. Anesthesiology 2011; 114:107685
10. Campbell TF, Dollaghan CA, Rockette HE, Paradise JL,

Feldman HM, Shriberg LD, Sabo DL, Kurs-Lasky M: Risk factors for speech delay of unknown origin in 3-year-old children. Child Dev 2003; 74:34657
In Reply:
We thank Dr. Drummond for his interest in our work.1
In response to his request, we have provided a table of
surgical procedures performed in the restricted cohort of
children who had data on all outcomes and covariates of
interest (table1). We also have applied the resource utilization band (RUB) comorbidity correction to our restricted
cohort as requested and found that the risks of deficit in
the restricted cohort were consistent with those reported
Table 1. Procedures Performed on Children Exposed

to Anesthesia in the Restricted Cohort (n = 781)
Procedure
Myringotomy
Inguinal and umbilical hernia
Dental Procedure
Minor skin and nail procedure
Orchiopexy, hydrocele, and varicocele
Tonsillectomy and adenoidectomy
Circumcision
Procedure on extraocular muscles
Hypo/epispadias repair and chordee release
Finger and hand surgery
Procedures on month/tongue and
cleft lip and palate repair
Nasolacrimal duct probe
Computed tomography scan
Foot and knee surgery
Lymph node excision
Minor rectal/anal procedure
Nasal airway procedure
Procedure on orbit, lens, or retina
Tracheostomy and removal
Bone marrow biopsy
Crainiectomy
Gastric and bowel repair and resection
Laparotomy and laparoscopy
Magnetic resonance imaging
PDA ligation
Total
n (%)
38 (24.5)
16 (10.3)
14 (9.0)
10 (6.5)
10 (6.5)
10 (6.5)
8 (5.2)
7 (4.5)
6 (3.9)
5 (3.2)
5 (3.2)
4 (2.6)
3 (1.9)
3 (1.9)
2 (1.3)
2 (1.3)
2 (1.3)
2 (1.3)
2 (1.3)
1 (0.6)
1 (0.6)
1 (0.6)
1 (0.6)
1 (0.6)
1 (0.6)
155 (100)
Due to patients with multiple exposures, the number of procedures

exceeds the number of exposed patients.
PDA = patent ductus arteriosus.
in the full cohort. However, because of the smaller sample

size of the restricted cohort, two of the outcomes could
not be modeled after adding RUB, and the remaining outcomes had wider 95% CIs than those reported in the full
cohort. We would emphasize again that the primary purpose of the restricted cohort analysis was not to quantify
the relative risk of cognitive deficits associated with the
exposure to anesthesia, but rather to assess whether certain outcome measures were more sensitive than others in
measuring differences between the exposed and unexposed
children.
Comorbid illnesses including otitis media are potential
confounders and may play a role in the cognitive outcome
differences between children exposed and unexposed to
anesthesia. Although 33% of the procedures in the full Raine
cohort were otolaryngological in nature, they were unlikely
to sufficiently explain the observed excess risk of cognitive
deficits in the exposed children because similar results have
been reported in children who underwent inguinal hernia
surgery only.2 In addition, the association between otitis media and developmental outcomes is disputed, with a
meta-analysis of prospective studies finding the association
to be negligible.3
Correspondence
As an alternative to restricting our analysis to specific

procedures or illness profiles, we used RUB scores to
account for comorbid illnesses in all children in the Raine
cohort. The Johns Hopkins Adjusted Clinical Groups
Case-Mix System was used to generate the RUB scores.
The Adjusted Clinical Groups system is a widely accepted
method for measuring health resource utilization and was
used in this study after a comprehensive search to find the
most appropriate comorbidity adjustment method available for children and adults. The reason that the RUB
comorbidity correction had varying effects on the different outcomes is that different cognitive deficits may impact
health services resource utilization differently. However,
as emphasized in our Discussion, unmeasured and often
unknown differences in exposed and unexposed children
may still represent a source of residual confounding. Even
the most rigorous adjustment technique in an observational study cannot match the ability of randomization
to account for both known and unknown confounders
between the two groups.
Regarding the adjustment for sex, all medical and demographic covariates were included as categorical variables in
our modified multivariable Poisson regression model, which
is a standard method in regression modeling. Per Dr. Drummonds request, we have evaluated the association between
anesthetic exposure and cognitive deficit in boys in the
restricted and full cohorts and found that the point estimates of the risk ratios were consistent with those reported
in our article. Owing to decreased sample size in this subset
analysis, however, some of the models did not converge and
those that did provided wider CIs for the outcome variables.
We also tested for possible interaction by sex, and found
that there was no significant interaction with sex in any of
the associations between exposure to anesthesia and cognitive outcomes.
We agree with Dr. Drummond that the question of
whether exposure to anesthetics in early childhood has
any significant long-term adverse effect on neurodevelopment in children is of clinical and public health importance yet extremely difficult to tackle. We are determined
to continue our efforts to help answer this important
question through rigorously designed and carefully performed studies.
Competing Interests
Caleb H. Ing, M.D., M.S., Charles J. DiMaggio,
Ph.D., M.P.H., P.A.-C., Eva Malacova, Ph.D., Andrew
J. Whitehouse, Ph.D., Mary K. Hegarty, M.B.B.S.,
F.A.N.Z.C.A., Tianshu Feng, M.S., Joanne E. Brady,
M.S., Britta S. von Ungern-Sternberg, M.D., Ph.D.,
Andrew J. Davidson, M.D., Melanie M. Wall, Ph.D.,
Alastair J. J. Wood, M.D., Guohua Li, M.D., Dr.P.H.,
Lena S. Sun, M.D. C
olumbia University College of Physicians and Surgeons, New York, New York (C.H.I.).
ci2119@cumc.columbia.edu
References
2. DiMaggio C, Sun LS, Kakavouli A, Byrne MW, Li G: A retrospective cohort study of the association of anesthesia and
hernia repair surgery with behavioral and developmental
disorders in young children. J Neurosurg Anesthesiol 2009;
21:28691
3. Roberts JE, Rosenfeld RM, Zeisel SA: Otitis media and

speech and language: A meta-analysis of prospective studies.
Pediatrics 2004; 113(3 Pt 1):e23848
Old Guidelines or Methods Cannot

Insure Quality or Progress
To the Editor:
The recent article and editorial regarding the use of a proprietary Decision Support Tool extolled the need to quit
memorizing data, favoring medical interactive applications
in applying medical knowledge.1,2 The Decision Support
Tool was designed to increase adherence to an outdated yet
still current 2007 American College of Cardiology/American Heart Association perioperative evaluation consensus
guideline (PECG). It is clear that perioperative blockade
(PBB) and cost containment played a very large role in the
underlying assumptions of that guideline. It is also clear that
in 2008, the POISE study (PeriOperative ISchemic Evaluation trial [ClinicalTrials.gov Identifier: NCT00182039])
completely transformed the premise of PBB, finding PBB
stroke morbidity outweighed any cardiac morbidity prevention. PBB guideline revisions followed rapidly in 2009,
without corresponding PECG changes. Furthermore, the
reporting of Dr. Poldermans ethical violations, as a world
proponent of PBB, further publically raised significant questions undermining the 2007 PECG validity. Cardiac guidelines experience particularly rapid turnover for multiple
reasons.* Medical reversal is a rapidly emerging reality, indicating guidelines have limits to application, as well as potentially short shelf-lives, as PBB clearly demonstrated.3 This
may directly compromise the usage of any Decision Support
Tool, especially if failing to update rapidly while physicians
life-long learning does facilitate updates.
Assuming the PECG is correct in 2014, is a fundamental problem. Similarly, testing correct answers based
* Boyles S: Cardiac Practice Guidelines Have High Turnover, May
27, 2014. Available at: http://www.medpagetoday.com/Cardiology/
CHF/46004. Accessed June 13, 2014.
Correspondence
As an alternative to restricting our analysis to specific

procedures or illness profiles, we used RUB scores to
account for comorbid illnesses in all children in the Raine
cohort. The Johns Hopkins Adjusted Clinical Groups
Case-Mix System was used to generate the RUB scores.
The Adjusted Clinical Groups system is a widely accepted
method for measuring health resource utilization and was
used in this study after a comprehensive search to find the
most appropriate comorbidity adjustment method available for children and adults. The reason that the RUB
comorbidity correction had varying effects on the different outcomes is that different cognitive deficits may impact
health services resource utilization differently. However,
as emphasized in our Discussion, unmeasured and often
unknown differences in exposed and unexposed children
may still represent a source of residual confounding. Even
the most rigorous adjustment technique in an observational study cannot match the ability of randomization
to account for both known and unknown confounders
between the two groups.
Regarding the adjustment for sex, all medical and demographic covariates were included as categorical variables in
our modified multivariable Poisson regression model, which
is a standard method in regression modeling. Per Dr. Drummonds request, we have evaluated the association between
anesthetic exposure and cognitive deficit in boys in the
restricted and full cohorts and found that the point estimates of the risk ratios were consistent with those reported
in our article. Owing to decreased sample size in this subset
analysis, however, some of the models did not converge and
those that did provided wider CIs for the outcome variables.
We also tested for possible interaction by sex, and found
that there was no significant interaction with sex in any of
the associations between exposure to anesthesia and cognitive outcomes.
We agree with Dr. Drummond that the question of
whether exposure to anesthetics in early childhood has
any significant long-term adverse effect on neurodevelopment in children is of clinical and public health importance yet extremely difficult to tackle. We are determined
to continue our efforts to help answer this important
question through rigorously designed and carefully performed studies.
Competing Interests
Caleb H. Ing, M.D., M.S., Charles J. DiMaggio,
Ph.D., M.P.H., P.A.-C., Eva Malacova, Ph.D., Andrew
J. Whitehouse, Ph.D., Mary K. Hegarty, M.B.B.S.,
F.A.N.Z.C.A., Tianshu Feng, M.S., Joanne E. Brady,
M.S., Britta S. von Ungern-Sternberg, M.D., Ph.D.,
Andrew J. Davidson, M.D., Melanie M. Wall, Ph.D.,
Alastair J. J. Wood, M.D., Guohua Li, M.D., Dr.P.H.,
Lena S. Sun, M.D. C
olumbia University College of Physicians and Surgeons, New York, New York (C.H.I.).
ci2119@cumc.columbia.edu
References
2. DiMaggio C, Sun LS, Kakavouli A, Byrne MW, Li G: A retrospective cohort study of the association of anesthesia and
hernia repair surgery with behavioral and developmental
disorders in young children. J Neurosurg Anesthesiol 2009;
21:28691
3. Roberts JE, Rosenfeld RM, Zeisel SA: Otitis media and

speech and language: A meta-analysis of prospective studies.
Pediatrics 2004; 113(3 Pt 1):e23848
Old Guidelines or Methods Cannot

Insure Quality or Progress
To the Editor:
The recent article and editorial regarding the use of a proprietary Decision Support Tool extolled the need to quit
memorizing data, favoring medical interactive applications
in applying medical knowledge.1,2 The Decision Support
Tool was designed to increase adherence to an outdated yet
still current 2007 American College of Cardiology/American Heart Association perioperative evaluation consensus
guideline (PECG). It is clear that perioperative blockade
(PBB) and cost containment played a very large role in the
underlying assumptions of that guideline. It is also clear that
in 2008, the POISE study (PeriOperative ISchemic Evaluation trial [ClinicalTrials.gov Identifier: NCT00182039])
completely transformed the premise of PBB, finding PBB
stroke morbidity outweighed any cardiac morbidity prevention. PBB guideline revisions followed rapidly in 2009,
without corresponding PECG changes. Furthermore, the
reporting of Dr. Poldermans ethical violations, as a world
proponent of PBB, further publically raised significant questions undermining the 2007 PECG validity. Cardiac guidelines experience particularly rapid turnover for multiple
reasons.* Medical reversal is a rapidly emerging reality, indicating guidelines have limits to application, as well as potentially short shelf-lives, as PBB clearly demonstrated.3 This
may directly compromise the usage of any Decision Support
Tool, especially if failing to update rapidly while physicians
life-long learning does facilitate updates.
Assuming the PECG is correct in 2014, is a fundamental problem. Similarly, testing correct answers based
* Boyles S: Cardiac Practice Guidelines Have High Turnover, May
27, 2014. Available at: http://www.medpagetoday.com/Cardiology/
CHF/46004. Accessed June 13, 2014.
CORRESPONDENCE
on a Decision Support Tool adhering to the 2007 PECG,

presents simply a false premise for contemporary knowledge. The guideline should also fit the patient and not vice
versa. I would ask the researchers to publish their defined
as correct answers to the already published questions, to
facilitate assessment whether these answers are deemed
correct by modern readers! Knowledgeable physicians may
justifiably reject the guideline and the proposed Correct
answers in modern practice, especially when tailored to
the variable contemporary reality at hand (i.e., University
vs. rural hospital). External realities further impose, where
patients produce satisfaction scores and see themselves
deserving EVERY consideration, test and therapy, regardless of cost, when rare complications produce 100% morbidity and mortality to them personally as rare events.
The editorial goes even further, promoting the unproven
utility of recertification. Similarly, transferring simulation
and objective structured clinical examination applications
for medical student/resident educations onto Recertification testing of practicing and competent physicians is yet
another unproven leap of faith. Board certification, and
especially recertification, have never been proven or demonstrated to clearly improve quality in care in outcomebased studies.
The real problem emphasized by both study and editorial, is that while both support the use of internet-based
data acquisition in daily medical practice, certification,
and recertification tests forbid it completely. Similar to old
guidelines, simply believing that certification or recertification matters, may also be a mere historical relic, proprietary advertisement and/or simple false legacy assumption,
having emerged before modern licensure and extensive regulation of residency training programs.4 These and other
concerns have led to the significant opposition to maintenance of certification among physicians at large. It is time
for an open discussion of the risks and benefits of the cost
and unproven assumptions of recertification and maintenance of certification, as a Quality = Value/cost indicator.
Blind, computerized, adherence to aging guidelines, however, requires short-term revalidation of underlying programs to insure patient safety.
Competing Interests
Paul M. Kempen, M.D., Ph.D., Weirton Medical Center,
Weirton, West Virginia. kmpnpm@yahoo.com
References
1. Hand WR, Bridges KH, Stiegler MP, Schell RM, DiLorenzo AN,
Ehrenfeld JM, Nietert PH, McEvoy MD: Effect of cognitive aid
on adherence to perioperative assessment and management
guidelines for the cardiac evaluation of noncardiac surgical
2. Lockman JL, Schwartz AJ: Learn it-memorize it! Better

yetopen your smartphone and use the information!
3. Prasad V, Vandross A, Toomey C, Cheung M, Rho J, Quinn

S, Chacko SJ, Borkar D, Gall V, Selvaraj S, Ho N, Cifu A: A
decade of reversal: An analysis of 146 contradicted medical
practices. Mayo Clin Proc 2013; 88:7908
4. Kempen PM: Maintenance of certification and

licensure:
Regulatory capture of medicine. Anesth Analg 2014;
118:137886
In Reply:
Dr. Kempens Letter to the Editor provides his perspective on the study by Hand et al.1 and our accompanying
Editorial2 recently published in Anesthesiology. Kempen
asserts that medical knowledge changes rapidly and as
such, This may directly compromise the utility of any
DST [Decision Support Tool], especially if failing to
update rapidly-while physicians lifelong learning does
facilitate updates.
Were it so easy and simple! How wonderful it would be if
physicians were able to conduct their own lifelong learning
in a dedicated, systematic, rigorous, and comprehensive fashion. We all wish we could faithfully read all of the pertinent
journals in a most timely manner; ideally, we would also perform critical appraisals of each manuscript and incorporate
only the appropriate results into our daily medical practice.
Of course, this theoretical physician would also attend to
patient cares long hours while balancing all of lifes outside
demands. Professional and personal life just do not allow for
this ideal vision of lifelong learning.
Lifelong learning is essential for effective and efficient
patient care. Teunissen and Dorman3 remind us that medical schools do not adequately prepare new practitioners to
assume the responsibilities of patient care and simultaneously be the best lifelong learners. Panda and Desbiens4 present a strong case for incorporating lifelong learning concepts
into undergraduate and graduate medical education. Becker
et al.5 studied stress and burnout among physicians and
attribute difficulty in staying current with new knowledge as
one contributor to the problem. Perhaps most importantly,
Burden et al.6 have studied the use of cognitive aids during
simulated anesthetic emergencies (and specifically a designated Reader to assist the Leader during these events).
They found that whereas none of the subjects performed all
of the critical actions in the control group, introduction of
a Reader with a cognitive aid resulted in execution of all
described critical actions.
As aptly pointed out by Li et al.7 in their analysis of and
suggestions for lifelong learning strategies, multiple barriers stand in the way of physicians attempting to reach
their self-directed learning goals. Five themes emerged
that characterized barriers to achieving learning goals:
difficulty with personal reflection, environmental strain,
competing demands, difficulty with goal generation, and
problems with plan development and implementation.7
CORRESPONDENCE
on a Decision Support Tool adhering to the 2007 PECG,

presents simply a false premise for contemporary knowledge. The guideline should also fit the patient and not vice
versa. I would ask the researchers to publish their defined
as correct answers to the already published questions, to
facilitate assessment whether these answers are deemed
correct by modern readers! Knowledgeable physicians may
justifiably reject the guideline and the proposed Correct
answers in modern practice, especially when tailored to
the variable contemporary reality at hand (i.e., University
vs. rural hospital). External realities further impose, where
patients produce satisfaction scores and see themselves
deserving EVERY consideration, test and therapy, regardless of cost, when rare complications produce 100% morbidity and mortality to them personally as rare events.
The editorial goes even further, promoting the unproven
utility of recertification. Similarly, transferring simulation
and objective structured clinical examination applications
for medical student/resident educations onto Recertification testing of practicing and competent physicians is yet
another unproven leap of faith. Board certification, and
especially recertification, have never been proven or demonstrated to clearly improve quality in care in outcomebased studies.
The real problem emphasized by both study and editorial, is that while both support the use of internet-based
data acquisition in daily medical practice, certification,
and recertification tests forbid it completely. Similar to old
guidelines, simply believing that certification or recertification matters, may also be a mere historical relic, proprietary advertisement and/or simple false legacy assumption,
having emerged before modern licensure and extensive regulation of residency training programs.4 These and other
concerns have led to the significant opposition to maintenance of certification among physicians at large. It is time
for an open discussion of the risks and benefits of the cost
and unproven assumptions of recertification and maintenance of certification, as a Quality = Value/cost indicator.
Blind, computerized, adherence to aging guidelines, however, requires short-term revalidation of underlying programs to insure patient safety.
Competing Interests
Paul M. Kempen, M.D., Ph.D., Weirton Medical Center,
Weirton, West Virginia. kmpnpm@yahoo.com
References
Ehrenfeld JM, Nietert PH, McEvoy MD: Effect of cognitive aid
on adherence to perioperative assessment and management
guidelines for the cardiac evaluation of noncardiac surgical

3. Prasad V, Vandross A, Toomey C, Cheung M, Rho J, Quinn

S, Chacko SJ, Borkar D, Gall V, Selvaraj S, Ho N, Cifu A: A
decade of reversal: An analysis of 146 contradicted medical
practices. Mayo Clin Proc 2013; 88:7908
4. Kempen PM: Maintenance of certification and

licensure:
Regulatory capture of medicine. Anesth Analg 2014;
118:137886
In Reply:
Dr. Kempens Letter to the Editor provides his perspective on the study by Hand et al.1 and our accompanying
Editorial2 recently published in Anesthesiology. Kempen
asserts that medical knowledge changes rapidly and as
such, This may directly compromise the utility of any
DST [Decision Support Tool], especially if failing to
update rapidly-while physicians lifelong learning does
facilitate updates.
Were it so easy and simple! How wonderful it would be if
physicians were able to conduct their own lifelong learning
in a dedicated, systematic, rigorous, and comprehensive fashion. We all wish we could faithfully read all of the pertinent
journals in a most timely manner; ideally, we would also perform critical appraisals of each manuscript and incorporate
only the appropriate results into our daily medical practice.
Of course, this theoretical physician would also attend to
patient cares long hours while balancing all of lifes outside
demands. Professional and personal life just do not allow for
this ideal vision of lifelong learning.
Lifelong learning is essential for effective and efficient
patient care. Teunissen and Dorman3 remind us that medical schools do not adequately prepare new practitioners to
assume the responsibilities of patient care and simultaneously be the best lifelong learners. Panda and Desbiens4 present a strong case for incorporating lifelong learning concepts
into undergraduate and graduate medical education. Becker
et al.5 studied stress and burnout among physicians and
attribute difficulty in staying current with new knowledge as
one contributor to the problem. Perhaps most importantly,
Burden et al.6 have studied the use of cognitive aids during
simulated anesthetic emergencies (and specifically a designated Reader to assist the Leader during these events).
They found that whereas none of the subjects performed all
of the critical actions in the control group, introduction of
a Reader with a cognitive aid resulted in execution of all
described critical actions.
As aptly pointed out by Li et al.7 in their analysis of and
suggestions for lifelong learning strategies, multiple barriers stand in the way of physicians attempting to reach
their self-directed learning goals. Five themes emerged
that characterized barriers to achieving learning goals:
difficulty with personal reflection, environmental strain,
competing demands, difficulty with goal generation, and
problems with plan development and implementation.7
Correspondence
To assist physicians genuinely wanting to be current in

their knowledge and practice (this is the overwhelming
majority of practitioners), Li et al. suggested strategies to
offset the barriers, including the development of external mentorship and accountability. Experts in specific
medical practice areas and in medical education are the
mentors from whom we all benefit when they guide the
content to learn and suggest our learning expectations
to gain that education. Recognition of this is one of the
key messages from the study of decision support tools by
Hand et al.
Kempen claims that our Editorial promotes, the
unproven utility of recertification and supports,
transferring simulation and OSCE [Objective Structured
Clinical Examination] applications for medical student/
resident education onto Recertification As educators,
we champion a very different message than claimed by
Kempen; we asked a question and provided our answer,
How best then to teach and learn safe provider autonomy? provide perfect practice via simulation and
use of decision support tools 2 We champion simulation, decision support tools and other cognitive aids,
and all types of hands-on experiences for their ability to
facilitate relevant lifelong learning. We are especially supportive of utilizing the Internet to make the transmission
of new information instantaneous. We make no assertion
that these learning tools enhance any certification/recertification program, but decreasing practice variability and
improving adherence to published guidelines are beneficial
to our patients, and there are data to support the role of
cognitive aids in these goals.
Competing Interests
Justin L. Lockman, M.D., Alan Jay Schwartz, M.D.,
M.S.Ed. University of Pennsylvania, The Childrens Hospital of Philadelphia, Philadelphia, Pennsylvania (A.J.S.).
schwartza@email.chop.edu
References
Ehrenfeld JM, Nietert PJ, McEvoy MD: Effect of a cognitive
aid on adherence to perioperative assessment and management guidelines for the cardiac evaluation of noncardiac surgical patients. Anesthesiology 2014; 120:133953

3. Teunissen PW, Dornan T: The competent novice lifelong

learning at work. BMJ 2008; 336:6679
4. Panda M, Desbiens NA: An education for life requirement
to promote lifelong learning in an internal medicine residency program. J Grad Med Educ 2010; 2:5625
5. Becker JL, Milad MP, Klock SC: Burnout, depression, and
career satisfaction: Cross-sectional study of obstetrics
and gynecology residents. Am J Obstet Gynecol 2006;
195:14449
6. Burden AR, Carr ZJ, Staman GW, Littman JJ, Torjman MC:
Does every code need a reader? Improvement of rare event
management with a cognitive aid reader during a simulated
emergency: A pilot study. Simul Healthc 2012; 7:19
7. Li ST, Paterniti DA, Co JP, West DC: Successful self-directed
lifelong learning in medicine: A conceptual model derived
from qualitative analysis of a national survey of pediatric
residents. Acad Med 2010; 85:122936
In Reply:
We thank Dr. Kempen for his interest in our recent
study published in Anesthesiology about the effect of
a decision support tool (DST) on adherence to published
guidelines.1
Dr. Kempen endorses the importance of evidencebased practice, demonstrated by his thorough reiteration
of the narrative related to perioperative -blockade, and
this underlying premise to his letter is very important.
Physicians should try to practice with the most up-to-date
and clinically-applicable evidence available. The effort of
our study was not surrounding the validity of the claims
of the 2007 American College of Cardiologists/American
Heart Association perioperative guidelines,2 but rather
to test the ability of mobile health technology to help
physicians apply this guideline to patient scenarios. It is
not lost on the authors that evidence will continue to be
refined, and, in fact, we have already begun to modify the
DST application based on the 2014 update to the American College of Cardiologists/American Heart Association
guidelines released this fall.3 This point is perhaps the
most important reason we believe a DST to be superior
to memory alone. The DST can be updated centrally with
push updates sent to end users quickly; and, in theory,
practice patterns can be modified almost instantly when
this occurs.
However, Dr. Kempen also notes that there may be a
fundamental problem with assuming that the 2014 practice guidelines are correct. To this we would state that
we are aware that these guidelines, as the former ones, will
almost certainly require amendment in the future. However, the reality of this fact does not negate the validity of
the approach to producing a practice guideline founded on a
rigorous evidence-based review, as detailed in the guidelines.
Additionally, we believe that understanding such guidelines
can aid physicians in the very struggles that Dr. Kempen
notes concerning patient expectations and responsible testing considerations. Dr. Kempen points out that resources are
inconsistently available depending on the location and size of
a facility. Regarding the interventions indicated by the 2007
American College of Cardiologists/American Heart Association guidelines discussed, we agree that select patients might
be simply better served having surgery only where there is
Correspondence
To assist physicians genuinely wanting to be current in

their knowledge and practice (this is the overwhelming
majority of practitioners), Li et al. suggested strategies to
offset the barriers, including the development of external mentorship and accountability. Experts in specific
medical practice areas and in medical education are the
mentors from whom we all benefit when they guide the
content to learn and suggest our learning expectations
to gain that education. Recognition of this is one of the
key messages from the study of decision support tools by
Hand et al.
Kempen claims that our Editorial promotes, the
unproven utility of recertification and supports,
transferring simulation and OSCE [Objective Structured
Clinical Examination] applications for medical student/
resident education onto Recertification As educators,
we champion a very different message than claimed by
Kempen; we asked a question and provided our answer,
How best then to teach and learn safe provider autonomy? provide perfect practice via simulation and
use of decision support tools 2 We champion simulation, decision support tools and other cognitive aids,
and all types of hands-on experiences for their ability to
facilitate relevant lifelong learning. We are especially supportive of utilizing the Internet to make the transmission
of new information instantaneous. We make no assertion
that these learning tools enhance any certification/recertification program, but decreasing practice variability and
improving adherence to published guidelines are beneficial
to our patients, and there are data to support the role of
cognitive aids in these goals.
Competing Interests
Justin L. Lockman, M.D., Alan Jay Schwartz, M.D.,
M.S.Ed. University of Pennsylvania, The Childrens Hospital of Philadelphia, Philadelphia, Pennsylvania (A.J.S.).
schwartza@email.chop.edu
References

3. Teunissen PW, Dornan T: The competent novice lifelong

learning at work. BMJ 2008; 336:6679
4. Panda M, Desbiens NA: An education for life requirement
to promote lifelong learning in an internal medicine residency program. J Grad Med Educ 2010; 2:5625
5. Becker JL, Milad MP, Klock SC: Burnout, depression, and
career satisfaction: Cross-sectional study of obstetrics
and gynecology residents. Am J Obstet Gynecol 2006;
195:14449
6. Burden AR, Carr ZJ, Staman GW, Littman JJ, Torjman MC:
Does every code need a reader? Improvement of rare event
management with a cognitive aid reader during a simulated
emergency: A pilot study. Simul Healthc 2012; 7:19
7. Li ST, Paterniti DA, Co JP, West DC: Successful self-directed
lifelong learning in medicine: A conceptual model derived
from qualitative analysis of a national survey of pediatric
residents. Acad Med 2010; 85:122936
In Reply:
We thank Dr. Kempen for his interest in our recent
study published in Anesthesiology about the effect of
a decision support tool (DST) on adherence to published
guidelines.1
Dr. Kempen endorses the importance of evidencebased practice, demonstrated by his thorough reiteration
of the narrative related to perioperative -blockade, and
this underlying premise to his letter is very important.
Physicians should try to practice with the most up-to-date
and clinically-applicable evidence available. The effort of
our study was not surrounding the validity of the claims
of the 2007 American College of Cardiologists/American
Heart Association perioperative guidelines,2 but rather
to test the ability of mobile health technology to help
physicians apply this guideline to patient scenarios. It is
not lost on the authors that evidence will continue to be
refined, and, in fact, we have already begun to modify the
DST application based on the 2014 update to the American College of Cardiologists/American Heart Association
guidelines released this fall.3 This point is perhaps the
most important reason we believe a DST to be superior
to memory alone. The DST can be updated centrally with
push updates sent to end users quickly; and, in theory,
practice patterns can be modified almost instantly when
this occurs.
However, Dr. Kempen also notes that there may be a
fundamental problem with assuming that the 2014 practice guidelines are correct. To this we would state that
we are aware that these guidelines, as the former ones, will
almost certainly require amendment in the future. However, the reality of this fact does not negate the validity of
the approach to producing a practice guideline founded on a
rigorous evidence-based review, as detailed in the guidelines.
Additionally, we believe that understanding such guidelines
can aid physicians in the very struggles that Dr. Kempen
notes concerning patient expectations and responsible testing considerations. Dr. Kempen points out that resources are
inconsistently available depending on the location and size of
a facility. Regarding the interventions indicated by the 2007
American College of Cardiologists/American Heart Association guidelines discussed, we agree that select patients might
be simply better served having surgery only where there is
CORRESPONDENCE
access to echocardiography, stress tests, and perhaps medical

management, although we hope this doesnot limit access to
care as we expect they are nearly ubiquitous, even outside the
University hospital.
Concerning Dr. Kempens statements about board certification, we make no comment here as that was not the object
under consideration in our article.
Dr. Kempens concluding remarks should be heeded
revalidation on the premise of our study needs to be publishedthat a DST will improve adherence to published
guidelines, ideally in actual patient care. It then falls upon
the practitioner and software developer to ensure the guidelines are internally valid and up to date, representing what is
actually published and then allowing the clinical to make the
final decision in application.
Competing Interests
William R. Hand, M.D., Matthew D. McEvoy, M.D. Medical University of South Carolina, Charleston, South Carolina
(W.R.H.). handw@musc.edu
References
2. Fleisher LA, Beckman JA, Brown KA, Calkins H, Chaikof

E, Fleischmann KE, Freeman WK, Froehlich JB, Kasper
EK, Kersten JR, Riegel B, Robb JF, Smith SC Jr, Jacobs AK,
Adams CD, Anderson JL, Antman EM, Buller CE, Creager
MA, Ettinger SM, Faxon DP, Fuster V, Halperin JL, Hiratzka
LF, Hunt SA, Lytle BW, Nishimura R, Ornato JP, Page RL,
Tarkington LG, Yancy CW: ACC/AHA 2007 guidelines on
perioperative cardiovascular evaluation and care for noncardiac surgery: Executive summary: A report of the American
College of Cardiology/American Heart Association Task
Force on Practice Guidelines. Circulation 2007; 116: 197196
3. Fleisher LA, Fleischmann KE,Auerbach AD, Barnason SA, Beckman
JA, Bozkurt B, Davila-Roman VG, Gerhard-Herman MD, Holly TA,
Kane GC, Marine JE, Nelson MT, Spencer CC, Thompson A, Ting
HH, Uretsky BF, Wijeysundera DN: 2014 ACC/AHA Guideline
on Perioperative Cardiovascular Evaluation and Management
of Patients Undergoing Noncardiac Surgery: A Report of the
American College of Cardiology/American Heart Association Task
Force on Practice Guidelines. J Am Coll Cardiol 2014 [Epub ahead
of print]
REVIEWS OF EDUCATIONAL MATERIAL

Alan Jay Schwartz, M.D., M.S. Ed., Editor
Anesthesia and the Fetus. Edited by Yehuda
Ginosar, B.Sc., M.B.B.S., Felicity Reynolds, M.D.,
F.R.C.A., Stephen H. Halpern, M.D., M.Sc., Carl P.
Weiner, M.B.B.S., M.B.A., F.A.C.O.G. Chichester,
United Kingdom, Wiley-Blackwell, 2013. Pages: 418.
Price: Hardcover$149.95, E-book$119.99.
As the title suggests, Anesthesia and the Fetus is the first textbook focused specifically on anesthetic implications for the
fetus. Editors Yehuda Ginosar, Felicity Reynolds, Stephen
Halpern, and Carl Weiner recruited an international and
internationally renowned group of anesthesiologists, obstetricians, neonatologists, pharmacologists, toxicologists, teratologists, and ethicists to distill and integrate an extensive
and far-reaching body of evidence into a concise and clinically relevant resource.
The book is divided into six sections. The first section
reviews basic physiology and pharmacology of the pregnant
woman, the developing fetus, and the perinatal transition to
extra-uterine life. The chapters are generally focused, brief
and well organized.
Anesthesia and the Fetus distinguishes itself in Section 2,
which details the assessment of fetal well-being and neonatal
outcome, achieving a depth and breadth not seen in most
anesthesiology textbooks. Comprehensive knowledge of
cardiotocography and intrapartum fetal assessment can be a
powerful tool to improve intraprofessional collaboration and
care on the labor and delivery unit.
One striking conclusion from reading this book is how
little is known about the long-term impact of many anesthetic interventions on the child the fetus will later become.
Chapter 8 explores the validity of endpoint measurement and questions the usage of surrogate outcomes, such
as umbilical arterial blood gas measurements. Although
imploring obstetric and fetal anesthesia researchers to focus
their efforts on clinically meaningful outcome measures, the
chapter could go further by specifically advocating for longterm measures of neonatal and childhood development, or
even school and workforce performance.
While longitudinal research after anesthetic exposure will
be logistically difficult and expensive, its necessity is highlighted in Section 3. Chapters on neuraxial analgesia and
anesthesia are largely reassuring. But unanswered questions
about fetal safety remain. Together, Chapters 21 and 22
thoroughly explore both the hypothesized mechanisms and
the clinical implications of anesthesia-induced neuroapoptosis on the developing brain. This body of evidence was generated largely in animal models, leading some to question
the relevance for human fetuses. The authors of Chapter 22
Michael J. Avram, Ph.D., served as Handling Editor for this book
review.
Copyright 2014, the American Society of Anesthesiologists, Inc. Lippincott
Williams & Wilkins. Anesthesiology 2015; 122:222-3
instead call for research in neuroprotective strategies, and

anesthetic protocols (e.g., maternal neuraxial anesthesia) that
limit fetal and neonatal exposure to general anesthetics.
Likewise, Chapter 30 delves into the fetal and neonatal effects
of fever attributed to neuraxial analgesia. In itself, hyperthermia
might be considered to be benign. The authors acknowledge the
possibility that only fever due to infection may lead to cerebral
palsy. But they cite additional evidence that pyrexia on its own
exacerbates damage due to hypoxia and ischemia without any
involvement of an infectious process.(p. 280) The second half
of the chapter describes strategies to limit epidural-related fever
and concludes with the possibility that maternal neck warmers
could be used to heat blood flowing to the maternal hypothalamic temperature receptors, thereby attenuating any hyperthermic effects of neuraxial analgesia.
Almost every fetal therapy is administered via maternal
interventions. The exception may be drugs and fluids administered directly to the fetus during an in utero intervention. We
had hoped this book would contain a chapter on the pharmacokinetics of drugs and fluids administered directly to the
fetusintramuscularly by spinal needle through the maternal abdomen, intravascularly through the umbilical artery, or
through a peripheral intravenous line inserted during an ex utero
intrapartum treatment procedure. Neonatal dosing guidelines
may be insufficient for the fetus, when dosages and volumes are
selected based on estimated body weight. This lesson was driven
home while observing our pediatric anesthesiologists struggle to
resuscitate a distressed fetus during an in utero sacrococcygeal
teratoma resection. Multiple fluid boluses of 10ml/kg failed to
fill the tiny quivering heart, likely because fetal blood also perfuses the placenta.
Section 4 surveys the interactions between anesthetic
interventions and multiple gestation and preeclampsia.
Section 5 reviews resuscitation guidelines for trauma, and
for maternal, fetal, and neonatal resuscitation.
Tucked in the final section of the book are two excellent
chapters on medical ethics, research, and the law. The authors
navigate this controversial terrain with expertise, drawing on
fundamental ethical principles to clearly delineate working
principles to balance obligations to the maternal and fetal
patients. International perspectives are offered in exploring
wide variation in the definitions of fetal rights. Chervenak
and McCullough conclude in Chapter 40 that ethically
as well as clinically the fetus is not a separate patient. And
yet, the fetus remains central to the care of any pregnant
woman, and one finishes the book with the acute awareness
that more research, more knowledge, and more awareness
about clinically significant fetal effects are needed to ensure
that any therapies administered to the mother will appropriately balance fetal beneficence with maternal beneficence
and autonomy.
The book demonstrates characteristics of a first edition,
with some redundancy noted within a few of the chapters. Nevertheless, this is an extremely useful resource for
January 2015
Reviews of Educational Material
anesthesiologists who wish to quickly discern what is known

and what is not yet known about the fetal effects of anesthesia. In at least a half dozen of the chapters, the authors
have undertaken a systematic review and present tables of
relevant randomized studies in addition to summarizing and
synthesizing their findings. For the busy clinician, the chapter conclusions are a rapid way to systematically sample the
content of each chapter throughout the entire book.
Expertly administered anesthesia offers tremendous potential to improve the comfort of childbirth and obstetric surgery
while preserving both fetal and maternal well-being. Increasingly, pregnant women face a barrage of well-meaning advice
to limit environmental and pharmacologic exposures for their
growing fetus. Consequently, to be effective, anesthesiologists
need to do more than simply administer safe anesthesia; we
must serve as ambassadors and educators with comprehensive knowledge of the effects of our interventions on both the
mother and fetus. For clinicians who seek to fulfil this obligation, Anesthesia and the Fetus will be an excellent resource.
Baskar Rajala, M.B.B.S., F.R.C.A., Jill M. Mhyre, M.D.
University of Arkansas for Medical Sciences, Little Rock, Arkansas ( J.M.M). jmmhyre@uams.edu
Shnider and Levinsons Anesthesia for

Obstetrics, Fifth Edition. Edited by Maya S.
Suresh, M.D., B. Scott Segal, M.D., M.H.C.M.,
Roanne L. Preston, M.D., F.R.C.P.C., Roshan
Fernando, M.B., B.Ch., F.R.C.A., C. LaToya Mason.,
M.D. Philadephia, Lippincott Williams & Wilkins,
2013. Pages: 902. Price: $232.99.
Most practicing physician anesthesiologists in community
hospitals and teaching hospitals are obliged to cover obstetrical anesthesia during the work day and call coverage, which
may be the most stressful part of a practice. Keeping up with
the literature specific to obstetric anesthesia may be a challenge to the busy practitioner. The revised and updated 5th
edition of Shnider and Levinsons Anesthesia for Obstetrics will
significantly help with this.
It has been 11 yr since the 4th edition of Shnider and
Levinsons Anesthesia for Obstetrics was released. There have
been advances in the field of obstetric anesthesiology since
then, focusing on neonatal resuscitation and management
of neonatal neurologic injury. The editors in the new edition
were eager to improve the content by providing the most upto-date information, while preserving the original quality of
the text as they did in the 4th edition.
review.
The updated edition is divided into 11 parts. The

first section reviews physiology and pharmacology.
The initial topic focuses on the areas of uteroplacental circulation, physiological changes of pregnancy,
and placental transfer of drugs. The second, third, and
fourth parts cover assessment of fetus, labor analgesia, and anesthesia for cesarean delivery. This goes in
depth looking at intrapartum fetal monitoring as well
as regional anesthesia techniques. The fifth, sixth, and
seventh portions cover neonatal concepts of wellbeing,
anesthetic considerations for obstetric complications,
and management of anesthetic complications. There
is in depth look at abnormal fetal positions, obstetric
hemorrhage, NPO controversies, and postdural puncture headaches management.
The latter sections cover anesthetic management of parturients, ethical challenges, maternal morbidity and mortality.
Very difficult topics we face are covered, including the Jehovahs Witness patient, trauma resuscitation, and the global
perspective of obstetric anesthesia. Some detail is provided
for in vitro fertilization and nonobstetric surgery during
pregnancy. Fortunately, given its emergence and growing
popularity, the newer, ex utero intrapartum therapy surgical
procedure is discussed.
The appendix has updated guidelines for neuraxial anesthesia and practice guidelines for obstetric anesthesia along
with intrapartum fetal heart rate monitoring.
Shnider and Levinsons chapters are well organized, contain excellent depth, and are informative. The online access
offers a full version site with full images to view.
What sets this text apart is the detail in each of its chapters. For those of us who cover obstetric anesthesiology regularly while on call, a review of it is a must. This is an excellent
resource for the clinician in training, such as residency or
fellowship, or in practice. The chapters are detailed, with
complete images, and online access make this a must own
reference. The only area for future improvement would be
adding ultrasound-guided neuraxial techniques, given the
trend in anesthesiology.
I strongly recommend this book as a priceless addition to
the library of any well-rounded anesthesiologist who covers
obstetrics full time or part time.
Abed Rahman M.D., M.S., Oak Forest Health Center and
Stroger Hospital Cook County and Rush University Medical
College, Chicago, Illinois. arahman2@cookcountyhhs.org
anesthesiologists who wish to quickly discern what is known

and what is not yet known about the fetal effects of anesthesia. In at least a half dozen of the chapters, the authors
have undertaken a systematic review and present tables of
relevant randomized studies in addition to summarizing and
synthesizing their findings. For the busy clinician, the chapter conclusions are a rapid way to systematically sample the
content of each chapter throughout the entire book.
Expertly administered anesthesia offers tremendous potential to improve the comfort of childbirth and obstetric surgery
while preserving both fetal and maternal well-being. Increasingly, pregnant women face a barrage of well-meaning advice
to limit environmental and pharmacologic exposures for their
growing fetus. Consequently, to be effective, anesthesiologists
need to do more than simply administer safe anesthesia; we
must serve as ambassadors and educators with comprehensive knowledge of the effects of our interventions on both the
mother and fetus. For clinicians who seek to fulfil this obligation, Anesthesia and the Fetus will be an excellent resource.
Baskar Rajala, M.B.B.S., F.R.C.A., Jill M. Mhyre, M.D.
University of Arkansas for Medical Sciences, Little Rock, Arkansas ( J.M.M). jmmhyre@uams.edu
Shnider and Levinsons Anesthesia for

Obstetrics, Fifth Edition. Edited by Maya S.
Suresh, M.D., B. Scott Segal, M.D., M.H.C.M.,
Roanne L. Preston, M.D., F.R.C.P.C., Roshan
Fernando, M.B., B.Ch., F.R.C.A., C. LaToya Mason.,
M.D. Philadephia, Lippincott Williams & Wilkins,
2013. Pages: 902. Price: $232.99.
Most practicing physician anesthesiologists in community
hospitals and teaching hospitals are obliged to cover obstetrical anesthesia during the work day and call coverage, which
may be the most stressful part of a practice. Keeping up with
the literature specific to obstetric anesthesia may be a challenge to the busy practitioner. The revised and updated 5th
edition of Shnider and Levinsons Anesthesia for Obstetrics will
significantly help with this.
It has been 11 yr since the 4th edition of Shnider and
Levinsons Anesthesia for Obstetrics was released. There have
been advances in the field of obstetric anesthesiology since
then, focusing on neonatal resuscitation and management
of neonatal neurologic injury. The editors in the new edition
were eager to improve the content by providing the most upto-date information, while preserving the original quality of
the text as they did in the 4th edition.
review.
The updated edition is divided into 11 parts. The

first section reviews physiology and pharmacology.
The initial topic focuses on the areas of uteroplacental circulation, physiological changes of pregnancy,
and placental transfer of drugs. The second, third, and
fourth parts cover assessment of fetus, labor analgesia, and anesthesia for cesarean delivery. This goes in
depth looking at intrapartum fetal monitoring as well
as regional anesthesia techniques. The fifth, sixth, and
seventh portions cover neonatal concepts of wellbeing,
anesthetic considerations for obstetric complications,
and management of anesthetic complications. There
is in depth look at abnormal fetal positions, obstetric
hemorrhage, NPO controversies, and postdural puncture headaches management.
The latter sections cover anesthetic management of parturients, ethical challenges, maternal morbidity and mortality.
Very difficult topics we face are covered, including the Jehovahs Witness patient, trauma resuscitation, and the global
perspective of obstetric anesthesia. Some detail is provided
for in vitro fertilization and nonobstetric surgery during
pregnancy. Fortunately, given its emergence and growing
popularity, the newer, ex utero intrapartum therapy surgical
procedure is discussed.
The appendix has updated guidelines for neuraxial anesthesia and practice guidelines for obstetric anesthesia along
with intrapartum fetal heart rate monitoring.
Shnider and Levinsons chapters are well organized, contain excellent depth, and are informative. The online access
offers a full version site with full images to view.
What sets this text apart is the detail in each of its chapters. For those of us who cover obstetric anesthesiology regularly while on call, a review of it is a must. This is an excellent
resource for the clinician in training, such as residency or
fellowship, or in practice. The chapters are detailed, with
complete images, and online access make this a must own
reference. The only area for future improvement would be
adding ultrasound-guided neuraxial techniques, given the
trend in anesthesiology.
I strongly recommend this book as a priceless addition to
the library of any well-rounded anesthesiologist who covers
obstetrics full time or part time.
Abed Rahman M.D., M.S., Oak Forest Health Center and
Stroger Hospital Cook County and Rush University Medical
College, Chicago, Illinois. arahman2@cookcountyhhs.org
ANNOUNCEMENTS
ANNOUNCEMENT AND CALL FOR ABSTRACTS
2015 ANNUAL JOURNAL SYMPOSIUM

The Anesthesiologist and Health Care Redesign
The American Society of Anesthesiologists (ASA) and its journal, Anesthesiology, announce the 24th annual Journal Symposium to be held at the ASA Annual Meeting on October 27, 2015, in San Diego, California. The 2015 Journal Symposium
will highlight feature the following presentations:
Presentation 1: Using Data to Improve Operating Room Throughput
Peter Dunn, M.D., Executive Medical Director of Perioperative Services, Massachusetts General Hospital
Retsef Levi, Ph.D., J. Spencer Standish Professor of Operations Management, Sloan School of Management, Massachusetts
Institute of Technology
Presentation 2: Redesigning Surgical Patient Care from Decision to Discharge
Jonathan Wanderer, M.D., M.Phil., Assistant Professor of Anesthesiology, Vanderbilt University School of Medicine
Presentation 3: Designing a New Hospital for Surgical Care
Brett Simon, M.D., Ph.D., Director of Memorial Sloan Ketterings Josie Robertson Surgery Center
A common axiom in architecture reads: If you want to change the way people work, change the building they work in. The
2015 Journal Symposium addresses the changes in healthcare design and delivery head on, with this axiom setting the context.
Anesthesiologists that are involved in the world of perioperative and procedural medicine are facing intense pressure to alter
systems of care to improve the value delivered to patients. Such pressures are not new, but the intensity of the pressure has risen
dramatically. The advent of the Affordable Care Act and payer initiatives like bundling and value-based purchasing provide
new impetus for healthcare delivery redesign. Recent descriptive publications have detailed elements of a nascent Perioperative Surgical Home. Does this construct represent a true reconsideration of perioperative healthcare delivery? How much of
redesign is required to truly affect the value a perioperative system brings to patients? Does the value accrue to patients, or does
the value of redesigning perioperative systems actually benefit healthcare systems and payers?
Four experts will introduce these topics for the first 90 minutes of the symposium, with 20-minute presentations and 10-minute discussions. The speakers will discuss three different approaches, of increasing scope, for redesigning healthcare delivery
systems to meet specific goals. Where available, empirical tests of the redesign efforts success at meeting their goals will be
presented.
We invite abstracts on this topic from all related fields: basic science, translational, clinical, and quality research. The top eight
abstracts will be presented orally during the second half of the symposium.
The authors of abstracts selected for the symposium will be offered an opportunity to submit their work to ANESTHESIOLOGY
for inclusion in an issue to be published in the spring of 2016.
224
January 2015
ANNOUNCEMENTS
NOMINATIONS SOUGHT FOR
2015 Award for Excellence in Research

The annual American Society of Anesthesiologists (ASA) Award for Excellence in Research recognizes an individual for outstanding achievement in research that has or is likely to have an important impact on the practice of anesthesiology.
The individuals work must represent a body of original, mature and sustained contribution to the advancement of the science
of anesthesiology. The nominee need not be a physician, an anesthesiologist, or a member of the ASA, but must be presently engaged in research related to anesthesiology, academically accomplished with peer-reviewed publications and funded
research, and nominated in response to a call for nominations. The completed application must include the nominees current
curriculum vitae, a letter of nomination, and a seconding letter from two individuals with an understanding of the research
contributions of the individual.
The 2014 Award for Excellence in Research was presented to Henrik Kehlet, M.D., Ph.D., at the ASA Annual Meeting in New
Orleans, on Monday, October 13. Dr. Kehlet is a Professor at Rigshospitalet, Copenhagen University, Copenhagen, Denmark.
The deadline for nominations for the 2015 Award for Excellence in Research is March 31, 2015. Please submit nominations
or any questions regarding this award to Vicki Vass Tedeschi, Managing Editor, ANESTHESIOLOGY; e-mail: managing-editor@
anesthesiology.org.
NOMINATIONS SOUGHT FOR

2015 Presidential Scholar Award
The American Society of Anesthesiologists (ASA) Presidential Scholar Award recognizes colleagues who dedicate their formative careers to research.
The deadline for nominations for the 2015 Presidential Scholar Award is March 31, 2015. Anesthesiologists who are within
10 years of their first appointment to a department of anesthesiology, who are Board-certified in their country of practice,
who are ASA members, and who are clinically active in anesthesia, intensive care, or pain medicine are eligible for the award.
Nominees must be academically accomplished with peer-reviewed publications and funded research. Candidates should be
nominated by their department chair or by the Committee on Research after review of the current years grant applicants of
the Foundation for Anesthesia Education and Research. The nominees department chair should submit a letter of support
and the nominees current curriculum vitae as well as one seconding letter from a senior faculty member. Only one nominee
per department will be accepted.
The recipient of the 2014 Presidential Scholar Award was Rebecca A. Aslakson, M.D., Ph.D., Assistant Professor, Johns Hopkins University School of Medicine, Department of Anesthesiology and Critical Care Medicine, Baltimore, Maryland, who
received the award at the ASA Annual Meeting in New Orleans, on Monday, October 13, 2014.
Please submit nominations or any questions regarding this award to Vicki Vass Tedeschi, Managing Editor, ANESTHESIOLOGY;
e-mail: managing-editor@anesthesiology.org.
225 Announcements
Announcements
FAER Call for Grant Applications

The board of directors of the Foundation for Anesthesia Education and Research (FAER) is pleased to
announce FAERs 2015 research grant funding opportunities for faculty members and trainees. The
application deadline is February 15, 2015.
FAER provides research grant funding for anesthesiologists and anesthesiology trainees to gain additional training in basic science, clinical and translational, health services, and education research. For early-career anesthesiologists interested in pursuing careers as physician-scientists, FAER grants can be an important starting point. These grants aim to help anesthesiologists
develop the skills and preliminary data they need to become independent investigators.
2015 Funding Opportunities

The following research grant funding opportunities are available to anesthesiologists and anesthesiology trainees. The application Web site for the 2015 grant funding cycle is open November 1, 2014, through February 15, 2015.
For more information regarding FAER grants and eligibility requirements, visit FAER.org/research-grants or e-mail Jody
Clikeman, Grant Program Coordinator, at JodyClikeman@faer.org.
Mentored Research Training Grants
Research Areas: Basic Science (MRTG-BS), Clinical and Translational (MRTG-CT), Health Services Research (MRTG-HSR)
Purpose: To help physician anesthesiologists develop the skills and preliminary data to become independent investigators
For Whom: Faculty members who completed their initial core anesthesiology residency within the past 10 years
Funding: $175,000
Duration: Two years
Percent Research: 75%
Research Fellowship Grant
Research Areas: Basic Science, Clinical and Translational, Health Services or Education
Purpose: To provide significant training in research techniques and scientific methods
For Whom: Anesthesiology trainee after the CA-1 year
Funding: $75,000
Duration: One year
Research in Education Grant
Research Areas: Education Research
Purpose: To improve the quality and impact of anesthesiology education research
For Whom: Faculty member of any rank (junior or senior faculty)
Funding: $100,000
Duration: Two years
226 Announcements
ANNOUNCEMENTS
Research Grant Application Deadlines

Online application opens November 1, 2014
Applications due February 15, 2015
Award notifications made by May 15, 2015
Project start date July 1, 2015, or January 1, 2016
Research Grant Eligibility CriteriaUpdated for 2015

The FAER Grant Management Committee has made a few changes and clarifications to the eligibility criteria and rules for
research grant funding from previous years.
Applicants may submit only one grant application per award cycle.
Tuition is not allowed in the budget for any grant.
The applicant and the primary mentor for the grant must be at the same institution.
To view the complete eligibility requirements and application guide, visit FAER.org/research-grants.
227 Announcements

Anestheology Latest PDF

Uploaded by

Document Information

Original Description:

Original Title

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

Anestheology Latest PDF

Uploaded by

Copyright:

Available Formats

balt6/z7i-anesth/z7i-anesth/z7i00812/contents panickes S12 7/2/12 22:40 Art: toc Input-citi

et al.: Characterizing the Epidemiology of Perioperative Transfusion-associated

Pinheiro de Almeida et al.: Transfusion Requirements in Surgical Oncology Patients:

. A Public-Private Strategy to Advance the Use of Clinical Registries

Science, Medicine, and the Anesthesiologist 

. Making Sedation Safer: Is Simulation the Answer?

and Donald M. Berwick

. What Are We Looking For? The Question of Resident Selection

Pay Toll on the Bridge from Innate Immunity to Ventilator-induced

J.W.Simmons and J.-F.Pittet

Ultrasound versus Fluoroscopy in Image-guided Pain Treatment: Use Caution

Anaphylaxis to Neuromuscular-blocking Drugs: All Neuromuscular-blocking Drugs

Big Brain, Small World? 

 Refers to This Month in Anesthesiology

Characterizing the Epidemiology of Postoperative Transfusion-related Acute Lung Injury 

L.Clifford, Q.Jia, A.Subramanian, H.Yadav, G.A.Wilson, S.P.Murphy, J.Pathak,

Characterizing the Epidemiology of Perioperative Transfusion-associated Circulatory

L.Clifford, Q.Jia, H.Yadav, A.Subramanian, G.A.Wilson, S.P.Murphy, J.Pathak,

J. Pinheiro deAlmeida, J.-L.Vincent, F. R. B. GomesGalas, E.PintoMarinhodeAlmeida,

Anaphylaxis Is More Common with Rocuronium and Succinylcholine than with

J.I.Reddy, P.J.Cooke, J.M.vanSchalkwyk, J.A.Hannam, P.Fitzharris, and S.J.Mitchell

Postoperative Bladder Catheterization Based on Individual Bladder Capacity:

T.A.Brouwer, P.F.W.M.Rosier, K.G.M.Moons, N.P.A.Zuithoff, E.N.vanRoon, and C.J.Kalkman

Accuracy of Malignant Hyperthermia Diagnoses in Hospital Discharge Records 

T.Pinyavat, H.Rosenberg, B.H.Lang, C.A.Wong, S.Riazi, J.E.Brady, L.S.Sun, and G.Li

Relationship between Chronic Intermittent Hypoxia and Intraoperative Mean Arterial

Long-term Effects of Single or Multiple Neonatal Sevoflurane Exposures on Rat

L.G.Amrock, M.L.Starner, K.L.Murphy, and M.G.Baxter

CRITICAL CARE MEDICINE

Modulation of Stress versus Time Product during Mechanical Ventilation Influences

P.M.Spieth, P.L.Silva, C.S.N.B.Garcia, D.S.Ornellas, C.S.Samary, L.Moraes,

T. Fink, A. Wolf, F. Maurer, F. W. Albrecht, N. Heim, B. Wolf, A. C. Hauschild, B. Bdeker,

Extracellular Histones Play an Inflammatory Role in Acid Aspiration-induced Acute

Y. Zhang, Z. Wen, L. Guan, P. Jiang, T. Gu, J. Zhao, X. Lv, and T. Wen

Disruption of Cortical Connectivity during Remifentanil Administration Is Associated

A. Khodayari-Rostamabad, S. S. Olesen, C. Graversen, L. P. Malver, G. P. Kurita, P. Sjgren,

Immediate Rescue Designs in Pediatric Analgesic Trials: A Systematic Review and

J. Kossowsky, C. Donado, and C. B. Berde

A Randomized Control Trial of Bupivacaine and Fentanyl versus Fentanyl-only for

M. G. Craig, E. N. Grant, W. Tao, D. D. McIntire, and K. J. Leveno

Y. Fujimoto, T. Funao, K. Suehiro, R. Takahashi, T. Mori, and K. Nishikawa

Inversion of the Right Hemidiaphragm due to Massive

A. F. Simpao, J. A. Galvez, A. Jay Schwartz, and M. A. Rehman

Clinical Concepts and Commentary

Pretransfusion Testing and Transfusion of Uncrossmatched Erythrocytes 

M. L. Boisen, R. A. Collins, M. H. Yazer, and J. H. Waters

Regulation of Cerebral Autoregulation by Carbon Dioxide 

Is the Standard Supplied by the Association for the Advancement of Medical

J. Fortin, K. Lerche, D. Flotzinger,and T. OBrien

Inotrope Use in Cardiac Surgery: ACause of Worse Outcomes, or Just a Marker

B. G. Maxwell, J. O. Wasey, and E. S. Heitmiller

Lung Ultrasonography for the Detection of Anesthesia-induced Lung Atelectasis 

M. Girard, V.Gnreux, and A.Monastesse

Old Guidelines or Methods Cannot Insure Quality or Progress 

ANESTHESIOLOGY REFLECTIONS FROM

Katz Oxygen Treatment for Catarrh

Laughing Gas from Knoxvilles Dr. H. F. Huffaker

Figuiers Forlorn Figure: Horace Wells and the Humbug Affair

REVIEWS OF EDUCATIONAL MATERIAL

CAREERS & EVENTS

INSTRUCTIONS FOR AUTHORS

Address for non-member subscription information, orders, or change of address:

balt6/z7i-anesth/z7i-anesth/z7i00812/contents panickes S12 7/2/12 22:40 Art: toc Input-citi

Pinheiro de Almeida et al.: Transfusion Requirements in Surgical Oncology Patients:

Science, Medicine, and the Anesthesiologist

Big Brain, Small World?

Refers to This Month in Anesthesiology

Characterizing the Epidemiology of Postoperative Transfusion-related Acute Lung Injury

Accuracy of Malignant Hyperthermia Diagnoses in Hospital Discharge Records

Pretransfusion Testing and Transfusion of Uncrossmatched Erythrocytes

Regulation of Cerebral Autoregulation by Carbon Dioxide

Lung Ultrasonography for the Detection of Anesthesia-induced Lung Atelectasis

Old Guidelines or Methods Cannot Insure Quality or Progress

Katz Oxygen Treatment for Catarrh

Laughing Gas from Knoxvilles Dr. H. F. Huffaker

Figuiers Forlorn Figure: Horace Wells and the Humbug Affair

REVIEWS OF EDUCATIONAL MATERIAL

CAREERS & EVENTS

balt6/z7i-anesth/z7i-anesth/z7i00812/thismonth panickes S245 6/29/12 22:41 Art:

46Postoperative Bladder Catheterization Based on Individual Bladder Capacity: A Randomized Trial

140Disruption of Cortical Connectivity during Remifentanil Administration Is

196Regulation of Cerebral Autoregulation by Carbon Dioxide (Review Article)

191Pretransfusion Testing and Transfusion of Uncrossmatched Erythrocytes

balt6/z7i-anesth/z7i-anesth/z7i00812/thismonth panickes S245 6/29/12 22:41 Art:

balt6/z7i-anesth/z7i-anesth/z7i00812/thismonth panickes S245 6/29/12 22:41 Art: