Zab Mohsenifar, Guy W. Soo Hoo Lung Biology in Health & Disease Volume 213 Practical Pulmonary and Critical Care Medicine Respiratory Failure PDF

Practical Pulmonary
and Critical Care Medicine

LUNG BIOLOGY IN HEALTH AND DISEASE
Executive Editor
Claude Lenfant
Former Director, National Heart, Lung, and Blood Institute
National Institutes of Health
Bethesda, Maryland
1. Immunologic and Infectious Reactions in the Lung,

edited by C. H. Kirkpatrick and H. Y. Reynolds
2. The Biochemical Basis of Pulmonary Function, edited by
R. G. Crystal
3. Bioengineering Aspects of the Lung, edited by J. B. West
4. Metabolic Functions of the Lung, edited by Y. S. Bakhle
and J. R. Vane
5. Respiratory Defense Mechanisms (in two parts), edited by
J. D. Brain, D. F. Proctor, and L. M. Reid
6. Development of the Lung, edited by W. A. Hodson
7. Lung Water and Solute Exchange, edited by N. C. Staub
8. Extrapulmonary Manifestations of Respiratory Disease,
edited by E. D. Robin
9. Chronic Obstructive Pulmonary Disease, edited by T. L. Petty
10. Pathogenesis and Therapy of Lung Cancer, edited by
C. C. Harris
11. Genetic Determinants of Pulmonary Disease, edited by
S. D. Litwin
12. The Lung in the Transition Between Health and Disease,
edited by P. T. Macklem and S. Permutt
13. Evolution of Respiratory Processes: A Comparative Approach,
edited by S. C. Wood and C. Lenfant
14. Pulmonary Vascular Diseases, edited by K. M. Moser
15. Physiology and Pharmacology of the Airways, edited by
J. A. Nadel
16. Diagnostic Techniques in Pulmonary Disease (in two parts),
edited by M. A. Sackner
17. Regulation of Breathing (in two parts), edited by T. F.
Hornbein
18. Occupational Lung Diseases: Research Approaches
and Methods, edited by H. Weill and M. Turner-Warwick
19. Immunopharmacology of the Lung, edited by H. H. Newball
20. Sarcoidosis and Other Granulomatous Diseases of the Lung,
edited by B. L. Fanburg
21. Sleep and Breathing, edited by N. A. Saunders
and C. E. Sullivan
22. Pneumocystis carinii Pneumonia: Pathogenesis, Diagnosis,
and Treatment, edited by L. S. Young
23. Pulmonary Nuclear Medicine: Techniques in Diagnosis of
Lung Disease, edited by H. L. Atkins
24. Acute Respiratory Failure, edited by W. M. Zapol
and K. J. Falke
25. Gas Mixing and Distribution in the Lung, edited by L. A. Engel
and M. Paiva
26. High-Frequency Ventilation in Intensive Care and During
Surgery, edited by G. Carlon and W. S. Howland
27. Pulmonary Development: Transition from Intrauterine to
Extrauterine Life, edited by G. H. Nelson
28. Chronic Obstructive Pulmonary Disease: Second Edition,
edited by T. L. Petty
29. The Thorax (in two parts), edited by C. Roussos
and P. T. Macklem
30. The Pleura in Health and Disease, edited by J. Chrtien,
J. Bignon, and A. Hirsch
31. Drug Therapy for Asthma: Research and Clinical Practice,
edited by J. W. Jenne and S. Murphy
32. Pulmonary Endothelium in Health and Disease, edited by
U. S. Ryan
33. The Airways: Neural Control in Health and Disease,
edited by M. A. Kaliner and P. J. Barnes
34. Pathophysiology and Treatment of Inhalation Injuries,
edited by J. Loke
35. Respiratory Function of the Upper Airway, edited by
O. P. Mathew and G. SantAmbrogio
36. Chronic Obstructive Pulmonary Disease: A Behavioral
Perspective, edited by A. J. McSweeny and I. Grant
37. Biology of Lung Cancer: Diagnosis and Treatment, edited by
S. T. Rosen, J. L. Mulshine, F. Cuttitta, and P. G. Abrams
38. Pulmonary Vascular Physiology and Pathophysiology,
edited by E. K. Weir and J. T. Reeves
39. Comparative Pulmonary Physiology: Current Concepts,
edited by S. C. Wood
40. Respiratory Physiology: An Analytical Approach,
edited by H. K. Chang and M. Paiva
41. Lung Cell Biology, edited by D. Massaro
42. HeartLung Interactions in Health and Disease,
edited by S. M. Scharf and S. S. Cassidy
43. Clinical Epidemiology of Chronic Obstructive Pulmonary
Disease, edited by M. J. Hensley and N. A. Saunders
44. Surgical Pathology of Lung Neoplasms, edited by
A. M. Marchevsky
45. The Lung in Rheumatic Diseases, edited by G. W. Cannon
and G. A. Zimmerman
46. Diagnostic Imaging of the Lung, edited by C. E. Putman
47. Models of Lung Disease: Microscopy and Structural
Methods,
edited by J. Gil
48. Electron Microscopy of the Lung, edited by D. E.
Schraufnagel
49. Asthma: Its Pathology and Treatment, edited by M. A. Kaliner,
P. J. Barnes, and C. G. A. Persson
50. Acute Respiratory Failure: Second Edition, edited by
W. M. Zapol and F. Lemaire
51. Lung Disease in the Tropics, edited by O. P. Sharma
52. Exercise: Pulmonary Physiology and Pathophysiology,
edited by B. J. Whipp and K. Wasserman
53. Developmental Neurobiology of Breathing, edited by
G. G. Haddad and J. P. Farber
54. Mediators of Pulmonary Inflammation, edited by M. A. Bray
and W. H. Anderson
55. The Airway Epithelium, edited by S. G. Farmer and D. Hay
56. Physiological Adaptations in Vertebrates: Respiration,
Circulation, and Metabolism, edited by S. C. Wood,
R. E. Weber, A. R. Hargens, and R. W. Millard
57. The Bronchial Circulation, edited by J. Butler
58. Lung Cancer Differentiation: Implications for Diagnosis
and Treatment, edited by S. D. Bernal and P. J. Hesketh
59. Pulmonary Complications of Systemic Disease, edited by
J. F. Murray
60. Lung Vascular Injury: Molecular and Cellular Response,
edited by A. Johnson and T. J. Ferro
61. Cytokines of the Lung, edited by J. Kelley
62. The Mast Cell in Health and Disease, edited by M. A. Kaliner
and D. D. Metcalfe
63. Pulmonary Disease in the Elderly Patient, edited by
D. A. Mahler
64. Cystic Fibrosis, edited by P. B. Davis
65. Signal Transduction in Lung Cells, edited by J. S. Brody,
D. M. Center, and V. A. Tkachuk
66. Tuberculosis: A Comprehensive International Approach,
edited by L. B. Reichman and E. S. Hershfield
67. Pharmacology of the Respiratory Tract: Experimental
and Clinical Research, edited by K. F. Chung and P. J. Barnes
68. Prevention of Respiratory Diseases, edited by A. Hirsch,
M. Goldberg, J.-P. Martin, and R. Masse
69. Pneumocystis carinii Pneumonia: Second Edition, edited by
P. D. Walzer
70. Fluid and Solute Transport in the Airspaces of the Lungs,
edited by R. M. Effros and H. K. Chang
71. Sleep and Breathing: Second Edition, edited by
N. A. Saunders and C. E. Sullivan
72. Airway Secretion: Physiological Bases for the Control
of Mucous Hypersecretion, edited by T. Takishima
and S. Shimura
73. Sarcoidosis and Other Granulomatous Disorders, edited by
D. G. James
74. Epidemiology of Lung Cancer, edited by J. M. Samet
75. Pulmonary Embolism, edited by M. Morpurgo
76. Sports and Exercise Medicine, edited by S. C. Wood
and R. C. Roach
77. Endotoxin and the Lungs, edited by K. L. Brigham
78. The Mesothelial Cell and Mesothelioma, edited by
M.-C. Jaurand and J. Bignon
79. Regulation of Breathing: Second Edition, edited by
J. A. Dempsey and A. I. Pack
80. Pulmonary Fibrosis, edited by S. Hin. Phan and R. S. Thrall
81. Long-Term Oxygen Therapy: Scientific Basis and Clinical
Application, edited by W. J. ODonohue, Jr.
82. Ventral Brainstem Mechanisms and Control of Respiration
and Blood Pressure, edited by C. O. Trouth, R. M. Millis,
H. F. Kiwull-Schne, and M. E. Schlfke
83. A History of Breathing Physiology, edited by D. F. Proctor
84. Surfactant Therapy for Lung Disease, edited by B. Robertson
and H. W. Taeusch
85. The Thorax: Second Edition, Revised and Expanded (in three
parts), edited by C. Roussos
86. Severe Asthma: Pathogenesis and Clinical Management,
edited by S. J. Szefler and D. Y. M. Leung
87. Mycobacterium aviumComplex Infection: Progress in
Research and Treatment, edited by J. A. Korvick
and C. A. Benson
88. Alpha 1Antitrypsin Deficiency: Biology Pathogenesis
Clinical Manifestations Therapy, edited by R. G. Crystal
89. Adhesion Molecules and the Lung, edited by P. A. Ward
and J. C. Fantone
90. Respiratory Sensation, edited by L. Adams and A. Guz
91. Pulmonary Rehabilitation, edited by A. P. Fishman
92. Acute Respiratory Failure in Chronic Obstructive Pulmonary
Disease, edited by J.-P. Derenne, W. A. Whitelaw,
and T. Similowski
93. Environmental Impact on the Airways: From Injury to Repair,
edited by J. Chrtien and D. Dusser
94. Inhalation Aerosols: Physical and Biological Basis for
Therapy, edited by A. J. Hickey
95. Tissue Oxygen Deprivation: From Molecular to Integrated
Function, edited by G. G. Haddad and G. Lister
96. The Genetics of Asthma, edited by S. B. Liggett
and D. A. Meyers
97. Inhaled Glucocorticoids in Asthma: Mechanisms and Clinical
Actions, edited by R. P. Schleimer, W. W. Busse,
and P. M. OByrne
98. Nitric Oxide and the Lung, edited by W. M. Zapol
and K. D. Bloch
99. Primary Pulmonary Hypertension, edited by L. J. Rubin
and S. Rich
100. Lung Growth and Development, edited by J. A. McDonald
101. Parasitic Lung Diseases, edited by A. A. F. Mahmoud
102. Lung Macrophages and Dendritic Cells in Health and Disease,
edited by M. F. Lipscomb and S. W. Russell
103. Pulmonary and Cardiac Imaging, edited by C. Chiles
and C. E. Putman
104. Gene Therapy for Diseases of the Lung, edited by
K. L. Brigham
105. Oxygen, Gene Expression, and Cellular Function, edited by
L. Biadasz Clerch and D. J. Massaro
106. Beta2-Agonists in Asthma Treatment, edited by R. Pauwels
and P. M. OByrne
107. Inhalation Delivery of Therapeutic Peptides and Proteins,
edited by A. L. Adjei and P. K. Gupta
108. Asthma in the Elderly, edited by R. A. Barbee and J. W.
Bloom
109. Treatment of the Hospitalized Cystic Fibrosis Patient,
edited by D. M. Orenstein and R. C. Stern
110. Asthma and Immunological Diseases in Pregnancy and Early
Infancy, edited by M. Schatz, R. S. Zeiger, and H. N. Claman
111. Dyspnea, edited by D. A. Mahler
112. Proinflammatory and Antiinflammatory Peptides, edited by
S. I. Said
113. Self-Management of Asthma, edited by H. Kotses
and A. Harver
114. Eicosanoids, Aspirin, and Asthma, edited by A. Szczeklik,
R. J. Gryglewski, and J. R. Vane
115. Fatal Asthma, edited by A. L. Sheffer
116. Pulmonary Edema, edited by M. A. Matthay and D. H. Ingbar
117. Inflammatory Mechanisms in Asthma, edited by S. T. Holgate
and W. W. Busse
118. Physiological Basis of Ventilatory Support, edited by
J. J. Marini and A. S. Slutsky
119. Human Immunodeficiency Virus and the Lung, edited by
M. J. Rosen and J. M. Beck
120. Five-Lipoxygenase Products in Asthma, edited by
J. M. Drazen, S.-E. Dahln, and T. H. Lee
121. Complexity in Structure and Function of the Lung, edited by
M. P. Hlastala and H. T. Robertson
122. Biology of Lung Cancer, edited by M. A. Kane
and P. A. Bunn, Jr.
123. Rhinitis: Mechanisms and Management, edited by
R. M. Naclerio, S. R. Durham, and N. Mygind
124. Lung Tumors: Fundamental Biology and Clinical
Management, edited by C. Brambilla and E. Brambilla
125. Interleukin-5: From Molecule to Drug Target for Asthma,
edited by C. J. Sanderson
126. Pediatric Asthma, edited by S. Murphy and H. W. Kelly
127. Viral Infections of the Respiratory Tract, edited by R. Dolin
and P. F. Wright
128. Air Pollutants and the Respiratory Tract, edited by D. L. Swift
and W. M. Foster
129. Gastroesophageal Reflux Disease and Airway Disease,
edited by M. R. Stein
130. Exercise-Induced Asthma, edited by E. R. McFadden, Jr.
131. LAM and Other Diseases Characterized by Smooth Muscle
Proliferation, edited by J. Moss
132. The Lung at Depth, edited by C. E. G. Lundgren
and J. N. Miller
133. Regulation of Sleep and Circadian Rhythms, edited by
F. W. Turek and P. C. Zee
134. Anticholinergic Agents in the Upper and Lower Airways,
edited by S. L. Spector
135. Control of Breathing in Health and Disease, edited by
M. D. Altose and Y. Kawakami
136. Immunotherapy in Asthma, edited by J. Bousquet
and H. Yssel
137. Chronic Lung Disease in Early Infancy, edited by R. D. Bland
and J. J. Coalson
138. Asthmas Impact on Society: The Social and Economic
Burden, edited by K. B. Weiss, A. S. Buist, and S. D. Sullivan
139. New and Exploratory Therapeutic Agents for Asthma,
edited by M. Yeadon and Z. Diamant
140. Multimodality Treatment of Lung Cancer, edited by
A. T. Skarin
141. Cytokines in Pulmonary Disease: Infection and Inflammation,
edited by S. Nelson and T. R. Martin
142. Diagnostic Pulmonary Pathology, edited by P. T. Cagle
143. ParticleLung Interactions, edited by P. Gehr and J. Heyder
144. Tuberculosis: A Comprehensive International Approach,
Second Edition, Revised and Expanded, edited by
L. B. Reichman and E. S. Hershfield
145. Combination Therapy for Asthma and Chronic Obstructive
Pulmonary Disease, edited by R. J. Martin and M. Kraft
146. Sleep Apnea: Implications in Cardiovascular
and Cerebrovascular Disease, edited by T. D. Bradley
and J. S. Floras
147. Sleep and Breathing in Children: A Developmental Approach,
edited by G. M. Loughlin, J. L. Carroll, and C. L. Marcus
148. Pulmonary and Peripheral Gas Exchange in Health
and Disease, edited by J. Roca, R. Rodriguez-Roisen,
and P. D. Wagner
149. Lung Surfactants: Basic Science and Clinical Applications,
R. H. Notter
150. Nosocomial Pneumonia, edited by W. R. Jarvis
151. Fetal Origins of Cardiovascular and Lung Disease, edited by
David J. P. Barker
152. Long-Term Mechanical Ventilation, edited by N. S. Hill
153. Environmental Asthma, edited by R. K. Bush
154. Asthma and Respiratory Infections, edited by D. P. Skoner
155. Airway Remodeling, edited by P. H. Howarth, J. W. Wilson,
J. Bousquet, S. Rak, and R. A. Pauwels
156. Genetic Models in Cardiorespiratory Biology, edited by
G. G. Haddad and T. Xu
157. Respiratory-Circulatory Interactions in Health and Disease,
edited by S. M. Scharf, M. R. Pinsky, and S. Magder
158. Ventilator Management Strategies for Critical Care, edited by
N. S. Hill and M. M. Levy
159. Severe Asthma: Pathogenesis and Clinical Management,
Second Edition, Revised and Expanded, edited by
S. J. Szefler and D. Y. M. Leung
160. Gravity and the Lung: Lessons from Microgravity, edited by
G. K. Prisk, M. Paiva, and J. B. West
161. High Altitude: An Exploration of Human Adaptation, edited by
T. F. Hornbein and R. B. Schoene
162. Drug Delivery to the Lung, edited by H. Bisgaard,
C. OCallaghan, and G. C. Smaldone
163. Inhaled Steroids in Asthma: Optimizing Effects in the Airways,
edited by R. P. Schleimer, P. M. OByrne, S. J. Szefler,
and R. Brattsand
164. IgE and Anti-IgE Therapy in Asthma and Allergic Disease,
edited by R. B. Fick, Jr., and P. M. Jardieu
165. Clinical Management of Chronic Obstructive Pulmonary
Disease, edited by T. Similowski, W. A. Whitelaw,
and J.-P. Derenne
166. Sleep Apnea: Pathogenesis, Diagnosis, and Treatment,
edited by A. I. Pack
167. Biotherapeutic Approaches to Asthma, edited by J. Agosti
and A. L. Sheffer
168. Proteoglycans in Lung Disease, edited by H. G. Garg,
P. J. Roughley, and C. A. Hales
169. Gene Therapy in Lung Disease, edited by S. M. Albelda
170. Disease Markers in Exhaled Breath, edited by N. Marczin,
S. A. Kharitonov, M. H. Yacoub, and P. J. Barnes
171. Sleep-Related Breathing Disorders: Experimental Models
and Therapeutic Potential, edited by D. W. Carley
and M. Radulovacki
172. Chemokines in the Lung, edited by R. M. Strieter,
S. L. Kunkel, and T. J. Standiford
173. Respiratory Control and Disorders in the Newborn,
edited by O. P. Mathew
174. The Immunological Basis of Asthma, edited by
B. N. Lambrecht, H. C. Hoogsteden, and Z. Diamant
175. Oxygen Sensing: Responses and Adaptation to Hypoxia,
edited by S. Lahiri, G. L. Semenza, and N. R. Prabhakar
176. Non-Neoplastic Advanced Lung Disease, edited by
J. R. Maurer
177. Therapeutic Targets in Airway Inflammation, edited by
N. T. Eissa and D. P. Huston
178. Respiratory Infections in Allergy and Asthma, edited by
S. L. Johnston and N. G. Papadopoulos
179. Acute Respiratory Distress Syndrome, edited by
M. A. Matthay
180. Venous Thromboembolism, edited by J. E. Dalen
181. Upper and Lower Respiratory Disease, edited by J. Corren,
A. Togias, and J. Bousquet
182. Pharmacotherapy in Chronic Obstructive Pulmonary Disease,
edited by B. R. Celli
183. Acute Exacerbations of Chronic Obstructive Pulmonary
Disease, edited by N. M. Siafakas, N. R. Anthonisen,
and D. Georgopoulos
184. Lung Volume Reduction Surgery for Emphysema, edited by
H. E. Fessler, J. J. Reilly, Jr., and D. J. Sugarbaker
185. Idiopathic Pulmonary Fibrosis, edited by J. P. Lynch III
186. Pleural Disease, edited by D. Bouros
187. Oxygen/Nitrogen Radicals: Lung Injury and Disease,
edited by V. Vallyathan, V. Castranova, and X. Shi
188. Therapy for Mucus-Clearance Disorders, edited by
B. K. Rubin and C. P. van der Schans
189. Interventional Pulmonary Medicine, edited by J. F. Beamis, Jr.,
P. N. Mathur, and A. C. Mehta
190. Lung Development and Regeneration, edited by
D. J. Massaro, G. Massaro, and P. Chambon
191. Long-Term Intervention in Chronic Obstructive Pulmonary
Disease, edited by R. Pauwels, D. S. Postma, and S. T. Weiss
192. Sleep Deprivation: Basic Science, Physiology, and Behavior,
edited by Clete A. Kushida
193. Sleep Deprivation: Clinical Issues, Pharmacology, and Sleep
Loss Effects, edited by Clete A. Kushida
194. Pneumocystis Pneumonia: Third Edition, Revised
and Expanded, edited by P. D. Walzer and M. Cushion
195. Asthma Prevention, edited by William W. Busse
and Robert F. Lemanske, Jr.
196. Lung Injury: Mechanisms, Pathophysiology, and Therapy,
edited by Robert H. Notter, Jacob Finkelstein,
and Bruce Holm
197. Ion Channels in the Pulmonary Vasculature,
edited by Jason X.-J. Yuan
198. Chronic Obstuctive Pulmonary Disease: Cellular and
Molecular Mechanisms, edited by Peter J. Barnes
199. Pediatric Nasal and Sinus Disorders, edited by Tania Sih
and Peter A. R. Clement
200. Functional Lung Imaging, edited by David Lipson
and Edwin van Beek
201. Lung Surfactant Function and Disorder, edited by
Kaushik Nag
202. Pharmacology and Pathophysiology of the Control
of Breathing, edited by Denham S. Ward, Albert Dahan
and Luc J. Teppema
203. Molecular Imaging of the Lungs, edited by Daniel Schuster
and Timothy Blackwell
204. Air Pollutants and the Respiratory Tract: Second Edition,
edited by W. Michael Foster and Daniel L. Costa
205. Acute and Chronic Cough, edited by Anthony E. Redington
and Alyn H. Morice
206. Severe Pneumonia, edited by Michael S. Niederman
207. Monitoring Asthma, edited by Peter G. Gibson
208. Dyspnea: Mechanisms, Measurement, and Management,
Second Edition, edited by Donald A. Mahler
and Denis E. O'Donnell
209. Childhood Asthma, edited by Stanley J. Szefler
and Sren Pedersen
210. Sarcoidosis, edited by Robert Baughman
211. Tropical Lung Disease, Second Edition, edited by Om Sharma
212. Pharmacotherapy of Asthma, edited by James T. Li
213. Practical Pulmonary and Critical Care Medicine: Respiratory
Failure, edited by Zab Mosenifar and Guy W. Soo Hoo
214. Practical Pulmonary and Critical Care Medicine: Disease
Management, edited by Zab Mosenifar and Guy W. Soo Hoo
The opinions expressed in these volumes do not necessarily represent the

views of the National Institutes of Health.
Practical Pulmonary
and Critical Care
Medicine
Respiratory Failure
Edited by
Zab Mosenifar
CedarsSinai Medical Center
Los Angeles, California, U.S.A.
Guy W. Soo Hoo

VA Greater Los Angeles Healthcare System
Geffen School of Medicine at UCLA
New York London

Published in 2006 by
Taylor & Francis Group
270 Madison Avenue
New York, NY 10016
2006 by Taylor & Francis Group, LLC
No claim to original U.S. Government works
Printed in the United States of America on acid-free paper
10 9 8 7 6 5 4 3 2 1
International Standard Book Number-10: 0-8493-6663-1 (Hardcover)
International Standard Book Number-13: 978-0-8493-6663-5 (Hardcover)
Library of Congress Card Number 2005044634
This book contains information obtained from authentic and highly regarded sources. Reprinted material is
quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts
have been made to publish reliable data and information, but the author and the publisher cannot assume
responsibility for the validity of all materials or for the consequences of their use.
No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic,
mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and
recording, or in any information storage or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.com
(http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive,
Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration
for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate
system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only
for identification and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
Practical pulmonary and critical care medicine / edited by Zab Mosenifar, Guy W. Soo Hoo
p. ; cm. -- (Lung biology in health and disease ; v. 213-214)
Includes bibliographical references and index.
Contents: v. 1. Respiratory failure-- v. 2. Disease management.
ISBN-13: 978-0-8493-6663-5 (v. 1 : alk. paper)
ISBN-10: 0-8493-6663-1 (v. 1 : alk. paper)
ISBN-13: 978-0-8247-2597-6 (v. 2 : alk. paper)
ISBN-10: 0-8247-2597-2 (v. 2 : alk. paper)
1. Lungs--Diseases. 2. Respiratory intensive care. I. Mosenifar, Zab, 1951- II. Soo Hoo, Guy W.
III. Series.
[DNLM: 1. Lung Diseases--therapy. 2. Acute Disease--therapy. 3. Critical Care--methods. WF 600
P8947 2006]
RC941.P73 2006
616.2'4--dc22 2005044634
Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com
Taylor & Francis Group
is the Academic Division of Informa plc.
Introduction
Surely, all of us would certainly concur with the rst sentence of the Preface pre-
pared by Drs. Mosenifar and Soo Hoo: Over the past decade, the pace of pro-
gress in medicine has been astounding. New developments in diagnosis,
management, and therapeutics occur at a breathtaking rate. On the other hand,
one could argue with the time frame, that is, why only the past decade? This is
a signicant point, especially when it comes to pulmonary and critical care. Pul-
monary medicine started to blossom many, many decades ago but especially
during the last 30 years or so. Critical care started its course, to what it is
today, with the development of the oxygen electrode (and later oximetry) and
with the recognition of the value of blood pH measurements in the late 1950s.
At the same time, the use of ventilators and respiratory assist devices became
more common.
However, in reality it matters little when progress began. Today, what
counts is that new developments are being introduced at such a fast pace that pro-
gress may exceed the ability to fully transfer and utilize all this knowledge in the
practice of medicine. Sure enough, tertiary hospitals have the facilities and work
force to adjust to and quickly adapt the changes as they occur. On the other hand,
hospitals which are community based and distant from academic centers may nd
iii
iv Introduction
it more difcult to take full advantage of the newer approaches for the manage-
ment of pulmonary patients and/or critical care situations.
What could be considered a dual standard between tertiary and community
hospitals is well recognized today, and has raised concerns of policy makers,
politicians, and medical leaders. The question is frequently raised about how
fast, and how well, research advances are translated into the practice of medicine,
including pulmonary and critical care medicine.
This monograph, Practical Pulmonary and Critical Care Medicine:
Respiratory Failure and its companion volume titled Practical Pulmonary and
Critical Care Medicine: Disease Management represent important steps to mini-
mize translation difculties. Indeed, they both present a very practical approach
to pulmonary and critical care medicine. The rst volume addresses, and very
carefully describes, the tools to provide the most optimal care and management
of respiratory failure. The second volume examines a number of situationsboth
pulmonary and non-pulmonarythat require the use of these tools. Together,
these volumes will not only enrich the knowledge of the readers, but also will
provide a wealth of practical information that has the potential to positively
impact on the care of their patients.
The editors, Dr. Zab Mosenifar and Dr. Guy W. Soo Hoo, have assembled
contributors with outstanding expertise and a wealth of practical experience,
coming from institutions/environments with large patient populations, and a
wide variety of cases and medical situations. They share their very practical
and real experiences throughout the volumes.
As the Executive Editor of the series of monographs Lung Biology in
Health and Diseases, I am proud to present these two volumes and to express
my most sincere thanks to the editors and the authors.
Claude Lenfant, MD
Gaithersburg, Maryland, U.S.A.
Preface
Over the past decade, the pace of progress in medicine has been astounding. New
developments in diagnosis, management, and therapeutics occur at a breathtaking
rate. A new disease emerges and, aided by the ease of transcontinental travel,
threatens to become the next pandemic. And just as quickly, the causative
agent is identied with diagnostics and therapeutics soon to follow. Old diseases
have their mysteries unraveled, and targeted therapeutics provide hope where
there was once despair. Some conditions come under control in less than a
generations span, while others plod along inexorably to the end, with little
available to alter their course.
Just as the conditions of disease have changed, so have the conduits for
information. The speed and availability of electronic databases now allow access
to vast warehouses of information at the click of a mouse or ick of a stylus. Where
once the resident or houseofcer carried a worn copy of a spiral bound manual,
handheld computers are now the essential accessories. The grand old textbooks
have followed suit, available in versions abbreviated to t le-size limits or
available in their own electronic internet-based versions. This electronic world
not only allows but mandates frequent content changes. Information can be
updated daily and even more frequently if necessary. The online resource is
now predominant in an arena that was once the domain of the print journal.
v
vi Preface
Whereas attendance at national or international meetings once offered the latest

developments, this information can now be accessed from a remote site and
disseminated at near-instantaneous speed, certainly before one can return home.
Why then, in this information rich era, would there be a need for this book?
First of all, even though there is instant access and availability to volumes of
information, there remains a dearth of practical information. No one functions
in the vacuum of cyberspace or isolation of an information warehouse. Everyone
faces the limitations of available technology, restricted formularies, time press-
ures, and treatment preferences. In addition to the wealth of knowledge, one
requires the wisdom of experience and expediency of practical management.
The best technology or therapeutic is only as effective as the treatment that
can be instituted by the lone practitioner. Treatment available only to the few
or the very specialized usually has no role in general management.
Filling this void is the guiding premise of this book. Even if mutations
threaten to render current antimicrobials ineffective or new disease entities
emerge, there remains a need for comprehensive and effective supportive care.
This care is best provided not by the disease specialist, but by physicians who
carry a broader perspective while also maintaining focus on the most pressing
problems. In the new lexicon of medicine, this is the hospitalist or intensivist.
Because many of the most immediate life-threatening conditions involve the
respiratory tract, management often falls under the domain of pulmonary and
critical care, either as the pulmonologist and/or intensivist. The need for coordi-
nated comprehensive care is further highlighted by increasingly vulnerable
patients as a result of increased longevity, treatment modalities that strip a
patients immune system, and emerging disease entities. The modern pulmonary
and critical care physician not only has to deal with the many complexities of
illness, but also must choose the best available approach to facilitate recovery.
This has required the intensivist to assume many roles in patient manage-
ment, but none more important than as the functional equivalent of a chief
executive with oversight over total management. This has required familiarity
with areas where they are not generally considered expert, and reliance on a
multi-disciplinary approach to care. These sections in the text are authored by
experts in the area who provide a broad overview, but highlight those issues
most important in managing critically ill patients. This brings the intensivists
view to the specialists world.
As every bibliophile knows, the advantage of a textbook often lies in the
additional or complementary information that is often encountered through
perusal of its pages. Electronic sources are often unforgiving in their searches,
limiting access to pre-determined and pre-dened categories. With a book, one
can often identify the needed content within its covers even if one is not quite
certain of the initial focus of inquiry. The topic may not be in the rst section,
but is invariably covered in a subsequent area.
Practical Pulmonary and Critical Care is divided into two volumes based
partly on space limitations and partly by design. This rst volume focuses
Preface vii
on immediate management and diagnosis. For the practicing physician, this is

often referred to in billing as the rst hour or more of critical care management.
This often involves the patient with respiratory failure, one of the most common
entry conditions into a hospital or critical care unit. These patients often present
in extreme distress and the underlying diagnosis may not be immediately appar-
ent or may require further investigation. Other patients are intubated as part of
support measures while therapy is directed at other organs.
This volume includes chapters and strategies on immediate management of
these patients. This includes an extensive treatise on oxygen strategies, helium
oxygen, and non-invasive ventilation. If the patient fails these conservative
measures, the focus shifts to intubation and ventilator management. The alphabet
soup of ventilator modes and waveform interpretation is claried with a section
that allows comparisons between multiple vendors. These patients demand close
monitoring, and a chapter is dedicated to the nuances of monitoring their
response to therapy as well as mechanical ventilation. Other sections deal with
the delicate task of discontinuing mechanical ventilation and patients who
require long-term ventilatory support.
These patients often have multi-organ system dysfunction. In addition
to ventilatory support, they often require other invasive procedures, either
therapeutically or to provide or guide therapeutics. A comprehensive section is
dedicated to the most commonly performed procedures, with an abundance of
detailed gures and tables. This includes coverage of procedures once reserved
for surgical colleagues, such as tube thoracostomy or tracheostomy. Equally
important are imaging studies that help diagnose and monitor a patients response
to therapy. The section on radiology covers the gamut with respect to patient
management, including the use of the radiograph for conrmation of location
of invasive devices, diagnosis, and illustrative cases. The radiographs are all
derived from recent cases and reproduce well, providing important detail
useful in management. Each topic is covered with a focus on details that facilitate
its implementation and potential pitfalls, as well as a practical perspective on its
role in overall management.
It is this perspective that not only denes this book, but hopefully lends an
enduring quality to its content. There have been very few scientic developments
that have completely altered the management of patients. Changes are often
incremental and incorporated over several years of practical experience, although
medications may be the one exception. The experience of time allows techniques
to undergo further renement. Practical aspects of management are highlighted
with detailed descriptions of procedures, protocols, or guidelines. There is
often much to gain from historical perspective, and this information is judiciously
included as well as that which is evidence based. Therefore, even though technol-
ogy, medications, and formularies have changed over the past decade, patients
with acute respiratory failure still have the same basic requirements in their
management. Ventilators must provide the best support while inicting the
least amount of ventilator-associated injury. And once the road to recovery is
viii Preface
reached, assisted ventilation should be removed as soon as possible. These basic

tenets will not change in the foreseeable future, only the details in implemen-
tation. Although a paper-based textbook may not have the allure of electronic
media, it can provide a roadmap and framework for efcient and pragmatic
care. Once in place, additional information can only enhance the work and
nal product. We share a common goal to enhance the recovery of patients
from a critical illness. We hope this textbook can contribute to that end result.
Zab Mosenifar
Guy W. Soo Hoo
Contributors
Janet Au Division of Pulmonary and Critical Care, Olive View UCLA

Medical Center and Geffen School of Medicine at UCLA, Los Angeles,
California, U.S.A.
Bruce M. Barack Department of Imaging, VA Greater Los Angeles
Healthcare System and Department of Radiological Science, Geffen School
of Medicine at UCLA, Los Angeles, California, U.S.A.
Sharad Dass Division of Pulmonary and Critical Care Medicine,
Cedars Sinai Medical Center and Geffen School of Medicine at UCLA,
Scott K. Epstein Department of Medicine, Caritas St. Elizabeths Medical
Center and Tufts University School of Medicine, Boston, Massachusetts, U.S.A.
Dani Hackner Transitional Critical Care Service, Cedars Sinai Medical
Center and Geffen School of Medicine at UCLA, Los Angeles, California, U.S.A.
Dean R. Hess Department of Respiratory Care, Massachusetts General
Hospital and Department of Anesthesia, Harvard Medical School, Boston,
Massachusetts, U.S.A.
ix
x Contributors
C. Matilda Jude Department of Imaging, VA Greater Los Angeles Healthcare

System and Department of Radiological Science, Geffen School of Medicine at
UCLA, Los Angeles, California, U.S.A.
Nader Kamangar Division of Pulmonary and Critical Care, Olive

View UCLA Medical Center and Geffen School of Medicine at UCLA,
Hsin-Yi Lee Department of Imaging, VA Greater Los Angeles Healthcare

System and Department of Radiological Science, Geffen School of Medicine at
UCLA, Los Angeles, California, U.S.A.
Michael I. Lewis Division of Pulmonary and Critical Care Medicine,

Cedars Sinai Medical Center and Geffen School of Medicine at UCLA,
Zab Mosenifar Division of Pulmonary and Critical Care Medicine,

Cedars Sinai Medical Center, Los Angeles, California, U.S.A.
Michael L. Nevins Division of Pulmonary and Critical Care and Department

of Medicine, Group Health Permanente, University of Washington School of
Medicine, Seattle, Washington, U.S.A.
Brian Richards Division of Pulmonary and Critical Care Medicine,

Cedars Sinai Medical Center, Los Angeles, California, U.S.A.
Antoinette R. Roth Department of Radiological Science, Geffen School of

Medicine at UCLA, Los Angeles, California, U.S.A.
Ammar Sakkour UCLA Santa Monica Specialties, Geffen School of

Medicine at UCLA, Santa Monica, California, U.S.A.
Silverio Santiago Pulmonary and Critical Care Section, West Los Angeles
Healthcare Center, VA Greater Los Angeles Healthcare System and Geffen
School of Medicine at UCLA, Los Angeles, California, U.S.A.
Nikhil Shah Pulmonary and Critical Care Section, West Los Angeles
Guy W. Soo Hoo Pulmonary and Critical Care Section, West Los Angeles
Irawan Susanto UCLA Santa Monica Specialties, Geffen School of Medicine

at UCLA, Santa Monica, California, U.S.A.
Contributors xi
Sanjay Vadgama Pulmonary and Critical Care Section, West Los Angeles
Steve Y. Wong Mission Hospital, Mission Viejo, Long Beach VA, Long Beach,
California, U.S.A.
This page intentionally left blank
Contents
Introduction Claude Lenfant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
1. Oxygen Therapy and Airway Management . . . . . . . . . . . . . . . 1

Steve Y. Wong and Dean R. Hess
I. Introduction . . . . 1
II. Oxygen Therapy . . . . 1
III. Airway Management . . . . 13
IV. Conclusions . . . . 25
References . . . . 25
2. Non-Invasive Ventilation in Critical Care . . . . . . . . . . . . . . . 33

Guy W. Soo Hoo
II. Rationale for Use . . . . 34
III. Clinical Conditions . . . . 35
IV. Other Conditions . . . . 45
xiii
xiv Contents
V. Clinical Aspects of NIMV . . . . 50

VI. Monitoring the Response to
Treatment . . . . 61
VII. Conclusion . . . . 67
3. Modes of Mechanical Ventilation . . . . . . . . . . . . . . . . . . . . . . 77

Brian Richards and Zab Mosenifar
II. The Breath . . . . 78
III. Breath Phases . . . . 78
IV. Breath Types . . . . 87
V. Breath Patterns or Modes . . . . 94
VI. Fine Tuning the Ventilator Breath . . . . 99
4. Monitoring During Mechanical Ventilation . . . . . . . . . . . . . 105

Dean R. Hess
II. Arterial Blood Gases . . . . 105
III. Arterialized Capillary Blood Gases . . . . 108
IV. Point-of-Care Testing . . . . 108
V. Blood Gas Monitors . . . . 109
VI. Mixed Venous Blood Gases . . . . 110
VII. Pulse Oximetry . . . . 112
VIII. Gastric Tonometry and Sublingual PCO2 . . . . 116
IX. Capnography . . . . 117
X. Lung Mechanics and Graphics . . . . 127
XI. Monitoring in Perspective . . . . 135
5. Weaning from Mechanical Ventilation . . . . . . . . . . . . . . . . . 153

Dani Hackner, Sharad Dass, and Michael I. Lewis
II. Pathophysiologic Factors Determining Weaning
Success or Failure: An Introductory Overview . . . . 154
III. Predicting Weaning Success . . . . 160
IV. Protocols and Pathways . . . . 164
V. Weaning from Prolonged Ventilation . . . . 167
VI. Sedation . . . . 168
VII. Goals of Therapy: Conversations
About Ventilation . . . . 174
Contents xv
6. Prolonged Mechanical Ventilation . . . . . . . . . . . . . . . . . . . . 187

Scott K. Epstein and Michael L. Nevins
II. How Is Prolonged Mechanical Ventilation
Dened and How Often Does It Occur? . . . . 188
III. Who Requires Prolonged Mechanical
Ventilation? . . . . 189
IV. Site of Care for Patients with Prolonged
Mechanical Ventilation . . . . 190
V. What Are the Outcomes for Patients Requiring
Prolonged Mechanical Ventilation? . . . . 191
VI. Why Do Patients Become Ventilator-Dependent? . . . . 192
VII. What Assessment Tools Are Available to Predict
Weaning Outcome for Patients with Prolonged
Mechanical Ventilation? . . . . 202
VIII. What Is the Best Approach to Weaning Patients
with Prolonged Mechanical Ventilation? . . . . 204
IX. Conclusion . . . . 207
7. Procedures in the Intensive Care Unit . . . . . . . . . . . . . . . . . 219

Sanjay Vadgama, Janet Au, and Nader Kamangar
II. Arterial Catheters . . . . 219
III. Central Venous Catheters . . . . 225
IV. Pulmonary Artery Catheters . . . . 236
V. Chest Tube Thoracostomy . . . . 261
8. Bronchoscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Nikhil Shah, Irawan Susanto, and Silverio Santiago
II. Specications, Indications, and Contraindications . . . . 286
III. Patient Preparation, Sedation, and Anesthesia . . . . 288
IV. Insertion Techniques . . . . 289
V. Diagnostic Techniques . . . . 289
VI. Post-Procedure Care . . . . 291
VII. Complications . . . . 291
VIII. Specic Diagnostic Indications . . . . 292
IX. Specic Diagnostic and Therapeutic Indications . . . . 297
X. Therapeutic Bronchoscopy:
Indications and Options . . . . 298
XI. Potential Therapeutic Indications . . . . 305
xvi Contents
9. Percutaneous Tracheostomy . . . . . . . . . . . . . . . . . . . . . . . . . 313

Ammar Sakkour and Irawan Susanto
II. Indications, Benets, and Timing of
Tracheostomy . . . . 315
III. Contraindications . . . . 316
IV. The Technique of Bedside PDT:
How We Do It . . . . 317
V. Complications . . . . 321
VI. OST vs. PDT . . . . 324
VII. Care of Patients with Tracheostomy . . . . 324
VIII. Future . . . . 325
10. Radiology in the Intensive Care Unit . . . . . . . . . . . . . . . . . . 331

Bruce M. Barack, C. Matilda Jude, Hsin-Yi Lee,
and Antoinette R. Roth
II. The Portable Chest Radiograph . . . . 331
III. Devices Used in the ICU . . . . 338
IV. Common Thoracic Abnormalities
Encountered in the ICU . . . . 364
V. Computerized Tomography . . . . 385
VI. Image-Guided Interventional
Procedures in the ICU . . . . 387
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
1
Oxygen Therapy and Airway Management
STEVE Y. WONG DEAN R. HESS
Mission Hospital, Mission Viejo Department of Respiratory Care

Long Beach VA, Long Beach, Massachusetts General Hospital
California, U.S.A. and Department of Anesthesia
Harvard Medical School
Boston, Massachusetts, U.S.A.
I. Introduction
Oxygen administration and airway management are two of the fundamental

aspects of management in a patient with acute respiratory failure. Proper appli-
cation of technical aspects of oxygen therapy and airway management can be
life saving. Despite the importance of these therapies and their frequent use in
the acute care setting, their nuances are often under-appreciated.
II. Oxygen Therapy
Oxygen (O2) is an elemental gas that is necessary for life in aerobic organisms.
In the absence of O2 (hypoxia), cellular respiration ceases and irreversible
cellular injury and death occur within minutes. At normal atmospheric pressure
and temperature, O2 exists as an odorless, tasteless, and colorless gas. It rep-
resents one-fth of the earths atmosphere by volume (20.96%).
A. Medical Oxygen
Medical grade O2 is manufactured by fractional distillation of liqueed air (1,2).
It is stored as a liquid to reduce the size of the storage container (1 L of liquid O2
produces 860 L of gaseous O2). The liquid O2, stored outside the hospital in
a cryogenic container, is converted to gaseous O2 and delivered to the patient
1
2 Wong and Hess
care areas via a bulk gas delivery system. By convention, O2 is supplied at a

pressure of 50 lbs/in2. Oxygen can also be delivered from a medical gas cylinder
at a pressure of 2000 lbs/in2 when full. Cylinders are frequently used to supply O2
during transport of patients to remote locations of the hospital, such as diagnostic
areas. Cylinders are identied by letter codes to indicate size and gas capacity.
The cylinders most frequently used in the hospital are small E-cylinders. They
are typically green in color, and a label afxed to the side of the cylinder indicates
the cylinder contents. When cylinders are used, it is important to calculate the
duration of ow to avoid inadvertent loss of O2 supply (Table 1).
B. Indications for Oxygen Therapy

The most important indication for O2 therapy is to treat hypoxemia. The alveolar
gas equation illustrates how increasing the inspired O2 fraction (FIO2) increases
the alveolar PO2 (PAO2) and subsequently the arterial PO2 (PaO2):
PA O2 FI O2 EBP 1:25 Pa CO2
where EBP is the barometric pressure corrected for water vapor pressure. The
effect of increasing FIO2 on the PaO2 is a function of the physiologic cause of
hypoxemia. In cases of shunt (V/Q 0), supplemental O2 therapy has little
effect on PaO2. If the cause of hypoxemia is low V/Q or diffusion defect,
supplemental O2 therapy will effectively increase the PaO2.
Table 1 Duration of Flow (in Hours) from an E-Cylinder as a Function of O2 Flow and
the Pressure of Gas Remaining in the Cylinder
Cylinder Flow (L/min)

pressure
(lbs/in2) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
500 2.3 1.2 0.8 0.6 0.5 0.4 0.3 0.3 0.3
600 2.8 1.4 0.9 0.7 0.6 0.5 0.4 0.4 0.3 0.3 0.3
700 3.3 1.6 1.1 0.8 0.7 0.5 0.5 0.4 0.4 0.3 0.3 0.3 0.3
800 3.7 1.9 1.2 0.9 0.7 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.3 0.3
900 4.2 2.1 1.4 1.1 0.8 0.7 0.6 0.5 0.5 0.4 0.4 0.4 0.3 0.3 0.3
1000 4.7 2.3 1.6 1.2 0.9 0.8 0.7 0.6 0.5 0.5 0.4 0.4 0.4 0.3 0.3
1100 5.1 2.6 1.7 1.3 1.0 0.9 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.4 0.3
1200 5.6 2.8 1.9 1.4 1.1 0.9 0.8 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.4
1300 6.1 3.0 2.0 1.5 1.2 1.0 0.9 0.8 0.7 0.6 0.6 0.5 0.5 0.4 0.4
1400 6.5 3.3 2.2 1.6 1.3 1.1 0.9 0.8 0.7 0.7 0.6 0.5 0.5 0.5 0.4
1500 7.0 3.5 2.3 1.8 1.4 1.2 1.0 0.9 0.8 0.7 0.6 0.6 0.5 0.5 0.5
1600 7.5 3.7 2.5 1.9 1.5 1.2 1.1 0.9 0.8 0.7 0.7 0.6 0.6 0.5 0.5
1700 7.9 4.0 2.6 2.0 1.6 1.3 1.1 1.0 0.9 0.8 0.7 0.7 0.6 0.6 0.5
1800 8.4 4.2 2.8 2.1 1.7 1.4 1.2 1.1 0.9 0.8 0.8 0.7 0.6 0.6 0.6
1900 8.9 4.4 3.0 2.2 1.8 1.5 1.3 1.1 1.0 0.9 0.8 0.7 0.7 0.6 0.6
2000 9.3 4.7 3.1 2.3 1.9 1.6 1.3 1.2 1.0 0.9 0.8 0.8 0.7 0.7 0.6
Oxygen Therapy and Airway Management 3
An unusual, but occasionally effective, use of O2 therapy is the

re-absorption of air leaks, such as a small pneumothorax, subcutaneous emphy-
sema, or pneumocephalous (3,4). If the patient is breathing room air, the compo-
sition of the leaked gas is presumably 21% O2 and 78% nitrogen. Breathing 100%
O2 clears nitrogen from the blood, creates a diffusion gradient for nitrogen to
diffuse from the extra-vascular space, and thus effectively reduces the volume
of the leaked gas. In a theoretical assessment of normobaric O2 therapy to treat
pneumocephalus, Dexter and Reasoner (3) reported that modestly increasing
the FIO2 from 0.21 to 0.4 decreased the total air absorption time by 67%,
whereas increasing the FIO2 from 0.8 to 1.0 decreased the total air absorption
time by only an additional 3%.
Another important use of O2 is treatment of carbon monoxide poisoning.
The half-life of carboxyhemoglobin is about ve hours breathing 21% O2 at
ambient pressure, a little more than one hour breathing 100% O2 at ambient
pressure, and ,30 min breathing 100% O2 at 3 atm of pressure. Weaver et al.
(5) conducted a double-blind, randomized trial to evaluate the effect of hyper-
baric-oxygen treatment on cognitive sequelae that commonly occur following
carbon monoxide poisoning. Patients with symptomatic acute carbon monoxide
poisoning were assigned to three hyperbaric oxygen treatments or one normo-
baric-oxygen treatment plus two sessions of exposure to normobaric room air.
Cognitive sequelae at six weeks were less frequent in the hyperbaric-oxygen
group (25%) than in the normobaric-oxygen group (46%).
Anaerobic infections and necrotizing fasciitis (6) may also be responsive to
O2 therapy. However, the treatment is usually with hyperbaric oxygen, rather
than normobaric oxygen, and this application of hyperbaric O2 therapy is
controversial.
C. Monitoring Oxygen Therapy

Oxygen should not be administered without an objective assessment of its effect.
Because the most frequent indication for O2 therapy is to treat hypoxemia,
the clinical effect of O2 administration is usually monitored with pulse oximetry
or arterial blood gas analysis. Most commonly, pulse oximetry is used and a
target O2 saturation (SpO2) of .92% is acceptable. If the O2 therapy is assessed
by arterial blood gas analysis, a PaO2 . 60 mm Hg is usually acceptable.
SpO2 or PaO2 should always be assessed relative to the amount of inspired
O2. In the case of low-ow O2 delivery systems (e.g., nasal cannula), the arterial
O2 level is assessed relative to the administered O2 ow. In the case of high-
ow O2 delivery systems and during mechanical ventilation, the arterial O2
level is assessed relative to the FIO2.
Oxygen analyzers use one of several methods to measure O2 concen-
tration(2). Polarographic and galvanic cell analyzers use an electrochemical prin-
ciple to measure changes in PO2, which is then converted to a display of %O2.
Paramagnetic analyzers use the Pauling principle, which is based on the fact that
4 Wong and Hess
O2 is a paramagnetic gas. A wheatstone bridge O2 analyzer uses the principle of

thermoconductivity. In the zirconium analyzer, an electric potential is developed
across heated zirconium oxide that is proportional to the PO2. Zirconium analyzers
are very precise and used in applications such as indirect calorimetry.
D. Physiologic Complications of Oxygen Therapy

Worsening hypercapnia with O2 administration in patients with chronic obstruc-
tive pulmonary disease (COPD) has been of concern for many years. This has
resulted in conservative use of O2 in these patients and has likely produced
unnecessary hypoxemia in some patients. In patients with stable COPD, continu-
ous O2 administration has been shown to be life-prolonging (7). Oxygen should
never be withheld from a hypoxemic patient in acute respiratory failure. More-
over, the risk may be relatively low (8), and the mechanism for hypercarbia
when O2 is administered to the patient with COPD is controversial. When hyper-
carbia occurs with O2 administration, this may be due to the Haldane effect (9),
an increase in alveolar dead space due to the effects of O2 on pulmonary blood
ow and V= _ Q _ mismatch, or suppression of the respiratory drive. Aubier et al.
(10) reported that an increased PaCO2 following O2 administration in patients
with COPD occurred with no change in minute ventilation. Crossley et al. (11)
reported that an increased FIO2 in mechanically ventilated patients with COPD
(PaO2 increase to about 200 mm Hg) did not change PaCO2, respiratory drive,
or dead space. Gomersall et al. (12) reported no signicant difference in
requirement for mechanical ventilation or mortality for patients with COPD in
acute respiratory failure in whom O2 was administered with a goal PaO2 . 50
or .70 mm Hg. However, Robinson et al. (13) reported a reduction in minute
ventilation and an increase in alveolar dead space in patients with COPD who
had worsening hypercapnia with O2 administration. These results, as well as
those by Dick et al. (14) and Dunn et al. (15), suggest that a decrease in venti-
latory drive may contribute to the hypercapnia that occurs with O2 administration
in some patients with COPD.
Concern for hypercapnia is also reported in patients with asthma exacer-
bation. Chien et al. (16) reported an increase in PaCO2 in patients with acute
asthma who received an FIO2 of 1.0, but the increase in PaCO2 was modest
(maximum of 10 mm Hg) and the clinical signicance of this is unclear.
Absorption atelectasis has been demonstrated in patients breathing 100%
O2 (17). Absorption of gas behind occluded airways is more likely to produce
atelectasis if the alveolar gas composition is 100% O2 rather than room air. If
the alveolar gas is room air, uptake of O2 results in an alveolar volume change
of 20%, delaying the onset of atelectasis. Although this effect has been demon-
strated primarily during anesthesia, it is likely that it might also occur during
acute respiratory failure. In mechanically ventilated patients, this complication
may be avoided by the use of positive end-expiratory pressure (PEEP).
Experimental evidence implicates the formation of reactive O2 species

(superoxide anion, hydroxyl radical, and hydrogen peroxide) in the pathogenesis
of pulmonary O2. The clinical importance of this, however, has been debated for
many years (18,19). It is common teaching that maintaining the FIO2 , 0.6
avoids O2 toxicity, although there are few data to support this. Early human
reports of respirator lung showed septal edema, endothelial cell damage,
hyaline membrane deposition, and interstitial brosis. However, these ndings
may have been the result of high ventilating pressure rather than hyperoxia
per se. Normal human volunteers exposed to hyperoxia develop chest discomfort,
dyspnea, cough, headache, and paresthesia within 12 hours of exposure and these
effects persist for several days. However, similar effects have not been shown to
occur in patients with acute respiratory failure.
Patients treated with bleomycin have an increased risk for oxygen-induced
lung toxicity (20). An FIO2 of 0.35 to 0.4 is sufcient to cause toxicity in this
setting. Bleomycin may impair the lungs antioxidant defenses, increasing the
burden of reactive O2 species produced in the lung by a high FIO2. Bleomycin
O2 toxicity may be steroid responsive.
Retinopathy of prematurity (21) and bronchopulmonary dysplasia (22) are
complications that are associated with O2 therapy in infants. Similar disorders
are not known to occur in patients older than neonates.
E. Technical Hazards of Oxygen Therapy

An important physical characteristic of O2 is its ability to support combustion.
Accordingly, the potential re hazard of O2 should be appreciated when sup-
plemental O2 is administered. Improper handling and storage of high-pressure
gas cylinders can result in rapid escape of gas, turning the cylinder into a danger-
ous projectile. Because O2 is stored dry, it must be humidied when delivered
at high concentrations. This is of particular concern when the upper airway
is bypassed with an endotracheal tube (ETT) or tracheostomy tube. When low-
ow O2 is administered, the need for humidication is less, and humidiers for
this application have not been shown to be useful (23 25).
F. Oxygen Administration Devices

Oxygen therapy systems are generally categorized as either low- or high-ow
devices (1,2). Low-ow devices deliver O2 at ow rates insufcient to meet
the inspiratory ow demand of the patient. The additional required ow is
inhaled from the room air. The FIO2 from low-ow devices may vary from
breath-to-breath, depending on the breathing pattern. High-ow devices
provide ow sufcient to meet the patients inspiratory ow demand and thus
maintain a precise FIO2 that is unaffected by changes in breathing pattern.
Relevant characteristics of O2 delivery devices are shown in Table 2.
The nasal O2 cannula (Fig. 1) is the most widely used O2 delivery device.
It consists of two short, soft, pliable plastic prongs, each about one-half inch in
6 Wong and Hess
Table 2 Oxygen Delivery Devices for Adult Applications
Approximate
Usual ow oxygen
Device range concentration Comments
Nasal 1 6 L/min 24 40% FIO2 is reduced with

cannula nasal obstruction;
can be less effective
with mouth breathing;
FIO2 varies with
breathing pattern
Simple 5 10 L/min 30 60% Flows ,5 L/min
mask result in rebreathing;
FIO2 varies with
breathing pattern
Non- Flow must Theoretically, a If SpO2 remains low
rebreathing be high enough non-rebreathing despite use of a
mask to prevent mask will deliver non-rebreathing
full collapse of close to 100% O2; mask, consider using
reservoir bag in reality, however, a high-ow O2
during inhalation; it delivers delivery device
ows 12 L/min concentrations of
are often required 60 80% because the
mask does not t
tightly over the face
Air- Use at least the O2 concentration is When mask is
entrainment ow stamped stamped on the removed, administer
mask on colored colored adapter O2 by nasal cannula
adapter to provide target SpO2
High-ow .30 L/min 24 100%, set by Gas should be
oxygen air and O2 ow humidied with
system meters on high-ow system
blender
length. The cannula is easily applied and well tolerated by most patients when used
with ows of 6 L/min. It is held in place by an elastic band around the head or,
more commonly, by looping the delivery tubing over the ears and holding it in
place with an adjustable slide under the chin. Oxygen delivery by nasal cannula
produces an FIO2 of approximately 0.24 at 1 L/min to about 0.40 at 6 L/min.
The FIO2 from a nasal cannula is highly variable even if the O2 ow remains con-
stant (Table 3). A number of studies using various methodologies conrm that the
FIO2 resulting from use of a nasal cannula is highly variable and dependent upon
the O2 ow, inspiratory ow, and minute ventilation (2636). The amount of deliv-
ered O2 is reduced if O2 is breathed from only one of the prongs, as may occur if the
prongs are displaced or there is unilateral nasal obstruction (37). There are
nasal
cannula
simple
mask
Figure 1 Nasal cannula and simple mask.
conicting data regarding the effect of mouth breathing when a nasal cannula is
used (38,39). Although mouth breathing may decrease the FIO2, O2 therapy by
nasal cannula can be effective even with mouth breathing.
The majority of O2 is wasted when it is delivered by nasal cannula. Most of
the O2 ow during the expiratory phase is wasted. It is only during the initial
period of inspiration that the supplemental ow of O2 is needed. It is this rst
part of inspiration that is delivered to the alveoli and participates in gas exchange.
The remainder of the inspired gas lls the anatomic dead space and is exhaled
without contributing in gas exchange. Oxygen-conserving devices conserve O2
by using a reservoir, which provides O2 ow only during the inspiratory phase,
Table 3 Estimation of FIO2 from a Nasal Cannula
Cannula ow 2 L/min (33 mL/sec)

Tidal volume 500 mL
Anatomic reservoir 50 mL (nasal passages
and nasopharynx)
Inspiratory time 1 sec
Volume of O2 inspired
33 mL O2 ow
50 mL Anatomic reservoir
88 mL Amount of O2 in the
inspired air; 417 mL 0.21
Volume O2 inspired 171 mL
FIO2 171 mL (O2)/500 mL (tidal volume) 0.34. Similar calculations can

be used to calculate the effects of changes in O2 ow or inspiratory time.
8 Wong and Hess
Table 4 Oxygen Delivery Systems Designed to Conserve O2 by Eliminating Waste

During the Expiratory Phase
Reservoir O2 cannula (moustache and pendant types): A small reservoir lls with O2
(20 mL for the moustache style and 40 mL for the pendant) during exhalation. At the
beginning of inhalation, a bolus of O2 is drawn from the reservoir. This may allow a
reduction in O2 ow by 50 75%. Reservoir cannulae are not well accepted by patients
due to their appearance.
Transtracheal O2 catheter: This is a small-diameter catheter surgically inserted into the
trachea. It is connected to a small ange and held in place by an adjustable chain. As O2
is continuously delivered directly into the trachea, a reduction in the O2 ow by about
50% is possible. Complications of the catheter placement include infection, bleeding,
and subcutaneous emphysema. The catheter must be cleaned regularly to prevent mucus
accumulation. Catheter obstruction is prevented by instilling saline and inserting a
cleaning wire into the lumen of the catheter.
Demand oxygen conservers: These devices only deliver O2 during the inspiratory phase.
When the patient begins an inspiration, this creates a negative pressure in the supply
tubing and causes a demand valve to open and supply a dose of O2.
either by providing an O2 bolus at the beginning of inspiration or by admini-

stering O2 directly into the trachea (Table 4) (40,41). These devices are
not commonly used in the acute care settings, but patients receiving chronic
O2 therapy may use one of these systems. Considerable variability in the perform-
ance of these devices has been reported (42).
The simple O2 mask (Fig. 1) is used when a higher FIO2 is needed than can
be attained with a nasal cannula or when a cannula is not appropriate (e.g., with
nasal obstruction). The simple mask increases the FIO2 over that achieved
by a nasal cannula because it adds a volume of 100 to 200 mL over the
face, which serves as an O2 reservoir. Additional air is inhaled through small
holes in the mask. To avoid rebreathing, a minimum ow of 5 L/min must be
used with O2 delivery by face mask (43). The simple mask is a low-ow O2
delivery device capable of providing an FIO2 of 0.3 to 0.6 at ows of 5 to
10 L/min (44). The FIO2 is dependent on the size of the mask and the patients
breathing pattern. Simple masks are subjectively less appealing than nasal
cannulae and may cause claustrophobia, mufing of speech, and difculty
eating and drinking.
The non-rebreathing mask (Fig. 2) increases the FIO2 by adding a reservoir
bag. It has a one-way valve between the bag and the mask and another one-
way valve over one or both mask ports. All of the exhaled tidal volume is
directed through the mask ports because of the valve between the mask and the
bag. The bag lls with 100% O2 during exhalation. During inhalation, the mask
valves close and the bag valve on the reservoir opens. The O2 ow must be high
enough to prevent the bag from emptying during inhalation. Theoretically, the
valves
reservoir
reservoir
bag
bag
Figure 2 Non-rebreathing mask.
non-rebreathing mask will deliver 100% O2. However, these masks do not provide
an airtight t on the face, and the valves do not provide a perfect seal (45). At ows of
10 to 15 L/min, an FIO2 of 0.6 to 0.8 may be achieved. Moreover, the non-rebreath-
ing masks have a valve over only one of the exhalation ports. This allows inhalation
of room air if the O2 supply ow becomes inadequate.
lower
O2 O2
FIO2
O2 higher O2
FIO2
Figure 3 The air-entrainment mask and its principle of operation.

10 Wong and Hess
G. High-Flow Devices
An air-entrainment mask (Fig. 3) consists of a mask, a jet nozzle, and air-entrain-
ment ports (46 49). Oxygen is delivered through the jet nozzle, which increases
its velocity. The high-velocity O2 entrains ambient air into the mask due to the
viscous shearing forces between the gas traveling through the nozzle and
the stagnant ambient air. The FIO2 depends on the nozzle size and the size of
the entrainment ports. Commercially available systems use interchangeable
jets or adjustable entrainment ports. Obstruction of the entrainment port or down-
stream obstruction decreases entrainment and increases FIO2. To deliver a xed
FIO2, the ow to the mask must exceed the peak inspiratory ow of the patient.
This may be difcult to achieve with a higher FIO2 settings (Table 5).
A high-ow O2 delivery system can be used to deliver a precise FIO2 and
any concentration between 0.21 and 1.0. Such a system provides sufcient ow to
meet the inspiratory demands of the patient (50). Because the total ow is set
higher than the inspiratory ow of the patient (typically 30 60 L/min), the
gas should be humidied. An air and an O2 ow meter can be used to deliver
a precise FIO2 (Fig. 4) and a variety of patient interfaces can be used (Fig. 5).
The FIO2 can be calculated from the ow rates of air and O2 (Table 6). A
blender uses pressurized sources of air and O2 to deliver a precise FIO2. Blenders
are compact and convenient but more expensive than using two ow meters.
H. Heliox
Heliox is a gas mixture of helium and oxygen that is clinically useful in some
circumstances due to its low density. Because helium does not support life, it must
always be delivered in a gas mixture containing at least 20% oxygen. There is
an increasing interest in its therapeutic use in patients with obstructive lung
Table 5 FIO2, Minimum Flow Requirements, Outputs, and Entrainment Ratios for an
Air-Entrainment Mask
FIO2 Minimum O2 Entrainment Total ow

setting ow (L/min) ratio (O2:Air) (L/min)
0.24 4 1:25 104

0.28 4 1:10 44
0.31 6 1:7 48
0.35 8 1:5 48
0.40 8 1:3 32
0.50 12 1:1.7 32
0.60 12 1:1 24
0.70 12 1:0.6 19
Source: From Branson RD. The nuts and bolts of increasing arterial oxygenation: devices and
techniques. Respir Care 1993; 38:672 686.
Figure 4 High-ow O2 delivery system.
aerosol
mask
tracheostomy
mask
T-piece
Figure 5 Patient interfaces used with high-ow O2 delivery systems.

12 Wong and Hess
Table 6 Air:O2 Ratios and Determination of Air and O2 Flows
Air:O2 (1.0 2 FIO2)/(FIO2 2 0.21)

If an FIO2 of 0.4 is desired:
Air:O2 (1.0 2 0.4):(0.4 2 0.21) 0.6:0.2 3:1
If a total ow of 60 L/min is required, an air ow of 45 L/min and an O2 ow of 15 L/min
will produce a FIO2 of 0.4.
If an FIO2 of 0.6 is desired:
Air:O2 (1.0 2 0.6):(0.6 2 0.21) 0.4:0.4 1:1
If a total ow of 60 L/min is required, an air ow of 30 L/min and an O2 ow of 30 L/min
will produce a FIO2 of 0.6.
diseases, and this has been the source of several reviews (51,52) and meta-ana-
lyses (53,54). Although the role of heliox has been reported benecial in case
series and anecdotal reports, current evidence is insufcient to allow a
recommendation for the use of heliox as a standard therapy for any specic
patient population.
One use of heliox is to reduce resistance with upper airway obstruction
(55,56). An example is post-extubation stridor, where use of heliox has been
reported to be benecial (57). There is also interest in the use of heliox
for acute asthma (58,59). In spontaneously breathing patients with asthma,
heliox decreases PaCO2, increases peak ow, and decreases pulsus paradoxus.
There may be benet related to the combination of heliox with aerosol broncho-
dilator delivery in patients with acute asthma or COPD (60 63). When heliox
(rather than air or oxygen) is used to power the nebulizer, the ow should be
increased by about 50% to assure adequate output from the nebulizer (64). As
demonstrated in several meta-analyses, however, the benet of heliox in the man-
agement of patients with acute asthma has yet to be conclusively demonstrated
(53,54). The role of heliox in the treatment of COPD is unclear (65 67).
COPD is a disease of small airwaysa region of the lungs in which ow is
density independent. Heliox has been reported to decrease work of breathing
in some, but not all, patients with COPD when evaluated just prior to extubation
(68). Benet has been reported for the use of heliox with non-invasive ventilation
in patients with COPD (69 72), and methods to administer heliox with a BiPAP
ventilator have been described (73).
Care must be taken to administer heliox in a safe and effective manner.
To avoid administration of a hypoxic gas mixture, it is recommended that 20%
oxygen/80% helium is mixed with oxygen to provide the desired helium concen-
tration and FIO2. If an FIO2 greater than 0.40 is required, the limited concen-
tration of helium is unlikely to produce clinical benet. For spontaneously
breathing patients, heliox is administered by face mask with a reservoir bag
(Fig. 6). A Y-piece attached to the mask allows concurrent delivery of aero-
solized medications. Sufcient ow is required to minimize contamination of
Figure 6 Delivery system for heliox administration for spontaneously breathing

patients.
the heliox with ambient air. This is often at least 12 to 15 L/min and requires 3 to
6 H-size cylinders per day. When using an oxygen-calibrated ow meter for
heliox therapy, it must be remembered that the ow of heliox (80% helium
and 20% oxygen) will be 1.8 times greater than the indicated ow. Heliox admin-
istration during mechanical ventilation can be problematic (74 78). Ventilators
are designed to deliver a mixture of air and oxygen. The density, viscosity, and
thermal conductivity of helium affect the delivered tidal volume and the measure-
ment of exhaled tidal volume. With some ventilators (e.g., Puritan Bennett), no
reliable tidal volume is delivered with heliox, whereas there may be a much
higher delivered tidal volume than desired for other ventilators. One commer-
cially available ventilator has been approved for use with heliox (VIASYS
AVEA, Palm Springs, California, U.S.A.).
III. Airway Management
When oxygen delivery remains compromised without mechanical assistance,

endotracheal intubation should be considered. The goal of this section is to
describe the practical aspects of endotracheal intubation. Fundamental key
points useful in enhancing a successful endotracheal intubation will be discussed.
14 Wong and Hess
A. Indications
Endotracheal intubation is indicated to protect the airway and reduce the risks
of pulmonary aspiration. It also offers a pathway for patients who require prolon-
ged positive pressure ventilation and frequent suctioning. In addition, the ETT
can be used to administer emergency medications when intravenous access is
not available. Most frequently, endotracheal intubation enables the patient to
undergo surgical procedures and allow delivery of inhalational anesthetics.
Outside the operation room, endotracheal intubation typically involves patients
who are in respiratory failure, shock, or cardiopulmonary arrest (79). Despite
the increasing use of non-invasive ventilation, most patients receive mechanical
ventilation invasively via an ETT.
B. Nasotracheal vs. Orotracheal Intubation
The ETT can be placed either nasally or orally. Nasotracheal intubation is typi-
cally indicated for patients undergoing oral surgeries or when intubation through
the mouth is unsuccessful. It is contraindicated in patients with basilar skull frac-
ture, mid-facial trauma, coagulopathy, nasal polyps, and epistaxis. For long-term
mechanical ventilation, orotracheal intubation is preferred because a nasotracheal
tube may increase the work of breathing as well as the risk of sinusitis. Because
the orotracheal route usually permits placement of a larger ETT, clearance of
secretions may also be easier.
C. Airway Assessment
A thorough airway evaluation should be performed prior to intubation to assess
the degree of difculty in airway management. Careful review of the medical
history and a detailed examination of the anatomic characteristics allow
identication of potential difcult mask ventilation or tracheal intubation.
Anatomic features such as a short muscular neck, receding mandible,
prominent upper incisors, small mouth with a high arched palate, and limited move-
ment of the mandible suggest increased likelihood of potential airway problems (80).
D. Predicting Difcult Mask Ventilation
Langeron and colleagues (81) identied ve independent predictors for difcult

mask ventilation: age .55 years, body mass index .26 kg/m2, lack of teeth,
presence of a beard, and history of snoring. When two or more of the risk
factors are present, the likelihood of difculty with mask ventilation is high.
Furthermore, difcult intubation is encountered more frequently when difcult
ventilation is present.
Figure 7 An example of the Mallampati class I airway.
E. Predicting Difcult Tracheal Intubation

The Mallampati classication (82) is widely used in predicting the ease of laryn-
goscopy. It compares tongue versus pharyngeal size. The observation is divided
into four categories. The evaluation is performed while the patient is sitting
with mouth open and tongue protruding. Visualization of the oropharyngeal
structures is then noted (Fig. 7).
. Class I: the soft palate, uvula, and faucial pillars are all visible.
. Class II: the tonsillar pillars and the base of uvula are obscured by the
base of the tongue.
. Class III: only the soft palate is visible.
. Class IV: soft palate is not visible.
When the pharyngeal class is Mallampati III or IV, a more difcult intubation is
expected. Unfortunately, it is not easy to use the Mallampati classication for
urgent or emergent intubation.
Simple measurement of the thyromental distance, which is the length
between the prominence of the thyroid cartilage and the bony point of the
chin, can be used to aid in predicting difcult intubation. When the thyromental
distance is ,7 cm and the Mallampati class is III or IV, difcult intubation can be
anticipated (83). The performance of the Mallampati test and the measurement of
the thyromental distance should identify most cases of difcult intubation
and allow appropriate preparations.
16 Wong and Hess
F. Laryngeal Visualization by Direct Laryngoscopy

Laryngoscopic view of the glottic opening is also used to predict the ease of tracheal
intubation (84). The extent of laryngeal exposure is categorized into grades I to IV.
When the glottic opening is fully visualized (grade I view), endotracheal intubation
is relatively easy. If only the posterior portion of the glottic opening is seen (grade II
view), intubation is technically more difcult. When the laryngoscopic view is
grade III (only the tip of epiglottis is visible) or grade IV (only the soft palate is
seen), success of tracheal intubation by direct laryngoscopy is expected to be low.
G. Maneuvers to Improve Laryngoscopic Visualization

Better laryngeal visualization of the glottic opening facilitates endotracheal intu-
bation. Several simple techniques have been advocated to improve laryngoscopic
visualization of the glottis. The BACK maneuver (simple back-pressure on the
thyroid or cricoid cartilage) displaces the larynx posteriorly and reduces the
failure rate to visualize any part of the glottis from about 9.2% to 1.6% (85).
The BURP maneuver (backward, upward, and rightward pressure of the
larynx) has also been reported to improve laryngeal visualization (86). In a com-
parative study (87), both the BURP maneuver and the BACK maneuver were
found to be effective. However, the BURP maneuver was shown to be superior.
Neither maneuver was associated with signicant complications.
The jaw thrust maneuver is a technique commonly used to relieve laryngeal
obstruction caused by the base of the tongue. By grasping the angles of the jaw
with one hand on each side, the mandible can be displaced forward to keep the
pharyngeal airway patent (88). The mandibular advancement can also improve
Figure 8 Equipment for endotracheal intubation.

Oxygen Therapy and Airway Management
Table 7 Medications Commonly Used to Facilitate Endotracheal Intubation
Depolarizing Non-depolarizing
Induction agent muscle relaxant muscle relaxant
Medication Propofol Etomidate Succinylcholine Rocuronium Vecuronium

Concentration 10 mg/mL 2 mg/mL 20 mg/mL 10 mg/mL 1 mg/mL
Dosage 1 2 mg/kg 0.2 0.6 mg/kg 1 2 mg/kg 0.6 1.2 mg/kg 0.08 0.1 mg/kg
Absolute Hypersensitivity Known sensitivity Malignant Known hyper- Known
contraindications to propofol, to etomidate hyperthermia sensitivity hypersensitivity to
soybean oil, egg, susceptibility, to rocuronium vecuronium
and glycerol myopathies,
burns, increased
intracranial pressure
Side effects/ Hypotension, apnea, Myoclonus, Bradycardia, Increased sensitivity Increased sensitivity
clinical pain at injection site adrenal hyperkalemia, in patients with in patients with
considerations suppression increased myasthenia gravis, myasthenia gravis,
intracranial Eaton Lambert Eaton Lambert
pressure syndrome syndrome
Source: From Donnelly AJ, Cunningham FE, Baughman VL. Anesthesiology and Critical Care Drug Handbook. Ohio: Lexi-Comp Inc., 2000.
17
18 Wong and Hess
laryngeal inlet view during nasal beroptic laryngoscopy (89). Tamura and
colleagues (90) have also demonstrated that such a maneuver can improve
laryngeal view during laryngoscopy performed by inexperienced physicians.
When the mandibular advancement is combined with the BURP maneuver,
further improvement of laryngeal visualization can be obtained.
H. Preparation for Orotracheal Intubation
Before attempting endotracheal intubation, preparation is needed to ensure

that the equipment (Fig. 8) is working and medications (Table 7) are available.
1. Laryngoscope: a laryngoscope consists of a handle and a blade. The
number on the blade reects its length; therefore, an adult with a
long neck may require a larger blade. The blade can be curved
(MacIntosh), straight (Jackson Wisconsin), or straight with a curved
tip (Miller). A curved (MacIntosh) #3 blade is most commonly used
for intubation in adults; however, the choice of the blade is entirely
dependent on individual preference. A MacIntosh blade may provide
more room to pass the ETT, while the straight blade gives better
exposure of the glottic opening. Regardless, blades of different size
and shape should be readily available. When checking the laryngo-
scope, it is important to ensure that there are batteries in the handle
and the light in the blade illuminates well.
2. ETT: a proper-sized ETT should be determined prior to intubation,
usually an internal diameter of 7.5 mm for adult males and 7.0 mm for
adult females. If bronchoscopy is anticipated, endotracheal tubes with
8.0 mm or larger internal diameters are preferred, as the smaller ETTs
may hinder or prevent the passage of the adult bronchoscope. When
examining the ETT, the cuff should be gently inated with a syringe
and checked for leaks. A stylet can be inserted in the ETT to provide rigid-
ity, which can allow the practitioner to direct the ETT with more control.
3. Oral airway: 8.0 and 9.0 mm oral airways are commonly used in adults
to maintain a patent conduit above the laryngeal inlet, just cephalad
to the epiglottis. Choosing the proper size oral airway is important as
airway obstruction can occur if the inserted oral airway is too big or
too small. When an ideal size oral airway is placed on the side of the
patients face, the proximal end of the oral airway should be at the lip
and the distal tip of the oral airway at the angle of the jaw.
4. Medications: induction agents are commonly used to sedate or ensure
unconsciousness in patients prior to direct laryngoscopy. Muscle
relaxants, either depolarizing or non-depolarizing, are often utilized
to facilitate orotracheal intubation. Vasoconstrictors, such as phenyl-
ephrine, and antihypertensive drugs, such as labetalol, should be
readily available to maintain hemodynamic stability.
Figure 9 Technique for endotracheal intubation. (A) Snifng position. (B) Visualiza-
tion of the glottic opening. (C) Advancement of the endotracheal tube between the
vocal cords under direct visualization. Source: Adapted from Dr. R. Steadman, UCLA.
5. Suction: Yankauer handle, connected to suction, should be available.

It is needed to clear oropharyngeal secretions to improve visualization
during direct laryngoscopy.
6. Manual ventilation (Ambu) bag: a manual ventilation bag permits tem-
porary assisted ventilation and should be connected to 100% oxygen.
7. CO2 detector: a method of detecting CO2 is helpful in conrming ETT
placement, especially when auscultation is equivocal.
I. Technique for Orotracheal Intubation

The basic intubation procedure (91) for an adult is outlined below in a step-
by-step manner (Fig. 9A C):
20 Wong and Hess
1. With the manual ventilation bag connected to 100% oxygen, place the
mask over the patients nose and mouth. Assist ventilation as needed to
maintain adequate oxygen saturation.
2. Make sure the patient is completely supine and adjust the bed to a
comfortable height, usually at the level of the practitioners xiphoid
cartilage. Place the patient in a snifng position by putting pads
under the patients head and gently extending the neck. The head
elevation aligns the oral and pharyngeal axes, whereas the neck exten-
sion aligns the oral axis to create a nearly straight view of the glottic
opening. This position maximizes exposure of the glottis and greatly
enhances the chance of a successful intubation. However, in a
patient with suspected or conrmed neck injury, the neck should be
kept in a neutral position.
3. Administer induction agent of choice and insert the oral airway. It is
crucial that the ability to ventilate the patient is conrmed before
any muscle relaxant is given. In a patient at risk of pulmonary
aspiration, pressure should be applied at the cricoid cartilage to
occlude the esophageal lumen, prevent insufation of the stomach,
and minimize regurgitation of the gastric contents (92). This is
known as the Sellicks maneuver. Cricoid pressure should be discon-
tinued only after the endotracheal placement is conrmed and the
ETT cuff is inated.
4. When adequate anesthetic depth is achieved, remove the oral airway.
Hold the laryngoscope with the left hand and insert the blade from
the right side of the patients mouth.
5. Sweep the tongue to the left and advance the blade until the epiglottis
is visualized. If a MacIntosh blade is used, the tip of the blade should
be placed in the vallecula, which is the space between the base of the
tongue and the epiglottis. If a straight blade is used, the tip of the blade
should be passed beneath the laryngeal surface of the epiglottis.
6. With the blade properly positioned, lift the laryngoscope handle in a
forward and upward motion to visualize the glottic opening. To
prevent dental injury, the practitioner should maintain the wrist in a
neutral position and avoid levering the laryngoscope backward.
7. Clear any secretions with the tip of the Yankauer suction catheter.
Once the glottic opening is well exposed, insert the ETT, usually
with a stylet, from the right side of the blade. Advance the ETT
between the vocal cords under direct vision.
8. After the endotrachael tube is advanced to a depth of about 23 cm in
men and 21 cm in women, inate the cuff to obtain a seal in the
presence of 20 to 30 cm H2O positive airway pressure. With the
right hand holding the ETT, remove the stylet with the left hand.
9. Connect the ETT to the oxygen source and conrm the ETT placement
before securing it with adhesive tape.
J. Conrmation of Proper ETT Placement

Conrming proper ETT position is a crucial step in airway management. Unrec-
ognized misplacement of the ETT (either into the esophagus or a mainstem
bronchus) can lead to serious complications and even death. Evidence suggestive
of proper endotracheal intubation includes equal and bilateral breath sounds,
absence of epigastric airow, condensation in the ETT, and chest movement
with assisted ventilation.
Most practitioners consider the detection of exhaled CO2 as the gold
standard for conrming intra-tracheal intubation. This method is based on the
principle that CO2 is eliminated through the lungs and no CO2 is present in the
stomach. While a CO2 analyzer is part of the anesthesia equipment in the operat-
ing room, a disposable colorimetric end-tidal CO2 detector (93) is often used
in the intensive care unit and the emergency room. This commercially available
device contains a pH-sensitive paper that changes color when CO2 is detected,
such as the single-use Easy Capw II CO2 detector (Fig. 10), which turns from
purple to yellow when CO2 is sensed. However, in the presence of low cardiac
output state or in the absence of pulmonary blood ow, CO2 may not be detected
despite a properly positioned ETT (94). In a multicenter study of a portable
colorimetric end-tidal CO2 detection device, Ornato et al. (95) reported that
endotracheal placement of the ETT was conrmed in a cardiac arrest patient if
more than 0.5% end-tidal CO2 was detected. In contrast, if it detected less than
0.5% end-tidal CO2 during CPR, esophageal intubation remained a possibility.
Figure 10 Single-use CO2 detector.

22 Wong and Hess
Cuff palpation at the suprasternal notch is a reliable method used to reduce

the risk of endobronchial intubation or impingement on the carina (96). After
conrming intra-tracheal intubation, repetitive gentle pressure is applied with
the ngers to the suprasternal notch while the ETT is advanced and withdrawn.
At the same time, the pilot balloon, held in the other hand, is palpated for
expansion. If the balloon distends with the pressure at the suprasternal notch,
the ETT is secured. It is important to keep in mind that this technique is used
not to verify endotracheal location but to prevent endobronchial intubation.
Other useful adjuncts to establish proper ETT position include chest X-ray
examination, beroptic visualization of the carina through the ETT, and use of
the esophageal detector device. The esophageal detector device uses a syringe
or squeeze bulb to aspirate air from the ETT. Easy aspiration of air suggests
tracheal placement, whereas difcult aspiration of air suggests esophageal
placement.
K. Difcult Airway
When endotracheal intubation is unsuccessful with direct laryngoscopy, it is
crucial to call for help from an experienced intubator. Even though the American
Society of Anesthesiologists (97) provides practice guidelines for management
of the difcult airway (Fig. 11), difcult intubation should be handled by
clinicians who are most experienced in airway management. A practitioner not
accustomed to difcult airway management may have trouble remembering or
following the complicated difcult airway algorithm.
As insertion of the laryngeal mask airway (LMA) (Fig. 12) is easy to learn,
it has been utilized in the difcult airway algorithm as a temporary approach
to ventilation (98). Despite its usefulness, the LMA has two major limitations.
First, it does not protect against aspiration. Second, it may fail to provide
adequate ventilation in patients with high airway resistance or low pulmonary
compliance. Therefore, it is contraindicated in individuals with gastroesophageal
reux, full stomach, pregnancy, obesity, or bronchospasm.
Other techniques of intubation or ventilation, such as beroptic intubation
(99), esophageal tracheal Combitube (97,100), intubating LMA (101), and trans-
tracheal jet ventilation (102), all require extensive practice to achieve prociency
and should not be attempted by anyone unfamiliar with them.
Commercially available kits for cricothyrotomy and retrograde tracheal
intubation can also be utilized in the management of the difcult airway. Both
involve entering the cricothyroid membrane to establish a conduit for ventilation.
In a life-threatening situation, emergency cricothyrotomy should be performed
to provide an airway quickly until tracheostomy can be established.
When extubation is planned in a patient with a known difcult airway, a well-
formulated plan must be established for possible re-intubation. A tube exchanger
can be placed to serve as a guide over which the new ETT can be threaded.
Figure 11 Difcult airway algorithm. Source: American Society of Anesthesiologists.

24 Wong and Hess
Figure 12 The laryngeal mask airway.
Although this approach does not guarantee a successful re-intubation, the tube
exchanger can be used to insufate oxygen until a denitive airway is secured.
L. Care of the ETT
Once the endotracheal position is established, the ETT should be secured so that
the distal end of the ETT is approximately 5 cm above the carina, with the head in
a neutral position (103). Taping the ETT at the 23 cm mark in men and the 21 cm
mark in women seems to reduce the risk of endobronchial intubation (104). Once
the proper position is determined, the centimeter mark on the tube at the level of
the patients lip should be recorded and checked on a regular basis. In addition,
cuff pressure should be monitored closely, as excessive pressure can lead to
tracheal mucosa injury and subsequently tracheal stenosis (105,106). Routine
morning radiographic examination should be used to verify ETT position (107).
Furthermore, meticulous efforts and attention should focus on preventing
unplanned self-extubation, especially in patients who have a lengthy stay in
the intensive care unit (108).
M. Complications of Orotracheal Intubation
Complications of tracheal intubation are rare, especially in the controlled

environment of the operating room (109). In contrast, emergency intubation in
critically ill adults carries a much higher rate of complication (110). During
direct laryngoscopy, transient hypertension and tachycardia are frequently
observed. Aspiration of regurgitated gastric contents can occur when the patients
protective reexes are blunted or abolished. If difcult intubation is encountered,
dental injury can result with repeated attempts and increased physical force.
Malpositioning of the ETT is a serious complication and has been reported to
occur more frequently in women after emergency intubation (111). Patients
who are hemodynamically unstable and receiving vasopressor therapy prior to
intubation have the highest mortality rate associated with intubation (110).
IV. Conclusions
Oxygen administration is a basic part in the management of patients with respir-

atory failure. Attention must be directed to select the appropriate oxygen delivery
device to meet the needs of the patient and thus avoid sequelae related to hypoxia.
Airway management plays an important role in ensuring adequate oxygenation
and ventilation for patients. The clinician should be familiar with airway evalu-
ation and orotracheal intubation using direct laryngoscopy. With meticulous
preparation and practice, prociency in basic airway management can be
achieved. If difcult intubation is encountered, assistance from an experienced
clinician in airway management may prove to be life saving.
References
1. Hess DR, MacIntyre NR, Mishoe SC, Galvin WF, Adams AB, Saposnick AB.
Respiratory care: principles and practice. Philadelphia: WB Saunders, 2001.
2. Branson RD, Hess DR, Chatburn RL. Respiratory care equipment. Philadelphia:
J.B. Lippincott Company, 1999.
3. Dexter F, Reasoner DK. Theoretical assessment of normobaric oxygen therapy to
treat pneumocephalus. Anesthesiology 1996; 84:442 447.
4. Annane D, Troche G, Delisle F, et al. Effects of mechanical ventilation with normo-
baric oxygen therapy on the rate of air removal from cerebral arteries. Crit Care
Med 1994; 22:851 857.
5. Weaver LK, Hopkins RO, Chan KJ, et al. Hyperbaric oxygen for acute carbon
monoxide poisoning. N Engl J Med 2002; 347:1057 1067.
6. Green RJ, Dafoe DC, Rafn TA. Necrotizing fasciitis. Chest 1996; 110:219 229.
7. Continuous or nocturnal oxygen therapy in hypoxemic chronic obstructive lung
disease: a clinical trial. Nocturnal Oxygen Therapy Trial Group. Ann Intern Med
1980; 93:391 398.
8. Moloney ED, Kiely JL, McNicholas WT. Controlled oxygen therapy and carbon
dioxide retention during exacerbations of chronic obstructive pulmonary disease.
Lancet 2001; 357:526 528.
9. Hanson CW, Marshall BE, Frasch HF, Marshall C. Causes of hypercarbia with
oxygen therapy in patients with chronic obstructive pulmonary disease. Crit Care
Med 1996; 24:23 28.
10. Aubier M, Murciano D, Milic-Emili J, et al. Effects of the administration of O2 on
ventilation and blood gases in patients with chronic obstructive pulmonary disease
during acute respiratory failure. Am Rev Respir Dis 1980; 122:747 754.
26 Wong and Hess
11. Crossley DJ, McGuire GP, Barrow PM, Houston PL. Inuence of inspired oxygen
concentration on deadspace, respiratory drive, and PaCO2 in intubated patients with
chronic obstructive pulmonary disease. Crit Care Med 1997; 25:1522 1526.
12. Gomersall CD, Joynt GM, Freebairn RC, Lai CK, Oh TE. Oxygen therapy for hyper-
capnic patients with chronic obstructive pulmonary disease and acute respiratory
failure: a randomized, controlled pilot study. Crit Care Med 2002; 30:113116.
13. Robinson TD, Freiberg DB, Regnis JA, Young IH. The role of hypoventilation and
ventilation-perfusion redistribution in oxygen-induced hypercapnia during acute
exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care
Med 2000; 161:1524 1529.
14. Dick CR, Liu Z, Sassoon CS, Berry RB, Mahutte CK. O2-induced change in
ventilation and ventilatory drive in COPD. Am J Respir Crit Care Med 1997;
155:609 614.
15. Dunn WF, Nelson SB, Hubmayr RD. Oxygen-induced hypercarbia in obstructive
pulmonary disease. Am Rev Respir Dis 1991; 144:526 530.
16. Chien JW, Ciufo R, Novak R, et al. Uncontrolled oxygen administration and
respiratory failure in acute asthma. Chest 2000; 117:728 733.
17. Hedenstierna G, Rothen HU. Atelectasis formation during anesthesia: causes and
measures to prevent it. J Clin Monit 2000; 16:329 335.
18. White AC. The evaluation and management of hypoxemia in the chronic critically
ill patient. Clin Chest Med 2001; 22:123 134.
19. Carvalho CR, Schettino GP, Maranhao B, Bethlem EP. Hyperoxia and lung disease.
Curr Opin Pulm Med 1998; 4:300 304.
20. Ingrassia TS, Ryu JH, Trastek VF, Rosenow EC. Oxygen-exacerbated bleomycin
pulmonary toxicity. Mayo Clin Proc 1991; 66:173 178.
21. Allen J, Zwerdling R, Ehrenkranz R, et al. Statement on the care of the child with
chronic lung disease of infancy and childhood. Am J Respir Crit Care Med 2003;
168:356 396.
22. Jobe AH, Bancalari E. Bronchopulmonary Dysplasia. Am J Respir Crit Care Med
2001; 163:1723 1729.
23. Campbell EJ, Baker MD, Crites-Silver P. Subjective effects of humidication of
oxygen for delivery by nasal cannula. A prospective study. Chest 1988; 93(2):
289 293.
24. Hess D, Figaszewski E, Henry D, Hoffman S, Pino D. Subjective effects of dry
versus humidied low ow oxygen on the upper respiratory tract. Respir Ther
1982(Nov/Dec); 12:71 75.
25. Estey W. Subjective effects of dry versus humidied low ow oxygen. Respir Care
1980; 25:1143 1144.
26. Stausholm K, Rosenberg-Adamsen S, Skriver M, Kehlet H, Rosenberg J.
Comparison of three devices for oxygen administration in the late postoperative
period. Br J Anaesth 1995; 74:607 609.
27. McBrien ME, Sellers WF. A comparison of three variable performance devices for
postoperative oxygen therapy. Anaesthesia 1995; 50:136 138.
28. Nolan KM, Winyard JA, Goldhill DR. Comparison of nasal cannulae with
face mask for oxygen administration to postoperative patients. Br J Anaesth
1993; 70:440 442.
29. Ooi R, Joshi P, Soni N. An evaluation of oxygen delivery using nasal prongs.
Anaesthesia 1992; 47:591 593.
30. Williams AB, Jones PL, Mapleson WW. A comparison of oxygen therapy devices
used in the postoperative recovery period. Anaesthesia 1988; 43:131 135.
31. Redding JS, McAfee DD, Gross CW. Oxygen concentrations received from com-
monly used delivery systems. South Med J 1978; 71:169 172.
32. Schacter EN, Littner MR, Luddy P, Beck GJ. Monitoring of oxygen delivery
systems in clinical practice. Crit Care Med 1980; 8:405 409.
33. Bazuaye EA, Stone TN, Corris PA, Gibson GJ. Variability of inspired oxygen
concentration with nasal cannulas. Thorax 1992; 47:609 611.
34. Gibson RL, Comer PB, Beckham RW, McGraw CP. Actual tracheal oxygen concen-
trations with commonly used oxygen equipment. Anesthesiology 1976; 44:7173.
35. Waldau T, Larsen VH, Bonde J. Evaluation of ve oxygen delivery devices
in spontaneously breathing subjects by oxygraphy. Anaesthesia 1998; 53:256263.
36. Ward JJ, Gracey DR. Arterial oxygen values achieved by COPD patients breathing
oxygen alternately via nasal mask and nasal cannula. Respir Care 1985; 30:
250 255.
37. Hess D, DAgostino D, Magrosky S, Myers C, Shuman J. Effect of nasal cannula
displacement on arterial PO2. Respir Care 1984; 29:21 24.
38. Faireld JE, Goroszeniuk T, Tully AM, Adams AP. Oxygen delivery systemsa
comparison of two devices. Anaesthesia 1991; 46:135 138.
39. Dunlevy CL, Tyl SE. The effect of oral versus nasal breathing on oxygen
concentrations received from nasal cannulas. Respir Care 1992; 37:357 360.
40. Christopher KL. Transtracheal oxygen catheters. Clin Chest Med 2003;
24:489 510.
41. Hoffman LA. Novel strategies for delivering oxygen: reservoir cannula,
demand ow, and transtracheal oxygen administration. Respir Care 1994;
39:363 379.
42. Bliss PL, McCoy RW, Adams AB. Characteristics of demand oxygen delivery
systems: maximum output and setting recommendations. Respir Care 2004;
49:160 165.
43. Jensen AG, Johnson A, Sandstedt S. Rebreathing during oxygen treatment with face
mask. The effect of oxygen ow rates on ventilation. Acta Anaesthesiol Scand
1991; 35:289 292.
44. Milross J, Young IH, Donnelly P. The oxygen delivery characteristics of the
Hudson Oxy-one face mask. Anaesth Intensive Care 1989; 17:180 184.
45. Boumphrey SM, Morris EA, Kinsella SM. 100% inspired oxygen from a Hudson
maska realistic goal? Resuscitation 2003; 57:69 72.
46. Campbell DJ, Faireld MC. The delivery of oxygen by a venturi T piece.
Anaesthesia 1996; 51:558 560.
47. Jeffrey AA, Warren PM. Should we judge a mask by its cover? Thorax 1992;
47:543 546.
48. Hunter J, Olson LG. Performance of the Hudson Multi-Vent oxygen mask. Med J
Aust 1988; 148:444 447.
49. Hill SL, Barnes PK, Hollway T, Tennant R. Fixed performance oxygen masks: an
evaluation. Br Med J (Clin Res Ed) 1984; 288:1261 1263.
50. Foust GN, Potter WA, Wilons MD, Golden EB. Shortcomings of using two jet
nebulizers in tandem with an aerosol face mask for optimal oxygen therapy.
Chest 1991; 99:1346 1351.
28 Wong and Hess
51. Hess D, Chatmongkolchart S. Techniques to avoid intubation: noninvasive

positive pressure ventilation and heliox therapy. Int Anesthesiol Clin 2000;
38(3):161 187.
52. Chevrolet JC. Helium oxygen mixtures in the intensive care unit. Crit Care 2001;
5(4):179 181.
53. Rodrigo GJ, Rodrigo C, Pollack CV, Rowe B. Use of helium oxygen mix-
tures in the treatment of acute asthma: a systematic review. Chest 2003;
123(3):891 896.
54. Ho AM, Lee A, Karmakar MK, Dion PW, Chung DC, Contardi LH. Heliox vs. air
oxygen mixtures for the treatment of patients with acute asthma: a systematic
overview. Chest 2003; 123(3):882 890.
55. Grosz AH, Jacobs IN, Cho C, Schears GJ. Use of helium oxygen mixtures to
relieve upper airway obstruction in a pediatric population. Laryngoscope 2001;
111(9):1512 1514.
56. Weber JE, Chudnofsky CR, Younger JG, et al. A randomized comparison of
helium oxygen mixture (Heliox) and racemic epinephrine for the treatment of
moderate to severe croup. Pediatrics 2001; 107(6):E96.
57. Jaber S, Carlucci A, Boussarsar M, Fodil R, Pigeot J, Maggiore S, Harf A, Isabey D,
Brochard L. Helium oxygen in the postextubation period decreases inspiratory
effort. Am J Respir Crit Care Med 2001; 164(4):633 637.
58. Reuben AD, Harris AR. Heliox for asthma in the emergency department: a review
of the literature. Emerg Med J 2004; 21(2):131 135.
59. Bag R, Bandi V, Fromm RE, Jr, Guntupalli KK. The effect of heliox-driven bronch-
odilator aerosol therapy on pulmonary function tests in patients with asthma.
J Asthma 2002; 39(7):659 665.
60. Piva JP, Menna Barreto SS, Zelmanovitz F, Amantea S, Cox P. Heliox versus
oxygen for nebulized aerosol therapy in children with lower airway obstruction.
Pediatr Crit Care Med 2002; 3(1):6 10.
61. Kress JP, Noth I, Gehlbach BK, et al. The utility of albuterol nebulized with
heliox during acute asthma exacerbations. Am J Respir Crit Care Med 2002;
165(9):1317 1321.
62. Rose JS, Panacek EA, Miller P. Prospective randomized trial of heliox-driven
continuous nebulizers in the treatment of asthma in the emergency department.
J Emerg Med 2002; 22(2):133 137.
63. deBoisblanc BP, DeBleiux P, Resweber S, Fusco EE, Summer WR. Randomized
trial of the use of heliox as a driving gas for updraft nebulization of bronchodilators
in the emergent treatment of acute exacerbations of chronic obstructive pulmonary
disease. Crit Care Med 2000; 28(9):3177 3180.
64. Hess DR, Acosta FL, Ritz RH, Kacmarek RM, Camargo CA, Jr. The effect of
heliox on nebulizer function using a beta-agonist bronchodilator. Chest 1999;
115(1):184 189.
65. Pecchiari M, Pelucchi A, DAngelo E, Foresi A, Milic-Emili J. Effect of heliox
breathing on dynamic hyperination in COPD patients. Chest 2004;
125(6):2075 2082.
66. Gerbeaux P, Gainnier M, Boussuges A, et al. Use of heliox in patients with severe
exacerbation of chronic obstructive pulmonary disease. Crit Care Med 2001;
29(12):2322 2324.
67. Tassaux D, Jolliet P, Roeseler J, Chevrolet JC. Effects of helium oxygen on intrin-
sic positive end-expiratory pressure in intubated and mechanically ventilated
patients with severe chronic obstructive pulmonary disease. Crit Care Med 2000;
28(8):2721 2728.
68. Diehl JL, Mercat A, Guerot E, et al. Helium/oxygen mixture reduces the work
of breathing at the end of the weaning process in patients with severe chronic
obstructive pulmonary disease. Crit Care Med 2003; 31(5):1415 1420.
69. Jolliet P, Tassaux D, Roeseler J, et al. Helium oxygen versus air oxygen nonin-
vasive pressure support in decompensated chronic obstructive disease: a prospec-
tive, multicenter study. Crit Care Med 2003; 31(3):878 884.
70. Jaber S, Fodil R, Carlucci A, et al. Noninvasive ventilation with helium oxygen in
acute exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit
Care Med 2000; 161(4 Pt 1):1191 1200.
71. Jolliet P, Tassaux D, Thouret JM, Chevrolet JC. Benecial effects of helium:oxygen
versus air:oxygen noninvasive pressure support in patients with decompensated
chronic obstructive pulmonary disease. Crit Care Med 1999; 27(11):2422 2429.
72. Austan F, Polise M. Management of respiratory failure with noninvasive positive
pressure ventilation and heliox adjunct. Heart Lung 2002; 31(3):214 218.
73. Chatmongkolchart S, Kacmarek RM, Hess DR. Heliox delivery with non-
invasive positive pressure ventilation: a laboratory study. Respir Care 2001;
46(3):248 254.
74. Tassaux D, Jolliet P, Thouret JM, Roeseler J, Dorne R, Chevrolet JC. Calibration of
seven ICU ventilators for mechanical ventilation with helium oxygen mixtures.
Am J Respir Crit Care Med 1999; 160(1):22 32.
75. Jolliet P, Tassaux D. Usefulness of helium oxygen mixtures in the treatment of
mechanically ventilated patients. Curr Opin Crit Care 2003; 9(1):45 50.
76. Berkenbosch JW, Grueber RE, Dabbagh O, McKibben AW. Effect of helium
oxygen (heliox) gas mixtures on the function of four pediatric ventilators. Crit
Care Med 2003; 31(7):2052 2058.
77. Devabhaktuni VG, Torres A Jr, Wilson S, Yeh MP. Effect of nitric oxide, peruor-
ocarbon, and heliox on minute volume measurement and ventilator volumes
delivered. Crit Care Med 1999; 27(8):1603 1607.
78. Oppenheim-Eden A, Cohen Y, Weissman C, Pizov R. The effect of helium on ven-
tilator performance: study of ve ventilators and a bedside Pitot tube spirometer.
Chest 2001; 120(2):582 588.
79. Natanson C, Shelhamer JH, Parrillo JE. Intubation of the trachea in the critical care
setting. JAMA 1985; 253:1160 1165.
80. Murrin KR. Intubation prcedure and causes of difcult intubation. In: Latto IP,
Rosen M, eds. Difculties in Tracheal Intubation. London: Baillere Tindall,
1985:75 89.
81. Langeron O, Masso E, Huraux C, et al. Prediction of difcult mask ventilation.
Anesthesiology 2000; 92:1229 1236.
82. Mallampati SR, Gatt SP, Gugino LD, et al. A clinical study to predict difcult
tracheal intubation. A prospective study. Can J Anaesth 1985; 32:429 434.
83. Frerk CM. Predicting difcult intubtion. Anaesthesia 1991; 46:1005 1008.
84. Cormack RS, Lehane J. Difcult tracheal intubation in obstetrics. Anaesthesia
1984; 39:1105 1111.
85. Wilson M, Spiegelhalter D, Robertson J, Lesser P. Predicting difcult intubation.
Br J Anaesth 1988; 61:211 216.
86. Knill RL. Difcult laryngoscopy made easy with a BURP. Can J Anaesth 1993;
40:279 282.
30 Wong and Hess
87. Takahata O, Kubota M, Mamiya K, et al. The efcacy of the BURP maneuver
during a difcult laryngoscopy. Anesth Analg 1997; 84:419 421.
88. Isono S, Tanaka A, Sho Y, et al. Advancement of the mandible improves velophar-
yngeal airway patency. J Appl Physiol 1995; 79:2132 2138.
89. Murashima K, Fukutome T. Effect of jaw-thrust maneuver on the laryngeal inlet.
Anaesthesia 1998; 53:203 204.
90. Tamura M, Ishikawa T, Kato R, et al. Mandibular advancement improves the lar-
yngeal view during direct laryngoscopy performed by inexperienced physicians.
91. Stoelting RK, Miller RD. Basics of Anesthesia. New York: Churchill Livingstone,
1994.
92. Sellick B. Cricoid pressure to control regurgitation of stomach contents during
induction of anesthesia. Lancet 1961; 2:404 406.
93. Anton WR, Gordon RW, Jordan TM, Posner KL, Cheney FW. A disposable
end-tidal detector to verify endotracheal intubation. Ann Emerg Med 1991;
20:271 275.
94. Sayah AJ, Peacock WF, Overton DT. End-tidal measurement in the detec-
tion of esophageal intubation during cardiac arrest. Ann Emerg Med 1990;
19:857 860.
95. Ornato J, Shipley J, Racht E, et al. Multicenter study of a portable, hand-size,
colorimetric end-tidal carbon dioxide detection device. Ann Emerg Med 1992;
21:518 523.
96. Pollard RJ, Lobato EM. Endotracheal tube location veried reliably by cuff palpa-
tion. Anesth Analg 1995; 81:135 138.
97. Practice guidelines for management of the difcult airway. An updated report by the
American Society of Anesthesiologists Task Force on Management of the difcult
airway. Anesthesiology 2003; 98:1269 1277.
98. Benumof, JL. Laryngeal mask airway and the ASA difcult airway algorithm.
99. Benumof JL. Management of the difcult adult airway, with special emphasis on
awake tracheal intubation. Anesthesiology 1991; 75:1087 1110.
100. Baskett PJ, Bossaert, Carli P, et al. Guidelines for the advanced management of the
airway and ventilation during resuscitation. Resuscitation 1996; 31:201 230.
101. Baskett PJ, Parr MJ, Nolan JP. The intubating laryngeal mask: results of a multicen-
ter trial with experience of 500 cases. Anaesthesia 1998; 53:1174 1179.
102. Benumof JL, Scheller MS. The importance of transtracheal jet ventilation in the
management of the difcult airway. Anesthesiology 1989; 71:769 778.
103. Henschke CI, Yankelevitz DF, Wand A, et al. Accuracy and efcacy of chest
radiography in the intensive care unit. Radiol Clin North Am 1996; 34:21 31.
104. Owen RL, Cheney FW. Endobronchial intubation: a preventable complication.
105. Cooper JD, Grillo HC. The evolution of tracheal injury due to ventilatory assistance
through cuffed tubes: a pathologic study. Ann Surg 1969; 169:334 348.
106. Cooper JD, Grillo HC. Analysis of problems related to cuffs on intratracheal tubes.
Chest 1972; 62:21S 27S.
107. Bekemeyer WB, Crapo RO, Calhoon S, et al. Efcacy of chest radiography in a
respiratory intensive care unit. Chest 1985; 88:691 696.
108. Atkins PM, Mion LC, Mendelson W, et al. Characteristics and outcomes of patients
who self-extubate from ventilatory support. Chest 1997; 112:1317 1323.
109. Keane WM, Rowe LD, Denneny JC, et al. Complications of intubation. Ann Otol
Rhinol Laryngol 1982; 91:584 587.
110. Schwartz DE, Matthay MA, Cohen NH. Death and other complications of
emergency airway management in critically ill adults. Anesthesiology 1995;
82:367 376.
111. Schwartz DE, Lieberman JA, Cohen NH. Women are at greater risk for malposi-
tioning of the endotracheal tube after emergent intubation. Crit Care Med 1994;
22:1127 1131.
2
Non-Invasive Ventilation in Critical Care
GUY W. SOO HOO
Pulmonary and Critical Care Section

West Los Angeles Healthcare Center, VA Greater Los Angeles
Healthcare System and
I. Introduction
Non-invasive ventilation mechanical (NIMV), broadly dened as ventilatory

support without an endotracheal tube, has gained an increasing acceptance in
the management of patients with acute respiratory failure. Its role has evolved
over the past decade to the point that it may be part of rst-line management
in those who do not require immediate intubation and has been endorsed in
numerous reviews and consensus recommendations (1 3). Most of the experi-
ence with NIMV has been with non-invasive positive pressure ventilation
(NPPV), either pressure or volume cycled, but ventilatory support can also be
provided with continuous positive airway pressure (CPAP) or negative pressure
ventilation (NPV).
Respiratory failure is generally characterized by its effects on gas exchange
and reects conditions in which the demands on the ventilatory system exceed its
ability to maintain adequate gas exchange. This is broadly separated into type I or
hypoxemic respiratory failure and type II or hypercapnic respiratory failure, and
patients may present with a combination of the two. Assisted ventilatory support
delivered through an endotracheal tube or tracheostomy tube connected to a
volume or pressure ventilator has been the standard of therapy since its introduc-
tion during the height of the polio epidemics of the 1950s (4). Although most
patients are successfully treated with intubation and mechanical ventilation,
the risks and associated adverse effects are legion and have always tempered
33
34 Soo Hoo
their benecial effects (5,6). It is these complications (airway trauma, baro-

trauma, nosocomial sinusitis, pneumonia, tracheal stenosis, etc.) that spurred
investigation into alternative methods of support. This was facilitated by the
success of nasal CPAP in patients with obstructive sleep apnea. Success in
patients with neuromuscular disease using a mask or mouthpiece led to its appli-
cation in other conditions (7). In addition, there was increasing recognition that
many of the patients with severe underlying lung disease and respiratory
failure may not recover from the episode or be destined for ventilator dependence
(8). This provided additional impetus for NIMV as an effective alternative that
also spared patients the long-term morbidities associated with intubation.
The following sections focus on diseases most amenable to NIMV, high-
lighting crucial clinical investigations to provide a better understanding of the
role of NIMV in the management of the critically ill patient. Later sections
will review practical issues in the application of NIMV.
II. Rationale for Use

A. Demonstration of Efcacy
Ventilatory support can be effectively delivered through a nasal or face mask
with improvement in both physiologic and clinical variables. Short-term appli-
cation of nasal NPPV in normal and stable patients with obstructive or restrictive
lung disease provided ventilatory support and unloaded the ventilatory muscles,
with reduction in transdiaphragmatic pressure, diaphragmatic electromyogram
(EMG), and other indices of muscle work (9,10). In patients with exacerbations
of underlying chronic obstructive pulmonary disease (COPD), face mask NPPV
with pressure support ventilation (PSV) decreased transdiaphragmatic pressure
and diaphragmatic EMG. This produced increases in tidal volume and a reduction
in respiratory rate associated with correction of hypoxemia, hypercapnia, and res-
piratory acidosis (11). This also decreased the intubation rate, duration of
ventilation, and intensive care unit (ICU) stay compared with historical controls,
but had no impact on mortality. Others have reported comparable results and
noted improvement in pH and PaCO2 after as little as one hour of treatment (12).
These reports spurred a variety of investigations, utilizing nasal and orona-
sal face masks for pressure and volume ventilation with nearly uniform success in
correcting a patients respiratory acidosis and avoiding intubation. In an early
summary of these reports, the success rate for NIMV was 67% with an aggregate
initial pH of 7.30 in treated patients (13). These trials primarily involved COPD
patients, and most experience has continued to be in those with COPD.
A few general principles that emerged from this initial experience are sum-
marized in Table 1. Patients who require immediate intubation, are deeply coma-
tose, or are close to cardiopulmonary arrest are not candidates for NIMV.
Denitive therapy should not be delayed by a trial of NIMV. Another crucial
and often unappreciated element is that the most favorable results with NIMV
Non-Invasive Ventilation in Critical Care 35
Table 1 Non-Invasive Ventilation: General Guidelines for

Patient Selection
Absolute contraindications
Coma
Respiratory arrest
Cardiac arrest
Any condition requiring immediate intubation
Clinical conditions
Rapidly reversible clinical conditions
COPD exacerbations
Cardiogenic pulmonary edema
After discontinuation of mechanical ventilation
Early extubation
? Post-extubation
? Pneumonia
COPD and pneumonia
Use non-invasive ventilation concomitantly with other therapy
Abbreviation: COPD, chronic obstructive pulmonary disease.
have been in patients with rapidly reversible conditions. The greatest experience
has accrued in these conditions. This virtually eliminates candidates with con-
ditions that may require an extended period (weeks) of ventilatory support or
an illness that may require weeks for resolution. Of course, patients with mild
disease may be candidates for NIMV, but must be evaluated on a case by case
basis. Patients also need to be treated concomitantly with therapy directed at
their underlying disease. NIMV provides an important adjunct to therapy, best
suited for those who require a brief period of support to prevent frank respiratory
failure while allowing the other components of therapy to take effect. These clini-
cal conditions will be examined in detail in the following sections.
III. Clinical Conditions

A. Chronic Obstructive Pulmonary Disease
Most experience with NIMV has been in patients with acute exacerbations of
COPD. These patients often develop dynamic hyperination, ineffective venti-
lation, incipient ventilatory muscle fatigue, and hypercapnia in association
with their illness. NIMV has been demonstrated to be particularly effective in
these patients, providing ventilatory support and unloading the ventilatory
muscles to facilitate recovery. The often rapid reversibility in these exacerbations
seems particularly suited for NIMV. There are over a dozen prospective random-
ized trials comparing NPPV to standard therapy in the treatment of these patients.
This section will focus on NPPV, though there are reports of efcacy, but no ran-
domized trials, with NPV (14).
36 Soo Hoo
Randomized Trials
The rst prospective randomized trial, reported by Bott et al. (15), involved 60
COPD patients and compared NPPV via nasal mask volume-cycled ventilation
plus conventional treatment with conventional treatment administered in a
specialized ward setting. They demonstrated greater improvement in respiratory
acidosis (pH and PaCO2) in treated patients compared with conventional therapy,
along with lower dyspnea scores. Mortality was lower but did not reach statistical
signicance. Those who died were more acidemic (pH 7.31 vs. 7.35) and hyper-
capnic (PaCO2 71 vs. 64 mmHg) on admission than survivors.
Another single-center study reported by Kramer et al. (16), compared nasal
mask bi-level positive airway pressures (BiPAP) plus standard therapy to standard
therapy alone in 31 patients (23 COPD). The intubation rate was markedly reduced
in the BiPAP group, 31% versus 73%, and 9% versus 67% for COPD patients, but
with no difference in mortality. They noted marked improvement in PaCO2, with a
decline in heart rate, respiratory rate, and dyspnea score after just one hour of
BiPAP.
In a multi-center, multi-national study reported by Brochard et al. (17), 85
patients with COPD exacerbations were randomized to either NPPV with face
mask pressure support of 20 cmH2O plus conventional therapy or to conventional
therapy alone. Endotracheal intubation was signicantly lower in the non-invasive
group (26% vs. 74%). The response rates were not different between the participat-
ing institutions, and this treatment modality was demonstrated effective in more
than one institution. The non-invasive group had fewer complications, specically
fewer episodes of nosocomial pneumonia and sepsis, lower mortality rate, and
shorter length of stay. It should be pointed out, though, that the one-hour decline
in PaCO2 noted by others was not evident in this study.
The largest multi-center, prospective, randomized trial, reported by Plant
et al., compared BiPAP using either face or nasal masks and randomized 236
patients (7,18). This occurred in a non-specialist ward setting in contrast to the
previously reported ward and ICU settings. They focused on those with mild
to moderate acidosis with a pH 7.25 as an exclusion criterion. Patients were
treated according to an established protocol, primarily initiated by nursing
staff. The primary outcome measure was the need for intubation, but once
this criterion was met, intubation was at the discretion of the attending physician
and management could continue with NPPV or even standard medical care. The
non-invasive group had a statistically signicantly lower number of patients who
met the need for intubation, 15% compared with 27%, as well as a lower mor-
tality rate, 10% versus 20%. However, there was much less difference in the
actual intubation rate (NPPV, 6% vs. 10%). In a sub-group analysis, the differ-
ence was most evident in those with a pH . 7.30. There was no benet of
NPPV in patients with an initial pH , 7.30.
The methodology of these and other trials have varied widely, ranging from
single-center to multi-center trials, different mask interfaces (nasal, face, or
both), different modes of assisted ventilation (volume vs. pressure), with different
inclusion and exclusion criteria, outcome measures, intubation criteria, and even
cross-over therapy, making consistent comparisons difcult. These studies often
represent narrowly selected groups of patients. In the trial reported by Brochard
et al., only 31% of those evaluated were enrolled in their study. Intubation is
probably the most important outcome measure, but conclusions are tempered
when one outcome is meeting criteria for intubation and the other is intubation.
Some trials enrolled relatively small numbers of patients, and although with sig-
nicant differences, they are limited by potential problems with statistical power.
Meta-analyses may help to provide better insight into the effects of therapy by
pooling patients and ameliorating differences in study design.
Meta-Analyses
Three meta-analyses have been reported, largely focusing on the use of NPPV in
COPD patients, although other causes of respiratory failure are included (19 22).
They have all reached the same general conclusion that NPPV reduces the intu-
bation rate, mortality, and length of stay in COPD patients. Peter et al. analyzed
15 studies, eight involving patients with COPD exacerbations. Focusing on the
reports of COPD patients, they noted a signicant reduction in mortality of
13% (95%CI: 6 21%), need for intubation of 18% (95%CI: 3 33%), and
decrease in hospital length of stay of 5.7 days (95%CI: 1.2 10.1 days).
Keenan et al. restricted their analysis to COPD patients and found a risk reduction
of 10% (95%CI: 5 15%) in mortality, a risk reduction of 28% (95%CI: 15 40%)
for endotracheal intubation, and a decrease in the length of hospital stay by 4.6
days (95%CI: 2.3 6.8 days). They also performed sub-group analysis and
found NPPV had the most impact in those with moderately severe exacerbations
dened by a pH , 7.30. It should be noted that patients in two studies reporting
no benet with NPPV had very mild disease with an average pH of 7.33 + 0.01
and 7.40 + 0.04, respectively. This may have skewed their analysis against those
with less severe disease, as they did not identify any differences in outcome in
those with less severe exacerbations. This conclusion runs contrary to the experi-
ence reported by Plant et al., who found that the greatest benet appeared to be in
those with less severe disease dened by a pH . 7.30.
In summary, the majority of reported data supports the use of NIMV in
COPD. It is particularly effective in patients with hypercapnic respiratory
failure. The optimal thresholds of treatment remain to be dened. There are
reports matching our own experience that COPD patients with an initial pH
range 7.10 can be successfully managed with NIMV, with the best results prob-
ably in patients with an initial pH . 7.25 (moderate severity). It does not add to
the management of patients with mild exacerbations (dened by a pH . 7.35).
When successful, it decreases the intubation rate, length of stay, and mortality
of these patients. In addition to reversing respiratory failure, mortality seems to
be decreased because of fewer episodes of nosocomial pneumonia. There
38 Soo Hoo
exists great heterogeneity in the study design of these trials, and generalization is
tempered by differences in many areas, including threshold for intubation.
B. Cardiogenic Pulmonary Edema

Respiratory failure due to cardiogenic pulmonary edema is a rapidly reversible
condition, much like decompensated COPD, that is amenable to treatment with
NIMV. The adverse effects of cardiogenic pulmonary edema are related to the
development of pulmonary vascular congestion and interstitial and alveolar
edema. This leads to hypoxemia with metabolic acidosis and hypercapnia preced-
ing frank respiratory failure. The benecial effects of positive pressure venti-
lation include increases in functional residual capacity, breathing on a more
compliant portion of the lungs pressure volume curve, recruitment of alveolar
lung units, improving ventilation/perfusion relationships, and gas exchange (23).
The sum of these effects is improved oxygenation and ventilation, unloading of
the respiratory muscles, and lower work of breathing. Positive intrathoracic
pressure also improves left ventricular function (24). Positive intrathoracic
pressure decreases transmural systolic left ventricular pressure, which reduces
the afterload to the left heart, improving left ventricular emptying and cardiac
output. Reductions in venous return also decrease left ventricular preload. This
has a salutary effect in patients with pulmonary edema who are volume over-
loaded. It is important to note that the benecial effects of positive airway
pressure are present as long as preload is elevated. If preload decreases, the detri-
mental effects of positive pressure on venous return may outweigh the benecial
effects on left ventricular function, resulting in decreased cardiac output and
arterial pressure.
The benets of positive pressure ventilation in cardiogenic pulmonary
edema have been noted in endotracheally intubated and ventilated subjects.
Improvement might be expected with CPAP, which represents augmented
airway pressure in its most basic form. This represents the most rudimentary
form of NIMV but requires an intact ventilatory drive, as it will not support
apneic patients. Investigations using nasal CPAP in the 5 to 15 cm H2O range
in heart failure patients have demonstrated reduction in the work of breathing,
with over 30% reduction in esophageal pressures, and improvement in cardiovas-
cular function, with greater than 15% improvement in stroke volume and cardiac
index (25 27).
Randomized Trials: CPAP
There has been a long-standing recognition of the benet of CPAP in cardiogenic

pulmonary edema (28), but randomized trials were not performed for almost
50 years following the initial reports. Three noteworthy trials compared face
mask CPAP either xed at 10 cm H2O or titrated from 2.5 to 12.5 cm H2O
along with conventional therapy (oxygen, nitrates, diuretics, etc.) to conventional
therapy alone (29 31). Rasanen et al. utilized face mask CPAP at 10 cm H2O in a
total of 40 patients and noted improvement in respiratory rate, heart rate, blood
pressure, PaCO2, and oxygenation after as little as 10 minutes of support. The
intubation rate in CPAP patients was 30%, compared with 60% in control
patients. Bersten et al. compared face mask CPAP at 10 cm H2O in a total of
39 patients with severe pulmonary edema. Their series is noteworthy for the
severity of illness of their patients, with mean APACHE II scores of 20 and arte-
rial blood gases in the CPAP-treated patients of a pH of 7.18 + 0.08, PaCO2 of
58 + 8 mmHg, and PaO2/FIO2 of 138 + 32. Control patients had a similar
severity of disease. None of the CPAP patients required intubation, compared
with 35% of the controls. Lin et al. reported the largest series with CPAP, invol-
ving a total of 100 patients, all with invasive hemodynamic monitoring, treated
with face mask CPAP and titrated over a three-hour period from 2.5 to
12.5 cm H2O. Their results mirrored that of the two prior studies but also demon-
strated a reduction in shunt fraction and improvement in cardiac indices with
CPAP. The intubation rate in CPAP patients was 16%, compared with 36% in
controls.
The improvement in outcome measures did not achieve statistical signi-
cance in these studies, but there were strong trends towards a reduction in intuba-
tion rate and mortality. In a meta-analysis of the three studies, pooled results
demonstrated a risk reduction in intubation of 26% (95%CI: 13 38%), with a
trend towards decreased mortality of 6.6% (95%CI: 23 16%) (32). The duration
of hospitalization was not different, although there was a decrease in the ICU stay
of patients in the study by Bersten et al. by about a day. In another randomized
study in patients with pulmonary edema complicating acute myocardial infarc-
tion, Takeda et al. provided 4 to 10 cm H2O nasal CPAP to 11 patients and
demonstrated a reduction in intubation rate (18% vs. 73%), mortality (6% vs.
64%), and cardiac hemodynamics (pulmonary artery wedge pressure, cardiac
index, stroke volume index) with CPAP.
Randomized Trials: NPPV
Given this experience with CPAP, it follows that NPPV would be equally or even
more efcacious in cardiogenic pulmonary edema. This might be expected
because NPPV breaths are augmented and assisted, unlike CPAP breaths.
Work of breathing would be expected to be less with NPPV than CPAP. While
several investigators have included patients with cardiogenic pulmonary
edema, four have focused on the addition of NPPV to conventional therapy in
cardiogenic pulmonary edema (33 35). All used a combination of pressure
cycled ventilation with positive end-expiratory pressure (PEEP), either PSV
with PEEP or BiPAP with bedside or portable ventilators. Three were conducted
at a single institution, with the largest a multi-center study. Demonstration of
benet with NPPV was not as evident as in the other studies that used CPAP.
Masip et al. (33) randomized a total of 40 patients and noted a lower
intubation rate (5% vs. 33%) with more rapid improvement in tachypnea,
40 Soo Hoo
oxygenation, and hypercapnia with pressure support NPPV than conventionally

treated patients (oxygen, diuretics, etc.). Improvement occurred in as little as
15 minutes. Hypercapnia in the control group was associated with intubation
(44%) as opposed to none of the hypercapnic patients treated with NPPV. The
mean amount of pressure support was 15.2 + 2.4 cm H2O in addition to
5 cm H2O of PEEP. No differences were noted in the length of stay and mortality.
This contrasts with the ndings by Sharon et al. (34), who randomized 40 patients
to either BiPAP and intravenous nitroglycerin or high-dose intravenous nitrogly-
cerin. The mean level of ventilatory support was an inspiratory positive airway
pressure (IPAP) 9.3 + 2.3 cm H2O and expiratory positive airway pressure
(EPAP) 4.2 + 3.1 cm H2O. This is relatively low compared with other inves-
tigators, and they never improved their study patients mean oxygenation. The
investigators found that 80% of the BiPAP group required intubation compared
with 20% of controls, with a BiPAP group myocardial infarction rate of 55%
compared with 10% in controls. The amount of ventilatory support was probably
suboptimal in this study, and the high intubation rate and infarction rate is quite
unusual. In another study initiated in the Emergency Department, Leavitt ran-
domized 38 patients with congestive heart failure to either BiPAP or high-ow
oxygen, in addition to conventional therapy. There was a trend toward a lower
intubation and myocardial infarction rate in the BiPAP group, but no outcome
measures were statistically signicant (36).
Nava et al. (35) conducted a multi-center study, randomizing a total of 130
patients to either conventional therapy with face mask PSV and PEEP or conven-
tional therapy. They used a portable ventilator and provided an average inspira-
tory pressure support of 14.5 + 2.1 cm H2O and PEEP of 6.1 + 3.2 cm H2O.
While NPPV provided more rapid improvement in tachypnea, dyspnea, and oxy-
genation, no differences were noted in the intubation rate or mortality. Sub-group
analysis indicated a reduction in the intubation rate (6% vs. 29%) in hypercapnic
patients (PaCO2 . 45 mmHg).
Randomized Trials: CPAP vs. NPPV

Both of these modes of NIMV have undergone direct comparison in patients with
cardiogenic pulmonary edema. In a small efcacy study, the two modes were
compared in a cross-over study in six patients that included esophageal pressure
monitoring and pressure monitoring with a pulmonary artery catheter (37).
Evaluation was made in 20-minute blocks, and patients treated with NPPV deliv-
ered at an inspiratory pressure support of 10 cm H2O and PEEP of 5 cm H2O had
greater reduction in indices of ventilatory muscle work than CPAP of 10 cm H2O,
but no difference was noted in cardiac indices between the two modes.
Mehta et al. (38) randomized a total of 27 patients to either nasal CPAP at
10 cmH2O or BiPAP with settings of IPAP/EPAP 15/5 cm H2O, both in
addition to standard therapy. The ventilatory support was delivered using
the same device. Within 30 minutes, patients randomized to BiPAP had
improvement of more indices (tachypnea, hypercapnia, dyspnea, pH, and mean

arterial pressure) than the CPAP group (tachypnea), but the differences
between the two groups were not signicant. However, over the course of the
hospitalization, there was a higher incidence of myocardial infarction (71%) in
the BiPAP group compared with the CPAP group (31%) and a group of
matched historical controls (38%). The latter nding resulted in termination
of the investigation. On review of patient characteristics, a greater number of
patients in the BiPAP group had chest pain on study entry (71% vs. 31%),
suggesting an unequal distribution of patients with active ischemia. In addition,
there may have been issues with the bedside ventilator with respect to its mech-
anisms for limiting the inspiratory time and triggering expiratory cycling. Air
leaks, if present, may not trigger expiratory cycling and would result in continued
inspiratory pressures. This failure to cycle off could subject patients to prolonged
increased intrathoracic pressures and potentially worsen intravascular hemody-
namics and coronary blood ow. This group has reported preliminary results
of a comparable study using a later generation bedside ventilator with a more
sophisticated expiratory triggering system that eliminates the risk of a prolonged
inspiratory phase (39). They randomized 27 patients to NPPV or CPAP and,
similar to their initial report, noted greater improvement in oxygenation and
dyspnea with NPPV, with few patients requiring ICU admission, but without
any differences in intubation rate, mortality, or incidence of myocardial
infarction.
In summary, the evidence to support the use of NPPV in cardiogenic pul-
monary edema is not as compelling as that with CPAP. While symptoms and oxy-
genation are more rapidly corrected with NPPV and possibly to a greater extent
than CPAP, this has not translated into improvement in the important clinical out-
comes of intubation and mortality. There are several issues that have contributed
to this, including small sample size, unequal randomization, and technical issues
(operator and mechanical). At this time, it does not appear that there is much
advantage in using NPPV over CPAP in patients with cardiogenic edema,
except possibly more rapid symptom relief. There may be a role for NPPV in
those with hypercapnia who do not improve with CPAP of 10 cm H2O given
the benet noted in sub-group analysis.
C. NIMV in the Post-Extubation Period

Many of the initial series of patients treated with NIMV included patients in the
post-operative or post-extubation period (40,41). NIMV was used either to facili-
tate discontinuation of mechanical ventilation or as a means to prevent re-
intubation in those developing post-extubation respiratory distress. This is a
particularly vulnerable period, and post-extubation respiratory insufciency
and failure may occur in 20% or more of the patients (42). A number of mech-
anisms may contribute singly or in unison to difculty with discontinuation of
42 Soo Hoo
mechanical ventilation. The removal of positive airway pressure after mechanical

ventilation may produce an increase in left ventricular preload and afterload,
which may worsen cardiac failure and lead to respiratory failure as previously
described. Respiratory muscle weakness or fatigue and increased work of breath-
ing are common. Other mechanisms leading to failure included increased CO2
production, upper airway obstruction, and inadequate ventilatory drive. These
are manifested as either hypercapnic or hypoxemic respiratory failure.
The possibility of post-extubation upper airway obstruction is particularly
intriguing. Upper airway edema and trauma are acknowledged complications of
prolonged endotracheal intubation, although the incidence and timeframe for res-
olution are not well characterized. In patients intubated an average of seven days,
Nathan and colleagues (43,44) noted increased work of breathing following extu-
bation. A sawtooth pattern on ow volume loops suggestive of upper airway
abnormalities was also noted. This was attributed to redundant upper airway
tissue and may represent upper airway or laryngeal edema in these patients. It
follows that NIMV would be effective in correcting these abnormalities as
well as the aforementioned disturbances. NPPV is effective in reducing the
work of breathing in the post-extubation period. Vitacca et al. (45) evaluated
12 patients with respiratory failure, comparing the same level of ventilatory
support (PSV 18.2 + 2.4 cm H2O plus PEEP 3.6 + 1.3 cm H2O) delivered
through an endotracheal tube prior to extubation and immediately post-
extubation using a face mask. There was equivalent reduction in the indices of
respiratory work with both endotracheally administered ventilatory support as
well as NPPV when compared with T-tube trials or spontaneous breathing.
During NPPV treatment, subjects had less dyspnea and were more comfortable
with this modality.
Randomized Trials
Investigations have evaluated NPPV in two basic strategies in the post-extubation
period. One can be designated an early extubation approach in which patients
who do not meet standard criteria for extubation are extubated early in their
course and managed with NPPV. Nava et al. (46) rst reported on this early extu-
bation approach in 50 COPD patients with hypercapnic respiratory failure after
48 hours of intubation. Those unable to tolerate a T-tube weaning trial were ran-
domized to either continued ventilator support or extubation with face mask PSV.
The NPPV patients received an average pressure support of 17.6 + 2.1 cm H2O,
comparable to those continuing with mechanical ventilation. Those treated with
NPPV had greater success at discontinuation of mechanical ventilation (88% vs.
68%) as shorter duration of mechanical ventilation, few ICU days, and reduced
mortality. These patients had no episodes of nosocomial pneumonia compared
with seven in the invasive ventilation group.
In a similar trial, Girault et al. (47) randomized 33 mostly COPD patients to
this early extubation approach if they had met established weaning criteria but
had failed a two-hour T-tube trial. Those randomized to NPPV had been
intubated an average of 4.6 + 1.9 days. By study design, the duration of invasive
ventilation was shorter for the NPPV group, but the NPPV patients actually
required a longer period of ventilatory assistance related to weaning
(11.5 + 5.2 days vs. 3.5 + 1.4 days) than invasively ventilated patients albeit
for a shorter time period each day. No differences were noted in the numbers
of patients eventually able to discontinue mechanical ventilation, duration of
ICU and hospital stay, complications, and mortality. Ferrer et al. (48) also
used this strategy but chose patients who had failed T-tube trials for three con-
secutive days prior to randomization to either BiPAP or continued mechanical
ventilation. In their 43 mostly COPD patients, those randomized to NPPV had
greater success with eventual discontinuation of mechanical ventilation, fewer
episodes of nosocomial pneumonia, fewer ICU days, and improved mortality.
As expected from their study design, the NPPV patients also had a shorter dur-
ation of invasive ventilation. They noted that hypercapnia (PaCO2 . 45 mmHg)
during their T-tube trials was an indicator of prolonged ICU stay and decreased
survival. On the other hand, using a similar approach, Hill et al. (49) reported a
trial involving 21 patients, which constituted less than 10% of their ICU patients
with respiratory failure. While the NPPV-treated patients had fewer days of endo-
tracheal intubation, the re-intubation rate was also higher in this group. All of the
patients in these trials were intubated on average a week or less prior to extuba-
tion and treatment with NPPV.
In conclusion, while this approach is promising, the experience is somewhat
mixed. Nosocomial pneumonia may be averted by this approach, but patients may
still require continued ICU care because of a tenuous respiratory status. The ideal
candidate for this approach remains to be dened but might be an otherwise
improving COPD patient who just requires a small amount of assistance to
regain the balance between ventilatory load and strength. This early extubation
approach seems to be effective up to a week following intubation.
The other approach in this arena involves the use of NPPV to treat post-
extubation respiratory distress and thereby avoid re-intubation and its adverse
effects. The available data casts some doubt on the efcacy of NPPV under
these circumstances. Jiang et al. (50) randomized 93 patients after extubation
to either face mask BiPAP or oxygen therapy. It should be noted that extubation
in 37 of these patients was unplanned (self-extubation). There was a 22% re-intu-
bation rate but no signicant difference between the two treatment groups,
although the re-intubation rate was higher in those receiving BiPAP. Keenan
et al. (51) randomized 81 patients who developed post-extubation respiratory dis-
tress within 48 hours of extubation to either NPPV or standard therapy. Patients
developed respiratory distress on average about 10 + 10 hours following extuba-
tion. No differences were noted in the re-intubation rate, duration of mechanical
ventilation, episodes of pneumonia, duration of ICU or hospital stay, or mortality.
The study patients represented only about 10% of screened patients who met
inclusion criteria. In addition, they identied 24 patients who met inclusion
44 Soo Hoo
criteria but were treated with NPPV outside the study during the study period.
Their re-intubation rate was lower (44%) as opposed to 70% in the study patients.
It was unclear if selection bias may have skewed the outcome of this trial. Sub-
sequently, a multi-center trial involving 221 patients, also did not identify any
difference in the re-intubation rate (48%) with NPPV and noted increased mor-
tality in those assigned to NPPV (25% vs 14%) (52). These ndings temper
the use of NPPV in this group of patients, but sub-group analysis suggests poss-
ible benet in those with COPD.
In summary, the support for NPPV after discontinuation of mechanical
ventilation is inconclusive. It appears to work best in an early extubation strategy
in which patients who do not meet conventional weaning criteria are extubated
and treated with NPPV within a week of intubation. Patients with COPD may
be the sub-group most likely to benet from this approach. The evidence to
support the use of NPPV in patients who meet weaning criteria and then
develop respiratory distress after extubation is disappointing.
D. Pneumonia
Although this has been the primary focus of only one randomized trial, patients
with community-acquired pneumonia constitute a large proportion of study
patients in several other investigations. As in other conditions, patients with pneu-
monia may benet from the improvement in oxygenation and relief of dyspnea
with NIMV, but there may be concerns with hemodynamic instability, secretion
clearance, and refractory respiratory failure. While the gas exchange defects of
pneumonia can be corrected, there may not be the rapid recovery that is character-
istic of the other conditions for which NIMV has demonstrated efcacy.
Confalonieri et al. (53) randomized fty-six consecutive patients with
severe community-acquired pneumonia to either NPPV with PSV (14.8 +
4.7 cmH2O) and PEEP (4.9 + 1.7 cmH2O) plus standard therapy or standard
therapy alone. There was a reduction in the intubation rate and duration of
ICU stay, but the improvement was noted only in those patients with concomitant
COPD. There was no improvement in any outcome measures in patients without
obstructive lung disease. It was unclear if the benet of NPPV may actually have
been due to the effect of NPPV on COPD, as opposed to pneumonia. The COPD
patients had a respiratory acidosis and actually had a higher PaO2/FIO2 than their
non-COPD counterparts.
As noted earlier, CPAP has potential to be of benet in patients with hypoxe-
mic respiratory failure, most notably cardiogenic pulmonary edema. Evaluation of
its efcacy in a larger group of heterogeneous patients with acute hypoxemic res-
piratory failure has not conrmed this benet. Delclaux et al. (54) randomized 123
patients to face mask CPAP (up to 10 cm H2O) in addition to standard therapy.
Sixty-seven of these patients had pneumonia, and as a group, there was improve-
ment in PaO2/FIO2 and symptoms at one hour, but ultimately no differences in
these outcome measures, intubation rate, duration of ICU stay, or mortality.
There were four cardiac arrests in the CPAP group, possibly related to delays in
intubation. There was no separate analysis presented about pneumonia patients.
CPAP may not have been the optimal mode of NIMV. In another prospective
trial of patients with acute hypoxemic respiratory failure, Ferrer et al. (55) random-
ized 105 patients to NPPV with either BiPAP (IPAP 16 + 3 cm H2O;
EPAP 7 + 2 cm H2O) plus standard therapy or standard therapy in a hetero-
geneous group of patients with hypoxemic respiratory failure. Thirty-four patients
had pneumonia without COPD, and they were able to demonstrate a difference in
the intubation rate (26% vs. 73%) with NPPV. NPPV patients also had greater
improvement in hypoxemia and tachypnea, as well as a decrease in the incidence
of septic shock, presumably due to decreased nosocomial pneumonia, decreased
ICU mortality, and 90-day mortality. Although data were presented for the
whole cohort, the benet of NPPV seemed to be restricted to the pneumonia
patients. Of note, there was no difference in the intubation rate for patients with
cardiogenic edema, mirroring the experience reported by Nava et al. Other ran-
domized investigations of NPPV that include pneumonia patients have not had
more than a handful of patients with pneumonia, making it difcult to reach any
meaningful conclusions (56 58).
These ndings urge caution in the use of NPPV in patients with pneumonia.
It is important to note that in analyses of large groups of patients treated with
NPPV, pneumonia has been identied as a predictor of failure with NPPV
(59,60). This concern may lessen as improvements in ventilatory devices
occur and clinicians gain experience and understanding of the limitations of
NPPV. As in other conditions, careful patient selection and close monitoring is
warranted.
The above constitutes the bulk of the favorable experience with NIMV.
These conditions and the primary benets of NIMV are summarized in
Table 2. NIMV has also been used in other conditions, but the evidence to rec-
ommend its routine use is limited. Those conditions that have undergone
single randomized controlled trials are discussed in the following section and
also noted in Table 3. Endorsement of the use of NIMV may be limited to
sub-groups or limited by relatively modest experience, thereby tempering its
widespread use in these conditions.
IV. Other Conditions

A. Immunocompromised Patients
Immunocompromised patients are difcult to manage, especially those with
hematologic malignancies or post-transplant, as they can have a very high sever-
ity of illness and have an extraordinarily poor prognosis if they develop respirat-
ory failure requiring intubation. Bone marrow transplant patients with respiratory
failure requiring intubation are a group with a particularly dismal prognosis (61).
Nevertheless, there is the possibility that a select group of patients may benet
from NPPV and avoid intubation.
46 Soo Hoo
Table 2 Non-Invasive Ventilation in Clinical Conditions with Randomized Controlled

Trials: Summary of Effects
Outcome variable
Gas Intubation Length

Condition exchange rate of stay Mortality
Chronic obstructive
pulmonary disease
Cardiogenic pulmonary edema
CPAP 2 +
PSV + 2 2
BiPAP + 2 2 2
CPAP vs. BiPAP (BiPAP) 2 2 2
After discontinuation of
mechanical ventilation
Early extubation
Post-extubation + 2 2 2
Pneumonia
Notes: , Uniform benet; , Benet in selected groups; +, Possible benet; 2, No difference.

Abbreviations: BiPAP, Bi-level positive airway pressure; CPAP, continuous positive airway
pressure; PSV, pressure support ventilation.
In a group of mostly liver and renal transplant patients, Antonelli et al. (62)
randomized 51 with acute hypoxemic respiratory failure to either NPPV with
pressure support and PEEP plus supportive therapy or supportive therapy
alone. The major causes of respiratory failure were pneumonia, cardiogenic
Table 3 Non-Invasive Ventilation: Experience in Other Conditions
Demonstrated benet (Single randomized controlled trial)

Asthma
Immunocompromised patients
Solid organ transplants (liver, renal, lung)
Febrile neutropenic patients
Post-operative lung resection
Rib fractures (trauma with non-penetrating chest injuries/ail chest)
Demonstrated efcacy (Case series)
Acute respiratory distress syndrome
Cystic brosis
Do not intubate patients
Kyphoscoliosis
Muscular dystrophy
Obstructive sleep apnea (decompensated)
Pneumocystis carinii pneumonia (mild)
Post-polio syndrome
pulmonary edema, acute respiratory distress syndrome (ARDS), and mucus plug-
ging. They were able to demonstrate greater improvement in oxygenation, intu-
bation rate, incidence of fatal complications, most notably in septic shock,
duration of ICU stay, and mortality. However, the benet was most evident in
those with cardiogenic pulmonary edema, as the benet was not substantially
different from standard therapy in the other sub-groups. In another study of
mostly neutropenic patients with hematologic malignancies or bone marrow
transplants, Hilbert et al. randomized 52 of these patients with febrile hypoxemic
respiratory failure and pulmonary inltrates to NPPV with pressure support
(15 + 2 cm H2O) and PEEP (6 + 1 cm H2O). About half of the patients had a
microbiologic diagnosis of pneumonia. They were able to demonstrate improved
oxygenation, with lower intubation rates, complications, ICU stay, and hospital
mortality. The benet was most pronounced in those with an identied cause
of their pneumonia. It should be noted that NPPV was only administered, on
average, eight hours a day for an average of four days. This indirect measure
of severity would suggest that these patients had a relatively modest case of pneu-
monia as highlighted in the accompanying editorial (63).
In summary, NPPV may benet a very select group of immunocompro-
mised patients. Experience from these single-center studies would suggest that
those with cardiogenic pulmonary edema and an identied cause of pneumonia
may be the best candidates for successful application. Caution should be used
in its application in any other sub-group of immunocompromised patients
because a delay in more appropriate intubation may increase their risk for
adverse events and death.
B. Asthma
Although the treatment of decompensated asthma closely mirrors that of COPD,
there is a striking paucity of reported experience with NIMV in this condition.
There are sufcient differences in the pathophysiology of these two conditions
that may not allow the experience in COPD patients to be translated to asthma
patients. Dynamic hyperination may play more of a role in COPD patients
than in asthma patients, who while subject to dynamic hyperination, are also
subject to increased airway resistance and sudden asphyxia not typically seen
in the COPD patients. Most of the reported experience in asthma is found in
case series, but there has been a randomized evaluation in asthma patients in
an emergency room setting (64). Soroksky et al. randomized 30 patients with
asthma to nasal BiPAP (maximum settings of IPAP 15 cm H2O; EPAP
5 cm H2O) plus standard therapy or standard therapy plus sham BiPAP with
settings of 1 cm H2O. They noted a more rapid and greater improvement in
lung function as well as a lower hospitalization rate. This experience is promis-
ing, but it should serve as the impetus for larger studies. As with other conditions,
this is a reasonable option provided there is careful patient selection and close
monitoring of response to therapy.
48 Soo Hoo
C. Post-Operative Lung Cancer Patients

Respiratory failure following lung resection for lung cancer can carry devastating
consequences, with mortality approaching 80% in the most severe cases (65). The
major causes of decompensation include cardiogenic and non-cardiogenic pul-
monary edema, mucus plugging, and pneumonia. Endotracheal intubation and
mechanical ventilation also carry a risk of exacerbating bronchial stump disrup-
tion, with attendant complications including broncho-pleural stula and infec-
tion. NIMV may ameliorate some of the risk factors as patients would not be
subject to the high peak airway pressures usually associated with mechanical
ventilation. In a focused evaluation of patients following lung resection,
Auriant et al. (66) randomized 48 patients who developed hypoxemic respiratory
failure after extubation following lung cancer resection (lobes or pneumo-
nectomy). They had been extubated in the operating room and developed
respiratory distress within two days of their ICU admission. Decompensation
was attributed to either pulmonary edema, atelectasis, or pneumonia. Patients
were treated with BiPAP at relatively modest levels (IPAP 9 + 2 cm H2O;
EPAP 4 + 0.1 cm H2O) and with a marked reduction in the intubation rate
(21% vs. 50%) as well as a reduction in overall mortality. It should be noted
that this experience represents a very select group of patients, comprising less
than 10% of their patients who required ICU admission and 0.06% of all patients
who underwent lung resection at their institution during the study period. While
this is a promising option, it requires further study and due caution must be exer-
cised when used in these patients.
D. Rib Fractures/Chest Wall Trauma

Rib fractures due to non-penetrating injuries to the chest (trauma) may predispose
to adverse complications due to pulmonary contusion, ineffective cough (pain),
atelectasis, and pneumonia. Although patients with multiple rib fractures and res-
piratory compromise have often been managed with intubation and mechanical
ventilation, NIMV may be able to provide this same pneumatic splint. Bolliger
et al. (67) randomized 70 patients with multiple rib fractures to face mask
CPAP (5 + 1 cm H2O) and regional analgesia or intubation and mechanical ven-
tilation. Only one of the CPAP patients subsequently required intubation. The
CPAP-treated patients recovered more quickly and had their hospitalization
decreased by almost a week, with fewer episodes of pneumonia and no difference
in mortality. While this appears promising, it should be noted that ventilator man-
agement has changed since this report. Patients undergoing intubation in the
study may not be intubated in current practice. However, this is an option that
can be considered in patients with chest wall trauma in whom physicians may
wish to avoid intubation.
E. Acute Respiratory Distress Syndrome

This has not been the focus of any prospective, randomized study of NIMV,
although patients with severe hypoxemia due to ARDS or non-cardiogenic
pulmonary edema have been included in many of the prospective studies. Case
series would suggest that patients with ARDS can be successfully treated with
NIMV, but the success seemed to be in those with less severe disease (68).
In their evaluation of hypoxemic respiratory failure, Ferrer et al. (55) enrolled
15 patients with ARDS and found no benet in any outcome variable in this
sub-group. In multi-variate analysis, ARDS was identied as a risk factor for
intubation. This mirrors the nding by Antonelli et al. (60), who identied
ARDS as a risk factor for NPPV failure. There is insufcient data to make any
recommendations on the use of this modality, but the time course of this
illness and many associated complications make it unlikely that it can be ef-
ciently managed with NIMV, except in very mild cases.
F. Do Not Intubate Status

Patients with advanced or end-stage disease, whether respiratory or otherwise,
constitute a sub-group of patients who may also benet from NIMV. These
patients may have previously designated a treatment preference of do not
resuscitate or do not intubate or are otherwise considered poor candidates
for invasive ventilatory support. These are frequently advanced stage COPD
patients. They often present with profound hypercapnic respiratory failure and
would expire without further supportive care. Intubation and mechanical venti-
lation is tempered by the risk of ventilator dependence, which is why a trial of
NIMV is a consideration. While a prospective, randomized trial is not feasible,
several case series suggest successful treatment is possible in these patients.
Early investigations in two series (30 and 11 patients, respectively), suggested
that NIMV can successfully treat 60% and 64% of these patients, although the
numbers discharged home were 53% and 45%, respectively (69,70). In another
series of 37 COPD patients, the median survival after an episode of hypercapnic
respiratory failure treated with NIMV was 179 days, and one year survival was
30% (71). Despite some modest benet, patients or family of patients treated
with NIMV in this situation must be counseled carefully to understand that this
still constitutes life support. Even if this provides an opportunity for recovery,
there remains the distinct possibility that application only prolongs the inevitable.
Some patients will experience dyspnea relief, but in others the mask may be
uncomfortable and add to their distress. While this approach does afford some
patients some extra time to resolve their affairs, this therapy also consumes valu-
able health-care resources (staff, monitored beds, supplies, etc.) in what may be
an ultimately futile situation (72,73). Drawing from the general experience with
NIMV, the do not intubate COPD patient may comprise the group best suited
for these trials. Trials in other patients would necessarily require a case by case
evaluation.
G. Other Conditions
There are many other respiratory conditions for which NIMV may be benecial
during episodes of respiratory failure, and they are listed in Table 3. However,
50 Soo Hoo
many of these are reported in case series format and not randomized, controlled
studies. These conditions include cystic brosis, kyphoscoliosis, neuromuscular
disease, decompensated obstructive sleep apnea, and mild Pneumocystis carinii
pneumonia. Except for Pneumocystis carinii pneumonia, most represent acute
decompensation of chronic respiratory failure. With cystic brosis, extensive
experience has accumulated and been reported by Madden et al. (74).
They have treated over 100 patients, and in 90 patients awaiting or undergoing
evaluation for lung transplantation, 38 (42%) have been stabilized with
NIPPV, with 28 having received a lung transplant and the others still on
the list. Some patients were maintained for over a year and a half with
NIMV. Even in a sub-group not composed of lung transplant candidates,
some were able to survive using NIMV for over a year. This experience indi-
cates that NIPPV can help provide a bridge of support until transplantation
and thereby avoid the complications associated with intubation and mechanical
ventilation.
V. Clinical Aspects of NIMV
Despite over a decade of experience with NIMV, its use remains limited and
often conned to medical centers with expertise or a special interest in its appli-
cation. Recent surveys on its actual use are telling. In a questionnaire survey,
Doherty and Greenstone (75) found that NIMV was available in 48% of the hos-
pitals surveyed in the United Kingdom. Even then, about 70% of hospitals treated
,20 patients per year. Less than 10% treated .60 patients per year. In another
month long, multi-country survey on ventilator use, over 4000 patients were ven-
tilated for .12 hours (76). Of that group, only 202 (4.7%) were treated with
NIMV, and the use was even less (1%) in a point prevalence study (77). In
another European survey of mostly French institutions, some with familiarity
with NIMV, only 16% of their ICU patients were treated with NIMV (78).
They found a wide range in the use of NIMV in the 42 ICUs surveyed, with
eight ICUs never using NIMV and one ICU using NIMV in 67% of their ICU
patients requiring ventilatory support. Hypercapnia (COPD), post-operative res-
piratory failure, pneumonia, and cardiogenic pulmonary edema were the frequent
conditions treated with NIMV. Therefore, despite the success reported, relatively
few patients with respiratory failure are placed on this treatment. There are
areas of its application that require close attention. Patient selection, mask
leaks, patient synchrony, ventilator efciency, nursing, and respiratory therapy
time are all potential limiting factors in successful application of NIMV.
These were potential stumbling blocks for initial investigators and are issues
that continue to limit the use of NIMV. A better understanding of these factors
not only helps dene the optimal treatment group, but also helps predict its
success.
A. Patient Selection
Patient selection is probably the most important factor to consider prior to initi-
ating NIMV. This is an added element that encompasses all of the clinical con-
ditions being treated. Once it is determined that the patient has a condition readily
treatable with NIMV (i.e., COPD exacerbation, cardiogenic pulmonary edema,
pneumonia, etc.), then a determination needs to be made if other aspects of
their presentation may preclude treatment with NIMV. This is cruial as those des-
tined to fail or only achieve a partial response with NIMV may be better served
with immediate intubation and mechanical ventilation. This spares what may
be ineffective therapy, avoids further progression of respiratory failure, and
limits associated complications from incompletely treated respiratory failure or
intubation under duress. Factors that inuence the success of NIMV can be
grouped into areas that deal with severity of illness, co-morbidities, technical
issues, and patient cooperation.
One of the unanswered questions involving patient selection (and in turn
predicting the success of NIMV) has been the limits imposed by the patients
severity of illness. This applies to both ends of the spectrum. Patients may
have a relatively mild severity of illness, destined to improve with medical
therapy, and not require NIMV. NIMV would be unnecessary and excessive.
There is general agreement that the most severely ill patients are better served
by immediate intubation without a trial of NIMV. Dening the limits of appli-
cation has been elusive. The severity of acidosis correlates closely with the like-
lihood for intubation in those presenting with hypercapnic respiratory acidosis.
Thresholds have been identied based on prospective and retrospective analyses.
A pH of 7.25 has been associated with a .50% intubation rate. It is noteworthy
that a pH of 7.25 has been used to limit enrollment in several NIMV studies. Plant
et al. (79) analyzed their database of patients with respect to pH and PaCO2 as
predictors of failure and meeting criteria for intubation. These two variables
are obviously linked and co-dependent. The pH may be the more important vari-
able. The relative risk of failure was greater with worsening pH for the same level
of hypercapnia than hypercapnia for the same level of pH. The relative risk for
failure with a pH 7.25 and PaCO2 45 was 3.89 for those treated with NIMV
and 9.98 for those treated with standard therapy. This compares to a relative
risk of unity for standard therapy with a pH 7.35, and 0.39 with NIMV. Higher
APACHE II scores indicative of an increased severity of illness have predicted
NIMV failures in several series (59,80).
As medical staff gain more expertise with this modality, the success rate
improves as expected. Carlucci et al. (81) have analyzed their experience with
COPD patients during 208 episodes of acute respiratory failure. While the
failure rate was constant (17.2%) over an eight-year period, patients were being
treated with a greater severity of respiratory acidosis (mean pH: 7.20 + 0.08
vs. 7.25 + 0.07) and with greater successful treatment of the more severe patients
52 Soo Hoo
during the latter periods of their survey compared with the early years. The failures
during the early period of their survey (1992 1996) were very similar in
terms of severity of illness to those successfully treated during the latter years
(1997 1998). One consistent nding in both periods was that successfully
treated patients had a marked improvement with NIMV after one hour of
therapy (pH improvement: 0.06 0.09), while failures had no improvement at
all. The lower limit of effectiveness remains to be dened, but limiting
NIMV to patients with hypercapnic respiratory acidosis to a pH . 7.10 is a
reasonable common sense limitation. It should be noted that there is very little
experience in these severely acidemic patients since most have been excluded
from clinical trials.
B. Co-Morbidities
In addition to severity of illness, co-morbidities also limit effective NIMV. These
are summarized in Table 4. These limitations generally involve issues that render
optimal management difcult without intubation and mechanical ventilation.
These include cardiac instability involved with acute myocardial infarction, ven-
tricular dysrhythmias, shock, GI bleeding, coma, profound lethargy, and status
epilepticus. The main limiting factor revolves around the need to protect the
airway from aspiration or the need to control the airway during treatment. Hemo-
dynamically unstable patients are best managed with intubation, eliminating a
potentially confounding factor in already critically ill patients. Those who
require airway protection or with mechanical upper airway obstruction are
Table 4 Non-Invasive Ventilation: Other Aspects of

Patient Selection Co-Morbidities that Prevent Effective
Non-Invasive Ventilation
Hemodynamic instability
Acute complicated myocardial infarction
Hemodynamically signicant cardiac dysrrhythmias
Severe septic shock
GI bleeding
Need for airway protection
Coma or any condition with an unstable respiratory drive
Extensive cerebral vascular accident or hemorrhage
Status epilepticus
Potential for life-threatening airway compromise
Head and neck tumors
Any tumor with extrinsic airway compression
Angioedema or anaphylaxis
Abbreviation: GI, gastrointestinal.

better served with endotracheal intubation to limit further aspiration of

oropharyngeal or gastic contents to the extent possible with endotracheal intuba-
tion. Patients with upper airway obstruction because of tumors, extrinsic masses,
anaphylaxis, or angioedema should be intubated to maintain the patency of the
airway. Comatose, severely obtunded patients or those with unstable respiratory
drive are not candidates for NIMV. Patients need to be able to cooperate with
mask tting and be able to coordinate with the ventilator. They can be lethargic
but arousable and need to be able to cooperate and provide feedback.
Those with severe pneumonia and copious secretions are problematic
because mask ventilation makes it difcult to control and clear secretions;
however, there has been successful management in selected patients. Craniofacial
defects may limit the ability to achieve an effective mask seal. Patients with bony
deformities following resection of head and neck tumors fall into this category.
Edentulous patients may have problems with nasal ventilation, but this can be
solved by using an oronasal mask. Beards may prevent an adequate mask seal,
but they can be shaved. Nasogastric tubes were once considered essential to
prevent gastric insufation. However, the pressures utilized with NIMV rarely
exceed the lower esophageal sphincter pressure (25 cmH2O) (82). They are no
longer recommended as their presence makes it difcult to achieve an adequate
mask seal and predisposes to air leaks.
C. Technical Issues: Masks

In the most basic conguration, NIMV substitutes the endotracheal tube with a
mask. However, this transforms the ventilatory system from a closed system to
one with potential for substantial leaks. The leaks usually occur along the
patient mask interface or from the mouth if nasal masks are used. Substantial
mask leaks lead to patient discomfort and ineffective ventilation. In addition to
improperly tted masks, the masks may become dislodged during therapy or
leaks may occur around nasogastric tubes or any other device that prevents an
adequate seal between the skin and the mask. Fitting and maintaining a proper
mask may be as important as patient selection in successful NIMV. Failure
rates of NIMV can be as high as 40%, and many are due to technical issues
related to the mask leaks and mask intolerance (60). Over the past decade,
there has been signicant progress in mask development. Masks are clear to
allow visualization of face and facilitate diagnosis in patients with emesis or
excess secretions. They are softer than the past, usually with silicone or an air-
lled cushion in the contact points with the face. There may be other cushions
at the greatest pressure points, and straps have become less cumbersome and
more secure. Other design improvements prevent circuit rebreathing in the
event of machine failure and allow rapid mask removal in the event of emesis.
Masks can be custom tted, but may be problematic in those with acute respira-
tory distress, and most of them are available in a variety of sizes.
54 Soo Hoo
Generally, the most effective mask is the smallest mask that provides an
adequate t. Patients should be upright, with the head of the bed elevated 458
or more at the time of mask placement. Patients may hold the mask up to their
face, or the mask and attached head straps are loosely placed over the patients
head prior to strapping it in place. Allowing patients to hold the mask also
allows them to become accustomed to the mask. It is important that the head
straps are in place prior to mask placement as this facilitates placement.
For nasal masks, the mask should allow adequate clearance for the nose, but
not cover the vermillion of the upper lip. The soft palate and teeth allow proper
closure of the oronasal cavity during positive pressure and help minimize leaks.
Edentulous patients may have difculty with maintaining proper mask position
and are often better served with a face mask. A face mask should t comfortably
on the mandible, just below the lower vermillion. A mask that encroaches on the
chin is subject to large leaks and ineffective ventilation. The straps should be
tight but allow easy passage of one to two ngers between the face and straps.
Very tight straps are not only uncomfortable, but they increase the likelihood of
developing pressure ulcers over the nose or face. Chin straps are available if
mouth leaks are unavoidable. A small amount of leak may be unavoidable despite
multiple manipulations. Leaks around the nose into the eyes are the most distressing
for patients. Nasal plugs or nasal pillows are not suitable for acutely ill patients. A
mouthpiece is an alternative, although that would require more active cooperation
by the patient than a mask that is anchored to the patient by head straps. Most
masks have the option of bleeding oxygen directly into the mask via oxygen
tubing, but this runs the risk of creating excess pressure in this enclosed space. If
oxygen is necessary, it is better delivered via an adaptor attached to ventilator
tubing. In a bench study assessing the inuence of the site of oxygen delivery
(mask or circuit), with various combinations as to the leak port site (mask,
plateau exhalation value, or circuit) and increasing levels of IPAP and EPAP
(1025 and 510 cm H2O, respectively), the highest concentrations of oxygen
were delivered with oxygen delivered at the mask and the leak in the circuit with
the lowest levels of pressure (83). This might be an issue with some of the older,
smaller bi-level ventilatory assist devices. Newer ventilators are able to blend
oxygen and deliver through the ventilator tubing to allow precise control of FIO2.
Nasal and oronasal (including full face) masks are the primary interfaces
used in NIMV. A summary of their advantages and disadvantages is provided
in Table 5. For most patients, the nasal mask is more comfortable, less obtrusive,
has less dead space, and allows easier communication. The nasal mask does
require a more coordination by the patient. Patients must be able to maintain a
mouth seal and are subject to more mouth leaks. Some patients may be unable
to prevent mouth leaks (edentulous, mouth breathing pattern, pursed lips breath-
ing, and obtundation). The oronasal mask allows for mouth breathing, creates
more difculty with speech and expectoration, and may cause claustrophobia
in some patients. Most experience has been with an oronasal mask that covers
the nose and the mouth. Face masks that cover the whole face including the
Table 5 Non-Invasive Ventilation: Comparison of Masks
Nasal Face (oronasal)
Better suited for

More cooperative patients Less cooperative patients
Lower severity of illness Higher severity of illness
Claustrophobic patients Mouth breathing
Pursed lips breathing pattern
Edentulous patient
Physiologic benets
Better tolerated, More effective ventilation,
? better compliance higher tidal volumes
More efcient CO2 removal
Possibly improved outcome measures
(? Lower intubation rate)
Other notes
Less frequently used (30%) Most frequently used (70%)
eyes are available as an alternative to nasal and oronasal masks, although with
limited experience (84). Helmet masks have also been developed, but likewise
with limited experience (85,86).
It follows that nasal masks are better suited for the less dyspneic patient and
the oronasal face mask better for the more dyspneic patient who is more likely to
be a mouth breather and subject to mouth leaks. Because this comprises the
majority of candidates for NIMV, face mask NIMV is the modality of choice
at our institution. Face mask ventilation is also the predominant interface for
NIMV in reported studies, used in about 70% of treated patients (87). Nasal ven-
tilation would be reserved for those cooperative patients who are unable to toler-
ate the face mask. It has been observed that issues with claustrophobia or mask
discomfort are often outweighed by the relief obtained with unloading of the res-
piratory system.
As might be expected, differences in randomized comparisons of the masks
in acute respiratory failure primarily reect difference related to leaks and efcacy.
In a prospective, randomized trial, Kwok et al. (88) compared nasal and oronasal
masks in 70 patients with acute respiratory failure (primarily pulmonary edema
and COPD). They noted comparable effects on physiologic indices (dyspnea,
comfort, oxygenation, and CO2 reduction) with both masks, but there was
greater mask intolerance with the nasal mask (34% vs. 11%). They also noted
more deaths (11% vs. 6%) and less success in the nasal mask group (49% vs.
66%), but these differences were not statistically different, with the same intuba-
tion rate (23%) in both groups. This mirrors other evaluations in acutely ill
patients, although there have been trends toward more efcient removal of CO2.
In a study of stable patients, there was greater CO2 reduction and higher minute
ventilation with face masks, but better tolerance with nasal masks (89,90).
56 Soo Hoo
In summary, once properly tted, effective ventilation can be accomplished

using either a nasal or an oronasal mask. Most patients who are candidates for
NIMV are best treated rst with face mask (oronasal) ventilation because their
dyspnea often mandates mouth breathing. It follows that less acutely ill patients
can be successfully ventilated with a nasal mask.
D. Ventilators
There has been great diversity in the types of ventilators used to provide venti-
latory support in NIMV. There is experience with both volume- and pressure-
cycled ventilators, large bedside and portable devices, and a variety of ventilatory
modes, including assist control, CPAP, pressure support, proportional assist ven-
tilation (PAV), or some combination of all of these modalities. In the appropriate
setting, all of these can provide effective ventilatory support with relief of
dyspnea, correction of gas exchange abnormalities, and successful prevention
of endotracheal intubation. The success rates are probably comparable, provided
that patients receive an adequate level of support. In this sense, one modality
cannot be endorsed above another, but in reality, pressure support with PEEP
in one survey was used in almost 75% of patients treated with NIMV (78). In
the survey, volume ventilation was used in only 15% of patients. Similarly, the
vast majority of the experience with NIMV has been accumulated with pressure
ventilation with volume ventilation used in only one major randomized trial,
comprising less than 10% of patients analyzed in a recent meta-analysis.
Volume ventilators may pose more problems during NIMV, primarily
because of mask or mouth leaks and potentially wide uctuations in peak
airway pressures. The leaks obviously impede delivery of adequate tidal
volume and can trigger more alarms, which may require more attention from
medical staff. This may occur even with drastic changes in alarm limits. Compen-
sation for the leaks often requires tightening of mask straps, which can contribute
to skin breakdown and pressure sores. High pressures may also lead to gastric dis-
tension. On the other hand, volume ventilators may be better suited for patients
with rapidly changing respiratory system compliance or drive who would require
some minimum minute ventilation. However, despite these potential limitations,
no signicant differences have been demonstrated between the two modes,
although there appears to be better compliance with pressure ventilation (91,92).
Pressure-cycled ventilators, whether provided by larger bedside critical
care ventilator or smaller, more mobile bi-level ventilatory assist devices,
seem to be better adapted for NIMV. The pressure ventilators are more tolerant
of system leaks, and some of the bi-level models have leak compensation
features, which result in fewer alarms requiring medical staff attention. The
bi-level devices are designed specically with NIMV in mind, whereas NIMV
is one of the many options available in the bedside critical care ventilator.
There are signicant cost differences between the two types of ventilators, and
versatility among patients may also inuence the use of one over the other.
There have been improvements in other areas. Earlier models of bi-level

pressure ventilators were limited by the maximal pressures that could be delivered,
possible CO2 rebreathing through single-hose tubing, and imprecise FIO2 delivery
(93,94). Renements have allowed higher levels of pressure support delivery (up
to 40 cm H2O of inspiratory pressure) and limited potential for CO2 rebreathing
(use of PEEP and modications of the expiratory valve), and newer models are
equipped with blenders that allow precise FIO2. The devices have become more
responsive to patient efforts with adjustable triggering sensitivities, involving both
the inspiratory and expiratory cycles. Flow triggered breaths provide for better
patientventilator synchroncy and lower work of breathing (95). The termination
of the inspiratory cycle or initiation of expiratory cycling may potentially adversely
affect patients. Common criteria include termination at a threshold representing a
2540% decrease from peak inspiratory ow rates or at a set time. There is the
potential for continued cycling due to mask leaks, and in a mathematical and lung
model, these lead to marked variations in the duration of the inspiratory cycle as
well as the development of intrinsic PEEP (96). This may have adverse consequence
on patientventilator synchrony and possibly even some hemodynamic conse-
quences. Modications in newer generation devices have eliminated this potential
problem. Perhaps the key to these ventilators is that their breaths are patient initiated.
This certainly facilitates patientventilator synchrony and may improve patient
comfort. There has been further blurring of the differences between these critical
care ventilators and bi-level devices. There are models that now incorporate
both invasive and non-invasive modes in their conguration. Table 6 provides
Table 6 Non-Invasive Ventilation: Comparison of Ventilators
Full function critical care ventilators Bedside bi-level assist devices
Multiple ventilatory options Ventilatory options (maximum pressures)

Assist control/volume Inspiratory positive airway pressure
Pressure support (40 cm H2O)
Pressure control Expiratory positive airway pressure
Synchronized intermittent (20 cm H2O)
mandatory ventilation Continuous positive airway pressure
Positive end-expiratory pressure (20 cm H2O)
Inverse ratio ventilation
Higher inspiratory ow rates
Multiple alarms Limited number of alarms
Not very leak tolerant Leak tolerant and leak compensation
Adjuncts Adjuncts
Humidication, nebulization, Some models with features incorporated
oxygen, in line Other models require addition or
Patient waveforms available adaptation for use
58 Soo Hoo
a comparison of the key differences between the major types of ventilators used in
NIMV.
As the distinction between ventilators has blurred, it becomes more import-
ant to recognize that different masks are used based on the ventilator used. This
would probably never be an issue except for those unfamiliar with NIMV. The
bi-level devices use a single-hose to deliver ventilatory support with exhalation
of gases usually through a hole in the mask or circuit. This exhalation port
should never be blocked. Likewise, there is also a apper valve that allows deliv-
ery of pressure and entrainment of air. If it is not functional, the mask should be
replaced. A separate mask without any openings in the mask or circuit is used for
those critical care ventilators that can also deliver NIMV. The critical care ven-
tilators use a closed two-hose system (inspiratory and expiratory) and therefore
are not designed to be used with a conguration with a xed leak. This type of
mix-up may result in ineffective ventilation, and there is always the potential
for untoward consequences. Figures 1 and 2 demonstrate the differences in the
masks used with different ventilators and also illustrate the similarities
between the two congurations.
PAV may provide even better patient ventilator synchrony and patient
comfort. It uses an inline pneumotachometer to provide ventilatory assistance
proportional to the patients efforts. It incorporates ow and volume assistance
with each breath, and expiratory cycling occurs with cessation of inspiratory
effort. It is more responsive to changes in patient effort than the previously
Figure 1 Patient supported with face mask non-invasive ventilation using a bedside
bi-level ventilator with a single-hose tubing attached to the mask. The connection has an
exhalation port and a apper valve attached at the interface between the mask and tubing.
Figure 2 Patient supported with face mask non-invasive ventilation using a critical care
ventilator that can be used in patients who are also endotracheally intubated. The mask is
the same as the mask used for the patient in Figure 1, but the head straps and ventilator
tubing attachments are different. The attachment to the mask is of a single-piece construc-
tion, different conguration, without any exhalation ports and is meant to be used with
double-hose tubing.
discussed models. This conguration should drastically reduce or even eliminate

problems with patient ventilator synchrony. Clinical experience indicates ef-
cacy equal to pressure support NIMV, but better overall tolerance and comfort
(97,98). The ventilator is not yet commercially available in the United States.
E. Ventilators: Other Issues
Initial settings are goal directed to achieve adequate tidal volumes. A goal tidal
volume of 5 to 7 mL/kg is appropriate with additional support added as necessary
to reduce the respiratory rate to 25/min. Appropriate initial settings in most
patients would be an IPAP or pressure support equal to 10 cmH2O and EPAP
or PEEP equal to 5 cm H2O. These pressures are tolerated by most, although
some will require lower pressures. However, it is not recommended that initial
inspiratory pressures be lower than 8 cm H2O or expiratory pressures much
lower than 4 cm H2O. Settings with lower pressures are usually inadequate for
ventilatory support, subjecting the patient to more dyspnea and distress, and
lead to premature termination of therapy. Maximal inspiratory pressures should
be limited to about 20 to 25 cm H2O. Patients do not tolerate these higher press-
ures, and in this range, there is a risk for overcoming the lower esophageal
sphincter pressure with subsequent gastric distension. Changes in settings
60 Soo Hoo
should target the primary respiratory defect. Hypoxemic patients are best treated
with an increase in the PEEP (CPAP) or EPAP with a proportionate increase in
the IPAP effectively maintaining the same tidal volume. Hypercapnic patients
benet most from an increase in tidal volume resulting from increases in pressure
support or IPAP without much change in the EPAP. If using BiPAP, the differ-
ence between IPAP and EPAP represents the level of pressure support supplied
to the patient. A back-up respiratory rate can be set to achieve a minimum
minute ventilation.
NIMV is best delivered in three to four hours blocks with 30 to 60 minutes
breaks. Early in the course of treatment, some patients develop intolerable
dyspnea off NIMV and receive near continuous support. The breaks in therapy
are encouraged to minimize ischemia and necrosis that can develop at the
skin mask interface. This also allows a chance for the patient to clear secretions,
eat, and drink. These breaks can serve as short weaning trials, and the duration
of the break can be extended at any time based on the comfort of the patient. In
this sense, ventilatory support and weaning occur at the same time, and venti-
latory support is continued only as long as absolutely necessary. In some patients,
NIMV is required for most of the rst day (20 hours), but some may sufce with
only eight hours of support. Patients are often very cognizant of the benet with
NIMV, and it is not infrequently that they not only dictate the duration of venti-
latory support, but can also self-wean themselves from NIMV.
Humidication may be an issue if NIMV is administered using one of the
earlier models of the portable bi-level devices. The drying effects of unhumidi-
ed air may dry and thicken airway secretions as well as mucosal surfaces.
Drawing from the experience with CPAP masks in patients with obstructive
sleep apnea, humidication prevents this complication and improves compliance
(99). It is a reasonable addition to patients during NIMV, especially those who
require more than eight hours of daily support. This is not a signicant issue
with the latest models of NIMV ventilators because most have inline humidi-
cation systems.
Treatment of COPD and asthma patients includes bronchodilators. There
may be an advantage of delivering these nebulized bronchodilators inline
during NIMV ventilation. There is potential for more drug to be delivered, and
in one small clinical study, this seemed to translate to more rapid improvement.
Potentially more drug may be able to be delivered with an inline nebulizer.
However, it should be noted that improvement may not be due to improved
drug delivery and distribution, but rather the benets of respiratory support
with NIMV (100,101). While this requires further investigation, there are no
known major adverse effects with this approach. Along the same line, a
helium oxygen mixture used in conjunction with NIMV may also enhance the
benet of NIMV in selected patients. The lower density gas is known to be effec-
tive by reducing the resistive load of breathing in obstructed patients, thereby
decreasing dyspnea and lowering the work of breathing while improving gas
exchange. In a randomized, cross-over study of primarily COPD patients, the
helium oxygen mixture during pressure support NIMV reduced indices of work
of breathing by one-third or more with greater improvement in hypercapnia than
with NIMV (102). These are both important adjuncts in the use of NIMV in res-
piratory failure and have the potential of possibly averting intubation in some
patients. More extensive studies are required to gain a better appreciation of
the role of these adjuncts in management.
VI. Monitoring the Response to Treatment

A. Predictors of Success and Failure
Once NIMV has been initiated, there needs to be close attention to the patients
response to therapy. This is crucial so as to avoid any unnecessary delay in intu-
bation and associated untoward complications. The main goal of NIMV has been
to avoid intubation, and this has been a long-standing outcome measure by which
to gauge the success of NIMV. Other factors may also help in this assessment.
This is a crucial area and the subject of a recent review of data from multiple
studies (103).
Patient cooperation is crucial for the successful NIMV. This in turn serves
as an indirect reection of a patients severity of illness and level of conscious-
ness. Some of this has been addressed in the section Patient Selection. Patients
need to be able to coordinate their respiratory efforts with the ventilator, ensuring
patient ventilator synchrony to allow effective ventilatory support. Patients with
severe tachypnea (RR . 35) may be so dyspneic that they are unable to coordi-
nate with the ventilator. At these extremes, the mask can be claustrophobic, heigh-
tening their sense of dyspnea. Airway pressures are low, making it easy for some
patients to successfully ght the ventilator, leading to ineffective ventilation
and failure. This may be more of an issue with volume ventilators, especially if
with signicant mask leaks. Some patients are so obtunded that effective venti-
lation is impossible. Measures of level of consciousness or neurologic scores
have been demonstrated to correlate with successful NIMV (21 22, 59,104).
The response of physiologic variables, specically pH, PaCO2, respiratory
rate, and heart rate, provide additional support for the effectiveness of NIMV and
its eventual success. This response is an indirect reection of aforementioned
variables, including patient selection, patient cooperation, and patient ventilator
synchrony. The response of these variables to a short period of NIMV (30
minutes to a few hours) has been cited as an important factor in predicting the
successful application of NIMV. This can be considered a trial of therapy, and
improvement represents success on several fronts. Improvement in hypercapnia,
respiratory acidosis, and oxygenation reects adequate ventilation and therefore
patient synchronization. This has been a near uniform nding, although there are
successfully treated patients who do not demonstrate this pattern of response. The
magnitude of the decline in PaCO2 is a variable based on the initial severity of
hypercapnia. Pooled studies suggest an average decline of 8 mm Hg after one
62 Soo Hoo
to three hours of ventilatory support (1). Patients who eventually fail and require
intubation either have an increase, no change, or a minimal decline in PaCO2.
This can cause somewhat of a conundrum in management. Unless patients
require immediate intubation, a case can be made for some groups of patients
(COPD) with respiratory failure to undergo a trial of NIMV. However, it may
not be clear whether failure to improve represents a failure of NIMV or a
patient who is slow to improve. There must be a time frame that constitutes an
effective trial of NIMV. In some patients, it will be clear that NIMV is ineffective
within a few minutes to an hour of initiation. In others, there may still be uncer-
tainty after three to six hours or longer. One would like to avoid extended trials in
patients destined to fail because problems may be encountered at the time of
intubation, with associated morbidity and mortality. In the two largest studies,
(17, 18) failure of NIMV occurred in the vast majority within 12 and 24 hours
of initiation, in 82% and 61% of patients, respectively.
Moretti et al. (105) identied predictors of late failure (dened as
.48 hours after initiation of NIMV). In their multi-center study of 186 patients,
31 were designated as late failures after an initial response to NIMV. These
patients experienced a relapse on average 8.2 + 2.8 days after initial therapy.
Lower initial functional status scores, hospital complications (pneumonia,
shock, and coma), and lower initial pH (7.22 + 0.08) were predictive factors.
These late failures also had an increased mortality rate (68% vs. 15%). In
summary, a three- to four-hour trial is a reasonable time period to assess the
response to NIMV. There should be some improvement in physiologic
parameters within this time frame, although resolution may require a much
longer period. If patients fail to demonstrate any improvement within this time
frame, intubation may be likely within the initial 12 to 24 hours of presentation.
The greatest success with NIMV has been in COPD and cardiogenic pul-
monary edema. While patients with other conditions can also be treated with
NIMV, they may not respond as favorably, and there may be a lower threshold
for intubation in these conditions. Higher failure rates or slower responses to
therapy (compared with COPD or cardiogenic pulmonary edema) have been
noted in patients with ARDS, pneumonia, and underlying restrictive lung
disease (60,106,107). This information provides additional perspective on the
use of NIMV in these conditions and may inuence not only the decision to
use NIMV, but also the duration of NIMV prior to proceeding to endotracheal
intubation. Table 7 provides a summary of the most useful measures of the
success or failure of NIMV. Of course, each patient must be evaluated individu-
ally, but these parameters should provide a framework for guidance.
The next great hurdle involves the decision to terminate NIMV in favor of
intubation. Reported criteria for intubation have included a pH , 7.20, pH 7.20 to
7.25 on two occasions one hour apart, hypercapnic coma (Glasgow coma scale ,8
and PaCO2 . 60 mm Hg, PaO2 , 45 mm Hg), or cardiopulmonary arrest (18).
Others have used a combination of major or minor clinical criteria (17). Major
criteria have included respiratory arrest, loss of consciousness with respiratory
Table 7 Non-Invasive Ventilation: Predictors of Success or Failure
Severity of illness
Acidosis (pH , 7.25)
Hypercapnia (PaCo2 . 80 and pH ,7.30)
APACHE II (.20)
Level of consciousness
Neurologic score (.4, stuporous, arousal only after vigorous stimulation;
inconsistently follows commands)
Encephalopathy score (.3, major confusion, daytime sleepiness, or agitation)
Glasgow coma score (,8)
Disease conditions
Acute respiratory distress syndrome
Pneumonia
Restrictive lung disease
Predictors of success
Response to brief trial of NIMV (1 3 hr)
Decrease in PaCO2 . 8 mm Hg
Improvement in pH . 0.06
Correction of respiratory acidosis
Time frame for failure requiring intubation
Failure of improvement with NIMV
Within 12 24 hr
Late failures (.48 hr after initiation of non-invasive ventilation)
Admission predictors of failure
Lower functional status (Activity score ,2, dyspnea with light activity)
Initial acidosis (pH 7.22)
Hospital complications (pneumonia, shock, coma)
Abbreviation: NIMV, non-invasive ventilation.
pauses, gasping for air, psychomotor agitation making nursing impossible

requiring sedation, heart rate ,50 with loss of alertness, and hemodynamically
instability with systolic blood pressure ,70 mm Hg. Minor criteria require two
of the following: respiratory rate .35, pH , 7.30 and decreased from onset,
PaO2 , 45 mm Hg despite oxygen, and increase in encephalopathy or decreased
level of consciousness. These recommendations are outlined in Table 8, and
additional guidelines used in our institution are included for comparison.
However, despite established criteria, there still exist different thresholds
for intubation by the individual physician. Arterial blood gases are a major pre-
dictor of respiratory failure, yet there is quite a difference between a pH of 7.30
and 7.20. Many patients with severe respiratory distress are able to maintain
acceptable gas exchange and oxygenation, yet undergo intubation because of
other factors or because of the potential for developing frank respiratory
failure. Investigations in NIMV are not amenable to treatment blinding for
obvious reasons. Therefore, there exists an extraordinary study bias with
64 Soo Hoo
Table 8 Non-Invasive Ventilation: Intubation Criteria
Brochard et al. (17)

Major (any one of the following)
Respiratory arrest
Loss of consciousness with respiratory pauses
Gasping for air
Psychomotor agitation requiring sedation
Heart rate , 50 with loss of alertness
Hemodynamic instability with systolic blood pressure , 70 mm Hg
Minor (two of the following)
Respiratory rate . 35
pH , 7.30 and decreased from onset
PaO2 , 45 mm Hg despite oxygen
Increase in encephalopathy or decreased level of consciousness
Plant et al. (18)
pH , 7.20
pH 7.20 7.25 on two occasions one hour apart
Hypercapnic coma (Glasgow coma scale , 8 and PaCO2 . 60 mm Hg)
PaO2 , 45 mm Hg
Cardiopulmonary arrest
Intubation guidelines (two or more in the context of respiratory distress)
RR . 35/min or ,6/min
Tidal volume , 5 mL/kg
Blood pressure changes: systolic ,90 mm Hg
Oxygen desaturation ,90% despite adequate supplemental oxygen
Hypercapnia (PaCO2 . 10 mm increase) or acidosis (pH decline .0.08) from
baseline
Obtundation
Diaphoresis
Abdominal paradox
respect to the outcome measure of intubation. There is often a natural tendency to

intubate early if there is a decision to intubate, but to defer intubation as long as
possible in those patients who wish to avoid intubation. This can lead to signi-
cant discrepancies in treatment. It should be noted that in the multi-center trial
that used major and minor criteria for intubation, major criteria for intubation
were met in 73% of the non-invasive group, but only 32% of the control
group. The guidelines are not meant to be absolute criteria, but to supplement
other clinical data and to provide some guidance, especially in cases where
laboratory data or measures of gas exchange may not be immediately available.
B. Complications
There are markedly different proles reported for complications associated with
intubation and mechanical ventilation and NIMV (82). The vast majority of these
complications can be attributed to direct injury from the patient interface, and in
the case of NIMV, represent pressure injury and skin necrosis from the mask.
This is by far the most frequent complication, occurring in about 10% of
reports. Some of these lesions can be quite extensive, with erosions of the
nasal bridge to the nasal cartilage or even bone. Gauze or other types of protective
skin dressing have been used to minimize this complication. Softer masks, redu-
cing strap pressure, and padding in key pressure points are other adjustments that
may lessen the incidence of this complication. This underscores the need to
provide breaks off NIMV for patients. Other reported complications are relatively
minor and include problems with nasal congestion, sinusitis, gastric distension,
and eye irritation. Major complications such as barotrauma, nosocomial pneumo-
nia, and sepsis are distinctly uncommon. There have been case reports of unusual
complications such as esophageal perforation, re-opening of an esophageal
pleural stula, and orbital herniation (108 110).
C. Reduction in Nosocomial Pneumonia
The reduction in nosocomial pneumonia is particularly noteworthy. This is related

to the advantage NIMV has over endotracheal intubation in preserving speech
function and upper airway defenses and permitting oral intake. This prevents
aspiration or microaspiration and allows more effective clearance of secretions.
Patients do not require sedation. This has the net effect of reducing the risk of
acquiring nosocomial pneumonia. In their randomized trial, Brochard et al. noted
a reduced incidence of nosocomial pneumonia and sepsis. His group has extended
this observation to retrospectively analyze, in a matched case-control study, 50
patients treated with NIMV and 50 endotracheally intubated patients (111). They
found a marked reduction in the nosocomial infections (18% vs. 60%), especially
nosocomial pneumonia (4% vs. 11%), in those treated with NIMV. The reduction
in the other sites of nosocomial infection were attributed to shorter ICU stays for the
NIMV patients. There was also less antibiotic use, fewer days of ventilatory support
(6 + 6 days vs. 10 + 12 days), and lower mortality. They concluded that this
benet was due in part to elimination of the adverse effects of endotracheal intuba-
tion and mechanical ventilation and reduction in nosocomial infections. NIMV
would also eliminate any post-extubation risks for nosocomial pneumonia (upper
airway edema, stridor, and aspiration). They extended this observation to 479
patients with acute respiratory failure (COPD or cardiogenic pulmonary edema),
of which 313 (65%) were treated with NIMV (112). In their retrospective analysis
of this cohort, they noted that NIMV was associated with a reduction in the inci-
dence of nosocomial infections, especially nosocomial pneumonia, reduction in
ICU stay by two days, and a threefold reduction in mortality.
D. Personnel and Systems Issues

The early experience with NIMV was tempered by an inordinate amount of
time (over 90%) spent by nursing staff in the care of some patients during
66 Soo Hoo
NIMV (113). This was attributed to alarms, mask adjustments, constant monitor-
ing, and equipment issues (nasal masks and portable volume ventilators). This
was likely a reection of the initial learning curve that all personnel must encoun-
ter with this modality. This is clearly a technique that only improves with repeated
administration and experience. Patients treated with NIMV also comprise a
somewhat tenuous group of patients. Intubation and mechanical ventilation
may be perceived as easier, requiring less time and attention, increasing resistance
to its use. However, several controlled trials have subsequently demonstrated that
the time commitment is very comparable to that required for those treated with
conventional therapy undergoing intubation and mechanical ventilation.
Kramer et al. (16) were the rst to address this issue and found an average
greater initial time commitment by respiratory therapists of 56 minutes during
the rst eight hours of therapy, which declined the next eight hours, so that
they actually spent less time in NIMV patients compared with control patients.
This experience mirrors that of Nava et al. (114), who found that the respiratory
therapist spent on average an extra 74 minutes during the initial eight hours of
NIMV, decreasing to equivalent time commitments at the end of 48 hours. In
an analysis of workload by Plant et al. (18), they found that nursing workload
increased by an average of only 26 minutes during the rst eight hours of NIMV,
but by 48 hours the difference was only 21 minutes and not signicant. This
represents a special situation as nursing staff provided the primary respiratory
care for the patients in the study. Other investigators have also reported greater
efciencies over time, evidenced by improving success rates with NIMV (81).
The other systems issue with NIMV involves the optimal location for its
applications. There is a wide range of experience that probably reects differ-
ences in health-care systems and practices, as well as local familiarity with the
technique. Although this has been demonstrated to a feasible technique in an
unmonitored ward setting, with or without specialty expertise, the vast majority
of patients are treated in a monitored setting. This includes the Emergency
Department, where treatment can be initiated, and it can be continued in a
specialized respiratory unit (with continuous monitoring), step-down unit, or
ICU. The need for monitoring stems from the potential need for intubation and
mechanical ventilation in those that fail therapy.
This has obvious implications with respect to health care costs. While
early studies indicate that the resources and costs required for NIMV were com-
parable to that for endotracheal intubation and mechanical ventilation, other
analysis suggest cost savings with this technique (16,114). One evaluation
reected the experience of utilizing NIMV in a ward setting, but another ana-
lyzed its use in critical care units (19,115). These analyses are probably not
applicable to the United States, where the use of NIMV requires a monitored
setting. The other unrecognized element of this analysis is that the health-care
reimbursement system has not yet recognized NIMV as a viable treatment. It
is difcult to code under current guidelines, resulting in suboptimal reimburse-
ment when compared with intubation and mechanical ventilation. In some
centers, the code for CPAP (used to treat obstructive sleep apnea) is used to
code for NIMV. It may be similar in concept but vastly different in personnel
and resource utilization. This serves as a negative incentive for its use and
may explain why this technique seems more prevalent in Europe and Canada
than in the United States.
VII. Conclusion
Although the efcacy of NIMV has been clearly demonstrated and is endorsed by
many, its utilization remains limited. The enthusiasm for treatment is tempered
by some research and practical limitations. Despite well-designed and well-
conducted prospective randomized studies, the actual numbers of patients under-
going study are relatively small. The studies cannot be blinded for obvious
reasons, and the comparison intubation rates in those managed with standard
therapy is inuenced by a treatment bias inherent in this type of study. There
is likely a lower threshold for the intubation of standard therapy patients than
NIMV patients despite the presence of intubation guidelines. Issues with
increased nursing staff and respiratory therapy time have been cited as possible
limiting factors. It is acknowledged that there exists a learning curve with this
technique, and it may be steeper in some institutions or situations. This treatment
does require patience and commitment by all involved. In some circumstances, it
is often easier for the patient to be endotracheally intubated.
On the other hand, once staff has gained sufcient experience with NIMV,
it is an invaluable treatment option. It is ideally suited for patients with COPD
with hypercapnic respiratory failure because this is often a rapidly reversible
condition. Patients with cardiogenic pulmonary edema are probably the other
major group that can be managed with NIMV. It allows dyspnea relief and ven-
tilatory support, preventing further deterioration while allowing other treatments
to take effect. It is intermittently applied, and treatment is discontinued once the
mask is removed. Therefore, short breaks off NIMV serve a dual purpose.
Patients are provided relief from skin and facial ischemia and necrosis, and the
breaks serve as weaning trials. Patients who do not tolerate these breaks can
be placed immediately back on support. This reduces much of the time and uncer-
tainty involved with discontinuing mechanical ventilation in those who are endo-
tracheally intubated. Patients are spared the potential complications and adverse
effects associated with re-intubation. Preservation of upper airway function
reduces the nosocomial/ventilator-associated pneumonia rate. All of these
factors serve to reduce the duration of ventilatory support, ICU stay, and possibly
duration of hospitalization. Reduction in mortality has not been universally
noted. However, the positive experiences call for continued study in the modality,
rening issues regarding patient selection and delivery of ventilatory support. As
it becomes easier to provide NIMV, barriers to its use will disappear, and it will
be the treatment of choice in select patients.
68 Soo Hoo
References
1. Mehta S, Hill NS. Noninvasive ventilation. Am J Respir Crit Care Med 2001;
163(2):540 577.
2. British Thoracic Society Standards of Care Committee. Non-invasive ventilation in
acute respiratory failure. Thorax 2002; 57(3):192 211.
3. International Consensus Conferences in Intensive Care Medicine: noninvasive
positive pressure ventilation in acute Respiratory failure. Am J Respir Crit Care
Med 2001; 163(1):283 291.
4. Severinghaus JW, Astrup P, Murray JF. Blood gas analysis and critical care medi-
cine. Am J Respir Crit Care Med 1998; 157(4 Pt 2):S114 S122.
5. Sandur S, Stoller JK. Pulmonary complications of mechanical ventilation.
Clin Chest Med 1999; 20(2):223247.
6. Chastre J, Fagon JY. Ventilator-associated pneumonia. Am J Respir Crit Care Med
2002; 165(7):867 903.
7. Ellis ER, Bye PT, Bruderer JW, Sullivan CE. Treatment of respiratory failure
during sleep in patients with neuromuscular disease. Positive-pressure ventilation
through a nose mask. Am Rev Respir Dis 1987; 135(1):148 152.
8. Menzies R, Gibbons W, Goldberg P. Determinants of weaning and survival among
patients with COPD who require mechanical ventilation for acute respiratory
failure. Chest 1989; 95(2):398 405.
9. Carrey Z, Gottfried SB, Levy RD. Ventilatory muscle support in
respiratory failure with nasal positive pressure ventilation. Chest 1990;
97(1):150 158.
10. Belman MJ, Soo Hoo GW, Kuei JH, Shadmehr R. Efcacy of positive vs
negative pressure ventilation in unloading the respiratory muscles. Chest 1990;
98(4):850 856.
11. Brochard L, Isabey D, Piquet J, Amaro P, Mancebo J, Messadi AA,
Brun-Buisson C, Raws A, Lemaire F, Harf A. Reversal of acute exacerbations of
chronic obstructive lung disease by inspiratory assistance with a face mask. N
Engl J Med 1990; 323(22):1523 1530.
12. Meduri GU, Abou-Shala N, Fox RC, Jones CB, Leeper KV, Wunderink RG.
Noninvasive face mask mechanical ventilation in patients with acute hypercapnic
respiratory failure. Chest 1991; 100(2):445 454.
13. Meyer TJ, Hill NS. Noninvasive positive pressure ventilation to treat respiratory
failure. Ann Intern Med 1994; 120(9):760 770.
14. Corrado A, Confalonieri M, Marchese S, Mollica C, Villela G, Gorini M,
Bella Porta R. Iron lung vs mask ventilation in the treatment of acute on chronic
respiratory failure in COPD patients: a multicenter study. Chest 2002;
121(1):189 195.
15. Bott J, Carroll MP, Conway JH, Keilty SE, Ward EM, Brown AM, Paul EA,
Elliot MW, Godfrey RC, Wodzicha JA. Randomised controlled trial of nasal ven-
tilation in acute ventilatory failure due to chronic obstructive airways disease.
Lancet 1993; 341(8860):1555 1557.
16. Kramer N, Meyer TJ, Meharg J, Cece RD, Hill NS. Randomized, prospective trial
of noninvasive positive pressure ventilation in acute respiratory failure. Am J
Respir Crit Care Med 1995; 151(6):1799 1806.
17. Brochard L, Mancebo J, Wysocki M, Lofaso F, Conti G, Rauss A, Simonncau G,

Benito S, Gasparetto A, Lemaire F. Noninvasive ventilation for acute exacerbations
of chronic obstructive pulmonary disease. N Engl J Med 1995; 333(13):817 822.
18. Plant PK, Owen JL, Elliott MW. Early use of non-invasive ventilation for acute
exacerbations of chronic obstructive pulmonary disease on general respiratory
wards: a multicentre randomised controlled trial. Lancet 2000; 355(9219):19311935.
19. Keenan SP, Gregor J, Sibbald WJ, Cook D, Gafni A. Noninvasive positive pressure
ventilation in the setting of severe, acute exacerbations of chronic obstructive pul-
monary disease: more effective and less expensive. Crit Care Med 2000;
28(6):2094 2102.
20. Keenan SP, Kernerman PD, Cook DJ, Martin CM, McCormack D, Sibbald WJ.
Effect of noninvasive positive pressure ventilation on mortality in patients
admitted with acute respiratory failure: a meta-analysis. Crit Care Med 1997;
25(10):1685 1692.
21. Peter JV, Moran JL, Phillips-Hughes J, Warn D. Noninvasive ventilation in acute
respiratory failurea meta-analysis update. Crit Care Med 2002; 30(3):555 562.
22. Keenan SP, Sinuff T, Cook DJ, Hill NS. Which patients with acute exacerbation of
chronic obstructive pulmonary disease benet from noninvasive positive-pressure
ventilation? A systematic review of the literature. Ann Intern Med 2003;
138(11):861 870.
23. Hopewell PC, Murray JF. Effects of continuous positive-pressure ventilation in
experimental pulmonary edema. J Appl Physiol 1976; 40(4):568 574.
24. Luce JM. The cardiovascular effects of mechanical ventilation and positive end-
expiratory pressure. J Am Med Assoc 1984; 252(6):807 811.
25. Naughton MT, Rahman MA, Hara K, Floras JS, Bradley TD. Effect of continuous
positive airway pressure on intrathoracic and left ventricular transmural pressures
in patients with congestive heart failure. Circulation 1995; 91(6):1725 1731.
26. Baratz DM, Westbrook PR, Shah PK, Mohsenifar Z. Effect of nasal continuous
positive airway pressure on cardiac output and oxygen delivery in patients with
congestive heart failure. Chest 1992; 102(5):1397 1401.
27. Bradley TD, Holloway RM, McLaughlin PR, Ross BL, Walters J, Liu PP. Cardiac
output response to continuous positive airway pressure in congestive heart failure.
Am Rev Respir Dis 1992; 145(2 Pt 1):377 382.
28. Barach AL MJEM. Positive pressure respiration and its application to the treatment
of acute pulmonary edema. Ann Intern Med 1938; 12:754 795.
29. Rasanen J, Heikkila J, Downs J, Nikki P, Vaisanen I, Viitanen A. Continuous posi-
tive airway pressure by face mask in acute cardiogenic pulmonary edema. Am J
Cardiol 1985; 55(4):296 300.
30. Bersten AD, Holt AW, Vedig AE, Skowronski GA, Baggoley CJ. Treatment of
severe cardiogenic pulmonary edema with continuous positive airway pressure
delivered by face mask. N Engl J Med 1991; 325(26):1825 1830.
31. Lin M, Yang YF, Chiang HT, Chang MS, Chiang BN, Cheitlin MD. Reappraisal
of continuous positive airway pressure therapy in acute cardiogenic pulmonary
edema. Short-term results and long-term follow-up. Chest 1995; 107(5):13791386.
32. Pang D, Keenan SP, Cook DJ, Sibbald WJ. The effect of positive pressure airway
support on mortality and the need for intubation in cardiogenic pulmonary edema: a
systematic review. Chest 1998; 114(4):1185 1192.
70 Soo Hoo
33. Masip J, Betbese AJ, Paez J, Vecilla F, Canizares R, Padro J, Paz MA, de Otero J,
Ballus J. Non-invasive pressure support ventilation versus conventional oxygen
therapy in acute cardiogenic pulmonary oedema: a randomised trial. Lancet
2000; 356(9248):2126 2132.
34. Sharon A, Shpirer I, Kaluski E, Moshkovitz Y, Milovanov O, Polak R,
Blatt A, Simovitz A, Shaham D, Feigenberg Z, et al. High-dose intravenous
isosorbide-dinitrate is safer and better than Bi-PAP ventilation combined with
conventional treatment for severe pulmonary edema. J Am Coll Cardiol 2000;
36(3):832 837.
35. Nava S, Carbone G, DiBattista N, Bellone A, Baiardi P, Cosentini R, Marenco M,
Giostra F, Borasi G, Groff P. Noninvasive ventilation in cardiogenic pulmonary
edema: a multicenter randomized trial. Am J Respir Crit Care Med 2003;
168(12):1432 1437.
36. Levitt MA. A prospective, randomized trial of BiPAP in severe acute congestive
heart failure. J Emerg Med 2001; 21(4):363 369.
37. Chadda K, Annane D, Hart N, Gajdos P, Raphael JC, Lofaso F. Cardiac and respiratory
effects of continuous positive airway pressure and noninvasive ventilation in acute
cardiac pulmonary edema. Crit Care Med 2002; 30(11):24572461.
38. Mehta S, Jay GD, Woolard RH, Hipona RA, Connolly EM, Cimini DM,
Drinkwine JH, Mill NS. Randomized, prospective trial of bilevel versus continuous
positive airway pressure in acute pulmonary edema. Crit Care Med 1997;
25(4):620 628.
39. Liesching TN, Cromier K, Nelson D, Short K, Sucov A, Hill NS. Bilevel noninva-
sive ventilation vs continuous positive airway pressure to treat acute pulmonary
edema. Am J Respir Crit Care Med 2004; 167(7):A864.
40. Pennock BE, Kaplan PD, Carlin BW, Sabangan JS, Magovern JA. Pressure
support ventilation with a simplied ventilatory support system administered
with a nasal mask in patients with respiratory failure. Chest 1991;
100(5):1371 1376.
41. Wysocki M, Tric L, Wolff MA, Gertner J, Millet H, Herman B. Noninvasive
pressure support ventilation in patients with acute respiratory failure. Chest 1993;
103(3):907 913.
42. Epstein SK, Ciubotaru RL, Wong JB. Effect of failed extubation on the outcome of
mechanical ventilation. Chest 1997; 112(1):186 192.
43. Nathan SD, Ishaaya AM, Koerner SK, Belman MJ. Prediction of minimal
pressure support during weaning from mechanical ventilation. Chest 1993;
103(4):1215 1219.
44. Ishaaya AM, Nathan SD, Belman MJ. Work of breathing after extubation. Chest
1995; 107(1):204 209.
45. Vitacca M, Ambrosino N, Clini E, Porta R, Rampulla C, Lanini B, Nava S.
Physiological response to pressure support ventilation delivered before and after
extubation in patients not capable of totally spontaneous autonomous breathing.
46. Nava S, Ambrosino N, Clini E, Prato M, Orlando G, Vitacca M, Brigada P,
Frachia C, Rubini F. Noninvasive mechanical ventilation in the weaning of patients
with respiratory failure due to chronic obstructive pulmonary disease. A random-
ized, controlled trial. Ann Intern Med 1998; 128(9):721 728.
47. Girault C, Daudenthun I, Chevron V, Tamion F, Leroy J, Bonmarchand G.

Noninvasive ventilation as a systematic extubation and weaning technique in
acute-on-chronic respiratory failure: a prospective, randomized controlled study.
48. Ferrer M, Esquinas A, Arancibia F, Bauer TT, Gonzalez G, Carillo A, Rodriguez-
Roisin R, Torres A. Noninvasive ventilation during persistent weaning failure: a
randomized controlled trial. Am J Respir Crit Care Med 2003; 168(1):70 76.
49. Hill N, Lin D, Levy M, OBrien A, Klinger J, Houtchens J, Nelson D,
McCormick J. Noninvasive positive pressure ventilation (NPPV) to facilitate extu-
bation after acute respiratory failure: a feasibility study. Am J Respir Crit Care Med
2000; 161(3):A263.
50. Jiang JS, Kao SJ, Wang SN. Effect of early application of biphasic positive airway
pressure on the outcome of extubation in ventilator weaning. Respirology 1999;
4(2):161 165.
51. Keenan SP, Powers C, McCormack DG, Block G. Noninvasive positive-pressure
ventilation for postextubation respiratory distress: a randomized controlled trial.
J Am Med Assoc 2002; 287(24):3238 3244.
52. Esteban A, Frutos-Vivar F, Ferguson ND, Arabi Y, Apezteguia C, Gonzalez M,
Epstein SK, Hill NS, Nava S, Soares MA, et al. Noninvasive positive-pressure
ventilation for respiratory failure after extubation. N Engl J Med 2004;
350(24):2452 2460.
53. Confalonieri M, Potena A, Carbone G, Porta RD, Tolley EA, Umberto MG. Acute
respiratory failure in patients with severe community-acquired pneumonia. A pro-
spective randomized evaluation of noninvasive ventilation. Am J Respir Crit Care
Med 1999; 160(5 Pt 1):1585 1591.
54. Delclaux C, LHer E, Alberti C, Mancebo J, Abroug F, Conti G, Guerin C,
Schortgen F, Lefort Y, Antonelli M, et al. Treatment of acute hypoxemic nonhyper-
capnic respiratory insufciency with continuous positive airway pressure delivered
by a face mask: a randomized controlled trial. J Am Med Assoc 2000;
284(18):2352 2360.
55. Ferrer M, Esquinas A, Leon M, Gonzalez G, Alarcon A, Torres A. Noninvasive
ventilation in severe hypoxemic respiratory failure: a randomized clinical trial.
56. Wysocki M, Tric L, Wolff MA, Millet H, Herman B. Noninvasive pressure support
ventilation in patients with acute respiratory failure. A randomized comparison with
conventional therapy. Chest 1995; 107(3):761 768.
57. Antonelli M, Conti G, Rocco M, Bu M, De Blasi RA, Vivino G, Gasparetto A,
Meduri GV. A comparison of noninvasive positive-pressure ventilation and con-
ventional mechanical ventilation in patients with acute respiratory failure. N
Engl J Med 1998; 339(7):429 435.
58. Martin TJ, Hovis JD, Costantino JP, Bierman MI, Donhoe MP, Rogers RM,
Kreit JW, Sciurba FC, Stiller RA, Sanders MH. A randomized, prospective evalu-
ation of noninvasive ventilation for acute respiratory failure. Am J Respir Crit Care
Med 2000; 161(3 Pt 1):807 813.
59. Ambrosino N, Foglio K, Rubini F, Clini E, Nava S, Vitacca M. Non-invasive mech-
anical ventilation in acute respiratory failure due to chronic obstructive pulmonary
disease: correlates for success. Thorax 1995; 50(7):755 757.
72 Soo Hoo
60. Antonelli M, Conti G, Moro ML, Equinas, A, Gonzabz-Diaz G, Confalonieri M,

Pelaia P, Principi T, Gregoretti C, Beltrame F, et al. Predictors of failure of nonin-
vasive positive pressure ventilation in patients with acute hypoxemic respiratory
failure: a multi-center study. Intensive Care Med 2001; 27(11):1718 1728.
61. Rubenfeld GD, Crawford SW. Withdrawing life support from mechanically venti-
lated recipients of bone marrow transplants: a case for evidence-based guidelines.
Ann Intern Med 1996; 125(8):625 633.
62. Antonelli M, Conti G, Bu M, Costa MG, Lappa A, Rocco M, Gasparetto A,
Meduri GV. Noninvasive ventilation for treatment of acute respiratory failure in
patients undergoing solid organ transplantation: a randomized trial. J Am Med
Assoc 2000; 283(2):235 241.
63. Hill NS. Noninvasive ventilation for immunocompromised patients. N Engl J Med
2001; 344(7):522 524.
64. Soroksky A, Stav D, Shpirer I. A pilot prospective, randomized, placebo-controlled
trial of bilevel positive airway pressure in acute asthmatic attack. Chest 2003;
123(4):1018 1025.
65. Kutlu CA, Williams EA, Evans TW, Pastorino U, Goldstraw P. Acute lung injury
and acute respiratory distress syndrome after pulmonary resection. Ann Thorac
Surg 2000; 69(2):376 380.
66. Auriant I, Jallot A, Herve P, Cerrina J, Le Roy Cadurie F, Fournier JL,
Lescot B, Parguin F. Noninvasive ventilation reduces mortality in acute respiratory
failure following lung resection. Am J Respir Crit Care Med 2001; 164(7):1231
1235.
67. Bolliger CT, Van Eeden SF. Treatment of multiple rib fractures. Randomized con-
trolled trial comparing ventilatory with nonventilatory management. Chest 1990;
97(4):943 948.
68. Rocker GM, Mackenzie MG, Williams B, Logan PM. Noninvasive positive
pressure ventilation: successful outcome in patients with acute lung injury/
ARDS. Chest 1999; 115(1):173 177.
69. Benhamou D, Girault C, Faure C, Portier F, Muir JF. Nasal mask ventilation in acute
respiratory failure. Experience in elderly patients. Chest 1992; 102(3):912917.
70. Meduri GU, Fox RC, Abou-Shala N, Leeper KV, Wunderink RG. Noninvasive
mechanical ventilation via face mask in patients with acute respiratory failure
who refused endotracheal intubation. Crit Care Med 1994; 22(10):1584 1590.
71. Chu CM, Chan VL, Wong IW, Leung WS, Lin AW, Cheung KF. Noninvasive ven-
tilation in patients with acute hypercapnic exacerbation of chronic obstructive pul-
monary disease who refused endotracheal intubation. Crit Care Med 2004;
32(2):372 377.
72. Freichels TA. Palliative ventilatory support: use of noninvasive positive
pressure ventilation in terminal respiratory insufciency. Am J Crit Care 1994;
3(1):6 10.
73. Clarke DE, Vaughan L, Rafn TA. Noninvasive positive pressure ventilation for
patients with terminal respiratory failure: the ethical and economic costs of delay-
ing the inevitable are too great. Am J Crit Care 1994; 3(1):4 5.
74. Madden BP, Kariyawasam H, Siddiqi AJ, Machin A, Pryor JA, Hodson ME. Non-
invasive ventilation in cystic brosis patients with acute or chronic respiratory
failure. Eur Respir J 2002; 19(2):310 313.
75. Doherty MJ, Greenstone MA. Survey of non-invasive ventilation (NIPPV) in

patients with acute exacerbations of chronic obstructive pulmonary disease
(COPD) in the UK. Thorax 1998; 53(10):863 866.
76. Esteban A, Anzueto A, Frutos F, Alia I, Brochard L, Stewart TE, Benito S, Epstein
SK, Apezteguia C, Nightingale P, et al. Characteristics and outcomes in adult
patients receiving mechanical ventilation: a 28-day international study. J Am
Med Assoc 2002; 287(3):345 355.
77. Esteban A, Anzueto A, Alia I, Gordo F, Apezteguia C, Palizas F, Cide D,
Goldwasser R, Soto L, Bugedo G, et al. How is mechanical ventilation employed
in the intensive care unit? An international utilization review. Am J Respir Crit
Care Med 2000; 161(5):1450 1458.
78. Carlucci A, Richard JC, Wysocki M, Lepage E, Brochard L. Noninvasive versus
conventional mechanical ventilation. An epidemiologic survey. Am J Respir Crit
Care Med 2001; 163(4):874 880.
79. Plant PK, Owen JL, Elliott MW. Non-invasive ventilation in acute exacerbations of
chronic obstructive pulmonary disease: long term survival and predictors of in-hos-
pital outcome. Thorax 2001; 56(9):708 712.
80. Soo Hoo GW, Santiago S, Williams AJ. Nasal mechanical ventilation for hyper-
capnic respiratory failure in chronic obstructive pulmonary disease: determinants
of success and failure. Crit Care Med 1994; 22(8):1253 1261.
81. Carlucci A, Delmastro M, Rubini F, Fracchia C, Nava S. Changes in the practice of
non-invasive ventilation in treating COPD patients over 8 years. Intensive Care
Med 2003; 29(3):419 425.
82. Abou-Shala N, Meduri U. Noninvasive mechanical ventilation in patients with
acute respiratory failure. Crit Care Med 1996; 24(4):705 715.
83. Schwartz AR, Kacmarek RM, Hess DR. Factors affecting oxygen delivery with
bi-level positive airway pressure. Respir Care 2004; 49(3):270 275.
84. Criner GJ, Travaline JM, Brennan KJ, Kreimer DT. Efcacy of a new full
face mask for noninvasive positive pressure ventilation. Chest 1994; 106(4):
1109 1115.
85. Antonelli M, Pennisi MA, Pelosi P, Gregoretti C, Squadrone V, Rocco M, Cecchini L,
Chiumello D, Severgnini P, Proietti R, et al. Noninvasive positive pressure venti-
lation using a helmet in patients with acute exacerbation of chronic obstructive pul-
monary disease: a feasibility study. Anesthesiology 2004; 100(1):16 24.
86. Chiumello D, Pelosi P, Carlesso E, Severgnini P, Aspesi M, Gamberoni C,
Antonelli M, Conti G, Chiaranda M, Gattinoni L. Noninvasive positive pressure
ventilation delivered by helmet vs. standard face mask. Intensive Care Med
2003; 29(10):1671 1679.
87. Schonhofer B, Sortor-Leger S. Equipment needs for noninvasive mechanical
ventilation. Eur Respir J 2002; 20(4):1029 1036.
88. Kwok H, McCormack J, Cece R, Houtchens J, Hill NS. Controlled trial of oronasal
versus nasal mask ventilation in the treatment of acute respiratory failure. Crit Care
Med 2003; 31(2):468 473.
89. Navalesi P, Fanfulla F, Frigerio P, Gregoretti C, Nava S. Physiologic evaluation of
noninvasive mechanical ventilation delivered with three types of masks in
patients with chronic hypercapnic respiratory failure. Crit Care Med 2000;
28(6):1785 1790.
74 Soo Hoo
90. Anton A, Tarrega J, Giner J, Guell R, Sanchis J. Acute physiologic effects of nasal
and full-face masks during noninvasive positive-pressure ventilation in patients
with acute exacerbations of chronic obstructive pulmonary disease. Respir Care
2003; 48(10):922 925.
91. Vitacca M, Rubini F, Foglio K, Scalvini S, Nava S, Ambrosino N. Non-invasive
modalities of positive pressure ventilation improve the outcome of acute exacer-
bations in COLD patients. Intensive Care Med 1993; 19(8):450 455.
92. Girault C, Richard JC, Chevron V, Tamion F, Pasquis P, Leroy J, Bonmerchand G.
Comparative physiologic effects of noninvasive assist-control and pressure support
ventilation in acute hypercapnic respiratory failure. Chest 1997; 111(6):1639
1648.
93. Ferguson GT, Gilmartin M. CO2 rebreathing during BiPAP ventilatory assistance.
94. Lofaso F, Brochard L, Touchard D, Hang T, Harf A, Isabey D. Evaluation of carbon
dioxide rebreathing during pressure support ventilation with airway management
system (BiPAP) devices. Chest 1995; 108(3):772 778.
95. Nava S, Ambrosino N, Bruschi C, Confalonieri M, Rampulla C. Physiological effects
of ow and pressure triggering during non-invasive mechanical ventilation in patients
with chronic obstructive pulmonary disease. Thorax 1997; 52(3):249254.
96. Hotchkiss JR, Adams AB, Dries DJ, Marini JJ, Crooke PS. Dynamic behavior
during noninvasive ventilation: chaotic support? Am J Respir Crit Care Med
2001; 163(2):374 378.
97. Gay PC, Hess DR, Hill NS. Noninvasive proportional assist ventilation for acute
respiratory insufciency. Comparison with pressure support ventilation. Am J
98. Wysocki M, Richard JC, Meshaka P. Noninvasive proportional assist ventilation
compared with noninvasive pressure support ventilation in hypercapnic acute res-
piratory failure. Crit Care Med 2002; 30(2):323 329.
99. Massie CA, Hart RW, Peralez K, Richards GN. Effects of humidication on nasal
symptoms and compliance in sleep apnea patients using continuous positive airway
pressure. Chest 1999; 116(2):403 408.
100. Pollack CV Jr, Fleisch KB, Dowsey K. Treatment of acute bronchospasm with beta-
adrenergic agonist aerosols delivered by a nasal bilevel positive airway pressure
circuit. Ann Emerg Med 1995; 26(5):552 557.
101. Chatmongkolchart S, Schettino GP, Dillman C, Kacmarek RM, Hess DR. In vitro
evaluation of aerosol bronchodilator delivery during noninvasive positive pressure
ventilation: effect of ventilator settings and nebulizer position. Crit Care Med 2002;
30(11):2515 2519.
102. Jaber S, Fodil R, Carlucci A, Boussarsand M, Pigeot J, Lemaire F, Hharf A, Lofuso F,
Isabey D, Brochard L. Non invasive ventilation with helium oxygen in acute
exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care
Med 2000; 161(4 Pt 1): 1191 1200.
103. Nava S, Ceriana P. Causes of failure of noninvasive mechanical ventilation.
Respir Care 2004; 49(3):295 303.
104. Anton A, Guell R, Gomez J, Serrano J, Castellano A, Carrasco JL, Sanchis J.
Predicting the result of noninvasive ventilation in severe acute exacerbations of
patients with chronic airow limitation. Chest 2000; 117(3):828 833.
105. Moretti M, Cilione C, Tampieri A, Fracchia C, Marchioni A, Nava S. Incidence and

causes of non-invasive mechanical ventilation failure after initial success. Thorax
2000; 55(10):819 825.
106. Robino C, Faisy C, Diehl JL, Rezgui N, Labrousse J, Guerot E. Effectiveness of
non-invasive positive pressure ventilation differs between decompensated chronic
restrictive and obstructive pulmonary disease patients. Intensive Care Med 2003;
29(4):603 610.
107. Domenighetti G, Gayer R, Gentilini R. Noninvasive pressure support ventilation in
non-COPD patients with acute cardiogenic pulmonary edema and severe commu-
nity-acquired pneumonia: acute effects and outcome. Intensive Care Med 2002;
28(9):1226 1232.
108. Van de LA, Brocas E, Boiteau R, Perrin-Gachadoat D, Tenaillon A. Esophageal
perforation associated with noninvasive ventilation: a case report. Chest 2002;
122(5):1857 1858.
109. Hurst JR, Polkey MI, Goldstraw P, Yung B. Oesophagopleural stula as a novel
cause of failed non-invasive ventilation. Thorax 2003; 58(7):642 643.
110. Lazowick D, Meyer TJ, Pressman M, Peterson D. Orbital herniation associated with
noninvasive positive pressure ventilation. Chest 1998; 113(3):841 843.
111. Girou E, Schortgen F, Delclaux C, Brun-Buisson C, Blot F, Lefort Y, Lemaire F,
Brochard L. Association of noninvasive ventilation with nosocomial infections
and survival in critically ill patients. J Am Med Assoc 2000; 284(18):2361 2367.
112. Girou E, Brun-Buisson C, Taille S, Lemaire F, Brochard L. Secular trends in
nosocomial infections and mortality associated with noninvasive ventilation in
patients with exacerbation of COPD and pulmonary edema. J Am Med Assoc
2003; 290(22):2985 2991.
113. Chevrolet JC, Jolliet P, Abajo B, Toussi A, Louis M. Nasal positive pressure
ventilation in patients with acute respiratory failure. Difcult and time-consuming
procedure for nurses. Chest 1991; 100(3):775 782.
114. Nava S, Evangelisti I, Rampulla C, Compagnoni ML, Fracchia C, Rubini F. Human
and nancial costs of noninvasive mechanical ventilation in patients affected by
COPD and acute respiratory failure. Chest 1997; 111(6):1631 1638.
115. Plant PK, Owen JL, Parrott S, Elliott MW. Cost effectiveness of ward based non-
invasive ventilation for acute exacerbations of chronic obstructive pulmonary
disease: economic analysis of randomised controlled trial. Br Med J 2003;
326(7396):956.
3
Modes of Mechanical Ventilation
BRIAN RICHARDS and ZAB MOSENIFAR
Division of Pulmonary and Critical Care Medicine, CedarsSinai Medical Center,

I. Introduction
In recent years, advances in ventilator technology have given rise to many new
and often confusing terms and acronyms used to describe modes of ventilator
function and patient interface. In an effort to bring order to this new terminology,
a system for understanding modes of mechanical ventilation has been developed
(1) and recently included in textbooks of respiratory care (2).
This new classication scheme enables logical description of new
ventilator modes that function in the same way but are given different brand
names. It also allows for grouping of ventilators in terms of the number of
breathing pattern options each ventilator offers. Despite the simplicity and
logic of this new system for describing ventilator modes, the terms and
abbreviations used in the system are not necessarily those used by physicians
who order mechanical ventilation or by respiratory care practitioners who
implement these orders.
Instead, when new ventilator modes are applied, they are usually ordered
by the terminology the manufacturers coin (and often trademark), which is
reinforced by their own clinical consultants when taught to end-users. In short,
the end-users of newer ventilator systems use a mix of community-accepted
terms and ventilator-specic labels to describe, order, and implement ventilator
mode and breath control settings. To expect otherwise would presume years of
re-educating physicians, nurses, and respiratory care practitioners in the
meaning and application of the new ventilator taxonomy and in the utility of
the new ventilator modes it describes.
77
78 Richards and Mosenifar
Accordingly, the reader is referred to the new system for classifying venti-
lator modes, but herein familiar terms will be used to describe popular ventilator
modes whenever possible and, when not, the manufacturer-labeled modes will be
described in simple terms.
II. The Breath
In the context of mechanical ventilation, the breath is the building block of breath
patterns called ventilator modes. A breath is dened as one inspiratory ow event
paired with one expiratory ow event. The areas circumscribed by each of these
ow events (i.e., the integral of ow) are dened as the inspiratory tidal volume
(Vt insp) and the expiratory tidal volume (Vt exp). A zero-ow inspiratory tidal
volume hold may or may not precede the expiratory ow event. A breath may or
may not include a passive expiratory phase that precedes the next breath (Fig. 1).
III. Breath Phases
The ventilator breath can be divided into four phases: the trigger phase, the limit
phase, the cycle phase, and the expiratory phase (2).
80
60
40 Zero Flow
20
.
VL/min 0
20 Passive Exp
40
60
0.5
80 0.5 1.5 sec sec
1.0 sec
sec
IT ET
(1.5 sec) (2.0 sec)
500
V
400 O
L
Vt ml 300
H
200 O
L
100 Vt insp D Vt exp
0
Figure 1 Machine breath, volume control with square ow wave. Computer

rendering of sketch drawn using common test lung, resistor, and Nellcor Puritan
Bennett 840 Ventilator graphics. Source: Courtesy of Tyco Health Care, Nellcor Puritan
Bennett, 2004.
Modes of Mechanical Ventilation 79
A. The Trigger Phase

The trigger phase denes what initiates an inspiratory ow event. In mechanical
ventilation, inspiration may be time-triggered or patient-triggered (2).
Time-Triggering
Time-Triggering is typically associated with control mode ventilation of the
patient who is apneic due to pathology or pharmacologic control of respiration.
The lapse of the ventilators breath cycle time [60 sec/min divided by set respir-
atory rate (RR)] automatically triggers inspiration.
Patient-Triggering
Patient-triggering occurs when patient inspiratory effort is translated into a
negative pressure signal (i.e., a pressure-trigger) or a negative ow signal
(i.e., a ow-trigger) that activates ventilator gas delivery.
1. Pressure-triggering occurs when a small user-set 1 to 2 cmH2O
pressure drop caused by patient inspiratory effort is communicated
to an upstream demand valve, which then opens to deliver ventilator
gas. Small pressure-trigger settings are more sensitive to patient
inspiratory efforts and make it easier for the patient to trigger ventilator
gas delivery. However, when positive end expiratory pressure (PEEP)
is set and a ventilator system leak causes expiratory baseline pressures
to fall below set PEEP, small pressure-trigger settings can result in
ventilator auto-triggering and ventilatorpatient dyssynchrony (3).
Conversely, large pressure-trigger settings are less sensitive to
patient breathing effort, increase patient trigger work, and delay venti-
lator demand valve activation. Large pressure-trigger settings are
especially problematic where expiratory air trapping results in auto-
PEEP that must be exceeded by patient-generated negative pressures
sufcient to overcome the auto-PEEP plus the user-set trigger pressure
(Fig. 2). Reducing the trigger pressure and setting a low-level PEEP
that approximates the auto-PEEP can decrease patient-trigger effort
and better synchronize it with machine breath delivery (4 7).
2. Flow-triggering employs a low level background (or bias) ow of gas
that is measured by ow sensors located at the ventilator circuits gas
inow port and the ventilator circuits gas outow port. When the
patient makes an inspiratory effort, part of this background ow of gas
goes to the patient, causing circuit outow to fall below circuit inow.
This difference in ow translates into a user-set ow signal, usually 1
to 3 lpm, that triggers inspiratory gas delivery to meet machine breath
settings or spontaneous breath demands (Fig. 3). In current generation
ventilators, the user can select ow-triggered or pressure-triggered
mechanisms to enable patient triggering of all breath types. Some studies
40
30 3cm H2O
20 Auto PEEP*
10
PcmH2O (-5 cm)Actual trigger
0
(-2 cm) Set Trigger threshold; add PEEP
Pressure of 3 cm to restore
-10
trigger threshold to
2cm below baseline.
-20
80
60
40 Air Trapping
20
.
VL/min 0
20
40
60
80
Time (sec)
Figure 2 Machine breath, volume control with auto-PEEP, resulting in increased

trigger work. Next machine breath electronically delayed so that rising pressure of
next breath does not mask auto-PEEP. Modeled after Nellcor Puritan Bennett 840 Venti-
lator graphics. Source: Courtesy of Tyco Health Care, Nellcor Puritan Bennett, 2004.
suggest that ow-triggering at minimum settings requires less trigger time

and trigger effort than pressure-triggering at minimum settings (8,9), but
the overall effect on total inspiratory effort generated throughout the
machine breath is small (7). Moreover, ow-triggers which are set low
to minimize trigger work are particularly sensitive to circuit leaks that
cause background gas outow to fall below background gas inow. In
such cases, ventilator ow sensors cannot distinguish between patient-
initiated trigger ow and leak ow. Thus, low-ow trigger settings
activated by chest tube leak (10) or articial airway leak may result in
ventilator auto-triggering and resultant ventilatorpatient dyssynchrony.
If such leaks cannot be xed, auto-triggering can be stopped by gradually
increasing trigger ow until it exceeds leak ow (11). Low-ow trigger
settings can also be activated by uctuations in background ow
caused by agitation of circuit condensate or strong cardiac contractions
(12). Again, resultant auto-triggering can be remedied by correcting the
problem and, if necessary, setting a higher trigger ow.
Prior to Patient Inspiratory Effort

Inspiratory 5L
pm
Flow Sensor
VENTILATOR Patient
Expiratory (Zero Flow)
m
Flow Sensor 5 Lp
At onset of Patient Inspiratory Effort

Inspiratory 5 Lp
m
Flow Sensor
VENTILATOR Patient
(2 Lpm
Expiratory Trigger Flow)
Flow Sensor m
3 Lp
Figure 3 Conceptual depiction of ow triggering.
B. Limit Phase
The limit phase denes the limit to which a user-set inspiratory control, like ow
or pressure, can rise without terminating inspiration (2).
Volume Control Ventilation

In volume control (VC) ventilation (VCV), a maximum (or peak) ow is reached
and a xed-pattern ow waveform is sustained until the user-set tidal volume is
delivered. The ow waveform set in VCV is usually a xed-pattern square
ow wave or a xed-pattern descending ow wave. When compared with the
square ow wave at the same peak ow setting, use of the xed-pattern descending
ow wave necessarily increases inspiratory time to preserve the area under the ow
curve and the constancy of the user-set Vt. In sedated patients on VCV, the use of
the xed-pattern descending ow wave has been reported to result in higher mean
airway pressures (MAWP) and better oxygenation at lower peak inspiratory
pressures (PIP) (13). However, in actively breathing patients who trigger high
mechanical respiratory rates, conversion from the square ow wave to the longer
xed-pattern descending ow wave increases risks of ow starvation and air trap-
ping (Fig. 4). A major problem in VCV is matching user-set Vt, peak ow and
xed-pattern ow wave settings with breath-to-breath changes in the inspiratory
demand of the actively breathing, tachypneic patient (13,14).
Pressure Control Ventilation

In pressure control (PC) ventilation (PCV), a constant pressure is sustained until
the user-set inspiratory time (I-Time) elapses. Because ow rates generated
during PCV can rise quickly to peak ows sometimes exceeding 100 lpm,
initial proximal airway pressures can spike above set control pressures (15). To
120
Square Wave
Decelerating Wave
60
.
VL/min 0
Air Trapping from
60 prolongation of
I-Time when
Decelerating Flow
120 curve used.
120
Scooped Pressure Curve
due to inadequate flow output
PcmH2O 60 relative to patient demand
Time (sec)
Figure 4 Volume control: square ow versus decelerating ow. Source: Adapted from
Chang DW. Clinical Application of Mechanical Ventilation. 2nd ed. Delmar Thomson
Learning, 2001; 11:304 (Fig. 11.29).
minimize pressure spiking, to stabilize set control pressures, and to properly

reference high-pressure alarm settings to set control pressures, some ventilator
manufacturers include a Rise Time control. The Rise Time control modies
the initial ow output from the ventilator that determines the rate of rise to set
control pressure (16). Higher initial ows result in rapid rise to control pressure,
while lower initial ows prolong rise to control pressure. An appropriate Rise
Time setting is one in which initial ow output is high enough to meet
patient demand (17), but slow enough to eliminate pressure spiking (Fig. 5).
The decreasing ow prole that follows from peak ow is characteristic of
PC breaths and varies with patient demand and lung mechanics. Thus, PC breaths
may better match breath-to-breath changes in patient demand than the constant Vt
peak ow and xed-pattern ow wave settings of VCV (14,18). Further, the
decreasing ow prole of PC breaths minimizes the resistive components of
proximal airway pressure, allowing for same tidal volume ventilation at
lower PIPs than associated with VC square ow wave breaths. However,
during VC of the acute respiratory distress syndrome (ARDS) patient, a high
alveolar plateau pressure (Pplat . 32 cmH2O), not a high PIP, is the risk factor
associated with ventilator-induced lung injury (19,20). Plateau pressure, in
turn, is measured under conditions of no ow, where airway resistance pressure
is not a factor, and equals the sum of the alveolar pressure caused by tidal volume
80 Rapid Rise Time Slower Rise Time

Pressure Spike
60 Control Pressure
PcmH2O 40
20
200
150
100
50
.
VL/min 0
50
100
150
200 Time (sec)
Figure 5 Effect of rise time control on pressure control breath. Computer rendering of
sketch drawn using common test lung, resistor, and Nellcor Puritan Bennett 840 Ventilator
graphics. Source: Courtesy of Tyco Health Care, Nellcor Puritan Bennett, 2004.
stretch alveolar (PEEP) (Fig. 6). Thus, Vt and PEEP combine to determine
alveolar stretch and related lung damage risks. While it remains controversial
as to which combination of Vt and PEEP provides the best pulmonary gas
exchange with the least risk of ventilator-induced lung injury (21), the pressure
generated by the sum of the set PC pressure and PEEP should be maintained at
no more than 30 35 cmH2O whenever feasible (21 23).
C. Cycle Phase
The cycle phase determines the inspiratory variable that terminates inspiration (2).
Primary Cycling Mechanisms

There are four primary cycling mechanisms (24): pressure cycling (as when an
intermittent positive pressure breath terminates inspiration upon reaching a
user-set pressure); volume cycling (as when a volume control breath terminates
inspiration upon delivery of a user-set tidal volume); time cycling (as when a PC
breath terminates inspiration upon lapse of a user-set I-Time); Flow cycling
[as when a pressure support (PS) breath terminates inspiration upon reaching a
factory- or user-set decreasing ow threshold].
120
. 60
VL/min
40 Zero Flow
VT
0
60
PIP = 40 cm H2O 2 Unit Lung Model
PcmH2O
40
Pplat =28cm H2O Raw = (PIP - Pplat ) / Flow
PEEP = C = V / (P -PEEP) VT + PEEP =>
20 5cm H2O ST T plat
PA ~ Pplat
Time (sec)
Figure 6 Conceptual relationship between peak inspiratory pressure and plateau

pressure estimate of peak alveolar pressure in a volume control machine breath.
Secondary Cycling Mechanisms

Secondary cycling mechanisms are often used as alarm thresholds or safety limits
to terminate inspiration when primary intended cycling mechanisms fail (2). For
example, the high-pressure alarm level set during VCV may be activated and ter-
minate inspiration when airway resistance (RAW) and/or static lung thorax com-
pliance (CST) worsen acutely. When high-pressure alarm cycling occurs, the
undelivered portion of the set tidal volume is vented to atmosphere, often
leading to air hunger, tachypnea, and ventilator patient dyssynchrony. Similarly,
delivered Vt may fall when a normally ow-cycled PS breath is pressure-cycled
off at a few cmH2O above set control pressure because the patient coughs,
sneezes, or tenses the abdomen during a PS breath (24). Flow-cycled PS
breaths will also time-cycle off after three to ve seconds if a circuit leak pre-
vents inspiratory ow from decreasing to the terminal ow-cycling threshold
(24). Some ventilators offer a terminal ow control that enables the user to
ensure ow-cycling during pressure support ventilation (PSV) by increasing
terminal ow-cycling thresholds above leak ows (24) (Fig. 7). An adjustable
terminal ow control also enables the clinician to better synchronize ow-
cycling of PS breaths in chronic obstructive pulmonary disease patients who,
because of long pulmonary lling times, often display expiratory muscle activity
before PS ow can decrease to low ow-cycling thresholds (7).
D. Expiratory Phase
The expiratory phase denes the time it takes for expiratory ow to return to
baseline ow (i.e., active expiration) plus the time in which the variable
known as PEEP or continuous positive airway pressure (CPAP) is applied
25
20 18 cm H2O PS
PcmH2O 15
10
80
Max Flow
60 50% Terminal
(60 Lpm) 25% Terminal
40 Flow (15 Lpm) Flow (30 Lpm)
. 20
VL/min 0
20
40
60
80 Time (sec)
Figure 7 Effect of terminal ow control on pressure support breath. Computer render-

ing of sketch drawn using common test lung, resistor, and Nellcor Puritan Bennett 840
Ventilator graphics. Note: Area under ow curve and therefore V1 decrease as terminal
ow cycling threshold is increased. Source: Courtesy of Tyco Health Care, Nellcor
Puritan Bennett, 2004.
(i.e., passive expiration). PEEP is an acronym used to label above-ambient base-

line pressures set in ventilator modes in which machine breaths appear. CPAP is
an acronym used to label above-ambient baseline pressures set in modes in which
only spontaneous breaths appear. Both PEEP and CPAP are used to keep margin-
ally inated air sacs from collapsing between mechanical or spontaneous breaths,
thereby reducing pulmonary shunt and improving oxygenation. Systemic PaO2s
thus rise and enable clinicians to decrease high FIO2s (.0.5) that are commonly
believed to predispose to cellular lung damage (i.e., O2 toxicity) (25).
However, improved PaO2s and a decrease in O2 toxicity risk are not the
only goals of PEEP therapy. Recent evidence suggests that PEEP and low-tidal
volumes should be used in acute lung injury and ARDS as part of a lung protec-
tive strategy to maximize alveolar recruitment and minimize uneven tidal volume
distribution, focal hyperination, interalveolar shear stress, and resultant lung
damage (26,27). Within the past few years, three approaches to applying lung
protective levels of PEEP have emerged.
FIO2: PEEP Relationship

The rst approach uses the FIO2:PEEP relationships (Table 1) that were used to
achieve a PaO2 of 55 to 80 mmHg (or an SpO2 88 95%) in the ARDS Network
Study (28,29).
Table 1 PEEP Values
FO2 0.3 0.4 0.4 0.5 0.6 0.7 0.7 0.7 0.8 0.9 0.9 1.0
PEEP 5 5 8 8 10 10 12 14 14 14 16 20 24
Source: From Ref. 29.
In ARDS patients, this PEEP strategy, when used in conjunction with a

tidal volume of 6 cc/kg and Pplat 30 cmH2O, resulted in patient outcome
data that demonstrated less days on the ventilator and improved survival.
Pressure Volume Curve

The second approach to setting PEEP uses the patients static pressure volume
(P V) curve to identify a low inection point (LIP) and a high inection point
(HIP) that correspond, respectively, to the most effective and safest levels of
PEEP and tidal volume (30). The LIP is a slight concavity that appears on the
initial portion of the inspiratory static P V curve. The point just beyond this
concave zone represents the critical opening pressure at which a majority of
alveoli are recruited and, therefore, the approximate x-axis value at which
PEEP should be set (31,32). The straight segment of the curve that follows the
LIP is the zone of maximum compliance and may give rise to a convex zone
or HIP near the end of the inspiratory P V curve that indicates an excessive
tidal volume (30). If excessive, the Vt setting of the ventilator should be decreased
to the y-axis value on the static inspiratory P V curve that just precedes the HIP
zone (Fig. 8).
Problems with this inection point PEEP strategy include: having to sedate
the patient so that long low-ow (non-resistive) static P V curves can be gener-
ated; determining when to make these P V curve measurements with respect to
rapidly changing disease status; reconciling differences in inection point
locations that appear on inspiratory versus expiratory portions of the static
P V curve; and setting PEEP when the LIP is absent (33).
Open Lung Approach

The third approach to determining how PEEP should be set is often referred to as
the open lung approach. Proponents of this PEEP strategy recommend using a
sustained alveolar recruitment maneuver in which 30 50 cmH2O CPAP (i.e., no
machine-generated breaths) is applied over 20 90 seconds to recruit atelectatic
alveoli (34,35), then decreased to a level that preserves oxygenation without
hemodynamic compromise [e.g., 15 cmH2O PEEP (35) or LIP 2 cmH2O (36)
with low Vt mechanical ventilation]. It is noteworthy that these alveolar recruit-
ment maneuvers used in the management of ARDS (albeit temporarily)
employed CPAP levels in excess of the 30 cmH2O plateau pressure limit used
in the ARDS Network Study. Additional concerns with the open lung approach
50 Lpm P-V Curve

Vol ml
10 Lpm P-V Curve
Pplat (22cmH2O)
500 Expiration PIP=32 cmH2O
400
HIP 300 HIP (300ml)
200
Inspiration
100
LIP (2 cmH2O) Pcirc cmH2O

-20 -10 10 20 30 40 50
LIP
Figure 8 Quasi-static pressure volume curve at 10 Lpm ow using test lung and resis-
tor. Computer rendering of sketch drawn using common test lung, resistor, and Nellcor
Puritan Bennett 840 Ventilator graphics. Source: Courtesy of Tyco Health Care.
include a limited number of randomized human studies, inconsistency of method,

limited efcacy in consolidative ARDS, and mixed conclusions as to patient
safety and outcomes (37).
IV. Breath Types
There are two types of breaths possible during mechanical ventilation, the
machine breath and the spontaneous breath.
A. Machine Breath
The machine breath is a breath in which inspiration is time- or patient-triggered
and cycles into expiration when a user-set Vt or I-Time is delivered by the
machine. Examples of machine breath are discussed subsequently.
Volume Control Ventilation

In VC ventilation, the VC breath is considered a machine breath because inspiration
is either time- or patient-triggered on and cycles off after the machine delivery
of a user-set tidal volume at a user-set ow or I-Time. Pressure, the dependent vari-
able in VCV, rises as lung mechanics worsen or ventilator circuit resistance
increases and falls as lung mechanics improve or ventilator circuit resistance

decreases.
Pressure Control Ventilation

In PCV, the PC breath is considered a machine breath because inspiration is
either time- or patient-triggered on and cycles off after the machine delivery
of a user-set control pressure sustained over a user-set I-Time interval. Flow, the
dependent variable, decelerates as alveolar-lling pressures approach set control
pressure as a function of patient demand and lung mechanics. However, in PC
ventilation, the user-set I-Time, not the patient, determines when inspiration
cycles to expiration. Thus, too short an I-Time setting can cut off inspiratory
ow before alveolar lling is complete, causing delivered tidal volume to fall.
Or, too long an I-Time setting can exceed alveolar-lling time, resulting in a
period of no ow or a volume hold (Fig. 9). This volume hold increases
mean airway pressure and may result in improved PaO2 (38). However, a
decrease in cardiac output from increasing I-Time settings and mean airway
pressures can result in decreased oxygen transport (38,39) that may offset
increases in PaO2. A prolonged I-Time setting that persists into neural expira-
tion also leads to patient agitation (7) and the need for patient sedation.
600
500
400 V
O
Vtml 300 L
H
200 Vt Vt O
L
100
Insp. Exp. D
80
60
40
V
tc
20 ut
. of
f
Vt hold*
VL/min 0
I-Time I-Time
20
too short too long,
40 for complete now
60 filling exceeding
fill time
80 TIME
Figure 9 Effect of I-Times setting on Vt delivery during pressure control ventilation.

Note: too long an I-Time may also result in air trapping and auto-PEEP if resultant alveo-
lar emptying time is shorter than alveolar lling time.
Bi-Level Ventilation
In bi-level ventilation (Nellcor-PB 840), like PC ventilation, the breath deliv-
ered is considered a machine breath in which inspiratory ow is either time- or
patient-triggered on and cycled off after the machine delivers a sustained,
pressure-limited breath over a user-set I-Time interval. In the absence of spon-
taneous breathing, the ow and pressure waveform graphics generated during a
bi-level machine breath resemble those generated by a time-cycled PC breath.
The bi-level machine breath differs from a PC breath in that the bi-level inspira-
tory pressure limit is labeled as the High PEEP from which the Low PEEP is
subtracted to determine the ventilating pressure (DP). For example, if the high
PEEP is set at 30 cmH2O and the low PEEP is set at 5 cmH2O, then the
maximum pressure reached is 30 cmH2O with a ventilating pressure (DP) of
25 cmH2O. In PC ventilation, the user-set inspiratory pressure limit is the venti-
lating pressure (DP) and is added to set PEEP to determine PIP. For example, if
the PC pressure limit is set at 30 cmH2O and the PEEP at 5 cmH2O, then the
maximum pressure reached is 35 cmH2O with a ventilating pressure (DP) of
30 cmH2O. The bi-level machine breath is further differentiated from the PC
breath in that it allows for spontaneous breathing atop the user-set inspiratory
pressure time curve; a decrease in sedation is an alleged advantage (40). PS is
also available to augment spontaneous breathing on both inspiratory and expira-
tory phases of the bi-level breath (41) (Fig. 10).
50 Inspiratory Effort
Expiratory
40 Effort
PcmH2O
30 HI PEEP
(30 cmH2O)
20 P=25cmH2O PS (15 cm H2O)
Lo PEEP
10 (5 cm H2O)
SET IT SET ET
80
60
. 40
VL/min
20
0
20
40
60
80
SET IT SET ET
Figure 10 Bi-Level machine breath with spontaneous breathing atop and between
machine breaths. Computer rendering of sketch drawn using common test lung, resistor,
and Nellcor Puritan Bennett 840 Ventilator graphics. Source: Courtesy of Tyco Helath
Care, Nellcor Puritan Bennett, 2004.
Airway Pressure-Release Ventilation

In Airway Pressure-Release Ventilation (APRV), the breath delivered is a
variation on the bi-level machine breath in which the user-set inspiratory pressure
limit (High PEEP) is sustained over a user-set I-Time that exceeds expiratory time.
Using more familiar terminology, APRV is pressure controlled inverse ratio ven-
tilation (PCIRV), with the added feature that the patient is able to breathe spon-
taneously atop the user-set inspiratory pressure time curve (42). Spontaneous
breathing atop the APRV inspiratory pressure curve may enhance hemodynamic
function and reduce the need for sedation when compared with PCIRV (43,44).
Pressure-Regulated Volume Control Ventilation
In Pressure-Regulated Volume Control Ventilation (PRVC; Maquet Siemens

Servo 300, Maquet GmbH & Co Kg, Rastatt, Germany), like PC ventilation,
the breath delivered is considered a machine breath in which inspiration is
either time- or patient-triggered on and cycled off after the machine delivers
a sustained pressure-limited breath over a user-set I-Time interval. Inspiratory
ow is decelerating and variable as a function of lung mechanics and patient
demand within the user-set I-Time window. However, unlike PC ventilation, in
PRVC ventilation the control pressure limit is not user-set, but instead is auto-
regulated in small increments or decrements between machine breaths, and in
response to changing lung mechanics, to achieve a user-set tidal volume at a
minimum control pressure (45).
Adaptive Support Ventilation
In Adaptive Support Ventilation (ASV; Hamilton Medical Galileo, Hamilton

Medical AG, Rhazuns, Switzerland), the breath delivered is a variation on
PRVC in which each machine breath is pressure-limited, with the pressure
limit and RR automatically adjusted between breaths, in response to changing
lung mechanics to achieve a user-set minute Ventilation (VE) (46). While the
targeted VE can be achieved with a variety of Vt RR combinations, ASV
breaths allegedly provide the safest combinations by imposing program Vt
boundaries to prevent over- or under-ination and RR boundaries to prevent bra-
dypnea or tachypnea and breath stacking (46).
Volume-Assured Pressure Support Ventilation

In Volume-Assured Pressure Support Ventilation (VAPS; VIASYS Healthcare,
Bird VIP GOLD), PC changes to VC within the same machine breath, rather
than over a series of breaths (2). In the VAPS breath, a low constant ow is
user-set along with a Vt target. The inspiratory ow event begins as a patient-
triggered or time-triggered PC breath with the pressure limit sustained as ow
decelerates towards a ow-cycling threshold set just above the constant ow. If
the set Vt is not delivered when the PC ow-cycling threshold is reached, then
the breath switches to a VC breath in which the pre-set, low level, constant ow is
delivered (and the PC pressure limit overridden) until the target Vt is reached or (in
the event of leak) a VAPS time-limit elapses (46). If the targeted Vt is delivered
before PC ow decays to the constant set ow of VC, then inspiratory ow is
ow-cycled off like a spontaneous PS breath (Fig. 11). However, unlike a spon-
taneous PS breath which must be patient-triggered, subsequent VAPS breaths may
be time-triggered as a function of a user-set RR.
B. Spontaneous Breath
The spontaneous breath is a breath in which inspiration is always patient-

triggered and cycles into expiration as a function of patient demand and lung
mechanics, not machine control settings. Examples of ventilator breaths
considered to be spontaneous in nature follow.
Pressure Support Breath

A PS breath is considered a spontaneous breath in which inspiratory ow is
patient triggered on, decelerates at a rate determined by patient demand and
lung mechanics, and cycles off when alveolar lling is nearly complete (i.e.,
when a low ow-cycling threshold is reached). As with normal spontaneous
PS Limit Override
25
20
Pressure
15 Support
PcmH2O
Limit
10
5
0
45 Vt on Target Vt below Target
Flow-cycled Constant flow
.
VL/min 0
-45
Time (sec)
Figure 11 Effect of failure to meet Vf target in VAPS breath. Computer rendering of

sketch drawn using common test lung, resistor, and Viasys Health Systems, VIP Gold
Ventilator graphics. Source: Courtesy of Viasys Health Care Inc., 2004.
breathing, during PS, inspiratory ow, inspiratory time, and inspiratory tidal
volume vary breath-to-breath as a function of proximal airway pressure,
patient demand, airway resistance, and pulmonary compliance (47). During
PSV, only the user-set control pressure and the low terminal ow-cycling
threshold distinguish PS inspiration from true spontaneous inspiration. The
main drawback of PSV is that the set control pressure, in the absence of the
clinician, does not vary in response to acute changes in lung mechanics or
breath-to-breath changes in inspiratory effort that reect changing pulmonary
workload. Moreover, during PSV there is no ventilator back-up respiratory rate
to safeguard against bradypnea or apnea (47). Setting ventilator alarms to indicate
inadequacy or excess in monitored volumes and respiratory rate is crucial to safe
patient management during PSV.
Volume Support Breath

A volume support (VS) breath (Maquet Siemens Servo 300, Maquet GmbH & Co
Kg, Rastatt, Germany) is a PS breath in which inspiratory ow is patient-
triggered on, decelerates at a rate determined by patient demand and lung
mechanics, and cycles off when a low, ow-cycling threshold is reached.
Unlike a PS breath, in which the inspiratory control pressure is user-set and con-
stant, in VS the inspiratory control pressure is auto-regulated over several breaths
to achieve a user-set Vt. As lung mechanics improve during VS, PS levels needed
to maintain the target Vt are down-regulated, which, according to the manufac-
turer, should facilitate weaning (45). When the Automode control option is
set to On, if the patient does not trigger a VS breath within a user-set time
frame (12 seconds for adults, 8 seconds for pediatrics, and 5 seconds for new-
borns), then the ventilator will switch back to the applicable, user-set machine
breath mode: VC or PRVC (45). When the patient triggers two consecutive
breaths in the applicable machine breath mode, the ventilator switches back to
spontaneous VS breaths (45). Automode also enables the ventilator to switch
from PS to PC breaths based on the same age-specic time intervals (45).
Resumption of spontaneous PS breathing occurs when the patient triggers two
consecutive PC breaths. Concern has been expressed about the risk of deleterious
changes in mean airway pressure that may occur as breath types uctuate
between time-cycled machine breaths and ow-cycled spontaneous breaths
during Automode application (46).
Tube Compensation Breath
A Tube Compensation breath (TC; Nellcor-Puritan Bennett 840) is another

form of auto-adjusting PS. In TC, the PS level auto-adjusts to overcome the
resistive work of breathing imposed by the articial airway (48). In the
patient-triggered TC breath, the PS compensates for the ow-dependent pressure
drop across an articial airway of a known internal diameter. As a patients
inspiratory ow increases through a tube of a known diameter, the PS level is
up-regulated to match the increased pressure drop across the airway (PS DP
RAW Flow). Conversely, as a patients inspiratory ow decreases, the PS level
is down-regulated to match the decreased pressure drop across the articial
airway. A second control option offered in TC breathing enables the user to set
fractional portions of the calculated TC PS level, thus gradually increasing the
patients resistive work of breathing (48). Patients unable to tolerate these incre-
mental increases in resistive work due to the articial airway may be weanable
only if the airway is removed. If the airway is not removable, then the possibility
of ventilator dependence becomes a consideration (49,50).
Proportional Assist Ventilation Breath

A Proportional Assist Ventilation breath (PAV; available in Europe on the Drager
Evita 4, Drager Medical Inc, Telford, PA, U.S.A.) is yet another form of
auto-adjusting PS. In a PAV breath, the PS level varies with patient inspiratory
effort to offset the resistive load imposed by the patients airways [DPRaw (PIP 2
Pplat) RAW Flow], plus the elastic load imposed by the patients alveoli and
chest wall [DPCST (Pplat 2 PEEP) Vt 1/CST] (51). The greater the patient
effort, the higher the PS level generated by the ventilator. In the decelerating ow
prole characteristic of a PAV PS breath, the high initial ow generates signicant
airway resistance that can be offset by setting the ventilator control to deliver a percen-
tage of the measured pressure drop (DPRAW) across the airway. In the lower portion
of the decelerating ow prole, where airway resistance is minimal and increasing
elastic forces reect increasing volume delivery, the ventilator control can be indepen-
dently set to deliver a percentage of the pressure (DCST) generated during tidal
stretch of the alveolar compartment and chest wall (51). Setting PAV elastance
and airway resistance control pressures requires a practical and reproducible
bedside measurement of the spontaneously breathing patients respiratory system
elastance and airway resistance, a measurement task as yet not standardized and com-
plicated by circuit leak that the ventilator interprets as increased patient effort (51).
C. Closed-Loop Control Systems

Both machine and spontaneous breaths may be further described in terms of
closed-loop control systems in which ventilator output variables are automati-
cally adjusted in response to physiologic measurements to achieve inputted,
breath targets. A review of some of the more commonly used closed-loop,
breath types follows.
In the volume-controlled, ow-limited machine breath of VCV, tidal volume
output from the ventilator auto-adjusts in response to compressible volume loss
to maintain a user-inputted target tidal volume. Compressible volume loss is a
ventilator-calculated variable in which a pre-determined circuit stretch and gas
compression factor is multiplied by proximal airway pressures that increase or
decrease as lung mechanics worsen or improve. Thus, as lateral tubing stretch
increases at higher circuit pressures, patient tidal volume would fall if this
compressible volume loss were not automatically offset by an increase in ventilator

output.
In the pressure-controlled, time-cycled machine breath of PCV, the clini-
cian inputs a pressure target, and ow output from the ventilator auto-adjusts
in response to patient demand and lung mechanics to sustain the target pressure
until a user set I-Time elapses.
In the PRVC machine breath, a more complex closed-loop control system
combines the main advantage of PCV (a decelerating ow that varies with
patient demand and lung mechanics) with the main advantage of VCV (delivery
of consistent tidal volumes). In PRVC, the clinician sets a minute volume and res-
piratory rate combination that results in input of a target tidal volume. At the onset
of PRVC, the ventilator imposes a pressure-controlled test breath in which venti-
lator ow output peaks and decelerates as a function of patient demand and lung
mechanics. The output ow signal over time translates into an output tidal
volume that is compared with the user inputted target tidal volume. The PC
level is then automatically regulated in small increments or decrements between
breaths to achieve the targeted tidal volume at minimal control pressures. ASV
is a variation on PRVC that also involves complex closed-loops designed to
achieve a targeted VE using output variables that minimize inspiratory work
while limiting RR and Vt combinations to lung protective levels of support.
Spontaneous breath types such as PS and VS incorporate closed-loop control
mechanisms similar to their machine breath counterparts, PCV and PRVC.
The complexity of closed-loop control systems that typify ventilator
breaths increases as multiple target inputs are achieved with multiple ventilator
output variations, based on multiple physiologic measurements. The goal
driving design of these increasingly complex closed-loop control systems is
the development of computer-based auto-adjusting ventilator breaths that use
rapid measurement and feedback of pulmonary function variables to achieve as
comfortable, safe, and brief a time on mechanical ventilation as possible. Yet iro-
nically, rapid liberation of the adult patient from mechanical ventilation currently
appears to depend less on computer-driven automatic adjustment of breath types
and breath patterns that minimize the human element, than it does on clinician
driven weaning protocols that use best evidence, weaning readiness criteria,
and a once-daily spontaneous breathing trial (59).
D. Breath Types, Vendors, and Special Features

Breath types, vendors and special features are shown in Table 2.
V. Breath Patterns or Modes
Machine and spontaneous breath types, as described earlier, are sequenced into
three traditional breathing patterns or modes of mechanical ventilation:
Assist/Control (A/C), intermittent mandatory ventilation (IMV), and CPAP.
(Text continues on p. 98.)
Table 2 Breath Types and Special Features

Breath typea Availabilitya Special features
Volume control Ubiquitous Machine breath type in which set Vt and ow remain constant while pressure varies in
ventilation response to changing lung mechanics and patient demand. For better control over delivered
Vt, many current generation ventilators auto-compensate for volume loss due to gas
compression and circuit stretch during the VCV breath. Patientventilator dyssynchrony
can occur when patient ow and Vt demands exceed ventilator ow and Vt control settings.
Pressure control Ubiquitous Machine breath type in which control pressure remains constant while ow and Vt vary as a
ventilation function of patient demand and lung mechanics. Tends to better match the ventilatory
pattern of patients with high, variable inspiratory demands than the VCV breath, but at the
risk of excessive Vt delivery. Conversely, acute deterioration in CST and/or RAW may
result in inadequate Vt delivery. Often valued for decreasing ow prole that minimizes
the resistive component of peak airway pressure, thus enabling same tidal volume
ventilation at lower PIPs than typically associated with VC square ow wave breaths.
Pressure control Ubiquitous A PCV breath in which the I-time is set to exceed E-time, thus inverting the physiologic I/E
inverse ratio ratio for purposes of increasing volume hold, mean airway pressure, and PaO2. As set
ventilation I-time exceeds E-time, risks of auto-PEEP and hemodynamic compromise may offset any
improvements in PaO2. Moreover, as set I-time encroaches on neural expiration, patient
agitation and the need for sedation increase.
Bi-level Nellcor Puritan Bennett 840; Resembles a PCV machine breath when spontaneous breathing is absent. But, is more like a
ventilation Drager Evita Series (i.e., distinct ventilator mode during spontaneous breathing, as an active exhalation valve
Evita, Evita 2 Dura, and enables the patient to breathe in an unrestricted manner atop the user-set, inspiratory
Evita 4) pressuretime curve. A decrease in sedation is a manufacturer-alleged advantage. PS is also
available to augment spontaneous breathing atop and between Bilevel machine breaths.
(Continued)
95
Table 2 Breath Types and Special Features (Continued)
96
Breath typea Availabilitya Special features
Airway pressure Nellcor Puritan Bennett 840; Is a variation on the Bi-level breath in which I-time is set to exceed E-time, resulting in a
release Drager Evita Series (i.e., PCIRV breath atop which the patient is able to breath spontaneously. Like PCIRV, APRV
ventilation Evita, Evita 2 Dura, Evita is used to improve oxygenation, but unrestricted spontaneous breathing atop the high
4), Hamilton Galileo Gold Control Pressure time curve is alleged to reduce the patient-ventilator dyssynchrony and
the high cardio-inhibitory mean airway pressures associated with the conventional
PCIRV breath. PS is also available to augment spontaneous breathing atop and between
APRV machine breaths.
Pressure Maquet Siemens Servo 300; The PRVC machine breath is a PCV breath in which the control pressure is automatically
regulated Maquet Siemens Servo i, regulated between breaths in small increments or decrements and in response to changing
volume Adaptive Pressure lung mechanics to achieve a target tidal volume. Thus, the advantage of the PCV breath, a
control Ventilation on the decreasing ow that varies with patient demand and enables same Vt ventilation at lower
Hamilton Galileo; Auto- peak pressures than the VCV square wave breath are combined with the volume guarantee
Flow on the Drager Evita of the VCV breath.
4; VCV on the Nellcor
Puritan Bennett 840
Adaptive Hamilton Galileo A variation on the PRVC machine breath in which the control pressure is up-regulated or
support down-regulated between breaths and in response to changing lung mechanics to achieve a
ventilation targeted VE comprised of multiples of safe RR and Vt combinations. Program breath
Richards and Mosenifar

boundaries prevent over- or under-ination and tachypnea or bradypnea from occurring
during VE targeted automatic control pressure adjustments.
Volume assured Viasys Bird 8400Sti; Viasys Intended for use on patients with an intact respiratory drive, the VAPS breath combines the
pressure T-Bird AVS III; Viasys high initial ow and ow decay pattern of a PS breath with, if necessary to achieve the
support Bird VIP Gold; Pressure targeted Vt, the constant ow of a VCV breath. If the target Vt is not reached as the PS
Augment on Viasys Bear breath decreases to its ow-cycling threshold, then the breath switches to a volume
1000 control breath in which a low constant ow is delivered until the target Vt is reached or a
safety time-limit elapses. While intended for patients capable of triggering the ventilator,
a user-set RR technically qualies the VAPS breath as a machine breath.
Pressure support Ubiquitous The PSV breath is considered a spontaneous breath because it is always patient-triggered, never
ventilation time-triggered. PSV control pressure is user-set and sustained while ow decreases as alveolar
lling pressures approach set control pressures as a function of patient demand and lung
mechanics. PSV is often used during the weaning process to offset the work of spontaneous
breathing through restrictive articial airways. However, a xed level of PS, in the absence of
the clinician, does not vary in response to acute changes in lung mechanics or breath to
breath changes in inspiratory effort. Thus, consistent Vt delivery cannot be guaranteed.
Volume support Maquet Siemens Servo 300; A variation on the PSV breath in which the inspiratory control pressure is not user-set, but
ventilation Maquet Siemens Servo i; instead is auto-regulated over several breaths and in response to changing lung mechanics
Nellcor Puritan Bennett 840. and patient demand to achieve a target Vt. Manufacturer literature suggests that weaning
should be expedited as the pressure support levels used to maintain the target Vt are
automatically down-regulated in response to improving lung mechanics.
Tube Nellcor Puritan Bennett 840; An auto-adjusting PSV breath in which control pressures automatically increase or decrease
compensation Automatic Tube to overcome ow-dependent resistive work of breathing through an articial airway of a
Compensation on Draeger known diameter. As a patients inspiratory ow demand increases, the PS level is
Evita 4; Tube Resistance up-regulated to match the increased pressure drop across the articial airway. Patient
Compensation on resistive work is thus minimized. However, Vt delivery is not guaranteed and may become
Hamilton Galileo Gold inadequate if the patients inspiratory effort wanes in response to worsening lung thorax
compliance or declining nutritional status.
Proportional Available in Europe on the An auto-adjusting PSV breath in which control pressure varies directly with patient inspiratory
assist Drager Evita 4 as effort to offset the resistive workload imposed by gas ow through the airways and elastic load
ventilation Proportional Pressure generated during tidal stretch of alveoli and chest wall. A Flow Assist control is used to
Support (aka: PPS) offset the pressure change that occurs over the upper portion of the PSV ow curve where
measured resistance is greatest and a Volume Assist control is used to offset the pressure
change that occurs over the lower portion of the PSV curve where airway resistance is
minimal and measured elastance highest. As with the Tubing Compensation breath, Vt
delivery is not guaranteed and may become inadequate if patient effort wanes or excessive if a
ventilator circuit leak is misinterpreted as increased patient effort.
97
a
Breath type brand names and their manufacturers from Pilbeam SP, Mechanical Ventilators: General-Use Devices. Pilbeam SP, Cairo JM, eds. Mosbys Respiratory
Care Equipment, 7th ed., St. Louis: Mosby, 2004; 12:391 661.
A. Assist/Control Ventilation
A/C is a ventilator mode in which the patients entire minute volume is com-
prised of machine breaths delivered in response to small patient-initiated
inspiratory efforts (Assist mode) or at the user-set mechanical respiratory
rate (Control mode). The A/C mode machine breath may be a volume-con-
trolled breath, a pressure-controlled breath, or a pressure-regulated/volume-
targeted breath. However, A/C mode with volume control remains the mode
that physicians most commonly use in the initial ventilator management of
ARDS (23,52). A/C mode has also been used to rest ventilatory muscles in
patients who display signs of intolerance to daily two-hour spontaneous breathing
trials and, when used in this fashion, may result in quicker weaning from mech-
anical ventilation than protocols that gradually reduce levels of ventilator support
(7). When managing patients on the A/C mode of ventilation, the clinician
should remember that the ventilators set RR serves as an apnea safeguard and
should therefore be set at approximately three-fourths the patients assist RR.
It is especially important to set the machine RR only a few breaths below the
assist RR if severe metabolic acidemia is present. Depending on the ventilator,
examples of machine breath types available in the A/C mode of mechanical ven-
tilation include: VC, PC, PRVC, and VAPS.
B. Intermittent Mandatory Ventilation
IMV is a ventilator mode in which the patients minute volume is comprised of

a mix of machine breaths and spontaneous breaths. Most modern ventilators
offer a synchronized variant of IMV called synchronized intermittent mandatory
ventilation (SIMV). In SIMV, machine breath delivery is triggered by a spon-
taneous inspiratory effort that occurs within a synchronization window
immediately prior to a scheduled machine breath. SIMV machine breaths may
be volume-controlled or pressure-controlled. Spontaneous breaths are inter-
spersed between machine breaths and include spontaneous breaths augmented
by constant or auto-adjusting levels of PS (or TC). SIMV is frequently used to
wean patients off the ventilator by gradually decreasing the machine breath
rate as the patient is able to assume progressively more of the work of breathing.
However, when excessive spontaneous ventilatory loads are incurred at low
SIMV frequencies, machine breath delivery does not transiently lessen the
excess spontaneous load (53,54). Thus, the strategy of gradually reducing
SIMV frequencies may contribute to the persistence of respiratory muscle
fatigue in patients not ready to wean and unnecessarily prolong mechanical
assistance in patients who are ready to wean. In recent years, two highly regarded
randomized studies showed that most patients deemed medically ready for
weaning tolerated a two-hour spontaneous breathing trial and were extubated
(55,56). Moreover, in those failing the initial spontaneous breathing trial,
SIMV produced the poorest weaning outcomes, with one study nding the
greatest success with once daily T-piece breathing trials (55) and the other with
progressive PS reductions in CPAP mode (56).
C. Continuous Positive Airway Pressure

CPAP is a ventilator mode in which the patients entire minute volume is comprised
of spontaneous breaths. For this reason, one manufacturer refers to the CPAP mode
as the SPONT mode (Nellcor Puritan Bennett 840, Puritan Benett, Carlsbad, CA,
U.S.A.). The ventilator CPAP mode can be used with ow-triggering as an alter-
native to the traditional T-piece spontaneous breathing trial via large volume
O2/aerosol generator. Advantages of the CPAP mode version of the spontaneous
breathing trial include less labor, as there is no additional equipment to set-up,
and the availability of the ventilators continuous monitoring and alarm systems
to indicate patient intolerance of the spontaneous breathing trial (57). Failure to tol-
erate the CPAP mode spontaneous breathing trial may prompt trials with graded PS
(56), low-level PS (58), or auto-adjusting variants of PS (49,50) in the CPAP mode.
Examples of breath types available in the ventilator CPAP mode or labeled as spon-
taneous breathing modes unto themselves include: PS, VS, TC and PAV. Yet,
despite the spontaneous breath type selected to manage the ventilator weaning
process, liberation of the patient from the mechanical ventilator is best achieved
by consistent application of an evidenced-based weaning protocol in which the
rst and most important step is determination of patient readiness to wean (59).
VI. Fine Tuning the Ventilator Breath
Clinical management objectives embodied in the foregoing text include reduction

in the inspiratory work of breathing and in risk factors associated with ventilator-
induced lung injury. Suggested ventilator patient management guidelines for
achieving these objectives follow.
Use of a ow trigger (e.g., 1 3 lpm) instead of the traditional pressure
trigger (e.g., 1 2 cmH2O) may reduce trigger work, although the impact of
trigger work on total WOBi is small unless auto-PEEP is present. If auto-
PEEP must be overcome by the patients inspiratory effort before the set
trigger point can be reached, then setting a low-level PEEP (e.g., 2 5 cmH2O)
that elevates the set trigger point into the auto-PEEP zone should reduce
trigger work necessary for ventilator demand valve actuation. However, in the
actively breathing ventilator patient, the bulk of inspiratory work is incurred
after demand valve actuation, whenever ventilator ow output falls below
patient inspiratory ow demand. In VCV, the ow demand of the actively breath-
ing ventilator patient can usually be met with a peak ow setting of 40 80 lpm
and a xed-pattern square ow wave. But, should the clinician opt for a xed-
pattern descending ow wave at the same peak ow setting (presumably to
reduce high PIPs), decreasing ow output from the ventilator may fall below
patient demand, as indicated by concave curvature or scooping of the
inspiratory pressure time curve. In patients with high and variable ow

demands, even high peak ow and xed-pattern square ow wave settings
may be insufcient to avoid tell-tail scooping on the inspiratory pressure time
curve along with inspiratory ow time curve sequences that are missing their
expiratory ow curve components. When inspiratory pressure curve scooping
and inspiratory ow curve humping occur together on the VCV graphics
display, ow and tidal volume settings are too low to meet patient ow and
tidal volume demand. A trial of PCV wherein peak ow often tops 80 lpm and
a descending ow pattern conforms to breath-to-breath changes in patient
demand and lung mechanics may reduce inspiratory work and discomfort the
high demand patient experiences in VCV. The availability of a rise time
control to attenuate the initial blast of gas and resultant pressure overshoot
sometimes seen in PCV, along with a terminal ow control that synchronizes ter-
mination of inspiratory ow with the onset of patient expiratory activity within
the PCV I-Time window, provides further means for achieving patient comfort
and inspiratory workload reduction.
However, persistent variability in PCV tidal volume delivery, especially if
in excess of 4 8 mL/kg in heterogeneous lung diseases like ARDS, is associated
with alveolar stretch injury and may require sedation of the high demand patient
and conversion back to volume controlled breaths. The concomitant use of effec-
tive levels of PEEP between breaths in this tidal range prevents atelectasis and
related sheer stress that is also considered damaging to parenchymal tissue, but
controversy exists as to how much PEEP should be used to recruit and prevent
derecruitment of atelectatic alveoli. In managing the ARDS ventilator patient,
the best approach to this controversy seems to lie in the use of the 4 6 mL/kg
tidal volumes and FIO2:PEEP relationships used in the NIH ARDS Network
trial that resulted in reduced ventilator days and mortality. Yet in severe
ARDS, use of even these low-end tidal volumes and FIO2:PEEP combinations
may not yield adequate gas exchange without exceeding the lung protective
target Pplat of 30 35 cmH2O. At this point, high-frequency oscillatory ventilation
(HFOV) has been proposed as a ventilator strategy for managing severe ARDS
that may maximize lung recruitment and oxygenation while minimizing
ventilator-induced stretch injury. While relatively new to adult ventilator man-
agement, HFOV has been used for nearly two decades to manage severe respir-
atory failure in neonates and infants. In HFOV, MAWP is the control determinate
of oxygenation and is initially set at a few centimeters above the conventional
ventilator MAWP. A user-set control pressure (pressure amplitude) is then
oscillated at a high user-set frequency (e.g., 3 5 Hz in adults) above and
below the MAWP baseline to achieve visual chest wall vibration (wiggle
from shoulder to thigh) that determines CO2 clearance while minimizing
cyclic alveolar stretch. HFOV might be thought of as a vibrating CPAP and in
adults represents an evolutionary extension of conventional ventilation lung
protective strategies that aim to minimize alveolar stretch injury while maintain-
ing maximum alveolar recruitment and adequate gas exchange.
Acknowledgment
Computer rendering of waveform sketches by Jihane Bishara, Management

Assistant, Division of Pulmonary/Critical Care Medicine, CSMC.
References
1. Chatburn RL, Primiano FP. A new system for understanding modes of mechanical
ventilation. Respir Care 2001; 46(6):604 621.
2. Chatburn RL, Volsko TA. Mechanical ventilators. In: Wilkins RL, Stoller JK
Scanlan CL, eds. Egans Fundamentals of Respiratory Care. 8th ed., St. Louis:
Mosby 2003; 39:929 962.
3. Dennison F. Basic ventilator wave form analysis. In: Chang DW, ed. Clinical
Application of Mechanical Ventilators. 2nd ed., Delmar Thomson Learning 2001;
11:264 324.
4. Rossi A, Gottfried SB, Zocchi L, Higgs BD, Lennox S, Calverly PM, Begin P,
Grassino A, Milic-Emili J. Measurement of static compliance of the total respiratory
system in patients with acute respiratory failure. The effect of intrinsic positive and
expiratory pressure. Am Rev Respir Dis 1985; 131(5):672 677.
5. MacIntyre NR, Cheng KG, McConnell R. Applied PEEP during pressure support
reduces the inspiratory threshold of intrinsic PEEP. Chest 1997; 111(1):188 193.
6. Leung P, Jubran A, Tobin M. Comparison of assisted ventilator modes on triggering,
patient effort and dyspnea. Am J Respir Crit Care Med 1997; 155(6):1840 1848.
7. Tobin MJ, Jubran A, Laghi F. Patient ventilator interaction. Am J Respir Crit Care
Med 2001; 163(5):1059 1063.
8. Branson RD, Campbell RS, Davis K, Johnson DJ. Comparison of pressure and ow
triggering systems during continuous positive airway pressure. Chest 1994;
106(2):540 544.
9. Giuliani R, Mascia L, Recchia F, Caracciolo A, Fiore T, Ranieri VM. Patient
ventilator interaction during synchronized intermittent mandatory ventilation.
Effects of ow triggering. Am J Respir Crit Care Med 1995; 151(1):1 9.
10. Sager J, Eiger G, Fuchs B. Ventilator auto-triggering in a patient with tuberculous
bronchopleural stula. Respir Care 2003; 48(5):519 521.
11. Nellcor Puritan Bennett, 840 Ventilator System Operators and Technical Reference
Manual No. 4-070088-00 Rev. C, 1998; TR 12:1 5.
12. Imanaka H, Nishimura M, Takeuchi M, Kimball W, Yahagi N, Kumon K. Auto trig-
gering caused by cardiogenic oscillation during ow-triggered mechanical venti-
lation. Crit Care Med 2000; 28(2):402 407.
13. Davis K, Branson RD, Campbell RS, Porembka DT. Comparison of volume control
and pressure control: is ow wareform the difference? J Trauma Injury Infect Critical
Care 1996; 41(5):808 814.
14. MacIntyre NR, McConnell R, Cheng KC, Sane A. Patient ventilator ow dys-
synchrony: ow-limited versus pressure-limited breaths. Crit Care Med 1997;
25(10):1671 1677.
15. Dick CR, Sassoon CS. Patient-ventilator interactions. Clin Chest Med 1996;
17(3):423 438.
16. Chiumello D, Pelosi P, Croci M, Bigatello LM, Gattinoni L. The effects of pressur-
ization rate on breathing pattern, work of breathing gas exchange and patient comfort
in pressure support ventilation. Eur Respir J 2001; 18(1):107 114.
17. Bonmarchand G, Chevron V, Menard J, Girault C, Moritz-Berthelot F, Pasquis P,
Leroy J. Effects of pressure ramp slope values on the work of breathing during
pressure support ventilation in restrictive patients. Crit Care Med 1999;
27(4):715 722.
18. Kallet RH, Campbell AR, Alonso JA, Morabito DJ, Mackersie RC. The effects of
pressure control versus volume control assisted ventilation on patient work of breath-
ing in acute lung injury and acute respiratory distress syndrome. Respir Care 2000;
45(9):1085 1096.
19. Tobin MJ. Culmination of an era of research on the acute respiratory distress
syndrome. N Eng J Med 2000; 342(18):1360 1361.
20. Gattinoni L, Vagginelli F, Chiumello D, Taccone P, Carlesso E. Physiologic ration-
ale for ventilator setting in acute lung injury/acute respiratory distress syndrome
patients. Crit Care Med 2003; 31(4):S300 S304.
21. Petrucci N, Iacovelli W. Ventilation with lower tidal volumes versus traditional tidal
volumes in adults for acute lung injury and acute respiratory distress syndrome.
Cochraine Database Syst Rev 2003; (3):CD003844.
22. Tobin MJ. Medical progress: advances in mechanical ventilation. N Eng J Med 2001;
344(26):1986 1996.
23. Thompson BT, Hayden D, Matthay MA, Brower R, Parsons PE. Clinicians
approaches to mechanical ventilation in acute lung injury and ARDS. Chest 2001;
120:1622 1627.
24. Pilbeam SP. Introduction to ventilators. In: Pilbeam SP, Cairo JM, eds. Mosbys
Respiratory Care Equipment. 7th ed. St. Louis: Mosby 2004; 11:317 390.
25. Carmichael LC, Dorinsky PM, Higgins SB, Bernard GR, Dupont WD, Swindell B,
Wheeler AP. Diagnosis and therapy of acute respiratory distress syndrome in adults:
an international survey. J Critical Care 1996; 11(1):9 18.
26. Slutsky AS. Lung injury caused by mechanical ventilation. Chest 1999;
116:9S 15S.
27. Uhlig S. Mechanotransduction in the lung, ventilation-induced lung injury and
mechanotransduction: stretching it too far? AJP Lung Cell Mol Physiol 2002;
282(5):L892 L896.
28. 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(18):1301 1308.
29. NIH/NHLBI ARDS Clinical Network: Mechanical ventilation protocol summary,
2000: Available from http://hedwig.mgh.havard.edu/ardsnet/6 ml.html/:1 4.
30. Jonson B, Svantesson C. Elastic pressure volume curves: what information do they
convey? Thorax 1999; 54(1):82 87.
31. Amato MBP, Barbas CSV, Medeiros DM, Magaldi RB, Schettino GP, Lorenzi-Filho
G, Kairalla RA, Deheinzelin D, Munon C, Oliveira R et al. Effect of bioprotective
ventilation on mortality in the acute respiratory distress syndrome. N Engl J Med,
1998; 338(6):347 354.
32. Mancini M, Zavala E, Moncebo J, Fernandez C, Babera JA, Rossi A, Roca J,
Rodriguez-Roisin R. Mechanisms of pulmonary gas exchange improvement
during a protective ventilatory strategy in acute respiratory distress syndrome. Am J

33. Saura P, Blanch L. How to set positive end-expiratory pressure. Respir Care 2002;
47(3):279 292.
34. Lapinski SE, Aubin M, Mehta S, Boiteau P, Slutsky AS. Safety and efcacy of a sus-
tained ination for alveolar recruitment in adults with respiratory failure. Intensive
Care Med 1999; 25(11):1297 1301.
35. Richards GA, Hopley M. Successful recruitment. Crit Care Med 2002; 30(9):2169.
36. Takeuchi M, Goddon S, Dolhnikoff M, Shimaoka M, Hess D, Amato M, Kacmarek
R. Set positive end-expiratory pressure during protective ventilation affects lung
injury. Anesthesiology 2002; 97(3):682 692.
37. Hess DR, Bigatello L. Lung recruitment: the role of recruitment maneuvers. Respir
Care 2002; 47(3):308 317.
38. Armstrong BW Jr, MacIntyre NR. Pressure-controlled, inverse ratio ventilation that
avoids air trapping in the adult respiratory distress syndrome. Crit Care Med 1995;
23(2):279 285.
39. Mercat A, Graini L, Teboul JL, Lenique F, Richard C. Cardio-respiratory effects of
controlledpressure ventilation with and without inverse ratio in patients with adult
respiratory distress syndrome. Chest 1993; 104(3):871 875.
40. Rathgeber J, Schorn B, Falk V, Kazmaier S, Spiegel T, Burchardi H. The inuence of
controlled mandatory ventilation (CMV), intermittent mandatory ventilation (IMV)
and biphasic positive airway pressure (BIPAP) on duration of intubation and con-
sumption of analgesics and sedatives. A prospective analysis in 596 patients follow-
ing cardiac surgery. Eur J Anaesthesiol 1997; 14(6):576 582.
41. Nellcor Puritan Bennett. 840 Ventilator System Operators and Technical Reference
Manual. BiLevel Option/800 Series Ventilators Rev. A, 1998; No. 4-075816-
00:1 8.
TM
42. Mallinkrodt, Inc. Two ventilating strategies in one mode BiLevel Rev. A, 1999;
No. A-AAZZ58-00:1 8.
43. Putensen C, Zech S, Wrigge H, Zinserling J, Stuber F, Von Spiegel T, Mutz N. Long-
term effects of spontaneous breathing during ventilatory support in patients with
acute lung injury. Am J Respir Crit Care Med 2001; 164(1):43 49.
44. Hering R, Peters D, Zinserling J, Wrigge H, Von Spiegel T, Putensen C. Effects of spon-
taneous breathing during airway pressure release ventilation on renal perfusion and func-
tion in patients with acute lung injury. Intensive Care Med 2002; 28(10):14261433.
45. Siemens Servo 300/300. A Operating Manual 8.0/9.0. Ventilation modes. 2nd ed.,
1998; No. AG0498 1.5:1 53.
46. Branson RD, Johannigman JA, Campbell RS, Davis K Jr, Closed-loop mechanical
ventilation. Respir Care 2002; 47(4):427 451.
47. Shelledy DC. Discontinuing ventilatory support. In: Wilkins RL, Stoller JK, Scanton
CL, eds. Egans Fundamentals of Respiratory Care, 8th ed., St. Louis: Mosby 2003;
44:1121 1154.
48. Mallinckrodt, Inc. Improving spontaneous breathing-tube compensation. Rev. A,
2003, No. A-AA2289-00:1 3.
49. Guttmann J, Haberthur C, Mols G, Lichtwarck-Aschoff M. Automatic tube
compensation [German with English Abstract]. Minerva Anestesiologica 2002;
68(5):369 377.
50. Haberthur C, Mols G, Elsasser S, Bingisser R, Stocker R, Guttmann J. Extubation

after breathing trials with automatic tube compensation, T-tube or pressure
support ventilation. Acta Anaesthesiol Scand 2002; 46(8):973 979.
51. Ambrosino N, Rossi A. Proportional assist ventilation (PAV): a signicant advance
or a futile struggle between logic and practice? Thorax 2002; 57(3):272 276.
52. Esteban A, Anzueto A, Alia I, Gordo F, Apezteguia C, Palizas SF, Cide D,
Goldwaser R, Soto L, Bugedo G et al. How is mechanical ventilation employed in
the intensive care unit? An International Utilization Review. Am J Respir Crit
Care Med 2000; 161(5):1450 1458.
53. Marini JJ, Smith TC, Lamb VJ. External work output and force generation during
synchronized intermittent mechanical ventilation: effect of machine assistance on
breathing effort. Am Rev Respir Dis 1988; 138(5):1169 1179.
54. Ismand C, Feihl F, Perret C, Fitting JW. Regulation of inspiratory muscle output
during synchronized intermittent mandatory ventilation. Anesthesiology 1994;
80(1):13 22.
55. Esteban A, Frutos F, Tobin MJ, Alia I, Solsona JF, Valverdu I, Fernandez R, De La
Cal MA, Benito S, Thomas R et al. A comparison of four methods of weaning
patients from mechanical ventilation, Spanish Lung Failure Collaborative Group,
N Eng J Med 1995; 335:1864 1869.
56. Brochard L, Rauss A, Benito S, Conti G, Mancebo J, Rekik N, Gasparetto A,
Lemaire F. Comparison of three methods of gradual withdrawal from ventilatory
support during weaning from mechanical ventilation. Am J Respir Crit Care Med
1994; 150(4):896 903.
57. Hess D. Ventilator modes used in weaning. Chest 2001; 120:474S 476S.
58. Frutos-Vivar F, Esteban A. When to wean from a ventilator: an evidence-based strat-
egy. Cleveland Clin J Med 2003; 70(5):389, 392 393, 397.
59. Ely E, Meade M, Haponik E, Kollef M, Cook D, Guyatt G, Stoller JK. Mechanical
ventilator weaning protocols driven by non-physician healthcare professionals-
evidence based clinical practice guidelines. Chest 2001; 120:454S 463S.
4
Monitoring During Mechanical Ventilation
DEAN R. HESS
Department of Respiratory Care

Massachusetts General Hospital and Department of Anesthesia
Harvard Medical School
Boston, Massachusetts, U.S.A.
I. Introduction
Monitoring is the continuous, or nearly continuous, evaluation of the physiologic

function of a patient in real time to guide diagnosis and management decisions,
including when to make therapeutic interventions and assessment of those inter-
ventions (1). Many physiologic parameters can be monitored during mechanical
ventilation (Fig. 1). These include both invasive and non-invasive monitoring.
Although its impact on patient outcomes has not been well studied, monitoring
has become an integral part of the care of mechanically ventilated patients.
II. Arterial Blood Gases
Arterial blood gas (ABG) analysis is commonly regarded as the gold standard
for evaluation of arterial oxygenation and acid-base balance, althought the indi-
cations for ABG remain unclear (2). There are numerous problems associated
with the measurement and use of ABG (Table 1). ABGs are the most frequently
ordered laboratory test in many critical care units. A rate of 4.8 ABG/patient/day
was reported from a surgical ICU, and the presence of an arterial line was most
predictive of the number of ABGs drawn per patient (3). The presence of an arter-
ial line also affects the pattern of transfusion requirements related to phlebotomy
for diagnostic tests in adults (4 7). In one study, patients with an arterial line
were phlebotomized at a rate twice that of patients who did not have a line (5).
Another study reported increases in the number of blood tests (29% increase),
105
106 Hess
Gas exchange Hemodynamics

blood gases heart rate & rhythm
oximetry blood pressure
capnometry cardiac output
transcutaneous wedge pressure
Tissue
oxygen delivery
Oxygenation
Respiratory Monitoring
During Mechanical Ventilation
Ventilator function
Lung mechanics Sedation and disconnect
airway pressure, Neuromuscular leak
flow, volume Blockade occlusion
esophageal pressure humidification
pressure-volume curve FIO2
Figure 1 Many physiologic parameters can be monitored during mechanical

ventilation.
blood-drawing procedures (30% increase), and the amount of blood volume loss
(44% increase) from patients with arterial lines compared with those without (6).
Several strategies might be used to decrease blood loss because of ABGs
(and other laboratory testing) in critically ill patients (7). Routine-use arterial
lines should be avoided. Blood conservation systems can be used to eliminate
the discard volume from arterial line blood draws (8 11). Multiple diagnostic
tests may be performed on a single sample of blood. Clinically accurate and
precise invasive and non-invasive technology can be substituted for some diag-
nostic tests. Blood sample volume can be reduced, such as by using pediatric
phlebotomy tubes and reduced syringe volume (12).
Table 1 Problems Associated with Arterial Blood Gas Analysis
Pain
Pre-analytical errors: air contamination, heparin dilution, storage, delay before analysis
Analytic errors: accuracy differences between models of blood gas analyzers
Post-analytical errors: transcription errors, delays in reporting results
Infection risk to patient (particularly with arterial catheter)
Infection risk to clinician (particularly with arterial puncture)
Thrombosis and distal embolization (particularly with arterial catheter)
Blood loss (particularly with arterial catheter)
Intermittent information
Cost
Regulation: CLIA-88
Monitoring During Mechanical Ventilation 107
Beasley et al. (13) established procedural, clinical, and therapeutic indi-

cators for ABGs in a surgical ICU. A forced interaction between respiratory
therapists and nurses was established by making therapists responsible for
ABGs analysis. The result of this interaction between therapists and nurses
regarding ABGs has resulted in an improvement in the appropriateness of
ABGs requested by the nurses. Pilon et al. (14) developed institutional guidelines
for ABGs and reported a decrease in the number of blood gases by about 50%
with near-doubling of the appropriateness of ABGs. It is important to develop
a behavior among respiratory therapists, physicians, and nurses that carefully
considers the need for blood gas analysis. There is no indication for routine
blood gases (e.g., after each ventilator change).
One frequently unappreciated feature of intermittent ABGs is spontaneous
variability (i.e., changes without any intervention or interaction with the patient)
(15 18). In one report, the coefcient of variation in PaO2 between four ABGs
drawn at 20-minute intervals in stable mechanically ventilated trauma patients,
who were not disturbed during the study period, was 5.1 + 3.2%, median
3.6%, 95th percentile 9.8%) (15). Changes in PaO2 within the limits of spon-
taneous variability, and without an associated clinical change, might be the
result of spontaneous variability.
An important clinical issue relates to the amount of time required to estab-
lish a new steady state after changing the FIO2 provided for the patient. If too
little time is allowed, the blood gas results will not reect the new clinical con-
ditions. If too much time is allowed, this could subject the patient to a long period
of potentially harmful hypoxemia. The time required to reach a steady state after
an FIO2 change depends upon the alveolar ventilation and the functional residual
capacity. A small alveolar ventilation or a large functional residual capacity requires
a longer time for equilibration. Unless the patient has obstructive lung disease
(which results in a lower alveolar ventilation and a larger functional residual
capacity), ve minutes is usually an adequate time to reach a new steady state
after a change in FIO2 (19). For patients with obstructive lung disease,
30 minutes should be allowed to reach a new steady state after a change in FIO2.
The gold standard for measurement of hemoglobin oxygen saturation (O2Hb)
is CO-oximetry. CO-oximetry uses multi-wavelength spectrophotometry to
measure O2Hb, deoxyhemoglobin (HHb), carboxyhemoglobin (COHb), and methe-
moglobin (metHb). Fractional oxyhemoglobin (FO2Hb) is then calculated as:
O2 Hb
FO2 Hb
O2 Hb HHb metHb COHb
Oxygen saturation can also be calculated from the measured PO2 using an empirical
equation for the oxyhemoglobin dissociation curve, as is commonly done using soft-
ware in blood gas analyzers. Clinically important errors can result from use of
calculated oxygen saturation in applications such as the shunt equation (2023).
Several errors are related to CO-oximetry. Owing to differences in the light
108 Hess
absorption spectra of fetal hemoglobin (HbF) and adult hemoglobins, the presence
of HbF may result in false elevations of COHb (24). Correction factors for the pre-
sence of HbF have been published (25), but these may not be appropriate under all
conditions (such as transfusion) (26). Intralipid infusions (27) and propofol infu-
sions may result in falsely elevated metHb levels measured by CO-oximetry.
III. Arterialized Capillary Blood Gases
To obtain capillary blood gas samples, a puncture is made with a lancet or similar
device into the cutaneous layer of the skin at a highly vascularized area. To accel-
erate blood ow and reduce the difference between arterial and venous gas press-
ures, vasodilation is achieved by warming or use of a vasoactive cream. Common
sites used for arterialized capillary blood gases are the heel (in neonates), nger,
toe, or earlobe.
Several studies have reported the use of arterialized capillary blood gases in
adults. Pitkin et al. (28) reported good agreement between capillary PO2 and
arterial PO2 (bias 21 mm Hg, limits of agreement 28 to 6 mm Hg). Dar et al.
(29) reported trivial and non-signicant differences between arterial and capil-
lary PO2. In contrast to these reports, Sauty et al. (30) reported a signicant
underestimation (4 mm Hg) of PaO2 by earlobe capillary blood gases with
wide limits of agreement (24 to 13 mm Hg).
Arterialized capillary blood gases produce a mixture of blood from capil-
laries and venules (31). The difference between arterial and venous PO2 is
usually large (about 60 mmHg), but this difference can be reduced by vasodila-
tion. Furthermore, the difference between arterial and venous PO2 decreases with
hypoxemia. Thus, the accuracy of arterialized capillary blood PO2 improves as
the PaO2 decreases. None the less, arterialized capillary PO2 systematically
underestimates the PaO2. Moreover, capillary blood samples are difcult to
obtain anaerobically. Exposure of the blood to room air raises the PO2, and
this may explain the agreement between arterialized capillary PO2 and PaO2
reported in some studies. Arterialized capillary blood gases have also been advo-
cated because they are less painful, but the pain of arterial puncture can be les-
sened by the use of local anesthesia (32).
IV. Point-of-Care Testing
Samples for blood gas analysis are drawn from the patient, packaged and trans-
ported to the laboratory, and analyzed, and the results are then communicated
back to the patient care unit. This model may result in pre-analytical errors
(e.g., failure to properly transport the sample), post-analytical errors (e.g., tran-
scription errors in reporting the results), and delay in reporting the results.
There has been increasing interest in point-of-care (POC) testing, in which
blood gases are measured at or near the site of patient care (33 40). Several
studies have reported acceptably accurate blood gas results using POC testing
devices (41 45). Kendall et al. (46) performed a randomized controlled trial
of POC on clinical outcome in an emergency department and reported that
POC testing resulted in signicantly faster decision-making. However, POC
testing did not affect the amount of time spent in the emergency department,
the length of stay in the hospital, admission rates, or mortality.
POC testing devices are portable (some are hand-held), and they typically
require a small volume of blood for testing (a few drops). Blood is introduced into
a single-use disposable cartridge that is introduced into the portable analyzer. The
cartridge chosen determines the tests that will be run (e.g., blood gases, electro-
lytes, hematocrit, glucose, BUN, creatinine, ionized calcium, and others). The
sensor calibrates automatically, after which the blood sample is drawn over the
biosensors. The POC analyzer then calculates, displays, and stores the results.
The POC testing device can communicate with the central laboratory or hospital
information system for reporting and archiving results.
The role of POC testing is evolving and is likely to expand in the future.
Quality control and quality assurance must be appropriately addressed. This is
typically achieved by following the manufacturers recommendations for use
of the device. The costs of the base unit the and the test cartridges must be
balanced against a quicker test turn-around time (potentially resulting in more
rapid treatment and better patient outcomes), the lower overhead (compared to
the central laboratory), and the smaller blood volume required for testing (result-
ing in lower transfusion requirements).
V. Blood Gas Monitors
Blood gas monitors measure pH, PCO2, and PO2 without permanently removing
blood from the patient (47). Blood gas monitors use uorescent optodes, which
are optical biosensors. Light is used to activate electrons in a dye to a higher
energy state. When the light exposure ends, the electrons return to their basal
state and light is re-emitted. The lower frequency of re-emitted light is augmented
as the concentration of hydrogen ions or carbon dioxide increases, and the
re-emitted light is quenched as the concentration of oxygen increases. Photo-
sensors are used to quantify the amount of quenching, and a microprocessor is
used to translate the signal into a display of PO2, PCO2, and pH. Various con-
gurations of optodes for intra-arterial blood gas monitors have been developed.
Optode systems are electrically isolated from the patient, they can be miniatur-
ized so that they t through an arterial catheter without affecting other catheter
functions, they are stable over time, they respond rapidly to changes in substrate
concentration, and they are not consumed in the measurement process.
Several approaches can be taken to blood gas monitoring in conjunction
with an arterial catheter. The rst approach uses a probe that passes through
the arterial catheter and resides directly within the arterial lumen (i.e., in vivo
110 Hess
blood gas monitor). With the second approach, the optode system is connected to
the proximal arterial line but does not pass through the catheter. With this
method, when blood gas and pH levels are desired, blood is drawn into a
chamber containing the optodes (i.e., ex vivo blood gas monitor). After analysis,
the blood is ushed back into the artery. Thus, frequent (but not continuous)
blood gas measurements are possible without blood loss. Studies evaluating
these devices (48 64) have generally reported acceptable overall accuracy of
blood gas monitors. Some of these studies, however, report occasional discrepan-
cies between the blood gas monitor and a blood gas analyzer. The reasons for this
may relate to issues such as sensor position and catheter ushing.
One problem with these systems is their cost. Blood gas analyzers are
equally expensive, but they are not dedicated to a single patient. Although manu-
factures are striving to design user-friendly systems that are easy to use, the
amount of technical attention that these systems are likely to require in a busy
ICU remains to be determined. The life of the sensor in a busy ICU also
remains to be seen. The quality control and quality assurance requirements for
clinical and regulatory purposes are unclear. Whether or not the benets of con-
tinuous blood gas and pH measurements will outweigh the costs and technical
support required remains to be seen.
VI. Mixed Venous Blood Gases
Blood obtained from the central venous circulation or pulmonary artery is used to
assess mixed venous oxygenation. If the catheter is positioned in a distal branch
of the pulmonary artery, rapid aspiration can contaminate the mixed venous
blood with pulmonary capillary blood (65). Although the importance of this
has been questioned (66), it is prudent to use a slow aspiration rate (,1 mL/
30 sec) when blood is sampled from a pulmonary artery catheter.
The mixed venous oxygenation has been commonly thought to represent
tissue oxygenation. However, several points are important in relation to the
ability of mixed venous oxygenation to reect tissue oxygenation. SvO2 (or
PvO2) depends on the oxygen content derived from many vascular beds, and it
thus does not necessarily reect well the SvO2 of any individual vascular bed.
With septic shock, peripheral arteriovenous shunts open, resulting in an SvO2
higher than expected (67). SvO2 will also be falsely elevated (i.e., not reect
tissue oxygenation) in patients with ventricular septal defect or cyanide poison-
ing. During circulatory arrest, large concentrations of CO2 may accumulate in the
venous circulation. With cardiac arrest, very high PvCO2 (venous respiratory
acidosis) has been reported in the presence of normal or low PaCO2 (arterial
respiratory alkalosis) (68,69).
SvO2 can be mathematically derived from the Fick equation:
_ 2
VO
Sv O2 1
DO2
where VO2 is oxygen consumption and DO2 is oxygen delivery. The relationship
VO2/DO2 is the oxygen utilization coefcient or the oxygen extraction ratio.
SvO2 is thus determined by the balance between VO2 and DO2.
Mixed venous blood is obtained from the pulmonary artery. Oxygen satur-
ation is slightly higher in the inferior vena cava than the superior vena cava due to
the relatively low levels of oxygen extraction relative to perfusion in the renal and
mesenteric vasculature (70). During periods of stress (e.g., shock, hypovolemia,
exercise), however, there is a drop in the oxygen saturation of the inferior vena
cava, so the saturation in the superior vena cava is greater than that of the inferior
vena cava. This occurs as the result of decreased renal and splanchnic perfusion
during these conditions. Use of central venous blood other than that obtained
from the pulmonary artery may be an inadequate reection of true mixed
venous blood. Although mixed venous blood samples should ideally be obtained
from the pulmonary artery, a high correlation between pulmonary shunt calcu-
lations using pulmonary artery samples and central venous samples (e.g., right
atria or superior vena cava) has been reported (70). A peripheral venous blood
gas reects the PO2 and PCO2 of the tissue bed served by that vein. Thus periph-
eral venous blood gases are not useful as a proxy for either an arterial blood gas
sample or a central venous blood gas sample.
The ability to continuously measure mixed venous oxygen saturation via
oximetric methodology became commercially available in the early 1980s.
This system uses a microprocessor, an optical module with light sources and
photodetectors, and a ow-directed pulmonary artery catheter (71 73). Using
beroptics, wavelengths of light between 650 and 1000 nm are pulsed into the
pulmonary artery. This light is reected off red blood cells in the pulmonary
artery and returned to the optical module via another beroptic bundle. SvO2 is
determined by the ratio of transmitted and reected light. Before insertion, the
system is calibrated with an in vitro calibration standard, and this can be
updated periodically by in vivo calibration using a CO-oximetry determined
SvO2. Factors that affect the measurement of SvO2 using this method include
temperature, pH, blood ow velocity, hematocrit, and occlusion of the catheter
tip (e.g., clot or vessel wall).
A number of studies have evaluated the accuracy of SvO2 catheter systems
(74 81). Most of them reported that the three wavelength systems correlate
better with CO-oximeter SvO2 than two wavelength systems and that drift is
less with three wavelength systems. Studies reporting bias and precision analysis
have found greater degrees of inaccuracy than those claimed by the manufacturer.
However, if meticulous attention is paid to care of the catheter system (e.g.,
in vivo calibrations, updating the hematocrit setting), either of these systems
may be acceptable to detect changes or trends. These systems do require more
technical attention than other monitoring systems, such as pulse oximetry or a
conventional pulmonary artery catheter.
Regardless of their technical performance, the clinical usefulness of SvO2
monitoring remains controversial. The results of a large multi-center study
112 Hess
(.10,000 patients in .50 ICUs) showed that targeting SvO2 . 70% did not
affect mortality (82). More recently, Rivers et al. (83) evaluated the efcacy of
early goal-directed therapy before admission to the ICU. Patients who arrived
at an emergency department with severe sepsis or septic shock were randomized
to receive either six hours of early goal-directed therapy or standard therapy
before admission to the ICU. In-hospital mortality was 30.5% in the group
assigned to early goal-directed therapy, compared with 46.5% in the group
assigned to standard therapy. In this study, a central venous catheter capable of
measuring oxygen saturation was used, with a targeted central venous oxygen
saturation 70%.
VII. Pulse Oximetry
Continuous pulse oximetry has become a standard of care for critically ill
patients. With pulse oximetry, two wavelengths of light (660 and 940 nm) are
passed through a pulsating vascular bed (84,85). This is accomplished using
two light-emitting diodes (LEDs) and a photodetector. There is some error in
the wavelength of light emitted by the LEDs (+30 nm) that can affect accuracy.
Also, the photodetector is not specic (i.e., if will respond to any wavelength of
light, which can result in interference). Although some of the light emitted from
the LEDs is absorbed by each constituent of the tissue, the only variable absorp-
tion is due to arterial pulsations. This is translated into a plethysmographic
waveform. The ratio of the amplitudes of these two plethysmographic waveforms
is translated into oxygen saturation.
A probe is used to pass light from the LEDs through a pulsating vascular bed.
Disposable and reusable probes are available and equally accurate (86). Pulse oxi-
meter probes can become contaminated with pathogenic bacteria and serve as a
source of nosocomial infection. A protective sheath can be used to allow the
same disposable probe to be used on multiple patients (87). The use of disposable
pulse oximeter probes can become very expensive for the hospital, and it has been
shown that disposable probes can be gas sterilized and recycled (88).
A number of limitations of pulse oximetry should be recognized, appreci-
ated, and understood by everyone who uses pulse oximetry data. Most pulse oxi-
meter errors can be explained as too little signal (e.g., low perfusion, improper
probe placement) or too much noise (e.g., motion, ambient light).
. Accuracy: Many clinical evaluations of the accuracy of pulse oxi-
meters have been published, including a meta-analysis (89). At satur-
ations .80%, the accuracy of pulse oximetry is +4% to +5%.
Below 80%, the accuracy is worse, but the clinical importance of this
is questionable. To appreciate the implications of the limits of accuracy
of pulse oximetry, one must consider the oxyhemoglobin dissociation
curve (Fig. 2). If the pulse oximeter displays a SpO2 of 95%, the true
saturation could be as low as 90% or as high as 100%. If the true
100
90 pH 7.25
pH 7.40
80
70
SO2 (%)
60
50
40
30
20
10
0
0 20 40 60 80 100 120 140 160
PO2 (mm Hg)
Figure 2 Oxyhemoglobin dissociation curve. If a pulse oximeter indicates a SpO2 of

95%, the true HbO2 may be as low as 90% or as high as 100% due to an accuracy of
pulse oximetry of +5%. As indicated in the gure, the SpO2 will change if the oxyhemoglo-
bin dissociation curve shifts (e.g., a change in pH), even though there is no change in PaO2.
saturation is 90%, the PO2 will be about 60 mmHg. If the true satur-
ation is 100%, however, one does not know how high the PO2 might
be. Also, a shift of the oxyghemoglobin dissociation curve can
change the SpO2, although no change in PaO2 has occurred.
. Misunderstanding by those who use pulse oximetry: Although pulse
oximetry is commonly used, it may be misunderstood by users.
Several studies (90,91) have reported that physicians and nurses
lacked adequate knowledge of pulse oximetery and made serious
errors in the interpretation of SpO2.
. Differences between devices and probes: The pulse oximeter is
unique in that it requires no user calibration. However, manufacturer-
derived calibration curves programmed into the software vary
from manufacturer to manufacturer, and they can vary among pulse
oximeters of a given manufacturer. Moreover, the output of LEDs
can vary from probe to probe. The accuracy of pulse oximetry has
been shown to vary among devices (91 97). The same pulse oximeter
and probe should ideally be used for each SpO2 determination on a
given patient.
. Penumbra effect: If the nger pulse oximeter probe does not t cor-
rectly, light can be shunted from the LEDs directly to the photodetector
(98). Theoretically, this will cause a falsely low SpO2 if SaO2 . 85%,
and a falsely elevated SpO2 if SaO2 , 85%.
. Dyshemoglobinemias: Because commercially available pulse oxi-
meters use only two wavelengths of light, they are only able to evaluate
O2Hb and HHb. Pulse oximeters assume that COHb and metHb con-
centrations are low. Abnormal elevations of COHb (99 102) and
114 Hess
metHb (103) both result in signicant inaccuracy in pulse oximetry.

Pulse oximetry is reliable with sickle cell anemia (104).
. Endogenous and exogenous dyes and pigments: Vascular dyes can
affect the accuracy of pulse oximetry, with methylene blue having
the greatest effect (105,106). Nail polish can also affect the accuracy
of pulse oximetry, and it should be removed before monitoring with
pulse oximetry is begun (107). Hyperbilirubinemia does not affect
the accuracy of pulse oximetry (108,109).
. Skin pigmentation: Several studies have found that the accuracy and
performance of pulse oximeters is affected by deeply pigmented skin
(110 112).
. Perfusion: Pulse oximeters require a pulsating vascular bed to function
correctly. Under conditions of low ow (e.g., cardiac arrest or severe
peripheral vasoconstriction), pulse oximetry becomes unreliable.
Under these conditions, an ear probe may be more reliable than a
nger probe.
. Anemia: Although pulse oximeters are generally reliable over a wide
range of hemoglobin levels, they become less accurate and reliable
with conditions of severe anemia (Hb , 8 g/dL at low saturations,
and hematocrit ,10% at all saturations) (113,114).
. Motion: Motion of the probe can produce considerable artifact, and
produce unreliable and inaccurate pulse oximetry readings. This
problem can sometimes be corrected by using an alternate probe site
(such as the ear or toe rather than the nger).
. High-intensity ambient light: Because the photodetector of the pulse
oximeter is non-specic, high-intensity ambient light can produce inter-
ference. The problem can be corrected by wrapping the probe with a
light barrier. The current generation of pulse oximeters may be affected
less by the presence of ambient light (115).
. Abnormal pulses: Venous pulses and a large dicrotic notch have been
shown to affect the accuracy of pulse oximetry (116 118).
. Safety: Pulse oximetry is generally considered safe. However, burns as
the result of defective probes and pressure necrosis have been reported
during the use of pulse oximetry (119 122).
Motion artifact and low perfusion are common causes of pulse oximetry
errors (123,124). A technique that uses mathematical manipulation of the pulse
oximeter signals to measure and subtract the noise components of motion and
low perfusion is commercially available (Masimo Corporation, Mission Viejo,
California, U.S.) (125). Barker (126) performed a human volunteer experiment
to compare the performances of 20-pulse oximeters during combinations of
motion and hypoxemia. A motorized table produced different hand motions,
and each motion was studied during both room air breathing and hypoxemia.
Pulse oximeters on the non-moving hand were used to provide control
measurements for comparison. The Masimo Signal Extraction Technology (SET)

pulse oximeter exhibited the best overall performance, with a performance index
(percentage of time in which the SpO2 reading is within 7% of control value) of
94%. Durbin and Rostow (127) evaluated the impact on clinical care of the
Masimo pulse oximetry technology. Post-cardiac surgery patients were moni-
tored with two oximeters, one employing conventional oximetry and one using
the Masimo technology, on different ngers of the same hand. The amount and
percentage of non-functional monitoring time was collected and found to be
much greater for the conventional pulse oximetery. Time to extubation was not
different between the two groups, but clinicians managing patients with the
Masimo technology weaned patients faster to an FIO2 of 0.40 and obtained
fewer arterial blood gas measurements.
The impact of pulse oximetry on patient outcomes is unclear. A large study
(.20,000 patients) of pulse oximetry use in anesthesia and postanesthesia care
found no difference in outcome (128,129). Pulse oximetry is indicated in unstable
patients likely to desaturate, in patients receiving a therapeutic intervention
that is likely to produce hypoxemia (such as bronchoscopy), and in patients
having interventions likely to produce changes in arterial oxygenation (such as
changes in FIO2 or PEEP). The pulse oximeter is no better at detection of a
ventilator disconnect than the alarms already available on the ventilator.
None the less, pulse oximetry is routinely used in mechanically ventilated
patients in the ICU. Jubran and Tobin (111) evaluated the use of pulse oximetry
in titrating supplemental oxygen in 54 critically ill ventilator-dependent
patients (Fig. 3). In caucasian patients, they found that a SpO2 of 92% was reliable
Figure 3 (Left): In caucasian patients (open circles), a SpO2 92% is reliable in pre-
dicting a PaO2 60 mmHg. (Right): In patients with deeply pigmented skin (lled
circles), a SpO2 95% was required to reliably predict a PaO2 60 mmHg. Source:
Adapted from Ref. 111.
116 Hess
Figure 4 Pulse oximeter tracings from a 60-year-old woman with an exacerbation of

chronic obstructive pulmonary disease who was admitted to the ICU in ventilatory
failure. (A) The patients pulse oximetry tracing at the time of admission, revealing the
respiratory variability in the pulse oximeter plethysmography tracing. Her measured
pulsus paradoxus at this time was 16 mm Hg. (B) The patients pulse oximetry tracing
after 12 hours of aggressive therapy. Her pulsus paradoxus at this time was 8 mm Hg.
Note the absence of RWV in the baseline of the oximeter tracing after the clinical improve-
ment in airow and the resolution of elevated pulsus paradoxus. Abbreviation: RWV,
respiratory waveform variation. Source: Adapted from Ref. 130.
in predicting a PaO2 60 mmHg. In patients with deeply pigmented skin,

however, a SpO2 of 95% was required.
There may be important non-oxygenation monitoring applications for pulse
oximetry. For example, Hartert et al. (130) reported the effect of pulsus para-
doxus, and therefore the severity of air trapping in obstructive airway disease,
on the pulse oximetry plethysmographic waveform (Fig. 4). They reported
that, in patients with obstructive lung disease and elevated pulsus paradoxus,
there was an altered pulse oximetry baseline tracing manifested as the respiratory
waveform variation. Pulsus paradoxus was signicantly correlated with the
degree of respiratory waveform variation of the pulse oximetry tracing and the
amount of auto-PEEP.
VIII. Gastric Tonometry and Sublingual PCO2
Gastric tonometry has been suggested as a minimally invasive method to assess

tissue oxygenation in critically ill patients (131 146). Perfusion of the gut is
affected early in the course of systemic hypoxia, which should result in a decrease
in gastric intramucosal pH (pHi). Normal pHi is 7.38 + 0.03. Hypoperfusion of
the gut also increases gastric intramucosal PCO2, which is referred to as regional
PCO2 (PrCO2). The PrCO2 can increase with either decreased perfusion or
increased PaCO2. With decreased perfusion, the gap between PrCO2 and
PaCO2, the Pr aCO2 increases. Levy et al. reported an increased mortality when
the Pr aCO2 gap was .20 mm Hg (145).
The gastric tonometer consists of a nasogastric tube with a distal CO2-
permeable silastic balloon. The balloon is lled with 2.53 mL of physiologic
saline solution. An equilibration time of one to two hours allows CO2 in the
gastric lumen to equilibrate with the saline inside the balloon. After discarding
11.5 mL of aspirate from the gastric tube (dead volume), the remaining
11.5 mL is analyzed for PCO2 using a blood gas analyzer. A simultaneous arterial
blood sample is analyzed to determine HCO2 3 . Gastric pHi is calculated from
the HendersonHasselbalch equation, using the PCO2 of the saline from the
gastric balloon and the HCO2 3 of arterial blood. Several studies have reported
substantial preanalytical errors related to air contamination of the saline sample
(147149). This has led to the development of gas tonometry techniques
(146,150153). Another advantage of these newer techniques is that they are
automated and can be used for nearly continuous monitoring. Even with improve-
ments in the technical aspects of gastric tonometry, its use remains controversial
(154,155).
Sublingual PCO2 (PSLCO2) correlates well with gastric intramucosal PCO2
and is relatively easy to perform (156 158). PSLCO2 is measured by using a
CapnoProbe subingual PCO2 measurement system (Nellcor, Pleasanton, CA,
U.S.A.). The CapnoProbe consists of a disposable PCO2 sensor and a battery-
powered handheld instrument. The sensor is a CO2-sensing optode. The optode
has a CO2-permeable capsule lled with a uorescent indicator at the distal
end of an optical ber. The uorescent indicator is excited by light conducted
through the optical ber, with the uorescent emission of the indicator being
monitored by the optical ber. As CO2 enters the capsule, carbonic acid is
formed, resulting in a lower pH. The pH change causes a shift in the uorescence
of the indicator, which is detected through the ber optics. The change in uor-
escence is then converted to a PCO2 via the Henderson Hasselbalch equation.
IX. Capnography
Capnometry is the measurement of CO2 at the airway opening during the venti-
latory cycle (159). Capnography is the graphic waveform display of CO2 as a
function of volume or time. The waveform displayed by a capnograph is called
a capnogram. Most bedside capnometers measure CO2 by infrared absorption,
which uses the absorption peak for CO2 at 4.26 mm. The CO2 in the sample
cell decreases the radiation transmitted to the detector. The increased radiation
transmitted from the reference cell (relative to that from the sample cell) is
detected and used to determine the CO2 concentration. The mass spectrometer
can also be used to measure CO2, in which the sample is aspirated into a
vacuum chamber where the gas is ionized by an electron beam. The charged frag-
ments are accelerated into a dispersion chamber where they are separated
118 Hess
according to mass and charge. When light strikes gas molecules, energy is
absorbed and reemitted at the same wavelength and direction, and a small frac-
tion of the absorbed energy is reemitted at new wavelengths in a phenomenon
known as Raman scattering. Raman scattering results in a red-shifted spectrum,
and the wavelength shift and amount of scattering can be used to measure the
PCO2. A portable non-electronic single-patient-use device is commonly used
to produce a color change (colorimetric end-tidal CO2 detection) in the presence
of exhaled CO2 (i.e., tracheal intubation). The color changes from purple with a
low CO2 concentration to yellow with a CO2 concentration of 2.0% to 5.0%.
Capnometers can be congured as mainstream or sidestream devices. With
the mainstream capnometer, the measurement chamber is placed directly at the
airway. With the sidestream capnometer, gas from the airway is aspirated
through ne bore tubing to the measurement chamber inside the device. Some
devices (e.g., colorimetric capnometers) can only be used as mainstream
devices, whereas other devices (e.g., mass and Raman spectrometers) can only
be used as sidestream devices. Infrared capnometers can be congured as
either mainstream or sidestream devices. There are advantages and disadvantages
of each approach (Table 2).
A new capnography technology, Microstream, has recently been intro-
duced (160). Microstream features low ow rates, reduced dead space, and
lack of moisture-associated occlusion problems. Microstream technology uses
a novel molecular correlation spectroscopy source that operates at room tempera-
ture and emits only CO2 specic radiation. The breath sample is brought to the
measuring cell at a ow of 50 mL/min via a specially designed Microstream
breath sampling circuit using a miniature diaphragm pump.
Carbon dioxide homeostasis is affected by volume of CO2 production
(VCO2), CO2 transport from the tissues to the lungs, and alveolar ventilation.
Conditions that increase VCO2 include fever, activity, sepsis, hyperthyroidism,
Table 2 Mainstream and Sidestream Capnometers
Advantages Disadvantages
Mainstream Capnometer
Sensor at patient airway Secretions and humidity block sensor
Fast response (crisp waveform) Sensor heated to prevent condensation
Short lag time (real-time readings) Bulky sensor at patient airway
No sample ow to reduce tidal volume Does not measure N2O
Difcult to use with non-intubated patients
Cleaning and sterilization of reusable sensor
Sidestream Capnometer
No bulky sensors or heaters at airway Secretions block sample tubing
Ability to measure N2O Water trap required
Disposable sample line Slow response to CO2 changes
Can be used with nonintubated patients Sample ow may decrease tidal volume
trauma and burn injuries, and high carbohydrate intake. Conditions that decrease
VCO2 include hypothyroidism, hypothermia (if shivering is controlled), sedation,
and paralysis. Carbon dioxide from tissue metabolism diffuses into the circula-
tion, producing a mixed venous PCO2 (PvCO2) of about 45 mmHg. The PCO2
of an individual lung unit depends upon the relationship between ventilation
and perfusion (V/Q ) (Fig. 5) (161 167). With no perfusion (dead space;
V/Q 1), the PaCO2 is the same as the inspired PCO2 (i.e., zero). With a
normal V/Q , the PACO2 is the same as the arterial PCO2 (i.e., 40 mm Hg).
With a low V/Q , the PACO2 increases towards the PvCO2 (i.e., 45 mm Hg).
The PACO2, and thus the end-tidal PCO2, must always be between zero and
the PvCO2. With a variety of V/Q throughout the lungs, the end-tidal PCO2
(PETCO2) is normally several mm Hg less than the PaCO2. However, the relation-
ship between the PaCO2 and PETCO2 will vary depending upon the relative
contributions of various V/Q units of the lungs.
The volume-based capnogram is displayed with PCO2 on the ordinate and
volume on the abscissa (Fig. 6). At the beginning of exhalation, PCO2 remains
zero as gas from the anatomic dead space leaves the airway (Phase I). The cap-
nogram then rises sharply as alveolar gas mixes with dead space gas (Phase II).
The capnogram then forms a plateau during most of exhalation (Phase III). Phase
III represents gas ow from alveoli and is thus called the alveolar plateau.
The slope of Phase III is determined by the V/Q status of the lung (163,168
171). The slope of Phase III is increased in patients with airway obstruction
(e.g., COPD, asthma). The PCO2 at end-exhalation is the PETCO2.
Airway dead space volume (anatomic dead space), physiologic dead space
fraction (VD/VT), and VCO2 can be determined from the volume-based capno-
gram (Fig. 6). If VCO2 is known, it is possible to calculate metabolic rate:
_
REE VCO 2 5:52 1440
where REE is resting energy expenditure (kcal/d), VCO2 is in L/min, 5.52 is the
caloric equivalent for CO2, and 1440 is the number of minutes in a day. The
Figure 5 , normal V/Q

PETCO2 with low V/Q , and high V/Q
.
120 Hess
%CO2in
Arterial Blood Alveolar dead space
% CO2 Carbon dioxide

production
VD VALV
Exhaled Tidal Volume
Figure 6 Components of volume-based capnogram. Abbreviations: VD, anatomic dead

space; VALV, alveolar gas volume.
mixed exhaled PCO2 (PE CO2) can be calculated from VCO2 if the minute ven-
tilation (VE) is known:

PE CO2 _
VCO 2
_
VCO _
2 VE or PE CO2 Pbar
Pbar V_ E
where Pbar is barometric pressure. The VD/VT can then be calculated if the PaCO2
is known:
VD Pa CO2 PE CO2

VT Pa CO2
Increased VD/VT has been associated with mortality in patients with ARDS
(172). Increased VD/VT has also been associated with a lower rate of weaning
from mechanical ventilation (173).
For critical care applications, the time-based capnogram is often displayed
(Fig. 7). Unlike the volume-based capnogram, the time-based capnogram has an
inspiratory segment and an expiratory segment. The PCO2 is zero during the
inspiratory phase. At the beginning of exhalation, PCO2 remains zero as gas
from the anatomic dead space leaves the airway (Phase I). The capnogram
then rises sharply as alveolar gas mixes with dead space gas (Phase II). The
curve then forms an alveolar plateau during most of exhalation (Phase III).
The PCO2 at the end of the alveolar plateau is the PETCO2. Note that Phase I,
Phase II, and Phase III are similar for the time based capnogram and the
volume-based capnogram. The capnogram with airow obstruction is character-
ized by an increased slope of Phase III (Fig. 8). The lack of an alveolar plateau
end
III exhalation
PCO2
II
I
time
begin
exhalation
Figure 7 Time-based capnogram. I, anatomic dead space; II, the transition from
anatomic dead space to the alveolar plateau; III, the alveolar plateau.
occurs because of the V/Q abnormalities that result from the airow obstruction.
In asthmatic patients with acute bronchospasm, the slope of Phase III has been
shown to correlate with peak expiratory ow rate, and it normalizes with beta-
agonist therapy (170,171).
The PETCO2 represents alveolar PCO2, and the PACO2 is the result of the
V/Q . With a normal V/Q , the PACO2 will approximate the PaCO2. If ventilation
is decreased compared with perfusion, there will be more time for equilibration
between PvCO2 and PACO2, and thus the PACO2 will rise towards PvCO2. With a
high V/Q (i.e., dead space), PACO2 will approach the inspired PCO2 (PICO2),
end
exhalation
III
PCO2
II
I
time
begin
exhalation
Figure 8 Capnogram produced with airow obstruction.

122 Hess
which is usually zero. The PETCO2 is a mixture of gas from millions of alveoli
and thus represents the mixture of many potentially different PACO2. Theoreti-
cally, PETCO2 could be as low as the PICO2 (zero) or as high as the PvCO2
(but not higher than this). An increase or decrease in PETCO2 can be the result
of changes in VCO2 and delivery to the lungs (167,174), changes in alveolar ven-
tilation, or an equipment malfunction (Table 3). However, because of homeosta-
sis, compensatory changes may occur so that PETCO2 does not change despite
these changes. For example, if VCO2 increases (such as with fever) and alveolar
ventilation increases proportionately (the normal homeostatic response), then
PETCO2 may not change. Thus, PETCO2 is a non-specic indicator of cardiopul-
monary homeostasis and often does not indicate a specic problem or
abnormality.
If the PaCO2 is measured, the gradient between PaCO2 and PETCO2
(Pa2ETCO2) can be calculated. This gradient is normally small, being less than
5 mm Hg. With dead space-producing disease (high V/Q ), the PETCO2 may be
considerably less than the PaCO2 (Table 4) (174 185). Although shunting may
result in a large gradient between PAO2 and PaO2, it only has a small effect on
the Pa2ETCO2. On occasion, the PETCO2 may be greater than the PaCO2. The
reason for a PETCO2 greater than PaCO2 is not well understood (186) and may
relate to low (but nite) V/Q regions within the lung. Fletcher and Jonson
(162) have reported that PETCO2 is more often greater than PaCO2 when the
tidal volume is high. Tulou and Walsh (187) found that the Pa2ETCO2 decreased
signicantly when measured at maximal exhalation in patients with airway
obstruction. With a larger tidal volume, the greater expiratory time may allow
Table 3 Causes of Increased and Decreased PETCO2
Increased PETCO2 Decreased PETCO2
Increased CO2 production and Decreased CO2 production and

delivery to the lungs delivery to the lungs
Fever, sepsis, bicarbonate administration, Hypothermia, pulmonary
increased metabolic rate, seizures hypoperfusion, cardiac arrest,
pulmonary embolism, hemorrhage,
hypotension
Decreased alveolar ventilation Increased alveolar ventilation
Respiratory center depression, Hyperventilation
muscular paralysis,
hypoventilation, COPD
Equipment malfunction Equipment malfunction
Rebreathing, exhausted CO2 absorber, Ventilator disconnect, esophageal
leak in ventilator circuit intubation, complete airway
obstruction, poor sampling,
leak around endotracheal tube cuff

Table 4 Causes of Increased Pa2ETCO2
Pulmonary hypoperfusion
Pulmonary embolism
Cardiac arrest
Positive pressure ventilation (especially PEEP)
High-rate low-tidal volume ventilation
Abbreviation: PEEP, positive end-expiratory pressure.
lung units with a low V/Q (and thus a longer time constant) to empty (188). Jones
et al. (189) found PETCO2 greater than PaCO2 during exercise and attributed this
to an increase in PACO2 because of increased CO2 that is emitted into a lung
volume becoming smaller during exhalation.
Stability of the Pa2ETCO2 is necessary if PaCO2 is to be reliably predicted
from PETCO2. Hoffman et al. (178) evaluated PETCO2 in critically ill patients
during changes in mechanical ventilation and found that changes in PETCO2
did not correlate well with changes in PaCO2 and that the trends in PETCO2
were opposite of the trends in PaCO2 in some patients. Raemer et al. (181)
found that the Pa2ETCO2 was too variable during anesthesia to allow precise pre-
diction of PaCO2 from PETCO2. Russell et al. (182 184) found that changes in
PaCO2 and PETCO2 were in opposite directions in a sizeable proportion of
patients. Because of a uctuating and unpredictable Pa2ETCO2, caution must
be used when predicting PaCO2 from PETCO2.
Measurement of PETCO2 is a standard of care to determine proper endotra-
cheal tube position (190 196). Esophageal PCO2 may be high (5%) after
exhaled gas ventilation following inadvertent gastric distention (192), but it
drops to ,2% following six ventilations of the stomach. Ingestion of a carbo-
nated beverage also results in an increased esophageal PCO2, but this rapidly
decreases following 10 to 15 seconds of gastric ventilation (197). Capnography
has also been reported useful to verify feeding tube placement (198).
Typically, the onset of cardiac arrest results in a drop of PETCO2 to zero.
With the initiation of CPR, there is an increase in PETCO2. PETCO2 correlates
with cardiac output (199 202) (i.e., pulmonary blood ow) and coronary per-
fusion pressure (203,204) during CPR. Kalenda (205) observed that PETCO2
decreased when a resuscitator became fatigued and that it increased when resus-
citation was continued with another resuscitator. Garnett et al. (206) reported that
PETCO2 increased immediately in patients who had a return of spontaneous cir-
culation. Similar results were reported by Falk et al. (207) and Sanders et al.
(208). Patients who were resuscitated had a PETCO2 of 15 + 4 mm Hg during
resuscitation, whereas patients who were not resuscitated had a PETCO2 of
only 7 + 5 mm Hg (208,209). Grmec et al. (210) reported that the initial
PETCO2 was greater in patients with asphyxial cardiac arrest than in those with
cardiac arrest due to ventricular brillation or ventricular tachycardia.
124 Hess
The relationship between the Pa2ETCO2 and VD/VT has been reported
(161,162,185,211 213). Fletcher et al. (162) reported that patients with
increased alveolar dead space had an increased slope of the alveolar plateau on
the capnogram. Yamanaka and Sue (185) reported that the Pa2ETCO2 correlated
closely with VD/VT. A common clinical cause of increased dead space is pulmon-
ary embolism, and there has been interest in the use of capnography in this setting
(214 218). In a comparison of capnometry to angiography in the diagnosis of
pulmonary embolism in 44 adult patients with chronic obstructive pulmonary
disease (COPD), Chopin et al. (217) reported a sensitivity of 100% but a speci-
city of only 65% (i.e., a false positive rate of 35%). Although the negative pre-
dictive value was 100%, the positive predictive value was only 74%. The
Pa2ETCO2 is usually increased when pulmonary embolism is present, but the gra-
dient is also increased for a variety of other causes when pulmonary embolism is
not present (e.g., any dead space-producing disease). Hatle (214) reported that
Pa2ETCO2 measured at forced exhalation was more useful in the evaluation of
acute pulmonary embolism. Eriksson et al. (216) and Verschuren et al. (218)
extrapolated phase III of the volumetric capnogram to determine the PCO2 at
15% of the predicted total lung capacity and found that this was useful in the
diagnosis of pulmonary embolism (Fig. 9). With maximal exhalation, the
Figure 9 The capnogram in a patient with a diagnosis of pulmonary embolism. This

patient has a VT of 612 mL. The PaCO2 was 29 mm Hg, the PETCO2 was 22.7 mm Hg,
and 15% of the predicted TLC was calculated to be 699 mL. The PETCO2 at this
volume of was 23.5 mm Hg after extrapolation of phase III. This relatively high gradient
theoretically separates a patient with pulmonary embolism from a healthy patient or a
chronic obstructive pulmonary disease patient. Source: Vershuren F. Liistro G, Coffeng
R, et al. Volumetric capnography as a screening test for pulmonary embolism in the emer-
gency department. Chest 2004; 125:841 850.
gradient approached zero in patients with obstructive lung disease but remained
high in patients with pulmonary embolism.
There has been interest in the use of capnometry during weaning from mech-
anical ventilation (219227). Most evaluations of the use of capnometry during
weaning have been conducted on post-operative patients. It can be argued that
the weaning of these patients is usually uncomplicated, that such patients require
little monitoring during weaning, and that extrapolation of these data to difcult-
to-wean patients is not valid. Perhaps more important, hypoventilation is often a
late nding with respiratory muscle fatigue (such as might occur during
weaning). Other clinical ndings such as tachypnea, use of accessory muscles of
breathing, and thoraco-abdominal paradox may be more sensitive indicators than
an increase in PCO2 to indicate fatigue during weaning from mechanical venti-
lation. On the basis of the available evidence, however, routine capnometry
cannot be recommended during weaning from mechanical ventilation. However,
the use of capnometry in long-term weaning unit has been reported useful (228).
Avoidance of hypercapnia remains a common treatment objective in the
care of patients with head injury. Mackersie and Karagianes (229) evaluated
the use of PETCO2 in 36 head-injured patients and reported good correlation
between PETCO2 and PaCO2. Sharma et al. (230) reported a stable Pa2ETCO2
during neuroanesthesia. In contrast to these ndings, others have reported that
PETCO2 is not a useful predictor of PaCO2 in neurosurgical patients (231,232).
Because many head injured patients are young and free of lung disease (at
least early in their hospital course), PETCO2 may be useful to monitor PaCO2.
However, PETCO2 may not be a very good predictor of PaCO2 in these patients
if acute or chronic lung disease is present or if pulmonary blood ow is reduced
secondary to hypotension or pulmonary embolism. Kerr et al. (233) found that
PETCO2 correlated well with PaCO2 in patients without respiratory compli-
cations, but its clinical validity is questionable in patients who have the greatest
need for PETCO2 monitoring (those patients with respiratory failure).
Palmon et al. (234) reported that PCO2 could be more tightly controlled
during transport when PETCO2 was monitored, but their data did not support
routine use of capnometry during short transport times. In trauma patients intu-
bated in the eld, Helm et al. (235) reported tighter control of PCO2 with the
use of capnography. In this study, the incidence of normoventilation was
greater and the incidence of hypoventilation was less with the use of
capnography.
Using volume-based capnography, it is possible to noninvasively measure
cardiac output with the partial CO2 rebreathing technique (NICO, Novametrix,
Wallingford, Connecticut, U.S.) (236,237). VCO2 is calculated on a breath-by-
breath basis, and the differential Fick equation is applied to establish the relation
between VCO2 and cardiac output (Q ):
_
VCO _
2 Q (Cv CO2 Ca CO2 )
126 Hess
where CvCO2 represents the CO2 content in mixed venous blood, and CaCO2 rep-
resents the CO2 content in arterial blood. CO2 rebreathing is performed for
35 seconds every 3 minutes (Fig. 10). Assuming that Q remains constant
during the rebreathing procedure yields the following:
_
DVCO _
2 Q (DCv CO2 DCa CO2 )
where DVCO2 is the change in VCO2 between normal breathing and rebreathing,
DCvCO2 the change in mixed venous carbon dioxide content, and DCaCO2 the
change in arterial carbon dioxide content. If CvCO2 remains constant during
rebreathing, the following equation is used:
_
DVCO _
2 Q (DCa CO2 )
When end-capillary content (CcCO2) is used in place of CaCO2, pulmonary capil-

lary blood ow (PCBF), the blood ow that participates in alveolar gas exchange,
Figure 10 Rebreathing cycle used by NICO to measure cardiac output using the partial
rebreathing technique. Source: Courtesy of Novametrix, Wallingford, Connecticut, U.S.A.
, and the following equation is used:

is measured rather than Q
_
DVCO 2 PCBF (Cc CO2 )
Assuming that 2CcCO2 is proportional to DPETCO2, the following equation can

be used:
_
PCBF DVCO 2 =(S DPET CO2 )
where DPETCO2 is the change in PETCO2 between normal breathing and

rebreathing and S is the slope of the carbon dioxide dissociation curve for
hemoglobin. Because cardiac output is the sum of PCBF and intrapulmonary
shunt ow:
_ PCBF
Q
_ s =Q
(1 Q _ t)
where Q s/Q
t is the intrapulmonary shunt fraction. The noninvasive method for
estimating Q s/Q t is adapted from Nunns iso-shunt plots, which are a series of con-
tinuous curves indicating the relation between PaO2 and FIO2 for different levels
of shunt. PaO2 is non-invasively determined using a pulse oximeter. There are
several potential limitations of partial rebreathing for the measurement of
cardiac output. In non-paralyzed patients, rebreathing increases respiratory rate,
which reduces the magnitude of the signal and limits the ability to detect
changes in PETCO2 and VCO2. Noise is increased by respiratory pattern irregula-
rities that produce unstable PETCO2 and VCO2 that may impair accuracy.
Additional cardiac output not calculated due to shunt fraction is estimated from
SpO2 and FIO2, which introduces errors. Several studies have reported mixed
results of accuracy of this method for measuring cardiac output (236 245).
X. Lung Mechanics and Graphics
Pulmonary mechanics is the expression of lung function through measures of

pressure and ow (246). From these measurements, a variety of derived
indices can be determined, such as volume, compliance, resistance, and work-
of-breathing. Pulmonary graphics are derived when one of the parameters of pul-
monary mechanics is plotted as a function of time or as a function of one of the
other parameters. This produces scalar pressure time, ow time, and volume
time graphics as well as ow volume and pressure volume (PV) loops. Current
generation ventilators provide monitoring of pulmonary mechanics and graphics
in real time. Bedside monitoring of mechanics and graphics during positive
pressure ventilation typically portrays the lungs as a single compartment with a
linear response over the range of tidal volume.
128 Hess
Airway pressure is universally measured during mechanical ventilation.

Peak inspiratory pressure (PIP) is predicted mathematically by the Equation of
Motion (247,248):
VT
PAW Pmus R V_ PEEP
C
where PAW is proximal airway pressure, Pmus the pressure generated by the res-
piratory muscles, C the respiratory system compliance, VT the tidal volume, R the
airways resistance, V the inspiratory ow, and PEEP the end-expiratory alveolar
pressure set on the ventilator. The Equation of Motion predicts that proximal
airway pressure will increase with a higher tidal volume, lower respiratory
system compliance, higher airways resistance, higher inspiratory ow, higher
PEEP, and the presence of auto-PEEP.
Because of the airways resistance, proximal airway pressure will always be
greater than alveolar pressure during inspiration if ow is present. During volume-
controlled ventilation, plateau pressure (Pplat) is obtained by applying an end-
inspiratory breath-hold for 0.5 to 2 seconds, during which pressure equilibrates
throughout the system so that the pressure measured at the proximal airway
approximates the peak alveolar pressure (Fig. 11) (249,250). Pplat is not valid
during active breathing and thus cannot be measured with ventilator modes such
as pressure support ventilation. During pressure-controlled ventiation, the ow
may decrease to zero before the end of the inspiratory phase, in which case PIP
and Pplat are equal.
Pplat is determined by tidal volume, PEEP and respiratory system
compliance: Pplat VT/C PEEP. Pplat indicates a greater risk of alveolar
PIP
resistance
flow
Pplat
compliance
tidal volume
PEEP time
Figure 11 Schematic representation of a pressure waveform during volume-controlled

ventilation. The difference between PIP and Pplat is determined primarily by airways
resistance and ow. The difference between Pplat and the level of PEEP is determined
by compliance and tidal volume. Abbreviations: PEEP, positive end-expiratory pressure;
PIP, peak inspiratory pressure; Pplat, plateau pressure.
over-distension during mechanical ventilation and should be maintained

30 cmH2O (251). However, a higher Pplat may be safe (and necessary) if
chest wall compliance is decreased. During volume-controlled ventilation, the
alveolar pressure at any time after the initiation of inspiration is determined by
the volume delivered by that point in the inspiratory phase and respiratory
system compliance: PALV V/C. For pressure ventilation, PALV at any time
after the initiation of inspiration is (247):
PALV (DP) (1 et=t )
where DP is the pressure applied to the airway above PEEP, e the base of the
natural logarithm, t the elapsed time after initiation of the inspiratory phase,
and t the time constant of the respiratory system.
Incomplete emptying of the lungs occurs if the expiratory phase is termi-
nated prematurely (252). When this occurs, alveolar pressure does not equilibrate
with proximal airway pressure at end-exhalation, and gas trapping results. The
pressure produced by this trapped gas is called auto-PEEP. Auto-PEEP increases
end-expiratory lung volume (hyperination). It is measured by applying an
end-expiratory pause for 0.5 to 2 seconds (Fig. 12). The pressure measured at
the end of this maneuver that is in excess of the PEEP set on the ventilator is
auto-PEEP. For a valid measurement, the patient must be relaxed and breathing
in synchrony with the ventilator. Many patients with COPD contract their
abdominal muscles during exhalation (253 255). This is an important determi-
nant of auto-PEEP in these patients but likely does not produce hyperination.
It has also been shown that the end-expiratory pause method can underestimate
auto-PEEP with complete airway closure during expiration, as may occur
during mechanical ventilation of patients with severe asthma (256).
Auto-PEEP is a function of ventilator settings (tidal volume and expiratory
PIP PIP
pressure
auto PEEP
set PEEP
time
Figure 12 Measurement of auto-PEEP with an end-expiratory pause maneuver. The

difference between the pause pressure and the set PEEP level is the amount of auto-PEEP.
Abbreviations: PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressure.
130 Hess
time) and lung function (airways resistance and lung compliance). It can be
expressed mathematically as (247):
VT
auto-PEEP K
C (e x Te 1)
where Kx is the inverse of the expiratory time constant (1/t) and Te is the expira-
tory time. Note that auto-PEEP is increased with increased resistance and com-
pliance (e.g., chronic obstructive lung disease), increased respiratory rate or
increased inspiratory time (both decrease expiratory time), and increased tidal
volume. Clinically, auto-PEEP can be decreased by decreasing minute venti-
lation (rate or tidal volume), increasing expiratory time, or decreasing airways
resistance (e.g., bronchodilator administration). Because set-PEEP may counter-
balance auto-PEEP in patients with ow limitation, auto-PEEP should be meau-
sured with PEEP 0.
Esophageal pressure is measured from a balloon inated with a small
volume of air (,1 mL) placed into the lower esophagus (257). Esophageal
pressure changes reect changes in pleural pressure. However, the absolute eso-
phageal pressure does not reect absolute pleural pressure. Changes in
esophageal pressure can be used to assess respiratory effort and work-of-breath-
ing during spontaneous breathing, to assess chest wall compliance during full
ventilatory support, and to assess auto-PEEP during spontaneous breathing.
If an esophageal balloon is not present, changes in pleural pressure can be esti-
mated by observing the respiratory variability of the central venous pressure
(Fig. 13) (258).
If exhalation is passive, the change in esophageal pressure required to
reverse ow at the proximal airway (i.e., trigger the ventilator) reects the
amount of auto-PEEP. Negative esophageal pressure changes that produce no
ow at the airway indicate failed trigger efforts (Fig. 14). Clinically, this is recog-
nized as a patient respiratory rate that is greater than the trigger rate on the ven-
tilator (observed by inspecting chest wall movement).
The airway pressure waveform is affected by active breathing efforts of the
patient. Patient ventilator dyssynchrony results in an airway pressure waveform
that varies from breath to breath, particularly during volume-controlled ventilation
(250,257). A special form of patient ventilator dyssynchony can occur during
pressure support ventilation, in which the patient actively exhales to terminate
the inspiratory phase (259,260). This is seen as a pressure spike at end-inspiration,
causing the ventilator to pressure-cycle to the expiratory phase (Fig. 15).
During volume-controlled ventilation, the inspiratory ow is that set on the
ventilator. During pressure-controlled or pressure-support ventilation, ow is
determined by the pressure applied across the lungs, airways resistance, and
the time constant (261,262):

_ DP
V (et=t )
R
Inhalation Exhalation
positive
pressure mm Hg
ventilation 18
10 Chest wall compliance
Inhalation Exhalation
mm Hg
spontaneous 18 Inspiratory muscle effort
breathing
10
Figure 13 (Top) During positive pressure ventilation in a relaxed patient, the increase
in CVP during the inspiratory phase is determined by chest wall compliance. (Bottom)
During spontaneous breathing, the decrease in CVP during the inspiratory phase is deter-
mined by inpiratory muscle effort. Abbreviation: CVP, central venous pressure.
flow 0
(L/s)
0.8
volume
(L)
0
30
Paw
(cm H2O)
0
20
Peso
(cm H2O) 0
estimation of missed
auto-PEEP trigger effort
Figure 14 Esophageal pressure measurements in a patient with auto-PEEP. Note that an

esophageal pressure drop of about 10 cm H2O is necessary to trigger the ventilator. This
represents an auto-PEEP level of about 10 cm H2O. Also note an inspiratory effort
that is not great enough to overcome auto-PEEP and trigger the ventilator. It can also
be seen that expiratory ow does not return to zero before the subsequent breath is
initiated.
132 Hess
Figure 15 With pressure support ventilation, patients may terminate the inspiratory
phase by actively exhaling. This produces an end-inspiratory pressure spike. In this
case, the ow does not decrease to the termination ow, as represented by the dashed
line. Source: Adapted from Ref. 259.
where DP is the sum of the pressure applied to the airway and the pressure gen-
erated by the respiratory muscles, R the airways resistance, t the elapsed time
after initiation of the inspiratory phase, e the base of the natural logarithm, and
t the product of airways resistance and respiratory system compliance (the
time constant of the respiratory system). Expiratory ow is determined by alveo-
lar driving pressure (PALV), airways resistance, the elapsed time since initiation
of exhalation, and the time constant of the respiratory system (248):

_ PALV
V (et=t )
R
A useful application of the airway ow waveform is detection of auto-

PEEP. If the expiratory ow does not return to baseline, this indicates the pre-
sence of auto-PEEP. Although the ow waveform is useful to detect
auto-PEEP, it does not quantitatively indicate the amount of auto-PEEP.
Dips in expiratory ow during patient-triggered ventilation (e.g., pressure
support) indicate trigger efforts in which the patients inspiratory effort was insuf-
cient to overcome auto-PEEP to trigger the ventilator (Fig. 14). The use of ow
graphics to recognize the presence of auto-PEEP is useful, although the
clinical examination can also be used to reliably detect the presence of auto-
PEEP (263).
Critical care ventilators derive volume by integration of ow. Because ow
is usually not measured directly at the proximal airway, volume output from the
ventilator is less than the volume delivered to the patient due to conditioning of
the inspired gas (warming and humidication) and gas compression in the circuit.
The volume waveform may be useful to detect a leak (e.g., bronchopleural
stula), which results in a difference between the inspiratory and expiratory
tidal volume.
Respiratory system compliance (CRS) is assessed in mechanically venti-
lated patients as the tidal volume divided by the pressure required to produce
that volume:
DV VT
CRS
DP Pplat PEEP
Respiratory system compliance is typically 50 to 100 mL/cm H2O in mechani-

cally ventilated patients and is determined by the compliance of the lungs and
chest wall. Chest wall compliance is calculated from the change in esophageal
pressure (pleural pressure) during passive ination. Chest wall compliance is nor-
mally 200 mL/cm H2O and can be decreased due to abdominal distension, chest
wall edema, chest wall burns, thoracic deformities (e.g., kyphoscoliosis), and an
increase in muscle tone (e.g., a patient who is bucking the ventilator). Chest wall
compliance is increased with ail chest and paralysis. Lung compliance is calcu-
lated using the transpulmonary pressure; in other words, the difference between
alveolar pressure (Pplat) and pleural pressure (esophageal). Normal lung compli-
ance is 100 mL/cm H2O and is decreased with pulmonary edema (cardiogenic or
noncardiogenic), pneumothorax, consolidation, atelectasis, pulmonary brosis,
pneumonectomy, mainstrem intubation, and hyperination. Lung compliance is
increased with emphysema.
During volume-controlled ventilation, inspiratory airways resistance can
be estimated:
PIP Pplat
RI
V_ I
134 Hess
where VI is the end-inspiratory ow. A simple way to make this measurement is

to set the ventilator for a constant inspiratory ow of 60 L/min (1 L/sec). Using
this approach, the inspiratory airways resistance is PIP Pplat. Other methods used
to measure airways resistance (264) include interrupter techniques, least squares
tting technique, and time constant of the respiratory system. Common causes of
an increased airways resistance are bronchospasm, secretions, and a small inner
diameter endotracheal tube. For intubated and mechanically ventilated patients,
airways resistance should be ,10 cm H2O/(L/sec) at a ow of 1 L/sec. Expiratory
airways resistance is typically greater than inspiratory airways resistance.
Mean airway pressure is determined by peak inspiratory pressure (PIP), the
fraction of time devoted to the inspiratory phase (Ti/Ttot), and PEEP (265,266).
For constant ow volume ventilation, in which the airway pressure waveform
is triangular, mean airway pressure can be estimated as (267).

Ti
Mean airway pressure 0:5 (PIP PEEP) PEEP
Ttot
During pressure ventilation, in which the airway pressure waveform is rectangu-

lar, mean airway pressure can be estimated as:

Ti
Mean airway pressure (PIP PEEP) PEEP
Ttot
The PV curve represents the static relationship between pressure and

volume of the respiratory system (lungs, abdomen, rib cage, and respiratory
muscles). It can be constructed using a number of techniques (268) that
measure pressure as the lungs are inated or deated. A super-syringe can be
used to inate the lungs with step changes in volume, or the lungs can be inated
with a constant slow ination (269,270). Volume is plotted as a function of
pressure (Fig. 16). This measurement requires a completely relaxed chest wall,
and thus the patient must be paralyzed for the best results. The respiratory
system PV curve can be separated into the lung and chest wall curves by estimat-
ing pleural pressure with an esophageal balloon. In the normal respiratory system,
the shape of the PV curve is nearly linear above the resting volume. The ination
and deation curves demonstrate differences in their shape and pressure for a
given volume (hysteresis). The ination curve with acute lung injury begins
with a at portion followed by a transition to a steeper, more compliant region
(Fig. 16). This transition has been called the lower inection point (PFlex). The
curve continues with a linear progression and at its upper end undergoes
another transition to a at region. This transition has been called the upper inec-
tion point. On deation a similar shape is achieved, but at a lower pressure than
the ination curve. In normal lungs, the lower PFlex has been equated with the
closing volume, and the upper PFlex has been equated with overdistension
(271). Determination of inection points is often arbitrary and inaccurate.
Methods have ranged from eyeball approximation to graphical curve-tting
Figure 16 Pressure volume curve during mechanical ventilation. Note that the curve is
nearly linear in the normal condition. In ARDS, the curve demonstrates a lower inection
point and an upper inection point. Abbreviation: ARDS, acute respiratory distress
syndrome.
methods. Inter-observer variability in the determination of PFlex has been

reported (272,273). Methods based on curve-tting equations may provide
more accurate estimates of the inection points (273,274).
There has been enthusiasm for the use of PV curves to optimize ventilator
settings with PEEP set above the PFlex and Pplat set below the upper PFlex. Recent
observations and analysis of the PV curve in acute respiratory distress syndrome
(ARDS) have changed its interpretation and implications for management. Two
studies reported that the chest wall affects the PFlex and determination of the
upper inection point (275,276). These observations imply that the PV curve
should be measured with an esophageal balloon to determine the inection
points for the lung alone. Because the PV curve represents the sum behavior of
all lung units, and given the heterogeneity of lung injury, it might not be possible
to determine an ideal point of recruitment or overdistention. Mathematical mod-
eling data suggest that PEEP settings based on PFlex may not be adequate to
ensure an open lung (277,278).
XI. Monitoring in Perspective
How much monitoring is needed? This is an important question for both clini-
cians and administrators. Clinicians often want to monitor everything possible,
with a more is better attitude. On the other hand, administrators and
managed care providers become justiably concerned with the costs associated
with monitoring. The presence of many monitors at the bedside can be very
136 Hess
distracting to clinicians (279,280). Many monitoring systems beep, buzz, and

blink constantly. The most frequent false-positive alarms have been reported
from the pulse oximeter (281 284). Bentt et al. (283) found that a pulse oximeter
alarm was present up to 47% of the time (28 min/hr) in a 10-bed surgical ICU,
and many of these were false alarms and required no intervention (283). During
anesthesia monitoring, Kestin et al. (284) found that 75% of all alarms that
sounded were spurious, and only 3% indicated risk to the patient. In an adult
ICU, monitor alarms were present 20% of the time, with an average peak
sound level of nearly 80 dBA (the EPA recommends that noise levels in hospitals
not exceed 45 dBA).
The nancial impact of monitoring is difcult to assess. For non-invasive
monitors such a pulse oximeters, no prospective randomized controlled trials
have evaluated cost effectiveness. Using a historical control and a relatively
short evaluation period (2 mo), Kellerman et al. (285) reported a reduction in
the number of ABGs ordered in an Emergency Department following the intro-
duction of pulse oximetry. However, others have shown an increased number of
ABGs when pulse oximetry was used (3). For mixed venous oximetry, Orlando
(286) reported a savings of $75/patient. Pearson et al. (287), however, found that
monitoring mixed venous saturation added signicantly to the cost incurred with
a routine pulmonary artery catheter. Jastremski et al. (288) also found that mixed
venous oximetry may not be cost-effective, particularly in a xed-payment reim-
bursement system.
Monitoring should not be done just because it is technically feasible.
Technical capability must be balanced against clinical usefulness and cost effec-
tiveness. The decision to monitor, like any other clinical decision, should be
based on therapeutic objectives.
References
1. Hess D. An overview of noninvasive monitoring in respiratory care: present, past,

and future. Respir Care 1990; 35:482 499.
2. Rafn TA. Indications for arterial blood gas analysis. Ann Intern Med 1986; 105:
390 398.
3. Muakkassa FF, Rutledge R, Fakhry SM, Meyer AA, Sheldon GF. ABGs and arterial
lines: the relationship to unnecessarily drawn arterial blood gas samples. J Trauma
1990; 30:1087 1095.
4. Smoller BR, Kruskall MS. Phlebotomy for diagnostic laboratory tests in adults.
Pattern of use and effect on transfusion requirements. N Engl J Med 1986;
314:1233 1235.
5. Low LI, Harrington GR, Stoltzfus DP. The effect of arterial lines on blood-drawing
practices and costs in intensive care units. Chest 1995; 108:216 219.
6. Zimmerman JE, Seneff MG, Sun X, et al. Evaluating laboratory usage in the inten-
sive care unit: patient and institutional characteristics that inuence frequency of
blood sampling. Crit Care Med 1997; 25:737 748.
7. Corwin HL, Parsonnet KC, Gettinger A. RBC transfusion in the ICU: Is there a
reason? Chest 1995; 108:767 771.
8. Peruzzi WT, Parker MA, Lichtenthal PR, et al. A clinical evaluation of a blood con-
servation device in medical intensive care unit patients. Crit Care Med 1993;
21:501 506.
9. Silver MJ, Jubran H, Stein S, et al. Evaluation of a new blood conserving arterial
line system for intensive care units. Crit Care Med 1993; 21:507 511.
10. Silver MJ, Li YH, Gragg LA, et al. Reduction of blood loss from diagnostic
sampling in critically ill patients using a blood conserving arterial line system.
Chest 1993; 104:1711 1715.
11. Peruzzi WT, Noskin GA, Moen SG, et al. Microbial contamination of blood con-
servation devices during routine use in the critical care setting: results of a prospec-
tive, randomized trial. Crit Care Med 1996; 24:1157 1162.
12. Foulke GE, Harlow DJ. Effective measures for reducing blood loss from diagnostic
laboratory tests in intensive care unit patients. Crit Care Med 1989; 17:1143 1145.
13. Beasley KE, Darin JM, Durbin CG. The effect of respiratory care department man-
agement of a blood gas analyzer on the appropriateness of arterial blood gas utiliz-
ation. Respir Care 1992; 37:343 347.
14. Pilon CS, Leathley M, London R, et al. Practice guideline for arterial blood gas
measurement in the intensive care unit decreases numbers and increases appropri-
ateness of tests. Crit Care Med 1997; 25:1308 1313.
15. Hess D, Agarwal NN. Variability of blood gases, pulse oximeter saturation, and
end-tidal carbon dioxide pressure in stable, mechanically ventilated trauma
patients. J Clin Monit 1992; 8:111 115.
16. Thorson SH, Marini JJ, Pierson DJ, Hudson LD. Variability of arterial blood gas
values in stable patients in the ICU. Chest 1983; 84:14 18.
17. Sasse SA, Chen PA, Mahutte CK. Variability of arterial blood gas values over time
in stable medical ICU patients. Chest 1994; 106:187 193.
18. Tsai YH, Lin MC, Hsieh MJ, et al. Spontaneous variability of arterial oxygenation
in critically ill mechanically ventilated patients. J Clin Monit 1999; 25:37 43.
19. Hess D, Good C, Didyoung R, et al. The validity of assessing arterial blood gases 10
minutes after an FIO2 change in mechanically ventilated patients without chronic
pulmonary disease. Respir Care 1985; 30:1037 1041.
20. Hess D, Elser RC, Agarwal NN. The effects on the pulmonary shunt value of using
measured versus calculated hemoglobin oxygen saturation and of correcting for
the presence of carboxyhemoglobin and methemoglobin. Respir Care 1984;
29:1001 1005.
21. Hess D, Maxwell C, Shefet D. Determination of intrapulmonary shunt: comparison
of an estimated shunt equation and a modied equation with the classic equation.
Respir Care 1987; 32:268 273.
22. Cohn JD, Engler PE. Shunt effect of carboxyhemoglobin. Crit Care Med 1979;
7:54 58.
23. Cane RD, Shapiro BA, Harrison RA, et al. Minimizing errors in intrapulmonary
shunt calculations. Crit Care Med 1980; 8:294 297.
24. Zijlstra WG, Buursma A, Meeuwsen-van der Roest WP. Absorption spectra of
human fetal and adult oxyhemoglobin, de-oxyhemoglobin, carboxyhemoglobin,
and methemoglobin. Clin Chem 1991; 37:1633 1638.
138 Hess
25. Mahoney JJ, Wong RJ, Vreman HJ, Stevenson DK. Fetal hemoglobin of transfused
neonates and spectrophotometric measurements of oxyhemoglobin and carboxy-
hemoglobin. J Clin Monit 1991; 7:154 160.
26. Cornelissen PJH, van Woensel CLM, van Oel WC, de Jong PA. Correction factors
for hemoglobin derivatives in fetal blood, as measured with the IL 282
CO-oximeter. Clin Chem 1983; 29:1555 1556.
27. Sehgal LR, Sehgal HL, Rosen AL, et al. Effect of intralipid on measurements of
total hemoglobin and oxyhemoglobin in whole blood. Crit Care Med 1984;
12:907 909.
28. Pitkin AD, Roberts CM, Wedzicha JA. Arterialized earlobe blood gas analysis:an
underused technique. Thorax 1994; 49:364 366.
29. Dar K, Williams T, Aitken R, et al. Arterial versus capillary sampling for analysing
blood gas pressures. Br Med J 1995; 309:24 25.
30. Sauty A, Uldry C, Debetaz LD, et al. Differences in PO2 and PCO2 between arterial
and arterialized earlobe samples. Eur Respir J 1996; 9:186 189.
31. Hughes JMB. Blood gas estimations from arterialized capillary blood versus arter-
ial puncture: are they different? Eur Respir J 1996; 9:184 185.
32. Giner J, Casan P, Belda J, et al. Pain during arterial puncture. Chest 1996;
110:1443 1445.
33. Harvey MA. Point-of-care laboratory testing in critical care. Am J Crit Care 1999;
8:72 85.
34. Castro HJ, Oropello JM, Halpern N. Point-of-care testing in the intensive care
unit. The intensive care physicians perspective. Am J Clin Pathol 1995;
104:S95 S99.
35. Kost GJ, Ehrmeyer SS, Chernow B, et al. The laboratory-clinical interface. Point-
of-care testing. Chest 1999; 115:1140 1154.
36. Smith BL, Vender JS. Point-of-care testing. Respir Care Clin NA 1995; 1:133 141.
37. Shapiro BA. Point-of-care blood testing and cardiac output measurement in the
intensive care unit. New Horizons 1999; 7:244 252.
38. Zaloga GP, Hill TR, Strickland RA, et al. Bedside blood gas and electrolyte
monitoring in critically ill patients. Crit Care Med 1989; 17:920 925.
39. Zaloga GP, Dudas L, Roberts P, et al. Near-patient blood gas and electrolyte
analyses are accurate when performed by non-laboratory trained individuals.
J Clin Monit 1993; 9:341 346.
40. Hicks JM. Near patient testing: is it here to stay? J Clin Pathol 1996; 49:191 193.
41. Zaloga GP, Roberts PR, Black K, et al. Hand-held blood gas analyzer is accurate in
the critical care setting. Crit Care Med 1996; 24:957 962.
42. MacIntyre NR, Lawlor B, Carsten D, Yetsko D. Accuracy and precision of a
point-or-care blood gas analyzer incorporating optodes. Respir Care 1996;
41:800 804.
43. Wahr JA, Lau W, Tremper KK, et al. Accuracy and precision of a new, portable,
handheld blood gas analyzer, the IRMA. J Clin Monit 1996; 12:317 324.
44. Shelledy DC, Smith WA. A comparison of the I-STAT system and Corning 278 for
the measurement of arterial blood gas values. Respir Care 1997; 42:693 697.
45. Murthy JN, Hicks JM, Soldin SJ. Evaluation of the I-STAT portable clinical
analyzer in a neonatal and pediatric intensive care unit. Clin Biochem 1997;
30:385 389.
46. Kendall J, Reeves B, Clancy M. Point of care testing: randomised controlled trial of
clincial outcome. Br Med J 1998; 316:1052 1057.
47. Shapiro BA. In-vivo monitoring of arterial blood gases and pH. Respir Care 1992;
37:165 169.
48. Peruzzi WT, Shapiro BA. Blood gas monitors. Respir Care Clin NA 1995;
1:143 156.
49. Shapiro BA. Clinical and economic performance criteria for intraarterial and extra-
arterial blood gas monitors, with comparison with in vitro testing. Am J Clin Pathol
1995; 104:S100 S106.
50. Shapiro BA, Cane RD, Chomka CM, Bandala LE, Peruzzi WT. Preliminary evalu-
ation of an intra-arterial blood gas system in dogs and humans. Crit Care Med 1989;
17:455 460.
51. Zimmerman J, Dellinger RP. Initial evaluation of a new intra-arterial blood gas
system in humans. Crit Care Med 1993; 21:495 500.
52. Shapiro BA, Mahutte CK, Cane RD, et al. Clinical performance of a blood gas
monitor. A prospective, multicenter trial. Crit Care Med 1993; 21:497 494.
53. Mahutte CK, Sasse SA, Chen PA, et al. Performance of a patient-dedicated, on-
demand blood gas monitor in medical ICU patients. Am J Respir Crit Care Med
1994; 150:865 869.
54. Haller M, Kilger E, Breigel J, et al. Continuous intra-arterial blood gas and pH
monitoring in critically ill patients with severe respiratory failure: a prospective,
criterion standard study. Crit Care Med 1994; 22:580 586.
55. Venkatesh B, Clutton TH, Hendry SP. A multiparameter sensor for continous intra-
arterail blood gas monitoring: a prospective evaluation. Crit Care Med 1994;
22:588 594.
56. Larson CP, Vender J, Leiver A. Multisite evaluation of a continuous intraarterial
blood gas monitoring system. Anesthesiology 1994; 81:543 552.
57. Venkatesh B, Clutton-Brock TH, Hendry SP. Evaluation of the Paratrend 7 intra-
vascular blood gas monitor during cardiac surgery: comparison with the C4000
in-line blood gas monitor during cardiopulmonary bypass. J Cardiovasc Vasc
Anesth 1995; 9:412 419.
58. Kurahashi K, Hirose Y, Yamada H, et al. Intra-arterial blood gas moni-
toring system: more accurate values can be obtained. J Clin Monit 1996;
12:141 147.
59. Franklin ML, Peruzzi WT, Moen SG, Shapiro BA. Evaluation of an on-demand,
ex vivo bedside blood gas monitor on pulmonary artery blood gas determination.
Anesth Analg 1996; 83:500 504.
60. Weiss IK, Fink S, Edmunds S, et al. Continuous arterial gas monitoring: initial
experience with the Paratrend 7 in children. Intensive Care Med 1996; 22:1414
1417.
61. Abraham E, Gallagher TJ, Fink S. Clinical evaluation of a multiparameter intra-
arterial blood-gas sensor. Intensive Care Med 1996; 22:507 513.
62. McKinnley BA, Parmley CL. Clinical trial of an ex vivo arterial blood gas monitor.
J Crit Care 1998; 13:190 197.
63. Bailey PL, McJames SW, Cluff NL, et al. Evaluation in volunteers of the Via V-
ABG automated bedside blood gas, chemistry, and hematocrit monitor. J Clin
Monit Comput 1998; 14:339 346.
140 Hess
64. Weiss IK, Fink S, Harrison R, et al. Clinical use of continuous arterial
blood gas monitoring in the pediatric intensive care unit. Pediatrics 1999;
103:440 445.
65. Suter PM, Lindauer JM, Fairley HB, Schlobohm RM. Errors in data derived from
pulmonary artery blood gas values. Crit Care Med 1975; 3:175 181.
66. Mihm F, Feeley TW, Rosenthal M, Rafn TA. The lack of effect of variable blood
withdrawal rates on the measurement of mixed venous oxygen saturation. Chest
1980; 78:452 455.
67. Nelson RD, Wilkins RL, Jacobsen WK, et al. Supranormal PvO2 in the presence of
tissue hypoxia: A case report. Respir Care 1983; 28:191 194.
68. Nowak RM, Martin GB, Carden DL, et al. Selective venous hypercarbia during
human CPR: implications regarding blood ow. Ann Emerg Med 1987; 16:527
530.
69. Weil MH, Rakow EC, Trevion R, et al. Difference in acid-base state between
venous and arterial blood during cardiopulmonary resuscitation. N Engl J Med
1986; 1986:153 156.
70. Tahvanainen J, Meretoja O, Nikki P. Can central venous blood replace mixed
venous blood samples? Crit Care Med 1982; 10:758 761.
71. Paulus DA. Invasive monitoring of respiratory gas exchange: continuous measure-
ment of mixed venous oxygen saturation. Respir Care 1987; 32:535 543.
72. Baraka A, Baroody M, Haroun S, et al. Continuous venous oximetry during cardi-
opulmonary bypass: inuence of temperature changes, perfusion ow, and hemato-
crit levels. J Cardiothorac Anesth 1990; 4:35 38.
73. Cariou A, Monchi M, Dhainaut J. Continuous cardiac output and mixed venous
oxygen saturation monitoring. J Crit Care 1998; 13:198 213.
74. Rouby JJ, Poete R, Bodin L, et al. Three mixed venous saturation catheters in
patients with circulatory shock and respiratory failure. Chest 1990; 98:954 958.
75. Gettinger A, DeTraglia MC, Glass DD. In vivo comparison of two mixed venous
saturation catheters. Anesthesiology 1987; 66:373 375.
76. Karis JH, Lumb PD. Clinical evaluation of the Edwards Laboratories and Oximetrix
mixed venous oxygen saturation catheters. J Cardiothorac Anesth 1988;
2:440 444.
77. Reinhart K, Moser N, Rudolph T, et al. Accuracy of two mixed venous saturation
catheters during long-term use in critically ill patients. Anesthesiology 1988;
69:769 773.
78. Hecker BR, Brown DL, Wilson D. A comparison of two pulmonary artery mixed
venous oxygen saturation catheters during the changing conditions of cardiac
surgery. J Cardiothorac Anesth 1989; 3:269 275.
79. VanWoerkens ECSM, Trouwborst A, Tenbrinck R. Accuracy of mixed venous sat-
uration catheter during acutely induced changes in hematocrit in humans. Crit Care
Med 1991; 19:1025 1029.
80. Scuderi PE, Bowton DL, Meredith JW, et al. A comparison of three pulmonary
artery oximetry catheters in intensive care unit patients. Chest 1992; 102:896 905.
81. Scuderi PE, MacGregor DA, Bowton DL, et al. A laboratory comparison of three
pulmonary artery oximetry catheters. Anesthesiology 1994; 81:245 253.
82. Gattinoni L, Brazzi L, Pelosi P, et al. A trial of goal-oriented hemodynamic therapy
in critically ill patients. N Engl J Med 1995; 333:1025 1032.
83. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in

the treatment of severe sepsis and septic shock. N Engl J Med 2001; 345:1368
1377.
84. Alexander CM, Teller LE, Gross JB. Principles of pulse oximetry: theoretical and
practical considerations. Anesth Analg 1989; 68:368 376.
85. Welch JP, DeCesare R, Hess D. Pulse oximetry: instrumentation and clinical appli-
cations. Respir Care 1990; 35:584 601.
86. Russell GB, Graybeal JM. Accuracy of laminated disposable pulse-oximeter
sensors. Respir Care 1995; 40:728 733.
87. Gerber DR, Santarelli RJ, Scott WE, et al. Evaluation of a protective sheath for dis-
posable pulse oximetry probes. Respir Care 1996; 41:197 201.
88. Salyer JW, Burton K, Lynch J, et al. Adventures in recycling:the reuse of dispo-
sable pulse oximeter probes. Respir Care 1993; 38:1072 1076.
89. Jensen LA, Onyskiw JE, Prasad NGN. Meta-analysis of arterial oxygen saturation
monitoring by pulse oximetry in adults. Heart Lung 1998; 27:387 408.
90. Stoneham MD, Saville GM, Wilson IH. Knowledge about pulse oximetry among
medical and nursing staff. Lancet 1994; 344:1339 1342.
91. Davies G, Gibson AM, Swanney M, et al. Understanding of pulse oximetry among
hospital staff. NZ Med J 2003; 116:U297.
92. Choe H, Tashiro C, Fukumitsu K, et al. Comparison of recorded values from six
pulse oximeters. Crit Care Med 1989; 17:678 681.
93. Taylor MB, Whitwam JG. The accuracy of pulse oximeters. A comparative clinical
evaluation of ve pulse oximeters. Anaesthesia 1988; 43:229 232.
94. Hannhart B, Michalski H, Delorme N, Chapparo G, Pulu JM. Reliability of
six pulse oximeters in chronic obstructive pulmonary disease. Chest 1991;
99:842 846.
95. Severinghaus JW, Naifeh KH, Koh SO. Errors in 14 pulse oximeters during pro-
found hypoxia. J Clin Monit 1989; 5:72 81.
96. Severinghaus JW, Naifeh KH. Accuracy of response of six pulse oximeters to pro-
found hypoxia. Anesthesiology 1987; 67:551 558.
97. Hannhart B, Haberer JP, Saunier C, et al. Accuracy and precision of fourteen pulse
oximeters. Eur Respir J 1991; 4:115 119.
98. Kelleher JF, Ruff RH. The penumbra effect:vasomotion-dependent pulse oximeter
artifact due to probe malposition. Anesthesiology 1989; 71:787 791.
99. Vegfors M, Lennmarken C. Carboxyhaemoglobinaemia and pulse oximetry. Br J
Anaesth 1991; 66:625 626.
100. Barker SJ, Tremper KK. The effect of carbon monoxide inhalation on pulse oxime-
try and transcutaneous PO2. Anesthesiology 1987; 66:677 679.
101. Raemer DB, Elliott WR, Topulos GP, Philip JH. The theoretical effect of carbox-
yhemoglobin on the pulse oximeter. J Clin Monit 1989; 5:246 249.
102. Hampson NB. Pulse oximetry in severe carbon monoxide poisoning. Chest 1998;
114:1036 1041.
103. Barker SJ, Tremper KK, Hyatt J. Effects of methemoglobinemia on pulse oximetry
and mixed venous oximetry. Anesthesiology 1989; 70:112 117.
104. Kress JP, Pohlman AS, Hall JB. Determinants of hemoglobin saturation in patients
with acute sickle chest syndrome. A comparison of arterial blood gases and pulse
oximetry. Chest 1999; 115:1316 1320.
142 Hess
105. Scott DM, Cooper MG. Spurious pulse oximetry with intrauterine methylene blue
injection. Anaesth Intens Care 1991; 19:267 284.
106. Scheller MS, Unger RJ, Kelner MJ. Effects of intravenously administered dyes on
pulse oximeter readings. Anesthesiology 1986; 65:550 552.
107. Cote CJ, Goldstein EA, Fuchsman WH, et al. The effect of nail polish on pulse oxi-
metry. Anesth Analg 1989; 67:683 686.
108. Veyckemans F, Baele P, Guillaume JE, et al. Hyperbilirubinemia does not interfere
with hemoglobin saturation measured by pulse oximetry. Anesthesiology 1989;
70:118 122.
109. Chelluri L, Snyder JV, Bird JR. Accuracy of pulse oximetry in patients with hyper-
bilirubinemia. Respir Care 1991; 36:1383 1386.
110. Zeballos RJ, Wiesman IM. Reliability of noninvasive oximetry in black subjects
during exercise and hypoxia. Am Rev Respir Dis 1991; 144:1240 1244.
111. Jubran A, Tobin MJ. Reliability of pulse oximetry in titrating supplemental oxygen
therapy in ventilator-dependent patients. Chest 1990; 97:1420 1425.
112. Cahan C, Decker MJ, Hoekje PL, et al. Agreement between noninvasive oximetric
values for oxygen saturation. Chest 1990; 97:814 819.
113. Severinghaus JW, Koh SO. Effect of anemia on pulse oximeter accuracy at low
saturation. J Clin Monit 1990; 6:85 88.
114. Lee S, Tremper KK, Barker SJ. Effects of anemia on pulse oximetry and continuous
mixed venous hemoglobin saturation monitoring in dogs. Anesthesiology 1991;
75:118 122.
115. Fluck RR, Schroeder C, Frani G, et al. Does ambient light affect the accuracy of
pulse oximetry? Respir Care 2003; 48:677 680.
116. Sami HM, Kleinman BS, Lonchyna VA. Central venous pulsations associated with
a falsely low oxygen saturation measured by pulse oximetry. J Clin Monit 1991;
7:309 312.
117. Wilkins CJ, Moores M, Hanning CD. Comparison of pulse oximeters:
effects of vasoconstriction and venous engorgement. Br J Anaesth 1989; 62:439
444.
118. Stewart KG, Rowbottom SJ. Inaccuracy of pulse oximetry in patients with severe
tricuspid regurgitation. Anaesthesia 1991; 46:668 670.
119. Chemello PD, Nelson SR, Wolford LM. Finger injury resulting from pulse
oximeter probe during orthognathic surgery. Oral Surg Oral Med Oral Pathol
1990; 69:161 163.
120. Murphy KG, Secunda JA, Rockoff MA. Severe burns from a pulse oximeter.
121. Stogner SW, Owens MW, Baethge BA. Cutaneous necrosis and pulse oximetry.
Cutis 1191; 48:235 137.
122. Bashein G, Syrovy G. Burns associated with pulse oximetry during magnetic
resonance imaging. Anesthesiology 1991; 75:382 383.
123. Plummer JL, Ilsley AH, Fronsko RRL, et al. Identication of movement
artefact by the Nellcor N-200 and N-3000 pulse oximeters. J Clin Monit 1997;
13:109 113.
124. Gehring H, Nornberger C, Matz H, et al. The effects of motion artifact and low per-
fusion on the performance of a new generation of pulse oximeters in volunteers
undergoing hypoxemia. Respir Care 2002; 47:48 60.
125. Goldman JM, Petterson MT, Kopotic RJ, et al. Masimo signal extraction pulse oxi-
metry. J Clin Monit 2000; 16:475 483.
126. Barker SJ. Motion-Resistant pulse oximetry: a comparison of new and old
models. Anesth Analg 2002; 95:967 972.
127. Durbin CG, Rostow SK. More reliable oximetry reduces the frequency of arterial
blood gas analyses and hastens oxygen weaning after cardiac surgery: a prospec-
tive, randomized trial of the clinical impact of a new technology. Crit Care Med
2002; 30:1734 1740.
128. Moller JT, Pedersen T, Rasmussen LS, et al. Randomized evaluation of pulse oxi-
metry in 20,802 patients. I. Design, demography, pulse oximetry failure rate, and
overall complication rate. Anesthesiology 1993; 78:436 444.
129. Moller JT, Johannessen NW, Espersen K, et al. Randomized evaluation of pulse
oximetry in 20,802 patients. II. Perioperative events and postoperative compli-
cations. Anesthesiology 1993; 78:445 453.
130. Hartert TV, Wheeler AP, Sheller JR. Use of pulse oximetry to recognize severity of
airow obstruction in obstructive airway disease: correlation with pulsus para-
doxus. Chest 1999; 115:475 481.
131. Gys T, Hubens A, Neels H, et al. Prognostic value of gastric intramucosal pH in
surgical intensive care patients. Crit Care Med 1988; 16:1222 1224.
132. Landow L, Phillips DA, Heard SO, et al. Gastric tonometry and venous oximetry in
cardiac surgery patients. Crit Care Med 1991; 19:1226 1233.
133. Hartmann M, Montgomery A, Jonsson K, et al. Tissue oxygenation in hemorrhagic
shock measured as transcutaneous oxgyen tension, subcutaneous oxygen
tension, and gastrointestinal intramucosal pH in pigs. Crit Care Med 1991;
19:205 210.
134. Silverman HJ, Tuma P. Gastric tonometry in patients with sepsis. Effects of dobu-
tamine infusions and packed red blood cell transfusions. Crit Care Med 1192;
102:184 188.
135. Gutierrez G, Bismar H, Danzker DR, et al. Comparison of gastric intramucosal pH
with measures of oxygen transport and consumption in critically ill patients. Crit
Care Med 1992; 20:451 457.
136. Gutierrez G, Palizas F, Doglio G, et al. Gastric intramucosal pH as a therapeutic
index of tissue oxygenation in critically ill patients. Lancet 1992; 339:195 199.
137. Roumen RMH, Vreugde JPC, Goris RJA. Gastric tonometry in multiple patient
patients. J Trauma 1994; 36:313 316.
138. Mohsenifar Z, Hay A, Hay J, et al. Gastric intramucosal pH as a predictor of success
or failure in weaning patients from mechanical ventilation. Ann Intern Med 1993;
119:794 798.
139. Marik PE. Gastric intramucosal pH. A better predictor of multiorgan dysfunction
syndrome and death than oxygen-derived variables in patients with sepsis. Chest
1993; 104:225 229.
140. Knichwitz G, Van Aken H, Brussel T. Gastrointestinal monitoring using measure-
ment of intramucosal PCO2. Anesth Analg 1998; 87:134 142.
141. Keenan SP, Guyatt GH, Sibbald WJ, et al. How to use articles about diagnostic
technology: Gastric tonometry. Crit Care Med 1999; 27:1726 1731.
142. Fiddian-Green RG. Associations between intramucosal acidosis in the gut and
organ failure. Crit Care Med 1993; 21:S103 S107.
144 Hess
143. Fiddian-Green RG, McGough E, Pittenger G, et al. Predictive value of intramural

pH and other risk factors for massive bleeding from stress ulceration. Gastroenter-
ology 1983; 85:613 620.
144. Levy B, Perrigault PF, Gawalkiewicz P, et al. Gastric versus duodenal feeding and
gastric tonometrics measurements. Crit Care Med 1998; 26:1991 1994.
145. Miller PR, Kincaid EH, Meredith JW, et al. Threshold values of intramucosal pH
and mucosal-arterial CO2 gap during shock resuscitation. J Trauma 1998;
45:868 872.
146. Levy B, Gawalkiewicz P, Vallet B, et al. Gastric capnometry with air-automated
tonometry predicts outcome in critically ill patients. Crit Care Med 2003;
31:474 480.
147. Knichwitz G, Kuhmann M, Brodner G, et al. Gastric tonometry: precision and
reliability are improved by a phosphate buffered solution. Crit Care Med 1996;
24:512 516.
148. Takala J, Parviainen I, Siloajho M, et al. Saline PCO2 is an important source of
error in the assessment of gastric intramucosal pH. Crit Care Med 1995;
23:1608 1610.
149. Riddington D, Venkatesh B, Clutton-Brock T, et al. Measured carbon dioxide
tension in saline and alternative solutions: quantication of bias and precision in
two blood gas analyzers. Crit Care Med 1994; 22:96 100.
150. Creteur J, de Backer D, Vincent J. Monitoring gastric intramucosal carbon dioxide
pressure using gas tonometry. In vitro and in vivo validation studies. Anesthesio-
logy 1997; 87:504 510.
151. Guzman JA, Kruse JA. Development and validation of a technique for continuous
monitoring of gastric intramucosal pH. Am J Respir Crit Care Med 1996; 153:
694 700.
152. Knichwitz G, Rotker J, Brussel T, et al. A new method for continuous
intramucosal PCO2 measurement in the gastrointestinal tract. Anesth Analg 1996;
83:6 11.
153. Gomersall CD, Joynt GM, Ho KM, et al. Gastric tonometry and prediction of
outcome in the critically ill: Arterial to intramucosal pH gradient and carbon
dioxide gradient. Anaesthesia 1997; 52:619 623.
154. Brown SD, Gutierrez G. Does gastric tonometry work? Yes. Crit Care Clin 1996;
12:569 585.
155. Benjamin E, Oropello JM. Does gastric tonometry work? No. Crit Care Clin 1996;
12:587 560.
156. Weil MH, Nakagawa Y, Tang W, et al. Sublingual capnometry: a new noninvasive
measurement for diagnosis and quantitation of severity of circulatory shock. Crit
Care Med 1999; 27:1225 1229.
157. Marik PE, Bankov A. Sublingual capnometry versus traditional markers of tissue
oxygenation in critically ill patients. Crit Care Med 2003; 31:818 822.
158. Marik PE: Sublingual capnography: A clinical validation study. Chest 2001;
120:923 927.
159. Hess DR. Capnometry. IN: Tobin MJ. Principles and practice of intensive care
monitoring. McGraw-Hill, New York, 1998.
160. Colman Y, Krauss B. Microstream capnography technology: A new approach to an
old problem. J Clin Monit 1999; 15:403 409.
161. Fletcher R, Jonson B, Cumming G, et al. The concept of deadspace with

special reference to the single breath test for carbon dioxide. Br J Anaesth 1981;
53:77 88.
162. Fletcher R, Jonson B. Deadspace and the single breath test for carbon dioxide
during anaesthesia and articial ventilation. Br J Anaesth 1984; 56:109 119.
163. Hoffbrand BI. The expiratory capnogram: a measure of ventilation-perfusion
inequalities. Thorax 1966; 21:518 523.
164. Jones NL, McHardy GJR, Naimark A, et al. Physiological dead space and alveolar-
arterial gas pressure differences during exercise. Clin Sci 1966; 31:19 29.
165. Nunn JF, Hill DW. Respiratory dead space and arterial to end-tidal CO2 tension
difference in anesthetized man. J Appl Physiol 1960; 15:383 389.
166. Severinghaus JW, Stupfel MA, Bradley AF. Alveolar dead space and arterial to
end-tidal carbon dioxide differences during hypothermia in dog and man. J Appl
Physiol 1957; 10:349 355.
167. Stockwell MA, Bruce W, Soni N. The inuence of CO2 production and physiologi-
cal deadspace on end-tidal CO2 during controlled ventilation: a study using a mech-
anical model. Anaesth Intensive Care 1989; 17:482 486.
168. Stromberg NO, Gustafsson PM. Ventilation inhomogeneity by nitrogen washout
and ventilation-perfusion mismatch by capnography in stable and induced airway
obstruction. Pediatr Pulmonol 2000; 29:94 102.
169. Sikand R, Cerretelli P, Farhi LE. Effects of VA and VA/Q distribution and of time
on the alveolar plateau. J Appl Physiol 1966; 21:518 523.
170. Yaron M, Padyk P, Hutsinpiller M, et al. Utility of expiratory capnogram in the
assessment of bronchospasm. Ann Emerg Med 1996; 28:403 407.
171. You B, Peslin R, Duvivier C, Vu VD, Grilliat JP. Expiratory capnography in
asthma: evaluation of various shape indices. Eur Respir J 1994; 7:318 323.
172. Nuckton TJ, Alonso JA, Kallet RH, et al. Pulmonary dead-space fraction as a risk
factor for death in the acute respiratory distress syndrome. N Engl J Med 2002;
346:1281 1286.
173. Hubble CL, Gentile MA, Tripp DS, et al. Deadspace to tidal volume ratio predicts
successful extubation in infants and children. Crit Care Med 2000; 28:2034 2040.
174. Patel AJ, Choi C, Giuffrida JG. Changes in end tidal CO2 and arterial blood gas
levels after release of tourniquet. S Med J 1987; 80:214 216.
175. Breen PH, Mazumdar B, Skinner SC. Comparison of end-tidal PCO2 and average
alveolar expired PCO2 during positive end-expiratory pressure. Anesth Analg
1996; 82:368 373.
176. Graybeal JM, Russell GB. Capnometry in the surgical ICU: an analysis
of the arterial-to-end-tidal carbon dioxide difference. Respir Care 1993; 38:923
928.
177. Hatle L, Rokeseth R. The arterial to end-expiratory carbon dioxide tension gradient
in acute pulmonary embolism and other cardiopulmonary diseases. Chest 1974;
66:352 357.
178. Hoffman RA, Krieger BP, Kramer MR, et al. End-tidal carbon dioxide in critically
ill patients during changes in mechanical ventilation. Am Rev Respir Dis 1989;
140:1265 1268.
179. Perrin F, Perrot D, Holzapfel L, et al. Simultaneous variations of PaCO2 and PaCO2
in assisted ventilation. Br J Anaesth 1983; 55:525 530.
146 Hess
180. Poppius H, Korhonen O, Viljanen AA, et al. Arterial to end-tidal CO2 difference in
respiratory disease. Scand J Resp Dis 1975; 56:254 262.
181. Raemer DB, Francis D, Philip JH, et al. Variation in PCO2 between arterial blood
and peak expired gas during anesthesia. Anesth Analg 1983; 62:1065 1069.
182. Russell GB, Graybeal JM, Strout JC. Stability of arterial to end-tidal carbon dioxide
gradients during postoperative cardiorespiratory support. Can J Anaesth 1990;
37:560 566.
183. Russell GB, Graybeal JM. Reliability of the arterial to end-tidal carbon dioxide gra-
dient in mechanically ventilated patients with multisystem trauma. J Trauma 1994;
36:317 322.
184. Russell GB, Graybeal JM. The arterial to end-tidal carbon dioxide difference in
neurosurgical patients during craniotomy. Anesth Analg 1995; 81:806 810.
185. Yamanaka MK, Sue DY. Comparison of arterial-end-tidal PCO2 difference and
dead space/tidal volume ratio in respiratory failure. Chest 1987; 92:832 835.
186. Piiper J. Blood-gas equilibrium of carbon dioxide in the lungs: a continuing contro-
versy. J Appl Physiol 1986; 60:1 8.
187. Tulou PP, Walsh PW. Measurement of alveolar carbon dioxide tension at maximal
expiration as an estimate of arterial carbon dioxide tension in patients with airway
obstruction. Am Rev Respir Dis 1970; 102:921 926.
188. Bhavani-Shankar K, Kumar AY, Moseley HS, et al. Terminology and the
current limitations of time capnography: a brief review. J Clin Monit 1995;
11:175 82.
189. Jones NL, Robertson DG, Kane JW. Difference between end-tidal and arterial
PCO2 in exercise. J Appl Physiol 1979; 47:954 960.
190. Birmingham PK, Cheney FW, Ward RJ. Esophageal intubation: a review of detec-
tion of techniques. Anesth Analg 1986; 65:886 891.
191. Guggenberger H, Lenz G, Federle R. Early detection of inadvertent oesophageal
intubation: pulse oximetry vs. capnography. Acta Anaesthesiol Scand 1989;
33:112 115.
192. Holland R, Webb RK, Runciman WB. The Australian Incident Monitoring Study.
Oesophageal intubation: an analysis of 2000 incident reports. Anaesth Intensive
Care 1993; 21:608 610.
193. Linko K, Paloheimo M, Tammisto T. Capnography for detection of accidental
oesophageal intubation. Acta Anaesthesiol Scand 1983; 27:199 202.
194. Murray IP, Modell JH. Early detection of endotracheal tube accidents by monitor-
ing carbon dioxide concentration in respiratory care. Anesthesiology 1983;
59:344 346.
195. Szekely SM, Webb RK, Williamson JA, et al. Problems related to the endo-
tracheal tube: an analysis of 2000 incident reports. Anaesth Intens Care 1993;
21:611 616.
196. Vaghadia H, Jenkins LC, Ford RW. Comparison of end-tidal carbon dioxide,
oxygen saturation and clinical signs for the detection of oesophageal intubation.
Can J Anaesth 1989; 36:560 564.
197. Garnett AR, Gervin CA, Gervin AS. Capnographic waveforms in esophageal
intubation: effect of carbonated beverages. Ann Emerg Med 1989; 18:387 390.
198. Araujo-Preza CE, Melhado ME, Gutierrez FJ, et al. Use of capnometry to verify
feeding tube placement. Crit Care Med 2002; 30:2255 2259.
199. Bosman RJ, Stoutenbeek CP, Zandstra DF. Non-invasive pulmonary blood ow
measurement by means of CO2 analysis of expiratory gases. Intensive Care Med
1991; 17:98 102.
200. Isserles SA, Breen PH. Can changes in end-tidal PCO2 measure changes in cardiac
output? Anesth Analg 1991; 73:808 814.
201. Ornato JP, Garnett AR, Glauser FL. Relationship between cardiac output and the
end-tidal carbon dioxide tension. Ann Emerg Med 1990; 19:1104 1106.
202. Weil MH, Bisera J, Trevino RP, et al. Cardiac output and end-tidal carbon dioxide.
Crit Care Med 1985; 13:907 909.
203. Sanders AB, Atlas M, Ewy GA, et al. Expired PCO2 as an index of coronary per-
fusion pressure. Am J Emerg Med 1985; 3:147 149.
204. Lewis LM, Stothert J, Standeven J, et al. Correlation of end-tidal CO2 to cerebral
perfusion during CPR. Ann Emerg Med 1992; 21:1131 1134.
205. Kalenda Z. The capnogram as a guide to the efcacy of cardiac massage. Resusci-
tation 1978; 6:259 263.
206. Garnett AR, Ornato JP, Gonzalez ER, Johnson EB. End-tidal carbon dioxide
monitoring during cardiopulmonary resuscitation. J Am Med Assoc 1987;
257:512 515.
207. Falk JL, Rackow EC, Weil MH. End-tidal carbon dioxide concentration during
cardiopulmonary resuscitation. N Engl J Med 1988; 318:607 611.
208. Sanders AB, Kern KB, Otto CW, et al. End-tidal carbon dioxide monitoring during
cardiopulmonary resuscitation. A prognostic indicator for survival. J Am Med
Assoc 1989; 262:1347 1351.
209. Levine RL, Wayne MA, Miller CC. End-tidal carbon dioxide and outcome of out-
of-hospital cardiac arrest. N Engl J Med 1997; 337:301 306.
210. Grmec S, Lah K, Tusek-Bunc K. Difference in end-tidal CO2 between asphyxia
cardiac arrest and ventricular brillation/pulseless ventricular tachycardia
cardiac arrest in the prehospital setting. Crit Care 2003; 7:R139 R144.
211. Julian DG, Travis DM, Robin ED, et al. Effect of pulmonary artery occlusion upon
end-tidal CO2 tension. J Appl Physiol 1960; 15:87 91.
212. Bermudez J, Lichtiger M. Increases in arterial to end-tidal CO2 tension differences
after cardiopulmonary bypass. Anesth Analg 1987; 66:690 692.
213. Poppius H, Korhonen O, Viljanen AA, et al. Arterial to end-tidal CO2 difference in
respiratory disease. Scand J Resp Dis 1975; 56:254 262.
214. Hatle L, Rokeseth R. The arterial to end-expiratory carbon dioxide tension gradient
in acute pulmonary embolism and other cardiopulmonary diseases. Chest 1974;
66:352 357.
215. Nutter DO, Massumi RA. The arterial-alveolar carbon dioxide tension gradient in
diagnosis of pulmonary embolus. Dis Chest 1966; 50:380 387.
216. Eriksson L, Wollmer P, Olsson C, et al. Diagnosis of pulmonary embolism based
upon alveolar dead space analysis. Chest 1989; 96:357 362.
217. Chopin C, Fesard P, Mangalaboyi J, et al. Use of capnography in diagnosis of pul-
monary embolism during acute respiratory failure of chronic obstructive pulmonary
disease. Crit Care Med 1990; 18:353 357.
218. Verschuren F, Liistro G, Coffeng R, et al. Volumetric capnography as a scree-
ning test for pulmonary embolism in the emergency department. Chest 2004;
125:841 850.
148 Hess
219. Healey CJ, Fedullo AJ, Swinburne AJ, Wahl GW. Comparison of noninvasive
measurements of carbon dioxide tension during withdrawal from mechanical ven-
tilation. Crit Care Med 1987; 15:764 767.
220. Weinger MB, Brimm JE. End-tidal carbon dioxide as a measure of arterial carbon
dioxide during intermittent mandatory ventilation. J Clin Monit 1987; 3:73 79.
221. Niehoff J, DelGuercio C, LaMorte W, et al. Efcacy of pulse oximetry and capno-
metry in postoperative ventilatory weaning. Crit Care Med 1988; 16:701 705.
222. Smith RA, Novak RA, Venus B. End-tidal CO2 monitoring utility during weaning
from mechanical ventilation. Respir Care 1989; 34:972 975.
223. Thrush DN, Mentis SW, Downs JB. Weaning with end-tidal CO2 and pulse oxime-
try. J Clin Anesth 1991; 3:456 460.
224. Withington DE, Ramsay JG, Saoud AT, et al. Weaning from ventilation after car-
diopulmonary bypass:evaluation of a non-invasive technique. Can J Anaesth 1991;
38:15 19.
225. Hess D, Schlottag A, Levin B, et al. An evaluation of the usefulness of end-tidal
PCO2 to aid weaning from mechanical ventilation following cardiac surgery.
Respir Care 1991; 36:837 843.
226. Morley TF, Giaimo J, Maroszan E, et al. Use of capnography for assessment of
the adequacy of alveolar ventilation during weaning from mechanical ventilation.
Am Rev Respir Dis 1993; 148:339 344.
227. Saura P, Blanch L, Lucangelo U, et al. Use of capnometry to detect hypercapnic
episodes during weaning from mechanical ventilation. Intensive Care Med 1966;
22:374 381.
228. Dasgupta A, Rice R, Mascha E, et al. Four-year experience with a unit for long-term
ventilation (respiratory special care unit) at the Cleveland Clinic Foundation. Chest
1999; 116:447 455.
229. Mackersie RC, Karagianes TG. Use of end-tidal carbon dioxide tension for
monitoring induced hypocapnia in head-injured patients. Crit Care Med 1990;
18:764 765.
230. Sharma SK, McGuire GP, Cruise CJ. Stability of the arterial to end-tidal carbon
dioxide difference during anaesthesia for prolonged neurosurgical procedures.
Can J Anaesth 1995; 42:498 503.
231. Isert P. Control of carbon dioxide levels during neuroanaesthesia: current practice
and an appraisal of our reliance upon capnography. Anaesth Intensive Care 1994;
22:435 441.
232. Christensen MA, Bloom J, Sutton KR. Comparing arterial and end-tidal carbon
dioxide values in hyperventilated neurosurgical patients. Am J Crit Care 1995;
4:116 121.
233. Kerr ME, Zwmpsky J, Sereika S, et al. Relationship between arterial carbon dioxide
and end-tidal carbon dioxide in mechanically ventilated adults with severe head
trauma. Crit Care Med 1996; 24:785 790.
234. Palmon SC, Liu M, Moore LE, et al. Capnography facilitates tight control of ven-
tilation during transport. Crit Care Med 1996; 24:608 611.
235. Helm M, Schuster R, Hauke J, et al. Tight control of prehospital ventilation by cap-
nography in major trauma patients. Br J Anaesth 2003; 90:327 332.
236. Tachibana K, Imanaka H, Takeuchi M, et al. Noninvasive cardiac output measure-
ment using partial carbon dioxide rebreathing is less accurate at settings of reduced
minute ventilation and when spontaneous breathing is present. Anesthesiology

2003; 98:830 837.
237. Tachibana K, Imanaka H, Miyano H, et al. Effect of ventilatory settings on accuracy
of cardiac output measurement using partial CO2 rebreathing. Anesthesiology
2002; 96:96 102.
238. Yem JS, Tang Y, Turner MJ, et al. Sources of error in noninvasive pulmonary blood
ow measurements by partial rebreathing. A computer model study. Anesthesiol-
ogy 2003; 98:881 887.
239. de Abreu MG, Geiger S, Winkler T, et al. Evaluation of a new device for noninva-
sive measurement of nonshunted pulmonary capillary blood ow in patients with
acute lung injury. Intensive Care Med 2002; 28:318 323.
240. Odenstedt H, Stenqvist O, Lundin S. Clinical evaluation of a partial CO2
rebreathing technique for cardiac output monitoring in critically ill patients. Acta
Anaesthesiol Scand 2002; 46:152 159.
241. Mielck F, Buhre W, Hanekop G, et al. Comparison of continuous cardiac output
measurements in patients after cardiac surgery. J Cardiothoracic Vasc Anesth
2003; 17:211 216.
242. Maxwell RA, Gibson JB, Slade JB, et al. Noninvasive cardiac output by partial CO2
rebreathing after severe chest trauma. J Trauma 2001; 51:849 853.
243. Brandi LS, Bertolini S, Pieri M, et al. Comparison between cardiac output measured
by thermodilution technique and calculated by O2 and modied CO2 ck methods
using a new metabolic monitor. Intensive Care Med 1997; 23:908 915.
244. Murias GE, Villagra A, Vatua S, et al. Evaluation of a noninvasive method for
cardiac output measurement in critical care patients. Intensive Care Med 2002;
28:1470 1474.
245. de Abreu MG, Winkler T, Pahlitzsch T, et al. Performance of the partial CO2
rebreathing technique under different hemodynamic and ventilation/perfusion
matching conditions. Crit Care Med 2003; 31:543 551.
246. Hess DR, Medoff MD, Fessler MB. Pulmonary mechanics and graphics during
positive pressure ventilation. Int Anesthesiol Clin 1999; 37(3):15 34.
247. Marini JJ, Crooke PS. A general mathematical model for respiratory dynamics rel-
evant to the clinical setting. Am Rev Respir Dis 1993; 147:14 24.
248. Chatburn RL. Classication of mechanical ventilators. In: Branson RD, Hess DR,
Chaburn RL, eds. Respiratory Care Equipment. Philadelphia: Lippincott Williams
& Wilkins, 1999.
249. Dhand R, Jubran A, Tobin MJ. Bronchodilator delivery by metered-dose inhaler in
ventilator-supported patients. Am J Respir Crit Care Med 1995; 151:1827 1833.
250. Truwit JD, Marini E. Evaluation of thoracic mechanics in the ventilated patient.
Part I: Primary measurements. J Crit Care 1988; 3:133 150.
251. Anonymous. 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; 342:1301 1308.
252. Ranieri VM, Grasso S, Fiore T, et al. Auto-positive end-expiratory pressure and
dynamic hyperination. Clin Chest Med 1996; 17:379 394.
253. Massimo G, Misuri G, Duranti R, et al. Abdominal muscle recruitment and PEEP
during bronchoconstriction in chronic obstructive pulmonary disease. Thorax 1997;
52:355 361.
150 Hess
254. Lessard MR, Lofaso F, Brochard L. Expiratory muscle activity increases intrinsic
positive end-expiratory pressure independently of dynamic hyperination in
mechanically ventilated patients. Am J Respir Crit Care Med 1995; 151:562 569.
255. Ninane V, Yerault J, de Troyer A. Intrinsic PEEP in patients with chronic obstruc-
tive pulmonary disease. Role of expiratory muscles. Am Rev Respir Dis 1993;
148:1037 1042.
256. Leatherman JW, Ravenscraft SA. Low measured auto-positive end-expiratory
pressure during mechanical ventilation of patients with severe asthma: hidden
auto-positive end-expiratory pressure. Crit Care Med 1996; 24:541 546.
257. Truwit JD, Marini JJ. Evaluation of thoracic mechanics in the ventilated patient.
Part II: applied mechanics. J Crit Care 1988; 3:199 213.
258. Chieveley-Williams S, Dinner L, Puddicombe A, et al. Central venous and bladder
pressure reect transdiaphragmatic pressure during pressure support ventilation.
Chest 2002; 121:533 538.
259. Parthasarathy S, Jubran A, Tobin MJ. Cycling of inspiratory and expiratory muscle
groups with the ventilator in airow limitation. Am J Respir Crit Care Med 1998;
158:1471 1478.
260. Jubran A, Van de Graaff WB, Tobin MJ. Variability of patient-ventilator interaction
with pressure support ventilation in patients with chronic obstructive pulmonary
disease. Am J Respir Crit Care Med 1995; 152:129 136.
261. Marini JJ, Crooke PS, Truwit JD. Determinants and limits of pressure-preset
ventilation: a mathematical model. J Appl Physiol 1989; 67:1081 1092.
262. Burke WC, Crooke PS, Marcy TW, et al. Comparison of mathematical and mech-
anical models of pressure-controlled ventilation. J Appl Physiol 1993; 74:922 933.
263. Kress JP, OConnor MF, Schmidt GA. Clinical examination reliably detects intrin-
sic positive end-expiratory pressure in critically ill, mechanically ventilated
patients. Am J Respir Crit Care Med 1999; 159:290 294.
264. Hess D, Tabor T. Comparison of six methods to calculate airway resistance during
mechanical ventilation. J Clin Monit 1993; 9:275 282.
265. Marini JJ, Ravenscraft SA. Mean airway pressure: physiologic determinants and
clinical importance. Part 1: physiologic determinants and measurements. Crit
Care Med 1992; 20:1461 1472.
266. Marini JJ, Ravenscraft SA. Mean airway pressure: physiologic determinants and
clinical importance. Part 2: clinical implications. Crit Care Med 1992;
20:1604 1616.
267. Primiano FP, Chatburn RL, Lough MD. Mean airway pressure: theoretical con-
siderations. Crit Care Med 1982; 10:378 383.
268. Brochard L. Respiratory pressure-volume curves. In: Tobin MJ, ed. Principles and
Practice of Intensive Care Monitoring. New York: McGraw-Hill, 1998.
269. Lu Q, Vieira SRR, Richecoeur J, et al. A simple automated method for measuring
pressure volume curves during mechanical ventilation. Am J Respir Crit Care Med
1999; 159:275 282.
270. Servillo G, Svantesson C, Beydon L, et al. Pressure volume curves in acute respir-
atory failure: automated low ow ination versus occlusion. Am J Respir Crit Care
Med 1997; 155:1629 1656.
271. Bigatello LM, Hurford WE, Pesenti A. Ventilatory management of severe acute
respiratory failure for Y2K. Anesthesiology 1999; 91:1567 70.
272. OKeefe GE, Gentilello LM, Erford S, et al. Imprecision in lower inection point
estimation from static pressure-volume curves in patients at risk for acute respirat-
ory distress syndrome. J Trauma 1998; 44:1064 1068.
273. Harris RS, Hess DR, Venegas JG. An objective analysis of the pressure volume
curve in the acute respiratory distress syndrome. Am J Respir Crit Care Med.
2000; 161:432 439.
274. Venegas JG, Harris RS, Simon BA. A comprehensive equation for pulmonary
pressure volume curve. J Appl Physiol 1998; 84:389 395.
275. Ranieri VM, Brienza N, Santostasi S, et al. Impairments of lung and chest wall
mechanics in patients with acute respiratory distress syndrome. Am J Respir Crit
Care Med 1997; 156:1082 1091.
276. Mergoni M, Martelli A, Volpi A, et al. Impact of positive end-expiratory pressure
on chest wall and lung pressure volume curve in acute respiratory failure. Am J
Respir Crit Care Med 1997; 156:846 854.
277. Hickling KG. The pressure-volume curve is greatly modied by recruitment. Am J
278. Hickling KG. Best compliance during a decremental, but not incremental, positive
end-expiratory pressure trial is related to open-lung positive end-expiratory
pressure: a mathematical model of acute respiratory distress syndrome lungs.
Am J Respir Crit Care Med 2001; 163:69 78.
279. Morrison WE, Haas EC, Shaffner DH, et al. Noise, stress, and annoyance in a pedi-
atric intensive care unit. Crit Care Med 2003; 31:113 119.
280. Chambrin MC. Alarms in the intensive care unit: how can the number of false
alarms be reduced? Critical Care 2001; 5:184 188.
281. Lawless ST. Crying wolf: false alarms in a pediatric intensive care unit. Crit Care
Med 1994; 22:981 985.
282. Tsein CL, Fackler JC. Poor prognosis for existing monitors in the intensive care
unit. Crit Care Med 1997; 25:614 619.
283. Bentt LR, Santora TA, Laverle BJ, et al. Accuracy and utility of pulse oximetry in
the surgical intensive care unit. Curr Surg 1990; 47:267 268.
284. Kestin IG, Miller BR, Lockhart CH. Auditory alarms during anesthesia monitoring.
285. Kellerman AL, Cofer CA, Joseph S, et al. Impact of portable pulse oximetry on
arterial blood gas test ordering in an urban emergency department. Ann Emerg
Med 1991; 20:130 134.
286. Orlando R. Continuous mixed venous oximetry in critically ill patients. High-tech
cost effectiveness. Arch Surg 1986; 121:470 471.
287. Pearson KS, Gomez MN, Moyers JR, et al. A cost/benet analysis of randomized
invasive monitoring for patients undergoing cardiac surgery. Anesth Analg 1989;
69:336 341.
288. Jastremski MS, Chelluri L, Beney KM, et al. Analysis of the effects of continuous
on-line monitoring of mixed venous oxygen saturation on patient outcome and cost
effectiveness. Crit Care Med 1990; 17:148 153.
5
Weaning from Mechanical Ventilation
DANI HACKNER SHARAD DASS and MICHAEL I. LEWIS
Transitional Critical Care Service Division of Pulmonary and Critical Care

CedarsSinai Medical Center and Medicine, Cedars Sinai Medical Center
Geffen School of Medicine at UCLA and Geffen School of Medicine at UCLA
Los Angeles, California, U.S.A. Los Angeles, California, U.S.A.
I. Introduction
The process of weaning (liberating) patients from mechanical ventilation

occupies a considerable portion of the total ventilator time, with estimates
ranging from 40% to 60% (1). This chapter provides a comprehensive overview
of important key topics related to weaning, such as pathophysiologic insights into
weaning failure, predicting and optimizing weaning success, and methods of
weaning, including protocol-driven approaches to weaning. A separate section
addresses sedation in relation to weaning. Weaning from chronic mechanical
ventilation includes the topics of tracheostomy and airway devices. A guide to
discussing goals of therapy for the ventilated patient is also addressed.
This chapter spans the mechanisms underlying weaning to the practical
clinical steps for liberation from ventilation. The rst section addresses patho-
physiologic processes that inform the weaning process and failure to wean.
The second section, on Predicting Weaning Success, addresses the strengths
and weaknesses of indices developed and attempts to place them in a clinical per-
spective. The third section, Protocols and Pathways, describes efforts to
systematize the weaning process. The fourth section describes one tested protocol
153
154 Hackner, Dass, and Lewis
for Weaning from Prolonged Mechanical Ventilation. The fth section encom-
passes topics of comfort, analgesia, sedation, and delirium during weaning,
which have great impact on duration of ventilation and patient satisfaction.
The sixth and nal section, Goals of Therapy: Conversations about Ventilation,
provides a bioethical rubric for initiation or terminating ventilation and weaning.
II. Pathophysiologic Factors Determining Weaning

Success or Failure: An Introductory Overview
A. Functional Force Reserve of the Respiratory Muscles
The concepts underlying a sustainable level of force by the respiratory muscles

are keys to appreciating those factors determining the functional force reserve
and endurance capacity of the respiratory pump. Skeletal muscles, including
the respiratory muscles, can sustain repetitive loads for prolonged periods of
time or indenitely, provided the forces generated during recruitment are
below a critical threshold percentage of the maximum force generating capacity
of the muscles in question. Transdiaphragmatic pressure (Pdi) is an index of dia-
phragm strength for which a critical threshold of 40% (Pdi breath/Pdimax) has
been determined, above which there is a denite time limit to fatigue or task
failure (2). For maximum inspiratory mouth pressure (an index of global inspira-
tory muscle strength), the critical ratio is 60%. The pressure (P) generated per
breath reects loads imposed, while the maximum pressure reects the maximum
force reserve. In the critical care arena and in the process of weaning, P breath
may be increased, Pmax decreased, or both may co-exist, which can augment the
P breath/Pmax ratio to levels at which sustainable ventilation is not possible, with
ensuing task failure, unless the level of load is appropriately decreased and the
weaning trial terminated.
The tension time index (TTdi) is the product of Pdi/Pdimax and duty cycle
(inspiratory time/total respiratory time) and was reported by Bellemare et al. to
be strongly related to diaphragm endurance capacity, blood ow, and oxygen
consumption (3 5). In normal subjects, the resting TTdi was 0.02, with 0.15
to 0.18 representing a fatigue threshold beyond which respiratory muscle
fatigue ensued in less than 30 minutes (3). The pressure time index utilizes
inspiratory mouth pressures (in lieu of Pdi), embodies similar concepts, and
has been reported in the literature in the context of weaning (6).
B. Load vs. Capacity
A delicate balance exists in critically ill patients requiring mechanical ventilation

between neuromuscular respiratory capacity and load, alterations of which can
severely impact on respiratory pump endurance capacity. Respiratory neuro-
muscular capacity can be decreased by depressed respiratory drive and conditions
or disorders producing weakness of the respiratory muscles. Enhanced loading
Weaning from Mechanical Ventilation 155
Figure 1 Relationship between the tension-time index of the diaphragm, predicted time
to task failure, and duration of trial of spontaneous breathing in weaning failure patients.
Note progressive increment in tension-time index of the diaphragm as the weaning trial
continued. Overt diaphragm contractile fatigue would likely have ensued if the trial had
continued for another 13 minutes. Source: Adapted from Refs. 14,15.
of the respiratory system may be secondary to augmented respiratory drive,

increased minute ventilation, and imposed elastic threshold or resistive loads
(7) (Fig. 1).
C. Respiratory Muscle Fatigue
Issues related to the development of overt contractile (peripheral) muscle

fatigue and central fatigue (central reduction in respiratory drive) with venti-
latory failure or with weaning failure are not well-dened. Muscle fatigue is
dened as a condition in which there is loss in the capacity for developing
force and velocity, resulting from muscle activity under load and which is
reversible by rest (8). Clinically, respiratory muscle fatigue may be viewed
as a condition in which the ability of the respiratory muscles to generate
pressure and airow is impaired. The latter is the result of heightened contrac-
tile activity of the respiratory muscles relative to strength and a decrease in
energy supply. Finally, the impairment of respiratory muscle function would
be expected to be improved by rest (i.e., complete unloading). By contrast,
muscle weakness was dened as a condition in which the capacity of a
rested muscle to generate force is impaired (8). It should be noted that the
fatiguing process may ensue prior to the inability to generate the required
force; secondly, this point of exhaustion or task failure is dependent on a

dened end point which, in turn, depends on the balance between pump
capacity and load. From the clinical perspective, task failure of the respiratory
muscle pump may be viewed as the onset of ventilatory failure.
Respiratory muscle fatigue has been classied as peripheral fatigue
(includes contractile fatigue and neurotransmission fatigue) or central fatigue.
Contractile fatigue refers to impaired contractile function of the muscle related
to mechanisms occurring within the muscle bers themselves. These include
disturbances in the sarcolemma, metabolic aberrations such as accumulation of
H and Pi, critical substrate depletion, and alterations in excitationcontraction
coupling mediated by mechanical or chemical factors such as reactive oxygen
species, nitric oxide, peroxynitrites, and cytokines, which can produce muscle
ber injury including sarcolemmal rupture, sarcomeric disruption, and cellular
organelle dysfunction (9 11). Fiber injury may be associated with prolonged
muscle impairment, peaking at 3 4 days (12,13).
Clinical data are sparse. Efthimiou et al. (14) reported data compatible
with contractile fatigue of the sternomastoid muscles in 10% of patients with
chronic obstructive pulmonary disease (COPD) admitted with acute exacer-
bations. With regard to weaning, Zakynthinos et al. (6) reported P breath/
PImax ratios of 0.42 + 0.11 in patients with acute respiratory failure upon dis-
continuation by mechanical ventilation, with pressure-time indices clustered
around or exceeding a critical threshold of 0.15 to 0.18. Similar results have
been noted for patients with COPD. Recently, Laghi et al. (15) evaluated dia-
phragmatic contractile responses non-volitionally (by phrenic nerve stimulation)
in patients undergoing spontaneous breathing trials as part of a weaning regimen.
Of interest, no change in twitch Pdi was observed in either successful or weaning
failure patients, suggesting that in the failure group, no evidence of contractile
diaphragm failure occurred at the point of termination of the weaning trial. It
was also highlighted in the paper that many of the weaning failure patients
had low twitch Pdis (i.e., diaphragm weakness). The data do not exclude a
central component, as the increased loads presented to the weaning failure
patients were not met with a concomitant increase in diaphragm recruitment.
The authors also point out that the reason weaning failure patients did not
develop low-frequency contractile fatigue was that mechanical ventilation was
instituted prior to this becoming manifest (15). On the basis of some of the con-
siderations summarized above (3), the authors predicted that fatigue would have
been evident had the trial of spontaneous breathing continued for a further 13
minutes (14) (Fig. 2).
D. Pathophysiology of Weaning Failure

Jubran and Tobin (16) studied respiratory mechanics in ventilated patients with
COPD either passing or failing a trial of spontaneous breathing. Rapid shallow
respirations were noted in the failure group at the onset of the trial. Further
Figure 2 Interrelationships between respiratory pump capacity, load, and peripheral

and central inuences in relation to weaning from mechanical ventilation. Source:
Adapted from Goldstone J and Moxham J. Weaning from mechanical ventilation.
Thorax 1991; 46:56 62.
progressive increments in elevated inspiratory resistance, elastance, and intrinsic

(auto) positive end-expiratory pressure (PEEPi) were noted between onset and
end of the trial in the weaning failure patients. This culminated in a signicant
increase in respiratory muscle energy expenditure as noted by a progressive
rise in the pressure time product (the time integral of the difference between eso-
phageal pressure and chest wall recoil pressure). At the conclusion of the weaning
trial, 76% of the weaning failure patients developed hypercapnia, which likely
reected inefcient clearance of carbon dioxide due to abnormal breathing
pattern and worsened pulmonary mechanics. In another study of weaning
success or failure following a spontaneous breathing trial in ventilated patients,
a progressive fall in mixed venous oxygen saturation (SvO2) was noted in the
failure group, but not in those successfully weaned (17). The fall of SvO2 with
weaning failure was because of a lack of increase in cardiac index and oxygen
delivery (increased right and left ventricular afterload during spontaneous breath-
ing trial), and enhanced oxygen extraction by the tissues.
E. Evaluation and Treatment of Reversible Factors

A number of important factors need to be considered when liberation from mecha-
nical ventilation is contemplated or when difculties are encountered with the
weaning process. These are outlined in Table 1. While the factors outlined in
Table 1 are self-explanatory, a number of key issues will briey be emphasized.
Respiratory muscle weakness and pump dysfunction are discussed in detail in
the Chapter 7, Vol. 2 on neuromuscular respiratory failure, as are specic treat-
ment modalities.
PEEPi develops because of insufcient expiratory time for exhalation to the
normal relaxation volume of the respiratory system prior to the delivery of the
subsequent mechanical breath (18). This results in dynamic hyperination and
positive end-expiratory alveolar pressure. PEEPi increases the work of breathing,
as the respiratory muscles must generate sufcient force to overcome positive
recoil pressures to initiate airow. The level of PEEPi and imposed work may
be reduced by measures that prolong expiratory time. In addition, in certain cir-
cumstances, trigger sensitivity may be reduced by the application of external
PEEP to levels not exceeding 85% of PEEPi (19). Gottfried (20) provided prac-
tical recommendations for the use of external PEEP in patients with PEEPi.
Although PEEPi can easily be measured on new generation ventilators,
the measurement needs to be made under completely passive conditions, and
increased abdominal muscle recruitment can falsely elevate measured levels
(21). Further, inappropriate application of external PEEP can result in
further dynamic hyperination and hemodynamic compromise (22). Also,
increasing ow rates to increase expiratory time and limit air trapping can
result in a tachypneic response negating the positive benet (23). Dynamic
hyperination secondary to air trapping imposes an increased elastic load on
the respiratory system and can alter optimal diaphragm geometric congu-
ration, resulting in inefciency of its action and reduced force generating
potential. Thus, efforts to limit PEEPi and air trapping may signicantly
curtail work of breathing.
Recent studies have shown that heat and moisture exchanges for
airway humidication may increase work of breathing, PEEPi, PaCO2, and dis-
comfort, requiring high levels of pressure support to offset (24). The use of a
heat and moisture exchanger should, thus, be avoided in patients with border-
line reserve and more conventional methods of heated humidication
employed.
The issue of carbohydrate excess as a limiting factor in weaning success is
a controversial one. This relates to high CO2 production with diversion of meta-
bolic pathways to fat synthesis in which the RQ for lipogenesis is 8.0. Studies
have shown, however, that high caloric feeding was more important than high
percentage carbohydrate provision (25). Thus, moderate caloric intake, particu-
larly in patients with limited reserve, is prudent. If necessary, energy require-
ments can be assessed using indirect calorimetry.
Table 1 Factors to Consider with Weaning Failure
Depressed respiratory drive

Sedatives
Analgesics including opiates
Signicant metabolic alkalosis
Hypothyroidism
Central hypoventilation
Obesityhypoventilation
Neurologic damage
Respiratory muscle weakness
Acute neuromuscular respiratory failure (e.g., Guillain Barre, myasthenic crisis)
Chronic neuromuscular respiratory failure (e.g., amyotrophic lateral sclerosis)
Critical illness polyneuropathy + myopathy
Diaphragm paralysis
High cervical cord injury
Stressed and unstressed malnutrition/cachexia
Electrolyte disorders (e.g., hypophosphatemia, hypomagnesemia, hypokalemia,
hypocalcemia)
Metabolic disturbances (e.g., respiratory acidosis, metabolic acidosis, compensated
metabolic acidosis, hypoxemia, thyroid dysfunction)
Drug inuences (e.g., prolonged effects of paralytic agents, drug-induced myopathy or
rhabdomyolysis, drug-induced myasthenic syndromes)
Resistive loads
Conditions associated with airow obstruction
Bronchospasm
Thick copious secretions
Blood clot obstruction
Articial airway (small endotracheal tube, kinking, biting, balloon herniation, mucus
plugs or blood clot)
Ventilator circuit and demand valves
Elastic loads
Reduced lung compliance (e.g., pulmonary edema, consolidation, brosis)
Reduced chest wall compliance (e.g., kyphoscoliosis, pleural disease, morbid obesity,
abdominal distension/ascites)
Dynamic hyperination
Threshold loads
Intrinsic (auto) PEEP
Increased minute ventilation
Increased dead space (wasted ventilation)
Increased metabolic rate (fever, hypermetabolic states, excess calorie/carbohydrate
provision)
Pain
Psychiatric disorders
Anxiety
Withdrawal states
Cardiac disease
LV dysfunction and failure
Cardiac ischemia
Inadequate reversal of original disease state
Figure 3 Pathophysiologic inuences impacting on the cardiovascular system upon with-

drawal of mechanical ventilatory support in patients with underlying cardiac dysfunction.
Abbreviations: PEEPi, intrinsic PEEP; Ppl, pleural pressure; PCWP, pulmonary capillary
wedge pressure; VO2, oxygen consumption; MVO2, myocardial oxygen consumption.
Source: Adapted from Ref. 7.
Patients with left ventricular dysfunction may benet from mechanical

ventilation, which reduces preload, afterload, and possibly improves contracti-
lity. Cardiac compromise and ischemia may ensue upon liberation from mechan-
ical ventilation (26,27). The complex pathophysiologic factors involved are well
illustrated in Figure 3.
III. Predicting Weaning Success
The publication of over 400 putative weaning predictors has given way to
increasing skepticism regarding the clinical utility of such parameters (28).
Indeed, the purported benets of weaning parameters have recently been high-
lighted as a myth in an editorial entitled Weaning the patient or weaning
old-fashioned ideas(29). Over 30 years of publications in this area have not
yielded strongly reproducible results. Furthermore, the study groups have been
small, and the data have generated unhelpful likelihood ratios (LR) and low pre-
dictive power (28). It has been suggested that physicians a priori clinical prob-
abilities of success of weaning may affect the selection of patients and in turn the
accuracy of the tests (28,29).
A. Weaning Parameters
Traditional weaning parameters have entailed indications of gas exchange and
indices of chest mechanics or breathing pattern. Gas exchange criteria have
spanned dead space calculations to hypoxemia ratios. In general, the minimal
acceptable PaO2 is 60 mmHg on a FIO2 of 0.4 and a PEEP less than 5 cmH2O.
Other parameters of oxygenation that can be used include PaO2/PAO2
(.0.35), A2a O2 (,350 mmHg), or PaO2/FIO2 (.200) (30). Although ade-
quate arterial oxygenation is essential to initiate weaning, the predictive value
of these indices for the outcome of the weaning trial is usually low (31). Strict
adherence to these criteria should not preclude liberating a patient from mechan-
ical ventilation, if deemed clinically appropriate.
Historically, indices of chest mechanics and respiratory pattern during
spontaneous breathing have included vital capacity (VC) greater than
10 15 mL/kg, spontaneous tidal volume (VT . 5 mL/kg), respiratory rate
(RR , 35/ min), minute ventilation (VE , 10 15 L/min), maximum voluntary
ventilation (MVV) at least twice VE, and maximal inspiratory mouth pressure
(PImax using, 215 to 230 cmH2O) (28,32,33). VC reects not only lung and
chest wall mechanics, but also respiratory muscle strength, albeit a non-specic
marker particularly of the latter (see chapter 7, vol. 2 on neuromuscular respirat-
ory failure). Pooled data for RR , 38 breaths/min has been analyzed with a LR
of 1.1, indicating unchanged post-test probability and thus an unhelpful measure
(28,34,35). By contrast, a RR . 38 breaths/min yielded a LR of 0.32, indicating
a somewhat reduced probability of weaning success (28,34,35). A high VE may
reect an increased dead space, load, and work of breathing. PImax is a measure
of global inspiratory muscle strength and is limited by the need for patient
cooperation, coordination, inuences of lung volume, and wide inter- and
intra-subject variability (see chapter 7, vol. 2 on neuromuscular respiratory
failure). While a one-way valve has been utilized to circumvent some of these
hindrances, its use has been limited (36). (This valve permits only expiration,
while the most negative pressure generated by inspiration against an occluded
airway over a 20-second period is recorded as PImax.) Table 2 depicts sensitivity
and specicity for several of these traditional indices.
Newer studies have attempted to improve the robustness of weaning tests
by either grouping variables or by seeking more specic physiological correlates.
Several integrative weaning indices and unique parameters will be briey
discussed. Yang and Tobin developed the commonly employed rapid shallow
breathing index (RSBI) after discovering that patients who failed weaning devel-
oped tachypnea and decreasing VT almost immediately upon discontinuation of
ventilatory support (37). The RSBI was measured after one minute as the ratio
of frequency to VT while breathing spontaneously (33). Using 105 (breaths/
min)/L as the threshold, the sensitivity was 97% for predicting weaning
success, whereas the specicity was 64%, indicating that 36% with an
RSBI , 105 failed the weaning trial (i.e., false-positives) (33). While pooled
data from six studies yielded similar sensitivities for RSBI, with a range of 87
97%, specicities were generally lower (mean: 47%; range: 33 60%) (28). Fur-
thermore, sensitivity for RSBI decreased with increased duration of mechanical
ventilation (33), whereas specicity was inuenced by disease state (38). For
example, specicity of RSBI in patients with COPD was 65%, while in patients
with neurologic disease or acute respiratory failure of varying etiology, the speci-
city was 26% and 28%, respectively (38). An analysis of extubation failure by
Epstein (39) showed that when weaning failed, despite an RSBI ,100 (i.e., false-
positive test), the mechanism was secondary to other comorbid processes and not
the underlying respiratory disease alone. RSBI has been reported to be higher in
females and patients with smaller endotracheal tubes, suggesting that the index
threshold needs appropriate adjustment to specic patient populations and con-
ditions (40). A threshold of 130 has been suggested as a more appropriate cut
point in elderly patients (41). Krieger and Isber suggested that stability of the
RSBI when measured serially from the onset of the weaning trial and for the sub-
sequent two to three hours gains sensitivity and specicity.
A number of integrative indices have been reported. The compliance, rate,
oxygen, and pressure (CROP) index incorporates gas exchange, chest wall mech-
anics, and respiratory strength.
CROP index [Dynamic compliance PImax (PaO2/PAO2)]/RR (33).
This index had a lower sensitivity and specicity than the RSBI (Table 2). A
weaning index based on ventilatory endurance and efciency of gas exchange
by Jabour et al. (42) yielded high predictive value; however, it is cumbersome
to calculate and was based on a limited number of patients. With airway occlu-
sion, the ratio of the pressure generated with the rst breath (PI) to the pressure
Table 2 Predictive Indexes of the Outcome of Weaning From Mechanical Ventilation
Index Threshold value Sensitivity Specicity
Spontaneous minute 15 L/min 0.78 0.18

ventilation (VE)
Spontaneous Respiratory 38/min 0.92 0.36
frequency (RR)
Spontaneous tidal volume (VT) 325 mL 0.97 0.54
Spontaneous VT/body weight 4 mL/kg 0.94 0.39
Maximal inspiratory 15 cmH2O 1.00 0.11
pressure (PImax)
Dynamic compliance 22 mL/cmH2O 0.72 0.50
Static compliance 33 mL/cmH2O 0.75 0.36
PaO2/PAO2 ratio 0.35 0.81 0.29
RR/VT (RSBI) 105 0.97 0.64
Compliance, rate, oxygen, and pressure 13 mL/(breath/min) 0.81 0.57
(CROP) index

generated after 20 seconds of airway occlusion (PImax) has been evaluated as an

index of respiratory pump reserve with a threshold value less than 0.3 (43). Com-
bining this index with RSBI, there was improved sensitivity of 81% and speci-
city of 93% (43). The occlusion pressure (P0.1), measured as inspiratory
pressure 100 milliseconds after airway occlusion, is in part an index of respiratory
drive and has been used as a weaning tool (44,45). Several studies in patients with
COPD report weaning failure when P0.1 exceeded 6 cmH2O (44,45). The normal
value is less than 2 cmH2O. One prospective study showed that P0.1 was a com-
parable predictor to RSBI and suggested that the combination of RSBI and P0.1
would even provide a better predictor (44). In contrast, Fernandez et al. (46)
found bedside P0.1 and P0.1 RSBI of little value in predicting extubation
failure. Experimentally, neural networks have been employed with some
success (47).
B. Novel Approaches Studied

A number of novel approaches have been studied but have not been adopted into
common clinical practice. The oxygen cost of breathing (OCB), (dened as
the difference in oxygen consumption during spontaneous breathing and on
mechanical ventilation), was evaluated with an OCB , 30% deemed a promising
indicator (48,49). Evaluation of work of breathing (WOB) is theoretically
attractive, with bed-side equipment available (50,51). WOB is fraught with
possible errors deriving from assumptions of chest wall compliance, endotracheal
tube resistance, and auto-peep. Further, it is invasive, requiring esophageal
pressure measurements, and no general or patient population-specic thresholds
are known.
A non-invasive indicator of ventilatory load, quantitative assessment of rib
cage abdominal motion, uses respiratory plethysmography and has been studied
by Tobin et al. (52) in a cohort of patients weaning from mechanical ventilation.
While abdominal paradox was seen in both weaning success and failure patients,
combined indices of abnormal rib cage and abdominal motion were elevated in
weaning failure patients. A subsequent study suggested that abnormal rib cage
and abdominal motion was related to increased load (53).
Another somewhat more invasive approach requires an oral- or naso-
gastric tube. Gastric intra-mural tonometry is well suited to environments such
as the intensive care unit. Intra-mural pH (pHi) may be estimated using tonome-
try devices or analysis of gastric juice (54 56). With high loads, respiratory work
load and blood ow demand is high with blood ow diverted from the splanchnic
bed. A fall in pHi , 7.3 or a fall of .0.09 during spontaneous breathing suggests
weaning failure (54 56).
VE recovery time has recently been touted as a new potential indicator of
extubation outcome in preliminary studies (57). The shorter the recovery time
to within 110% of baseline VE upon return to full ventilatory support, the
greater the chance of weaning success.
While weaning parameters are of limited value, especially when applied

broadly, there may be a role for selective use of certain parameters. In some
cases, diagnosis groups and demographics selection may inform the choice
of parameters. Patients with recovering neuromuscular disease may benet
from tracking of vital capacities, whereas these same patients may not
benet from routine application of the rapid shallow breathing index, which
lacks specicity. In contrast, emphysematous patients may benet from screen-
ing for tachypnea (RR . 38) and RSBIs. In some cases, thresholds values may
need adjustment, such as RSBIs in women and the elderly. In more selected
populations, such as long-term weaning patients, weaning parameters may
facilitate non-physician-run protocols. In any application, diligent bedside
clinical evaluation cannot be supplanted by current or historical weaning
parameters.
IV. Protocols and Pathways

A. Techniques and Best Practice
Attempts at standardization of the weaning process have been touted for over a
decade (58). Early reports focused on the best practice techniques in weaning
from ventilation. In one prospective, randomized trial of weaning by intermittent
mandatory ventilation (IMV), outcomes of time to extubation were similar to T-
tube (59). The authors concluded that protocols involving both approaches may
be applied to clinically stable patients with success. Despite some early experi-
ence in acute surgical populations with IMV trials, weaning protocols involving
T-piece trials or pressure support ventilation now appear to be superior to inter-
mittent mandatory and are preferred for acute medical and surgical patients (60).
However, for post-ICU long term weaning cohorts, protocols involving synchro-
nized IMV plus pressure support have enjoyed success (61). Similar approaches
have achieved promising results in certain special populations, such as spinal
cord injury patients, with expected long duration of weaning (62).
B. Practice Variation
Personal preference and variability of weaning practice has frequently been
reported, and several randomized clinical trials have produced conicting
results regarding the best technique for carrying out the weaning process (63).
Work done by the Spanish Lung Failure Collaborative Group in the 1980s
demonstrated wide variation in regional weaning practice as well as reliance
on AC and SIMV techniques during weaningapproaches linked to long dur-
ations of weaning (64). Esteban et al.s analysis of international ventilation prac-
tices found remarkable similarity in approaches to maintenance ventilation but
wide practice variation around weaning (65). Their group and others have
demonstrated the superiority of not one but several approaches over AC or
SIMV, including daily and multiple daily spontaneous T-tube trials and pressure
support weaning (66).
Some groups have applied multi-disciplinary processes aimed at the
leading factors prolonging ventilator stay, such as ventilator-associated pneumo-
nia (67). In other cases, technology for monitoring weaning progress has also
been promoted. In one series, esophageal manometry was used adjunctively to
keep patients on protocol (68). No differences in steps to weaning were observed;
however, the monitoring technology appeared to give clinicians a greater con-
dence with staying on protocol without an increase in protocol intolerance. In
short, while QI tools, technology, and protocols may reduce variability in prac-
tice, a disciplined process and not a particular approach appears to be the
salient message.
C. To Wean or to Extubate?
Although improvements in time to wean have modestly improved on protocols,
overall duration of ventilation has markedly decreased on some protocols (69).
One explanation is that protocols enable practitioners to identify patients ready
for rapid extubation without weaning. Saura et al. (69) studied the clinical con-
sequences of a protocol that enabled direct extubation without a weaning tech-
nique, which was applied in 10% of their cohort and resulted in decreased
duration of mechanical ventilation and ICU stay. Safe, direct extubation, and
not weaning methods per se, appeared to be the factor linked to timely extubation.
Similar rapid extubation strategies have been applied with nurse-directed proto-
cols following cardiac surgery, and such protocols under various specialist direc-
tion are now standard of care (70).
D. Substance over Style

Employing a systematic approach may be more important than the particular
technique selected. Protocol-guided weaning usually performed by non-
physicians, with improvements in patient outcomes (durations of mechanical
ventilation, incidence of ventilator-associated pneumonia, and frequency of
complications), has gained popularity (63). Protocol-guided weaning by nurses
and respiratory therapists appears safe and more rapid than physician-directed
weaning (71). Respiratory therapists appear to safely and efciently wean
cardiac surgery patients from mechanical ventilation with a two-hour difference
in ventilation time (favoring therapists over physicians) (72). Physicians, includ-
ing housestaff, appear to accept these protocols when integrated into multidisci-
plinary practice (73).
Protocols do produce dramatic cost reductions (71,74). Nursing care
facilities are increasingly interested in such protocols as a regular part of a
cost-saving and patient ow strategy (74 76). These cost reductions have been
achieved by reducing hours of ventilation by 40% to 50%, length of stay by
approximately 30%, and acute physiology scores (presumably affecting acuity
and stafng). Interestingly, protocols do not appear to affect rates of reintubations

or mortality (71,74).
Nurses and respiratory therapists have been successfully involved in
protocols involving educational interventions as well as standing orders for spon-
taneous breathing trials and pressure support reduction. Medical intensivists have
also performed in a manner consistent with successful protocols, but respiratory
therapists and nurses were able to implement protocols without physician super-
vision. In children, one report noted reductions in total ventilation time, weaning
time, pre-weaning time, time to extubation, and time on FIO2 . 0.40; even after
stratication by diagnosis, differences in weaning time remained signicant (77).
As in adults, differences in the incidence of reintubation were not found, and rates
of new-onset tracheitis, subglottic stenosis, and pneumonia were not increased in
children.
Ely et al. identied common themes to these protocols: stabilization of a
patients condition should prompt timely spontaneous breathing trials when
safe. Strategies for patients who fail a trial of spontaneous breathing should
focus on reversible factors, use of a well-monitored mode of mechanical venti-
lation such as pressure support, and repeat spontaneous trials daily: (1) consider-
ation of all remediable factors (such as electrolyte derangements, bronchospasm,
malnutrition, patient positioning, and excess secretions) to enhance the prospects
of successful liberation from mechanical ventilation; (2) use of a comfortable,
safe, and well-monitored mode of mechanical ventilation (such as pressure
support ventilation); and (3) repeating a trial of spontaneous breathing on the fol-
lowing day (78). Through a careful, slow implementation process involving daily
reminders, improved compliance with weaning protocols can be achieved (79).
Calhoun and Specht describe three phases: optimizing the patients condition,
assessing and evaluating the patients progress, and diagnostic testing (spontaneous
breathing trials). In addition, Calhoun and Specht spell out the need for a pro-
fessional and responsible monitor (nurse or respiratory therapist) (80). Where
these elements are present, even without an explicit protocol, respiratory therapist,
or nurse monitoring, such as in a closed, well-staffed teaching team run by phys-
icians, outcomes have been reported as equivalent by Krishnan et al. (81). In a coun-
terpoint to the Krishnan article, Tobin argues for caution in studies that compare
against usual care (82). In contrast to the period when Esteban reported wide prac-
tice variation, during the current era, many physicians and other clinicians may have
integrated elements of a successful protocol into daily practice.
E. Specic Approaches and Examples
Several protocols are available that have been shown in well-designed trials to
efciently wean most patients. After a spontaneous breathing trial, patients
who appear to be doing well by clinical criteria or weaning parameters may be
extubated, placed on t-tube, or ventilated with low-level pressure support.
Usually the level of pressure support is 7 cmH2O. Trials on t-tube or pressure
support may last 30 to 120 minutes. Little information is known about the optimal
duration of spontaneous breathing trials, and these may safely last from 30 to 120
minutes (83). Many advocate a exible approach of spontaneous trials of 30
minutes or low-level pressure support (84).
Where patients do not tolerate immediate low level pressure support or t-tube,
a pressure wean or titration can take place. A pressure support titration can establish
the minimal level to provide an adequate respiratory rate, tidal volume, and minute
ventilation. Then titration can be followed by gradual reduction by 2 cmH2O every
one to two hours. Patients weaned on consecutive days can often return to the last
pressure support level tolerated. Two randomized studies found that, in difcult-to-
wean patients, synchronized intermittent mandatory ventilation may be even more
effective (85,86). Other studies indicate that patients with prolonged ventilation be
placed on a chronic weaning protocol, which usually involves slower, graded IMV
reductions followed by pressure support reductions.
Non-invasive ventilation has been touted as a means to improve weaning
success, and studies have shown improvements in mortality (87). Unfortunately,
weaning failures or the duration of mechanical support related to weaning were
unimproved (88). One subgroup, chronic obstructive lung disease patients,
appeared to demonstrate the greatest mortality reduction (88). In all, given the
mixed results on ventilation, the Cochrane group still considers the data on
non-invasive ventilation insufcient to recommend its routine use in weaning.
Additional detail is described in the AHRQ technology assessment (89).
Automated, adaptive support ventilation techniques have been reported to
facilitate fewer ventilator manipulations but produced no improvement in out-
comes (90). In summary, reduction of practice variation at a single site
through the use of protocols may produce reductions in duration of ventilation,
despite differences in practitioners and details of protocols.
Expert and evidence-based recommendations for acute weaning protocols
have been summarized by Ely et al. in Table 3. The group provides guidance on
essential content, development strategies, and implementation of protocol.
V. Weaning from Prolonged Ventilation
The long-term acute care (LTAC) setting may be the best setting for protocolized
approaches. Scheinhorn et al. (61) reported signicantly shorter time to
weaning than benchmarks and historical controls through therapist-implemented
patient-specic (TIPS) weaning protocols. Scheinhorns protocol, trialed in over
250 patients and 9000 ventilator days at Barlow Respiratory Hospital and
Research Center, emphasizes adherence with a rigid set of steps, after assessing
for patient-specic issues with readiness and tolerance screens. While retaining
patient-specic adaptations, such as undiagnosed hypoxemia (with blood gases),
the overall emphasis is a rigid codes of procedure and a mantra: assess-
ment ! intervention ! assessment. This fusion of process engineering and
clinical judgment allows the expertise of the physicians to be present at the
Table 3 Weaning Protocol Recommendations
Non-physicians should be included in the development and utilization of protocols related

to weaning
ICU physicians should utilize protocols for liberating patients from mechanical ventilation
Clinicians should conduct SBTs at least daily to identify patients ready for liberation
When a patient fails SBT, clinicians should consider reversible factors and attempt SBTs
at least once daily with head of bed elevated 30 458
If faced with recurrent failures, consider early tracheostomy
When a patient passes an SBT, prompt extubation should be considered
For analgesia, agitation, and delirium, a protocol should be considered, including daily
lifting of sedating medications
For development of weaning protocols, consider: evidence-based goals, interactive
education, audit, reminders, opinion leaders, and feedback
Abbreviation: SBT, spontaneous breathing trial.

bedside of weaning patients continuously, (61,91). The algorithm is shown in

Table 4. The algorithm features synchronized IMV weaning, followed by
graded pressure support reductions, and increasing duration spontaneous
breathing trials.
VI. Sedation
Modern state-of-the-art approaches to weaning include proactive approaches to

deal with agitation, delirium and pain, which all affect the efciency of the
weaning process. Recently much emphasis has been placed on these key
issues. Sedation of the ventilated patient is both a goal and process. For the
ventilator-dependent patient, adequate oxygenation and safe ventilation can
better be achieved with a comfortable and calm patient. Agitation plays a signi-
cant role in ventilator dyssynchrony, increased oxygen demands, accidental self-
extubation, and catheter removals (92,93). Furthermore, the delirium and disor-
ientation, seen in 15% to 20% of general medical surgical patients and as many
as 80% of ventilated intensive care patients, must be resolved in order to ade-
quately ventilate and eventually liberate the patient from the respirator (94,95).
Severity of delirium has been associated with duration of ventilation and
increased cost, independent of severity of illness, degree of organ dysfunction,
admitting diagnosis, or the presence of nosocomial infection (96). Pain and
analgesia may also complicate the sedation and weaning plan.
In practice, tensions arise between the need to acutely respond to discomfort
and the dangers of prolonged ventilation. Prolonged sedation may lead to muscle
atrophy, venous stasis and thrombosis, pressure damage to soft tissues, and pro-
longed obtundation (97). Oversedation may increase duration of ventilation and
costs (97,98). Daily reassessment of sedation may be as important as daily
Table 4 TIPS Long-Term Weaning Protocol
Initial ventilator setting

If SIMV . 10 or PSV . 20, change to A/C
Adjust PSV according to tidal volume measured
Daily evaluation for do not wean signs
Pulse
Temperature
FIO2 requirement
Vasopressor (dopamine) requirement
PEEP requirement
Other clinical reason
Weaning assessment (5 min into each attempt)
Respiratory rate
Tidal volume
O2 saturation
Pulse
Accessory muscle use
Post-weaning RSBI
After rst weaning attempt,
For low RSBI (,80), advance to SBTs of increasing duration immediately or
Graded SIMV or PSV reduction
Graded SIMV reduction rst
Graded PSV reduction
Each step with weaning assessment
Advance to SBTs of increasing duration
breathing trials to reduce the duration of mechanical ventilation (99,100). Close

attention to appropriate dosing intervals of sedatives and re-addressing continuous
infusions may avoid prolonged ventilation and escalating drug costs (97). Cost-
effective rational use guidelines may be safely applied in a combination with
academic detailing by pharmacists and critical-care consultants (101).
A. Establishing a Comfortable Environment

Institutional, and especially ICU-related, causes of agitation should be con-
sidered for ventilated patients (93). Noise, light, temperature, insertion of
suction catheters, and re-positioning are among the common disruptions for the
ventilated patient, and the subsequent loss of contiguous, quality sleep may con-
tribute to patient anxiety and possibly immune function (102).
For the ventilated patient, frequent orientation and attempts to control
environmental variables have been suggested but have not been well-studied.
B. The Agitated, Ventilated Patient

In the uncomfortable patient, after maximizing non-pharmacological approaches,
rst exclude common physiologic causes. Hypoxemia, hypoglycemia, hypoten-
sion, pain, ischemia, and arrhythmias can present with non-specic agitation in
the uncommunicative patient and may affect weaning success (93,103). Next,
address pain and discomfort, associated with ventilation or other health con-
ditions, prior to initiating sedatives that may preclude a pain assessment.
Visual analog scales may be helpful, as family and caregivers may misapprehend
patient pain levels (93,104). Unchecked stress responses to pain may also have
negative clinical consequences (93). Because continuous sedation itself has
been linked to poor outcomes, routine lifting or lightening of sedation should
be attempted (103).
C. Dening the Sedation Goal for the Ventilated Patient

The Task Force of the American College of Critical Care Medicine (ACCM) of
the Society of Critical Care Medicine (SCCM), the American Society of Health-
System Pharmacists (ASHP), and the American College of Chest Physicians
(ACCP) recommend goal-oriented therapy for sedation. Keys to success in reach-
ing a sedation goal are frequent assessments and use of common language to
describe levels of sedation and anxiety (Table 5) (105). Validated subjective
scales, such as the Ramsay, Riker Sedation Agitation Scale (SAS) and the
Motor Activity Assessment Scale (MAAS), have been effectively used as part
of protocol-driven sedation plans. Figure 4 shows a ow chart with the
SCCM/ASHP/ACCP approach.
D. Choice of Sedating Agents for Ventilation

Several principles guide the choice of sedating medications. First, analgesics are
indicated prior to initiating other agents that impair the expression or assessment
of pain. Splinting, guarding, spasm, or rigidity may affect coordinated chest wall,
abdominal, and diaphragmatic movement and lead to pulmonary dysfunction
(92,93). Appropriate analgesia has been linked to reduced post-operative compli-
cations. Second, amnestic agents may prevent recall of painful or traumatic
experiences during mechanical ventilation (106).
Common rst line agents include benzodiazepines with excellent antero-
grade amnestic effects, as well as propofol and haloperidol (Table 4). Midazolam
and lorazepam to a lesser extent have the potential for prolonged drug effect in a
continuous infusion. Midazolam is rapid acting, but its metabolite may have a
prolonged half-life after days of infusion, and lorazepam crystallizes easily and
has been associated with acute tubular necrosis. Propofol, in a goal-oriented, pro-
tocolized approach, can provide comparable sedation to benzodiazepines and
may allow rapid awakening (107). With propofol, clinicians should be on the
alert for bradycardia, hypotension, myocardial depression, drops in intracranial
Table 5 Recommendations for Analgesia, Sedation, and Delirium
Analgesia
Pain assessment with an appropriate approach on a regular basis
Patients who can communicate: Encourage subjective (visual analog) scale
Patients who cannot communicate: observe pain-related behaviors and the change in
these parameters following analgesia
Establish therapeutic plan and communicate to all caregivers
Opioid analgesics: fentanyl, hydromorphone, and morphine are the recommended
agents
Scheduled opioid doses or a continuous infusion is preferred over an as needed
regimen
Fentanyl is preferred for a rapid onset of analgesia in acutely distressed patients
Fentanyl or hydromorphone are preferred for patients with hemodynamic instability or
renal insufciency
Morphine and hydromorphone are preferred for intermittent therapy because of their
longer duration of effect
NSAIDs or acetaminophen may be used as adjuncts to opioids in selected patients
Ketorolac therapy should be limited to a maximum of ve days, with close monitoring
for the development of renal insufciency or gastrointestinal bleeding
Other NSAIDs may be used via the enteral route in appropriate patients
Sedation
Sedation of agitated critically ill patients should be started only after
providing adequate analgesia and treating reversible physiological causes
A sedation goal or endpoint should be established and regularly redened for each
patient
Regular assessment and response to therapy should be systematically documented
The use of a validated sedation assessment scale (SAS, MAAS, or VICS) is
recommended
After seven days, consider the potential for opioid, benzodiazepine, and propofol
withdrawals
Doses should be tapered systematically to prevent withdrawal symptoms
Delirium
Haloperidol is the preferred agent for the treatment of delirium in critically ill
patients
Patients should be monitored for electrocardiographic changes when receiving
haloperidol
(QT interval prolongation and arrhythmias)
Sleep
Sleep promotion should include optimization of the environment and
non-pharmacologic methods to promote relaxation with adjunctive
use of hypnotics

Is patient comfortable
or at goal?
Rule out and treat Reassess Daily Goal

reversible causes Daily Lifting of Sedation
Taper if > 1 week and monitor
for withdrawal
Consider Continuous
Infusion Versus
Nonpharmacologic Intravenous Push
Treatment
Fentanyl,
Assess pain / analgesia
Hydromorphone,
(Visual Analog Scale)
Morphine
Assess agitation / Midazolam,

sedation Lorazepam,
(Subjective Scale) Propofol
Assess delirium Haloperidol

(CAM-ICU Scale) UNLESS alcohol-related
Figure 4 Simplied ow chart for sedation of mechanically ventilated patients. Source:

Adapted from Ref. 93.
pressure, and hypertriglyceridemia. Etomidate, a non-benzodiazepine, non-

barbiturate agent used for intubations and anesthesia induction has been associ-
ated on repeat use with Addisonian crisis (108).
Third, agents that may be used for anxiety and analgesia may precipitate or
exacerbate delirium in the young or elderly. In general, clinical status permitting,
sedation should be routinely lifted during ventilation to evaluate mental status.
E. Cost Implications of Sedation for the Ventilated Patient

A wide variety of acquisition costs are reported for these sedating medications,
but studies on cost are mixed (109,110). On the other hand, sedation guidelines
and systematic, multi-disciplinary approaches have been shown to reduce
drug costs (from $81.54 to $18.12 per patient per day), ventilator time
(317 167 hr), and the lengths of ICU stay (19.1 9.9 days) and total stay without
a change in mortality (92).
F. Preventing Anticipated Pain and Discomfort and

Addressing the Unanticipated
Endotracheal tubes, non-invasive ventilation devices, thoracostomy tubes, catheters,
airway suctioning, chest percussive therapy, dressing changes, immobilization,
transfers, and re-positioning are all potentially painful and uncomfortable experi-
ences as well. Interventions to improve comfort include: avoidance of endotracheal
and thoracostomy tube traction, proper tting of non-invasive masks/adaptors,
avoidance of pressure sores (especially with masks), allowing small leaks on leak
compensating non-invasive devices (to avoid tight straps), and catheter infection
and clot prevention techniques. Some small studies suggest specialized beds to
prevent decubitus ulcers or to facilitate mobility (111). Head of bed elevation to
prevent nosocomial pneumonia and acid suppression or mucosal protection for
peptic ulcer prevention are now widely accepted to avoid ventilator-associated com-
plications and attendant discomfort (112114).
G. Selection and Titration Analgesic Medications

Analgesics include opioids, non-steroidal anti-inammatory drugs (NSAIDs),
and acetaminophen. Agent selection in ventilated patients should be based on
side-effects, potency, and cost. Spontaneously breathing patients on non-invasive
mechanical ventilation or patients with inadequate mandatory backup rates may
have precipitous declines in minute ventilation with opioid analgesia. Without
pro-motility agents, patients may not tolerate enteral feeding. Multiple studies
have reported the negative consequences of failure to feed early in ventilation.
Gastric retention in those being fed may increase the risk of aspiration, especially
in patients not placed in a semi-recumbent position. When protocols incorporat-
ing daily awakening from analgesia and sedation have been employed, analgesic
titration have been more precise, total doses of narcotics were lower, and duration
of ventilation and ICU stay were lower (115).
H. Tolerance, Withdrawal, and Sedative Reversal

After prolonged ventilation and sedation, adaptation, tolerance, dependence, and
withdrawal may occur. One report noted the phenomenon in ARDS patients, who
presumably require large doses of sedatives (116). Studies of pediatric patients
recommend that small daily opioid decrements may avoid withdrawal in this
population. Long-acting agents may also mitigate withdrawal. Subcutaneous
infusions have successfully been used for gradual weaning in children (92).
Outside of acute respiratory compromise, class-specic reversal agents, such as
naloxone and umazenil, should be avoided after prolonged treatment, in
known analgesic dependence, and in patients prone to dysrhythmias (92).
Similarly, one should avoid partial opioid agonists, such as butorphanol, bupre-
norphine, and nalbuphine.
In addition to opiate or benzodiazepine withdrawal, clinicians should be
vigilant about alcohol withdrawal. Overlooking the diagnosis of alcohol withdra-
wal and empiric treatment with neuroleptics has been shown to lead to adverse
clinical outcomes (117). Vigilance for withdrawal signs and symptoms may
avoid costly clinical misadventures, prolonged ventilation in an attempt to re-
sedate, or potentially increased mortality.
I. Delirium
Delirium in the ICU has been associated with fewer median days alive and without
mechanical ventilation (118). Delirium is also linked to higher incidence of cogni-
tive impairment at hospital discharge and higher six-month mortality, even after
adjusting for relevant covariates (118). The condition is also associated with 39%
higher ICU costs (94). Long-term neuropsychological impairment is increasingly
recognized following mechanical ventilation (119). Severely ill patients, elderly
patients, children, and psychotic patients may become more confused with benzo-
diazepine sedatives. Other risk factors for delirium are listed in Table 6.
The need for reliable interdisciplinary measures has led to increasing use of
tools such as the Confusion Assessment Method (CAM-ICU) (120) For its treat-
ment, haloperidol is generally well-tolerated and has anxiolytic properties. In
patients with prior cardiac disease, the benets and risks of alternatives should
be weighed. In general, patients on prolonged neuroleptics should be monitored
for behavioral changes and for electrocardiographic abnormalities. Routine
assessment and treatment of delirium, along with sedation and analgesia,
should take place for the ventilated patient.
VII. Goals of Therapy: Conversations About Ventilation
Weaning is rarely an immediately successful procedure and should be viewed as a

process taking days, weeks, or months. While many patients willingly pursue treat-
ment of reversible causes of respiratory failure over many days or weeks, that sen-
timent is not universally true. Moreover, clinicians cannot consistently predict
whether failure can be reversed or what patients would wish. Even where respirat-
ory dysfunction can be improved, the time and complications attending ventilation
and weaning may not be acceptable to all patients. In the end, the weaning process
may yield unanticipated low levels of physical function. In order to initiate and
continue a course of weaning, it is best practice to address the goals of therapy
and the values related to those goals early in the weaning process (or even prior
to ventilation if possible). This section provides a rubric to discuss patient
wishes and construct clinical plans consistent with patient values.
Most patients with known chronic respiratory disease have no discussion of
goals of therapy prior to initiation of mechanical support (121). However, one
Table 6 Risk Factors for Delirium
Age over 70 yr
Alcohol abuse (within a month)
Analgesics or sedatives (benzodiazepines or narcotics)
BUN/creatinine ratio .18
Cardiogenic or septic shock
Central venous catheters
Drug overdose or illicit drug use (within week)
History of congestive heart failure
History of stroke, epilepsy
HIV infection
Hypo- or hyperglycemia
Hypo- or hypernatremia
Hypo- or hyperthermia (fever)
Hypo- or hyperthyroidism
Liver disease (T-bilirubin .2 mg/dL)
Malnutrition
Physical restraints (including posey vest)
Prior history of deperession
Rectal or bladder catheters
Renal failure (Cr . 2 mg/dL)
Transfer from a nursing home
Tube feeding or parenteral nutrition
Visual or hearing impairment
study suggests that the vast majority of patients in the inpatient setting do wish to
discuss advance directives or goals of therapy (122). Some may have thought
carefully about the possibilities but will usually not have had a prior experience
with mechanical ventilation (Fig. 5). Moreover, studies report a dissociation
between surrogate statements of wishes and those of the patient, especially in
cases where a patient wishes not to be intubated/resuscitated (123).
One approach to addressing goals is to use a structured approach to discuss-
ing goals and values with the patient to elicit specic values regarding venti-
lation. In a structured approach, the clinician explores the patients health
status (baseline), minimal acceptable health status/outcome, willingness to
sustain burdensome therapy, and expressed wishes regarding chances of
success or duration of therapeutic trials. Then ventilation plans, intubations, extu-
bations, and sedation strategies are adapted to the matrix of values (Table 7).
Advance directives can be recommended based on a patients goals and values
and may change as clinical probabilities or feelings about burden or duration
of therapy changes.
The rst concept in laying out goals is the burden of therapy. Prior to initi-
ating mechanical ventilation, a patient should be willing to sustain the burdens of
Figure 5 Flow chart of patients responses to questions about prolonged mechanical

ventilation. Shown are the results of patients responses to three questions on preferences
about prolonged mechanical ventilation. Similar questions asked about resuscitation.
Source: Adapted from Hofmann: Ann Intern Med 1997; 127(1):1 12.
mechanical ventilation for the expected duration of mechanical support. For a

patient with late-stage amyotrophic lateral sclerosis, the decision to ventilate
may be a lifelong one. On the other hand, with non-invasive options, tracheos-
tomies, and small home ventilators, the option may be favorable for some patients
(124,125). For an otherwise healthy young patient with pneumococcal pneumo-
nia, the ventilation may be a brief, reversible occurrence.
A second concept, minimal acceptable health status or minimal acceptable
outcome, does not depend on a particular therapeutic intervention. For some
patients, long-term tracheostomy and mechanical ventilation may be an accepta-
ble and tolerable outcome. For others, nothing short of independence of mechan-
ical devices and independent living is acceptable.
For many patients, minimal acceptable health status can be expressed in
terms of activities of daily living or as a setting (to be home, in assisted living,
in a long-term ventilator facility, etc.). Certain patients may state their wishes in
terms of a level of cognitive or neurological function (full cognition, the ability
to walk, and the ability to communicate with family or loved ones). If the
minimal acceptable outcome cannot be achieved, ventilation is not indicated.
It is especially important to take into account religious and cultural values,
as some patients and families make distinctions on initiating and withdrawing
mechanical ventilation. When given the clinical probabilities prior to pursuing
a path of mechanical ventilation, even those patients may not choose to
embark on ventilation. The clinicians role in the case of ventilation is to articu-
late whether the outcomes of ventilation are likely to achieve the health status.
Table 7 Matrix for Addressing Goals and Values for the Ventilated Patient
Concept (Status, goal,

value) Common choices Ventilation plan
Baseline: Baseline Unacceptable according to Do not initiate ventilation;
health status or patient or valid health encourage Advance
quality of life agent Directive
Unacceptable but patient/ Do not initiate; do not
proxy express religious withdraw; may be
belief that distinguishes appropriate to place DNAR
between initiating and order and not to initiate
withdrawing other therapies
Acceptable Proceed
Declining (e.g., ALS patient) Enter discussions to establish
minimal acceptable
outcome
Minimal acceptable Life at any level of function Proceed only if: medically
outcome: Minimal for any duration achievable, medically
acceptable health LTAC dependence, stable indicated, burden of
status or quality of medically (e.g., end-stage therapy is acceptable along
life emphysema with dialysis) path to goal; consult
Limited ADLs bioethics where futility
Cognition, assisted living, considered; unachievable
independent living, and outcomes: recommend
other choices Advance Directive
Freedom from ventilation
Maximal burden of Comfort is paramount: Discuss ventilation and
therapy: Maximal unwilling to experience establish plan for sedation.
tolerable pain and pain or discomfort If patient or proxy still
suffering and expresses unhappiness with
possibly a specied plan or experienced trial of
duration ventilation, consider
ceasing ventilation/
Advance Directive
Willing to sustain painful, Proceed with ventilation if
invasive, or debilitating outcome is achievable
processes to reach
outcome
Willing to proceed when Ventilate if clinical odds
chances of success are_% exceed patients chances of
success. Re-evaluate and
adjust Advance Directive
based on clinical prognosis.
Willing to proceed for a Ventilate and meet again to
duration of time (days) discuss with patient/proxy
Thirdly, the concept of a patient baseline informs the discussion of venti-

lation goals. While a debilitated, end-stage emphysematous patient may
recover from an episode of acute respiratory failure, mechanical support may
not be acceptable if the baseline itself is unacceptable. For some patients, even
a very low level of function, a poor health status, and a declining baseline are
acceptable. For these cases, personal values of preservation of life at any cost
may place providers in a quandary with respect to futile care and distributive
justice.
Finally, goals of therapy may change. With more data available to the
medical team about clinical prognosis and probabilities of success, or with
additional experience available to the patient or proxy about the conditions of
mechanical ventilation, both recommendations and decisions may change. It is
important to consider that clinical odds of success with ventilation may change
bedside decisions. For instance, the cystic brosis patient with multiple airway
complications, recurrent and resistant nosocomial infections, and requirements
for frequent chest physiotherapy may initially pursue all available respiratory
support; after experiencing prolonged ventilator dependence and with a reason-
able expectation of weeks or months of mechanical ventilation, the patient
may now elect palliative care. Over time and with exposure to the medical
setting, patients or their spokespersons may decide to forgo or initiate mechanical
ventilation with a better understanding of their own values and the interventions
clinicians offer.
Acknowledgment
Special thanks to Lawrence Maldonado, MD, for his assistance with the Goals
of Therapy: Conversations About Ventilation section and Table 7.
References
1. Esteban A, Alia I, Ibanez J, Benito S, Tobin MJ. Modes of mechanical ventilation

and weaning. A national survey of Spanish hospitals. The Spanish Lung Failure
Collaborative Group. Chest 1994; 106(4):1188 1193.
2. Roussos C, Macklem PT. Diaphragmatic fatigue in man. J Appl Physiol 1977;
43:189 197.
3. Bellemare F, Grassino A. Effect of pressure and timing of contraction on human
diaphragm fatigue. J Appl Physiol 1982; 53:1190 1195.
4. Bellemare F, Wright D, Levine CM, Grassino A. Effect of tension and timing
of contraction on the blood ow of the diaphragm. J Appl Physiol 1983;
54:1597 1606.
5. Field S, Sanci S, Grassino A. Respiratory oxygen consumption estimated by the
diaphragm pressure time index. J Appl Physiol 1984; 57:44 51.
6. Zakynthinos SG, Vassilakopoulos T, Roussos CT. The load of inspiratory muscles

in patients needing mechanical ventilation. Am J Respir Crit Care Med 1995;
152:1248 1255.
7. Lessard MR, Brochard LJ. Weaning from ventilatory support [Review] [96 refs].
Clin Chest Med 1996; 17(3):475 489.
8. NHLBI W. Respiratory muscle fatigue: report of the respiratory muscle fatigue
workshop group. Am Rev Respir Dis 1990; 142:474 480.
9. El Dwairi Q, Comtois A, Guo Y, Hussain SN. Endotoxin-induced skeletal muscle
contractile dysfunction: contribution of nitric oxide synthases. Am J Physiol 1998;
274:C770 C779.
10. Fitts RH. Cellular mechanisms of muscle fatigue. Physiol Rev 1994; 74:49 94.
11. Stofan DA, Callahan LA, DiMarco AF, Nethery DE, Supinski GS. Modulation of
release of reactive oxygen species by the contracting diaphragm. Am J Respir
Crit Care Med 2000; 161:891 898.
12. Laghi F, DAlfonso N, Tobin MJ. Pattern of recovery from diaphragmatic fatigue
over 24 hours. J Appl Physiol 1995; 79:539 546.
13. Reid WD, Belcastro AN. Time course of diaphragm injury and calpain activity
during resistive loading. Am J Respir Crit Care Med 2000; 162:1801 1806.
14. Efthimiou J, Fleming J, Spiro SG. Sternomastoid muscle function and fatigue in
breathless patients with servere respiratory disease. Am Rev Respir Dis 1987
Nov; 136(5):1099 1105.
15. Laghi F, Cattapan SE, Jubran A, et al. Is weaning failure caused by low-frequency
fatigue of the diaphragm? Am J Respir Crit Care Med 2003; 167(2):120 127.
16. Jubran A, Tobin MJ. Pathophysiologic basis of acute respiratory distress in patients
who fail a trial of weaning from mechanical ventilation. Am J Respir Crit Care Med
1997; 155(3):906 915.
17. Jubran A, Mathru M, Dries D, Tobin MJ. Continuous recordings of mixed venous
oxygen saturation during weaning from mechanical ventilation and the ramica-
tions thereof. Am J Respir Crit Care Med 1998; 158(6):1763 1769.
18. Pepe PE, Marini JJ. Occult positive end-expiratory pressure in mechanically venti-
lated patients with airow obstruction: the auto-PEEP effect. Am Rev Respir Dis
1982; 126:166 170.
19. Ranieri VM, Giuliani R, Cinnella G, et al. Physiologic effects of positive end-
expiratory pressure in patients with chronic obstructive pulmonary disease during
acute ventilatory failure and controlled mechanical ventilation. Am Rev Respir
Dis 1993; 147:5 13.
20. Gottfried S. The role of PEEP on CPAP in the mechanically unilateral COPD
patient. In: Roussos C, ed. The Thorax. New York: Marcel, 1995.
21. Zakynthinos SG, Vassilakopoulos T, Zakynthinos E, Roussos C, Tzelepis GE.
Correcting static intrinsic positive end-expiratory pressure for expiratory muscle con-
traction validation of a new method. Am J Respir Crit Care Med 1999; 160:785790.
22. Tuxen DV. Detrimental effects of positive end-expiratory pressure during
controlled mechanical ventilation of patients with severe airow obstruction. Am
Rev Respir Dis 1989; 140:5 9.
23. Corne S, Gillespie D, Roberts D, Younes M. Effect of inspiratory ow rate on
respiratory rate in intubated ventilated patients. Am J Respir Crit Care Med
1997; 156:304 308.
24. Girault C, Breton L, Richard JC, et al. Mechanical effects of airway humidication
devices in difcult to wean patients. Crit Care Med 2003; 31:1306 1311.
25. Talpers SS, Romberger DJ, Bunce SB, Pingleton SK. Nutritionally associated
increased carbon dioxide production: excess total calories vs high proportion of
carbohydrate calories. Chest 1992; 102:551 555.
26. Oh TE, Bhatt S, Lin ES, Hutchinson RC, Low JM. Plasma catecholamines and
oxygen consumption during weaning from mechanical ventilation. Intensive Care
Med 1991; 17:199 203.
27. Lemaire F, Teboul JL, Cinotti L, et al. Acute left ventricular dysfunction during
unsuccessful weaning from mechanical ventilation. Anesthesiology 1988;
69:171 179.
28. Meade M, Guyatt G, Cook D, et al. Predicting success in weaning from mechanical
ventilation. Chest 2001; 120(suppl 6):400S 424S.
29. Hess DR. Liberation from mechanical ventilation: weaning the patient or weaning
old-fashioned ideas? Crit Care Med 2002; 9:2154 2155.
30. Lessard MR, Brochard LJ. Weaning from ventilatory support. Clin Chest Med
1996; 17(3):475 489.
31. Wright PE, Marini JJ, Bernard GR. In vitro versus in vivo comparison of endotra-
cheal tube airow resistance. Am Rev Respir Dis 1989; 140:10 16.
32. Sahn SA, Lakshminarayan S. Bedside criteria for discontinuation of mechanical
ventilation. Chest 1973; 63:1002 1005.
33. Yang KL, Tobin MJ. A prospective study of indexes predicting the outcome of trials
of weaning from mechanical ventilation. N Engl J Med 1991; 324(21):14451450.
34. Stroetz RW, Hubmayr RD. Tidal volume maintenance during weaning with
pressure support. Am J Respir Crit Care Med 1995; 152:1034 1040.
35. Rivera L, Weissman C. Dynamic ventilatory characteristics during weaning in post-
operative critically ill patients. Anesth Analg 1997; 84:1250 1255.
36. Marini JJ, Smith TC, Lamb V. Estimation of inspiratory muscle strength in
mechanically ventilated patients: the measurement of maximal inspiratory pressure.
J Crit Care 1986; 1:32 38.
37. Tobin MJ, Perez W, Guenther SM, et al. The pattern of breathing during successful
and unsuccessful trials of weaning from mechanical ventilation. Am Rev Respir Dis
1986; 134(6):1111 1118.
38. Vallverdu I, Calaf N, Subirana M, Net A, Benito S, Mancebo J. Clinical character-
istics, respiratory functional parameters, and outcome of a two-hour T-piece trial in
patients weaning from mechanical ventilation. Am J Respir Crit Care Med 1998;
158:1855 1862.
39. Epstein SK. Etiology of extubation failure and the predictive value of the rapid
shallow breathing index. Am J Respir Crit Care Med 1995; 152(2):545 549.
40. Epstein SK, Ciubotaru RL. Inuence of gender and endotracheal tube size on
preextubation breathing pattern. Am J Respir Crit Care Med 1996; 154(6 Pt
1):1647 1652.
41. Krieger BP, Isber J. Serial measurements of the rapid-shallow-breathing index as
a predictor of weaning outcome in elderly medical patients. Chest 1997; 11:10291034.
42. Jabour ER, Rabil DM, Truwit JD, Rochester DF. Evaluation of a new weaning
index based on ventilatory endurance and the efciency of gas exchange. Am
Rev Respir Dis 1991; 144:531 537.
43. Yang KL. Inspiratory pressure/maximal inspiratory pressure ratio: a predictive

index of weaning outcome. Intensive Care Med 1993; 19:204 208.
44. Sassoon C, Mahutte C. Airway occlusion pressure and breathing as predictors of
weaning outcome. Am Rev Respir Dis 1993; 148:860 866.
45. Murciano D, Boczkowski J, Lecocguic Y, Emili JM, Pariente R, Aubier M.
Tracheal occlusion pressure: a simple index to monitor respiratory muscle
fatigue during acute respiratory failure in patients with chronic obstructive pulmon-
ary disease. Ann Intern Med 1988; 108:800 805.
46. Fernandez R, Raurich JM, Mut T, Blanco J, Santos A, Villagra A. Extubation
failure: diagnostic value of occlusion pressure (P01) and P01-derived parameters.
Intensive Care Med 2004; 30:234 240.
47. Ashutosh K, Lee H, Mohan CK, Ranka S, Mehrotra K, Alexander C. Prediction
criteria for successful weaning from respiratory support: statistical and connection-
ist analyses. Crit Care Med 1992; 20:1295 1301.
48. Miwa K, Mitsuoka M, Takamori S, Hayashi A, Shirouzu K. Continuous monitoring
of oxygen consumption in patients undergoing weaning from mechanical venti-
lation. Work of breathing: reliable predictor of weaning and extubation. Respiration
2003; 70:623 630.
49. Shikora SA, Bistrian BR, Borlase BC, Blackburn GL, Stone MD, Benotti PN. Work
of breathing: reliable predictor of weaning and extubation. Crit Care Med 1990;
18(2):157 162.
50. Petros AJ, Lamond CT, Bennett D. The bicore pulmonary monitor. A device to
assess the work of breathing while weaning from mechanical ventilation. Anaesthe-
sia 1993; 48:985 988.
51. Banner MJ, Kirby RR, Gabrielli A, Blanch PB, Layon AJ. Partially and totally
unloading respiratory muscles based on real-time measurements of work of breath-
ing: a clinical approach. Chest 1994; 106:1835 1842.
52. Tobin MJ, Guenther SM, Perez W, et al. Konno-Mead analysis of ribcage-abdomi-
nal motion during successful and unsuccessful trials of weaning from mechanical
ventilation. Am Rev Respir Dis 1987; 135(6):1320 1328.
53. Tobin MJ, Perez W, Guenther SM, Lodato RF, Dantzker DR. Does rib cage-
abdominal paradox signify respiratory muscle fatigue? J Appl Physiol 1987;
63:851 860.
54. Mohsenifar Z, Hay A, Hay J, Lewis MI, Koerner SK. Gastric intramural pH as a
predictor of success or failure in weaning patients from mechanical ventilation.
Ann Intern Med 1993; 119:794 797.
55. Bouachour G, Guiraud MP, Gouello JP, Roy PM, Alquier P. Gastric intramucosal
pH: an indicator of weaning outcome from mechanical ventiolation in COPD
patients. Eur Resp J 1996; 9:1868 1873.
56. Hurtado FJ, Beron M, Olivera W, et al. Gastric intramucosal pH and intra-
luminal PCO2 during weaning from mechanical ventilation. Crit Care Med 2001;
29:70 76.
57. Martinez A, Seymour C, Nam M. Minute ventilation recovery time: a predictor of
extubation outcome. Chest 2003; 123(4):1214 1221.
58. Calhoun CJ, Specht NL. Standardizing the weaning process. AACN Clin Issues Crit
Care Nurs 1991; 2(3):398 404.
59. Tomlinson JR, Miller KS, Lorch DG, Smith L, Reines HD, Sahn SA. A prospective
comparison of IMV and T-piece weaning from mechanical ventilation. Chest 1989;
96(2):348 352.
60. Dries DJ. Weaning from mechanical ventilation. J Trauma 1997; 43(2):372
384.
61. Scheinhorn DJ, Chao DC, Stearn-Hassenpug M, Wallace WA. Outcomes in post-
ICU mechanical ventilation: a therapist-implemented weaning protocol. Chest
2001; 119(1):236 242.
62. Thomas E, Paulson SS. Protocol for weaning the SCI patient. SCI Nurs 1994;
11(2):42 45.
63. Kollef MH, Horst HM, Prang L, Brock WA. Reducing the duration of mechanical
ventilation: three examples of change in the intensive care unit. New Horiz 1998;
6(1):52 60.
64. Esteban A, Alia I, Ibanez J, Benito S, Tobin MJ. Modes of mechanical ventilation
and weaning: a national survey of Spanish hospitals. Spanish Lung Failure Colla-
borative Group Chest 1982; 106:1188 1193.
65. Esteban A, Anzueto A, Alia I, et al. How is mechanical ventilation employed in the
intensive care unit? An international utilization review. Am J Respir Crit Care Med
2000; 161(5):1450 1458.
66. Esteban A, Frutos F, Tobin MJ, et al. A comparison of four methods of weaning
patients from mechanical ventilation. Spanish Lung Failure Collaborative Group.
N Engl J Med 1995; 332(6):345 350.
67. Joiner GA, Salisbury D, Bollin GE. Utilizing quality assurance as a tool for redu-
cing the risk of nosocomial ventilator-associated pneumonia. Am J Med Qual 1996;
11(2):100 103.
68. Gluck EH, Barkoviak MJ, Balk RA, Casey LC, Silver MR, Bone RC. Medical
effectiveness of esophageal balloon pressure manometry in weaning patients
from mechanical ventilation. Crit Care Med 1995; 23(3):504 509.
69. Saura P, Blanch L, Mestre J, Valles J, Artigas A, Fernandez R. Clinical conse-
quences of the implementation of a weaning protocol. Intensive Care Med 1996;
22(10):1052 1056.
70. Davies N. Nurse-initiated extubation following cardiac surgery. Intensive Crit Care
Nurs 1997; 13(2):77 79.
71. Kollef MH, Shapiro SD, Silver P, et al. A randomized, controlled trial of protocol-
directed versus physician-directed weaning from mechanical ventilation. Crit Care
Med 1997; 25(4):567 574.
72. Wood G, MacLeod B, Moffatt S. Weaning from mechanical ventilation: physician-
directed vs a respiratory-therapist-directed protocol. Respir Care 1995;
40(3):219 224.
73. Hynes-Gay P, Brindley A, Chu W, et al. Implementation of an RN/RT-initiated
extubation protocol. Dynamics (Pembroke, Ont) 2000; 11(4):12 16.
74. Horst HM, Mouro D, Hall-Jenssens RA, Pamukov N. Decrease in ventilation time
with a standardized weaning process. Arch Surg 1998; 133(5):483 488.
75. Hospital cuts time on ventilators, ICU LOS. Health Benchmarks 1999;
6(5):57 59.
76. Hospital cuts length of stay for vent patients by 34%. Hosp Case Manag 1999;
7(9):152 156.
77. Randolph AG, Wypij D, Venkataraman ST, et al. Effect of mechanical ventilator
weaning protocols on respiratory outcomes in infants and children: a randomized
controlled trial.[see comment]. J Am Med Assoc 2002; 288(20):2561 2568.
78. Ely EW, Meade MO, Haponik EF, et al. Mechanical ventilator weaning protocols
driven by nonphysician health-care professionals: evidence-based clinical practice
guidelines. Chest 2001; 120(suppl 6):454S 463S.
79. Ely EW, Bennett PA, Bowton DL, Murphy SM, Florance AM, Haponik EF. Large
scale implementation of a respiratory therapist-driven protocol for ventilator
weaning. Am J Respir Crit Care Med 1999; 159(2):439 446.
80. Calhoun CJ, Specht NL. Standardizing the weaning process. AACN Clin Issues Crit
Care Nurs 1991 Aug 2002;398 404.
81. Krishnan JA, Moore D, Robeson C, Rand CS, Fessler HE. A prospective, controlled
trial of a protocol-based strategy to discontinue mechanical ventilation. Am J
82. Tobin MJ. Of principles and protocols and weaning. Am J Respir Crit Care Med
2004; 169(6):661 662.
83. Esteban A, Alia I, Tobin MJ, et al. Effect of spontaneous breathing trial duration on
outcome of attempts to discontinue mechanical ventilation. Spanish Lung Failure
Collaborative Group. Am J Respir Crit Care Med 1999; 159(2):512 518.
84. Frutos-Vivar F, Esteban A. When to wean from a ventilator: an evidence-based
strategy. Cleve Clin J of Med 1997; 70(5):389 393.
85. Alia I, Esteban A. Weaning from mechanical ventilation. Crit Care 2000 2004;
72 80.
86. Frutos-Vivar F, Esteban A. When to wean from a ventilator: an evidence-based
strategy. Cleve Clin J Med 2003; 70(5):389, 392 393, 397.
87. Girou E, Schortgen F, Delclaux C, et al. Association of noninvasive ventilation with
nosocomial infections and survival in critically ill patients. J Am Med Assoc 2000;
284(18):2361 2367.
88. Burns K, Adhikari N, Meade M. Noninvasive positive pressure ventilation as a
weaning strategy for intubated adults with respiratory failure. Cochrane Database
Systemat Rev 2004; 1, 2004.
89. Agency for Healthcare Research and Quality. Criteria for weaning from mechanical
ventilation. Evidence Report/Technology Assessment No. 23. [AHRQ Publication
No. 01-E010], 2003.
90. Petter AH, Chiolero RL, Cassina T, Chassot PG, Muller XM, Revelly JP. Automatic
respirator/weaning with adaptive support ventilation: the effect on duration of
endotracheal intubation and patient management. Anesth Analg 2003;
97(6):1743 1750.
91. Gluck E. Program and abstracts of Weaning96, Weaning from Prolonged
Mechanical Ventilation. 1996. Palm Springs, FL.
92. Woods JC, Mion LC, Connor JT, et al. Severe agitation among ventilated medical
intensive care unit patients: frequency, characteristics and outcomes. Intensive Care
Med 2004; 14.
93. Jacobi JPFB, Fraser GLPF, Coursin DBM, Riker RRM. Clinical practice guidelines
for the sustained use of sedatives and analgesics in the critically ill adult. Society of
Critical Care Medicine and the American Society of Health-System Pharmacists I,
ed. Crit Care Med 30(1), 119 141.
94. Milbrandt EB, Deppen S, Harrison PL, et al. Costs associated with delirium in
mechanically ventilated patients. Crit Care Med 2004; 1932:955 962.
95. Ely EW, Gautam S, Margolin R, et al. The impact of delirium in the
intensive care unit on hospital length of stay. Intensive Care Med 2001;
27(12):1892 1900.
96. Milbrandt EB, Deppen S, Harrison PL, et al. Costs associated with delirium in
mechanically ventilated patients. Crit Care Med 2004; 32(4):955 962.
97. Cheng E. The cost of sedating and paralyzing the critically ill patient. Crit Care Clin
1995; 11(4):1005 1019.
98. Barrientos-Vega R, Mar Sanchez-Soria M, Morales-Garcia C, Robas-Gomez A,
Cuena B, yensa-Rincon A. Prolonged sedation of critically ill patients with
midazolam or propofol: impact on weaning and costs [see comment]. Crit Care
Med 1997; 25(1):33 40.
99. Kress JP, Pohlman AS, OConnor MF, Hall JB. Daily interruption of sedative
infusions in critically ill patients undergoing mechanical ventilation. N Engl J
Med 2000; 342(20):1471 1477.
100. ACP J Club. Daily screening and breathing trials hastened the discontinuation of
mechanical ventilation. ACP J Club 1997; v126:74-Jun.
101. Mascia MF, Koch M, Medicis JJ. Pharmacoeconomic impact of rational use guide-
lines on the provision of analgesia, sedation, and neuromuscular blockade in critical
care. Crit Care Med 2000; 28(7):2300 2306.
102. Krachman SL DGCG. Sleep in the intensive care unit. Chest 1995; 107:
1713 1720.
103. Rothaar RC, Epstein SK. Extubation failure: magnitude of the problem, impact on
outcomes, and prevention. Curr Opin Crit Care 2003; 9(1):59 66.
104. Ferguson J GDPK. Dimensions of pain and analgesic administration associated
with coronary artery bypass in an Australian intensive care unit. J Adv Nurs
1997; 26:1065 1072.
105. NHS Centre for Reviews and Dissemination. Using and understanding
sedation scoring systems. Database of Abstracts of Reviews of Effectiveness
2004:2001.
106. Gamper G, Willeit M, Sterz F, et al. Life after death: posttraumatic stress disorder in
survivors of cardiac arrestprevalence, associated factors, and the inuence of
sedation and analgesia. Criti Care Med 2004; 378 383.
107. NHS Centre for Reviews and Dissemination. A lack of evidence of superiority of
propofol versus midazolam for sedation in mechanically ventilated critically ill
patients. Database of Abstracts of Reviews of Effectiveness 2004 2001.
108. Durbin CG. Sedation of the agitated, critically ill patient without an articial
airway. Critical Care Clinics 1995; 11(4):913 936.
109. Grounds RM, Lalor JM, Lumley J, Royston D, Morgan M. Propofol infusion for
sedation in the intensive care unit: preliminary report. Br Med J (Clin Res Ed)
1987; 294(6569):397 400.
110. Devlin JW, Holbrook AM, Fuller HD. The effect of ICU sedation guidelines and
pharmacist interventions on clinical outcomes and drug cost. Ann Pharmacother
1997; 31(6):689 695.
111. Jastremski CA. Pressure relief bedding to prevent pressure ulcer development in
critical care. J Crit Care 2002; 17(2):122 125.
112. Helman DL Jr, Sherner JH III, Fitzpatrick TM, Callender ME, Shorr AF. Effect of
standardized orders and provider education on head-of-bed positioning in mechani-
cally ventilated patients. Crit Care Med 2003; 31(9):2285 2290.
113. Berenholtz SM, Milanovich S, Faircloth A, et al. Improving care for the ventilated
patient. Jt Comm J Qual Saf 2004; 30(4):195 204.
114. Pitimana-aree S, Forrest D, Brown G, Anis A, Wang XH, Dodek P. Implementation
of a clinical practice guideline for stress ulcer prophylaxis increases appropriate-
ness and decreases cost of care. Intensive Care Med 1998; 24(3):217 223.
115. Kress JP, Pohlman AS, OConnor MF, Hall JB. Daily interruption of sedative infu-
sions in critically ill patients undergoing mechanical ventilation. N Engl J Med
2000; 342:1471 1477.
116. Cammarano WB, Pittet JF, Weitz S, Schlobohm RM, Marks JD. Acute withdrawal
syndrome related to the administration of analgesic and sedative medications in
adult intensive care unit patients. Crit Care Med 1998; 676 684.
117. Mayo-Smith MF, Beecher LH, Fischer TL, et al. Management of alcohol with-
drawal delirium: an evidence-based practice guideline. Arch Intern Med
164(13):1405 12, 2004.
118. Ely EW, Shintani A, Truman B, et al. Delirium as a predictor of mortality in
mechanically ventilated patients in the intensive care unit. J Am Med Assoc
2004 Apr 14 291; 1753 1762.
119. Jackson J, Hart R, Gordon Seal. Six-month neuropsychological outcome of medical
intensive care unit patients. Criti Care Med 2003; 31:1226 1234.
120. Ely EW, Margolin R, Francis J, et al. Evaluation of delirium in critically ill patients:
validation of the Confusion Assessment Method for the Intensive Care Unit (CAM-
ICU). Crit Care Med 2001; 29(7):1370 1379.
121. Travaline JM, Silverman HJ. Discussions with outpatients with chronic obstructive
pulmonary disease regarding mechanical ventilation as life-sustaining therapy.
Southern Med. Journal 1995; 88(10):1034 1038.
122. Reilly BM, Magnussen CR, Ross J, Ash J, Papa L, Wagner M. Can we talk?
Inpatient discussions about advance directives in a community hospital. Attending
physicians attitudes, their inpatients wishes, and reported experience [see
comment]. Arch Intern Med 1994; 154(20):2299 2308.
123. Sonnenblick M, Friedlander Y, Steinberg A. Dissociation between the wishes
of terminally ill parents and decisions by their offspring [see comment]. J Am
Geriatrics Soc 1993; 41(6):599 604.
124. Miller RG. Examining the evidence about treatment in ALS/MND [Review] [12
refs]. Amyotrophic Lateral Sclerosis Other Motor Neuron Disorders 2001;
2(1):3 7.
125. Simonds AK. Nasal ventilation in progressive neuromuscular disease: experience
in adults and adolescents [Review] [28 refs]. Monaldi Arch Chest Dis 2000;
55(3):237 241.
6
Prolonged Mechanical Ventilation
SCOTT K. EPSTEIN MICHAEL L. NEVINS
Department of Medicine Division of Pulmonary and Critical Care

Caritas St. Elizabeths Medical and Department of Medicine, Group
Center and Tufts University Health Permanente
School of Medicine University of Washington
Boston, Massachusetts, U.S.A. School of Medicine
Seattle, Washington, U.S.A.
I. Introduction
Most patients requiring invasive mechanical ventilation for acute respiratory

failure are rapidly liberated from the ventilator after improvement or resolution
of the acute precipitating illness. Nevertheless, approximately 5% to 20% of
patients prove more difcult to liberate taking many weeks to be successfully
weaned or, in a smaller number of cases, proving impossible to remove from mech-
anical ventilation. The denition of prolonged mechanical ventilation varies
whether one refers to those given by regulatory agencies or by investigators pub-
lishing in this eld. In earlier reports, patients with underlying obstructive lung
disease, severe chest wall deformities, or neuromuscular disease comprised the
majority of patients undergoing prolonged mechanical ventilation. More recently,
the demographics have changed, with patients suffering from acute lung injury,
prolonged acute hospitalization for multi-organ system failure, or post-operative
complications making up the majority requiring prolonged ventilation (1).
Prior to 1990, patients needing prolonged mechanical ventilation were
almost exclusively cared for in intensive care units (ICUs). Patients requiring
prolonged mechanical ventilation comprise a minority of the ICU patient popu-
lation yet account for a disproportionately high percentage of ICU days, mechan-
ical ventilation days, and ICU costs (2). Up to one quarter of patients requiring
prolonged support were eventually discharged home while still needing invasive
mechanical ventilation. Beginning in the early 1990s in the United States, the
187
188 Epstein and Nevins
locus of care began to shift to acute hospital-based long-term weaning units or in

free-standing, long-term acute care hospitals. A fundamental principal of these
centers is that patients have recovered from their acute illness prior to transfer
and thus need a lower intensity of care (including a lower nurse-to-patient stafng
ratio) (1).
Although some patients are referred for long-term custodial care and some
for training in domiciliary mechanical ventilation and airway care prior to their
discharge home, most are admitted with a goal of liberation from mechanical ven-
tilation. Therefore, this chapter will principally focus on weaning from prolonged
mechanical ventilation. Special emphasis will be placed on the pathophysiology of
weaning failure, identifying predictors of weaning outcome, and the techniques
used to liberate patients from prolonged mechanical ventilation. This chapter
will not examine patients who require prolonged mechanical ventilation for
chronic progressive respiratory or neuromuscular disease. Neither will it cover
the area of home mechanical ventilation. Recent reviews have expertly covered
this important subset of patients requiring prolonged mechanical ventilation.
II. How Is Prolonged Mechanical Ventilation Dened and

How Often Does It Occur?
An essential question concerning patients intubated for acute respiratory failure

concerns the time at which a patient shifts from requiring mechanical ventilation
as supportive care during the resolution of an acute process to requiring pro-
longed mechanical ventilation. Authors of earlier reports have dened a period
ranging from 24 hours to 29 days (3 6). Alternatively, DRG 483 (the need for
tracheotomy, not for head and neck cancer) has been used to dene prolonged
mechanical ventilation. If one excludes trauma patients, it is likely that the
vast majority of DRG 483 patients require at least 21 days of mechanical venti-
lation. Indeed, the majority of experts in the eld have adopted the denition
employed by the Centers for Medicare & Medicaid Services (CMS); a
minimum six hours per day for 21 days (7).
The denition of prolonged mechanical ventilation clearly inuences the
incidence of this entity. For example, a large international observational study
conducted in 361 ICUs found that 25% of patients required more than seven
days of mechanical ventilation (8). It has been estimated that 5% to 20% of all
patients who require invasive mechanical ventilation will go on to require
more than 21 days of ventilation (9 12). As an example, in a cohort of nearly
600 medical patients admitted to a tertiary care intensive care unit, approximately
10% remained invasively ventilated at day 21 (13).
Another approach to analysis is determining the absolute number of
patients requiring prolonged mechanical ventilation. As an example, in 1985
there were 147 patients in Massachusetts who required mechanical ventilation
for more than three weeks, which extrapolates to 6800 patients nationwide
(14). A study performed a decade later in Massachusetts identied a similar
Prolonged Mechanical Ventilation 189
number of patients requiring prolonged mechanical ventilation, subsequently

arriving at a national estimate of 7250 patients (15,16). In 1990, the American
Association for Respiratory Care and the Gallup Organization conducted a
cross-sectional study using a randomly selected sampling of respiratory care
directors and pulmonologists to determine the number of patients needing mech-
anical ventilation for at least six hours a day for 30 or more days (15,16). This
study estimated that more than 11,000 patients in the United States required
chronic mechanical ventilation. In contrast, the MEDPAR database of Medicare
discharges noted .43,000 DRG 483 discharges in the United States in 1998 (17).
Indeed, an estimated 88,000 patients (including those under age 65) were coded
DRG 483 in an analysis of a National Inpatient Sample database (18).
Numerous investigators have developed tools for predicting survival in the
acute care setting. Carson and Bach determined APACHE II, SAPS II, MPM II,
and the logistic organ dysfunction score in 182 patients with prolonged critical
illness admitted to a long-term acute care (LTAC) hospital in the Chicago area.
The investigators found that none of the four severity of illness indices
distinguished well between patients who lived and those who died (19). In
contrast, a retrospective study of four LTACs found that a model containing age
and organ failure provided adequate discrimination between survivors and non-
survivors (20). Examining predictors of one-year survival, Carson et al. (21) found a
model containing age and prior functional status identied a group of patients
at very high risk of death. Specically, one-year survival was 5% for patients
over 75 years old or those older than 65 who also had poor prior functional status.
Recent interest has also focused on quality of life for patients with pro-
longed mechanical ventilation. As an example, 90% of patients with prolonged
mechanical ventilation managed in the LTAC setting are not functionally inde-
pendent at one year (22). Of those who survive and are capable of responding
to survey interviews, the reported quality of life is good (23 25).
III. Who Requires Prolonged Mechanical Ventilation?
The etiology of acute respiratory failure likely plays a role in determining which
patients go on to require prolonged mechanical ventilation. In a study of 183
patients with chronic obstructive pulmonary disease (COPD) and acute respirat-
ory failure admitted to a medical ICU, 10.4% remained mechanical ventilated
21 days after intubation (26). In contrast, in the same ICU, the relative risk of
remaining ventilated at day 21 was twice as high in patients intubated
for acute lung injury (21 of 107) (27). In a prospective study of 23 long-term
acute care hospitals, 1411 patients were admitted after a median of 25 days
(range: 0 273 days) of invasive mechanical ventilation at the referring acute
care hospital (28). The average age was 72 years, and 50% were women.
Seventy percent of patients were smokers, averaging 57 pack years. In this
cohort, 43% had COPD, 54% coronary artery disease or congestive heart
failure, and 20% neurologic disease. Among patients with respiratory failure
secondary to medical illness (61% of the cohort), the precipitating etiology was
bacterial pneumonia (34%), sepsis (21%), acute neurological disease (20%),
acute exacerbation of COPD (17%), congestive heart failure (14%), and
aspiration pneumonia (13%). In 39% of cases, respiratory failure followed a sur-
gical procedure (44% cardiovascular and 22% gastrointestinal). Despite the
advanced age of this cohort, other investigators have not found older patients
to be at higher risk for prolonged mechanical ventilation compared with
younger patients, after controlling for confounding factors (29,30).
IV. Site of Care for Patients with Prolonged

Mechanical Ventilation
Prior to 1990, nearly all patients requiring prolonged mechanical ventilation were
cared for in the ICU, with 14% to 25% eventually being discharged home
while still requiring mechanical ventilation (14,16). Subsequently, there was a
proliferation of centers specializing in the care and weaning of patients with pro-
longed mechanical ventilation. These have taken the form of specialized units
within acute care hospitals and freestanding LTAC or post-acute care hospitals.
In a 10-year follow-up study, the percentage of patients receiving prolonged
mechanical ventilation in an acute hospital setting decreased (from 62% to
18%), while the number of patients in post-acute care (from 22% to 46%) and
long-term care (from 2% to 24%) facilities increased (15).
One principle used to support the expansion of these centers is that patients
have recovered from their acute illness prior to transfer and therefore require a
lower intensity of care. As an example, nurse-to-patient stafng ratios are typi-
cally lower in chronic ventilator facilities than in ICUs (31). This may have
important implications because appropriate nursing care appears crucial for
favorable weaning outcomes in COPD patients on invasive mechanical venti-
lation in the acute care setting (32). In a European investigation, the duration
of mechanical ventilation increased from 7.3 to 38.2 days when the index of
nursing (a measure of the effective workforce of nurses as dened by both
stafng ratios and nursing qualications) fell. Indeed, the average duration of
mechanical ventilation decreased to 9.9 days when the initial index of nursing
was restored. Additionally, patients are now transferred earlier and with a rela-
tively high severity of illness (33). As an example, Scheinhorn et al. (34)
originally reported on 421 patients admitted to their regional weaning center,
observing a mean duration of mechanical ventilation of 49 days prior to
LTAC transfer (34). In subsequent reports, the authors noted a signicant trend
toward shorter duration of pre-admission mechanical ventilation (median:
33 days) and higher admission severity of illness scores (median APS of
APACHE III: 32) (33). A recent preliminary update totaling 1600 patients
described a continued, rising trend in severity of illness scores in their patients
that had become comparable to those reported at the time of ICU admission
(35). These single-center observations are supported by the ndings in a

multi-center observational study, where an average APACHE III acute physi-
ology score of 36 was found at time of admission to the LTAC (28). These
APACHE III scores should be interpreted in the context of a study of 116,000
patients admitted to 28 acute care ICUs in the Cleveland area where the
median acute physiology score was only slightly higher at 41 (36).
As patients with prolonged mechanical ventilation comprise a minority of
those in acute ICUs, any one intensivists experience may be limited. The funda-
mental importance of the physicians role in the care of patients with prolonged
mechanical ventilation has been conrmed. Indeed, the high rates of weaning
success in certain venues indicate that some experts are adept at identifying
patients likely to be successfully liberated (4,37,38). Physicians also inuence
many objective outcomes, including the rate of weaning from and the duration
of mechanical ventilation, LTAC length of stay, and rates at which care is
withdrawn from patients with terminal illness (39). Indeed, with increasing
experience, reductions in the time for successful weaning from ventilation,
lower mortality, and higher rates of discharge to home can be achieved (40,41).
When considering patients transferred with prolonged mechanical venti-
lation, a minority of patients are referred for long-term custodial care (41), train-
ing in domiciliary mechanical ventilation, and airway care prior to their discharge
home (37). The overwhelming majority of patients are transferred with a goal of
liberation from mechanical ventilation. When full liberation from mechanical
ventilation is not feasible, achievement of the alternative goals of nocturnal inva-
sive mechanical ventilation or conversion to non-invasive ventilation may allow
for successful discharge to home. Despite all efforts, there remains a small, stable
subset of unweanable patients who are transferred home with complete mecha-
nical ventilation.
V. What Are the Outcomes for Patients Requiring

Prolonged Mechanical Ventilation?
The characteristics of individual populations (age, severity of illness, etiology of

acute respiratory failure, presence of multiple organ dysfunction, and comorbid
conditions) signicantly affect patient outcomes and must be considered when
analyzing this literature. Admission criteria can vary, with many facilities
requiring that a patient be on mechanical ventilation for a minimum of three
weeks and have failed at least two weaning attempts prior to their transfer to
the LTAC (38,42). Various institutions have different levels of care they can
administer, and their capabilities may affect their patient selection and outcomes.
Some are equipped with operating rooms capable of providing day-surgery level
services, whereas others maintain small ICUs with the ability for invasive
hemodynamic monitoring should a patient become unstable after transfer.
Centers lacking this capability may report a lower mortality if critically ill
patients are transferred (and die) in an outside acute care facility. There are no
randomized controlled trials comparing outcomes for patients cared for in
acute ICU versus those managed in an LTAC. Gracey et al. (4,37) compared
patients managed in their acute care ICU at the Mayo Clinic to similar patients
requiring prolonged mechanical ventilation managed in their chronic
ventilator-dependent unit (CVDU). Patients treated in the CVDU had a signi-
cantly lower mortality rate (9.8% vs. 28.6%, P 0.001) after excluding patients
with multi-organ failure to control for discrepancies between the two populations.
Mortality rates range from 7% to 61% in single-center reports depending on
the setting (acute care hospital vs. LTAC) and the denition of prolonged mechan-
ical ventilation (Table 1). In a large multi-center trial of more than 1400 patients
admitted to 23 LTACS around the United States, Scheinhorn et al. (28) noted a
26% mortality rate. Some investigators have focused on more long-term outcomes,
such as one-year survival, because of the limitations noted above. Indeed, one-year
survival ranges from just 23% to 38%, likely a reection of underlying comorbid
conditions, an ongoing debilitated state, and vulnerability to recurrent acute
disease experienced by short-term survivors of prolonged mechanical venti-
lation (21,33). Other studies found that long-term survival was correlated with
younger age and shorter pre-CVDU duration of mechanical ventilation (38).
Patients who fail extubation are more likely to require prolonged
mechanical ventilation. As an example, 12 additional days of invasive mecha-
nical ventilation were required for a cohort of medical patients who needed rein-
tubation after planned extubation (43). The need for reintubation also identies
patients more likely to need tracheostomy and eventual transfer to a post-acute
care facility (43 45).
Weaning success rates vary widely depending on the denition of success,
the population studied, and the admission criteria applied by the unit under
investigation. When looking at weaning success at hospital discharge, rates
range from 25% to 82% in single-center studies (Table 1). In a prospective
multi-center trial, 54% of 1411 patients (73% of hospital survivors) were charac-
terized as successfully liberated from mechanical ventilation at the time of
discharge from 23 long-term acute care units (28). Twenty-seven percent of
survivors (20% of all patients admitted) remained ventilator-dependent at the
time of hospital discharge.
VI. Why Do Patients Become Ventilator-Dependent?
There are numerous factors that may contribute to ventilator dependency in

patients intubated for acute respiratory failure (Table 2).
A. Systemic Factors
As noted above, patients with prolonged mechanical ventilation have persistently
high severity illness (as assessed by APACHE acute physiology scores) and
Prolonged Mechanical Ventilation
Table 1 Outcomes for Patients Requiring Prolonged Mechanical Ventilation
Duration of mechanical Weaned at

References Year Setting n ventilationb (days) Mortality (%) discharge (%)
ICU care
Sivak (150) 1980 ICU 15 14 7 66
Morganroth et al. (129) 1984 ICU 11 30 27 73
Spicher and white (5) 1987 ICU 245 10 61 ?
Gracey et al. (4) 1992 ICU 104 29 42 53
Non-ICU care
Indihar (41) 1991 PRCU 171 Median 36 33 34
Cordasco et al. (151) 1991 Non-ICU 99 ? 25 25
Gracey et al. (4) 1992 CVDU 61 21 9.5 82
Scheinhorn et al. (34) 1994 RWC 421 Mean 49 28 53
Nava et al. (123) 1994 IICU 42 21 29 36
Gracey et al. (37) 1995 CVDU 132 21 9.8 78
Latriano et al. (59) 1996 RCF 224 Mean 23 50 47
Gluck (42) 1996 LTVF 72 21 38 27 (at 42 days)
Gracey et al. (38) 1997 CVDU 206 21 8 70
Scheinhorn et al. (33) 1997 RWC 1123 Median 33 29 56a
(Continued)
193
194
Table 1 Outcomes for Patients Requiring Prolonged Mechanical Ventilation (Continued)
Duration of mechanical Weaned at

References Year Setting n ventilationb (days) Mortality (%) discharge (%)
Bagley and Cooney (40) 1997 RWC 278 ? 48 38

Scalise et al. (152) 1997 RWC 47 Mean 86 23 62
Bach et al. (39) 1998 LTAC 86 ? 52 34
Carson et al. (21) 1999 LTAC 133 14 50 35
Dasgupta et al. (58) 1999 ReSCU 212 ? 18 60
Gracey et al. (153) 2000 CVDU 549 ? 7 60
Schonhofer et al. (154) 2002 RWC 403 Mean 41 24 68
Scheinhorn et al. (28) 2003 LTAC 1411a Median 25 26c 54
a
Multi-center study.
b
Duration of mechanical ventilation minimum time on ventilation prior to transfer.
c
Outcomes reported on 1037 patients.
Abbreviations: CVDU, chronic ventilator-dependent unit; IICU, Intermediate intensive care unit; LTAC, long-term acute care facility; LTVF, long-term ventilator
Epstein and Nevins

facility; NIMU, non-invasive monitoring unit; PRCU, prolonged respiratory care unit; RCF, respiratory care oor; ReSCU, respiratory special care unit; RWC,
regional weaning center.
Table 2 Potential Causes for Ventilator Dependency
Mechanical factors
Increased work of breathing
Reduced respiratory muscle capacity
Critical illness polyneuropathy
Steroid myopathy
Disuse myopathy
Isolated phrenic nerve/diaphragmatic injury (e.g., post-cardiac surgery)
Imbalance between increased work of breathing and respiratory muscle capacity
Metabolic/systemic factors
Chronic comorbid conditions
(e.g., malignancy, chronic obstructive pulmonary disease, immunosuppression)
Overall severity of illness
Non-pulmonary organ failure
Poor nutritional status
Iatrogenic Factors
Imposed work of breathing from tracheotomy tubes
Upper airway obstruction (e.g., tracheal stenosis)
Recurrent aspiration
Infection (e.g., pneumonia, sepsis)
Psychological factors
Sedation
Delirium
Depression
Anxiety
Process of care factors
Absence of weaning protocols
Inadequate nursing stafng
Insufcient physician experience
frequently have important comorbid conditions (e.g., COPD) that likely contrib-
ute to persistent weaning failure. Indeed, a large proportion of patients transferred
to an LTAC with prolonged mechanical ventilation ultimately require transfer
back to acute care (46,47). In one study, of 97 patients (71 with prolonged mech-
anical ventilation) transferred to LTACs, 23% required readmission to acute
care within 30 days of transfer (47). Patients with prolonged mechanical venti-
lation frequently have non-pulmonary organ failure, the presence of which is
associated with poor outcome. In one study of 52 patients with prolonged mech-
anical ventilation and requiring hemodialysis, none were successfully weaned
and only three survived (48). Chao et al. (49) studied 63 patients with severe
renal dysfunction (creatinine .2.5 mg/dL) at the time of transfer to their
regional weaning center. Forty of these patients were on renal replacement
therapy at transfer, and hemodialysis was initiated in another 10 while at the
LTAC. Patients with severe renal dysfunction were less likely to be successfully
weaned (13% vs. 56%), and only four patients requiring renal replacement
therapy were liberated from the ventilator.
Cardiac disease often limits liberation from mechanical ventilation in the
acute setting. Myocardial ischemia (often occult), left ventricular dysfunction,
and pulmonary edema have all been documented during trials of spontaneous
breathing (50 55). Systematic examinations of cardiac factors have not
been published in patients requiring prolonged mechanical ventilation. In a pre-
liminary report, successful diuresis and weight loss were associated with weaning
success in patients transferred to a regional weaning center (56).
Poor nutritional status may contribute to prolonged mechanical ventilation
by adversely affecting respiratory muscle function and the ventilatory response
to gas exchange abnormalities, and by predisposing to infection (57). Indeed,
several studies of prolonged mechanical ventilation have noted an association
between hypoalbuminemia and weaning failure (58,59).
B. Mechanical Factors
An extensive body of research has identied key mechanical factors related to
weaning failure in the acute setting. These factors include abnormalities in respir-
atory drive, increased work of breathing (resistive, elastic, intrinsic PEEP),
decreased respiratory muscle capacity, cardiovascular factors (ischemia, pulmon-
ary edema), and psychological factors.
Numerous studies performed in acute hospital settings have shown patients
who fail weaning trials will usually have increased, not diminished, central res-
piratory drive (60 63). Nevertheless, there are some patients with central
nervous system disturbances or who are heavily sedated in whom weaning and
extubation may be delayed (64 66). In the large majority of patients who fail
weaning trials, the problem lies in an imbalance between the inspiratory load
and the capacity of the respiratory system (60,67,68). Patients failing weaning
trials often develop a pattern of rapid, shallow breathing in the face of increased
load (63). This pattern of breathing is seen within minutes of disconnection from
the ventilator suggesting that diaphragmatic fatigue is not the primary mechan-
ism involved, at least initially (69). Nevertheless, studies demonstrate that
weaning failure in the acute setting is associated with a tension time index
(TTI) of .0.15, which is thought to represent the threshold beyond which respir-
atory muscle fatigue ensues (60,68,70). As in patients with shorter durations of
ventilation, patients with a mean of 20 days who failed a trial of spontaneous
breathing also developed or maintained values of the TTI in the fatigue
zone (67). This possibility that fatigue might occur was further suggested by
Cohen et al. (71), who noted EMG evidence suggestive of high-frequency
diaphragmatic fatigue during failed weaning trials. Yet, in a study of eight
weaning success and 11 weaning failure patients, Laghi et al. (72) measured
twitch transdiaphragmatic pressure before and 30 minutes after spontaneous
breathing trials lasting up to 60 minutes. The weaning failure patients did not
develop low-frequency fatigue of the diaphragm despite considerable diaphrag-
matic weakness, and greater load and diaphragmatic effort. The absence of
fatigue may have resulted from increased recruitment of ribcage and expiratory
muscles or because of prompt reinstitution of ventilatory support (e.g., before
fatigue could ensue). The issue of fatigue is not trivial, as some studies in
normals indicate that it can take more than 24 hours to recover full diaphragmatic
function (73).
There is an emerging literature examining mechanical factors limiting
weaning in patients with prolonged mechanical ventilation. Appendini et al.
(74) monitored respiratory mechanics and diaphragmatic effort [using
pressure time product (PTP) and diaphragmatic TTI] in eight patients with
COPD requiring prolonged mechanical ventilation. Respiratory muscle strength
(maximal inspiratory, pleural, and transdiaphragmatic pressures) was found com-
patible with successful weaning but was countered by excessive load (increased
PEEPi, pulmonary resistance, and PTP) (74). Gluck (42) noted that patients with
prolonged mechanical ventilation demonstrated a pattern of rapid shallow breath-
ing, increased resistance, increased dead space, elevated dead space, and a
decreased compliance. Abnormalities in mechanics may not be overtly detect-
able. As an example, Reinoso et al. (75), using interrupter mechanics, detected
supramaximal ow transients after shutter valve release during passive
expiration, indicating occult airways disease in six of 25 patients tested.
Purro et al. (76) similarly evaluated 39 patients who had been intubated for
more than three weeks, 28 with COPD and 11 who had undergone cardiac surgery
[post-cardiac surgery (PCS)] complicated by diaphragmatic dysfunction. Control
groups were made up of nine patients with severe, stable COPD (maintained
with tracheotomies but not ventilator-dependent), and 11 PCS patients who
had been successfully liberated from mechanical ventilation within 48 hours of
their surgery. Eight of 28 COPD patients tolerated unsupported breathing for
60 minutes (WS-COPD), while the remaining 20 COPD patients (VD-COPD)
and all 11 PCS patients (VD-PCS) failed (mean time to failure: 40 min) and
needed reinstitution of ventilatory support. When compared with the stable
COPD and WS-COPD patients, the VD-COPD group demonstrated reduced
tidal volume, minute ventilation, maximal transdiaphragmatic pressure, and
maximal inspiratory pressure. The VD-COPD group demonstrated an increase
in central respiratory frequency, airway occlusion pressure at 0.1 seconds
(P0.1), PEEPi, and pulmonary resistance. Similarly, when compared with the
stable successfully extubated PCS patients, the VD-PCS group had reduced
tidal volume and maximum inspiratory pressure (MIP), while demonstrating
higher respiratory frequency and P0.1. Initially no differences were noted in the
mean respiratory rates between the stable COPD, WS-COPD, and VD-COPD
groups. In contrast, analysis of esophageal pressure tracings revealed untriggered
breaths (trigger asynchrony) in eight of 20 VD-COPD patients, resulting in a
central respiratory frequency 40% higher than measured from inspiratory
ow tracings (28 breaths/min vs. 20 breaths/min). As in investigations in the

acute setting, the majority of VD patients, but no weaned or stable patients,
had TTI values greater than 0.15 (Fig. 1) (77).
The ndings above demonstrate that ventilator dependence in prolonged
mechanical ventilation results from of an imbalance between inspiratory load
and respiratory capacity, that is, mechanisms similar to those operative in patients
failing weaning in the acute setting. Nevertheless, it is uncertain if the underlying
mechanisms leading to this imbalance are similar for the two populations. As an
example, the potential for disuse atrophy of the diaphragm and other inspiratory
muscles may be greater with prolonged mechanical ventilation. In fact, a large
number of animal studies document the adverse effect of controlled mechanical
ventilation on respiratory muscle function (78 82). These adverse effects, which
include decreased strength, increased fatigue, alterations in contractile properties,
remodeling, and myobril damage, can occur in less than 24 hours but appear
to become more signicant with increasing duration of controlled mechanical
ventilation. Sedated and paralyzed baboons demonstrated a 25% reduction in dia-
phragmatic strength and 36% in endurance after 11 days of controlled ventilation
with large tidal volumes (83).
Figure 1 Plots of respiratory duty cycle (Ti/Ttot) and respiratory muscle effort
(mean diaphragmatic pressure generated per breath divided by maximal diaphragmatic
pressure, Pdi/Pdimax) for VD, W, and stable patients with either COPD or PCS. The
product of the duty cycle and respiratory muscle effort is the TTI. Ventilator dependent
patients had higher TTI values, usually greater that the 0.15 threshold thought to represent
a fatiguing load on the respiratory system. Abbreviations: COPD, chronic obstructive
pulmonary disease; PCS, post-cardiac surgery; TTI; tension time index; VD, ventilator-
dependent; W, weaned. Source: Adapted from Ref. 76.
Steroid myopathy, alone (84) or in conjunction with paralytic agents

(85,86), delayed reversal of paralysis following the use of neuromuscular
blocking agents (87), and critical illness polyneuropathy have all been observed
in patients with prolonged mechanical ventilation (88). Critical illness polyneuro-
pathy may involve the diaphragm (89) and typically presents as failure to wean in
patients recovering from sepsis and multi-organ system failure (90).
Its occurrence has been tied to a higher in-hospital mortality rate and delayed
rehabilitation (91). Spitzer et al. (92) prospectively performed electromyographic
studies in 21 patients, without known history of neuromuscular disease, with
prolonged mechanical ventilation who failed weaning trials. Sixty-two percent
had evidence of acquired neuromuscular disease, half of which were consistent
with critical illness polyneuropathy or steroid myopathy. Similarly, Coakley
et al. (93) found that 95% of patients requiring seven or more days of intensive
care had myopathic or neurogenic changes on muscle biopsy. Leitjen et al. (91)
colleague studied 50 patients with at least seven days of invasive ventilation. The
58% of patients with EMG evidence of polyneuropathy were more likely to need
prolonged mechanical ventilation and post-acute care. De Jonghe et al. (94)
measured muscle strength at day 7 using the Medical Research Council Score,
considering a score ,48 indicative of ICU-acquired paresis in a prospective
clinical study of 95 patients without pre-existing neuromuscular disease (94).
Patients with ICU-acquired paralysis took twice as long to wean from mechanical
ventilation compared with those who did not acquire neuromuscular disease. The
role of respiratory muscle dysfunction is further supported by studies showing an
improvement over time in negative inspiratory force or maximal inspiratory
pressure when comparing patients failing weaning and then subsequently at the
time of weaning success (68,95).
C. Iatrogenic Factors
Tracheotomies have been shown to reduce the work of breathing imposed by an
articial airway when compared with endotracheal tubes, though the impact on
weaning success remains unclear (96). Whereas in the acute setting a majority
of patients are ventilated through oral or nasal endotracheal tubes, most patients
admitted to a chronic ventilator facility have tracheotomies in place (21,41).
Indeed, in a large multi-center trial, 95% transferred to 23 LTACs came with a
tracheotomy (28). Nevertheless, abnormalities of the upper airway resulting
from complications of the articial airway can contribute to ventilator depen-
dence. Ten percent of patients with prolonged mechanical ventilation will
develop tracheal injury despite the use of articial airways with low-pressure,
high-volume cuffs (97,98). Tracheal injury at or above the level of the tracheal
tube is typically clinically silent until the time of extubation or decannulation.
Rumbak et al. (99) identied 37 of 756 (5%) patients on prolonged mechanical
ventilation who had evidence for distal tracheal obstruction contributing to ven-
tilator dependence. All patients had transient elevations in peak airway pressures,
and difculty passing a suction catheter was uniformly present. Bypassing the
obstruction using either a longer tracheal tube or tracheal stent allowed 35 of
the 37 patients to be successfully liberated from mechanical ventilation within
one week.
Tracheostomies can also contribute to swallowing dysfunction and aspira-
tion in patients on prolonged mechanical ventilation. Indeed, 50% of 83 patients
had evidence of aspiration when studied using videoouroscopic tapes of modied
barium swallows. The majority of aspiration events were not accompanied by
clinical evidence of respiratory distress (100). In another investigation of 35
patients, 83% had swallowing dysfunction by videoouroscopy (101). Using scin-
tigraphy to identify aspiration, Schonhofer et al. (102) found a somewhat lower
rate of aspiration (30%) among patients requiring prolonged mechanical venti-
lation. Aspiration may contribute to weaning failure by compromising lung func-
tion, increasing respiratory secretions, predisposing to ventilator-associated
pneumonia, and possibly through compromising oral intake of nutrition.
In the acute setting, increased duration of mechanical ventilation has
been associated with increased risk of complications, including nosocomial
infection. In their prospective observation study, Scheinhorn et al. noted that
acquired infection was common: urinary tract infection (32%), lower respiratory
tract infection (28%), Clostridium difcile infection (18%), and central line infec-
tion (12%) (Fig. 2). Importantly, those who acquired lower respiratory tract infec-
tion had a longer length of stay, took longer to wean, were less likely to be
Figure 2 Effect of acquisition of lower respiratory tract infection on outcome for

patients with prolonged mechanical ventilation. Patients with lower respiratory tract infec-
tion (gray bars) were less likely to be successfully weaned (wean %), were more likely to
remain ventilator-dependent (vent %), and spent more days on mechanical ventilation and
in the hospital than patients without lower respiratory tract infection (black bars). Lower
respiratory tract infection occurred in 28% of 1411 patients admitted to 23 long-term acute
care facilities. Abbreviation: LOS, length of stay. Source: Adapted from Ref. 103.
liberated, and were more likely to remain ventilator-dependent at the time of

LTAC discharge (103).
Increasingly investigators have demonstrated that mechanical ventilation is
complicated by psychiatric disturbances, including anxiety, agitation, and deli-
rium (65,66,104 106). The presence of these entities correlates with increased
duration of mechanical ventilation and is frequently present in patients with
prolonged mechanical ventilation (107,108). The potential importance of these
entities is supported by studies showing that a weaning strategy using biofeed-
back techniques shortens the duration of mechanical ventilation (109).
Rothenhausler et al. (110) treated seven patients with prolonged mechanical
ventilation and psychomotor retardation associated with markedly depressed
mood (DSM-IV criteria) with methylphenidate, a psychostimulant. Five of the
seven patients had marked/moderate improvement in mood within three to
four days and were discontinued from ventilatory support within eight to
14 days (110).
D. Process of Care Factors

Data from several studies suggest that the ability of a center to liberate patients
from mechanical ventilation is related to the caregivers skills with patients
on prolonged mechanical ventilation (32,40,41). Bach et al. (39) compared
outcomes for patients managed by either university- or community-based
physicians. Comparable mortality rates were noted, but the duration of mecha-
nical ventilation was reduced (39 days vs. 57 days, P 0.02), and there was a
trend toward higher rates of weaning success (46% vs. 30%, P 0.14), for
patients managed by the university-based physicians. Similarly, some investi-
gators report improving rates of weaning success over time as caregivers gain
increasing experience in managing patients with prolonged mechanical venti-
lation (40,41). Unit nursing and respiratory stafng is important in determining
ventilator dependence. As an example, Thorens et al. (32) studied patients
with COPD with acute respiratory failure and noted a dramatic increase in the
duration of mechanical ventilation (7 38 days) when the effective nursing
staff decreased.
Absence of a structured approach to weaning (a weaning protocol) may
increase the number of patients requiring prolonged mechanical ventilation
(10,111,112). As an example, Ely et al. (10) found that patients weaned by pro-
tocol were less likely to require .21 days of mechanical ventilation when com-
pared to patients weaned by a traditional approach. Efforts to improve bedside
communication between healthcare workers (113) and placing a weaning proto-
col on a handheld computer have also proved benecial in avoiding prolonged
mechanical ventilation (114). Protocols directed at non-ventilator aspects
of care, including reducing the amount of sedation administered (65,66) and
improving glycemic control (115), also reduced the duration of mechanical
ventilation in the acute setting.
VII. What Assessment Tools Are Available to Predict

Weaning Outcome for Patients with Prolonged
Mechanical Ventilation?
More than 50 distinct weaning predictors have been studied in an effort to foretell
the outcome of weaning trials in the acute ICU setting (116,117). The rationale
for using objective physiologic parameters is strong as prediction based on clini-
cal gestalt alone is relatively inaccurate (118,119). More importantly, precise pre-
dictors would enable physicians to postpone weaning in patients likely to fail.
Alternatively, predictors could be used to better identify patients ready to initiate
weaning trials with a goal of hastening the process of liberation, thereby reducing
the duration of mechanical ventilation. The latter goal is based on the association
between increased duration of ventilation and complications (ventilator-associ-
ated pneumonia, airway injury, barotrauma, sinusitis, thromboembolism, and
gastrointestinal bleeding), which further increase duration of stay, costs, and mor-
tality (120,121). Recent analyses and evidence-based guidelines indicate that
weaning predictors are insufciently accurate to adequately inform decision-
making in the acute ICU (117,122).
Studies in patients requiring .21 days of mechanical ventilation have
shown that respiratory mechanics used to predict weaning outcomes in acute
patients may have some value among patients requiring prolonged mechanical
ventilation (76,123). Nava et al. observed that lower maximal inspiratory
pressure and higher central drive as measured by the airway occlusion pressure
were predictive of weaning success among patients with COPD (76,123).
Although the frequency tidal volume ratio was also signicantly different in
patients with weaning success when compared with those of weaning failure,
the mean values in the latter group were well below the threshold originally
described by Yang and Tobin (69). As suggested by recent work, patients with
COPD may be unable to increase their respiratory rates signicantly because
of the mechanical constraints of expiratory-ow limitation and hyperination.
In this setting, a lower f/Vt threshold may yield more accurate predictions
(124). Indeed, Scheinhorn et al. (125) used the frequency-to-tidal volume ratio
to accelerate weaning progress through a therapist-implement protocol (125).
Weaning success was nearly twice as likely when the f/Vt was ,80 compared
with .120 (breaths/L)/min (Fig. 3) (126). Failure to appreciate respiratory
efforts that do not trigger the ventilator or result in inspiratory ow
may lead to an underestimation of the respiratory rate and therefore the f/Vt
ratio (76).
Indeed, Chao et al. (127) noted an association between patient ventilator
trigger asynchrony and weaning failure in patients requiring prolonged mechan-
ical. Nineteen of the 200 patients screened were noted to have trigger asynchrony.
Interventions such as altering trigger sensitivity, changing to ow triggering,
and increasing external PEEP were unsuccessful in eliminating the trigger
asynchrony. Only decreasing the level of pressure support was successful in
Figure 3 Percentage of patients with weaning success based on the frequency to tidal
volume ratio determined prior to the initiation of weaning efforts. Source: Adapted
from Ref. 126.
extinguishing trigger asynchrony, but this maneuver imposed a fatiguing

respiratory load and was poorly tolerated. Of the 19 patients with trigger
asynchrony, only three (16%) were successfully liberated after 70 to 108 days
compared with patients without trigger asynchrony who were successfully
weaned 56% of the time after a median 33 days. In all likelihood, the presence
of trigger asynchrony serves as a marker for severe disease of the respiratory
system (abnormal mechanics, increased PEEPi, respiratory muscle weakness,
or abnormal control of breathing).
As noted earlier, an abnormal mental status has been associated with
prolonged mechanical ventilation. Hendra et al., studying a cohort not receiving
sedative infusions and half of whom had neurologic diagnoses, found that
patients with a modied Glasgow Coma Scale score of ,8 were 6.5 times
more likely to fail weaning (128).
Because ventilator dependence may be multi-factorial, an approach that
combines a number of factors and parameters (e.g., a scoring system) may
prove superior in predicting weaning outcome for patients with prolonged
mechanical ventilation. As an example, Morganroth et al. combined a ventilator
score based upon six simply measured factors (FIO2, PEEP, static and dynamic
compliance, set minute ventilation, and triggered respiratory rate) and an
adverse factor score, calculated using 21 components, including vital signs,
quantity and quality of secretions, level of consciousness, emotional state, and
medications (129). During 11 episodes of prolonged mechanical ventilation,
the investigators demonstrated an improvement in the combined score prior to
the initiation of a successful weaning period. In contrast, two patients who
died prior to weaning showed no improvements in their scores. To date this
score has not been validated and appears cumbersome to determine.
Gluck and Corrigan (42) studied 20 patients requiring prolonged mecha-

nical ventilation to identify variables in univariate analysis that correlated with
weaning outcome and determine threshold values (42). All patients were
weaned using a standardized protocol, and patients unable to be weaned from
mechanical ventilation after 42 days were classied as failures. The frequency-
to-tidal volume ratio measured ve minutes after removal of the patient from
mechanical ventilation, dead space to tidal volume ratio, static compliance,
airway resistance, and PaCO2 measured at the end of a trial of spontaneous
breathing were integrated into a scoring system. This weaning score was then
prospectively studied in an additional 72 patients, correctly classifying weaning
outcome 84% of the time. No patient who failed to wean by 42 days on protocol
was successfully weaned in the following six months. Scheinhorn et al. tested
several scoring system models in a 170-patient validation group and identied
the A B G score (P(A2a)O2, BUN, and Gender) as most predictive of
weaning outcome, correctly classifying patients 70% of the time (6). A group
of investigators has suggested that physicians are able to accurately predict
both liberation from mechanical ventilation and survival in patients
requiring prolonged mechanical ventilation (58). To date, none of these multi-
dimensional approaches to predicting weaning success has been independently
validated.
VIII. What Is the Best Approach to Weaning Patients with

Prolonged Mechanical Ventilation?
Fundamental criteria should be satised prior to weaning a patient from pro-

longed mechanical ventilation. Specically, there should be signicant recovery
from the acute illness that precipitated respiratory failure. The patient must
have adequate gas exchange, be hemodynamically stable, and have neurological
function that generates spontaneous respiration and airway maintenance. Unlike
the acute setting (e.g., when an endotracheal tube is present), the capacity to
protect the airway is not a limiting factor in weaning from prolonged mechanical
ventilation. Airway integrity becomes crucial when the patient has been liberated
from the ventilator and tracheotomy tube decannulation is under consideration. A
search for potentially correctable conditions that can contribute to ventilator
dependence needs to be undertaken in patients who repeatedly fail to tolerate
weaning efforts.
In the acute setting, daily spontaneous breathing trials (using T-piece) and
pressure support weaning have been shown to be superior to traditional IMV
weaning strategies (9,130). Uncontrolled experiences in patients with prolonged
mechanical ventilation have noted the successful use of several approaches to
weaning. Until recently there were no randomized, controlled trials investigating
the use of different weaning modes for patients requiring prolonged mechanical
ventilation. Vitacca et al. (131) randomized 52 COPD patients (with a minimum
duration of mechanical ventilation 15 days, range: 15 39 days) to weaning by

either progressive reduction in pressure support level (PSV) or by tracheotomy
trials of increasing duration. The authors found no difference when failure was
dened as either death or failure to wean by 30 days.
Studies performed in the acute setting have demonstrate that standardization
of the weaning process through the use of protocols can shorten the duration of
weaning and the time spent on mechanical ventilation (10,111,112,132,133). Of
note, the results of this approach in pediatric patients has produced conicting
results (134,135). Similarly, protocols shown to be effective in medical ICU
patients were not efcacious in neurosurgical patients (136) or in trauma patients
(137). Interestingly, a recent study showed no difference between a weaning
protocol and a standard approach in an acute ICU where intensivists were on site
for a large proportion of the day (138). This scenario is unlikely to occur in the
post-acute care setting where nurse- and respiratory therapist-to-patient ratios are
lower and on site physician presence is likely to be less. The benecial effect of
protocols may derive more from early identication of patients capable of breathing
spontaneously rather than an algorithmic reduction in ventilatory support, an obser-
vation that may also be applicable to patients with prolonged mechanical venti-
lation. Vitacca et al. (131) screened 75 COPD patients for their randomized
controlled trial by allowing an initial trial of spontaneous breathing (131).
Twenty-three (31%) patients did not need a slow process of liberation because
they tolerated this initial trial of spontaneous breathing for at least 48 hours.
Although several groups have investigated the application of weaning
protocols to patients requiring prolonged mechanical ventilation, randomized
controlled trials have not been reported. Gluck and Corrigan used a weaning pro-
tocol employing sequential reduction in both continuous positive airway pressure
and PSV as part of an evaluation of predictors of weaning (42). In an uncontrolled
study, Scheinhorn et al. (125) used a respiratory therapist-driven protocol and
noted a reduction in the duration of mechanical ventilation (median: 29 days
vs. 18 days, P , 0.05) and in the hospital length of stay (median: 53 days vs.
43 days, P , 0.05) after institution of the protocol when compared to historic
controls (125). This protocol is notable because the frequency tidal volume
ratio was used to identify patients in whom the weaning process could be accel-
erated. The protocol employed a 19-step gradual reduction of the synchronized
intermittent mandatory ventilation rate and PSV level. For patients with a
frequency to tidal volume ratio 80 breaths/min/L, the rst nine steps of
weaning were bypassed, with patients advancing directly to a one-hour trial of
spontaneous breathing. Vitacca et al. (131) compared 52 patients weaned by
protocol (either tracheotomy collar or pressure support) to historic controls and
found a higher 30-day weaning success rate and shorter duration of mechanical
ventilation in the former (131).
The frequent presence of respiratory muscle weakness in patients with
prolonged mechanical ventilation has led some to investigate strategies for
improving respiratory muscle strength and endurance. Aldrich et al. (95) used
periods of spontaneous breathing through an adjustable nonlinear resistor with

gradually increasing duration and resistance in patients on prolonged mechanical
ventilation. Respiratory muscle training resulted in improved mean negative
inspiratory force (37 46 cmH2O) and vital capacity (561 901 mL) in 27
patients. Twelve of the patients (44%) were successfully liberated from mechan-
ical ventilation over 10 to 46 days, with another ve weaned to nocturnal
ventilation (19%). A similar technique has been reported in the rehabilitation
of three high quadriplegic patients who were ventilator-dependent (139).
Martin et al. (140) studied 10 patients who had been ventilated for a mean of
34 days and had failed weaning for at least one week (140). Inspiratory muscle
training consisting of four sets of six breaths through a threshold resistor were
combined with tracheotomy collar trials of increasing duration. Using this
approach-nine of 10 patients were liberated from mechanical ventilation after a
total of 44 days. These studies did not utilize a control population, so it is difcult
to attribute the patients improvement to the muscle training where the same
number may have been liberated from mechanical ventilation using a more
conventional technique.
Non-invasive mechanical ventilation has recently been implemented in the
acute setting as an adjunct to weaning patients from invasive mechanical venti-
lation (141 143). Several authors have reported their experience with its use in
patients requiring prolonged mechanical ventilation because of various etiolo-
gies. Udwadia et al. (144) studied 22 patients with weaning failure who had
been ventilated for a median of 31 days (range: 2 219 days) (144). Eighteen
patients were successfully liberated from invasive mechanical ventilation and
discharged home a median of 11 days (range: 8 13 days) after starting non-
invasive ventilation. Restrick et al. (145) used non-invasive ventilation to
successfully wean 13 of 14 patients with prolonged mechanical ventilation.
Well-designed randomized controlled trials are needed before this approach
can be fully recommended.
Approximately 20% of all patients transferred to an LTAC will not be
successfully liberated from mechanical ventilation despite surviving the hospital-
ization (28). In the absence of ongoing but remediable disease, patients
with prolonged mechanical ventilation should not be classied as irreversibly
ventilator-dependent before 90 days of failed weaning attempts (122).
Efforts shift to determining the need for the tracheotomy tube for patients
who are liberated from the ventilator. Assessment of the upper airway becomes
crucial because the tracheotomy stoma can close within 48 to 72 hours, making
re-establishment of an airway difcult should post-decannulation respiratory
distress occur. The initial maneuver is to determine patient response to deation
of the cuff (e.g., does the patient aspirate?). If this is tolerated, the response to
tube occlusion (capping) is performed after placement of a fenestrated tracheot-
omy. Once tube occlusion is tolerated for 24 to 48 hours, decannulation may be
considered. Alternatively, a strategy of sequential downsizing of tracheotomy
tubes may be desirable (146 148). Ceriana et al. (149) demonstrated that a
decisional ow chart based on clinical and physiologic principles led to decannu-

lation in 56 of 72 tracheotomized survivors who had been weaned from the ven-
tilator after respiratory failure from a variety of causes. Only 3% of decannulated
patients required reintubation within the subsequent three months. The main
reasons precluding successful decannulation were either inability to manage
secretions or severe glottic stenosis. Fiberoptic bronchoscopy and laryngoscopy
should be performed to search for sources of upper airway obstruction (vocal
cord injury/dysfunction, granulation tissue, airway strictures) should respiratory
distress occur with tracheotomy tube occlusion.
IX. Conclusion
The need for prolonged mechanical ventilation is multi-factorial in origin, invol-

ving abnormalities of respiratory mechanics, respiratory muscle dysfunction, the
presence of systemic illness, and complications associated with critical illness. At
least in the short-term the majority of patients survive, with slightly more than
half being successfully liberated from mechanical ventilation. Although random-
ized trials are not yet available, strong observational investigations indicate that
an organized approach to liberation (e.g., use of a protocol) likely improves
weaning success and shortens the duration of ventilation. The literature further
demonstrates that satisfactory outcomes can be achieved at long-term acute
care facilities with extensive experience in caring for patients with prolonged
mechanical ventilation.
References
1. Nevins ML, Epstein SK. Weaning from prolonged mechanical ventilation. Clin
Chest Med 2001; 22:13 33.
2. Arabi Y, Venkatesh S, Haddad S, Al Shimemeri A, Al Malik S. A prospective study
of prolonged stay in the intensive care unit: predictors and impact on resource
utilization. Int J Qual Health Care 2002; 14:403 410.
3. Gillespie DJ, Marsh HM, Divertie MB, Meadows JA III. Clinical outcome of res-
piratory failure in patients requiring prolonged (greater than 24 hours) mechanical
ventilation. Chest 1986; 90:364 369.
4. Gracey DR, Viggiano RW, Naessens JM, Hubmayr RD, Silverstein MD, Koenig
GE. Outcomes of patients admitted to a chronic ventilator-dependent unit in an
acute-care hospital. Mayo Clin Proc 1992; 67:131 136.
5. Spicher JE, White DP. Outcome and function following prolonged mechanical
ventilation. Arch Intern Med 1987; 147:421 425.
6. Scheinhorn DJ, Hassenpug M, Artinian BM, LaBree L, Catlin JL. Predictors of
weaning after 6 weeks of mechanical ventilation. Chest 1995; 107:500 505.
7. Scheinhorn DJ, Chao DC, Stearn-Hassenpug M. Liberation from prolonged
mechanical ventilation. Crit Care Clin 2002; 18:569 595.
8. Esteban A, Anzueto A, Frutos F, Alia I, Brochard L, Stewart TE, Benito S, Epstein

SK, Apezteguia C, Nightingale P, Arroliga AC, Tobin MJ. Characteristics and out-
comes in adult patients receiving mechanical ventilation: a 28-day international
study. JAMA 2002; 287:345 355.
9. Brochard L, Rauss A, Benito S, Conti G, Mancebo J, Rekik N, Gasparetto A,
Lemaire F. Comparison of three methods of gradual withdrawal from ventilatory
support during weaning from mechanical ventilation. Am J Respir Crit Care Med
1994; 150:896 903.
10. Ely EW, Baker AM, Dunagan DP, Burke HL, Smith AC, Kelly PT, Johnson MM,
Browder RW, Bowton DL, Haponik EF. Effect on the duration of mechanical ven-
tilation of identifying patients capable of breathing spontaneously. N Engl J Med
1996; 335:1864 1869.
11. Pierson DJ. Current limitations of mechanical ventilation: what improvements
should the clinician expect? Respir Care 1995; 40:933 941.
12. Stauffer JL, Fayter NA, Graves B, Cromb M, Lynch JC, Goebel P. Survival follow-
ing mechanical ventilation for acute respiratory failure in adult men. Chest 1993;
104:1222 1229.
13. Epstein SK, Vuong V. Lack of Inuence of gender on outcome of mechanically
ventilated medical intensive care unit patients. Chest 1999; 116(3): 732 739.
14. Make B, Dayno S, Gertman P. Prevalence of chronic ventilator dependency.
Am Rev Respir Dis 1986; 133:A167.
15. Harris J, Haughton J, Celli B. A survey of ventilator dependent patients in
Massachusetts. Am J Respir Crit Care Med 1997; 155:A411.
16. Mulligan S. AARC and Gallup estimate numbers and costs of caring for chronic
ventilator patients. AARC Times 1991; 15:30 36.
17. HCFA. Bureau of data management and strategy. 100% MEDPAR inpatient
hospital scal year 1998, 6/99 update. 1999.
18. Dewar DM, Kurek CJ, Lambrinos J, Cohen IL, Zhong Y. Patterns in costs and out-
comes for patients with prolonged mechanical ventilation undergoing tracheost-
omy: an analysis of discharges under diagnosis-related group 483 in New York
State from 1992 to 1996. Crit Care Med 1999; 27:2640 2647.
19. Carson SS, Bach PB. Predicting mortality in patients suffering from prolonged critical
illness: an assessment of four severity-of-illness measures. Chest 2001; 120:928933.
20. Dematte DAmico JE, Donnelly HK, Mutlu GM, Feinglass J, Jovanovic BD,
Ndukwu IM. Risk assessment for inpatient survival in the long-term acute care
setting after prolonged critical illness. Chest 2003; 124:1039 1045.
21. Carson SS, Bach PB, Brzozowski L, Leff A. Outcomes after long-term acute care.
An analysis of 133 mechanically ventilated patients. Am J Respir Crit Care Med
1999; 159:1568 1573.
22. Carson SS, Bach PB. The epidemiology and costs of chronic critical illness. Crit
Care Clin 2002; 18:461 476.
23. Teno JM, Fisher E, Hamel MB, Wu AW, Murphy DJ, Wenger NS, Lynn J, Harrell
FE, Jr. Decision-making and outcomes of prolonged ICU stays in seriously ill
patients. J Am Geriatr Soc 2000; 48:S70 74.
24. Heyland DK, Konopad E, Noseworthy TW, Johnston R, Gafni A. Is it worthwhile
to continue treating patients with a prolonged stay (.14 days) in the ICU? An
economic evaluation. Chest 1998; 114:192 198.
25. Chatila W, Kreimer DT, Criner GJ. Quality of life in survivors of prolonged
mechanical ventilatory support. Crit Care Med 2001; 29:737 742.
26. Nevins ML, Epstein SK. Predictors of outcome for patients with COPD requiring
invasive mechanical ventilation. Chest 2001; 119:1840 1849.
27. Zilberberg MD, Epstein SK. Acute lung injury in the medical ICU: comorbid con-
ditions, age, etiology, and hospital outcome. Am J Respir Crit Care Med 1998;
157:1159 1164.
28. Scheinhorn D, Chao D, Stearn-Hassenpug M, Doig G, Epstein S, Knight E,
Petrak R, Pitt E, Votto J. Ventilator-dependent survivors of catastrophic illness: a
multicenter outcomes stuy. Am J Respir Crit Care Med 2003; 167:A458.
29. Ely EW, Evans GW, Haponik EF. Mechanical ventilation in a cohort of elderly
patients admitted to an intensive care unit. Ann Intern Med 1999; 131:96 104.
30. Esteban A, Anzueto A, Frutos-Vivar F, Alia I, Ely EW, Brochard L, Stewart TE,
Apezteguia C, Tobin MJ, Nightingale P, Matamis D, Pimentel J, Abroug F.
Outcome of older patients receiving mechanical ventilation. Intensive Care Med
2004; 30:639 646.
31. Elpern EH, Silver MR, Rosen RL, Bone RC. The noninvasive respiratory care unit.
Patterns of use and nancial implications. Chest 1991; 99:205 208.
32. Thorens JB, Kaelin RM, Jolliet P, Chevrolet JC. Inuence of the quality of nursing
on the duration of weaning from mechanical ventilation in patients with chronic
obstructive pulmonary disease. Crit Care Med 1995; 23:1807 1815.
33. Scheinhorn DJ, Chao DC, Stearn-Hassenpug M, LaBree LD, Heltsley DJ. Post-
ICU mechanical ventilation: treatment of 1,123 patients at a regional weaning
center. Chest 1997; 111:1654 1659.
34. Scheinhorn DJ, Artinian BM, Catlin JL. Weaning from prolonged mechanical
ventilation. The experience at a regional weaning center. Chest 1994;
105:534 539.
35. Scheinhorn DJ, Chao DC, Stearn-Hassenpug M. Post-ICU mechanical ventilation-
treatment of 1575 patients over 11 years at a regional weaning center. Am J Respir
Crit Care Med 2000; 161:A793.
36. Sirio CA, Shepardson LB, Rotondi AJ, Cooper GS, Angus DC, Harper DL,
Rosenthal GE. Community-wide assessment of intensive care outcomes using a
physiologically based prognostic measure: implications for critical care delivery
from Cleveland Health Quality Choice. Chest 1999; 115:793 801.
37. Gracey DR, Naessens JM, Viggiano RW, Koenig GE, Silverstein MD, Hubmayr
RD. Outcome of patients cared for in a ventilator-dependent unit in a general
hospital. Chest 1995; 107:494 499.
38. Gracey DR, Hardy DC, Naessens JM, Silverstein MD, Hubmayr RD. The Mayo
Ventilator-Dependent Rehabilitation Unit: a 5-year experience. Mayo Clin Proc
1997; 72:13 19.
39. Bach PB, Carson SS, Leff A. Outcomes and resource utilization for patients with
prolonged critical illness managed by university-based or community-based
subspecialists. Am J Respir Crit Care Med 1998; 158:1410 1415.
40. Bagley PH, Cooney E. A community-based regional ventilator weaning unit:
development and outcomes. Chest 1997; 111:1024 1029.
41. Indihar FJ. A 10-year report of patients in a prolonged respiratory care unit. Minn
Med 1991; 74:23 27.
42. Gluck EH. Predicting eventual success or failure to wean in patients receiving
long-term mechanical ventilation. Chest 1996; 110:1018 1024.
43. Epstein SK, Ciubotaru RL, Wong JB. Effect of failed extubation on the outcome of
mechanical ventilation. Chest 1997; 112:186 192.
44. Epstein SK, Ciubotaru RL. Independent effects of etiology of failure and time to
reintubation on outcome for patients failing extubation. Am J Respir Crit Care
Med 1998; 158:489 493.
45. Daley BJ, Garcia-Perez F, Ross SE. Reintubation as an outcome predictor in trauma
patients. Chest 1996; 110:1577 1580.
46. Chan M, Mehta R, Vasishtha N. Ventilator care in a nursing home. Am J Respir Crit
Care Med 1999; 159:A374.
47. Nasraway SA, Button GJ, Rand WM, Hudson-Jinks T, Gustafson M. Survivors of
catastrophic illness: outcome after direct transfer from intensive care to extended
care facilities. Crit Care Med 2000; 28:19 25.
48. Tafreshi M, Schneider R, Rosen M. Outcome of patients who require long-
term mechanical ventilation and hemodialysis (abstract). Chest 1995;
108(suppl):134S.
49. Chao DC, Scheinhorn DJ, Stearn-Hassenpug M. Impact of renal dysfunction on
weaning from prolonged mechanical ventilation. Crit Care (London) 1997;
1:101 104.
50. Hurford WE, Lynch KE, Strauss HW, Lowenstein E, Zapol WM. Myocardial
perfusion as assessed by thallium-201 scintigraphy during the discontinuation of
mechanical ventilation in ventilator-dependent patients. Anesthesiology 1991;
74:1007 1016.
51. Hurford WE, Favorito F. Association of myocardial ischemia with failure to wean
from mechanical ventilation. Crit Care Med 1995; 23:1475 1480.
52. Chatila W, Ani S, Guaglianone D, Jacob B, Amoateng-Adjepong Y, Manthous CA.
Cardiac ischemia during weaning from mechanical ventilation. Chest 1996;
109:1577 1583.
53. Abalos A, Leibowitz AB, Distefano D, Halpern N, Iberti TJ. Myocardial ischemia
during the weaning period. Am J Crit Care 1992; 1:32 36.
54. Jubran A, Mathru M, Dries D, Tobin MJ. Continuous recordings of mixed venous
oxygen saturation during weaning from mechanical ventilation and the ramica-
tions thereof. Am J Respir Crit Care Med 1998; 158:1763 1769.
55. Lemaire F, Teboul JL, Cinotti L, Giotto G, Abrouk F, Steg G, Macquin-Mavier I,
Zapol WM. Acute left ventricular dysfunction during unsuccessful weaning from
mechanical ventilation. Anesthesiology 1988; 69:171 179.
56. Scheinhorn D. Increase in serum albumin and decrease in body weight correlate
with weaning from prolonged mechanical ventilation (abstract). Am Rev Respir
Dis 1992; 145:A522.
57. Kreimer D, Berger B, Rahim P. Nutritional status of chronic ventilator dependent
patients on admission to a ventilator rehabilitation unit (abstract). Am J Respir
Crit Care Med 1998; 157:A307.
58. Dasgupta A, Rice R, Mascha E, Litaker D, Stoller JK. Four-year experience with a
unit for long-term ventilation (respiratory special care unit) at the Cleveland Clinic
Foundation. Chest 1999; 116:447 455.
59. Latriano B, McCauley P, Astiz ME, Greenbaum D, Rackow EC. Non-ICU care
of hemodynamically stable mechanically ventilated patients. Chest 1996; 109:
1591 1596.
60. Jubran A, Tobin MJ. Pathophysiologic basis of acute respiratory distress in patients
who fail a trial of weaning from mechanical ventilation. Am J Respir Crit Care Med
1997; 155:906 915.
61. Pourriat JL, Lamberto C, Hoang PH, Fournier JL, Vasseur B. Diaphragmatic
fatigue and breathing pattern during weaning from mechanical ventilation in
COPD patients. Chest 1986; 90:703 707.
62. Sassoon CS, Mahutte CK. Airway occlusion pressure and breathing pattern as pre-
dictors of weaning outcome. Am Rev Respir Dis 1993; 148:860 866.
63. Tobin MJ, Perez W, Guenther SM, Semmes BJ, Mador MJ, Allen SJ, Lodato
RF, Dantzker DR. The pattern of breathing during successful and unsuccessful
trials of weaning from mechanical ventilation. Am Rev Respir Dis 1986;
134:1111 1118.
64. Coplin WM, Pierson DJ, Cooley KD, Newell DW, Rubenfeld GD. Implications
of extubation delay in brain-injured patients meeting standard weaning criteria.
65. Kress JP, Pohlman AS, OConnor MF, Hall JB. Daily interruption of sedative
infusions in critically ill patients undergoing mechanical ventilation. N Engl J
Med 2000; 342:1471 1477.
66. Brook AD, Ahrens TS, Schaiff R, Prentice D, Sherman G, Shannon W, Kollef MH.
Effect of a nursing-implemented sedation protocol on the duration of mechanical
ventilation. Crit Care Med 1999; 27:2609 2615.
67. Capdevila X, Perrigault PF, Ramonatxo M, Roustan JP, Peray P, dAthis F,
Prefaut C. Changes in breathing pattern and respiratory muscle performance par-
ameters during difcult weaning. Crit Care Med 1998; 26:79 87.
68. Vassilakopoulos T, Zakynthinos S, Roussos C. The tension time index and the
frequency/tidal volume ratio are the major pathophysiologic determinants of
weaning failure and success. Am J Respir Crit Care Med 1998; 158:378 385.
69. Yang KL, Tobin MJ. A prospective study of indexes predicting the outcome
of trials of weaning from mechanical ventilation. N Engl J Med 1991; 324:
1445 1450.
70. Bellemare F, Grassino A. Effect of pressure and timing of contraction on human
diaphragm fatigue. J Appl Physiol 1982; 53:1190 1195.
71. Cohen CA, Zagelbaum G, Gross D, Roussos C, Macklem PT. Clinical manifes-
tations of inspiratory muscle fatigue. Am J Med 1982; 73:308 316.
72. Laghi F, Cattapan SE, Jubran A, Parthasarathy S, Warshawsky P, Choi YS, Tobin
MJ. Is weaning failure caused by low-frequency fatigue of the diaphragm? Am J
73. Laghi F, DAlfonso N, Tobin MJ. Pattern of recovery from diaphragmatic fatigue
over 24 hours. J Appl Physiol 1995; 79:539 546.
74. Appendini L, Purro A, Patessio A, Zanaboni S, Carone M, Spada E, Donner CF,
Rossi A. Partitioning of inspiratory muscle workload and pressure assistance
in ventilator-dependent COPD patients. Am J Respir Crit Care Med 1996;
154:1301 1309.
75. Reinoso MA, Gracey DR, Hubmayr RD. Interrupter mechanics of patients
admitted to a chronic ventilator dependency unit. Am Rev Respir Dis 1993;
148:127 131.
76. Purro A, Appendini L, De Gaetano A, Gudjonsdottir M, Donner CF, Rossi A.
Physiologic determinants of ventilator dependence in long-term mechanically ven-
tilated patients. Am J Respir Crit Care Med 2000; 161:1115 1123.
77. Roussos CS, Macklem PT. Diaphragmatic fatigue in man. J Appl Physiol 1977;
43:189 197.
78. Capdevila X, Lopez S, Bernard N, Rabischong E, Ramonatxo M, Martinazzo G,
Prefaut C. Effects of controlled mechanical ventilation on respiratory muscle
contractile properties in rabbits. Intensive Care Med 2003; 29:103 110.
79. Powers SK, Shanely RA, Coombes JS, Koesterer TJ, McKenzie M, Van Gammeren
D, Cicale M, Dodd SL. Mechanical ventilation results in progressive contractile
dysfunction in the diaphragm. J Appl Physiol 2002; 92:1851 1858.
80. Radell PJ, Remahl S, Nichols DG, Eriksson LI. Effects of prolonged mechanical
ventilation and inactivity on piglet diaphragm function. Intensive Care Med
2002; 28:358 364.
81. Sassoon CS, Caiozzo VJ, Manka A, Sieck GC. Altered diaphragm contractile
properties with controlled mechanical ventilation. J Appl Physiol 2002;
92:2585 2595.
82. Yang L, Luo J, Bourdon J, Lin MC, Gottfried SB, Petrof BJ. Controlled mechanical
ventilation leads to remodeling of the rat diaphragm. Am J Respir Crit Care Med
2002; 166:1135 1140.
83. Anzueto A, Peters JI, Tobin MJ, de los Santos R, Seidenfeld JJ, Moore G, Cox WJ,
Coalson JJ. Effects of prolonged controlled mechanical ventilation on diaphrag-
matic function in healthy adult baboons. Crit Care Med 1997; 25:1187 1190.
84. Hanson P, Dive A, Brucher JM, Bisteau M, Dangoisse M, Deltombe T. Acute
corticosteroid myopathy in intensive care patients. Muscle Nerve 1997;
20:1371 1380.
85. Hirano M, Ott BR, Raps EC, Minetti C, Lennihan L, Libbey NP, Bonilla E,
Hays AP. Acute quadriplegic myopathy: a complication of treatment with steroids,
nondepolarizing blocking agents, or both. Neurology 1992; 42:2082 2087.
86. Grifn D, Fairman N, Coursin D, Rawsthorne L, Grossman JE. Acute myopathy
during treatment of status asthmaticus with corticosteroids and steroidal muscle
relaxants. Chest 1992; 102:510 514.
87. Segredo V, Caldwell JE, Matthay MA, Sharma ML, Gruenke LD, Miller RD.
Persistent paralysis in critically ill patients after long-term administration of
vecuronium. N Engl J Med 1992; 327:524 528.
88. Wijdicks EF, Litchy WJ, Harrison BA, Gracey DR. The clinical spectrum of critical
illness polyneuropathy. Mayo Clin Proc 1994; 69:955 959.
89. Bolton CF, GrandMaison F, Parkes A, Shkrum M. Needle electromyography of the
diaphragm. Muscle Nerve 1992; 15:678 681.
90. Bolton CF, Gilbert JJ, Hahn AF, Sibbald WJ. Polyneuropathy in critically ill
patients. J Neurol Neurosurg Psychiatry 1984; 47:1223 1231.
91. Leijten FS, Harinck-de Weerd JE, Poortvliet DC, de Weerd AW. The role of
polyneuropathy in motor convalescence after prolonged mechanical ventilation.
JAMA 1995; 274:1221 1225.
92. Spitzer AR, Giancarlo T, Maher L, Awerbuch G, Bowles A. Neuromuscular causes

of prolonged ventilator dependency. Muscle Nerve 1992; 15:682 686.
93. Coakley JH, Nagendran K, Yarwood GD, Honavar M, Hinds CJ. Patterns of
neurophysiological abnormality in prolonged critical illness. Intensive Care Med
1998; 24:801 807.
94. De Jonghe B, Bastuji-Garin S, Sharshar T, Outin H, Brochard L. Does ICU-
acquired paresis lengthen weaning from mechanical ventilation? Intensive Care
Med 2004; 30:1117 1121.
95. Aldrich TK, Karpel JP, Uhrlass RM, Sparapani MA, Eramo D, Ferranti R. Weaning
from mechanical ventilation: adjunctive use of inspiratory muscle resistive training.
Crit Care Med 1989; 17:143 147.
96. Diehl JL, El Atrous S, Touchard D, Lemaire F, Brochard L. Changes in the work of
breathing induced by tracheotomy in ventilator-dependent patients. Am J Respir
Crit Care Med 1999; 159:383 388.
97. Law JH, Barnhart K, Rowlett W, de la Rocha O, Lowenberg S. Increased frequency
of obstructive airway abnormalities with long-term tracheostomy. Chest 1993;
104:136 138.
98. Rumbak MJ, Graves AE, Scott MP, Sporn GK, Walsh FW, Anderson WM,
Goldman AL. Tracheostomy tube occlusion protocol predicts signicant tracheal
obstruction to air ow in patients requiring prolonged mechanical ventilation.
Crit Care Med 1997; 25:413 417.
99. Rumbak MJ, Walsh FW, Anderson WM, Rolfe MW, Solomon DA. Signicant
tracheal obstruction causing failure to wean in patients requiring prolonged mech-
anical ventilation: a forgotten complication of long-term mechanical ventilation.
Chest 1999; 115:1092 1095.
100. Elpern EH, Scott MG, Petro L, Ries MH. Pulmonary aspiration in mechanically
ventilated patients with tracheostomies. Chest 1994; 105:563 566.
101. Tolep K, Getch CL, Criner GJ. Swallowing dysfunction in patients receiving
prolonged mechanical ventilation. Chest 1996; 109:167 172.
102. Schonhofer B, Barchfeld T, Haidl P, Kohler D. Scintigraphy for evaluating early
aspiration after oral feeding in patients receiving prolonged ventilation via tra-
cheostomy. Intensive Care Med 1999; 25:311 314.
103. Scheinhorn D, Chao D, Stearn-Hassenpug M, Doig G, Epstein S, Knight E, Petrak
R, Pitt E, Votto J. Infectious Complications in Weaning from Prolonged
Mechanical Ventilation at Long Term Hospitals: Preliminary Report from a
Multicenter Study (abstract). Am J Respir Crit Care Med 2004; 169:A44.
104. Ely EW, Gautam S, Margolin R, Francis J, May L, Speroff T, Truman B, Dittus R,
Bernard R, Inouye SK. The impact of delirium in the intensive care unit on hospital
length of stay. Intensive Care Med 2001; 27:1892 1900.
105. Ely EW, Shintani A, Truman B, Speroff T, Gordon SM, Harrell FE Jr, Inouye SK,
Bernard GR, Dittus RS. Delirium as a predictor of mortality in mechanically ven-
tilated patients in the intensive care unit. JAMA 2004; 291:1753 1762.
106. Kollef MH, Levy NT, Ahrens TS, Schaiff R, Prentice D, Sherman G. The use of
continuous i.v. sedation is associated with prolongation of mechanical ventilation.
Chest 1998; 114:541 548.
107. Coffman K, Scheinhorn D, Chao D. Diagnostic psychiatric consultation at a
long-term acute care hospital that specializes in weaning from prolonged
mechanical ventilation and pulmonary rehabilitation (abstract). Chest 2000; 118

(suppl):164S.
108. Repetz N, Ciccolella D, Criner G. Long-term outcome of patients with delirum on
admission to a multi-disciplinary ventilator rehabilitation unit (abstract). Am J
Respir Crit Care Med 2001; 163:A889.
109. Holliday JE, Hyers TM. The reduction of weaning time from mechanical venti-
lation using tidal volume and relaxation biofeedback. Am Rev Respir Dis 1990;
141:1214 1220.
110. Rothenhausler HB, Ehrentraut S, von Degenfeld G, Weis M, Tichy M, Kilger E,
Stoll C, Schelling G, Kapfhammer HP. Treatment of depression with methylpheni-
date in patients difcult to wean from mechanical ventilation in the intensive care
unit. J Clin Psychiatry 2000; 61:750 755.
111. Kollef MH, Shapiro SD, Silver P, St John RE, Prentice D, Sauer S, Ahrens TS,
Shannon W, Baker-Clinkscale D. A randomized, controlled trial of protocol-
directed versus physician-directed weaning from mechanical ventilation. Crit
Care Med 1997; 25:567 574.
112. Marelich GP, Murin S, Battistella F, Inciardi J, Vierra T, Roby M. Protocol weaning
of mechanical ventilation in medical and surgical patients by respiratory care prac-
titioners and nurses: effect on weaning time and incidence of ventilator-associated
pneumonia. Chest 2000; 118:459 467.
113. Henneman E, Dracup K, Ganz T, Molayeme O, Cooper C. Effect of a collaborative
weaning plan on patient outcome in the critical care setting. Crit Care Med 2001;
29:297 303.
114. Iregui M, Ward S, Clinikscale D, Clayton D, Kollef MH. Use of a handheld com-
puter by respiratory care practitioners to improve the efciency of weaning patients
from mechanical ventilation. Crit Care Med 2002; 30:2038 2043.
115. van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M,
Vlasselaers D, Ferdinande P, Lauwers P, Bouillon R. Intensive insulin therapy in
the critically ill patients. N Engl J Med 2001; 345:1359 1367.
116. Epstein SK. Weaning parameters. Respir Care Clin N Am 2000; 6:253 301,
v vi.
117. Meade M, Guyatt G, Cook D, Grifth L, Sinuff T, Kergl C, Mancebo J, Esteban A,
Epstein S. Predicting success in weaning from mechanical ventilation. Chest 2001;
120:400S 424S.
118. Afessa B, Hogans L, Murphy R. Predicting 3-day and 7-day outcomes of weaning
from mechanical ventilation. Chest 1999; 116:456 461.
119. Stroetz RW, Hubmayr RD. Tidal volume maintenance during weaning with
pressure support. Am J Respir Crit Care Med 1995; 152:1034 1040.
120. Heyland DK, Cook DJ, Grifth L, Keenan SP, Brun-Buisson C. The attributable
morbidity and mortality of ventilator-associated pneumonia in the critically ill
patient. The Canadian Critical Trials Group. Am J Respir Crit Care Med 1999;
159:1249 1256.
121. Keith RL, Pierson DJ. Complications of mechanical ventilation. A bedside
approach. Clin Chest Med 1996; 17:439 451.
122. MacIntyre NR, Cook DJ, Ely EW Jr, Epstein SK, Fink JB, Heffner JE, Hess D, Hub-
mayer RD, Scheinhorn DJ. Evidence-based guidelines for weaning and discontinu-
ing ventilatory support: a collective task force facilitated by the American College
of Chest Physicians, the American Association for Respiratory Care, and the
American College of Critical Care Medicine. Chest 2001; 120:375S 395S.
123. Nava S, Rubini F, Zanotti E, Ambrosino N, Bruschi C, Vitacca M, Fracchia C,
Rampulla C. Survival and prediction of successful ventilator weaning in COPD
patients requiring mechanical ventilation for more than 21 days. Eur Respir J
1994; 7:1645 1652.
124. Alvisi R, Volta CA, Righini ER, Capuzzo M, Ragazzi R, Verri M, Candini G,
Gritti G, Milic-Emili J. Predictors of weaning outcome in chronic obstructive
pulmonary disease patients. Eur Respir J 2000; 15:656 662.
125. Scheinhorn DJ, Chao DC, Stearn-Hassenpug M, Wallace WA. Outcomes in post-
ICU mechanical ventilation: a therapist-implemented weaning protocol. Chest
2001; 119:236 242.
126. Scheinhorn DJ, Chao DC, Stearn-Hassenpug M. Utility of RSBI in a therapist-
implemented protocol for weaning from prolonged mechanical ventilation in the
post-ICU setting. Am J Respir Crit Care Med 2000; 161:A791.
127. Chao DC, Scheinhorn DJ, Stearn-Hassenpug M. Patient ventilator trigger
asynchrony in prolonged mechanical ventilation. Chest 1997; 112:1592 1599.
128. Hendra KP, Bonis PA, Joyce-Brady M. Development and prospective validation of
a model for predicting weaning in chronic ventilator dependent patients. BMC
Pulm Med 2003; 3:3.
129. Morganroth ML, Morganroth JL, Nett LM, Petty TL. Criteria for weaning from
prolonged mechanical ventilation. Arch Intern Med 1984; 144:1012 1016.
130. Esteban A, Frutos F, Tobin MJ, et al. A comparison of four methods of weaning
patients from mechanical ventilation. Spanish Lung Failure Collaborative Group.
N Engl J Med 1995; 332:345 350.
131. Vitacca M, Vianello A, Colombo D, Clini E, Porta R, Bianchi L, Arcaro G, Vitale
G, Guffanti E, Lo Coco A, Ambrosino N. Comparison of two methods for weaning
patients with chronic obstructive pulmonary disease requiring mechanical venti-
lation for more than 15 days. Am J Respir Crit Care Med 2001; 164:225 230.
132. Cohen IL, Bari N, Strosberg MA, Weinberg PF, Wacksman RM, Millstein BH,
Fein IA. Reduction of duration and cost of mechanical ventilation in an intensive
care unit by use of a ventilatory management team. Crit Care Med 1991; 19:
1278 1284.
133. Saura P, Blanch L, Mestre J, Valles J, Artigas A, Fernandez R. Clinical conse-
quences of the implementation of a weaning protocol. Intensive Care Med 1996;
22:1052 1056.
134. Randolph AG, Wypij D, Venkataraman ST, Hanson JH, Gedeit RG, Meert KL,
Luckett PM, Forbes P, Lilley M, Thompson J, Cheifetz IM, Hibberd P, Wetzel
R, Cox PN, Arnold JH. Effect of mechanical ventilator weaning protocols on res-
piratory outcomes in infants and children: a randomized controlled trial. JAMA
2002; 288:2561 2568.
135. Schultz TR, Lin RJ, Watzman HM, Durning SM, Hales R, Woodson A, Francis B,
Tyler L, Napoli L, Godinez RI. Weaning children from mechanical ventilation: a
prospective randomized trial of protocol-directed versus physician-directed
weaning. Respir Care 2001; 46:772 782.
136. Namen AM, Ely EW, Tatter SB, Case LD, Lucia MA, Smith A, Landry S, Wilson
JA, Glazier SS, Branch CL, Kelly DL, Bowton DL, Haponik EF. Predictors of
successful extubation in neurosurgical patients. Am J Respir Crit Care Med 2001;

163:658 664.
137. Duane TM, Riblet JL, Golay D, Cole FJ, Jr., Weireter LJ, Jr., Britt LD. Protocol-
driven ventilator management in a trauma intensive care unit population. Arch
Surg 2002; 137:1223 1227.
138. Krishnan JA, Moore D, Robeson C, Rand CS, Fessler HE. A prospective, controlled
trial of a protocol-based strategy to discontinue mechanical ventilation. Am J
139. Lerman RM, Weiss MS. Progressive resistive exercise in weaning high quadriple-
gics from the ventilator. Paraplegia 1987; 25:130 135.
140. Martin AD, Davenport PD, Franceschi AC, Harman E. Use of inspiratory muscle
strength training to facilitate ventilator weaning: a series of 10 consecutive patients.
Chest 2002; 122:192 196.
141. Ferrer M, Esquinas A, Arancibia F, Bauer TT, Gonzalez G, Carrillo A,
Rodriguez-Roisin R, Torres A. Noninvasive ventilation during persistent
weaning failure: a randomized controlled trial. Am J Respir Crit Care Med 2003;
168:70 76.
142. Girault C, Daudenthun I, Chevron V, Tamion F, Leroy J, Bonmarchand G.
Noninvasive ventilation as a systematic extubation and weaning technique in
acute-on-chronic respiratory failure: a prospective, randomized controlled study.
143. Nava S, Ambrosino N, Clini E, Prato M, Orlando G, Vitacca M, Brigada P,
Fracchia C, Rubini F. Noninvasive mechanical ventilation in the weaning of
patients with respiratory failure due to chronic obstructive pulmonary disease. A
randomized, controlled trial. Ann Intern Med 1998; 128:721 728.
144. Udwadia ZF, Santis GK, Steven MH, Simonds AK. Nasal ventilation to facilitate
weaning in patients with chronic respiratory insufciency. Thorax 1992; 47:
715 718.
145. Restrick LJ, Scott AD, Ward EM, Feneck RO, Cornwell WE, Wedzicha JA. Nasal
intermittent positive-pressure ventilation in weaning intubated patients with
chronic respiratory disease from assisted intermittent, positive-pressure ventilation.
Respir Med 1993; 87:199 204.
146. Heffner JE, Hess D. Tracheostomy management in the chronically ventilated
patient. Clin Chest Med 2001; 22:55 69.
147. Heffner JE. The role of tracheotomy in weaning. Chest 2001; 120:477S 481S.
148. Heffner JE. Tracheotomy application and timing. Clin Chest Med 2003;
24:389 398.
149. Ceriana P, Carlucci A, Navalesi P, Rampulla C, Delmastro M, Piaggi G, De Mattia E,
Nava S. Weaning from tracheotomy in long-term mechanically ventilated patients:
feasibility of a decisional owchart and clinical outcome. Intensive Care Med 2003;
29:845 848.
150. Sivak ED. Prolonged mechanical ventilation; an approach to weaning. Cleve Clin Q
1980; 47:89 96.
151. Cordasco EM Jr, Sivak ED, Perez-Trepichio A. Demographics of long-term venti-
lator-dependent patients outside the intensive care unit. Cleve Clin J Med 1991;
58:505 509.
152. Scalise PJ, Gerardi DA, Wollschlager CM, Votto JJ. A regional weaning center for
patients requiring mechanical ventilation: an 18-month experience. Conn Med
1997; 61:387 389.
153. Gracey DR, Hardy DC, Koenig GE. The chronic ventilator-dependent unit: a lower-
cost alternative to intensive care. Mayo Clin Proc 2000; 75:445 449.
154. Schonhofer B, Euteneuer S, Nava S, Suchi S, Kohler D. Survival of mechanically
ventilated patients admitted to a specialised weaning centre. Intensive Care Med
2002; 28:908 916.
7
Procedures in the Intensive Care Unit
SANJAY VADGAMA JANET AU and NADER

KAMANGAR
Pulmonary and Critical
Care Section Division of Pulmonary and
West Los Angeles Healthcare Center Critical Care
VA Greater Los Angeles Healthcare System Olive ViewUCLA Medical Center
and Geffen School of and Geffen School of Medicine at UCLA
Medicine at UCLA Los Angeles, California, U.S.A.
I. Introduction
This chapter is divided into four sections that focus on major practical principles
pertaining to the management of patients with arterial catheters, central venous cath-
eters, pulmonary artery catheters, and chest tubes. Each section carefully reviews the
indications, equipment, techniques, and complications of these procedures.
II. Arterial Catheters

A. Background
Arterial catheters are routinely inserted into critically ill patients for monitoring
arterial pressure and arterial oxygenation. Indwelling arterial catheters provide a
continuous assessment of systolic, diastolic, and mean arterial pressures. As
patients become increasingly unstable in the presence of shock, arrhythmias,
and vasoactive drug infusions, the traditional cuff pressures lose their accuracy.
Non-invasive techniques in these patients have proven to be unreliable (1,2).
Arterial catheters provide beat to beat information, allowing for rapid interpret-
ation of therapeutic interactions, helping guide the clinician in providing and
assessing the appropriate therapy.
219
220 Vadgama, Au, and Kamangar
B. Indications
The indications for arterial cannulation are shown in Table 1. The appropriate use
limits discomfort caused to the patient and avoids unnecessary complications.
C. Equipment
The equipment used varies depending on the artery to be cannulated and the insti-
tutions preference. Commonly used sites for cannulation are the radial, femoral,
axillary, and dorsalis pedis artery. The equipment includes a 19- or 20-gauge
Teon catheter over a guidewire, 1% lidocaine solution without epinephrine,
and 3.0 or 4.0 silk suture. Monitoring equipment consists of the monitor with
amplier, electronic monitoring equipment, oscilloscope display screen and
recorder, constant ush device, transducer, and uid lled non-compliant
tubing with stopcocks. Accurate and reproducible data is dependent upon the
absence of technical problems. Management of the most common problems is
discussed in the forthcoming section.
D. Techniques
As stated earlier, there are many potential sites for the insertion of arterial cath-
eters. The radial and femoral arteries are the most commonly used. When choos-
ing the ideal site for catheter insertion, several factors have to be considered.
These include hand dominance, patency of collateral circulation, ease for
nursing care, and absence of infection.
Radial Artery Cannulation
The radial artery remains the most frequently cannulated artery. Concerns about
collateral blood ow stem from studies which have demonstrated incomplete
palmar arch anastomosis in 3% to 6% (3) of specimens, that with inadequate
or absent ulnar ow in 12% (4). The modied Allens test, rst described in
1929 (5), has been used to test for collateral ulnar ow. The patients hand is elev-
ated with clenched st, while both radial and ulnar arteries are compressed by the
examiner. This allows the blood to be drained from the hand. Once pallor is pro-
duced, the hand is lowered and the st opened. One artery is released and the time
Table 1 Indications for Arterial Line Catheters
Hemodynamic monitoring In unstable patients (acutely hypotensive or

hypertensive), including patients on
vasoactive or inotropic drug therapy
Frequent arterial blood sampling For blood gas determination, particularly in
mechanically ventilated patients
Frequent blood sampling In patients with limited venous access
Intra-aortic balloon pump use
Procedures in the ICU 221
for color to return to the palm is noted. The procedure is then repeated with the
other artery. Color should return to the hand within seconds, indicating a patent
ulnar artery and intact supercial arch. There remains some controversy over the
validity of Allens test. When compared with Doppler examination, it has been
shown to have a sensitivity of 87% and specicity of only 18% (6). In the
absence of peripheral vascular disease, Allens test does not appear to be a pre-
dictor of ischemia of the hand during or after cannulation (7). Doppler ultrasound
has been shown to be a safe and rapid method to localize the radial artery in
difcult cases (8,9).
Once the radial artery is located either by palpation or Doppler ultrasound,
the wrist is placed in hyperextension. The skin is prepped and draped using sterile
technique. Using a 25-gauge needle, approximately 0.5 mL of the 1% lidocaine is
inltrated down to the periosteum on both sides of the radial artery. A 20-gauge
non-tapered Teon catheter over needle is advanced into the artery at a 308 to 458
angle to the skin, approximately 3 to 5 cm from the proximal wrist crease. When
blood return is noted in the hub, the needle is held in place. The angle is decreased
to 258 and the catheter advanced into the artery. When the needle is removed,
correct position is conrmed by blood return. Use of the guidewire technique
has been shown to be more successful than direct puncture alone (10). Using
this method, once the artery is located the guidewire is advanced into the
vessel (Fig. 1). The catheter is then advanced over the guidewire into the
artery. Upon removal of the guidewire and needle, pulsatile blood return
should be seen. The cannula is then connected to the transducer tubing and
securely sutured into the skin, and dressing is then applied.
Femoral Artery Cannulation

The femoral artery may be the only palpable artery in cases of profound shock. It
is a large bore vessel, easily palpable midway between the anteriosuperior iliac
Figure 1 The modied Seldinger technique of percutaneous arterial line placement.

Once the artery is punctured, the tab is used to advance the guidewire into the artery.
The cannula is then advanced over the guidewire into the artery; the needle and guidewire
are then removed. Source: From Kruse JA, Fink MP, et al. Percutaneous cannulation of
the radial artery. Saunders Manual of Critical Care, 2003; 688.
Table 2 Common Complications of Arterial Cannulation
Hematoma and hemorrhage

Limb ischemia
Thrombosis and thromboembolism
Infections
Pseudoaneurysms
spine and the symphasis pubis, lateral to the femoral vein, and medial to the
femoral nerve. The artery is cannulated using the Seldinger technique. Difculty
in cannulation is normally because of atherosclerosis, or prior vascular pro-
cedures. Specic complications to this site include retroperitoneal hemorrhage
and bowel perforation in large inguinal hernias.
E. Complications
Arterial cannulation remains a relatively safe procedure (11), with a small

percentage of clinically relevant complications. Complications of arterial cannu-
lation regardless of site are listed in Table 2. Site-specic complications are listed
in Table 3.
Thrombosis and Embolization

Thrombosis remains the most common complication of intra-arterial catheters,
occurring most frequently in radial and dorsalis pedis arteries and rarely in
femoral or axillary artery catheters. The incidence of thrombosis increases
with both the size of catheter and duration of cannulation beyond 48 to 72
hours (7,12,13). The risk of thrombosis has been reduced by the use of Teon
catheters (14) and the use of continuous or intermittent heparin ush systems
(15). Despite the relatively high incidence of thrombosis (up to 10% of
20-gauge radial catheters), it is clinically signicant in less than 1% of patients
(16). Regular examination of the extremity for signs of ischemia and prompt
removal of the catheter can minimize thrombotic complications.
Distal ischemia can also be caused by embolization of air or atheromatous
plaque. Air embolism is caused by gas in the ush solution, the inltrate being
under a pressure of 300 mmHg. Depending upon the patients size (smaller
Table 3 Site Specic Complications of Arterial Cannulation
Radial artery Cerebral embolization; peripheral neuropathy

Brachial artery Median nerve damage; cerebral embolization
Femoral artery Retroperitoneal hemorrhage;
bowel perforation; arteriovenous stula
Axillary artery Cerebral embolization; brachial plexopathy
patients) and position (sitting or upright), air embolism may travel in a retrograde
direction into the cerebral circulation via the vertebral arteries. Radial artery cath-
eters appear to have a higher incidence of embolization (17). The risk is reduced
by ensuring all air is removed from the tubing prior to ushing, and opening the
ushing valve for no more than three seconds.
Infection
Catheter-related infections are discussed in detail in the Central Venous
Catheters section. As when placing all intravenous access, thorough site prepa-
ration and sterile technique is paramount. The risk of infection increases with dur-
ation of cannulation (18,19). The incidence of arterial catheter colonization has
been reported to be from 5% to 10%, and it is similar in both radial and
femoral catheters (20 22). Routine changing of arterial lines has not been
shown to lower infection rates. Guidewire exchange is not recommended.
Systemic infections may also be caused by contaminated ush solutions or
equipment such as stopcocks or transducer domes. Staphylococcus epidermidis
remains the most common organism causing line infection (23). Gram-negative
organisms and Candida have also been reported, the latter more so in immuno-
compromised patients. Catheter infections should be treated with the appropriate
antibiotic for 7 to 14 days, and complicated cases may require prolonged therapy.
Catheters should be removed immediately in suspected or conrmed catheter
infections.
Most institutions do not replace or relocate arterial catheters routinely for
infection control. Transducers and continuous ush devices are replaced at 72
hours, along with intravenous tubing and ush solution.
Diagnostic Blood Loss

To avoid contamination when obtaining a sample, 3 to 5 mL of blood is typically
wasted with each sample drawn. When frequent samples are required, this can
result in a signicant amount of blood loss over time (24,25) requiring blood
transfusion. Methods of minimizing diagnostic blood loss include the use of pedi-
atric tubes, tubing systems that use a reservoir for blood sampling (24,26), and
continuous intra-arterial blood gas monitoring (27).
F. Technical Sources of Error

Technical problems from the equipment may result in erroneous data collection.
The most common problems encountered with the pressure monitoring system
are inadequate dynamic response, zeroing drift, and transducer or monitor cali-
bration (28).
The dynamic response of the catheter system is determined by the resonant
frequency and damping coefcient. Stimulation of the system results in oscil-
lation known as the natural or resonant frequency. The resonant frequency of
the monitoring system must be greater than the natural frequency of the input
signal. This avoids progressive amplication of the output signal known as

ringing, which may result in exaggeration of systolic peak. The occurrence of res-
onance is limited by the use of short, wide bore, low compliance catheters and
extension tubing. The time taken for an oscillating system to come to rest is
known as the damping coefcient.
A bedside ush test can be applied to the arterial catheter system to deter-
mine the damping coefcient and resonant frequency. Briey, opening and
closing the valve in the continuous ush system produces a square wave displa-
cement in the oscilloscope, followed by ringing and return to baseline. In
a normal system, the pressure wave initially seen is followed by a few
oscillating waveforms (Fig. 2A). An underdamped system shows lower fre-
quency oscillations resulting in falsely high systolic readings (Fig. 2B). Over-
damping is demonstrated by a ush test that does not produce oscillations after
the initial pressure release (Fig. 2C). Underdamping can be caused by connecting
tubing and stopcocks and patient factors such as tachycardia and high
Figure 2 Square wave testing demonstrating optimal, underdamped, and overdamped

frequency responses. (A) Optimally damped. One to two oscillations before returning to
tracing. Readings obtained are reliable. (B) Underdamped. .2 oscillations. Overestimates
systolic blood pressures. Diastolic pressure may be underestimated. (C) Overdamped. , 1 12
oscillations. Underestimation of systolic blood pressures. Diastolic pressure may not be
affected. Source: From Lichtenthal R, ed. Quick Guide to Cardiopulmonary Care, 2002.
output states. Overdamping may occur with catheter kinking or thrombus occlu-
sion, blood in the transducer dome or line, and air bubbles in the tubing or
stopcocks.
Improper zeroing is the most common source of error. The transducer can
be zeroed by opening the transducer stopcock to air and aligning it at the fourth
intercostals space using a level. If the transducer is positioned too high, it will
underestimate the true readings, giving falsely low pressures. Conversely, if it
is positioned too low, it will record falsely high pressures. This process should
be repeated each time the patients position is changed; when signicant
changes in blood pressure occur; and routinely every six to eight hours (29).
III. Central Venous Catheters

A. Background
Central venous catheters are used extensively both in intensive care units (ICUs)
and general wards. Knowledge of indications for placement, the appropriate site
selection, and complications are essential for all physicians, not only intensive
care specialists.
It has been estimated that over ve million central venous and pulmonary
artery catheters are inserted each year in the U.S.A (30). Their uses include
providing secure access to the central circulation for infusion therapy, hemody-
namic monitoring, nutritional support, temporary transvenous cardiac pacing,
and hemodialysis. The complication rates of venous catheters have been reported
to be over 15% (31 33), and some have been associated with signicant morbid-
ity and economic consequences.
B. Indications
Common indications for central venous catheter placement are listed in Table 4.
The indication for placement and patient-related factors, such as hemodynamic
stability, pulmonary problems, and sites of infection, determine the most appro-
priate site for catheter placement. The most common sites used are the internal
Table 4 Common Indications for Central Venous Catheter Placement
Lack of peripheral access, especially in patients requiring

rapid volume replacement in the setting of hypovolemic shock
Hemodynamic monitoring
For the administration of irritant medications or vasopressors
Long-term total parental nutrition
Acute hemodialysis
Temporary transvenous pacing
jugular, the subclavian, and the femoral veins. Advantages and disadvantages of
each site are listed in Table 5.
C. General Technique
Central venous catheter infections remain a major cause of morbidity in critically

ill patients. Placement should be performed under strict sterile conditions.
The importance of hand washing, aseptic technique, and full barrier precautions
cannot be overstated. Central venous catheters are placed by using the Seldinger
technique of guidewire dilation. The vein is cannulated with a needle, through
which the guidewire is passed. The dilator is then passed over the guidewire,
Table 5 Advantages and Disadvantages of the Internal Jugular, Subclavian, and

Femoral Approaches
Approach Advantages Disadvantages
Internal jugular Successful cannulation less Accidental internal carotid

operator dependent; higher artery puncture; poor
success rate; low risk of landmarks in obese/
pneumothorax; direct path edematous patients; difcult
to superior vena cava; easily access in hypovolemia;
compressible bleeding difcult access during CPR;
vessels left sided thoracic duct
injuries; difcult dressing
care; uncomfortable for
patient; not preferred site for
long-term catheterization
Subclavian Easier landmarks; less High pneumothorax risk;
infection risk for long-term malposition common;
catheterization; vessels compressibility of bleeding
accessible in hypovolemia; vessels difcult; difcult
easier dressing care; avoids access during CPR
airway
Femoral Easy landmarks; high success High infection rates; increased
rate; does not interfere risk of thrombosis;
with CPR; no risk of unreliable drug delivery to
pneumothorax; supine or the central circulation.
Trendelenberg is necessary
for placement; easily
compressible bleeding
vessels
Abbreviation: CPR, cardiopulmonary resuscitation.

followed by the catheter itself. The wire is then removed, leaving the catheter
within the vessel. Triple-lumen catheters are generally used over single-lumen
catheters because of their multiple ports.
Sites other than the internal jugular vein, subclavian vein, and femoral
vein include the external jugular vein and brachiocephalic vein. Insertion tech-
niques at the three most frequently used sites are discussed in the following
sections.
Internal Jugular Vein

The internal jugular vein provides direct access to the superior vena cava for pul-
monary artery catheterization and transvenous pacemakers that enter the right
side of the heart. Cannulation of the right internal jugular vein is preferred to
the left because it provides direct access to the right heart. The right pleural
dome lies lower than the left, thus having a lower theoretical risk of pneu-
mothorax. The right-sided approach also avoids the thoracic duct. Internal
jugular vein cannulation generally carries a lower risk of pneumothorax com-
pared with the subclavian approach, with greater ease of compressibility in the
event of excessive bleeding or accidental arterial puncture. In the absence of
clear anatomic landmarks, cannulation at this site may be difcult without the
use of ultrasound localization.
When preparing to cannulate the internal jugular vein, the patient should be
placed in Trendelenbergs position both to distend the vein and to prevent the risk
of air embolism. The head should be turned to the contralateral side. The operator
should be wearing sterile gloves, gown, and a surgical mask. Preparation of the
guidewire and catheter is important prior to insertion. All ports of the catheter
should be ushed with saline. The skin should be cleaned with 2% chlorhexidine
and a sterile eld established.
Cannulation of the internal jugular vein can be achieved by one of
three methods: the anterior approach, the median (or central) approach, and
the lateral (or posterior) approach. The anterior and median approaches use the
carotid artery as a palpable landmark. The landmarks for both these appro-
aches are easily identied in thinner patients. The lateral (posterior)
approach has the highest risk of carotid artery puncture, but the lowest risk of
pneumothorax.
In the median (central) approach (Fig. 3A), the apex of the triangle formed
by the two heads of the stenocleidomastoid muscle and clavicle is identied. The
vein runs beneath this triangle in its medial position prior to entering the superior
vena cava. The pulsation of the internal carotid artery is usually felt just medial to
the apex of the triangle. With ngers of the left hand on the carotid artery, a
22-gauge needle is used to inltrate 1% lidocaine into the skin at the apex of
the triangle. The 22-gauge needle is used to locate the internal jugular vein.
This needle is advanced at 308 to 458 in a frontal plane toward the ipsilateral
Figure 3 (A) Catheterization of the right internal jugular vein. The apex of the triangle
formed by the two heads of the sternocliedomastoid and clavicle is identied. The internal
jugular vein lies lateral to the internal carotid artery. (B) Catheterization of the right sub-
clavian vein. The skin is punctured 2 3 cm caudal to the midpoint of the clavicle. The
needle is then advanced in the direction of the sternal notch, hugging the inferior
surface of the clavicle. Source: From Mc Gee DC, Gould MK. Pulmonary complications
of central venous catheterization. N Engl J Med 2003; 348:1123 1133.
nipple applying constant back pressure to the syringe. The vein is normally ident-
ied within 2 to 5 cm from the skin surface. If the vein is not located, the needle
should be withdrawn slowly, with continued gentle back pressure as the vein is
commonly located during withdrawal of the needle. Repeated attempts should
be just medial or lateral to the initial attempt. Once the vein is located with the
nder needle, the operator can either withdraw the nder needle and insert the
larger needle in the same plane or leave the nder needle in and insert the
larger needle alongside it into the vein. Finder needles may not be necessary if
the cannulation is performed with ultrasound guidance. The syringe is then
removed leaving the hub in place. Backow is checked, ensuring it is not pulsi-
tile, an indication of arterial blood. The J-tip guidewire is then passed through the
needle into the vein, where it should pass easily, encountering no resistance. The
needle is then removed while maintaining control of the guidewire. A small skin
nick contiguous with the guidewire is made using an upward facing scalpel blade.
The dilator is then advanced over the wire using a twisting motion; the operator
should hold on to the guidewire at all times. The dilator should then be with-
drawn, leaving the guidewire in place; gauze is placed at the puncture site to
control oozing. The catheter is threaded through the guidewire into the vein
15 17 cm. At no point should the operator let go of the actual wire. With
the catheter stabilized, the guidewire is removed. Each port of the catheter
should be aspirated and then ushed with normal saline. The catheter is securely
sutured or stapled into the skin and sterile dressing applied. A chest X-ray is then
obtained to check for complications and position of the tip of the catheter.
In the anterior approach, the needle is inserted along the medial edge of
the sternocleidomastoid muscle at the level of the inferior margin of the
thyroid cartilage. The needle is inserted 1 cm lateral to the pulsation of the
carotid artery.
For the posterior (lateral) approach, the needle is inserted at the posterio-
lateral margin of the sternocleidomastoid, approximately 5 cm above the sterno-
clavicular joint. The needle is directed toward the suprasternal notch at a 158
angle to the skin. Venopuncture normally occurs within 5 to 7 cm.
Subclavian Vein
The subclavian vein approach has several advantages. This approach has more
easily identiable bony landmarks, a lower incidence of infection, easier dres-
sing care and maintenance, and improved patient comfort. It is the preferred
route for long-term total parental nutrition and central access in hypotension
when placed by experienced operators. The subclavian vein is a 1 to 2 cm
diameter vessel, a continuation of the axillary vein beginning at the lateral
border of rst rib. It is usually xed by brous attachment directly beneath
the clavicle for 3 to 4 cm before becoming the brachiocephalic vein. The
brous attachments prevent the vein from collapsing even in severe
hypovolemia. The anterior scalene muscle separates the subclavian vein from
the subclavian artery.
As with the internal jugular approach, the patient is placed in
Trendelenbergs position. Additionally, a small rolled towel is placed between
the shoulder blades. There are two basic techniques for inserting the subclavian
catheter, the supraclavicular approach and the infraclavicular approach. Catheter
malposition and pneumothorax rates may be a little lower using the supraclavi-
cular approach. There is no difference in the success rates between both
approaches.
With the patients head turned towards the contralateral side, the area is
prepped with 2% chlorhexidine. Using sterile technique, the area is draped to
create a sterile eld. For the infraclavicular approach, the skin is punctured 2
to 3 cm caudal to the midpoint of the clavicle (Fig. 3B). Following inltration
with 1% lidocaine, an 18-gauge needle attached to a 10-mL syringe is advanced
in the direction of the sternal notch until the tip of the needle abuts the clavicle.
The needle is then walked down the clavicle until it slips under its inferior
edge. Keeping the bevel pointed cephalad, the needle is advanced towards the
suprasternal notch, hugging the inferior surface of the clavicle. If no blood
return is seen, the needle is gently withdrawn, maintaining gentle suction. If
venopuncture does not occur, the needle should be angled in a slightly more
cephalad direction for the next attempt. Once blood return is achieved, the
needle is rmly held, whereas the syringe is detached from the hub. A nger is
placed over the hub to prevent air embolism. The blood return should be non-pul-
sitile. The guidewire is then passed through the needle, turning the bevel as the
guidewire is advanced; the needle is then withdrawn. As with the internal
jugular approach, the catheter is passed over the wire following dilatation. The
catheter is sutured 16 to 17 cm if placed on the right side of the chest and 18
to 19 cm if placed on the left. A post-procedure chest X-ray is taken.
For the supraclavicular approach, the operator is positioned at the head of
the patient on the side to be cannulated. The venopuncture site is the claviculo-
sternocleidomastoid angle, found lateral to the insertion of the clavicular head of
the sternocleidomastoid muscle, above the clavicle. The needle is advanced
underneath the clavicle, at an angle of 158 toward the contralateral nipple. The
vein is usually found at a depth of 1 to 4 cm from the surface.
Femoral Vein
Cannulation of the femoral vein has several advantages. As the artery is
usually easy to palpate, the vein, which lies medially, is usually easy to
locate. If the femoral artery is accidentally punctured, it is easily compressible.
Patients can be cannulated without being placed in Trendelenbergs position,
important when severe orthopnea may be a problem. It also avoids the risk of
pneumothorax. Femoral access is useful during cardiopulmonary resuscitation
when chest compressions and airway management may make subclavian and
internal jugular vein cannulation difcult. Disadvantages of this site include the
higher infection rates and unreliable drug delivery to the heart in low ow states.
The femoral vein is a continuation of the popliteal vein, becoming the
external iliac vein at the inguinal ligament. The femoral vein lies within the
femoral sheath, medial to the femoral artery, which is usually easily palpable.
If possible, the patient is placed in the supine position. The femoral vein lays
approximately 1 to 1.5 cm medial to the femoral artery and 1 to 3 cm below the
inguinal ligament. Once again the site is meticulously sterilized and prepped as
described earlier. An 18-gauge needle is advanced at a 458 to 608 angle, pointing
the needle towards the head. In obese patients it may be necessary to advance the
needle up to the hub. Once venous return is conrmed, the syringe is removed.
Ensuring non-pulsatile blood ow, the guidewire is passed through the needle
and venous catheter placed as described previously.
D. Ultrasound-Guided Insertion of Central Venous Catheters
Failure rates for insertion of central venous catheters have been found to be as
high as 12%, with a complication rate of 10% (34). Commonest causes of
failure or complication include previously difcult catheterization, limited
sites, difcult landmarks, and thrombosed vessels. Studies have clearly demon-
strated the superiority of using ultrasound guidance over blind landmark-
guided techniques in the time taken to place the catheter, the number of attempts,
and the complication rate (35 38). The ultrasound also identies thrombosed or
unusually small vessels, thus avoiding inaccessible sites. Sterile ultrasound probe
covers and needle guidance clips allow cannulation of the vessel as it is seen in
real-time on the monitor (Fig. 4A). Although helpful, the guidance clips can be
cumbersome, and some operators prefer cannulation under real-time conditions.
Real-time ultrasound uses high-frequency sound waves (2 10 MHz), generating
a two-dimensional grey-scale image of the vein and surrounding tissues (Fig.
4B). Fluid such as blood transmits sound completely and is seen as a dark
image. Veins are identied by their non-pulsatile appearance, compressibility,
and distension with the patient in Trendelenbergs position (Fig. 4C). The
artery is generally not compressible on application of gentle pressure. Ultra-
sound-guided catheter placement appears to be the most benecial for the internal
jugular site and should be used whenever available.
E. Complications of Central Venous Catheters

Complications of central venous catheters can be divided into mechanical,
embolic, thrombotic, and infectious.
Mechanical complications include arterial puncture, hematoma, malposit-
ion, and pneumothorax. Accidental arterial puncture is the most frequent compli-
cation of the internal jugular site, with rates being reported to be around 3% (39).
Bleeding can usually be controlled by the application of local pressure for several
minutes. Occasionally a hematoma may form, which can hinder further catheter
placement attempts. Accidental cannulation of the internal carotid artery has been
reported (40), leading to retrograde dissection of the subclavian, innominate
artery, and ascending aorta (41). Subclavian artery puncture is less common,
usually managed by applying pressure above and below the clavicle. Femoral
artery puncture occurs in 5% to 10% of adults (42 44); however, this is
mostly uncomplicated, controlled by local pressure.
Cannulation of the left internal jugular vein has its own complications.
These include injury to the left innominate vein and thoracic duct. Because of
its anatomy, cannulation of the left innominate vein has a greater risk for
perforation. Care should be taken when dilation over the guidewire is attemp-
ted, being careful not to force the dilator beyond the internal jugular vein.
Damage to the thoracic duct can result in chylothorax and rarely chylo-
pericardium (45).
Catheter malposition rates have been reported at 5.3% versus 9.3% for
internal jugular versus the subclavian site (39) and lower for the femoral vein.
When using the internal jugular and subclavian sites, care should be taken to
ensure the catheter tip is not placed too low into the heart. The ideal tip position
is 3 to 5 cm proximal to the caval atrial junction. Fatal tamponade can occur sec-
ondary to perforation of the right atrium (RA) or ventricle. Arrhythmias have also
been reported (46).
Figure 4 (A) An illustration of the ultrasound probe with needle guide attached. Site-
rite ultrasound system. Source: C.R. Bard, Inc., (B) Real-time ultrasound demonstrating
the internal jugular vein and common carotid artery, with the patient in Trendelenbergs
position. (C) Upon application of gentle downward pressure from the ultrasound probe,
the vein collapses but the artery remains clearly seen. (Continued next page.)
Figure 4 (Continued.)
Malposition of the subclavian catheters into the ipsilateral jugular vein has
been well documented (47). Catheters can enter the contralateral brachiocephalic
vein more commonly in left-sided insertions. An example of malposition is
shown in Figure 5A; the central venous catheter was inserted into the left
internal jugular vein and is shown looping back and entering the right internal
jugular vein.
Figure 5 Examples of complications. (A) Anterio-posterior radiograph showing the tip

of the central venous catheter, inserted in the left internal jugular vein, entering the right
internal jugular vein. (B) Anterio-posterior radiograph showing a complication of femoral
artery catheterization, with the guidewire displaced distally into the aortic arch.
Pneumothorax is a recognized complication of both the internal jugular and

subclavian approaches. The incidence of pneumothorax ranges from 0.3% to
3.0% (48). The complication rates at these sites increase signicantly with the
number of attempts made at cannulation (34,39). The incidence of pneumothorax
is higher for the subclavian approach than internal jugular, where pneumothorax
only occurs if attempts at catheterization occur close to the clavicle. Treatment of
the pneumothorax depends on the patients clinical condition, severity of under-
lying disease, and the size of the pneumothorax. Although most pneumothoracies
are immediately apparent, they can also occur as a late complication, occurring
several days after subclavian vein catheterization (50).
Catheter and wire embolism occur most commonly when they are with-
drawn through the needle as a result of the shearing action over the barrel of
the needle. The fragments of the catheter can embolize, resulting in potentially
serious complications (51). Shearing of the wire is more common than of the
catheter. Air embolism occurs most commonly when catheters are accidentally
disconnected, becoming open to the atmosphere. Complications range from tran-
sient hypoxemia and chest pain, focal cerebral lesions with hemiparesis, or hemi-
anopia (52) to cardiovascular collapse and death (53). Placing the patient in
Trendelenbergs position during catheter insertion and removal is important in
the prevention of air embolism and its complications. Infrequently, the operator
may lose the guidewire during insertion as is shown in Figure 5B, where the
guidewire can be seen in the descending aorta.
Thrombosis
Although a high incidence of thrombosis has been reported in central venous

catheters, the clinical signicance is less clear. Catheter associated thrombosis
of the femoral vein appears to have more clinical signicance than either the
internal jugular or the subclavian vein (31,54,55) and may be partly responsible
for the higher incidence of infection (31,54 57) seen at this site.
Infection
Intravascular catheters play an essential role in the management of critically and
chronically ill patients, with millions purchased by healthcare institutions.
Catheter-related infections are associated with increased morbidity, longer
duration of hospitalization, and substantial nancial costs. The incidence of
catheter-related infections varies according to the type of catheter used, the site
of placement, the conditions under which the catheter was placed, frequency of
manipulation, and patient-related factors, such as the severity of their underlying
disease (58). The diagnosis of catheter-related infection is often difcult because
symptoms of fever, chills, and signs of sepsis may occur from other foci of infec-
tion. Catheter-related infection is suggested by several factors. These include
inammation or frank pus at the catheter insertion site or exit site, dysfunction
of the catheter secondary to intraluminal clot, isolation of coagulase negative sta-
phylococci, corynebacterium species or fungi, and clinical improvement on
removal of the catheter.
There should also be an absence of infection from other sites, such as the
lung and urine. The presence of blood stream infection is conrmed by two
positive blood cultures drawn from a peripheral vein, not the catheter itself,
which may be colonized with contaminants. Catheter-related blood stream infec-
tions are unlikely if blood cultures drawn from the catheter are negative (59).
Positive cultures of the catheter tip should demonstrate more than 15 colony
forming units of the same organism (60). The evaluation and management of
catheter-associated infections is discussed in more depth in the chapter on
infection control in the ICU.
With infection of central venous catheters being such a troublesome cause
of nosocomial infections, prevention of infections is paramount. Aseptic tech-
nique and strict adherence to hand washing have been found to signicantly
reduce the rates of catheter-related infections (61 63). Whenever placing a cath-
eter, full barrier precautions are necessary, including sterile gloves, long-sleeved
gowns, surgical masks, and sterile drapes (64). Two-percent chlorhexidine has
been shown to be a more effective cutaneous disinfectant than povidone
iodine (65). Use of the subclavian vein has been shown to have a lower infection
rate compared with both internal jugular and femoral veins (31,66,67). Femoral
veins have the highest rate of infection and hence the use of this site is generally
discouraged. Chlorhexidine silver impregnated catheters have been shown to
have a lower rate of catheter colonization and catheter-related bacteremia (68),
as have minocycline rifampin bonded catheters (69). At present, antimicrobial
impregnated catheters are reserved for use when rates of catheter-related infec-
tions remain high despite appropriate prevention measures.
Catheters should be removed as soon as they are no longer needed. Infec-
tion rates increase after the fth to seventh day (70 72). Routine exchange of
catheters over guidewire or new puncture has not been shown to reduce the
rates of catheter-related infections (73,74) and carries the risk of mechanical
complications. The catheter site should be examined daily for evidence of puru-
lence or erythema, suggestive of infection requiring removal of catheter. The
absence of these signs does not exclude the possibility of a catheter-related
infection.
Antibiotic therapy should be directed by blood and catheter tip cultures, the
severity of the patients underlying disease, and the hospitals infection and
resistance patterns.
Future developments will be aimed at the use of more effective preventa-
tive strategies, including more effective antiseptics and catheters made of
new polymer-antibiotic systems that will inhibit the formation of a bacterial
biolm.
IV. Pulmonary Artery Catheters

A. Background
Pulmonary artery (PA) catheterization was introduced in 1929 by a German phys-
ician, Werner Forssmann, who guided a urologic catheter through a patients vein
in his arm and into his heart. He took a chest radiograph to conrm the position of
the catheter and published the procedure as a brief report (75). In 1947, Dexter
et al. measured PA pressure by wedging a catheter in a distal branch of a pulmon-
ary artery. Subsequently, they found that the pressure in the wedge position was
the same as the lling pressure in the left ventricle (LV). By measuring the
pressure and oxygen content of blood in the RA, right ventricle (RV), and
pulmonary artery, they were able to diagnose congenital heart disease, valvular
heart disease, and left ventricular failure (76,77). Bedside catheterization of the
pulmonary artery was introduced by Swan et al. in 1970 (78). An inatable
balloon at the tip of a exible catheter enabled the catheter to be directed by
blood ow into the pulmonary artery, thus avoiding the need for uoroscopy-
guided placement. This ow-directed pulmonary artery catheter, also known as
the Swan Ganz catheter, was originally used to assess and guide therapy of
patients following acute myocardial infarction.
With subsequent improvements in PA catheter technology, it is now used in
both diagnosis and management of a wide range of conditions in critically ill
patients.
Despite its extensive use over the past 30 years, there has been increasing
concern over the use and interpretation of the data obtained by the PA catheter,
with some studies suggesting an increased mortality with its use. These contro-
versies will be discussed later.
B. Indications and Contraindications for PA Catheter Placement

Indications for PA catheter can be broadly divided into diagnostic and therapeutic
categories (Table 6). Diagnostic indications include differentiating between and
assessing different forms of shock, differentiating between cardiogenic and
non-cardiogenic pulmonary edema, distinguishing between primary and
secondary pulmonary hypertension, and diagnosing cardiac tamponade and intra-
cardiac shunts. Therapeutic indications include the management of shock,
complicated myocardial infarctions, guiding uid management in a variety of
conditions, and also guiding pharmacologic therapy. The use and interpre-
tation of data from the PA catheter in specic clinical scenarios will be
discussed later.
Contraindications for catheter placement are can be divided into absolute
or relative (Table 7) (79). Absolute contraindications include patients with
right-sided endocarditis and patients with demonstrated tumor or thrombus
within the RA or RV. Catheter placement is also contraindicated in prosthetic tri-
cuspid or pulmonic valves, where there is a danger of the catheter becoming
entangled in the valve mechanism, and in patients with terminal illnesses,
where aggressive management is futile. The balloons of most catheters are
made of latex; therefore, they are contraindicated in patients with a history of
latex allergy. Relative contraindications include patients with an underlying
Table 6 Indications for PA Catheter Use
Diagnosis
Differentiation and assessment of shock
Cardiogenic
Hypovolemic
Septic
Pulmonary embolus
Differentiation of pulmonary edema
Cardiogenic versus non-cardiogenic (e.g., ARDS)
Primary versus secondary pulmonary hypertension
Diagnosis of pericardial tamponade
Diagnosis of left to right intracardiac shunt
Therapeutic
Management of complicated myocardial infarction
Hypotension unresponsive to volume challenge
Hemodynamic instability (cadiogenic shock versus hypovolemia)
Ventricular septal rupture versus acute mitral regurgitation
Right ventricular infarction
Assessment of valvular heart disease
Guide to uid management
Gastrointestinal hemorrhage
Sepsis
Heart failure
Acute renal failure
Burns
Decompensated cirrhosis
Guide to pharmacological therapy
Afterload and preload reducing agents, inotropics, vasopressors, beta-blockers
Management of perioperative cardiac and noncardiac surgical patients with cardiac
instability
Post-operative management of open heart surgical patients
Management of severe pre-eclampsia
Ventilator management (assessment of best PEEP for oxygenation)
Abbreviations: ARDS, acute respiratory distress syndrome; PEEP, positive end-expiratory pressure.
severe coagulopathy (or receiving anticoagulation) at the time of catheter place-

ment. Fluoroscopic guidance is recommended for patients with temporary pacing
wires or recently placed permanent pacemakers or implantable debrillators,
dilated RA or RV, and left bundle branch block. Patients with left bundle
branch block are at risk of developing right bundle branch block leading to
complete heart block, although the occurrence of this complication in these
patients is very low. See the Complications of PA Catheters section for more
details.
Table 7 Contraindications for PA Catheter Placement
Absolute
Right sided endocarditis
Mechanical tricuspid (or pulmonic) valve
Tumor or thrombus in the right atrium or right ventricle
Terminal illness where aggressive is considered futile
Latex allergy
Relative
Coagulopathy (or anticoagulation)
Recently placed permanent pacemakers or implantable debrillators
C. Equipment and Technique

PA Catheters and Monitoring Equipment
A variety of catheters are now available. The typical PA catheter is 110 cm in
length, having a 1.5 cc balloon located just proximal to its tip. Double lumen
catheters allow balloon catheter ination through one lumen, with the distal
opening at the catheter tip used to measure intravascular pressures and sample
blood. Triple lumen catheters have a proximal port 30 cm from the catheter tip
allowing simultaneous measurement of right atrial, PA, and wedge pressures.
The four-lumen catheter is the most commonly used PA catheter. It contains a
thermistor 4-cm proximal to the tip that allows measurement of the PA blood
temperature and calculation of the cardiac output (CO) by thermodilution tech-
niques. Five-lumen catheters have a proximal port at 4 cm allowing delivery of
medications and venous blood sampling. PA catheters and electrode on its
surface for continuous ECG monitoring or temporary cardiac pacing are also
available. Special features in some pulmonary artery catheters allow continu-
ously measurement of mixed venous oxygen saturation (SvO2) via beroptics,
continuous cardiac output (CCO), continuous right ventricular end diastolic
volume (RVEDV), right ventricular ejection fraction (RVEF), and pacing.
These parameters have the ability to continuously guide the clinician as to the
status of oxygen delivery and consumption as well as guide on the assessment
of preload, afterload, and contractility.
Semi-rigid, non-compliant tubing lled with isotonic saline and heparin trans-
mits the pressure recorded to a uid-lled pressure transducer. Back-up of blood is
prevented by a constant infusion device. As the uid in the non-compliant tubing is
non-compressible, changes in intracardiac pressure are accurately transmitted to the
transducer membrane, generating an amplied signal seen on the monitor.
Changes in the height of the transducer or the patient will alter the pressure
measurements. The system is zeroed to ambient air pressure prior to recording
any readings. The reference point is the midpoint of the left atrium, estimated
as the fourth intercostal space in the mid axillary line with the patient in the
supine position. The transducer is xed at this height, the membrane exposed to
atmospheric pressure, and the monitor adjusted to zero. For most catheters, cali-
bration of monitors is not required.
It is important for the monitoring system to have an appropriate frequency
response. Flicking or gently shaking the catheter tip should elicit a brisk high-
frequency response in the waveform. Damped waveforms are caused by air
bubbles, long non-compliant tubing, vessel wall impingement, intra catheter
debris, and loose connections in the tubing. Following insertion, a rapid ush
test is performed similar to that described in the Arterial Catheters section.
D. Insertion
PA catheters can be inserted through a choice of sites. The risks and benets of
the most popular sites are listed in Table 8. Most physicians prefer the right
internal jugular or the left subclavian vein approach. The right internal jugular
is the shortest and straightest path to the heart. The curvature of the catheter uti-
lizes the course of the left subclavian vein into the SVC and right heart.
A wide sterile procedural eld is prepared and aseptic technique and full
barrier precautions are taken. The catheter is ushed in all ports and the balloon
inated to test for leaks. The balloon is then deated. Pressure transducer function
is checked by icking or waving the distal catheter tip gently. Using techniques
described in the central lines section, the appropriate site for insertion is identied,
and the vein is located and cannulated using the Seldinger technique. A dilator is
passed through the introducer, and together these are advanced over the guidewire
into the vein. The guidewire and dilator are then removed simultaneously, leaving
the introducer in the vessel. The catheter is introduced through the sterile sheath
adapter, which at the end of the procedure is pulled over the catheter. The distal
end of the sleeve adapter attaches to the introducer sheath limb. The catheter con-
tains markings spaced 10 cm apart from the distal tip. Using these markings, and
continuous pressure monitoring, the catheter is advanced into the RA. Table 8 illus-
trates the appropriate distances for the three most frequently used venous sites.
Table 8 Comparison of the Three Most Frequently Used PA Catheter Insertions Sites
Distance (cm) to:
Insertion site RA RV PA Advantages Disadvantages
Internal jugular 15 20 30 40 Shortest, most direct Risk of carotid artery

route to heart puncture, pneumothorax
Subclavian 15 20 30 40 Easier to oat from Higher risk
the left side of pneumothorax
Femoral 30 40 50 Easy cannulation, More difcult to oat,
fewer overall higher infection rate
complications
The catheter is advanced until it is in the RA; the appropriate right atrial
waveform should be seen with characteristic a and v waves. Right atrial blood
samples are obtained, if needed, and the pressure recorded. The balloon is then
inated with the recommended amount of air (no more than 1.5 mL) until a
small amount of resistance is felt. If no resistance is felt, one should suspect
balloon rupture and re-examine the catheter immediately. If signicant resistance
is encountered, then one should suspect malposition. The catheter is withdrawn
and re-advanced to a new position. With the balloon inated, the catheter is
advanced into the RV. As it crosses over the tricuspid valve into the RV, the
pressure waveform changes and a rise in systolic pressure is seen (Fig. 6). The
pressures within the RV are recorded. The catheter is then advanced until the
diastolic pressure increases above that seen in the RV. At this point, the catheter
tip has moved across the pulmonary valve into the PA. In this position, the
characteristic dicrotic notch appears in the waveform, indicating closure of
the pulmonic valve. If the RV tracing is still seen after 40 cm of the catheter is
introduced, it is likely that the catheter is coiled within the RV. If this occurs,
the balloon is deated and catheter withdrawn until the RA waveform is seen.
The balloon can then be inated and advanced again as described. The inated
catheter is advanced into the PA until a fall in pressure and a change in waveform
is seen. This is the pulmonary capillary wedge pressure (PCWP). The PCWP is
recorded as a mean value, not systolic or diastolic, and is measured at the end of
expiration. Once this is recorded, the balloon should be deated, at which point
the PA tracing should reappear. The balloon ination volume needed to change
the PA tracing to the PCWP should be noted. If the volume needed is signicantly
lower than the recommended volume of 1.5 mL or subsequent PCWP readings
require smaller volumes, then the catheter tip has migrated too far peripherally
and should be pulled back immediately (with the balloon deated). Characteristic
waveforms in the right heart chambers, pulmonary artery, and PCWPs are shown
in Figure 7.
As a rule, the catheter tip should never be advanced until the balloon is
inated and must always be deated if the catheter is withdrawn. In patients
with an enlarged RA and RV, it may be difcult to advance the PA catheter
into PA and wedge position. The catheter softens as it is exposed to body
temperature, which makes it more difcult to pass in patients with pulmonary
hypertension. The catheter stiffness can be maintained by placing it in a
refrigerator or freezer prior to insertion. Alternatively, PA catheters that have
a guidewire within one of the lumens to maintain its form may be used. In dif-
cult cases, uoroscopy is useful in guiding the PA catheter to the appropriate
position.
Once the optimum position has been located, the catheter is secured by
suturing or taping it to the skin. The position of the catheter based on the centi-
meter markings at the site of insertion is noted. An appropriate dressing is then
applied. Finally, a chest X-ray is taken to conrm the catheter position and
exclude a pneumothorax. Daily chest X-rays are recommended.
(A)
(B)
(C)
(D)
Figure 6 Posterio-anterior catheter position and the characteristic pressure waveforms

in relation to the EKG tracing, from the right atrium to pulmonary artery wedge position.
(A) Right atrium. (B) Right ventricle. (C) Pulmonary artery. (D) Pulmonary artery wedge
pressure. Abbreviations: a, atrial systole; c, closure of tricuspid valve; v, atrial lling,
ventricular systole. PA, pulmonary artery; PAW, pulmonary artery wedge pressure.
Source: From Lichtenthal R, ed. Quick Guide to Cardiopulmonary Care, 2002.
E. Waveforms and Hemodynamics

Once appropriately inserted and secured, the PA catheter waveforms are
measured and calculated parameters can be interpreted.
Figure 7 Normal pressures and waveforms obtained as the PA catheter is advanced

from the right atrium to a pulmonary artery wedge position. Abbreviations: RA, right
atrium; RV, right ventricle; PCWP, pulmonary capillary wedge pressure. Source: From
Lichtenthal R, ed. Quick Guide to Cardiopulmonary Care, 2002.
The right atrial pressure waveform reects the venous return to the RV and,
in the absence of tricuspid valve pathology, is an indicator of right ventricular end
diastolic pressure. The normal atrial pressure waveform consists of ve com-
ponentsthree positive deections, the a, c, and v waves, and two negative
deections, the x and y descents (Table 9). The normal pressures recorded by
the PA catheter are listed in Table 10. Changes in the right atrial pressure wave-
forms and pressures within the heart in various pathological conditions are dis-
cussed later in this section. Using data obtained from the PA catheter, useful
hemodynamic parameters can be derived. Table 11 shows how these values
are derived and their normal values.
Pulmonary Capillary Wedge Pressure
When the mitral valve is open, there is no barrier between the tip of the catheter
and the left heart. Ination and wedging of the balloon closes off ow from the
right heart, so the catheter only reects a backward pressure from the left heart.
Ideally, left ventricular end diastolic volume (LVEDV) is the desired value for
hemodynamic monitoring, but it unfortunately requires cardiac catheterization.
Table 9 Waveform Presentations
Waveform/ Correlation to EKG and cardiac cycle

descent
a wave Atrial systole, it follows P wave (within PR interval) on ECG
c wave Bulging of the tricuspid valve into the right atrium (RA) during
right ventricular (RV) systole
v wave Passive atrial lling with a closed tricuspid valve, near the end of T
wave, correlates with the beginning of QRS and later 1/2 of PR
interval on ECG in RA tracing, after T wave in the wedge
(PAWP) tracing
x descent Atrial diastole (RA relaxation)
y descent Atrial emptying by opening of the tricuspid valve with ow of
blood into RV
The left ventricular end diastolic pressure (LVEDP) is used as an alternative;

however, this value cannot be measured directly. The PA wedge pressure
approximates the left atrial pressure, which in turn is usually slightly less than
the left ventricular pressure in diastole, or intracavitary LVEDP (lling pressure)
(80 95). This relationship allows the PA wedge pressure to be used as a close
approximation of the LVEDP. Therefore, the wedge pressure does not directly
measure LVEDV but provides an acceptable estimate of the volume in the LV.
The relationship between LVEDV or preload (lling volume) and wedge
pressure (lling pressure) depends on:
1. Ventricular compliancemeaning the distensibility of the LV.

2. Transmural ventricular distending pressureequals the intracavitary
pressure minus the juxtacardiac pressure.
Changes in juxtacardiac pressure, as with positive end-expiratory pressure

(PEEP) or auto-PEEP, or changes in ventricular compliance, as with ischemia or
vasoactive agents, can change the values and therefore interpretation of the
Table 10 Normal Pressure Values Within the Heart
Measurement from PA catheter Normal range

Central venous pressure (mmHg) ,10
Right arterial pressure (mmHg) ,10
Right ventricular pressure, systolic (mmHg) 15 30
Right ventricular pressure, diastolic (mmHg) 08
Pulmonary artery pressure, systolic (mmHg) 15 28
Pulmonary artery pressure, diastolic (mmHg) 5 16
Pulmonary artery pressure, mean (mmHg) 10 22
Pulmonary artery wedge pressure, mean (mmHg) 6 12
Cardiac output (L/min) 47
Table 11 Hemodynamic Parameters Calculated From the PA Catheter and Their

Normal Values
Parameter Calculation Normal values
CO (Fick Equation) VO2/(CaO2-CvO2) 4 7 L/min

CI CO/BSA 2.8 4.2 L/min/m2
SV CO/HR 50 80 mL/beat
SI CO/HR BSA 30 65 mL/beat/m2
Left ventricular 0.136 (Stroke volume) 43 61 g/beat/m2
stroke work index (MAP 2 PCWP)
Right ventricular 0.0136 (stroke volume) 7 12 g/beat/m2
stroke work index (mean PAP 2 CVP)
SVR (MAP 2 CVP/CO) 80 11 18 mmHg/L/min
880 1440 dyne/sec/cm25
PVR (Mean PAP 2 1.5 to 3.0 mmHg/L/min
PCWP/CO) 80 150 to 240 dyne/ sec/cm25
% Shunt (CcO2 2 CaO2)/(CcO2 2 CvO2) ,5%
Oxygen uptake (CaO2 2 CvO2) CO 10 150 300 mL O2/min
Abbreviations: BSA, body surface area; CO, cardiac output; CVP, central venous pressure; CaO2,
arterial oxygen content; CcO2, pulmonary capillary blood oxygen content; CvO2, venous oxygen
content; CvO2, mixed venous oxygen content; HR, heart rate; MAP, mean arterial pressure;
PAP, pulmonary artery pressure; PCWP, pulmonary capillary wedge pressure; VO2, oxygen
consumption.
PCWP. PEEP causes a decrease in LV transmural pressure, so a falsely high

PCWP may be obtained. It will overestimate the LVEDP with PEEP or
auto-PEEP, but the extent is difcult to determine precisely (94,95). With stiff
ventricles, as seen in pericardial effusion, myocardial infarction (MI), and LV
hypertrophy, the PCWP may be less than the actual LVEDP. In aortic regurg-
itation, the LVEDP is higher than the PCWP. Conditions in which the PCWP
will overestimate the LVEDP are mitral valve regurgitation, left atrial
myxoma, pulmonary embolism, intracardiac left to right shunt, and non-zone 3
catheter position (92).
The PA catheter must be in a zone 3 lung segment to provide accurate
readings. There are three lung zones depicted by West (92), which are as follows:
Zone 1: The least gravity-dependent area of the lung where, theoretically,

there is no blood ow because pulmonary artery and venous pressure are
less than alveolar pressure.
Zone 2: Pulmonary artery systolic pressure is greater than alveolar
pressure, but alveolar pressure is greater than pulmonary venous pressure.
As pulmonary artery pressure (systolic and diastolic) and alveolar pressure
(inspiration and expiration) are phasic, the changing perfusion ventilation
pressures allow only intermittent ow.
Zone 3: It is the gravity-dependent area of the lung where, under normal

circumstances, pulmonary artery and venous pressure always exceed alveo-
lar pressures, and blood ow is constant. When the PA catheter is wedged
in zone 3, where both arterial and venous pressures are greater than alveolar
pressure, resulting in uninterrupted blood ow to the left heart, then the
PCWP reects the pulmonary venous and left heart pressures.
These are not anatomically xed zones but rather functional zones. Any
anatomic part of the lung may take on the characteristics of zone 1, 2, or 3,
depending on alterations in hemodynamic (increased or decreased pulmo-
nary blood volume and pressures) and ventilatory status (PEEP, alveolar air
trapping).
Relationship Between PCWP and Respiration

There can be considerable variation in the waveform tracings because of respir-
atory variation. The general pattern may resemble a sine wave. Of course, there
may be minimal variation in the waveform baseline and no need to make any
adjustments for respiratory variation. By convention, readings are made at the
end-expiration, but they will be at different points if the patient is spontaneously
breathing or is being mechanically ventilated. At end-expiration, the inuences of
pleural pressure during spontaneous breathing or airway pressures during mech-
anical ventilation are minimized. During spontaneous breathing, pleural pressure
usually approximates atmospheric pressure at end-expiration (Fig. 8). This
should represent a zero reference point with minimal inuence of external press-
ures. During mechanical positive pressure breathing, positive airway and
intrathoracic pressures are at their lowest at end-expiration (Fig. 9). If there is
signicant inuence due to respiratory effects, end-expiration will occur at the
highest portion of the waveform during spontaneous breathing and at the
Figure 8 Ventilatory effect of spontaneous breathing on PCWP. PCWP values should

be obtained at end-expiration when intrathoracic pressure inuence is minimal. Abbrevi-
ation: PCWP, pulmonary capillary wedge pressure.
Figure 9 The effect of controlled mechanical ventilation on PCWP. The PCWP is

recorded at end-expiration as shown. The positive deections following end-expiration
reects positive pressure inspirations. Abbreviation: PCWP, pulmonary capillary
wedge pressure. Source: From Lichtenthal R, ed. Quick Guide to Cardiopulmonary
Care, 2002.
lowest portion of the waveform during mechanical ventilation. The average of

three or more end-expiratory values should be used to minimize variation.
These inuences are especially important if the patient is rapid breathing or
being mechanically ventilated. In this situation, digital readout devices at bedside
are inadequate. These devices do not compensate for changes in waveform base-
line, and readings are not reliable and may be inaccurate. In order to correct this
error, it is better to read the PCWP at the highest or peak pressure instead of at the
average pressures for spontaneously breathing patients. Conversely, the digital
readout should be read at the lowest or trough pressures in mechanically venti-
lated patients. In some patients with severe dyspnea and marked respiratory
distress, there can be considerable inuence from actively contracting expiratory
muscles, and accurate readings may not be possible without eliminating respi-
ratory muscle activity either with heavy sedation or a short-acting paralytic. In
these situations, it is preferable to obtain pressure tracings on calibrated paper
and visually identify the pressures at end-expiration.
PEEP and PWCP

When a patient is receiving PEEP, the PCWP will increase because PEEP causes
a reduction in the left ventricular transmural pressure, as the pleural pressure and
juxtacardiac pressure at end-expiration rise more than the intravascular pressure.
However, this effect is not uniform in all conditions. At end-expiration, pleural

pressure will be less than one-half the PEEP in patients with poor lung compli-
ance (e.g., patients with ARDS). The net effect of PEEP is to cause measured
PCWP to overestimate transmural LVEDP to an uncertain degree. If the level
of PEEP is less than 10 cmH2O, then the effect is negligible. If PEEP is
greater than 10 cmH2O, then one can estimate the juxtacardiac pressure by
assuming that one half of the PEEP level is transmitted to the pleural and juxta-
cardiac space, and then subtract that value from the measured PCWP to approxi-
mate the real LVEDP. A similar effect on PCWP results when a patient has
air-trapping and generates auto-PEEP. It is generally not recommended that
patients be disconnected from PEEP for PAWP measurements. Although this
will probably eliminate any confounding effects of PEEP, this may further
compromise the patients respiratory status with loss of alveolar units, and
there may also be uid shifts with the removal of PEEP. It is unlikely that this
maneuver provides a clinical benet in management. In general, it is wise
to follow both the hemodynamic and clinical changes that occur with changes
in PCWP and not to pay too much signicance to an isolated PCWP value
(96 100).
Thermodilution CO
The PA catheter is equipped with a thermistor at the tip, allowing CO to be cal-

culated using differences in uid temperature in an indicator-dilution technique
(90). A bolus of cold (colder than body temperature) uid can be detected by
the thermistor, and the rate of blood ow or CO would be inversely proportional
to the change in the amount of cold uid detected over time. This thermodilution
CO measurement produces characteristic curves, with poor CO reected as long,
drawn out tracings. In practice, 5 to 10 mL of cold D5W or normal saline is
injected into the RA via the proximal port of the PA catheter within two
seconds. The uid can be room temperature or have been placed in ice. If 5 mL
of uid is used, this must be noted for analysis as the CO calculations will
need to be adjusted for the smaller uid bolus. It can be a source of error if
this adjustment to the bedside monitoring software program is not made. The
solution is rapidly injected, preferably at end-expiration to minimize any varia-
bility because of the respiratory cycle (101). Three separate measurements
should be obtained at one-minute intervals and reported as an average. The
measurements should be consistent within a 10% range, and the curves should
be inspected to conrm reproducibility and reliability of readings. Falsely
low measurements may be obtained in the presence of tricuspid regurgitation
or prolonged injection time exceeding four seconds (102). Falsely high measure-
ments may be noted in the presence of intracardiac shunts and low output
states (103).
Cardiac Output
The PA catheter is equipped with a thermistor at the tip, which allows CO to be
calculated using the Ficks principle (91). In practice, 5 to 10 mL of cold D5W or
saline is injected into the RA via the proximal port of the PA catheter. The
average of three separate measurements should be obtained at one-minute inter-
vals. The solution is rapidly injected, preferably at end-expiration.
Mixed Venous Oxygen Saturation

SvO2 reects the balance between tissue oxygen delivery (DO2) and oxygen con-
sumption or uptake (VO2). Hence, it is an indicator of systemic oxygen utiliz-
ation. It may provide some insight in the management of critically ill patients
as changes may signal shifts in this balance, specically problems with oxygen
delivery or increases in oxygen consumption, which in turn may prompt
further interventions. This relationship is further illustrated using the Fick prin-
ciple in describing CO as calculated by the following equation:
_ 2 (mL= min )
VO
CO mixed venous O2 content (Cv O2 )
arterial O2 content (Ca O2 )
or
_ 2
VO
CO Cv O
Ca O2 2
Rearranging for CvO2
_ 2
VO
Cv O2 Ca O2
CO
The oxygen content is the sum of both the oxygen bound to hemoglobin and that
in solution, therefore
Ca O2 (Hb Sa O2 1:34) Pa O2 0:003

Cv O2 (Hb Sv O2 1:34) Pv O2 0:003
SaO2 and SvO2 represent arterial oxygen saturations and mixed venous oxygen
saturations, respectively. The amount of oxygen (in milliliters) that can bind to
1 g of hemoglobin (Hb) is 1.34. The solubility of oxygen in the blood is 0.003.
Under normal conditions this is negligible, therefore
Cv O2 Hb Sv O2 1:34
Ca O2 Hb Sa O2 1:34
Substituting these simplications with SvO2 and SaO2 into the Fick
equation gives
_ 2
VO
Sv O2 Sa O2
CO Hb 1:34
It becomes easy to see that the four factors that can inuence SvO2 are arterial
oxygen saturation, oxygen consumption, CO, and hemoglobin level. Therefore
changes in SvO2 may reect changes in any or all of these parameters.
The normal SvO2 is 75% (range: 60 80%). When less than 60%, this
suggests anaerobic metabolism (e.g., lactic acidosis), and values below 40%
are associated with severely insufcient tissue oxygenation due to shock.
When the hemoglobin concentration, oxygen consumption, and arterial oxygen
saturation are constant, changes in SvO2 are due to changes in CO. Continuous
mixed venous oxygen saturation is possible with PA catheters by the use of ber-
optic reectance spectrophotometry. The continuous mixed oxygen saturation
reading from the catheter is therefore a useful monitor of CO in hemodynamically
unstable patients. Increases in SvO2 would indicate improvement in CO or
tissue perfusion, and conversely, a decline would represent further deterioration
of hemodynamics and tissue compromise. These changes may be evident prior to
other changes in routinely monitored variables (104). In addition, there may be
metabolic derangements, such as lactic acidosis, that may be persist despite
normal hemodynamics that can be identied with SvO2 monitoring. Despite the
potential benet of continuous monitoring of SvO2, there remains considerable
uncertainty of its role in the management of critically ill patients and advantage
over other available monitoring devices (105). The mixed venous oxygen satur-
ation is affected in different clinical conditions as shown in Table 12.
F. Use of the PA Catheter in Specic Clinical Scenarios

Myocardial Infarction
The PA catheter was initially used in patients with MI complicated by shock.
Measurements of CO, preload by obtaining the PAWP, and calculation of
systemic vascular resistance allow management of uids and inotropic or vasopr-
essor medications (106,107). It can also be used for diagnosing and managing
complications of MI, such as congestive heart failure, mitral regurgitation,
ventricular septal defect, right ventricular infarction, and cardiac tamponade.
In 1998, the American College of Cardiology (ACC) published a consensus
report stating the indications and recommendations for the use of bedside PA
catheter or right-side catheterization in patients with cardiac disease (79). For
acute MI, there are six conditions in which PA catheterization is warranted:
1. To differentiate between cardiogenic and hypovolemic shock.
2. For the management of cardiogenic shock with pharmacologic and
mechanical support in patients with and without coronary reperfusion
therapy.
Table 12 Changes in Mixed Venous Oxygen Saturation in Various Clinical Conditions
Mechanism Example
Increased SvO2
Increased arterial O2 supply by increased Sepsis
cardiac output
Decreased tissue demand by decreased VO2 Anesthesia, coma, hypothermia,
hypothyroid, cirrhosis, left to right
cardiac shunt
Decreased SvO2
Decreased cardiac output Hypovolemic shock, acute myocardial
infarction
Decreased arterial oxygen (CaO2) Hypoxemia
Decreased Hemoglobin Anemia, hemorrhage
Increased tissue demand by increased VO2 Fever, seizures, burns, hyperthyroid,
pain, shivering (34 37)
3. Short-term guidance of pharmacologic and mechanical management

(intra-aortic balloon pump) of acute mitral regurgitation (with or
without disruption of the mitral valve) before surgical correction.
4. To diagnose and manage left to right shunt and ventricular septal
rupture before surgical correction.
5. Guidance of therapy for right ventricular infarction with hypotension.
6. Guidance of therapy for acute pulmonary edema not responding to
treatment with diuretics, nitroglycerin, and other vasodilator agents
and low doses of inotropic medications (79).
Heart Failure
There are six conditions in patients with heart failure for which PA catheteriza-
tion is warranted:
1. To differentiate pulmonary edema caused by heart failure or by pul-
monary disease or both.
2. To distinguish and guide therapy between cardiogenic and non-
cardiogenic shock.
3. To guide therapy in patients with concomitant manifestations of
forward (hypotension, oliguria, and azotemia) and backward
(dyspnea and hypoxemia) heart failure.
4. To diagnose pericardial tamponade when clinical assessment is incon-
clusive and echocardiography is unequivocal.
5. To guide perioperative management in selected patients with decom-
pensated heart failure undergoing intermediate or high-risk non-
cardiac surgery.
6. To diagnose pulmonary vasoconstriction and determine its reversibil-

ity in patients being considered for heart transplantation (79).
Perioperative Use in Cardiac Surgery

There are four conditions in the management of cardiac surgery patients where
PA catheterization is warranted:
1. To differentiate between causes of low CO (hypovolemia vs.
ventricular dysfunction) when clinical and echocardiography is
inconclusive.
2. To distinguish between right and left ventricular dysfunction and diag-
nose pericardial tamponade when clinical and echocardiography is
equivocal.
3. To guide management of severely low CO syndromes.
4. To diagnose and guide therapy for pulmonary hypertension in patients
with systemic hypotension and evidence of inadequate organ perfusion
(79).
Pulmonary Hypertension
There are four conditions for which PA catheterization is warranted in the man-
agement of pulmonary hypertension according to the ACC expert consensus:
1. To exclude postcapillary (elevated PAWP) causes of pulmonary
hypertension.
2. To assess the diagnosis and severity of precapillary (normal PAWP)
pulmonary hypertension.
3. To determine the safety and efcacy of vasodilator therapy based on
acute hemodynamic response.
4. To evaluate different hemodynamic variables before lung transplan-
tation (79).
Although there were no conditions in which right heart catheterization was
not warranted, there were conditions in which reasonable differences in opinions
between the ACC experts existed. These conditions are the evaluation of the
long-term efcacy of vasodilator therapy, in particular prostacycline, and the
exclusion of signicant left to right or right to left intracardiac shunts.
Acute Mitral Regurgitation
Acute mitral regurgitation can occur as a complication of acute MI. PA catheter-

ization is very useful when echocardiography with Doppler is inconclusive or
unavailable. Elevated left atrial pressure and PAWP are noted and tall peaked
v waves are seen in the PAWP tracing (Fig. 10). A bid PA waveform com-
posed of the PA systolic wave and the v wave can also be seen. A large v
wave is not diagnostic of acute mitral regurgitation, as it may occur whenever
Figure 10 The tracing in a patient with mitral regurgitation. The characteristic tall v
waves are demonstrated.
the LA is distended and non-compliant due to left ventricular failure from any
cause, that is, dilated cardiomyopathy (108,109).
Tricuspid Regurgitation
Tricuspid regurgitation usually occurs in the setting of pulmonary hypertension
and right ventricular dilatation. It is associated with elevated right atrial and
right ventricular end-diastolic pressures. It usually represents chronic valvular
insufciency. A prominent early right atrial v wave with a steep y descent
is seen. In severe tricuspid regurgitation, the tracings of the right atrial and
right ventricular pressures may look similar (110).
Cardiac Tamponade
Echocardiography is usually diagnostic for pericardial effusion, and it is widely

used to guide the needle for therapeutic measures. In circumstances where the
echocardiogram is questionable, PA catheterization may be diagnostic. The
characteristic pattern of acute cardiac tamponade is elevation and equalization
of right- and left-sided diastolic lling pressures (RA, RV, PA diastolic, and
PAWP). A prominent systolic x descent and a blunted y descent on the
right atrial pressure tracing are seen. The right atrial pressure declines in inspi-
ration in contrast to right ventricular infarction and constrictive pericarditis (111).
Right Ventricular Infarction
RV infarction typically occurs in the setting of inferior myocardial infarction.

The right atrial pressure is elevated (usually .10 mmHg), and often in the
right atrial waveform, prominent x and y descents are seen. Sometimes the
y descent may exceed the x descent due to a dilated noncompliant RV con-
ned by a non-distensible pericardium (112).
Constrictive Pericarditis
Chronic brous thickening of the pericardium can lead to constrictive pericardi-
tis. Causes of constrictive pericarditis include tuberculosis, malignancy, post-
chest radiotherapy, and idiopathic causes. Acute or subacute presentations are
seen after cardiac surgery. On examination, the jugular venous pressure (JVP)
is commonly raised and may rise or fail to fall with inspiration (Kussmals
sign), which distinguishes this condition from cardiac tamponade, in which
there is a decline in the right atrial pressure (JVP) with inspiration. The JVP
also shows a prominent y descent. At catheterization, low CO is seen with
prominent x and y descents on the atrial waveform, producing m or w
waveform appearance. The waveforms have a characteristic diastolic dip and
plateau pattern (or square root sign) in the ventricular waveform, reecting
abrupt termination of the ventricular lling due to the rigid pericardium. Equili-
bration of right ventricular and left ventricular end diastolic pressures
(,5 mmHg) is often seen, although this can occur in patients with heart
failure or acute volume overload. Unlike pericardial tamponade, the PCWP
may be as high as 20 to 25 mmHg and similar in appearance to the RA waveform.
Acute Respiratory Distress Syndrome

The use of the PA catheter is ARDS is two-fold. These patients may have a chest
roentgenogram that suggests cardiogenic pulmonary edema, or it may be difcult
to discern whether bilateral diffuse airspace disease is due to congestion or capil-
lary leak. This uncertainty may only be claried after placement of a PA catheter.
The denition of ARDS includes PAWP readings ,18 mm Hg, thereby exclud-
ing a cardiogenic cause of respiratory failure (113). Judicious uid management
is also an integral part of the management of patients with severe respiratory
failure and ARDS. There may be a mortality benet in patients who receive
less overall uid or are treated with a conservative uid management strategy
(114). This may be best managed using a PA catheter. The use of PA catheteriza-
tion is being intensively investigated in the setting of acute lung injury (ALI) and
ARDS (115). A prospective, randomized, multi-center ARDS Network trial com-
paring PA catheter versus central venous catheter use in the management of ALI
and ARDS as well as uid management strategies entitled Fluid and Catheter
Treatment Trial (FACTT) is currently underway. This trial represents two
studies designed to compare the use of the PA catheter to a central venous line
and also conservative uid management with liberal uid management in these
patients. It was initiated in July 2000, then suspended in July 2002 because of
concerns by the Ofce for Human Research Protection (OHRP), but has since
resumed enrollment.
Shock States
In clinical practice, specic conditions are associated with characteristic ndings
in the PAWP, CO, systemic, and pulmonary vascular resistance. This can be
useful in conrming suspected diagnoses and provide guidance in management.

Cardiogenic and vasodilatory or septic shock are the two most frequently encoun-
tered shock states, and both have characteristic ndings. In cardiogenic shock
associated with a myocardial infarction, the primary pathology is that of poor
pump function. This is seen as an elevated PAWP, decreased CO and cardiac
index, and elevated systemic vascular resistance. Elevated pulmonary artery
pressures may be seen as a result of markedly elevated PAWP and a poorly com-
pliant LV. In vasodilatory or septic shock, the PAWP is typically in the normal
range or even low, with an increased CO and cardiac index and decreased sys-
temic vascular resistance. Of course these are generalities, but they may be
helpful in distinguishing these states from other similar conditions.
Acute Pulmonary Embolism

Massive pulmonary embolism may present as cardiogenic shock. At PA catheter-
ization, patients may have decreased CO and cardiac index, but with elevated
pulmonary artery diastolic pressures and normal PAWP, which distinguishes
massive pulmonary embolism from cardiogenic shock. As much as one-third
of a previously normal pulmonary vascular bed may be obstructed, but the
mean PA pressure rarely exceeds 50 mmHg unless there is a chronic component
of pulmonary hypertension (116). Pulmonary vascular resistance is elevated, and
the PA diastolic pressure is signicantly higher than the mean PAWP (by more
than 5 mmHg). The a and v waves of the PAWP tracing may disappear
because of the elevated pulmonary vascular resistance, and the catheter tip
may not be able to sense the pressure generated from the LA (117,118). CO
may be decreased if more than 60% of the pulmonary vascular bed is being
obstructed (116 118).
Pre- and Post-Operative Monitoring by Using the PA Catheter

This is one of the most controversial areas in the use of the PA catheter. It has
never been clear that the placement of a PA catheter will improve the manage-
ment of high-risk patients undergoing surgery. Sandham et al., in the Canadian
Critical Care Clinical Trials Group, examined the use of the PA catheter in
high-risk surgical patients in a randomized, controlled trial. Of 3803 eligible
patients, 52.4% (1994 patients) underwent randomization, comparing goal-
directed therapy with and without a PA catheter (88). The subjects were high-
risk according to the American Society of Anesthesiologists (ASA) class III or
IV, aged 60 or older, who were scheduled for urgent or elective major surgery,
followed by ICU stay. Each group had 997 patients; 78 patients who had a PA
catheter died in the hospital, compared with 77 patients without a PA catheter
(7.8% vs. 7.7%). There was no signicant difference in mortality rate between
the two groups. But there is a signicant increase in the incidence of
pulmonary embolism in the catheter group (8 events vs. 0 events). More patients
in the catheter group received inotropic agents, vasodilators, antihypertensive
medications, blood transfusions, and colloids. They concluded that routine

insertion of the PA catheter perioperatively in high-risk surgical patients is not
warranted (88).
The ASA, in 2003, published their practice guidelines for PA catheteriza-
tion (119). They reviewed their old guidelines from 1993 and incorporated infor-
mation from the American College of Physicians, ACC/American Heart
Association, American College of Chest Physicians, American Thoracic
Society, and the Society of Critical Care Medicine. They concluded that the
appropriateness of PA catheterization depends on the combination of three risk
factors: the health status of the patient, the specic type of surgery, and the prac-
tice settings. Patients are stratied according to the ASA: classes 1 and 2 are
patients who are unlikely to have organ dysfunction, class 3 are those with mod-
erate risk, and classes 4 and 5 are those patients who have hemodynamic instabil-
ity and a greater chance of organ dysfunction or death. Surgery was classied as
low-, moderate-, or high-risk procedures, based on expected morbidity and mor-
tality. Critical features of the practice setting included catheter insertion skills,
training, and experience of nursing staff and doctors in interpreting data; techni-
cal support; ancillary services; and availability of specialists and equipment
(120). In general, PA catheterization is appropriate in moderate to high-risk
patients undergoing moderate to high-risk surgery in a moderate to high-risk
setting. Patients with moderate to high-risk stratication (ASA 3 5, Table 13)
are expected to have a moderate to high predicatability of uid or hemodynamic
changes, increasing the risk of morbidity and mortality. Moderate- to high-risk
surgical procedures are those with an increased risk of complications from hemo-
dynamic changes. Examples include surgical patients undergoing elective or
urgent major surgery such as coronary artery bypass graft surgery, lung resection,
Table 13 Patient ASA Risk Stratication
ASA Patient stratication Examples
1 Low Healthy patient

2 Low Mild systemic disease (e.g., controlled
hypertension, diabetes)
3 Moderate Severe systemic illness, not incapacitated
(e.g., morbid obesity, COPD, angina post,
myocardial infarction)
4 High Incapacitating systemic illness, constant threat to life
(e.g., unstable angina, congestive heart failure,
hepatic, or renal failure)
5 High Moribund, not expected to survive for 24 hr
with or without procedure
Abbreviations: ASA, The American Society of Anesthesiologists; COPD, chronic obstructive

pulmonary disease.
or trauma. These patients usually will be followed by care in the ICU and with the
information provided by the PA catheter, intensivists are be able to provide
earlier goal-directed therapy. In low-risk patients undergoing low-risk surgery,
e.g., knee replacement, the PA catheter is not necessary. Hence, the use of the
PA catheter depends on clinician knowledge, patient classication, and the sever-
ity of the surgery.
In summary, there is no absolute requirement for the use of the PA catheter
in critically ill patients. Many studies over the last four decades have been done,
but none have provided conclusive support or against its routine use. The above
examples are suggestions and recommendations for the use of the catheter in
specic clinical situations. Clinician knowledge about the insertion and compli-
cations of the catheter, appropriate interpretation of the data, and appropriate
application of the data are crucial for the safe use of this device, and this expertise
may be lacking by some. Current training, credentialing, and quality improve-
ment issues need to be evaluated at each institution. Clinicians should continue
to carefully weigh the risks and benets of the PA catheterization in management
of individual patients before its insertion.
G. Complications of PA Catheters
Complications of PA Catheterization are listed in Table 14. These have been sep-
arated into complications related to placement of the catheter and complications
related to the catheter itself. Dysrhythmias, such as atrial brillation and utter,
premature ventricular tachycardia, or brillation, may occur during catheter
insertion or withdrawal but usually resolve spontaneously after the catheter is
advanced or withdrawn through the right heart chambers (121). Dysrhythmias
Table 14 Complications of PA Catheterization
Complication related to central venous cannulation

(refer to section on Central Venous Catheters)
Complications related to insertion of the PA catheter
Due to catheterization procedure
Dysrhythmias
Right bundle branch block
Complete heart block
Catheter knotting
Due to catheter presence
Thrombosis/embolism
Pulmonary infarction
PA rupture/perforation
Balloon rupture
Infection
Endocardial damage
Abbreviation: PA, pulmonary artery.

seldom need to be treated with anti-arrhythmic drugs or cardioversion. Risk

factors for ventricular tachycardia include hypocalcemia, myocardial ischemia
or infarction, hypotension, hypokalemia, hypocalcemia, hypoxemia, and acido-
sis. Catheter advancement can injure the right bundle branch, with a resultant
conduction block. In patients with a pre-existing left bundle branch block, this
may precipitate a complete heart block (122 128). Although acknowledged as
a potential complication, it rarely occurs and should not be considered a contra-
indication to PA catheter placement (127).
Complications of catheter residence inside pulmonary artery include
venous thrombosis, thrombophlebitis, pulmonary embolism, and infarction
(129 134). During the initial years of PA catheterization, pulmonary infarction
was thought to be one of the most common serious complications. Foote et al.
found a 7.2% incidence of infarction in 1974 (129). In 1983, Boyd et al. found
1.3% of 500 consecutive insertions (56). Sandham et al. (88) in 2003 stated that
they had eight events of pulmonary embolism in 78 of 997 patients in whom a
PA catheterization was used in the randomization of 1994 patients. Pulmonary
infarction can be caused by the persistent wedging of the catheter. Distal
migration of the catheter is common (134) and is most likely to occur within
12 to 24 hours of insertion. The catheter should be maintained in such a position
that nearly the full recommended balloon ination is required to produce a
wedge. Chest roentgenograms should be done daily to assess the position of
the tip of the catheter. Since the introduction of heparin-bonded catheters,
thromboembolism has been reduced, but other complications such as heparin-
induced thrombocytopenia and thrombosis remain (135). Fatal pulmonary
artery rupture is fortunately rare (136). The major risk factors for rupture or per-
foration are pulmonary artery hypertension, advanced age of the patient,
hypothermia, and anticoagulant therapy. Patients with PA perforation can
present with massive hemoptysis. Emergency management includes immediate
wedge tamponade followed by arteriogram, bronchoscopy, intubation of the
unaffected lung, and consideration of emergency lobectomy (136,137,138).
Balloon rupture should be suspected when attempted ination is not met with
the usual feeling of resistance (93,94). If the balloon ruptures during catheter
insertion, the protective cushioning effect of the balloon will be lost, and it
may cause endocardial damage and more arrhythymic complications. Knotting
of the catheter can occur when the cardiac chambers are dilated or the catheter
is repeatedly withdrawn and re-advanced (128). This can be avoided if care is
taken not to advance the catheter signicantly beyond the distances at which
entrance to the ventricle or PA would normally be anticipated. On some
occasions, it is advised to perform the PA catheterization under the guidance
of uoroscopy, especially when the RV is dilated.
Infection is a potentially serious complication of PA catheter insertion and
residence inside the pulmonary artery (139 142). It may be difcult to determine
whether the source of infection is from the catheter or from the patients pre-
existing or concomitant infection. Skin ora is a common source of infection.
The incidence of catheter-related septicemia increases with prolonged duration of

catheter placement (.3 4 days) (65). Right-sided septic endocarditis has been
reported (141). Incidence of catheter colonization or contamination from skin
ora varies from 5% to 20%, depending on the duration of catheter residence
and the criteria used to dene colonization (139,143). Aseptic techniques
should always be performed during catheter insertion and maintenance. A
sterile protective sleeve, antibiotic-bonded catheter, and empiric changing of
catheters over a guidewire have been used to reduce the risk of infection
(144 146).
The ASA Task Force concluded that serious complications (PA rupture,
serious ventricular dysrhythmias, endocardial lesions) occur in 0.1% to 0.5%
of PA catheter-monitored surgical patients (119). However, physicians dealing
with other patient populations have different clinical experiences. In a survey of car-
diologists and internists, the estimated probability of severe morbidity or death from
PA catheterization was noted to be 2% to 5% and 0.5% to 1%, respectively (132).
H. Controversy of PA Catheter Use

As PA catheterization is a monitoring device, its clinical utility depends on the
appropriate indication, health status of the patient, skill and technical support of
the equipment, interpretation of data, and availability of effective treatments.
Iberti et al. evaluated 496 physicians knowledge of the PA catheter in the U.S.A.
and Canada (147). The mean test scores varied independently by training, frequency
of insertion and interpretation of data, and whether the respondents hospital was a
primary medical school afliate. The conclusion was that physician understanding
of the use of the PA catheter is extremely variable and frighteningly low (148). The
main issue in the use of the PA is whether it impacts the mortality and safety of
patients. If follows that physicians who use this device should be fully knowledge-
able about the intricacies of the catheter. The authors believe that credentialing
requirements for PA catheter use should not only include a demonstration of inser-
tion skill, but competence in the interpretation and application of the data. They pro-
posed that 10 insertions would be required to attain the initial mastery and
subsequently ve insertions per year would be required to maintain competence.
This has not yet been adopted in routine credentialing.
Some of the controversy with the use of the PA catheter stems from favor-
able outcomes noted when it was used to guide therapy to reach normal or supra-
normal levels. Shoemaker et al. published several papers between 1985 and 1991,
using the PA catheter in the perioperative period to evaluate and guide therapy in
high-risk surgical patients (80,81,149,150). They optimized the CO and oxygen
delivery to supranormal levels with data obtained from the PA catheter. Schultz
et al. were able to reproduce similar favorable results in patients with hip
fractures (82). In the early treatment of severe sepsis and septic shock in the
emergency room, similar success was published by Rivers et al. in a early
goal-directed approach to management (83). A higher CO was associated with
lower mortality in patients in septic shock with the rationale that better tissue per-
fusion was obtained (81,84). In a meta-analysis of studies published between
1970 and 1996, there was decreased morbidity noted in patients with PA catheters
compared with those without, calculating a relative risk ratio of 0.78 (0.65 0.94)
for the incidence of organ failure (151).
However, there are several investigations that are contrary to this experience.
Gore et al. reported that the use of the PA catheter was associated with a higher
fatality rate in patients with congestive heart failure, 44.8% compared with
25.3% (P , 0.001) (85). In patients with hypotension, the case fatality rate was
48.3% in those with PA catheterization, compared with 32.2% for those without
(P , 0.001). Length of hospital stay was signicantly longer in patients receiving
a PA catheter. A study from Israel reported a similar lack of benet in an observa-
tional cohort of 5841 patients with acute myocardial infarction (86). In 1996,
Connors et al. (152) performed a post hoc analysis on 5735 critically ill patients
in nine disease categories admitted to the ICUs in ve medical centers as part of
the SUPPORT study. This was an observational study, and as part of their retro-
spective analysis, they developed a propensity score to adjust for disease severity
and other demographic factors. They identied two groups of patients with similar
propensity scores. Patients who underwent PA catheterization had higher mortality
rates, longer length of stay in the hospital, and subsequently higher hospital costs
than those who did not (87). This suggested that the PA catheter was associated
with increased morbidity and mortality and spawned considerable debate about
the use of the device in the management of critically ill patients (153). It should
be noted that the device was never subject to a prospective evaluation of efcacy
prior to its introduction into clinical use.
Subsequent, prospective randomized trials have not identied an increase in
mortality associated with the use of the PA catheter. The randomized, controlled
trial of PA catheters in high-risk surgical patients by Sandham et al. (88) has been
previously outlined. There was no signicant difference in mortality rate between
the two groups, but there is a signicant increase in the incidence of pulmonary
embolism in the catheter group (8 events vs. 0 events). The authors found no
benet to routine insertion of the PA catheter perioperatively in high-risk surgical
patients and concluded it was not warranted (88). Richard et al. randomized 676
with shock, ARDS, or both to early use (within 24 hours) of a PA catheter in man-
agement. The placement of a PA catheter did not signicantly affect mortality or
morbidity as dened by organ system failure, need for vasoactive agents, duration
of mechanical ventilation, ICU, or hospital stay. On day 28, the mortality rate in
the PA catheter group was 59.4%, and the control group was 61.0%. They con-
cluded that there was no observed benet with the insertion of a PA catheter in
these patients (89).
I. Ongoing Prospective, Randomized Trials

Subsequent trials addressing the utility of the PA catheter in patient management
are in progress. These include the previously described FACTT trial and a U.K.
sponsored PAC-Man trial (154). The latter trial has now been completed. It
enrolled 1041 patients from 65 United Kingdom intensive care units with a
broad range of diagnoses. The focus was on critically ill patients with a perceived
need for PA catheter placement by their physician. The study demonstrated no
difference in hospital mortality in critically ill patients managed with or
without pulmonary artery catheters. There were more complications related to
catheter insertion and changes in management based on results from the catheter
were modest. It is unclear whether PA catheter use improved outcome. Another
trial evaluated the utility of the PA catheter in the management of patients with
severe congestive heart failure (New York Heart Association classes III and IV)
(155). This is the Evaluation Study of Congestive Heart Failure and Pulmonary
Artery Catheterization Effectiveness (ESCAPE) trial, which has completed
enrollment and is currently undergoing analysis. The results from these three
major studies should further dene the role of the PA catheter in the management
of critically ill patients.
J. General Principles in PA Catheter Use
Until the results of these future studies are available, clinicians using the PA cath-
eters for hemodynamic monitoring should carefully assess the risk-to-benet
ratio on an individual patient basis. The indications, insertion techniques, equip-
ment, patients health status, and interpretation of data obtained from the catheter
should be taken into consideration before the PA catheter is used (156). PA cathe-
terization should only be used as a complementary device to diagnose and guide
therapy after clinical evaluation and treatment have been exercised. It is import-
ant to recognize that diagnostic tests and monitoring devices do not determine
clinical outcomes, but that outcomes are inuenced by therapy directed by infor-
mation obtained with the catheter. Minimizing the complications associated with
its placement and limiting the duration of placement should be the major guiding
principles in its use. The PA catheter should be removed as soon as information
derived from its placement is no longer clinically useful. In some cases, this may
occur within a few hours, and certainly placement should not exceed three to four
days in any case. Although there is great hope that pending research studies will
further dene the role of the PA catheter in patient management, it is important to
recognize that there will remain areas of uncertainty in its use. Even with these
trials, there may still be signicant debate surrounding the use of the PA catheter.
It is probably safe to conclude that there will always be a role for this device in the
management of critically ill patients. Therefore, it behooves all involved to be
cognizant of its potential and limitations.
V. Chest Tube Thoracostomy

A. Background
Chest tubes are used in many clinical settings, and their placement is one of the
most important skills to master in the management of life-threatening conditions.
Chest tube placement and management is important not only in the ICU but also
in Advanced Trauma and Life Support (157).
B. Anatomy
The parietal pleura lines the thoracic wall and covers the superior surface of the
diaphragm. It can be divided into four sections: the costal pleura, cervical
pleura, mediastinal pleura, and the diaphragmatic pleura. The parietal pleura is
supplied with blood from the intercostal arteries and branches of the internal thor-
acic artery. It is innervated by the intercostal nerves and the phrenic nerve. Pain
from these nerve bers may be referred to the thoracic and abdominal walls,
neck, and shoulder (phrenic nerve). At the hilum and pulmonary ligament, the
pleura becomes the visceral pleura as it is reected and adheres to the surface
of the lung. The visceral pleura is mainly perfused by the systemic circulation.
The pleural layers are in close opposition to each other, with the potential space
between them containing a thin layer of uid. This uid acts as a lubricant, allow-
ing the pleural membranes to slide easily against each other. Under normal
circumstances, there is a negative intrapleural pressure of 2 to 5 cm of water.
Disruption in visceral or parietal pleura results in pneumothorax or hemothorax.
Fluid can accumulate in the pleural space because of the changes in hydrostatic
or oncotic pressures due to impaired lymphatic drainage or inammatory
diseases.
C. Indications
Chest tubes are inserted to evacuate air or uid. The indications for chest tube
placement are listed in Table 15 and also discussed in the following sections.
Pneumothorax
Pneumothorax is the most common indication for chest tube placement. Patients
may present with symptoms of dyspnea, tachypnea, and chest pain. Clinical
examination reveals diminished breath sounds and hyper resonance to percus-
sion. The diagnosis is usually conrmed by chest X-ray. Inspiratory and expira-
tory lms are helpful in equivocal cases as is chest computed tomographic
scanning. Spontaneous or iatrogenic pneumothorax of less than 25% volume in
Table 15 Indications for Chest Tube Placement
Pneumothorax Spontaneous; traumatic; tension; iatrogenic; bronchopleural stula

Hemothorax Traumatic; iatrogenic; infections; ruptured thoracic aortic aneurysm
Empyema Parapneumonic; post traumatic; intraabdominal
Pleural effusion Transudate; exudates
Chylothorax
Pleurodesis
asymptomatic or minimally symptomatic patients without signicant underlying

disease can be observed with serial chest X-rays. Urgent chest tube placement is
necessary in patients with severe distress and expanding amounts of intrapleural
air, signicant underlying lung disease, and mechanically ventilation. Tension
pneumothorax requires immediate decompression with a 14- or 16-gauge cath-
eter in the second intercostal space, mid clavicular line, while preparations for
chest tube placement are being made.
Hemothorax
Hemothorax refers to a collection of blood in the pleural cavity. The pleural uid
hematocrit is greater than or equal to 50% of the blood hematocrit. Etiology of
hemothorax can be spontaneous, iatrogenic, or traumatic in origin. Spontaneous
hemothorax may result from primary or metastatic malignancy in the lung or
pleura, pulmonary infarcts, arteriovenous malformations, and necrotizing pul-
monary infections. Common traumatic and iatrogenic causes include rib frac-
tures, attempted thoracentesis, or chest tube placement resulting in damage to
intercostal or internal mammary arteries. Large bore drainage tubes in this
setting help in the assessment of the degree of blood loss. Evidence of persistent
bleeding following trauma, with hemodynamic instability, is an indication for
open thoracotomy. Late-stage complications of unresolved hemothorax include
empyema and brothorax.
Empyema and Parapneumonic Effusions

Pleural uid secondary to bacterial process in the thoracic cavity is classied as
parapneumonic effusions, complicated parapneumonic effusions, and empyema
(158). A parapneumonic effusion is any pleural effusion secondary to a bacterial
process, usually treated with antibiotics with or without thoracentesis. A compli-
cated parapneumoic effusion is a parapneumonic effusion that requires chest tube
placement for resolution as it is unlikely to clear with antibiotics alone. Empyema
is pus in the pleural space. Indications for chest tube placement are a pleural uid
with a pH , 7.20, a glucose level less than 40 mg/dL and a lactate dehydroge-
nase (LDH) level greater than 1000 U/L, and positive Gram stain or positive
pleural uid cultures. Simple complicated parapneumonic effusions can be
managed with relatively small chest tubes placed percutaneously. Complex com-
plicated parapneumonic effusions are multiloculated. CT scans or ultrasound
(Fig. 11) are used to help localize the loculated uid (159). These require
larger bore chest tubes and occasionally the addition of a thrombolytic agent,
usually streptokinase, urokinase, or alteplase (160,161). Large bore chest tubes
(28F or greater) are required for simple empyemas with free-owing uid or in
a single loculus. Effusions that do not resolve with chest tube placement and
thrombolytics require surgical decoritication (162) or open ap drainage.
Figure 11 Transthoracic sonogram demonstrating a loculated pleural effusion. Fibrin

strands are clearly seen.
Pleural Effusion
Fluid can accumulate in the pleural space for a number of reasons. These include
changes in hydrostatic and oncotic pressures, increased permeability of the
microvasculature by inammatory mediators, and impaired lymphatic drainage.
Thoracentesis is usually done to determine whether the effusion is transudative or
exudative in origin. Transudates result from imbalances in hydrostatic and
oncotic pressures. Causes of transudative effusions include cirrhosis, congestive
cardiac failure, and nephrotic syndrome. Management of transudative effusions is
aimed at identifying and treating the underlying cause. Chest tubes are usually
not indicated in the management of transudates except in selected cases of
massive effusions. Exudates are formed by the disruption in the integrity of the
endothelial membranes of pleural capillaries and venules. Inammatory pro-
cesses result in an increased permeability of these membranes with subsequent
protein leak. Impaired lymphatic drainage decreases the removal of protein in
the pleural space. Fluid from the peritoneum can also cause exudates. Infection,
malignancy, and inammatory diseases are all causes of exudative pleural
effusions.
Chemical analysis of the pleural uid distinguishes transudative effusions
from exudates. Measurement of both serum and pleural uid protein and LDH
is the most practical method of distinguishing the two. Lights criteria (163),
shown in Table 16, are used to distinguish exudative pleural effusions. If one
of these three criteria is present, the uid is almost always an exudate. If none
of these three are present, it is virtually always a transudate. If paired testing is
not an option, any one of the three tests on a single pleural uid sample shown
in Table 16 will identify an exudate (164).
Table 16 Tests Used to Distinguish Between Transudative and Exudative Pleural

Effusions
Lights criteria for dening an exudates

Pleural uid protein/serum protein ratio .0.5
Pleural uid LDH/serum LDH ratio .0.6
Pleural uid LDH .2/3 of the upper limit of normal of the serum LDH
Single pleural uid testing in dening an exdudate
Pleural uid protein .3.0 g/dL (3 g/L)
Pleural uid cholesterol .45 mg/dL (1.16 mmol/L)
Pleural uid LDH .60% of the upper limit of normal serum value
Abbreviation: LDH, lactate dehydrogenase.

Source: From Refs. 163 and 164.
Decubitus lms help determine if the pleural effusion is free owing. Safe
thoracentesis requires at least 1 cm of visible uid on the decubitus lm. Ultra-
sound, as already mentioned, can help determine the most appropriate site for
thoracentesis or chest tube placement. In malignant effusions, chest tubes are
placed for both symptomatic improvement and, often, subsequent instillation
of sclerosing agents for pleurodesis. Sclerosing agents, such as talc, tetracycline
derivatives, and bleomycin, are used once the lung has fully re-expanded and
pleural uid drained.
Chylothorax
Chyle is lymphatic uid of intestinal origin carried in the thoracic duct across the
chest. It has a high content of triglycerides, typically over 110 mg/dL, and lym-
phocytes. The thoracic duct enters the mediastinum on the right side of the chest
and crosses over to the left at the level of the fth thoracic vertebrae in most
adults. Between 1.5 and 2.4 L of uid ow through the thoracic duct in a day.
The causes of chylothrorax are divided into non-traumatic and traumatic
(Table 17). Lymphoma is the most common non-traumatic cause; surgical pro-
cedures account for most traumatic cases. Treatment in non-traumatic cases is
aimed at the underlying cause. Failure of chemotherapy or radiotherapy or the
development of symptoms due to an increasing pleural effusion require chest
tube drainage. Symptomatic effusions of idiopathic origin are treated with
chest tube drainage and either decreased oral intake or parental nutrition with
or without medium chain triglycerides. Most cases resolve within two to three
weeks. Treatment options for resistant cases include pleurodesis, pleurectomy,
and ligation of the thoracic duct at the hiatus.
Prolonged and copious drainage from chest drains can result in weight loss,
progressive hypoproteinemia, and immunosuppression. In these cases, earlier
surgical intervention is warranted.
Table 17 Etiologies of Chylothorax
Causes
Non-traumatic
Malignant Lymphomatous, non-Lymphomatous (e.g., pulmonary and
mediastinal malignancies)
Non-malignant Idiopathic, others (e.g., cirrhosis, tuberculous, venous thrombosis)
Traumatic
Surgical Cardiovascular, aortic lobectomy, pneumonectomy etc.
Nonsurgical Penetrating, nonpenetrating trauma to the neck,
thorax, upper abdomen
D. Contraindications
There are no absolute contraindications to chest tube placement. Ideally in non-

emergent cases, coagulopathies and low platelet counts should be corrected to
avoid excessive bleeding. CT imaging or ultrasound-guided location of the pla-
cement is helpful in patients with a history of pneumonectomy, lung transplant, or
complicated loculated effusions.
E. Equipment
The equipment necessary for chest tube placement is listed in Table 18. This
includes the appropriate-size chest tube, a drainage device with or without suction
source, connecting hoses with connectors, and a tray with insertion instruments.
F. Drain Size
Modern chest tubes are plastic, pliable, fenestrated tubes made of minimally
thrombogenic material. Chest tube sizes range from number 6 to 40 French.
Table 18 Equipment for Chest Tube Insertion
Povidone iodine solution

Sterile towels and drapes
Sterile sponges
2 Kelly clamps
1% lidocaine without epinephrine (40 50 mL)
1.0 non-absorbable sutures
18, 21, and 25 gauge needles
Scalpel, scissors
Needle holder
10 mL syringe
Sterile gown, gloves, mask, and cape
Sterile gauze
Chest tube and drainage system
Table 19 Internal Diameter Correlate to the French Grading System
French size Inch equivalent Millimeter equivalent
4F 0.053 1.35
8F 0.105 2.7
10F 0.131 3.3
16F 0.210 5.3
18F 0.236 6.0
22F 0.288 7.3
26F 0.341 8.7
28F 0.367 9.3
32F 0.419 10.7
36F 0.471 12.0
38F 0.498 12.7
40F 0.524 13.3
Table 19 lists the internal diameter correlate to the French grading system for a
range of chest tube sizes. The use of large bore tubes has previously been rec-
ommended (165 167) as it was felt they decreased the incidence of blockage,
particularly in thick malignant or infected uid. Most physicians now use
smaller tubes (10 14 French) as studies have shown these to be as effective as
the larger bore tubes (168,169), more comfortable, and better tolerated (170).
There remains intense debate over the optimal drain size. The 9 French tubes
have been used with success rates of up to 87% in pneumothoracies (171).
Ultrasound-guided insertion of small-bore tubes for treatment of malignant
pleural effusions for sclerotherapy has been well studied with good effect
(169,172,173). Small- and medium-bore chest tubes are usually placed using
the Seldinger technique.
Complicated parapneumonic effusions not amenable to a single large cath-
eter may require more than one small-bore tube placed under image guidance.
Larger bore tubes (32 40 French) are recommended for hemothorax, to allow
for successful drainage and assessment of continuing blood loss (174). The
larger bore tubes are inserted by blunt dissection, which allows the operator to
manually break down loculations if necessary.
G. Technique
Prior to inserting the chest tube, it is important to place the patient in the correct
position. The preferred position is on the bed, slightly rotated with the arm on the
side of the lesion behind the patients head, exposing the axillary area (166,175).
Alternatively, the patient can be sitting upright, leaning on a pillow over a table.
Aim for insertion in the triangle of safety (157), which lies between the
anterior and posterior axillary lines, through the fourth or fth intercostal
space, just above the rib, thus avoiding the intercostal neurovascular bundle.
Under sterile conditions, the area is prepped with pivodine iodine solution.
Local anesthesia is inltrated into the site of insertion. A small gauge needle is
used to raise a small bleb prior to inltration of the interocostal muscles and
pleural surface. Up to 30 40 mL of lidocaine may be needed to ensure adequate
anesthesia.
Thoracentesis is often initially performed to conrm the presence of air or
uid at the site of insertion. Chest tubes can be inserted by either blunt dissection
with or without the use of a trochar or the Seldinger technique. Whichever tech-
nique is used, the use of substantial force must be avoided to avoid penetration
and damage to vital intrathoracic structures. For blunt dissection, a 2-cm incision
is made parallel to the intercostal space, immediately above the rib. A Kelly
clamp is used for blunt dissection (Fig. 12), creating a subcutaneous tunnel to
the intercostal space. The closed clamp is gently inserted through the parietal
pleura over the superior aspect of the rib. Once through the parietal pleural,
the clamp is gently opened to spread the parietal pleura and intercostal
muscles. For a chest tube similar to the size of a nger, the tract should be
gently explored with a nger, performing a gentle sweep and taking care not
to disrupt rm organized adhesions. The proximal end of the tube, rmly held
Figure 12 Appropriate angle of penetration into the pleural space during the blunt dis-
section technique. Source: From Miller, KS, Sahn, SA. Chest tubes, indications, tech-
niques, management and complications. Chest 1987; 91(2):258 264.
by the Kelly clamp, is then inserted into the pleural space. If using a trochar, care
must be taken to ensure the sharp point of the trochar is retracted a few centi-
meters from the chest tube tip. Trochar insertion is relatively fast but runs the
risk of impaling the lung or other organs.
The position of the tip of the chest tube should ideally be aimed anteriorly
towards the apex of the lung in the treatment of a pneumothorax. For free-owing
effusions, chest tubes are aimed downwards and in the posterior direction, into
the costovertebral angle. For basal and lateral empyemas or loculated collections,
the chest tube is placed in the most dependant region. However, any tube position
can be effective at draining air or uid, and an effective drain should not be re-
positioned solely because of its radiographic position. The depth of insertion of
the chest tube ranges from 5 to 15 cm, ensuring all sideports are within the chest,
and the proximal port must be at least 2 cm beyond the rib margin. The centi-
meter markings on the side of the chest tube start from the most proximal port;
therefore, a tube at a 5-cm mark at the skin means it is 5 cm from the most prox-
imal port. The nal depth is dependant on the ease of aspiration of the pleural
uid and the patient size.
Correct insertion of the chest tube is conrmed by visualization of conden-
sation within the tube with respiration or the drainage of pleural uid. Two simple
horizontal mattress sutures are placed on either side of the insertion site to anchor
the tube. Complicated purse string sutures should be avoided as they convert a
linear incision into a circular wound, which is painful and can leave a prominent
scar (176). A sterile dressing is then applied over the site, which is covered and
secured with surgical tape. Occlusive petroleum dressings are generally not
recommended.
Small- and medium-bore catheters are usually inserted with the aid of guide
wire by the Seldinger technique (170,177,178) (Fig. 13). This technique is fast
and relatively safe, making a smaller incision and insertion, causing less dis-
comfort than the blunt dissection method. It is, however, difcult to control the
direction of the tube entry into the pleura, and advancement may be difcult if
adhesions are present. Using this technique, the introducer needle is advanced
over the superior border of the rib into the pleural space. Air or uid is aspirated
to conrm the correct position. The guide wire is then advanced through the
needle into the pleural space without any resistance. The needle is then
removed, leaving the guide wire in place. Ensuring adequate analgesia, a series
of dilators are passed over the guide wire into the pleural space. Introduction
into the pleural space is facilitated by rotating and advancing the dilators in
the same plane of the guidewire to prevent kinking. The chest tube assembly is
advanced into the pleural space over the guide wire. The guide wire and chest
tube inserter are removed, leaving the chest tube in place. All ports of the
chest tube should be in the pleural space.
Once the chest tube is inserted and secured, it is connected to the pleural
drainage system. All connections between the chest tube and its drainage
system should be tight and securely taped.
Figure 13 Percutaneous insertion of a chest tube using the Seldinger technique. Source:
Thal-Quick Chest Tube Instruction Manual. Cook Critical Care. Cook Incorporated, 1987.
H. Complications
Complications of chest tube placement can be divided into three categories:
placement, maintenance, and discontinuation (Table 20).
Complication rates from chest tube insertion in blunt trauma have been
shown to be between 9% and 21% (179,180). Insertion alone has a complication
Table 20 Complications of Chest Tube Placement
Placement Damage to vital structures (lung, heart, liver, spleen)

Hemorrhage (great vessels, heart, intercostal artery)
Neurological damage (phrenic nerve)
Chylothorax
Maintenance Re-expansion pulmonary edema
Subcutaneous emphysema
Lung entrapment
Pneumothorax with inadvertent disconnection
Empyema
Discontinuation Recurrence of pneumothorax
Pleurocutaneous stula
rate of 1% to 2% (179,181). Perforation of the RA, RV, and abdominal organs has
also been reported (182 185), as has chylothorax from thoracic duct injury
(186). Perforation of the diaphragm is avoided by inserting the chest tube no
lower than the fth intercostal space. Ipsilateral hemidiaphragmatic palsy from
isolated injury to the intrathoracic phrenic nerve has also been reported (187).
Subcutaneous emphysema from the thoracostomy site involving the chest wall
and extending into the neck can also occur, but this is usually a cosmetic problem.
Rapid re-expansion of previously collapsed lung may result in unilateral
pulmonary edema (188), which may be fatal (189). It can occur in young patients
with pneumothoracies (190) and also following removal of large volumes of uid
or after removal of an obstructive tumor (191 193). The development of
re-expansion pulmonary edema probably correlates with the amount of negative
intrathoracic pressure, which in turn is related to the rate of uid removal. The
onset of cough or chest tightness is an important warning sign to end the
procedure. Most experts recommend the removal of 1 to 1.5 L at any one time.
Supportive treatment is usually sufcient.
Secondary infection of the pleural space following chest tube insertion is a
rare complication that occurs most frequently following treatment of traumatic
hemothorax.
I. Chest Tube Management
Once secured, the chest tube is attached to a drainage device. Although there are
many commercially available devices, they consist of one or more of the follow-
ing: (1) water seal, (2) drainage tap, (3) pressure control change, and (4) connect-
ing hoses. The water seal prevents air from entering the pleural cavity whilst air
or uid is being drained. Most institutions use a three-chamber system (Fig. 14).
The closed water seal is placed under water at 3 cm. This allows the operator to
see air bubble out as the lung re-expands in pneumothorax or the uid evacuation
rate in effusions or empyemas. If suction is required, it is performed via the water
seal at a level of 15 to 20 cm of water. This water level should be checked daily to
ensure an adequate water seal and suction regulator is maintained. If suction is
applied, bubbling is seen in the suction control chamber. Periodic milking or
stripping of the connecting hoses is generally discouraged because as it has not
been shown to increase tube patency (194). If continuous bubbling through the
water seal is seen, intermittent clamping of the connecting hoses, starting at
the drainage device and moving proximally to the chest wall, should reveal the
site of leakage. Connecting sites are common sources of leaks. Persistent air
leaks may be due to bronchopleural stulas. If there is no variation of the
water seal with respiration, then occlusion of the chest tube should be suspected.
Saline irrigation should initially be attempted. Prolonged clamping of the chest
tube should be avoided as this may result in tension pneumothorax. Serial
chest X-rays should be obtained to assess tube position, exclude complications
Figure 14 Example of a three-chamber water seal chest drain. Source: Atrium Ocean
Water Seal Chest Drain, Atrium Medical Corporation, 2004.
such as pneumothorax, and evaluate the progress of drainage. Non-functioning

tubes should be removed and a new tube placed in a different site.
J. Chest Tube Removal

Chest tubes are removed when the pneumothorax or pleural uid has resolved or
when they are no longer functional. Although controversial, in the case of pneu-
mothorax, some physicians prefer clamping chest tubes for four to six hours prior
to removal in order to identify a persistent leak. Regardless of whether the chest
tube is clamped or not, most physicians wait ve to 12 hours after the last evi-
dence of an air leak to repeat a chest X-ray prior to removing the tube. For
pleural effusions, chest tubes can be removed when daily output is less than
100 mL on water seal or the tube is no longer functional. Chest tubes are
removed at the end of exhalation with the patient performing the Valsalva man-
euver. The tube is rapidly removed and an occlusive dressing applied. Indwelling
sutures are tightened. A follow-up chest X-ray should be performed.
References
1. Cohn, JN. Blood pressure measurement in shock: mechanism of inaccuracy auscul-

tatory and palpatory methods. J Am Med Assoc 1967; 199:972.
2. Bur A, Hirschl MM, Herkner H, et al. Accuracy of oscillometric blood pressure
measurement according to the relation between cuff size and upper arm circum-
ference in critically ill patients. Crit Care Med 2000; 28:371.
3. Coleman SS, Anson BJ. Arterial patterns in the hand based upon a study of 650
specimens. Surg Gynecol Obstet 1961; 113:409.
4. Clark CA, Harman EM. Hemodynamic monitoring: arterial catheters. In: Civetta
JM, Taylor RW, Kirby RR, eds. Critical Care. Philadelphia: Lipincott, 1988.
5. Allen EV. Thromboangitis obliterans: method diagnosis of chronic occlusive arter-
ial lesions distal to the wrist with illustrative cases. Am J Med Sci 1929; 178:237.
6. Glavin RJ, Jones HM. Assessing collateral circulation in the hand: four methods
compared. Anaesthesia 1989; 44:594.
7. Slogoff S, Keats AS, Arlund C. On the safety of radial artery cannulation. Anesthe-
siology 1983; 59(1):4247.
8. Nagabhushan S, Colella JJ Jr, Wagner R. Use of Doppler ultrasound in performing
percutaneous cannulation of the radial artery. Crit Care Med 1976; 4(6):327.
9. Maher JJ, Dougherty JM. Radial artery cannlation guided by Doppler ultrasound.
Am J Emerg Med 1989; 7(3):260 262.
10. Beards SC, Doedens L, Jackson A, Lipman J. A comparison of arterial lines and
insertion techniques in critically ill patients. Anaesthesia 1994; 49(11):968 973.
11. Mandel MA, Dauchot PJ. Radial artery cannulation in 1000 patients: precautions
and complications. J Hand Surg [Am] 1977; 2(6):482 485.
12. Bedford RF. Long-term radial artery cannulation: effects on subsequent vessel
function. Crit Care Med 1978; 6(1):64 67.
13. Davis FM, Stewart JM. Radial artery cannulation: a prospective study in patients
undergoing cardiothoracic surgery. Br J Anaesth 1980; 52(1):41 47.
14. Davis FM. Radial artery cannulation: inuence of catheter size and material on
arterial occlusion. Anaesth Intensive Care 1978; 6(1):49 53.
15. Oh T, Opie NJ, Davis NJ. Continuous ush system for radial artery cannulation.
Anaesth Intensive Care 1976; 4(1):29 32.
16. Weiss BM, Gattiker RI. Complications during and following radial artery cannula-
tion: a prospective study. Intensive Care Med 1986; 12(6):424 428.
17. Chang C, Dughi J, Shitabata P, Johnson G, Coel M, McNamara JJ. Air embolism
and the radial arterial line. Crit Care Med 1988; 16(2):141 143.
18. Raad I, Umphrey J, Khan A, Truett LJ, Bodey GP. The duration of placement as a
predictor of peripheral and pulmonary arterial catheter infections. J Hosp Infect
1993; 23(1):17 26.
19. Band JD, Maki DG. Infections caused by arterial catheters used for hemodynamic
monitoring. Am J Med 1979; 67(5):735 741.
20. Gronbeck C III, Miller EL. Nonphysician placement of arterial catheters. Experi-
ence with 500 insertions. Chest 1993; 104(6):1716 1717.
21. Singh S, Nelson N, Acosta I, Check FE, Puri VK. Catheter colonization and
bacteremia with pulmonary and arterial catheters. Crit Care Med 1982; 10(11):
736 739.
22. Thomas F, Burke JP, Parker J, et al. The risk of infection related to radial vs femoral
sites for arterial catheterization. Crit Care Med 1983; 11(10):807 812.
23. Maki DG. Nosocomial bacteremia. An epidemiologic overview. Am J Med 1981;
70(3):719 732.
24. Peruzzi WT, Parker MA, Lichtenthal PR, et al. A clinical evaluation of a blood con-
servation device in medical intensive care unit patients. Crit Care Med 1993;
21(4):501 506.
25. Smoller BR, Kruskall MS. Phlebotomy for diagnostic laboratory tests in adults.
Pattern of use and effect on transfusion requirements. N Engl J Med 1986;
314(19):1233 1235.
26. Silver MJ, Jubran H, Stein S, McSweeney T, Jubran F. Evaluation of a new blood
conserving arterial line system for patients in intensive care units. Crit Care Med
1993; 21(4):507 511.
27. Shapiro BA, Mahutte CK, Cane RD, Gilmour IJ. Clinical performance of a blood
gas monitor: a prospective, multicenter trial. Crit Care Med 1993; 21(4):487 494.
28. Veremakis C, Holloran TH. The technique of monitoring arterial blood pressure.
J Crit Illn 1989; 4:82.
29. Gardner RM, Hollingsworth KW. Optimizing the electrocardiogram and pressure
monitoring. Crit Care Med 1986: 14(7):651 658.
30. McGee DC, Gould MK. Preventing complications of central venous catheteriza-
tion. N Engl J Med 2003; 348(12):1123 1133.
31. Merrer J, De Jonghe B, Golliot F, et al. Complications of femoral and subclavian
venous catheterization in critically ill patients: a randomized controlled trial. J
Am Med Assoc 2001; 286(6):700 707.
32. Sznajder JI, Zveibil FR, Bitterman H, Weiner P, Bursztein S. Central vein catheter-
ization. Failure and complication rates by three percutaneous approaches. Arch
Intern Med 1986; 146(2):259 261.
33. Veenstra DL, Saint S, Saha S, Lumley T, Sullivan SD. Efcacy of antiseptic-
impregnated central venous catheters in preventing catheter-related bloodstream
infection: a meta-analysis. J Am Med Assoc 1999; 281(3):261 267.
34. Manseld PF, Hohn DC, Fornage BD, Gregurich MA, Ota DM. Compli-
cations and failures of subclavian-vein catheterization. N Engl J Med 1994;
331(26):1735 1738.
35. Hind D, Calvert N, McWilliams R, Davidson A, Paisley S, Beverley C, Thomas
S. Ultrasonic locating devices for central venous cannulation: meta-analysis.
BMJ 2003; 327(7411):361.
36. Randolph AG, Cook DJ, Gonzales CA, Pribble CG. Ultrasound guidance for place-
ment of central venous catheters: a meta-analysis of the literature. Crit Care Med
1996; 24(12):2053 2058.
37. Armstrong PJ, Cullen M, Scott DH. The SiteRite ultrasound machinean aid to
internal jugular vein cannulation. Anaesthesia 1993; 48(4):319 323.
38. Hateld A, Bodenham A. Portable ultrasound for difcult central venous access. Br
J Anaesth 1999; 82(6):822 826.
39. Ruesch S, Walder B, Tramer MR. Complications of central venous catheters:
internal jugular versus subclavian accessa systematic review. Crit Care Med
2002; 30(2):454 460.
40. Golden LR. Incidence and management of large-bore introducer sheath puncture of
the carotid artery. J Cardiothorac Vasc Anesth 1995; 9(4):425 428.
41. Applebaum RM, Adelman MA, Kanschuger MS, Jacobowitz G, Kronzon
I. Transesophageal echocardiographic identication of a retrograde dissection of
the ascending aorta caused by inadvertent cannulation of the common carotid
artery. J Am Soc Echocardiogr 1997; 10(7):749 751.
42. Getzen LC, Pollak EW. Short-term femoral vein catheterization. A safe alternative
venous access? Am J Surg 1979; 138(6):875 878.
43. Williams JF, Seneff MG, Friedman BC, McGrath BJ, Gregg R, Sunner J,
Zimmerman JE. Use of femoral venous catheters in critically ill adults: prospective
study. Crit Care Med 1991; 19(4):550 553.
44. Dailey RH. Femoral vein cannulation: a review. J Emerg Med 1985; 2(5):367
372.
45. Kurekci E, Kaye R, Koehler M. Chylothorax and chylopericardium: a complication
of a central venous catheter. J Pediatr 1998; 132(6):1064 1066.
46. Unnikrishnan D, Idris N, Varshneya N. Complete heart block during central venous
catheter placement in a patient with pre-existing left bundle branch block. Br J
Anaesth 2003; 91(5):747 749.
47. Wechsler RJ, Steiner RM, Kinori I. Monitoring the monitors: the radiology of thor-
acic catheters, wires, and tubes. Semin Roentgenol 1988; 23(1):61 84.
48. Rosen M, Latto IP, Ng WS. Handbook of percutaneous central venous catheteri-
zation. London: WB Sanders, 1981.
49. Lefrant JY, Muller L, De La Coussaye JE, Prudhomme M, Ripart J, Gouzes C,
Peray P, Saissi G, Eledjam JJ. Risk factors of failure and immediate complication
of subclavian vein catheterization in critically ill patients. Intensive Care Med 2002;
28(8):1036 1041.
50. Sivak SL. Late appearance of pneumothorax after subclavian venipuncture. Am J
Med 1986; 80(2):323 324.
51. Aronchick JM, Miller WT Jr. Tubes and lines in the intensive care setting. Semin
Roentgenol 1997; 32(2):102 116.
52. Heckmann JG, Lang CJ, Kindler K, Huk W, Erbguth FJ, Neundorfer B. Neurologic
manifestations of cerebral air embolism as a complication of central venous cathe-
terization. Crit Care Med 2000; 28(5):1621 1625.
53. Ely EW, Hite RD, Baker AM, Johnson MM, Bowton DL, Haponik EF. Venous air
embolism from central venous catheterization: a need for increased physician
awareness. Crit Care Med 1999; 27(10):2113 2117.
54. Trottier SJ, Veremakis C, OBrien J, Auer AI. Femoral deep vein thrombosis
associated with central venous catheterization: results from a prospective, random-
ized trial. Crit Care Med 1995; 23(1):52 59.
55. Joynt GM, Kew J, Gomersall CD, Leung VY, Liu EK. Deep venous thrombosis
caused by femoral venous catheters in critically ill adult patients. Chest 2000;
177:178 183.
56. Mian NZ, Bayly R, Schreck DM, Besserman EB, Richmand D. Incidence of deep
venous thrombosis associated with femoral venous catheterization. Acad Emerg
Med 1997; 4(12):1118 1121.
57. Durbec O, Viviand X, Potie F, Vialet R, Albanese J, Martin C. A prospective evalu-

ation of the use of femoral venous catheters in critically ill adults. Crit Care Med
1997; 25(12):1986 1989.
58. Centers for Disease Control and Prevention. Guidelines for the prevention of intra-
vsacular catheter-related infections. MMWR Recomm Rep 2002; 51(RR-10):
1 29.
59. DesJardin JA, Falagas ME, Ruthazer R, Grifth J, Wawrose D, Schenkein D, Miller
K, Snydman DR. Clinical utility of blood cultures drawn from indwelling central
venous catheters in hospitalized patients with cancer. Ann Intern Med 1999;
131(9):641 647.
60. Maki DG, Weise CE, Saran HW. A semiquantitative culture method for
identifying intravenous catheter-related infection. N Engl J Med 1977; 296(23):
1305 1309.
61. Mermel LA, McCormick RD, Springman SR, Maki DG. The pathogenesis and
epidemiology of catheter-related infection with pulmonary artery Swan Ganz
catheters: a prospective study utilizing molecular subtyping. Am J Med 1991;
91(3B):197S 205S.
62. Raad II, Hohn DC, Gilbreath BJ, Suleiman N, Hill LA, Bruso PA, Marts K,
Manseld PF, Bodey GP. Prevention of central venous catheter-related infections
by using maximal sterile barrier precautions during insertion. Infect Control
Hosp Epidemiol 1994; 15(4 Pt 1):231 238.
63. OGrady NP, Gerberding JL, Weinstein RA, Masur H. Patient safety and the
science of prevention: the time for implementing the Guidelines for the prevention
of intravascular catheter-related infections is now. Crit Care Med 2003; 31(1):
291 292.
64. Mermel LA. Prevention of intravascular catheter-related infections. Ann Intern
Med, 2000; 132(5):391 402.
65. Chaiyakunapruk N, Veenstra DL, Lipsky BA, Saint S. Chlorhexidine compared
with povidone iodine solution for vascular catheter-site care: a meta-analysis.
Ann Intern Med 2002; 136(11):792 801.
66. Richet H, Hubert B, Nitemberg G, Andremont A, Buu-Hoi A, Ourbak P, Galicier C,
Veron M, Boisivon A, Bouvier AM, et al. Prospective multicenter study of vascu-
lar-catheter-related complications and risk factors for positive central-catheter
cultures in intensive care unit patients. J Clin Microbiol 1990; 28(11):2520 2525.
67. Collignon P, Soni N, Pearson I, Sorrell T, Woods P. Sepsis associated with
central vein catheters in critically ill patients. Intensive Care Med 1988;
14(3):227 231.
68. Veenstra DL, Saint S, Saha S, Lumley T, Sullivan SD. Efcacy of antiseptic-
impregnated central venous catheters in preventing catheter-related bloodstream
infection: a meta-analysis. J Am Med Assoc 1999; 281(3):261 267.
69. Darouiche RO, Raad II, Heard SO, Thornby JI, Wenker OC, Gabrielli A, Berg J,
Khardori N, Hanna H, Hachem R, Harris RL, Mayhall G. A comparison of two anti-
microbial-impregnated central venous catheters. Catheter Study Group. N Engl J
Med 1999; 340(1):1 8.
70. Raad I, Darouiche R, Dupuis J, et al. Central venous catheters coated with minocy-
cline and rifampin for the prevention of catheter-related colonization and
bloodstream infections. A randomized, double-blind trial. The Texas Medical

Center Catheter Study Group. Ann Intern Med 1997; 127(4):267 274.
71. Maki DG, Stolz SM, Wheeler S, Mermel LA. Prevention of central venous catheter-
related bloodstream infection by use of an antiseptic-impregnated catheter. A ran-
domized, controlled trial. Ann Intern Med 1997; 127(4):257 266.
72. Collin GR. Decreasing catheter colonization through the use of an antiseptic-
impregnated catheter: a continuous quality improvement project. Chest 1999;
115(6):1632 1640.
73. Cobb DK, High KP, Sawyer RG, Sable CA, Adams RB, Lindley DA, Pruett TL,
Schwenzer KJ, Farr BM. A controlled trial of scheduled replacement of
central venous and pulmonary artery catheters. N Engl J Med 1992;
327(15):1062 1068.
74. Eyer S, Brummitt C, Crossley K, Siegel R, Cerra F. Catheter-related sepsis: pro-
spective, randomized study of three methods of long-term catheter maintenance.
Crit Care Med 1990; 18(10):1073 1079.
75. Martin L. Pulmonary Physiology in Clinical Practice. Ohio: The C.V. Mosby
Company, 1987:153.
76. Dexter L, Haynes FW, Burwell CS, Eppinger EC, Seibel RE, Evans JM. Studies of
congenital heart disease, I: technique of venous catheterization as a diagnostic pro-
cedure. J Clin Invest 1947; 26:547 553.
77. Dexter L, Haynes FW, Burwell CS, Eppinger EC, Sagerson RP, Evans JM. Studies
of congenital heart disease, II: the pressure and oxygen content of blood in the right
auricle, right ventricle, and pulmonary artery in control patients, with observations
on the oxygen saturation and source of pulmonary capillary blood. J Clin Invest
1947; 26:554 560.
78. Swan HJC, Ganz W, Forrester JS, Marcus H, Diamond G, Chonette D. Catheriza-
tion of the heart in man with use of a ow-directed balloon-tipped catheter. N Engl J
Med 1970; 283:447 451.
79. Mueller HS, Chatterjee K, Davis KB, et al. American College of Cardiology expert
consensus document: present use of bedside right heart catheterization in patients
with cardiac disease. J Am Coll Card 1998; 32:840 864.
80. Shoemaker WC, Kram HB, Appel PL. Therapy of shock based on pathophysiology,
monitoring, and outcome prediction. Crit Care Med 1990; 18(1):S19 S25.
81. Bland RD, Shoemaker WC, Abraham E, Cobo JC. Hemodynamic and oxygen trans-
port patterns in surviving and nonsurviving postoperative patients. Crit Care Med
1985; 13(2):85 90.
82. Schultz RJ, Whiteld GF, LaMura, JJ, Raciti A, Krishnamurthy S. The role of
physiologic monitoring in patients with fractures of the hip. J Trauma 1985;
25(4):309 316.
83. Rivers E, Nguyen B, Haystad S. Early goal-directed therapy in the treatment of
severe sepsis and septic shock. N Engl J Med 2001; 345:1368 1374.
84. Tuchschmidt J, Fried J, Astiz M, Rackow E. Elevation of cardiac output and oxygen
delivery improves outcome in septic shock. Chest 1992; 102:216 220.
85. Gore JM, Goldberg RJ, Spodick DH, Spodick DH, Alpert JS, Dalen JE. A community-
wide assessment of the use of pulmonary artery catheters in patients with acute
myocardial infarction. Chest 1987; 92:721 727.
86. Zion MM, Balkin J, Rosenmann D, Goldbourt U, Reicher-Reiss H, Kaplinsky E,

Behar S. Use of pulmonary artery catheters in patients with acute myocardial infarc-
tion: analysis of experience in 5841 patients in the SPRINT registry. Chest 1990;
98:1331 1335.
87. Boyd O. Effectiveness of right heart catheterization: time for a randomized trial
(Letter). J Am Med Assoc 1997; 277:110 111.
88. Sandham JD, Hull RD, Brant RF, et al. Canadian Critical Care Clinical Trials
Group. A randomized, controlled trial of the use of pulmonary artery catheters in
high-risk surgical patients. N Engl J Med 2003; 348:5 14.
89. Richard C, Warszawski, J, Anguel N, Deye N, Combes A, Barnoud D, Boulain T,
Lefort Y, Fartoukh M, Baud F, et al., the French Pulmonary Artery Catheter Study
Group. Early use of the pulmonary artery catheter and outcomes in patients with
shock and acute respiratory distress syndrome, a randomized controlled trial.
J Am Med Assoc 2003; 290:2713 2720.
90. Forrester JS, Ganz W, Diamond G, McHugh T, Chonette DW, Swan HJC. Thermo-
dilution cardiac output determination with a single ow-directed catheter. Am
Heart J 1972; 83:306 11.
91. Sprung CL, ed. The pulmonary artery catheter: methodology and clinical appli-
cations. Baltimore: University Park Press, 1983:73 104.
92. West JB, Dollery CT, Naimark A. Distribution of blood ow in isolated lung:
relation to vascular and alveolar pressures. J Applied Physiol 1964; 19:713 724.
93. Wiedemann HP, Matthay MA, Matthay RA. Cardiovascular pulmonary monitor-
ing in the intensive care unit (Part1). Chest 1984; 85:537 549.
94. Wiedemann HP, Matthay MA, Matthay RA Cardiovascular-Pulmonary monitoring
in the intensive care unit (Part 2). Chest 1984; 85:656 668.
95. Irwin RS, Rippe JM, Cerra SB, Curley FJ, Heard SO. Procedures and Techniques in
Intensive Care Medicine, 1999:46 71.
96. OQuin R and Marinii J. Pulmonary artery occlusion pressure: clinical physiology,
measurement, and interpretation. Am Rev Respir Dis 1983; 128:319 326.
97. Qvist J, Pontoppidan H, Wilson R, Lowenstein E, Laver M. Hemodynamic
responses to mechanical ventilation with PEEP: the effect of hypervolemia.
98. Zarins C, Virgilio R, Smith D, Peters R. The effect of vascular volume on positive
end-expiratory pressure-induced cardiac output depression and wedge left atrial
pressure discrepancy. J Surg Res 1977; 23:348 360.
99. Davison D, Parker M, Harrison RA. The validity of determinations of pulmonary
wedge pressure during mechanical ventilation. Chest 1978; 73:352 355.
100. Pepe PE, Marinii JJ. Occult positive end-expiratory pressure in mechanically-
ventilated patients with airow obstruction. Am Rev Respir Dis 1982; 126:
166 170.
101. Stevens JH, Rafn TA, Mihm FG, et al. Thermodilution cardiac output measure-
ment. Effects of the respiratory cycle on its reproducibility. J Am Med Assoc
1985; 253:2240 2242.
102. Hamilton MA, Stevenson, LW, Woo, M, et al. Effect of tricuspid regurgitation on
the reliability of the thermodilution cardiac ouput technique in congestive heart
failure. Am J Cardiol 1989; 64:945 948.
103. Nishikawa T, Dohi S. Errors in the measurement of cardiac output by thermo-

dilution. Can J Anaesth. 1993; 40(2):142 153.
104. Rady MY, Rivers EP, Nowak RM. Resuscitation of the critically ill in the ED:
responses of blood pressure, heart rate, shock index, central venous oxygen satur-
ation, and lactate. Am J Emerg Med 1996; 14:218 225.
105. Jastremski MS, Chelluri L, Beney KM, et al. Analysis of the effects of continuous
on-line monitoring of mixed venous oxygen saturation on patient outcome and cost-
effectiveness. Crit Care Med 1989; 17:148 153.
106. Vender JS. Invasive cardiac monitoring. Crit Care Clin 1988; 4:455 477.
107. Chatterjee K, Swan JHC, Ganz W, Gray R, Loebel H, Forrester JS, Chonette D. Use
of a balloon-tipped oatation electrode catheter for cardiac monitoring. Am J
Cardiol 1975; 36:56 61.
108. Fuchs RM, Heuser RR, Yin FC, Brinker JA. Limitations of pulmonary
wedge V waves in diagnosing mitral regurgitation. Am J Cardiol 1982;
49:849 854.
109. Picard AD, Kay R, Smith H, Rentrop P, Holt J, Gorlin R. Large V waves in the
pulmonary wedge pressure tracing in the absence of mitral regurgitation. Am J
Cardiol 1982; 50:1044 1050.
110. McAllister RG Jr, Friesinger GG, Sinclair-Smith BC. Tricuspid regurgitation
following inferior myocardial infarction. Arch Intern Med 1976; 136:95 99.
111. Reddy PS, Curtiss EI, OToole JD, Shaver JA. Cardiac tamponade: hemodynamic
observations in man. Circulation 1978; 58:265 272.
112. Cohn JN, Guiha NH, Broder MI, Limas CJ. Right ventricular infarction: clinical
and hemodynamic features. Am J Cardiol 1974; 33:209 214.
113. Bernard GR, Artigas A, Brigham KL, et al. The American European Consensus
Conference on ARDS: denitions, mechanisms, relevant outcomes, and clinical
trial coordination. Am J Respir Crit Care Med 1994; 149:818 824.
114. Mitchell JP, Schuller D, Calandrino FS, Schuster DP. Improved outcome based on
uid management in critically ill patients requiring pulmonary artery catheteriza-
tion. Am Rev Respir Dis 1992; 145(5):990 998.
115. ARDSNet. Prospective, randomized, multi-center trial of pulmonary artery catheter
(PAC) vs central venous catheter (CVC) for management of acute lung injury (ALI)
and acute respiratory distress syndrome (ARDS). Prospective, randomized multi-
center trial of uid conservative versus uid-liberal management of acute
lung injury (ALI) and acute respiratory distress syndrome (ARDS). Available
from http://ardsnet.org, March 2000. Last accessed on August 31, 2004. RFP
No. NIH-NHLBI-HR-01 01.cardiac disease.
116. McIntyre KM, Sasahara AA. The hemodynamic response to pulmonary
embolism in patients without prior cardiopulmonary disease. Am J Cardio 1971;
28:288 294.
117. Rice DL, Awe RJ, Gaasch WH, Alexander JK, Jenkins DE. Wedge pressure
measurements in obstructive pulmonary disease. Chest 1974; 66:628 632.
118. Moser KM, Spragg RG. Use of the balloon tipped pulmonary artery catheter in pul-
monary disease. Ann Intern Med 1983; 98:53 58.
119. American Society of Anesthesiologists Task Force. Practice guidelines for pulmon-
ary artery catheterization. Anesthesiology 2003; 99:988 1014.
120. Pulmonary Artery Catheter Consensus Conference. Consensus Statement. Crit Care
Med 1997; 25(6):910 925.
121. Shaw TJI. The Swan Ganz pulmonary artery catheter: incidence of complications,
with particular reference to ventricular dysrhythmias, and their prevention.
Anesthesia 1979; 34:651 656.
122. Thomson IR, Dalton BC, Lappas DG, Lowenstein E. Right bundle-branch block
and complete heart block caused by the Swan Ganz catheter. Anesthesiology
1979; 51:359 362.
123. Sprung CL, Jacobs LJ, Caralis PV, Karpf M. Ventricular arrhythmias during
Swan-Ganz catheterization of the critically ill. Chest 1981; 79:413 415.
124. Sprung CL, Elser B, Schein MH, Marcial EH, Shrager BR. Risk of right bundle-
branch block and complete heart block during pulmonary artery catheterization.
Crit Care Med 1989; 17:1 3.
125. Baldwin IC, Heland M. Incidence of cardiac dysrhythmias in patients during
pulmonary artery catheter removal after cardiac surgery. Heart Lung 2000;
29:155 160.
126. Iberti TJ, Benjamin E, Gruppi I, Raskin JM. Ventricular arrhythmias during
pulmonary artery catheterization in the intensive care unit: prospective study.
Am J Med 1985; 78:451 454.
127. Morris D, Mulvhill D, Lew WYW. Risk of developing complete heart block during
bedside pulmonary artery catheterization in patients with left bundle-branch block.
Arch Intern Med 1987; 147:2005 2010.
128. Lipp H, ODonoghue K, Resnekov L. Intracardiac knotting of a ow-directed
balloon catheter. N Engl J Med 1971; 284:220.
129. Foote GA, Schabel SI, Hodges M. Pulmonary complications of the ow-directed
balloon-tipped catheter. N Engl J Med 1974; 290:927 931.
130. Boyd KD, Thomas SJ, Gold J, Boyd AD. A prospective study of complications
of pulmonary artery catheterizations in 500 consecutive patients. Chest 1983;
84:245 249.
131. Carlson TA, Goldenberg IF, Murray PD, Tadavarthy SM, Walker M, Gobel FL.
Catheter-induced delayed recurrent pulmonary artery hemorrhage; J Am Med
Assoc 1989; 261:1943 1945.
132. Barash PG, Nardi D, Hammond G, Walker-Smith G, Capuano D, Laks H,
Kopriva CJ, Baue AE, Geha AS. Catheter-induced pulmonary artery perforation:
mechanisms, management and modications. J Thorac Cardiovasc Surg 1981;
82:5 12.
133. Sprung CL, Marcial EH, Garcia AA, Sequeira RF, Pozen RG. Prophylactic
use of lidocaine to prevent advanced ventricular arrhythmias during pulmonary
artery catheterization: prospective double blind study. Am J Med 1983;
75:906 910.
134. Kearney TJ, Shabot MM. Pulmonary artery rupture associated with the Swan Ganz
catheter. Chest 1995; 108:1349 1352.
135. Slaughter TF, Greenberg CS. Heparin-associated thrombocytopenia and throm-
bosis: Implications for perioperative management. Anesthesiology 1997;
87:667 675.
136. Thomas R., Siproudhis L, Laurent JF, Bouget J, Bousser J, Camus C, Michelet C.
Massive hemoptysis from iatrogenic balloon catheter rupture of pulmonary
artery: successful early management by balloon tamponade. Crit Care Med 1987;
15:272 273.
137. Kelly TF Jr, Morris GC, Crawford ES, Espada R, Howell JF. Perforation of the pul-
monary artery with Swan Ganz catheters: diagnosis and surgical management.
Ann Surg 1981; 193:686 692.
138. Slacken A. Complications of invasive hemodynamic monitoring in the intensive
care unit. Curr Probl Surg 1988; 25:69 145.
139. Michel L, Marsh M, McMichan JC, Southorn PA, Brewer NS. Infection of pulmon-
ary artery catheters in critically ill patients. J Am Med Assoc 1981; 245:
1032 1036.
140. Raad I, Umphrey J, Khan A, Truett LJ, Bodey GP. The duration of placement as a
predictor of peripheral and pulmonary arterial catheter infections. J Hosp Infect
1993; 23:17 26.
141. Rowley KM, Clubb KS, Smith GJW, Cabin HS. Right-sided infective endo-
carditis as a consequence of ow-directed pulmonary artery catheterization: A
clinicopathological study of 55 autopsied patients. N Engl J Med 1984;
311:1152 1156.
142. Mermel LA, McCormick RD, Springman SR, Maki DG. The pathogenesis and
epidemiology of catheter-related infection with pulmonary artery Swan Ganz
catheters: A prospective study utilizing molecular subtyping. Am J Med 1991;
91(3B):197S 205S.
143. Myers ML, Austin TW, Sibbald WJ. Pulmonary artery catheter infections: a
prospective study. Ann Surg 1985; 201:237 241.
144. Singh SJ, Puri VK. Prevention of bacterial colonization of pulmonary artery cath-
eters. Infect Surg 1984; 853 856.
145. Kopman EA, Sandza JG Jr. Manipulation of the pulmonary artery catheter after
placement: maintenance of sterility. Anesthesiology 1978; 48:373 374.
146. Heard SO, Davis RF, Sherertz RJ, Mikhail MS, Gallagher RC, Layon A, Gallagher
TJ. Inuence of sterile protective sleeves on the sterility of pulmonary artery cath-
eters. Crit Care Med 1987; 15:499 502.
147. Iberti T, Fischer E, Leibowitz AB, Panacek EA, Silverstein JH, Albertson TE, and
the Pulmonary Artery Catheter Study Group. A multicenter study of physicians
knowledge of the pulmonary artery catheter. J Am Med Assoc 1990; 264:2928
2932.
148. Bone RC. High-tech predicament: pulmonary artery catheter. J Am Med Assoc
1990; 264:2933.
149. Shoemaker WC, Appel PL, Kram HB, Waxman K, Lee TS. Prospective trial of
supranormal values of survivors as therapeutic goals in high-risk surgical patients.
Chest, 1988; 94:1176 1186.
150. Shoemaker WC, Appel PC, Kram HB. Oxygen transport measurements to evaluate
tissue perfusion and titrate therapy: dobutamine and dopamine effects. Crit Care
Med 1991; 19:672 688.
151. Ivanov R, Allen J, Calvin JE. The incidence of major morbidity in critically in
patients managed with pulmonary artery catheters: a meta-analysis. Crit Care
Med 2000 28:615 619.
152. Connors AF Jr, Speroff T, Dawson NV, Thomas C, Harrell FE Jr, Wagner D,
Desbiens, N, Goldman L, Wu AW, Califf RM, et al., for the SUPPORT
Investigators. The effectiveness of right heart catheterization in the initial care of

critically ill patients. J Am Med Assoc 1996; 276:889 897.
153. Dalen JE, Bone RC. Is it time to pull the pulmonary artery catheter? J Am Med
Assoc 1996; 276:916 918.
154. Harvey S, Harrison DA, Singer M, Aschcroft J, Jones CM, Elbourne D, Brampton
W, Williams D, Young D, Rowan K, et al. Assessment of the clinical effectiveness
of pulmonary artery catheters in management of patients in intensive care (PAC-
Man): a randomize controlled trial. Lancet 2005; 366:472 477.
155. Shah MR, OConnor CM, Sopko G, Hasselblad V, Califf RM, Stevenson LW.
Evaluation Study of Congestive Heart Failure and Pulmonary Artery Catheteriza-
tion Effectiveness (ESCAPE): design and rationale. Am Heart J 2001;
141(4):528 535.
156. Ginosar Y., Sprung CL. The Swan Ganz catheter: twenty-ve years of monitoring.
Crit Care Clin 1996; 12:771 776.
157. American College of Surgeons Committee on Trauma. Thoracic Trauma. In:
Student Manual of Advanced Trauma Life Support Course for Physicians. 5th
ed. American College of Surgeons, 1993.
158. Light RW. A new classication of parapneumonic effusions and empyema. Chest
1995; 108:299 301.
159. Krejci CS, Trent EJ, Dubinsky T. Thoracic sonography. Respir Care 2001;
46(9):932 939.
160. Bouros D, Schiza S, Patsourakis G, Chalkiadakis G, Panagou P, Siafakas NM. Intra-
pleural streptokinase versus urokinase in the treatment of complicated parapneu-
monic effusions: a prospective, double-blind study. Am J Respir Crit Care Med
1997; 155(1):291 295.
161. Robinson LA, Moulton AL, Fleming WH, Alonso A, Galbraith TA. Intrapleural
brinolytic treatment of multiloculated thoracic empyemas. Ann Thorac Surg
1994; 57(4):803 813, 813 814 (discussion).
162. Cassina PC, Hauser M, Hillejan L, Greschuchna D, Stamatis G. Video-assisted
thoracoscopy in the treatment of pleural empyema: stage-based management and
outcome. J Thorac Cardiovasc Surg 1999; 117(2):234 238.
163. Light RW, Macgregor MI, Luchsinger PC, Ball WC Jr. Pleural effusions: the
diagnostic separation of transudates and exudates. Ann Intern Med 1972;
77:507 513.
164. Heffner JE, Brown LK, Barbieri CA. Diagnostic value of tests that discriminate
between exudative and transudative pleural effusions. Chest 1997; 111:970 980.
165. Harriss DR, Graham TR. Management of intercostal drains. Br J Hosp Med 1991;
45(6):383 386.
166. Quigley RL. Thoracentesis and chest tube drainage. Crit Care Clin 1995;
11(1):111 126.
167. Hyde J, Sykes T, Graham T. Reducing morbidity from chest drains. BMJ 1997;
314(7085):914 915.
168. Clementsen P, Evald T, Grode G, Hansen M, Krag Jacobsen G, Faurschou
P. Treatment of malignant pleural effusion: pleurodesis using a small percutaneous
catheter. A prospective randomized study. Respir Med 1998; 92(3):593 596.
169. Parker LA, Charnock GC, Delany DJ. Small bore catheter drainage and sclero-
therapy for malignant pleural effusions. Cancer 1989; 64(6):1218 1221.
170. Patz EF Jr, Goodman PC, Erasmus JJ. Percutaneous drainage of pleural collections.
J Thorac Imaging 1998; 13(2):83 92.
171. Conces DJ Jr, Tarver RD, Gray WC, Pearcy EA. Treatment of pneumothoraces
utilizing small caliber chest tubes. Chest 1988; 94(1):55 57.
172. Goff BA, Mueller PR, Muntz HG, Rice LW. Small chest-tube drainage followed by
bleomycin sclerosis for malignant pleural effusions. Obstet Gynecol. 1993;
81(6):993 996.
173. Thompson RL, Yau JC, Donnelly RF, Gowan DJ, Matzinger FR. Pleurodesis with
iodized talc for malignant effusions using pigtail catheters. Ann Pharmacother
1998; 32(7 8):739 742.
174. Parry GW, Morgan WE, Salama FD. Management of haemothorax. Ann R Coll
Surg Engl 1996; 78(4):325 326.
175. Iberti TJ, Stern PM. Chest tube thoracostomy. Crit Care Clin 1992; 8(4):879 895.
176. Tomlinson MA, Treasure T. Insertion of a chest drain: how to do it. Br J Hosp Med
1997; 58(6):248 252.
177. Thal AP, Quick KL. A guided chest tube for safe thoracostomy. Surg Gynecol
Obstet 1988; 167(6):517.
178. Guyton SW, Paull DL, Anderson RP. Introducer insertion of mini-thoracostomy
tubes. Am J Surg 1988; 155(5):693 696.
179. Daly RC, Mucha P, Pairolero PC, Farnell MB. The risk of percutaneous chest tube
thoracostomy for blunt thoracic trauma. Ann Emerg Med 1985; 14(9):865 870.
180. Etoch SW, Bar-Natan MF, Miller FB, Richardson JD. Tube thoracostomy. Factors
related to complications. Arch Surg 1995; 130(5):521 525.
181. Millikan JS, Moore EE, Steiner E, Aragon GE, Van Way CW III. Complications of
tube thoracostomy for acute trauma. Am J Surg 1980; 140(6):738 741.
182. Chest drainage tubes. Surg Clin North Am 1989; 69(1):41 46.
183. Dalbec DL, Krome RL. Thoracostomy. Emerg Med Clin North Am 1986;
4(3):441 457.
184. Kopec SE, Conlan AA, Irwin RS. Perforation of the right ventricle: a complication
of blind placement of a chest tube into the postpneumonectomy space. Chest 1998;
114(4):1213 1215.
185. Meisel S, Ram Z, Priel I, Nass D, Lieberman P. Another complication of thora-
costomyperforation of the right atrium. Chest 1990; 98(3):772 773.
186. Kumar SP, Belik J. Chylothoraxa complication of chest tube placement in a
neonate. Crit Care Med 1984; 12(4):411 412.
187. Palomeque A, Canadell D, Pastor X. Acute diaphragmatic paralysis after chest tube
placement. Intensive Care Med 1990; 16(2):138.
188. Murphy K, Tomlanovich MC. Unilateral pulmonary edema after drainage of a
spontaneous pneumothorax: case report and review of the world literature.
J Emerg Med 1983; 1(1):29 36.
189. Gascoigne A, Appleton A, Taylor R, Batchelor A, Cook S. Catastrophic circulatory
collapse following re-expansion pulmonary oedema. Resuscitation 1996;
31(3):265 269.
190. Fitzpatrick S, Acheson J, Curran P. Re-expansion pulmonary edema and circulatory
shock in a 20-year-old man Eur J Emerg Med 2003; 10(2):146 148.
191. Mahfood S, Hix WR, Aaron BL, Blaes P, Watson DC. Reexpansion pulmonary
edema. Ann Thorac Surg 1988; 45(3):340 345.
192. Timby J, Reed C, Zeilender S, Glauser FL. Mechanical causes of pulmonary edema.
Chest 1990; 98(4):973 979.
193. Matsumiya N, Dohi S, Kimura T, Naito H. Reexpansion pulmonary edema after
mediastinal tumor removal. Anesth Analg 1991; 73(5):646 648.
194. Lim-Levy F, Babler SA, De Groot-Kosolcharoen J, Kosolcharoen P, Kroncke GM.
Is milking and stripping chest tubes really necessary? Ann Thorac Surg 1986;
42(1):77 80.
8
Bronchoscopy
NIKHIL SHAH IRAWAN SUSANTO
Pulmonary and Critical UCLASanta Monica Specialties

Care Section Geffen School of Medicine at UCLA
West Los Angeles Healthcare Center Santa Monica, California, U.S.A.
VA Greater Los Angeles Healthcare
System and Geffen School
of Medicine at UCLA
SILVERIO SANTIAGO
Pulmonary and Critical
Care Section
West Los Angeles Healthcare Center
VA Greater Los Angeles Healthcare
System and Geffen School
of Medicine at UCLA
I. Introduction
Bronchoscopy is the most commonly performed procedure by pulmonary physicians

today. The history of bronchoscopy dates back to the 19th century and was rst
performed in 1897 by Gustav Killian for extraction of an aspirated piece of bone
from a patients right main stem bronchus (1). The rst bronchoscopes developed
were rigid and used mostly for extraction of foreign bodies. Modications to the
rigid bronchoscope now allow for maintenance of ventilation, improved visualization
with the use of optical telescopes, and passage of various instruments for diagnostic
and therapeutic procedures (2,3). The major disadvantage of rigid bronchoscopy is
the inability to access the upper lobes and airways distal to the segmental orices.
The rigid bronchoscope continues to have many important uses today, although its
use has subsided with the introduction of the exible beroptic bronchoscope. The
rst beroptic bronchoscope was developed by Ikeda in 1964 (4). The properties of
285
286 Shah, Susanto, and Santiago
beroptics enable the bronchoscope to bend, and this allows for easy navigation
throughout the tracheo-bronchial tree. Flexible bronchoscopy currently plays an
important role for diagnosis and treatment in both outpatient and inpatient settings.
II. Specications, Indications, and Contraindications
Rigid bronchoscopes are straight metal hollow tubes that allow for diagnostic
procedures and therapeutic interventions. They have ports that allow for venti-
lation and administration of anesthetic gases and suction ports for removal of
secretions and blood. The hollow tube allows for the introduction of optical tele-
scopes or a exible beroptic bronchoscope. Other types of accessory equipment
are available, including biopsy forceps, baskets for foreign body removal, intro-
ducers for stent placement, and laser bers for treatment of endobronchial
lesions. Rigid bronchoscopy is typically performed under general anesthesia.
The indications for rigid bronchoscopy are listed in Table 1.
The exible beroptic bronchoscope has continued to evolve since its intro-
duction in 1967. The average adult bronchoscope has an external diameter of 4 to
6 mm and a typical viewing radius of 1208. The working channel for introduction of
accessory instruments or suction is typically 1 to 3 mm in diameter. Ultrathin
bronchoscopes for the evaluation of small distal airways are now available and
have an external diameter of 2.7 mm. Another relatively new development is the
videoscope that contains a video chip at the distal tip of the bronchoscope. These
chips have a lower propensity for damage when compared with ber bundles and
also allow for digital processing of images. A wide variety of ancillary equipment
may be used with the exible beroptic bronchoscope, including biopsy forceps,
protected and non-protected brushes for cytologic and microbiologic studies, and
devices for transbronchial needle aspiration. Flexible bronchoscopy is usually per-
formed with local anesthesia and may or may not require conscious sedation. The
indications for exible bronchoscopy are listed in Table 2 (5,6).
There are relatively few absolute contraindications to both rigid and ex-
ible bronchoscopy. These include lack of consent for the procedure and the pre-
sence of malignant cardiac arrhythmias and refractory hypoxemia. Additional
absolute contraindications for rigid bronchoscopy include the presence of cervi-
cal injury, an unstable neck, and microstomia.
Table 1 Indications for Rigid Bronchoscopy
Dilatation of strictures/stenosis
Placement of airway prostheses (stents)
Retrieval of foreign body
Massive hemoptysis
Resection of bulky tumors
Laser therapy
Cryotherapy
Photodynamic therapy
Bronchoscopy 287
Table 2 Indications for Flexible Bronchoscopy
Diagnostic indications
Abnormal chest roentgenogram
Unexplained cough, hemoptysis, wheeze, or stridor
Suspected pulmonary infections
Refractory lung abscess
Determination of the presence and extent of thermal burns
Evaluation of the airways for suspected injury after chest trauma
Unexplained vocal cord paralysis and diaphragmatic paralysis
Abnormal sputum cytology
Suspected tracheo-esophageal or broncho-pleural stulas
Carcinoma of the lung
Evaluation of problems associated with endotracheal tubes
To obtain diagnostic material from patients with non-infectious diffuse lung disease
Foreign bodies
Assessment of airway patency
Evaluation of lung transplant rejection
To investigate the etiology of SVC syndrome, chylothorax, or unexplained pleural
effusion
Therapeutic indications
Atelectasis secondary to retained secretions, mucous plugs, or clots not mobilized by
non-invasive techniques
Debulking of malignant neoplasm
Retrieval of foreign bodies
Hemoptysis
Treatment of Bronchopleural stula
Drainage of bronchogenic cysts
Endotracheal tube placement
Percutaneous dilatational tracheostomy
Electrocautery
Cryotherapy
Photodynamic therapy
Laser therapy
Brachytherapy
Placement of metal stents
Dilatation of strictures and stenosis
Therapeutic BAL for Pulmonary Alveolar Proteinosis
Abbreviations: SVC, superior vena cava; BAL, Bronchoalveolar lavage.
Conditions involving increased risk for both rigid and exible broncho-
scopy include lack of patient cooperation, recent myocardial infarction, unstable
angina, and active bronchospasm. Increased risk of bleeding after transbronchial
biopsy is seen in patients with pulmonary hypertension. Similarly, increased risk
of bleeding after biopsy is seen in patients with uremia, thrombocytopenia, and
coagulopathy. Bleeding and laryngeal edema may also occur in patients with
superior vena cava syndrome.
III. Patient Preparation, Sedation, and Anesthesia
Bronchoscopies may be performed in endoscopy suites, operating rooms, radi-

ology departments, or at the patients bedside. Pre-procedure preparation is
similar to that for other procedures. Informed consent should be obtained from
the patient. If the patient is unable to consent, the durable power of attorney
for healthcare or surrogate should give informed consent on behalf of the
patient. Patients should not take anything by mouth for six to eight hours prior
to the procedure. Pre-bronchoscopy laboratory tests, including complete blood
count, prothrombin time, partial thromboplastin time, and chemistries, are
ordered by most physicians. Imaging studies should be available for review,
including chest roentgenograms and computerized tomography (CT) scans,
especially if transbronchial needle aspiration is to be performed. Patients on
anti-coagulant therapy should be transitioned from coumadin to low molecular
weight heparin or unfractionated heparin so that coagulopathy can be correc-
ted prior to the procedure. Endocarditis prophylaxis is recommended in suscep-
tible patients undergoing rigid bronchoscopy but not beroptic bronchoscopy.
Nevertheless, each patient should be individualized based on the presence
or absence of signicant infection, prosthetic valves, or prior episodes of
endocarditis (7).
All patients undergoing bronchoscopy should have an intravenous line.
Blood pressure, oxygen saturation, and electrocardiogram should be monitored
and oxygen administered to maintain oxygen saturation greater than 90%.
Patients who undergo rigid bronchoscopy usually receive general anesthesia
under the guidance of an anesthesiologist. On the other hand, exible beroptic
bronchoscopy can be performed with or without conscious sedation. Most
pulmonologists, however, sedate patients for exible beroptic bronchoscopy
(8). Frequently used agents include benzodiazepines and opiates. The benzo-
diazepines provide sedation and an amnesic effect, whereas opiates provide
sedation and help suppress coughing. Over-sedation with benzodiazepines and
opiates can be reversed with umazenil and naloxone, respectively.
Local anesthesia is usually attained with topical agents, such as the amide
derivative lidocaine. Lidocaine is one of the safest topical agents for local
anesthesia and is applied to the nares, oropharynx, vocal cords, and tracheo-
bronchial tree (9). Caution should be exercised because lidocaine is easily
absorbed from the respiratory tract and excess levels can lead to seizures,
cardiac arrhythmias, and hypotension. It is, therefore, recommended that the
total dose of lidocaine be limited to 8.2 mg/kg in adults (10). As lidocaine is
metabolized in large part by the liver, patients with hepatic disease may be at
increased risk for developing toxic blood levels. Lidocaine has also been
reported to cause methemoglobinemia with hypoxemia (11), but when used judi-
ciously, it provides adequate anesthesia without signicant risk of adverse
effects. The routine use of antisialogogues, such as atropine, has come into
question (10).
Bronchoscopy 289
IV. Insertion Techniques
The rigid bronchoscope can be inserted through the mouth or through a tracheost-
omy stoma. On the other hand, the exible bronchoscope can be introduced
through the nose, mouth, tracheostomy, or endotracheal tube. The oral route is
preferred for exible bronchoscopy if frequent removal and reinsertion of the
bronchoscope is anticipated or if endotracheal intubation is to be performed.
The decision to intubate the patient is dependent on the clinical status of the
patient. Patients are frequently intubated over the bronchoscope if their respirat-
ory status is tenuous or if they are at increased risk for bleeding after biopsies.
The nasal route avoids the possible complication of damage to the bronchoscope
if the patient bites down in the absence of a bite block. In the intensive care unit
(ICU), a patient requiring bronchoscopy may already have an endotracheal tube
in place. This can pose problems if the tube is too small to accommodate an
average bronchoscope. A size 7.5 (mm, inner diameter) endotracheal tube will
allow passage of the bronchoscope but can limit adequate ventilation. A size
8.0 or greater endotracheal tube is recommended for ease of passage of the
bronchoscope and for adequate ventilation. In patients with signicant lung
disease, the bronchoscope may need to be withdrawn multiple times throughout
the procedure to allow the patient to oxygenate and ventilate adequately. Smaller
pediatric bronchoscopes can also be used. When compared with larger diameter
adult bronchoscopes, use of pediatric bronchoscopes offers advantages in the
mechanically ventilated patient, with less signicant cardiovascular and respirat-
ory side effects while providing comparable cell and microbiological yields (12).
When performing bronchoscopy on patients on mechanical ventilation, the FIO2
should be increased to 1.0 and maintained at that level throughout the procedure.
V. Diagnostic Techniques
Bronchoalveolar lavage (BAL) is a frequently utilized diagnostic procedure and

is also used occasionally as a therapeutic modality in patients with pulmonary
alveolar proteinosis. BAL uid is a marker for infectious and inammatory pro-
cesses occurring at the alveolar level. Proper technique of obtaining BAL
samples is important in order to ensure diagnostic accuracy. BAL uid should
be obtained prior to performing biopsies or brushings in order to avoid the admix-
ture of signicant amounts of blood with the sample. This is especially true if
bronchoscopy is being performed for diagnosis in patients with interstitial lung
disease. Samples should be obtained from localized areas of abnormalities. In
the absence of focal abnormalities, samples are usually obtained from the right
middle lobe or the lingula. The bronchoscope should be allowed to occlude the
airway (usually in the fourth- or fth-order bronchus) without over-wedging.
Sterile saline at room temperature or warmed to body temperature is instilled
via the working channel in aliquots of 50 to 60 mL to a total of 150 to 200 mL
per BAL site. After a short dwell time, the uid is suctioned into a sterile
container, with care being taken not to apply excessive suction to avoid collapse
of the airway or suction trauma. It is estimated that this allows for the sampling of
approximately one million alveoli. The uid can then be sent for a variety of ana-
lyses, including differential cell counts, cytologic examination, and microbiolo-
gic studies (13,14). Complications of BAL are uncommon but include fever,
pneumonitis, bleeding, and bronchospasm.
Endobronchial brushing is another frequently used technique for obtaining
samples for cytology and microbiology. Brushes may be protected (sheathed) or
unprotected. When attempting to obtain an uncontaminated material for culture,
the protected brush is typically favored. Discrete peripheral lesions suspicious for
malignancy can also be brushed to obtain specimens for cytology and is performed
under uoroscopy to ensure that the brush is in the vicinity of the lesion. The most
frequent complication seen with brushing is bleeding, which is usually self-limited.
Bronchoscopic biopsies can be obtained from both visible endobronchial
lesions and distal lesions. Different biopsy forceps are available, including
toothed, non-toothed, and spiked forceps. Endobronchial lesions can be biopsied
under direct visualization, thus minimizing risk of complications. Spiked forceps
with an impaler needle can facilitate biopsies of endobronchial lesions that are dif-
cult to grab. Four to six biopsies are generally recommended to optimize diagnostic
yield in bronchogenic carcinoma, although as many as 10 biopsies may be necessary
to maximize yield for peripheral carcinomas (15). For more peripheral lung lesions
or diffuse lung disease, samples are obtained using the transbronchial approach.
Transbronchial biopsy in diffuse lung disease can be performed with or without
uoroscopic guidance. Whether use of uoroscopy decreases the incidence of
pneumothorax after transbronchial biopsy is debated (10,16,17). The other major
complication that occurs with bronchoscopic biopsies is bleeding. Bleeding is
usually self-limited but may be severe. Thus, it is important to ensure that the
patient does not have coagulopathy, thrombocytopenia, signicant renal failure,
or severe pulmonary hypertension prior to proceeding with biopsies. Topical appli-
cation of 5 mL of 1:20,000 epinephrine prior to biopsy may minimize bleeding (18).
Transbronchial needle aspiration is a useful procedure that allows sampling
of mediastinal lymph nodes for diagnosis and staging of lung cancer. The bronch-
oscopist must be familiar with mediastinal anatomy prior to performance of
transbronchial needle aspiration. In addition, a CT scan of the chest is required
to determine the exact location of adenopathy or tumor. Transbronchial
needles come in different sizes, with smaller needles providing cytology speci-
mens and larger ones providing biopsy specimens. Complications from this pro-
cedure are rare and include minor bleeding (19).
Endobronchial ultrasound (EBUS) is a recent addition in the armamentar-
ium of the interventional pulmonologist. EBUS extends the imaging capability of
the bronchoscopist beyond the airway lumen in assessing bronchial wall tumor
invasion, locating lymph nodes or tumor for needle biopsy, and locating vascular
structures adjacent to the airway. EBUS requires transient occlusion of the airway
for 20 to 30 seconds, and it provides a 3608 view of the soft tissues surrounding
the airway. It can be used to identify the cause of an extrinsic compression and
Bronchoscopy 291
determine if it is a vascular structure or soft tissue. EBUS has been shown to

increase the yield of transbronchial needle aspiration of mediastinal lymph
nodes (20).
Fluorescence bronchoscopy is a diagnostic technique that utilizes charac-
teristic autouorescence emissions from lung tissues when illuminated by
light. When illuminated by a light of a given excitation wavelength, cancerous
and pre-cancerous lesions display lower autouorescence than normal tissue.
Fluorescence bronchoscopy is used to enhance the ability to detect and localize
dysplastic mucosa (premalignant) or carcinoma in situ. In a multi-center clinical
trial, uorescent bronchoscopy was shown to have a higher sensitivity and a
higher negative predictive value than regular white light bronchoscopy for the
diagnosis of early lung cancer (21). Mucosal dysplasia or airway cancer is not
likely to be present when uorescence bronchoscopy is negative. The presence
of an abnormal uorescent pattern is, however, non-specic and requires tissue
biopsy for histologic conrmation. Fluorescence bronchoscopy is FDA-approved
for patients who present with positive sputum cytology, hemoptysis, unresolved
pneumonia, persistent cough of unclear cause, or an abnormal chest X-ray suspi-
cious for lung cancer. Fluorescence bronchoscopy is also useful in the preopera-
tive assessment of lung cancer to determine the extent of endobronchial
involvement and the presence of synchronous airway lesions.
VI. Post-Procedure Care
Prior to discharge, patients should be monitored until sedation has worn off.
Post-procedure chest radiographs are routinely recommended after rigid
bronchoscopy or after bronchoscopy in patients on mechanical ventilation.
This may not be necessary after transbronchial biopsies, particularly if biopsies
are performed under uoroscopic guidance and the patient does not complain
of shortness of breath or chest pain. In a large survey, however, the majority
of pulmonologists replied that they routinely obtain chest radiographs after
transbronchial biopsy (8).
VII. Complications
When performed with appropriate care, bronchoscopy is a safe procedure. The

incidence of major and minor complications range from 0.08% to 0.12% and
0.27% to 6.5%, respectively (8,22 25). Mortality rates from surveys and pro-
spective studies range from 0.01% to 0.1% (22 26). As discussed earlier,
common complications include hypoxemia, bronchospasm, cardiovascular
events, pneumothoraces, bleeding, fever, pneumonitis, and complications
related to anesthesia and pre-medication. The number of major complications
was higher with rigid bronchoscopy than with beroptic bronchoscopy in a com-
parative study of 3449 rigid bronchoscopies and 1146 exible bronchoscopies
(26). In addition, higher rates of complications related to anesthesia were seen for
both procedures.
VIII. Specic Diagnostic Indications

A. Diagnosis and Staging of Lung Cancer
The utility of bronchoscopy in lung cancer is well established. As a diagnostic
tool, the yield is best if bronchoscopic techniques are combined, including
forceps biopsy, brush biopsy, BAL, and transbronchial needle aspiration when
appropriate. For visible endobronchial lesions, the diagnostic yield is higher
than 90% (15). For non-visible peripheral lesions, the yield will vary according
to the experience of the bronchoscopist, the location and size of the lesion, and
techniques utilized. Reports of diagnostic yield for non-visible lesions range
from 27% to 75% (15,27,28). The utility of bronchoscopy is likewise well estab-
lished in the staging of lung cancer. Again, accuracy will depend on the experi-
ence of the operator with transbronchial needle aspiration, the location of the
tumor, and the presence or absence of extrinsic compression of the airways. A
73% positive aspiration rate in patients with abnormal mediastinum on chest
roentgenograms with high specicity has been reported by Wang and Terry (19).
B. Suspected Pulmonary Infections
Although the etiologic agent of most infections can be determined based on clini-
cal history and non-invasive testing, bronchoscopy often serves as a useful
adjunct when therapy is ineffective or when prompt diagnosis is necessary. In
addition to providing specimens for microbiologic studies, bronchoscopy pro-
vides information regarding the severity of airway inammation and the quantity
and nature of secretions. BAL and protected brushings are particularly useful for
obtaining specimens for microbiologic studies. It is important to keep in mind
that although the isolation of certain organisms is diagnostic for the etiology of
disease (e.g., Nocardia), other organisms may be non-pathogenic colonizers
(14). All BAL culture results must, therefore, be interpreted in the proper clinical
context. Transbronchial biopsy is an additional technique that may add to the
diagnostic yield, particularly if invasive fungal or viral disease is suspected
(29). Transbronchial biopsies may also help establish the diagnosis of non-
infectious processes, such as Kaposi sarcoma or lymphoma, especially in immu-
nocompromised hosts.
Although the effect may be negligible, the inhibitory properties of lidocaine
on bacterial growth in vitro should be kept in mind when using liberal amounts of
lidocaine during diagnostic bronchoscopy for infection (9,30). On the other hand,
BAL or brush specimens may be falsely positive if withdrawn through a working
channel contaminated with upper airway secretions. Therefore, care must be
taken to obtain non-contaminated specimens. In addition, bronchoscopes must
Bronchoscopy 293
be thoroughly sterilized prior to use in order to prevent transmission of infection.

The latter is an infrequent but real risk (31,32).
Bronchoscopy is particularly useful in immunocompromised patients who
are susceptible not only to common pathogens, but also to pathogens that do not
typically cause illness in immunocompetent hosts. Timely diagnosis is crucial in
these patients because of their limited reserve and inability to mount an effective
immune response.
Among immunocompetent patients with pneumonia, bronchoscopy is
reserved for those with persistent or worsening inltrates despite appropriate
empiric agents. Bronchoscopy may assist not only in determining the etiologic
agent but may also uncover the presence of an obstructing endobronchial
lesion or foreign body leading to post-obstructive pneumonia. Bronchoscopy
may also assist in determining the cause of rapid deterioration leading to acute
respiratory distress syndrome (33).
Hospital-acquired (nosocomial) pneumonia is a serious illness with
signicant morbidity and mortality, and it is the leading cause of death from
hospital-acquired infections (34,35). It is dened as pneumonia that occurs
48 hours after admission, excluding any infection that is incubating at the
time of admission (36). Patients who develop hospital-acquired pneumonia
tend to be infected with pathogens different from those associated with
community-acquired pneumonias. Typical pathogens include staphylococcus
aureus and enteric Gram-negative bacilli. Ventilator-associated pneumonia, a
subset of nosocomial pneumonia, is diagnosed based on the appearance of
pneumonia .48 hours after intubation. Pneumonia in these patients is difcult
to diagnose and is based on radiographic evidence of a new or progressive inl-
trate with clinical ndings of fever, leukocytosis, and purulent secretions
(37,38). Cultures are often obtained from sputum or endotracheal aspirates in
intubated patients. These cultures are, however, frequently positive for bacteria
that colonize the oropharynx or the endotracheal tube as a biolm. Broncho-
scopy can be performed in these patients utilizing quantitative cultures from
BAL or protected brush specimens. The threshold for positive cultures is
usually .103 colony forming units (cfu)/mL for brushings and .104 cfu/mL
for BAL (39 41). Controversy exists regarding the impact of invasive testing
in patients with nosocomial pneumonia. Results of studies are mixed with
regards to the effects of bronchoscopy on outcomes such as mortality (42,43).
Additional randomized clinical studies are needed to determine whether
bronchoscopy should be routinely performed in patients with suspected nosoco-
mial pneumonia.
C. Specic Pathogens
Pneumocystis jiroveci (previously Pneumocystis carinii) is a common pulmonary
pathogen in patients with AIDS and other immunocompromised hosts. The diag-
nostic yield of BAL for Pneumocystis is as high as 86% to 97% in patients with
AIDS (44 46). Transbronchial biopsy, in addition, increases the diagnostic yield
with a reported sensitivity of 100% when combined with BAL (46). The yield
tends to be lower, however, in patients receiving aerosolized pentamidine (47)
or in patients with disease processes other than HIV. Site-directed multiple
lobe BAL combined with the use of indirect uorescent antibodies have also
been reported to increase the diagnostic yield for Pneumocystis (48). Bronchial
brushings, on the other hand, do not provide additional value to BAL and trans-
bronchial biopsy (49).
Mycobacterial diseases, including those caused by Mycobacterium tuber-
culosis as well as the many atypical mycobacteria, continue to be a major
health concern both in the United States and throughout the world (50). Diagnosis
of tuberculosis is typically made by obtaining sputum samples. In patients,
however, who cannot produce sputum even with induction, bronchoscopy with
BAL may provide the diagnosis. The diagnostic yield from BAL has been
reported to be comparable to that from sputum (51 53). Additional diagnostic
yield has been reported with transbronchial biopsy in conjunction with bronchial
brushings (54). Another report, however, did not nd transbronchial biopsies
helpful in establishing diagnosis (55). Following bronchoscopy, patients fre-
quently continue to produce sputum for one to two days. These samples should
be collected and sent to the laboratory for analysis as they will often be positive
for mycobacteria (56). It should be kept in mind that bronchoscopy in patients
with tuberculosis increases the risk of transmission of disease to medical person-
nel, and appropriate precautions should be taken.
Bronchoscopy is also useful in the diagnosis of infection caused by
non-tuberculous mycobacteria (NTM). The most frequent pulmonary pathogens
include Mycobacterium avium complex, Mycobacterium kansasii, Mycobacterium
abscessus, Mycobacterium fortuitum, Mycobacterium xenopi, and Mycobacterium
malmoense (57). NTM tend to cause disease in elderly patients, patients with
chronic diseases, and immunocompromised hosts. The presence of atypical
mycobacteria in bronchoscopic samples is not diagnostic of disease, however,
in the absence of radiographic and clinical criteria because of their ubiquitous
nature and the high prevalence of colonization (58,59).
Fungal infections have also been diagnosed by bronchoscopy. These
include histoplasmosis, coccidioidomycosis, blastomycosis, and aspergillosis.
In general, the diagnosis of pulmonary histoplasmosis requires a tissue sample
obtained via thoracoscopy or open lung biopsy. This is especially true if
disease is limited to a solitary pulmonary nodule. However, when utilized in
patients with other radiographic features, such as inltrates and cavitary
lesions, bronchoscopy samples are associated with a higher diagnostic yield com-
pared with sputum (60).
Similarly, bronchoscopy samples are more efcacious in the diagnosis of
coccidioidomycosis and blastomycosis compared with sputum. Fungal stains of
BAL samples for Coccidioidomycosis immitis tend to have a low yield, but cul-
tures and transbronchial biopsies have high yields. In addition, serologic tests for
Bronchoscopy 295
coccidioidomycosis may be positive when BAL cultures are negative. Serologic

tests and bronchoscopic studies should, therefore, be considered in unison
(61,62). In patients with blastomycosis, the diagnostic yield of cultures from
bronchoscopic samples is high, with a higher diagnostic efcacy reported with
bronchial washings compared with BAL specimens. The average time for
culture conrmation may, however, take up to ve weeks. Use of wet smears
and cytologic examination of respiratory specimens should, therefore, be per-
formed. Although lower in yield, KOH smears provide an early diagnosis of
blastomycosis when positive (63).
Cryptococcus neoformans is a ubiquitous fungal organism that typically
causes pulmonary, central nervous system, or disseminated disease in immuno-
compromised hosts. In cryptococcosis, fungal stains of BAL samples for the
organism have a higher sensitivity than transbronchial biopsies. No additive
benet is obtained with biopsy specimens if cryptococcus is suspected as the
pathogenic organism (64).
Aspergillus species are likewise ubiquitous in the environment. Aspergillus
fumigatus, Aspergillus avus, and Aspergillus niger are the most frequently
reported pathogens causing a variety of illnesses of the respiratory tract, includ-
ing invasive disease, tracheobronchitis, allergic bronchopulmonary aspergillosis,
and aspergilloma. Bronchoscopy is particularly useful in diagnosing tracheobron-
chitis and invasive disease that tend to occur in immunocompromised hosts.
Tracheobronchitis is typically evident on inspection of the airways where
mucous impaction or pseudomembrane formation is often seen. It is, however,
difcult to denitively ascribe tracheobronchitis to aspergillus infection despite
positive cultures of BAL and washings due to the frequent colonization of the
airways by aspergillus. Denitive diagnosis of invasive pulmonary aspergillosis
requires a tissue sample. However, these patients frequently have contra-
indications to transbronchial or open lung biopsy. Variable diagnostic yields
are seen with both BAL and transbronchial lung biopsies (65,66). In the
appropriate clinical setting (e.g., immunocompromised patient with neutropenia),
isolation of aspergillus from BAL cultures in the absence of other pathogens is
used as a marker for invasive aspergillosis (65,67). The utility of BAL PCR
and detection of galactomannan antigen from BAL samples from patients with
invasive aspergillosis has also been reported (68).
Herpes viruses and cytomegalic virus (CMV) are pathogenic mostly in
immunocompromised hosts. Herpes viruses can rarely cause tracheobronchitis
or pneumonia. Lesions may be visualized as vesicles or shallow ulcers during
bronchoscopy, and cultures from these lesions may yield positive results (69).
CMV isolation from BAL samples is not uncommon. The signicance of
this nding is, however, controversial as it may represent simple colonization
of the lungs (70). In patients with CMV pneumonia, the sensitivity of BAL cul-
tures may be as low as 50% (71). Transbronchial lung biopsy showing evidence
of viral inclusions is more specic for CMV related disease and has a sensitivity
of up to 90% (72).
D. Diffuse Non-Infectious Pulmonary Disease

Bronchoscopy can assist in the diagnosis of various non-infectious diffuse
pulmonary diseases. Techniques utilized will vary according to the suspected
diagnosis. For example, if sarcoidosis is suspected, transbronchial biopsies are
obtained to establish the presence of non-caseating granulomas. If eosinophilic
granuloma is suspected, BAL is performed to look for Langerhans cells. Dissem-
inated carcinoma, on the other hand, can be detected by transbronchial biopsy,
BAL, or brush biopsy. Although bronchoscopy can assist in the diagnosis of
some interstitial lung diseases, thoracoscopic or open lung biopsy remains the
standard method for obtaining adequate tissue samples.
E. Suspected Tracheoesophageal Fistula

Tracheoesophageal stulas most often occur as a result of prolonged intubation
with a cuffed endotracheal or tracheostomy tube. Bronchoscopy or esophago-
scopy can be performed to diagnose and localize tracheoesophageal stulas.
However, small communications may not be visualized, and the diagnosis may
require a barium swallow.
F. Chest Trauma
Blunt chest trauma occasionally results in injury to the tracheobronchial tree.
Patients typically complain of dyspnea and hemoptysis and may develop subcu-
taneous emphysema. Bronchoscopy, in conjunction with radiographic imaging, is
the procedure of choice in establishing the presence and extent of traumatic
airway injury.
G. Thermal Injuries
Thermal injuries are typically the result of exposure to hot smoke from burning
material. Edema, inammation, necrosis, and ulceration of the airways lead to
upper airway obstruction that can develop rapidly, necessitating emergent intuba-
tion. As such, serial bronchoscopy should be considered as edema can be pro-
gressive during the rst 24 hours. However, initial bronchoscopic assessment
does not predict the duration or degree of ventilatory support necessary during
hospitalization (73,74). In addition to damage caused by direct thermal burns,
the lung parenchyma can also be damaged by other products of combustion,
leading to pulmonary edema and contributing further to the mortality of patients
with inhalation injury.
H. Evaluation of Lung Transplant Rejection

Patients who have received lung transplants frequently present with respiratory
symptoms and fever that may represent either an infectious complication or
acute lung rejection. As these patients are immunocompromised and because
Bronchoscopy 297
respiratory failure can evolve rapidly, early diagnostic determination is crucial.

The optimal method to make this determination is to perform bronchoscopy
and to obtain BAL uid and transbronchial biopsies (72). Three to ve biopsies
may be necessary for establishing the diagnosis of acute rejection (75).
IX. Specic Diagnostic and Therapeutic Indications

A. Hemoptysis
Bronchoscopy plays an important role in the diagnostic evaluation and treatment
of patients with hemoptysis. Both focal and diffuse pulmonary processes may
cause hemoptysis, and early diagnosis and treatment are particularly important,
especially in patients with massive hemoptysis. Bronchoscopy is useful in loca-
lizing the site of bleeding and also allows for suctioning of the airways. In
addition, in cases of unilateral bleeding, selective intubation of the uninvolved
side can be facilitated using the exible bronchoscope. Flexible bronchoscopy
allows the operator to examine all lobes and segments of the lungs but is some-
what limited with regard to suction capability. The rigid bronchoscope has a
larger working channel that allows for improved suctioning and may be preferred
in patients with massive hemoptysis. After localizing the site of bleeding and
securing the airway, bronchoscopy can also be used as a therapeutic tool. The
use of topical epinephrine, balloon tamponade, and laser therapy are various treat-
ment modalities discussed in more detail in other sections of this book. If these
modalities are unsuccessful, the bronchoscopist can provide guidance to the inter-
ventional radiologist or thoracic surgeon regarding the source of bleeding (76).
B. Bronchopleural Fistula
Bronchopleural stula is a common complication seen in patients requiring
mechanical ventilation or following lung resection surgery. In patients with per-
sistent air leak, the bronchoscope can be used for localization of the site of leak
and for attempting therapeutic maneuvers. Localization is accomplished by
sequential ination and deation of a Fogarty balloon catheter in segmental
bronchi and watching for cessation of air leak in the water-seal apparatus. Alter-
natively, a bronchogram can be performed by injecting contrast material through
a catheter in the working channel of the bronchoscope. Various agents have been
utilized to seal the airways leading to the stula, including oxidized regenerated
cellulose (Surgicel), brin clot, silver nitrate, methyl-2-cyanoacrylate, tetra-
cycline, and doxycycline. In addition, laser therapy has been used to coagulate
the site of leak (77 80).
C. Foreign Bodies
Foreign body aspiration is a common problem seen in pulmonary medicine.
Aspiration most often occurs in children with equal prevalence of right and
left lung aspiration. In adults, the right side is more often involved than the left
because of the branch angle of the left main stem bronchus. Patients with foreign
bodies in the upper airway may be profoundly symptomatic, whereas those with
passage of the foreign body to more distal airways may be asymptomatic. Follow-
ing imaging studies of the chest, the next diagnostic and treatment modality is
bronchoscopy. In the pediatric population, rigid bronchoscopy is preferred. In
adults, exible bronchoscopy is usually attempted rst and is often successful
in removal of the foreign body with use of ancillary equipment, such as
forceps and wire baskets. In difcult cases, rigid bronchoscopy may be necessary
for extraction of the foreign body and control of the airway (81).
D. Endotracheal Tube Placement and Assessment of Placement
Endotracheal intubation is occasionally challenging in patients with difcult

airways. Patients with signicant amounts of redundant upper airway tissue or
cervical spine injury precluding extension of the neck are particularly difcult
to intubate. In these cases, the bronchoscope serves as an obturator for sliding
the endotracheal tube into place. Additionally, bronchoscopic inspection
ensures appropriate position of the endotracheal tube above the carina.
X. Therapeutic Bronchoscopy: Indications and Options
Alleviation of airway obstruction is the primary indication for therapeutic

bronchoscopy. Mucus plugs and blood clots in the central airway can be
removed using suction and cryotherapy probes, and foreign bodies within the
airway can be removed using forceps, snares, or baskets. In addition, central
airway obstruction caused by tumors, granulation tissues, or strictures can be
managed using laser resection, cryotherapy, electrocautery, photodynamic
therapy, and brachytherapy in conjunction with mechanical dilation using
balloons, bougies or rigid bronchoscopes, and airway stents.
A. Atelectasis
Atelectasis and retained secretions occur commonly among ICU patients.
Approximately 50 to 60% of bronchoscopies performed in the ICU are for
treatment of atelectasis. The efcacy of bronchoscopy for atelectasis varies in
different reports. A recent review found that patients with lobar or segmental
atelectasis benet more from bronchoscopy compared to patients with subseg-
mental atelectasis or air bronchograms on chest roentgenograms. There are,
however, insufcient studies at this time comparing the efcacy of bronchoscopy
to non-invasive treatments, such as chest physiotherapy, mucolytic agents, and
kinetic beds. The decision to proceed with bronchoscopy for atelectasis must
be made on an individual basis (82).
Bronchoscopy 299
B. Debulking of Obstructing Neoplasm

With few exceptions, most therapeutic airway interventions can be performed
using a exible bronchoscope. However, bulky lesions obstructing the central
airway can be managed much easier and faster using a rigid bronchoscope
with less overall risk of complications for the patient. A patient (Fig. 1) had
a complete obstruction of the right main stem bronchus by tumor arising out of
the right upper lobe. The rigid bronchoscope provided an excellent tool for
resecting the tumor in large pieces (Fig. 1D). If a exible bronchoscope had
been used in this patient, it would have taken much longer to evaporate the
tumor using laser heat alone, with the additional risk of bleeding, prolonged
anesthesia, smoke inhalation, and airway ignition. In many therapeutic airway
interventions, however, the exible and the rigid bronchoscopes are complemen-
tary. Although the rigid bronchoscope provides a large ventilating lumen for
therapeutic procedures in the central airway, the exible bronchoscope can aid
in interventions involving smaller lobar or segmental airways, especially in the
upper lobes. Unfortunately, few practicing pulmonologists routinely perform
rigid bronchoscopy. A 1999 survey by the American Association for Bronchology
revealed that 96% of the 744 respondents did not perform rigid bronchoscopy
(83). Only 8% performed laser airway procedures, and only 5% performed
airway stenting.
The rigid bronchoscope is a good conduit for Nd:YAG laser photocoagula-
tion and photoevaporation and is an ideal instrument for performing mechanical
resection of bulky airway tumors. Nd:YAG laser procedures can also be per-
formed through the exible bronchoscope. However, resection of bulky tumors
using the exible bronchoscope relies largely on laser photoevaporation, which
is time consuming. The prolonged laser time increases the risk of airway ignition
and amount of smoke production. An additional disadvantage of the exible
bronchoscope is limited suction capability for evacuation of smoke and blood.
Also, metal stents are the only type of stents that can be inserted through the ex-
ible bronchoscope, whereas placement of all types of stents, including silicon and
metal stents, are possible with the rigid bronchoscope.
Because most rigid bronchoscopy procedures are done with an open
bronchoscope, inhalational anesthesia is generally not advisable. General
anesthesia is usually accomplished with total intravenous anesthesia. Ventilatory
assistance is usually provided by means of positive pressure ventilation using
venturi or jet ventilation and, in some cases, assisted spontaneous ventilation.
The feasibility and ease of using negative pressure ventilation during rigid
bronchoscopy under general anesthesia has also been reported (84).
C. Laser Therapy
Laser bronchoscopy can be performed using either a exible or rigid broncho-
scope as described above. For most airway interventions below the vocal
cords, the Nd:YAG laser is generally the laser of choice. It is important to
Figure 1 A 52-year-old man with known metastatic colon carcinoma who presented
with worsening dyspnea and on chest roentgenography was found to have complete
right lung collapse. (A) Bronchoscopy shows complete obstruction of the right main
stem bronchus at the level of the major carina. This tumor was known to arise out of
the right upper lobe encroaching into the right main stem bronchus. (B) The chest roent-
genogram before tumor resection. (C) The chest roentgenogram immediately following
rigid bronchoscopic resection of tumor with opening of the right middle and lower
lobes and post obstructive pneumonia. (D) The size of tumor fragments resected from
the right main stem bronchus. (Continued.)
Bronchoscopy 301
Figure 1 (Continued.)
understand laser physics and the difference between photocoagulation and

photoevaporation during airway interventions. Photocoagulation is achieved at
low-power density to coagulate blood vessels of the soft tissue or the tumor
for prevention of bleeding. This is usually not accompanied by apparent or sig-
nicant tumor shrinkage or smoke production. Photoevaporation, on the other
hand, is achieved at high-power density to evaporate the soft tissue, resulting
in tissue shrinkage and smoke production. Laser photocoagulation is commonly

used as a prelude to mechanical tumor resection of the central airway using the
rigid bronchoscope. Laser photoevaporation can be used to cut benign strictures
and burn small tumors or the residual tumor bed following mechanical resection
of bulky tumors. Control of bleeding at any stage of the procedure can be accom-
plished readily with laser photocoagulation. Because laser photoevaporation
results in smoke production, smoke evacuation becomes important during the
procedure. The small suction channel of a exible bronchoscope can be easily
overwhelmed by a large amount of smoke if extensive laser photoevaporation
is performed. This is generally not a problem with rigid bronchoscopy due to
the ability to place a large suction catheter through the rigid bronchoscope. In
addition, the risk of re hazard during a laser airway procedure is higher with
the exible bronchoscope because the outer cover of the bronchoscope itself
and many of the common endotracheal tubes used can be ignited by the
laser itself.
For the most part, bronchoscopic interventions for malignant airway
obstruction are used for palliative purposes with a primary objective of improve-
ment in the quality of life of patients with far advanced or end-stage malignancy.
In a series of 2008 patients with malignant airway obstruction who underwent
laser bronchoscopy, median survival was only several months (85). More
recently, bronchoscopic management of tracheobronchial tumor obstruction
with curative intent has been reported (86 88).
Not all airway lesions are suitable for bronchoscopic resection. Best results
are obtained in patients with proximal or central airway tumors with a signicant
endobronchial component. The presence of airway patency distal to the site of
obstruction is necessary. In addition, the obstructed area has to have relatively
uncompromised perfusion. Opening an obstructed bronchus subtending an area
with limited perfusion because of tumor invasion will only result in dead space
ventilation and may potentially worsen hypoxemia. The segment of the
obstructed airway should have minimal vascular invasion in order to minimize
the risk of bronchovascular stulas and fatal hemoptysis after the tumor is suc-
cessfully resected. If laser bronchoscopy is performed with curative intent, the
lesion should be small, centrally located, and with limited base and depth
of invasion.
D. Cryotherapy
Endobronchial cryotherapy works by causing tissue destruction through rapid
freezing followed by slow thawing. Nitrous oxide (N20) gas is used to lower
the temperature to approximately 2898C. Rapid freezing results in the formation
of ice crystals within the cytoplasm that grind and abrade cellular organelles
during thawing, resulting in cell death. Cryo-sensitive tissues include skin,
mucous membranes, granulation tissues, tumors, and other tissues with a signi-
cant amount of water content. Cryo-resistant tissues include fat, cartilage, bone,
Bronchoscopy 303
and brous connective tissue with little water content. The exible cryotherapy
probe can be delivered through a exible or a rigid bronchoscope. The advan-
tages of endobronchial cryotherapy as opposed to Nd:YAG laser therapy or elec-
trocautery include its safety, low cost, ease of use, and the ability to remove
certain foreign bodies and blood clots. The disadvantage is that it is slow in
removing tumor or granulation tissue from the airways, with repeat bronchoscopy
being generally required within a week. Some success has also been reported with
the use of endobronchial cryotherapy with curative intent in the management of
airway carcinoma in situ (87).
E. Electrocautery
Endobronchial electrocautery involves the application of heat to the bronchial

tissue, resulting in tissue coagulation and vaporization. Electrocautery can be
used through either a exible or rigid bronchoscope, and in experienced hands,
results are comparable to that of laser therapy. It can be used as a direct
contact agent in the form of probes, snares, and scissors, or as a non-contact
agent using argon plasma as a gas conductor within an electrical eld [argon
plasma coagulator, (APC)]. APCs unique property is its ability to vaporize
tumors at an acute angle because the argon plasma burns off the closest tissue
within the available electrical eld, even if the closest tissue is at an angle.
This property of the APC is sometimes useful in removing small tumors in the
upper lobe bronchi. Endobronchial electrocautery has also been used to treat
carcinoid tumors with a curative intent (88).
F. Photodynamic Therapy
Photodynamic therapy (PDT) is an effective technique for the management of
small endobronchial tumors. It is used primarily in non-surgical candidates
with small lesions and early stages of lung cancer, with a reported cure rate of
approximately 75% and a recurrence rate of approximately 30% (89). PDT
involves intravenous injection of a tumor photosensitizing agent derived from
hematoporphyrin, which is retained in tumor cells and cleared from most
healthy tissues in six hours except for the lung, reticuloendothelial tissues, and
the skin. Airway tumor cells retain the photosensitizer 48 to 72 hours after its
administration and undergo selective destruction by photo-activation upon
exposure to 630 nm wavelength of laser light introduced in the airway using a
exible bronchoscope. Tumor necrosis occurs as a result of cellular destruction
from superoxide and hydroxyl radicals and vascular occlusions related to throm-
boxane-A2 release (90). A clean-up bronchoscopy two to three days later is
required to remove necrotic tumor debris from the airway. Photodynamic
therapy makes the skin and eyes sensitive to light for six weeks or more after
treatment, and patients are advised to avoid direct sunlight and bright indoor
light during this time period.
G. Brachytherapy
Endobronchial brachytherapy involves delivery of radiation therapy from a
radioactive source placed within or very near the tumor mass. Direct placement
of the radioactive source is accomplished through an airway catheter placed
bronchoscopically. When used in early stage endobronchial cancer, brachyther-
apy produces results comparable to that of photodynamic therapy, with reported
cure rates of 75% to 85% (89). In addition, the use of brachytherapy in conjunc-
tion with Nd:YAG laser treatment increases survival compared with laser therapy
alone (91).
H. Placement of Airway Stents

Airway prostheses are used primarily for the treatment of central airway
obstructions. These include tracheobronchial stenoses secondary to either
extrinsic compressions or intrinsic lesions (tumors and strictures), as well as tra-
chobronchial collapse (tracheomalacia). Although there are several commer-
cially available stents, each has its own advantages and disadvantages. The
design and sophistication of airway stents lag behind vascular stents, and the
ideal airway stent continues to remain elusive (92). An ideal airway stent
should t the airway lumen snugly and remain in place without external x-
ation. It should not cause mucosal, cartilage, or vascular injury and should
only cause minimal granulation tissue formation. It should create satisfactory
expansion of a stricture and should be capable of resisting external compression.
Finally, the stent should be easily tailored to t the entire stricture and should be
easy to remove or replace.
The common airway stents used in North America are Ultraexw,
Wallstentw, Dumonw, Polyexw and Dynamic-Yw. The Ultraex stent and
Wallstent are metallic stents that can be deployed using a exible bronchoscope.
The Dumon stents are silicone stents that are placed using a specially designed
rigid steel delivery device through a rigid bronchoscope. The Polyex stent is
made of a plastic mesh with silicone membrane and is deployed using a plastic
tubular delivery device directly into the airway, guided by a rigid
laryngoscope. The Dynamic-Y stent is made of silicone with interspersed
inverted U-shaped metal struts that mimic the shape of the tracheal cartilage.
The Dynamic-Y stent is used in the trachea and is deployed using a special
delivery forceps guided by a rigid laryngoscope.
The metal stents come as bare stents or covered with a silicone membrane
in the middle. Removal of these metal stents usually requires the use of a rigid
bronchoscope. Due to the growth of bronchial mucosa over the bare part of
these metallic stents, they are considered permanent stents, and they should be
used with extreme caution in benign strictures, especially in young patients
with long life expectancy. When compared with covered metallic stents or sili-
cone stents, the bare metal stents allow for better mucociliary transport but are
more prone to the development of granulation tissue or recurrent tumor growth
Bronchoscopy 305
through the wire mesh. Mucostasis and mucus plug formation tend to occur with
silicone stents and stents covered with a silicone membrane. The edges of the
stents may also elicit formation of granulation tissue, which may potentially
cause future obstruction.
XI. Potential Therapeutic Indications

A. Bronchoscopic Lung Volume Reduction
Advances in our understanding of the physiologic changes associated with lung
volume reduction surgery in patients with emphysema have spurred interest in
the development of bronchoscopic techniques for achieving lung volume
reduction. Ingenito et al. (93) studied bronchoscopic lung volume reduction in
an emphysema sheep model. Fibrin-based glue was introduced bronchoscopi-
cally to collapse, seal, and scar the target lung tissue. Comparable changes in
total lung capacity and residual volume were found between surgical and
bronchoscopic methods of lung volume reduction. Several sterile abscesses
were seen in some of the treated areas. In a follow-up study by the same group
using a modied agent, sterile abscesses were not found (94).
Human studies are now ongoing using bronchoscopically placed one-way
valves (Emphasysw and Spirationw valves) in segmental bronchi to achieve lung
volume reduction (95). These one-way valves allow air and mucus to exit the
lung segment of interest, resulting in passive atelectasis and achievement of
lung volume reduction. The valves are removable bronchoscopically. Prelimi-
nary human studies using the Emphasys valve to achieve non-surgical lung
volume reduction appear encouraging (96). The Spiration valve has been
studied only in animals, and pilot human studies are forthcoming.
B. Creation of Extra Anatomical Airway Tract
Because of the increased work of breathing associated with air trapping in

patients with emphysema, creation of an articial bypass airway tract between
a major airway and the emphysematous pulmonary parenchyma may lead to
improvements in respiratory mechanics by taking advantage of the extensive col-
lateral ventilation present in the emphysematous lung. In a feasibility human
study, lungs of patients scheduled for lobectomy and pneumonectomy were
studied for creation of bypass tracts prior to resection (97). Bronchoscopically-
induced extra anatomical airway tracts were successfully created using a
cautery probe, guided by a Doppler catheter to avoid blood vessels in the bron-
chial wall. An ex vivo study of resected emphysematous lungs showed signicant
improvements in lung mechanics after bypass tracts were successfully formed
between the bronchus and the lung parenchyma using a radiofrequency catheter
and expandable stents placed through a exible bronchoscope (98).
C. Future of Interventional Bronchoscopy

In the future, we may see several exciting new technologies and minimally inva-
sive bronchoscopic interventions. Bronchoscopic interventions on airway tumors
may be performed with curative intent rather than palliative intent. Improvements
in airway stent design may lead to the development of the ideal stent. We will
see more human data on strategies for managing advanced stages of emphysema,
such as articial extra anatomical airway bypass tract and bronchoscopic lung
volume reduction. There is reason for excitement and optimism for the future
of interventional bronchoscopy.
References
1. Patterson EJ. History of bronchoscopy and esophagoscopy for foreign body. The
Laryngoscope 1926; 36:157 175.
2. Jackson C. Bronchoscopy: past, present and future. N Engl J Med 1928; 199:758763.
3. Andersen HA, Fontana RS, Harrison EG, Jr. Transbronchoscopic lung biopsy in
diffuse pulmonary disease. Dis Chest 1965; 48:187 192.
4. Ikeda, S, Yanai N, Ishikawa S. Flexible bronchoberscope. Keio J Med 1968; 17:116.
5. Sokolowski JW Jr, Burgher LW, Jones FL Jr, et al. Guidelines for beroptic broncho-
scopy in adults. Position paper of the American Thoracic Society Board of Directors.
Am Rev Resp Dis 1987; 136:1066.
6. Prakash UBS. Bronchoscopy in the Critical Care Unit. Semin Respir Crit Care Med
1997; 6:583 591.
7. Dajani AS, Taubert KA, Wilson W, Bolger AF, Bayer A, Ferrieri P, Gewitz MH,
Shulman ST, Nouri S, Newburger JW, Hutto C, Pallasch TJ, Gage TW, Levinson
ME, Peter G, Zuccaro G. Prevention of Bacterial Endocarditis. J Am Med Assoc
1997; 277:1794 1801.
8. Prakash UBS, Offord KP, Stubbs SE. Bronchoscopy in North America: The ACCP
Survey. Chest 1991; 100:1668 1675.
9. Kirkpatrick MB. Lidocaine topical anesthesia for exible bronchoscopy. Chest 1989;
96:965 967.
10. British Thoracic Society Guidelines on diagnostic exible bronchoscopy. Thorax
2001; 56(suppl 1):i1 i21.
11. Karim A, Ahmed S, Siddiqui R, Mattana J. Methemoglobinemia complicating topical
lidocaine used during endoscopic procedures. Am J Med 2001; 111:150153.
12. Ricou B, Grandin S, Nicod L, Thorens J, Suter PM. Adult and paediatric size
bronchoscopes for bronchoalveolar lavage in mechanically ventilated patients:
yield and side effects. Thorax 1995; 50(3):290 293.
13. Reynolds, HY. Bronchoalveolar lavage. Am Rev Respir Dis 1987; 135:250 263.
14. American Thoracic Society. Clinical role of bronchoalveolar lavage in adults with
pulmonary disease. Am Rev Resp Dis 1990; 142:481 486.
15. Popovich J Jr, Kvale PA, Eichenhorn MS, Radke JR, Ohorodnik JM, Fine G.
Diagnostic accuracy of multiple biopsies from exible beroptic bronchoscopy.
Am Rev Resp Dis 1982; 125:521 523.
16. de Fenoyl O, Capron F, Lebeau B, Rochemaure J. Transbronchial biopsy without
uoroscopy: A ve year experience in outpatients. Thorax 1989; 44:956 959.
Bronchoscopy 307
17. Zavala DC. Transbronchial biopsy in diffuse lung disease. Chest 1978; 73(suppl):
727 733.
18. Zavala DC. Flexible Fiberoptic Bronchoscopy: A Training Handbook. Iowa City:
University of Iowa 1978.
19. Wang KP, Terry PB. Transbronchial needle aspiration in the diagnosis and staging of
bronchogenic carcinoma. Am Rev Respir Dis 1983; 127:344 347.
20. Herth F, Becker HD, Ernst A. Conventional vs endobronchial untrasound-guided
transbronchial needle aspiration: a randomized trial. Chest 2004; 125:322 325.
21. Lam S, Kennedy T, Unger M, Miller YE, Gelmont D, Rusch V, Gipe B, Howard D,
LeRiche JC, Coldman A, Gazdar AF. Localization of bronchial intraepithelial
neoplastic lesions by uorescence bronchoscopy. Chest 1998; 113:696 702.
22. Suratt PM, Smiddy JF, Gruber B. Deaths and complications associated with
beroptic bronchoscopy. Chest 1976; 69:747 751.
23. Credle WF, Smiddy JF, Elliott RC. Complications of beroptic bronchoscopy. Am
Rev Respir Dis 1974; 109:67 72.
24. Simpson FG, Arnold AG, Purvis A, et al. Postal survey of bronchoscopic practice by
physicians in the United Kingdom. Thorax 1986; 41:311 317.
25. Pereira W Jr, Kovnat DM, Snider GL. A prospective cooperative study of
complications following exible beroptic bronchoscopy. Chest 1978; 73:813 816.
26. Lukomsky GI, Ovchinnikov AA, Bilal A. Complications of bronchoscopy: com-
parison of rigid bronchoscopy under general anesthesia and exible beroptic
bronchoscopy under topical anesthesia. Chest 1981; 79:316 321.
27. Gellert AR, Rudd RM, Sinha G. Fiberoptic bronchoscopy: effect of multiple biopsies
on diagnostic yield in bronchial carcinoma. Thorax 1982; 37:684 687.
28. Zavala D. Diagnostic beroptic bronchoscopy: techniques and results of biopsy in
600 patients. Chest 1975; 68:12 19.
29. Salzman SH. Bronchoscopic techniques for the diagnosis of pulmonary compli-
cations of HIV infection. Semin Respir Infect 1999; 14:318 326.
30. Wimberley N, Willey S, Sullivan N, Bartlett JG. Antibacterial properties of
lidocaine. Chest 1979; 76:37 40.
31. Ramsey AH, Oemig TV, Davis JP, Massey JP, Torok TJ. An outbreak of broncho-
scopy-related Mycobacterium tuberculosis infections due to lack of bronchoscope
leak testing. Chest 2002; 121:976 981.
32. Pseudomonas aeruginosa infections associated with defective bronchoscopes.
MMWR 2002; 51:190.
33. Dakin J, Grifths M. Pulmonary investigations for acute respiratory failure. Thorax
2002; 57:79 85.
34. Chastre J, Fagon JY. Ventilator-associated pneumonia. Am J Respir Crit Care Med
2002; 165:867 903.
35. Craven DE, De Rosa FG, Thornton D. Nosocomial pneumonia: emerging
concepts in diagnosis, management, and prophylaxis. Curr Opin Crit Care 2002:8:
421 429.
36. American Thoracic Society. Hospital-acquired pneumonia in adults. Diagnosis,
assessment of severity, initial antimicrobial therapy, and preventive strategies. A
consensus statement. Am J Respir Crit Care Med 1995; 153:1711 1725.
37. Fagon JY, Chastre J, Hance AJ, Guiguet M, Trouillet JL, Domart Y, Pierre J, Gibert
C. Detection of nosocomial lung infection in ventilated patients: use of a protected
specimen brush and quantitative culture techniques in 147 patients. Am Rev Respir
Dis 1988; 138:110 116.
38. Meduri GU. Diagnosis and differential diagnosis of ventilator-associated pneumo-
nia. Clin Chest Med 1995; 16:61 93.
39. Meduri GU, Chastre J. The standardization of bronchoscopic techniques for
ventilator-associated pneumonia. Chest 1992; 102(suppl):557 564.
40. Grifn JJ, Meduri GU. New approaches in the diagnosis of nosocomial pneumonia.
Med Clin North Am 1994; 78:1091 1122.
41. Baselski VS, El-Torky M, Coalson JJ, Grifn JP. The standardization of criteria for
processing and interpreting laboratory specimens in patients with suspected venti-
lator-associated pneumonia. Chest 1992; 102(suppl):571 579.
42. Fagon JY, Chastre J, Wolff M, Gervais C, Parer-Aubas S, Stephan F, Similowski T,
Mercat A, Diehl JL, Sollet JP, Tenaillon A. Invasive and noninvasive strategies for
management of suspected ventilator-associated pneumonia: a randomized trial. Ann
Intern Med 2000; 132:621 630.
43. Ruiz M, Tores A, Ewig S, Marcos MA, Alcon A, Lledo R, Asenjo MA, Maldonaldo
A. Noninvasive versus invasive microbial investigation in ventilator-associated
pneumonia: evaluation of outcome. Am J Respir Crit Care Med 2000; 162:119 125.
44. Gal AA, Klatt EC, Koss MN, Strigle SM, Boylen CT. The effectiveness of broncho-
scopy in the diagnosis of Pneumocystis carinii and Cytomegalovirus pulmonary
infections in acquired immunodeciency syndrome. Arch Pathol Lab Med 1987;
111:238 241.
45. Golden JA, Hollander H, Stulbarg MS, Gamsu G. Bronchoalveolar lavage as the
exclusive diagnostic modality for Pneumocystis carinii pneumonia. A prospective
study among patients with acquired immunodeciency syndrome. Chest 1986;
90:18 22.
46. Broaddus C, Dake MD, Stulbarg MS, Blumenfeld W, Hadley WK, Golden JA,
Hopewell PC. Bronchoalveolar lavage and transbronchial biopsy for the diagnosis
of pulmonary infections in the acquired immunodeciency syndrome. Ann Intern
Med 1985; 102:747 752.
47. Jules-Elysee KM, Stover DE, Zaman MB, Bernard EM, White DA. Aerosolized pen-
tamidine: effect on diagnosis and presentation of Pneumocystis carinii pneumonia.
Ann Intern Med 1990; 112:750 757.
48. Levine SJ, Kennedy D, Shelhamer JH, Kovacs A, Feuerstein IM, Gill VJ, Stock F,
Solomon D, Boylen CT, Masur H, Ognibene FP. Diagnosis of Pneumocystis
carinii pneumonia by multiple lobe, site-directed bronchoalveolar lavage with
immunouorescent monoclonal antibody staining in human immunodeciency
virus-infected patients receiving aerosolized pentamidine chemoprophylaxis. Am
Rev Respir Dis 1992; 146:838 843.
49. Metersky ML, Harrell JH, Moser KM. Lack of utility of bronchial brush biopsy
in patients infected with the human immunodeciency virus. Chest 1992;
101:680 683.
50. American Thoracic Society. Diagnostic standards and classication of tuberculosis
in adults and children. Am J Respir Crit Care Med 2000; 161:1376 1395.
51. Anderson C, Inhaber N, Menzies D. Comparison of sputum induction with ber-
optic bronchoscopy in the diagnosis of tuberculosis. Am J Respir Crit Care Med
1995; 152:1570 1574.
Bronchoscopy 309
52. Conde MB, Soares SL, Mello FC, Rezende VM, Almeida LL, Reingold AL, Daley
CL, Kritski AL. Comparison of sputum induction with beroptic bronchoscopy in
the diagnosis of tuberculosis. Am J Respir Crit Care Med 2000; 162:2238 2240.
53. McWilliams T, Wells AU, Harrison AC, Lindstrom S, Cameron RJ, Foskin E.
Induced sputum and bronchoscopy in the diagnosis of pulmonary tuberculosis.
Thorax 2002; 57:1010 1014.
54. Palenque E, Amor E, Bernaldo de-Quiros JC. Comparison of bronchial washing,
brushing and biopsy for diagnosis of pulmonary tuberculosis. Eur J Clin Microbiol
1987; 6:191 192.
55. Stenson W, Aranda C, Bevelaque FA. Transbronchial biopsy culture in pulmonary
tuberculosis. Chest 1983; 83:883 884.
56. Sarkar SK, Sharma GS, Gupta PR, Sharma RK. Fiberoptic bronchoscopy in the diag-
nosis of pulmonary tuberculosis. Tubercle 1980; 61:97 99.
57. American Thoracic Society. Diagnosis and treatment of disease caused by non-TB
mycobacteria. Am J Respir Crit Care Med 1997; 156(suppl):1 25.
58. Good RC, Snider DE. Isolation of nontuberculous mycobacteria in the United States,
1980. J Infect Dis 1982; 146:829 833.
59. OBrien RJ, Geiter LJ, Snider DE. The epidemiology of nontuberculous mycobacter-
ial diseases in the United States: results from a national survey. Am Rev Respir Dis
1987; 135:1007 1014.
60. Prechter GC, Prakash UBS. Bronchoscopy in the diagnosis of pulmonary histopla-
mosis. Chest 1989; 95:1033 1036.
61. Sobonya RE, Barbee RA, Wiens J, Trego D. Detection of fungi and other pathogens
in immunocompromised patients by bronchoalveolar lavage in an area endemic for
coccidioidomycosis. Chest 1990; 97:1349 1355.
62. DiTomasso JP, Ampel NM, Sobonya RE, Bloom JW. Bronchoscopic diagnosis of
pulmonary coccidioidomycosis. Comparison of cytology, culture, and transbronchial
biopsy. Diagn Microbiol Infect Dis 1994; 18:83 87.
63. Martynowicz M, Prakash UBS. Pulmonary blastomycosis: an appraisal of diagnostic
techniques. Chest 2002; 121:768 773.
64. Malabonga VM, Basti J, Kamholz SL. Utility of bronchoscopic sampling techniques
for cryptococcal disease in AIDS. Chest 1991; 99:370 372.
65. Kahn FW, Jones JM, England DM. The role of bronchoalveolar lavage in the diag-
nosis of invasive pulmonary aspergillosis. Am J Clin Pathol 1986; 86:518 523.
66. Albelda SM. Talbot GH, Gerson SL, Miller WT, Cassileth PA. Role of beroptic
bronchoscopy in the diagnosis of invasive pulmonary aspergillosis in patients with
acute leukemia. Am J Med 1984; 76:1027 1034.
67. Lortholary O, Meyoha MC, Dupont B, Cadranel J, Salmon-Ceron D, Peyramond D,
Simonin D. Invasive aspergillosis in patients with acquired immunodeciency syn-
drome: Report of 33 cases. Am J Med 1993; 95:171 187.
68. Sanguinetti M, Posteraro B, Pagano L, Pagliari G, Fianchi L, Mele L, La Sorda M,
Franco A, Fadda G. Comparison of real-time PCR, conventional PCR, and galacto-
mannan antigen detection by enzyme-linked immunosorbent assay using bronchoal-
veolar lavage uid samples from hematology patients for diagnosis of invasive
pulmonary aspergillosis. J Clin Micro 2003; 41:3922 3925.
69. Graham BS, Snell JD Jr. Herpes simplex virus infection of the adult lower respiratory
tract. Medicine 1983; 62(6):384 393.
70. Mann M, Shelhamer J, Masur H, Gill VJ, Travis W, Solomon D, Manischewitz J,

Stock F, Lane HC, Ognibene FP. Lack of clinical utility of bronchoalveolar
lavage cultures for cytomegalovirus in HIV infection. Am J Respir Crit Care Med
1997; 155:1723 1728.
71. Wallace JM, Hannah J. Cytomegalovirus pneumonitis in patients with AIDS. Chest
1987; 92:198 203.
72. Trulock EP, Ettinger NA, Brunt EM, Pasque MK, Kaiser LR, Cooper JD. The role of
transbronchial lung biopsy in the treatment of lung transplant recipients: an analysis
of 200 consecutive procedures. Chest 1992; 102:1049 1054.
73. Moylan JA, Adib K, Birnbaum M. Fiberoptic bronchoscopy following thermal
injury. Surgery, Gynecol Obst 1975; 140(4):541 543.
74. Bingham HG, Gallagher TJ, Powell MD. Early bronchoscopy as a predictor of
ventilatory support for burned patients. J Trauma 1987; 27(11):1286 1288.
75. Tazelaar HD, Nilsson FN, Rinaldi M, Murtaugh P, McDougall JC, McGregor CGA.
The sensitivity of transbronchial biopsy for the diagnosis of acute lung rejection.
J Thorac Cardiovasc Surg 1993; 105:674 678.
76. Lordan JL, Gascoigne A, Corris PA. Assessment and management of massive
haemoptysis. Thorax 2003; 58:814 819.
77. Sprung J, Krasna MJ, Yun A, Thomas P, Bourke D. Treatment of a bronchopleural
stula with a Fogarty catheter and oxidized regenerated cellulose (Surgicel). Chest
1994; 105:1879 1881.
78. Nicholas JM, Dulchavsky SA. Successful use of autologous brin gel in traumatic
bronchopleural stula: case report. J Trauma 1992; 32:87 88.
79. Wang KP, Schaeffer L, Heitmiller R, Baker R. Nd:YAG laser closure of a broncho-
pleural stula. Monaldi Arch Chest Dis 1993; 48:301.
80. Varoli F, Roviaro G, Grignani F, Vergani C, Maciocco M, Rebuffat C. Endoscopic
treatment of bronchopleural stulas. Ann Thorac Surg 1998; 65:807 809.
81. Dikensoy O, Usalan C, Filiz A. Foreign body aspiration: clinical utility of exible
bronchoscopy. Postgrad Med J 2002; 78:399 403.
82. Kreider ME, Lipson DA. Bronchoscopy for atelectasis in the ICU: a case report and
review of the literature. Chest 2003; 124:344 350.
83. Colt HG, Prakash UBS, Offord KP. Bronchoscopy in North America: survey by the
American association for bronchology, 1999. J Bronchol 2000; 7:8 25.
84. Natalini G, Cavaliere S, Seramondi V, Foccoli P, Vitacca M, Ambrosino N,
Candiani A. Negative pressure ventilation vs external high-frequency oscilla-
tion during rigid bronchoscopy: a controlled randomized trial. Chest 2000; 118:
18 23.
85. Cavaliere S, Venuta F, Foccoli P, Toninelli C, La Face B. Endoscopic treatment of
malignant airway obstructions in 2008 patients. Chest 1996; 110:1536 1542.
86. Cavaliere S, Foccoli P, Toninelli C. Curative bronchoscopic laser therapy for
surgically resectable tracheobronchial tumors: personal experience. J Bronchol
2002; 9:90 95.
87. Deygas N, Froudarakis M, Ozenne G, Vergnon JM. Cryotherapy in early supercial
bronchogenic carcinoma. Chest 2001; 120:26 31.
88. Van Boxem TJ, Golding RP, Venmans BJ, Postmus PE, Sutedja TG. High resolution
CT in patients with intraluminal typical bronchial carcinoid tumors treated with
bronchoscopic therapy. Chest 2000; 117:125 128.
Bronchoscopy 311
89. Mathur PN, Edell E, Sutedja T, Vergnon JM. Treatment of early stage non-small cell
lung cancer. Chest 2003; 123:176s 180s.
90. Edell ES, Cortese DA. Photodynamic therapy: its use in the management of broncho-
genic carcinoma. Clin Chest Med 1995; 16:455 463.
91. Shea JM, Allen RP, Tharratt RS, Chan AL, Siefkin AD. Survival of patients
undergoing Nd:YAG laser therapy compared with Nd:YAG laser therapy and
brachytherapy for malignant airway disease Chest 1993; 103:1028 1031.
92. Colt HG, Dumon JF. Airway stents: present and future. Clin Chest Med 1995;
16:465 478.
93. Ingenito EP, Reilly JJ, Mentzer SJ, Swanson SJ, Vin R, Keuhn H, Berger RL,
Hoffman A. Bronchoscopic volume reduction: a safe and effective alternative to
surgical therapy for emphysema. Am J Respir Crit Care Med 2001; 164:295 301.
94. Ingenito EP, Berger RL, Henderson AC, Reilly JJ, Tsai L, Hoffman A. Broncho-
scopic lung volume reduction using tissue engineering principles. Am J Respir
Crit Care Med 2003; 167:771 778.
95. Maxeld RA. New and emerging minimally invasive techniques for lung volume
reduction. Chest 2004; 125:777 783.
96. Snell GI, Holsworth L, Borrill ZL, Thomson KR, Kalff V, Smith JA, Williams TJ.
The potential for bronchoscopic lung volume reduction using bronchial prostheses.
Chest 2003; 124:1073 1080.
97. Rendina EA, De Giacomo T, Venuta F, Coloni GF, Meyers BF, Patterson GA,
Cooper JD. Feasibility and safety of the airway bypass procedure for patients with
emphysema. J Thorac Cardiovasc Surg 2003; 125:1294 1299.
98. Lausberg HF, Chino K, Patterson GA, Meyers BF, Toeniskoetter PD, Cooper JD.
Bronchial fenestration improves expiratory ow in emphysematous human lungs.
Ann Thorac Surg 2003; 75:393 397.
9
Percutaneous Tracheostomy
AMMAR SAKKOUR and IRAWAN SUSANTO
UCLA Santa Monica Specialties

Santa Monica, California, U.S.A.
I. Introduction
Tracheostomy is one of the frequent surgical procedures in critically ill patients.

Open surgical tracheostomy (OST) has traditionally been performed by surgeons
in the operating room. In the past half-century, several methods of percutaneous
tracheostomy have been introduced. Over the last decade, percutaneous tracheo-
stomy has been performed primarily in the intensive care unit (ICU) on patients
who require a prolonged articial airway because of their critical illness or
progressive chronic lung disease.
In 1909, Jackson described the standardized surgical technique and indi-
cations for open tracheostomy (1). In 1957, Sheldon et al. (2) rst described per-
cutaneous tracheostomy. He introduced a special slotted needle percutaneously
into the trachea. The needle served to direct a three-bladed cutting stylet that
was combined with a metal tracheostomy tube. Because of several deaths, the
technique was abandoned.
In 1969, Toye and Weinstein (3) used the Seldinger technique for tracheo-
stomy placement. After cannulating the trachea with a needle, a tracheostomy tube
was loaded onto a stiff wire that contained a recessed blade. The device was
advanced into the needle tract until the tracheostomy tube was placed. Because
of the high risk of paratracheal insertion, the technique never gained popularity.
The most popular technique worldwide today is percutaneous dilatational
tracheostomy (PDT), described by Ciaglia et al. in 1985 (4) as a method utilizing
serial dilators (later modied to a single tapered single dilator) over a guidewire
313
314 Sakkour and Susanto
in a modied Seldinger technique. PDT is most commonly performed at bedside

in the ICU with or without bronchoscopic guidance (5). In the past decade, most
of the published literature on percutaneous tracheostomy has dealt with this
technique.
There are also other reported methods for tracheostomy placement.
In 1989, Schachner et al. (6) used a dilating tracheostome via the
Seldinger technique. The device had blades that slid over the guidewire into
the trachea, where they were forced open to make a stoma that would t
a tracheostomy tube (7). This method had a high incidence of posterior
tracheal wall tears, which led to its withdrawal from the United States
market.
In 1990, Griggs et al. (8) described tracheostomy with a grooved
Howard Kelly forceps over a guidewire using the Seldinger technique.
Once the tip of the forceps was in the trachea, the forceps were opened to
create a stoma. Sviri et al. (9) followed the long-term outcomes of percuta-
neous tracheostomy using the Griggs technique. Thirty-eight percent of the
patients complained of some degree of voice change, 12% of ongoing
severe cough, and 31% of shortness of breath, with more than half of these
having concomitant heart or lung disease. In 19.5%, there was some evidence
of upper airway obstruction on spirometry, but only 5% were symptomatic.
Borm and Gleixner (10) found bedside percutaneous tracheostomy with the
Griggs system safe and effective in patients with critical neurosurgical
disease, if done under bronchoscopic control. They also found no increase
in intracranial pressure. Anon et al. (11) observed similar rates of perioperative
complications comparing the Ciaglia and Griggs techniques. Van Heurn and
Nates (12,13) in separate studies found signicantly lower rate of compli-
cations in the Ciaglia technique. Today, the Griggs technique remains
popular in certain parts of the world.
In 1993, Fantoni and Ripamonti (14) described a translaryngeal tracheost-
omy method. In this approach, the endotracheal tube is pulled to the subglottic
area. Then, a guidewire needle is placed in the rst or second tracheal interspace,
and a guidewire is passed through the needle and directed to the mouth along-
side the endotracheal tube, similar to the technique of retrograde intubation.
The endotracheal tube is replaced with a smaller one. The guidewire that exits
the mouth is connected to the pointed tip of a special trocar-tracheostomy tube
combination. The guidewire is then pulled from the anterior tracheal side,
causing the trocar-tracheostomy tube to be pulled through the mouth and the
vocal cord. The tube nally exits the anterior tracheal wall. The relative complex-
ity of this method has kept it from gaining wider acceptance beyond parts of
Europe.
In 1998, Ciaglia rened his percutaneous dilatational technique to a single-
step dilation instead of multiple dilation using a single-tapered dilator coated
with a hydrophilic agent, named the Blue Rhinow. Since its introduction,
Percutaneous Tracheostomy 315
users of the Ciaglia PDT technique have largely adopted this Blue Rhino
technique over the serial dilator technique.
II. Indications, Benets, and Timing of Tracheostomy
Tracheostomy is indicated for critically ill patients who require prolonged

mechanical ventilation, for the relief of upper airway obstruction after initial
stabilization, or for airway protection from massive aspiration.
Converting translaryngeal intubation to tracheostomy in ventilator-
dependent patients offers several potential advantages (Table 1). It provides a
more secure airway and enhances the patients mobility, which is limited in
patients with translaryngeal intubation because of the risk of inadvertent extuba-
tion (15). Patients with tracheostomy have 1% incidence of inadvertent decannu-
lation, compared with 8% to 21% in patients with translaryngeal intubation.
Inadvertent extubation may lead to adverse cardiopulmonary and hemodynamic
events in a signicant number of cases (16,17).
Because tracheostomy is better tolerated in mentally alert patients, it
lessens the need for heavy sedation that has been associated with an increased
risk of nosocomial pneumonia (18). It also facilitates airway suctioning, oral
hygiene, and oral nutrition.
Tracheostomy may also expedite weaning in patients with borderline ven-
tilatory capacity by reducing the airway resistance (19). It facilitates early trans-
fer from ICU before the completion of weaning from mechanical ventilation.
In addition, tracheostomy may help avoid some of the complications of
prolonged translaryngeal intubation, including vocal cord paralysis, glottic and
subglottic stenosis, sinusitis, otitis, and tracheal injury, such as tracheomalacia
and tracheal stenosis (20 22) (Table 2).
The decision on the timing of tracheostomy should be individualized and
should take into account the anticipated duration of mechanical ventilation
after the initial period of stabilization and therapeutic trial. It should also take
into account patient preferences, the likelihood of benets versus harm, and
Table 1 Benets of Converting Translaryngeal Intubation

to Tracheostomy in ICU Patients
Sparing further laryngeal injury

Facilitating nursing care and airway suctioning
Increasing patient mobility by providing a more secure tube
Facilitating transfer from the ICU setting
Improving comfort
Permitting early return of speech
Facilitating oral feeding
Decreasing airway resistance
Table 2 Complications of Translaryngeal Intubation
Laryngeal injury
Vocal cord paralysis
Glottic and subglottic stenosis
Infectious complications: sinusitis, otitis, pneumonia
Tracheal injury (tracheomalacia, tracheal dilation, and
tracheal stenosis)
the contraindications at any particular time in the clinical course of the critically
ill patient. Some evidence suggests that the risk of long-term airway compli-
cations signicantly increase beyond the 10th day of translaryngeal intubation
(23). Many of the studies that compared outcomes of patients converted at
various times in their clinical course to tracheostomy vary in design and sampling
and suffer from multiple design problems (24). Armstrong et al. (25) found that
early use of tracheostomy in blunt trauma patients with ventilator dependence
reduces length of stay in ICU and in the hospital with no adverse effect on mor-
bidity and mortality rates. However, Sugerman et al. (26) failed to show any sig-
nicant difference with early tracheostomy. Brook et al. (27) concluded in their
study that early tracheostomy was associated with shorter hospital stay, lower
hospital cost, shorter length of stay in the ICU, and shorter duration of mechan-
ical ventilation. Kluger et al. (28) found that tracheostomy performed between
day 0 and day 3 of mechanical ventilation may decrease the incidence of pneu-
monia in critically ill trauma patients. A recent prospective randomized study of
120 critically ill medical patients by Rumbak et al. (29) demonstrated that early
(within 48 hr) percutaneous tracheostomy was associated with signicantly lower
rates of mortality, pneumonia, and accidental extubations compared with pro-
longed translaryngeal intubation, in which tracheostomy was performed on day
14 to day 16.
The 1989 ACCP consensus conference on articial airway in mechanically
ventilated patients recommended tracheostomy in patients whose anticipated
need for articial airway is more than 21 days. It also recommended that once
the decision for tracheostomy is made, the procedure should be performed as
early as possible to minimize the duration of translaryngeal intubation, based
on individual patient factors (30).
III. Contraindications
The relative and absolute contraindications for percutaneous tracheostomy

include uncorrectable coagulopathy, infection at the tracheostomy site, previous
neck surgery, non-intubated patients, infants, thyromegaly, emergent setting,
midline neck mass, positive end-expiratory pressure (PEEP) higher than
10 cmH2O, limited neck extension, poor oxygenation with FIO2 higher than
60%, or hemodynamic instability (Table 3). The line between relative and
Table 3 Relative Contraindications to Percutaneous

Dilatational Tracheostomy in the ICU
Immature airway in children

Emergency airway access
Anatomic abnormality of the trachea (e.g., midline neck mass)
Palpable blood vessel over the tracheostomy site
Active infection over the planned tracheostomy site
Short or obese neck
FIO2 .0.6
Positive end-expiratory pressure .10 cmH2O
Platelet count ,50,000/mm3
Bleeding time .10 min
Serum creatinine .3.0 mg/dL
PT or PTT .1.5 times control
Limited ability to extend the cervical spine or unstable neck
Abbreviations: PT, prothrombin time; PTT, partial thromboplastin time.
absolute, and what is feasible and what is not, becomes less clear with the increas-
ing experience and in the hands of highly skilled operators. Beiderlinden et al.
(31) found that bronchoscopically guided PDT in patients on high PEEP did
not jeopardize oxygenation one hour and 24 hours following PDT (31). Meyer
et al. (32) found that trained physicians can safely perform bedside percutaneous
tracheostomy in patients who had undergone previous tracheostomy. Although
not considered a standard practice, several studies have reported that emergency
percutaneous tracheostomy is feasible and safe (33,34). Morbid obesity has
traditionally been a contraindication for percutaneous tracheostomy because of
the presumed difculty in identifying landmarks. This assumption was never sup-
ported by any trials. In experience, percutaneous tracheostomy in morbidly obese
patients can be performed safely as long as proper precautions and adjustments
are made, including using a longer tracheostomy tube to accommodate the
additional soft tissue of the neck. Mansharamani et al. (35) reported 13
consecutive cases of PDT in patients with a body mass index of 28 to 62.
They reported no complications, technical difculty, or failure to place the
tracheostomy. Laryngeal mask airway has been shown to be as an effective
and reliable alternative to endotracheal intubation as a temporary articial
airway during percutaneous tracheostomy (36,37).
IV. The Technique of Bedside PDT: How We Do It
The most popular percutaneous tracheostomy technique worldwide is the

Ciaglia PDT (Figs. 1 5). Keep in mind, like any surgical technique, the PDT
itself evolves over time, and each practitioner develops and adapts the technique
Figure 1 A large bore needle with a catheter sheath over it is introduced into the tra-
cheal lumen at the level of the rst tracheal interspace. Source: Courtesy of Cook Critical
Care, Inc., Bloomington, Indiana, U.S.A.
in individual ways. We will focus on describing the Ciaglia PDT the way we
do it.
The preparatory steps before PDT are just as important as the tracheostomy
itself. Proper patient preparation cuts down on the potential complications that
may arise during the procedure. The mechanical ventilator settings are adjusted
Figure 2 After the needle is removed, a guidewire is inserted through the catheter
sheath into the tracheal lumen. Once the guidewire is in place, the catheter is removed.
Source: Courtesy of Cook Critical Care, Inc., Bloomington, Indiana, U.S.A.
Figure 3 Following dilation of the anterior tracheal wall with an 11 French dilator
(not shown), a guiding catheter is introduced over the guidewire. Source: Courtesy of
Cook Critical Care, Inc., Bloomington, Indiana, U.S.A.
Figure 4 A dilator is passed over the guiding catheter and guidewire to dilate the
anterior tracheal wall. Dilation can be achieved serially using a progressively larger
serial dilators or a single tapered dilator (Blue Rhinow). Source: Courtesy of Cook
Critical Care, Inc., Bloomington, Indiana, U.S.A.
Figure 5 A tracheostomy tube preloaded onto a tapered dilator/trocar is inserted over

the guiding catheter and guidewire into the trachea. The dilator/trocar, the guiding
catheter, and the guidewire are then removed, leaving the tracheostomy tube in place.
Source: Courtesy of Cook Critical Care, Inc., Bloomington, Indiana, U.S.A.
to an FIO2 of 1.0, with appropriate alarm settings, and placed on volume control
with appropriate rate and tidal volume to maintain adequate basal minute venti-
lation when a paralytic agent is used. The high-pressure limit is increased to
accommodate the increase in peak airway pressure due to the presence of the
bronchoscope in the endotracheal tube. Rolled towels are placed behind the
patients back in a supine position to extend the head and neck. The blood
pressure, heart rate, and pulse oximetry are continuously monitored throughout
the procedure. The surface markers of the trachea are carefully palpated and
marked. The rst or second tracheal interspace is the preferred tracheostomy
site. The subcricoid space may be used in patients with short neck or severe
kyphosis. We avoid the space below the third tracheal ring to decrease the risk
of accidental erosion into the innominate artery and minimize the potential
trauma to the thyroid isthmus. The skin is prepped and draped in a sterile fashion.
The patient is premedicated with a combination of intravenous benzo-
diazepine and opiate analgesics, for example, meperidine, morphine, or fentanyl.
Alternatively, the patient could be maintained on intravenous proprofol. A non-
depolarizing neuromuscular blocking agent is also used. A 2-cm skin incision
is made vertically or transversely at midline over the selected site, at least one
ngerbreadth above the suprasternal notch. A blunt dissection is then done
using a curved clamp until the pretracheal fascia is felt. The left middle nger
and thumb are used to secure the lateral edges of the trachea, while the index
nger is used to palpate the anterior tracheal wall to select the appropriate
tracheal interspace for insertion site. A exible bronchoscope is introduced
into the endotracheal tube. The tip of the bronchoscope is kept inside the tube
to prevent damage by the needle. The cuff of the endotracheal tube is then
deated, and the tube is slowly withdrawn to the subglottic space, guided by
the transillumination of the anterior trachea. Alternatively, the proper position of

the endotracheal tube can be bronchoscopically estimated by pushing on the
anterior tracheal wall using the index nger of the left hand and visualizing
the anterior tracheal lumen wall depression through the bronchoscope. Once
the rst or second tracheal interspace is identied, the introducer needle is
advanced into the tracheal lumen with continuous suction (Fig. 1). This step
and all subsequent steps are guided under direct bronchoscopic visualization.
The catheter sheath over the needle is then advanced, and the needle is
removed. A guidewire is then passed through the catheter (Fig. 2), followed by
passage of a short 11 French dilator. An 8 French guiding catheter is then
placed over the guidewire (Fig. 3). Dilatation of the stoma is performed either
by serial dilators or by using the tapered Blue Rhino single dilator. This dilator
has a hydrophilic coating that is activated by wetting the surface prior to use
(Fig. 4). The stoma is commonly dilated to 36 or 38 French. Finally, the
tracheostomy tube, which is tted with a curved dilator/trocar, is passed over
the guiding catheter and guidewire into the tracheal lumen (Fig. 5). After the
dilator/trocar is removed, the intratracheal position of the tracheostomy tube is
conrmed by passing the bronchoscope through the tracheostomy tube and
observing the normal tracheal anatomy. Any active bleeding into the trachea
should be noted and suctioned. It is a good practice to keep the endotracheal
tube in position and not remove it until the intratracheal position of the tracheost-
omy tube has been conrmed. Once the position is conrmed, the tracheostomy
tube is connected to the ventilator tubing and the cuff is inated. The tracheost-
omy tube is then secured with a Velcro neck strap, and the tube is sutured to skin
using 2-0 silk.
After the tracheostomy tube is successfully placed, we routinely examine
the suprastomal area, the subglottic space, and the vocal cords to record the
damages and abnormalities associated with the endotracheal intubation. To do
this, the bronchoscope is introduced through the endotracheal tube again. The
cuff of the endotracheal tube is then deated and the tube is pulled out over
the bronchoscope. The bronchoscope is then slowly pulled back from the supras-
tomal intratracheal position while carefully observing any abnormalities of the
proximal trachea, the subglottic space, the vocal cords with particular attention
to the posterior commissure, and the supraglottic structures. A chest X-ray is
done only if there is suspected pneumothorax or pneumomediastinum in our
clinical exam. A recent retrospective study found that routine chest radiography
following PDT that had been performed under bronchoscopic visualization
was unnecessary in the absence of clinical deterioration or the suspected post-
operative complications (38).
V. Complications
In general, the complication rates of PDT compares favorably to that of OST.

Mortality of tracheostomy is less than 1% in several large series (39 43). The
most common complications of PDT are bleeding, infection, and tracheal

stenosis (Table 4). In a series of 356 patients, Hill et al. (44) reported a mortality
rate of 0.3% and a 3.7% rate of symptomatic tracheal stenosis. Marx et al. (45)
reported an operative mortality of 0.39%, bleeding of 1.2%, infection of 0.2%,
tracheal stenosis of 0.8%, and pneumothorax of 0.4%. On the other end of
the spectrum, Friedman and Mayer (46) reported a 5% rate of bleeding, a 3%
rate of infection, and a 4% rate of asymptomatic stenosis.
The risk of bleeding is low in PDT because of the tamponading effect of the
tracheostomy tube and the dilators. Major bleeding complications are rare and
can occur if an artery or a vein lies in the tract of the tracheostomy. Also,
erosion into the innominate artery may result from low placement of a tracheost-
omy tube. Minor peristomal oozing can be managed with inltration of the tissue
with lidocaine and epinephrine. In our experience, persistent peristomal oozing
responds very well to packing of the peristomal area with sterile gelfoam strips
soaked in topical thrombin. Platelet transfusion and correction of coagulopathy
are also useful in correcting persistent slow peristomal bleeding in the setting
of thrombocytopenia and coagulopathy.
The rate of infection is between 1.5% and 4% in different series and is
usually minor. The rate of infection in PDT is comparable to, if not lower
than, that in surgical tracheostomy (47). Routine use of antibiotic prophylaxis
is not recommended, except in patients at risk for bacterial endocarditis.
Accidental extubation with loss of airway control may happen during
tracheostomy. Failure to re-establish a lost airway expediently during tracheost-
omy can be catastrophic, especially if neuromuscular blockade is used. An intu-
bation tray should be readily available at bedside. The risk of paratracheal
placement of the tracheostomy is minimized with the use of bronchoscopic
guidance.
In case of inadvertent decannulation in the early postoperative period, blind
reinsertion of the tracheostomy tube increases the risk of placement of the
Table 4 Complications of Percutaneous

Dilatational Tracheostomy
Bleeding
Infection
Accidental extubation
Paratracheal insertion
Esophageal perforation
Subcutaneous emphysema
Pneumothorax
Tracheal ring fracture
Tracheal stenosis
Tracheoamalacia
tracheostomy tube into the paratracheal space, resulting in upper airway

obstruction and hypoxemia (48). Inadvertant decannulation in the rst seven to
ten days should ideally be managed by oral intubation and elective reinsertion
of the tracheostomy tube with a guidewire and preferably with bronchoscopic
guidance.
Esophageal perforation should be avoided with bronchoscopic guidance
during PDT. Subcutaneous emphysema and pneumomediastinum may occur
with dissection of the paratracheal soft tissue through the stoma by the pres-
sured air from the ventilator (49). It is a good practice to inate the tracheostomy
tube cuff only after the ventilator has been connected to the tracheos-
tomy tube. The risk of pneumothorax may increase with a low tracheostomy
placement (50).
Long-term complications of tracheostomy include tracheal stenosis,
tracheomalacia, tracheo-esophageal stula, and tracheo-innominate stula.
Tracheal stenosis can be suprastomal, stomal, or infrastomal in location. Tracheal
stenosis may be caused by antecedent injury from endotracheal intubation
or from the tracheostomy itself. Distinguishing the cause of the stenosis from
the endotracheal tube or from the tracheostomy tube is often difcult and may
not be possible; therefore, this issue has to be kept in mind in interpreting
the studies on tracheal stenosis. The suprastomal or stomal stenosis can be
caused by damage to the tracheal mucosa and cartilage and the presence of the
articial airway itself, with subsequent granulation tissue formation. Tracheal
stenosis at the stoma site can be present one to six months after decannulation
(51). The risk of tracheal stenosis ranges in different studies between 0% and
16%. Most of these stenoses are less than 50% and are not symptomatic
(41,52 57). Infrastomal stenosis could result from pressure and damage
exerted by the tracheostomy tip or the excessive cuff pressures. These lead to
mucosal degeneration and chondritis that may be exacerbated by pooled
secretions and gastro-esophageal reux disease. Patients with tracheal stenosis
are often asymptomatic until the actual lumen is reduced by 50% to 70%,
especially in patients who are sedentary or bed-bound after a prolonged critical
illness and respiratory failure.
Tracheo-innominate stula is a life-threatening complication that requires
immediate surgical intervention. The incidence is 0.6% to 0.7%, and it most often
occurs within two to four weeks of the tracheostomy. Risk factors include low
placement of tracheostomy, high-pressure cuffs, and excessive tracheostomy
movement. The injury may be caused by pressure necrosis from any part of
the tracheostomy tube. Bleeding is usually the only symptom, with often non-
diagnostic angiography or bronchoscopy (58 60).
Tracheo-esophageal stula is another rare complication and may be due to
pressure necrosis or iatrogenic posterior wall puncture by the introducer needle.
Symptoms include coughing up feeding or food material, recurrent pulmonary
infection, copious secretions, or ventilator air leak. Treatment options include
surgical repair or double stenting of the esophagus and the trachea.
VI. OST vs. PDT
The trend toward minimally invasive surgery and the development of interven-
tional services within the non-surgical specialties has spurred considerable
interest in bedside PDT in the ICU. The economic issue was one of the important
aspects of PDT that led to its initial popularity. OST has traditionally been
performed in the operating room. In many institutions it remains so. However,
performing tracheostomy in the ICU avoids the potential hazards of transporting
critically ill patients to and from the operating room. When rst introduced,
bedside PDT was touted not only for its ease of performance with safety
prole comparable to OST, but also for a signicant decrease in hospital
charges and a more efcient utilization of ICU resources. The latter reects the
absence of operating room charges and anesthesia fees associated predominantly
with OST performed in the operating room. More recently, there are reports of a
comparable safety prole in performing bedside OST (61 63). Bedside OST is
usually performed with a standard reusable tracheostomy tray and electrocautery,
without operating room charges and anesthesia fees. The professional fee for
tracheostomy is the same, regardless of where or how tracheostomy is performed.
The cost of using the surgical instruments at bedside is minimal. PDT is usually
performed using disposable kits under bronchoscopic guidance; therefore, there
are additional costs and charges for these items, rendering PDT somewhat
more costly than OST when both are performed at bedside. Although most
studies have reported a shorter operative time for PDT when compared with
OST (39,47,62,64,65), this is only an issue when the procedure is performed in
the operating room, in which longer operating time entails higher charges. The
time from decision to performance of the tracheostomy is shorter in PDT
compared with OST (47,66). Whether this faster decision to performance time
will reect in overall decreased hospital and ICU days and the overall cost
remains to be seen. Some studies also report less hemorrhage and wound
infection with PDT, while others have shown no difference (47,65,67 69).
Most of the studies looking at the safety and outcome prole of the two
techniques lack the rigorous design for this comparison to be meaningful.
Many of the OST in the studies were performed in the operating room, not at
bedside. Even the meta-analyses comparing percutaneous tracheostomy and sur-
gical tracheostomy are plagued with inconsistencies, including the heterogeneity
of the techniques included under percutaneous tracheostomy and the wide
ranging quality of studies included (39,40). In general, the superiority of one
procedure over the other has not been established. What has been established
in multiple studies is that PDT has a comparable safety prole to OST (70 74).
VII. Care of Patients with Tracheostomy
In the early post-tracheostomy period, care should focus on keeping the wound
clean. Dressing changes are indicated twice daily or whenever dressing is
soiled. The site can be cleaned with a mixture of hydrogen peroxide and saline
solution. If the tracheostomy tube is secured with sutures, then sutures can be
removed in seven to ten days. The tracheostomy tube can be secured comfortably
with a foam-pad Velcrow neck strap. The inner cannula can be changed daily in
the early post-operative period. We do not routinely change the tracheostomy
tube after placement, unless a different tube is needed or there is a malfunction
of the existing tube. Dislodgement of the tracheostomy tube in the rst seven
to ten days before the tract has matured can be problematic. Endotracheal
intubation should be performed if the tracheostomy tube cannot be reinserted
safely. Bronchoscopic guidance and conrmation of placement is strongly
recommended.
Routine care for an established tracheostomy generally centers on suction-
ing, appropriate humidication, weekly cleaning of the tube, and daily change
of the inner cannula. Cuffed tubes are used when patients are still mechanically
ventilated, or when risks of aspiration are high. Cufess tubes promote better
clearance of secretions, and less likelihood of infection.
The intervals to downsizing of tracheostomy tubes and decannulation
should be individualized. Factors to consider include the initial indication for tra-
cheostomy and the current respiratory status, the amount of secretions and the
patients ability to clear the secretions, and the mental status. Once decannulated,
the stoma closes rapidly within days, especially if the stoma has been downsized.
VIII. Future
Many percutaneous tracheostomy methods have been introduced over the past
half-century. Almost two decades since PDT was rst introduced, a large body
of literature has conrmed that in skilled hands, the safety prole of the technique
is comparable to OST. PDT comes in the era of cost containment, increasing role
of minimally invasive surgery, and the growing intensivist movement, all of
which have contributed to the momentum towards its acceptance as a safe and
viable bedside technique in the ICU. PDT should not be viewed as a replacement
for OST because OST still has its vital role and place. In the future, we will likely
see more renements of percutaneous tracheostomy techniques beyond PDT.
In our opinion, PDT will likely be the benchmark to which future percutaneous
tracheostomy techniques will be compared.
References
1. Frost EA. Tracing the tracheostomy. Ann Otol Rhinol Laryngol 1976; 85:618 624.
2. Sheldon CH, Pudenz RH, Feshwater DB, Crue BL. A new method for tracheotomy.
J Neurosurg 1957; 12:428 431.
3. Toye FJ, Weinstein JD. A percutaneous tracheostomy device. Surgery 1969;
65:384 389.
4. Ciaglia P, Firshing R, Syniec C. Elective percutaneous dilational tracheostomy: a

new simple bedside procedure. Chest 1985; 87:715 719.
5. Ciaglia P. Technique, complications, and improvements in percutaneous dilational
tracheostomy. Chest 1999; 115:1229 1230.
6. Schachner A, Ovil J, Sidi J, Rogev M, Heilbronn. Percutaneous tracheostomy: a new
method. Crit Care Med 1989; 17:1052 1056.
7. Schanchner A, Ovil Y, Sidi J, Avram A, Levy MJ. Rapid percutaneous tracheostomy.
Chest 1990; 98:1266 1270.
8. Griggs WM, Wothley LI, Gilligan JE, Thomas PD, Myburg JA. A simple percuta-
neous tracheostomy technique. Surg Gynecol Obstet 1990; 170:543 545.
9. Sviri S, Samie R, Roberts BL, van Heerden PV. Long-term outcomes following per-
cutaneous tracheostomy using the Griggs technique. Anaesth Intensive Care 2003;
31(4):40 47.
10. Borm W, Gleixner M. Experience with two different techniques of percutaneous dila-
tional tracheostomy in 54 neurosurgical patients. Neurosurg Rev 2003; 26(3):188191.
11. Anon JM, Gomez V, Paz Escuela M, De Paz, V, Solana LF, De La Casa RM, Perez
JC, Zeballos E, Navarro L. Percutaneous tracheostomy: comparison of Ciaglia and
Griggs techniques. Crit Care 2000; 4(2):124 128.
12. Nates NL, Cooper DJ, Myles PS, Scheinkestel CD, Tuxen DV. Percutaneous
tracheostomy in critically ill patients: a prospective randomized comparison of
two techniques. Crit Care Med 2000; 28(11):3734 3739.
13. Van Heurn LW, Mastboom WB, Scheeren CI, Brink PR, Ramsay G. Comparative
clinical trial of progressive dilatational and forceps dilatational tracheostomy. Inten-
sive Care Med 2001; 27(1):292 295.
14. Fantoni A, Ripamonti D. A non-derivative, non-surgical tracheostomy: the translar-
yngeal method. Intensive Care Med 1997; 23:386 392.
15. Heffner JE, Hess D. Tracheostomy management in the chronically ventilated patient.
Clin Chest Med 2001; 22:555 569.
16. Stauffer JL, Olson DE, Petty TL. Complications and consequences of endotracheal
intubation and tracheostomy. Am J Med 1981; 70:65 76.
17. Coppolo DP, May JJ. Self extubation. A 12 month experience. Chest 1990;
98:165 169.
18. Kollef MH, Levy NT, Ahrens TS, Schaiff R, Prentice D, Sherman G. The use of
continuous iv sedation is associated with prolongation of mechanical ventilation.
Chest 1998; 114:541 548.
19. Friedman Y. Indications, timing, techniques, and complications of tracheostomy in
the critically if patient. Curr Opin Crit Care 1996; 2:47 53.
20. White RE. A prospective study of laryngotracheal sequelae in long term intubation.
21. Heffner JE, Miller KS, Sahn SA. Tracheostomy in the intensive care unit: part
2. Complications. Chest 1986; 90:430 436.
22. Stauffer JL, Olson DE, Petty TL. Complications and consequences of endotracheal
intubation and tracheostomy. Am J Med 1981; 70:65 76.
23. White RE. A prospective study of laryngotracheal sequelae in long term intubation.
24. Maziak DE, Meade MO, Todd TR. The timing of tracheotomy: a systemic review.
Chest 1998; 114:605 609.
25. Armstrong PA, Mc Carthy MC, Peoples JB. R. Reduced use of resources by early
tracheostomy in ventilator dependant patients with blunt trauma. Surgery 1998;
124:763 767.
26. Sugerman HJ, Wolfe L, Pasquale MD, Rogers FB, OMalley KF, Knudson M,
DiNardo L, Gordon M, Schaffer S. Multicenter, randomized, prospective trial of
early tracheostomy. J trauma 1997; 43:741 747.
27. Brook A, Sherman G, Malen J, Kollef MH. Early vs late tracheostomy in
patients who require prolonged mechanical ventilation. Am J Crit Care 2000;
9:352 359.
28. Kluger Y, Paul DB, Lucke J, Cox P, Colella JJ, Townsend RN, Raves JJ, Diamond
DL. Early tracheostomy in trauma patients. Eur J Emerg Med 1996; 3:95 101.
29. Rumbak KJ, Newton M, Truncale T, Schwartz SW, Adams JW, Hazard PB. A pro-
spective, randomized study comparing early percutaneous dilational tracheostomy to
prolonged translaryngeal intubation (delayed tracheotomy) in critically ill medical
patients. Critical Care Med 2004; 32(8):1689 1694.
30. Plummer AL, Gracey DR. Consensus conference on articial airways in patients
receiving mechanical ventilation. Chest 1989; 96:178 180.
31. Beiderlinden M, Groeben H, Peters J. Safety of percutaneous dilatational tracheost-
omy in patients ventilated with high positive end-expiratory pressure (PEEP).
Intensive Care Med 2003; 29(7):1190.
32. Meyer M, Critchlow J, Manasharamani N, Angel LF, Garland R, Ernst A. Repeat
bedside percutaneous dilatational tracheostomy is a safe procedure. Crit Care Med
2002; 30(5):986 988.
33. Ben-Nun A, Altman E, Best LA. Emergency percutaneous tracheostomy in trauma
patients: an early experience. Ann Thorac Surg 2004; 77(3):1045 1047.
34. Ault MJ, Ault B, Ng PK. Percutaneous dilatational tracheostomy for emergent
airway access. J Intensive Care Med 2003; 18(4):222 226.
35. Mansharamani NG, Koziel H, Garkland R, LoCicero J III, Critchlow J, Ernst
A. Safety of bedside percutaneous tracheostomy in obese patients in the ICU.
Chest 2000; 118:1412 1418.
36. Dosemeci L, Yilmaz M, Gurpinar F, Ramazanoglu A. The use of the laryngeal mask
airway as an alternative to the endotracheal tube during percutaneous dilatational
tracheostomy. Intensive Care Med 2002; 28(9):1368.
37. Craven RM, Laver SR, Cook TM, Nolan JP. Use of pro-seal LMA facilitates percu-
taneous dilational tracheostomy. Can J Anaesth 2003; 50(7):718 720.
38. Datta D, Onyirimba F, McNamee MJ. The Utility of Chest Radiographs following
percutaneous dilatational tracheostomy. Chest 2004; 123(5):1603 1606.
39. Freeman BD, Isabella K, Lin N, Buchman TG. A meta-analysis of prospective trials
comparing percutaneous and surgical tracheostomy in critically ill patients. Chest
2000; 118:1412 1418.
40. Dulguerov P, Gysin C, Perneger TV, Chevrolet JC. Percutaneous or surgical tra-
cheostomy: a meta-analysis. Crit Care Med 1999; 27:1617 1625.
41. Goldenberg D, Ari EG, Golz A, Danino J, Nezer A, Joachims HZ. Tracheotomy
complications: a retrospective study of 1130 cases. Otolaryngol Head Neck Surg
2000; 123:495 500.
42. Polderman KH, Spijkstra JJ, de Bree R, et al. Percutaneous dilatational tracheostomy
in the ICU. Chest 2003; 123(5):1336 1338.
43. Gambale G, Cancellieri F, Bladini U, Vacchi Suzzi M, Baroncini S, Ferrari F,

Petrini F. Ciaglia Percutaneous Dilatational Tracheostomy. Early and late compli-
cation and follow up. Minerva Anestesiol 2003; 69(11):825 830.
44. Hill BB, Zweng TN, Maley RH, Charash WE, Toursarkissian B, Kearne PA.
Percutaneous dilational tracheostomy: report of 356 cases. J Trauma 1996;
41(2):238 243.
45. Marx WH, Ciaglia P, Graniero KD. Some important details in the technique of per-
cutaneous dilational tracheostomy via the modied Seldinger technique. Chest 1996;
101:762 766.
46. Friedman Y, Mayer AD. Bedside percutaneous tracheostomy in critically ill patients.
Chest 1993; 104:532 535.
47. Friedman Y, Fildes J, Mizock B, Samuel J, Patel S, Appavu S, Robert F. Comparison
of percutaneous and surgical tracheostomies. Chest 1996; 110:480 485.
48. Reibel JF.Tracheostomy. Respir Care 1999; 44:820 823.
49. Kaylie DM, Wax MK. Am J Otolaryngol 2002; 23(5):300 302.
50. Marx WH, Ciaglia P, Graniero KD. Some important details in the technique of
percutaneous dilational tracheostomy via the modied Seldinger technique. Chest
1996; 101:762 766.
51. Heffner JE. Upper airway dysfunction. In: Marini JJ, Slutsky AS, eds. Physiological
Basis of Ventilatory Support. New York: Marcel Dekker, 1998:533 574.
52. Rosenbower TJ, Morris JA Jr, Eddy VA, Ries WR. The long-term complications of
percutaneous dilatational tracheostomy. Am Surg 1998; 131:265 271.
53. Ciaglia P, Graniero KD. Percutaneous dilational tracheostomy. Results and long
term follow up. Chest 1992; 101:464 467.
54. Law RC, Carney AS, Manara AR. Long-term outcome after percutaneous
dilational tracheostomy. Endoscopic and spirometric ndings. Anesthesia 1997;
52:51 56.
55. Van Heurn LW, Goei R, De Ploeg I, Ramsay G, Brink PR. Late complications of
percutaneous dilatational tracheostomy. Chest 1996; 110:1572 1576.
56. Van Heurn LW, Van Geffen GJ, Brink PR. Clinical experience with percutaneous
dilatational tracheostomy: report of 150 cases. Eur J Surg 1996; 162:531 535.
57. Callanan V, Gilmore K, Field S, Beaumont A. The use of magnetic imaging to asses
tracheal stenosis following percutaneous dilatational tracheostomy. J Laryngol Otol
1997; 953 957.
58. Yang FY, Criado E, Schwartz JA, et al. Tracheo-innominate artery stula: retrospec-
tive comparison of treatment methods. South Med J 1988; 81(6):701 706.
59. Anthony R, Newbegin CJR. Emergency management of tracheo-brachiocephalic
artery stula. J Laryngol Otol 2003; 117: 985 986.
60. Gasparri MG, Nicolosi AC, Almassi H. A novel approach to the management of
trachoinnominate artery stula. Ann Thorac Surg 2004; 77(4):1424 1426.
61. Wang SJ, Sercarz JA, Blackwell KE, Aghamohammadi M, Wang MB. Open bedside
tracheotomy in the intensive care unit. Laryngoscope 1999; 109:891 893.
62. Porter JM, Ivatury RR. Preferred route of tracheostomy- percutaneous vs open at the
bedside: a randomized, prospective study in the surgical intensive care unit. Am Surg
1999; 65:142 146.
63. Grover A, Robbins J, Bendick P, Gibson M, Villalba M. Open vs percutaneous
dilatational tracheostomy: efcacy and cost analysis. Am Surg 2001; 67:297 301.
64. Freeman BD, Isabella K, Cobb JP, Boyle WA III, Schmieg RE Jr, Kolleff MH, Lin N,
Saak T, Thompson EC, Buchman TG. A prospective, randomized study comparing
percutaneous with surgical tracheostomy in critically ill patients. Crit Care Med
2001; 29(5):926 930.
65. Hazard P, Jones C, Benitone J. Comparative clinical trial of standard operative tra-
cheostomy with percutaneous tracheostomy. Crit Care Med 1991; 19(8):1018 1024.
66. Friedman Y, Mizock BA. Percutaneous vs surgical tracheostomy: procedure of
choice or choice of procedure. Crit Care Med 1999; 27(8):1684 1685.
67. Griggs WM, Myburg JA, Worthley LI. A prospective comparison of a percutaneous
tracheostomy technique with standard surgical tracheostomy. Intensive Care Med
1991; 17(5):261 263.
68. Cobean R, Beals M, Moss C, Brendenberg CE. Percutaneous dilatational tracheost-
omy. A safe, cost-effective bedside procedure. Arch Surg 1996; 131(3):265 271.
69. Holdgaard HO, Pedersen J, Jensen RH, Outzen KE, Midtgaard T, Johansen LV,
Moller J, Paaske PB. Percutaneous dilatational tracheostomy versus conventional
surgical tracheostomy. A clinical randomized study. Acta Anaesthesiol Scand
1998; 42(5):545 550.
70. Sustic A, Krstulovic B, Eskinja N, Zelic M, Ledic D, Turina D. Surgical tracheost-
omy vs percutaneous dilational tracheostomy in patients with anterior cervical spine
xation: preliminary report. Spine 2002; 27(17):1942 1945.
71. Khalili TM, Koss W, Margulies DR, Morriaon E, Shabot MM. Percuateous dilational
tracheostomy is as safe as open tracheostomy. Am Surg 2002; 68(1):92 94.
72. Lukas J, Stritesky M. Tracheostomy in critically ill patients. Bratisl Lek Listy 2003;
104(7 8):239 242.
73. Kaylie DM, Andersen PE, Wax MK. An analysis of time and staff utilization for
open versus percutaneous tracheostomies. Otolaryngol Head Neck Surg 2003;
128(1):109 114.
74. Wu JJ, Huang MS, Tang GJ, Shih SC, Yang CC, Kao WF, Huang MH, Lee CH.
Percutaneous dilatational tracheostomy vs open tracheostomya prospective,
randomized controlled trial. J Chin Med Assoc 2003; 66(8):467 473.
10
Radiology in the Intensive Care Unit
BRUCE M. BARACK, ANTOINETTE R. ROTH

C. MATILDA JUDE, and
HSIN-YI LEE Department of Radiological Science
Department of Imaging Los Angeles, California, U.S.A.
VA Greater Los Angeles
Healthcare System and
Department of Radiological Science
I. Introduction
The care of the intensive care unit (ICU) patient requires frequent radiological
consultation. Daily review of radiographs is an integral part of critical care
rounds. This chapter will address the role of thoracic radiology in the ICU patient.
The indications, technique, interpretation, and limitations of the portable
chest radiograph will be discussed. The proper positioning and complications
of devices used in the ICU and the radiologic ndings of common thoracic
abnormalities encountered will also be described. Lastly, the role of multi-
detector computed tomography (MDCT) imaging and the role of interventional
radiology will be summarized.
II. The Portable Chest Radiograph

A. Indications
The accepted indications for portable chest radiographic examination of the ICU
are shown in Table 1.
331
332 Barack et al.
Table 1 Indications for Portable Radiographic Examination of the ICU Patient
Admission to ICU as baseline

Prior to any major diagnostic or surgical procedure in the absence of a previous lm within
24 hr
Immediately following any major diagnostic or surgical procedure
Immediately following any tube, line, or pacemaker attempt (insertion or removal)
Any rapid clinical deterioration
As needed to monitor response to therapy
Although not universally accepted as an indication, the mere presence of

the patient in the ICU frequently generates requests for daily portable chest
examinations. There is ample evidence in the literature to support this practice
in the medical ICU, the respiratory ICU, and intubated, mechanically ventilated
patients (1 4) when uncomplicated cardiac patients and those without heart or
lung disease are excluded. Data for surgical ICU (SICU) patients are mixed
with some investigations that support this practice and others that argue
against routine daily studies (5 7).
In support of daily routine examinations, new ndings, including cardio-
pulmonary abnormalities and line or tube malpositions, led to changes in diag-
nostic approach or therapeutic measures in 14% to 37% of chest X-rays.
Changes in therapy were more frequent in intubated, pulmonary, and unstable
cardiac patients (1), patients with two or more catheters and tubes (2), and
patients in whom a change in clinical condition prompted the exam (3).
In a study of SICU patients arguing against daily routine chest radiography,
only 13 of 1028 (1.3%) devices central venous pressure (CVP) lines, endotra-
cheal tubes, etc. required repositioning, and only three of 525 chest X-rays
(0.6%) (two pneumothoraces and one effusion) had potential clinical impact
(6). In another multivariate analysis of 1003 chest X-rays in SICU patients,
only patients with Swan Ganz catheters were felt to be justied in having
routine daily radiographs (7). The authors recommend that routine daily radio-
graphs be obtained only on admission to the SICU and after placement of an inva-
sive device. The American College of Radiology recommends daily portable
chest radiographs on patients with acute cardiopulmonary problems and those
receiving mechanical ventilation (8).
B. Technique
Portable Antero-Posterior View
Portable chest radiography accounts for 50% of all chest radiographs performed
in hospitals (9). The key to obtaining optimal repeat portable chest radiographs
on the same ICU patient is to minimize variation in the technical parameters
of successive radiographs (10 12). At our institution, all portable ICU chest
radiographs are performed at a 50 in. source image distance (SID), with the
Radiology in the ICU 333
patient in the supine position, during peak inspiration with 85 kVp. The exposure
is timed with respect to peak inspiratory pressure if the patient is on mechanical
ventilation. The technique is recorded on an adhesive label, which is placed at
the patients bedside for use on all subsequent examinations. It has been found
that the supine position minimizes position errors secondary to rotation,
distortion errors secondary to kyphosis, and unwanted body artifacts secondary
to head or extremity obscuring the thoracic anatomy.
Portable chest radiographs result in the patient receiving two to four times
the radiation dose of a radiograph obtained with stationary equipment, primarily
because of increased scatter radiation (12). Scatter radiation at ICU nursing
stations has been demonstrated to be signicantly below the maximal permissible
dose for non-occupational workers (13,14).
Special Views
Special portable views that may be useful in the ICU patient include lateral,
transthoracic lateral, and lateral decubitus views of the chest, left lateral decubi-
tus view of the abdomen, and an antero-posterior (AP) view of the lower chest
and upper abdomen.
In one series, the portable lateral chest radiograph yielded an 11% inci-
dence of either unexpected conditions, such as catheter or tube malpositions,
or improved interpretation on the AP radiograph (15). The transthoracic lateral
view may conclusively demonstrate a pneumothorax. The lateral decubitus
view also demonstrates a pneumothorax and differentiates free from loculated
pleural uid and pneumothorax from pneumomediastinum. The left lateral decu-
bitus view of the abdomen is frequently used to demonstrate free intraperitoneal
air, which collects superiorly to the liver. An AP view of the lower chest and
upper abdomen is useful in localizing the tip of a nasoenteric tube.
Oral Contrast
Oral contrast may be used to diagnose a tracheoesophageal stula or an esopha-

geal perforation in conjunction with a portable chest radiograph in an ICU patient
too unstable to be transported to the radiology department. It may also be used to
localize the tip of a nasoenteric or gastrostomy tube. The choice of oral contrast
agent is dictated by the site of the suspected pathology. Barium provokes
an extensive inammatory reaction in the mediastinum and hinders surgical
exploration. It should not be used when mediastinal extravasation is a possibility.
On the other hand, barium is relatively inert in the airway and lung parenchyma.
Water soluble contrast is better tolerated in the mediastinum than barium.
However, water soluble contrast results in a severe inammatory response in
the lungs. Furthermore, hypertonic water soluble contrast may result in uid
shifts between compartments.
Barium should be used in cases of suspected tracheoesophageal stula
(Fig. 1) and water soluble contrast in suspected esophageal rupture or perforation
334 Barack et al.
Figure 1 Tracheoesophageal stula. Antero-posterior view of barium swallow in a

patient with a bulky tumor in the upper mid esophagus demonstrates opacication of
the distal trachea, left main stem bronchus, and left lower lobe bronchus (black arrow-
heads), indicating the presence of a tracheoesophageal stula. The proximal and distal eso-
phagus are well coated with barium (white arrowheads).
(Fig. 2). Because of the compromised condition of the ICU patient, isoosmolar
water soluble contrast is used to avoid possible uid compartmental shifts.
Advantages of Digital Film and PACS

A digital radiograph is any radiographic image acquired without the use of photo-
graphic lm. Digital radiography includes computed radiography (CR), direct
digital radiography (DR), indirect digital radiography, and optically coupled
direct radiography. At present, CR is the only system used for bedside chest
radiographs. DR systems offer high image quality and the potential for dose
reduction when compared with images obtained with conventional lm-screen
systems (16).
Advantages of a CR system over conventional lm-screen systems are
the inherent wide latitude of the CR system with relatively constant image
Figure 2 Esophageal perforation. Axial computed tomography image (mediastinal

window) in a patient post-dilatation of an esophageal stricture demonstrates evidence of
perforation with extravasation (arrows) of oral contrast from the distal esophagus
through the mediastinum into the right pleural space.
quality over a wide range of exposures, the ability to postprocess the image, and
advanced imaging processing algorithms that minimize scatter and reduce radia-
tion dose. The major disadvantages of slightly higher image noise and lower
spatial resolution are not signicant problems with portable chest radiographs
(17). Although the classic technical errors of patient motion and malpositioning,
incorrect patient identication, incorrect examination, and double exposure occur
with the same frequency as with conventional lm-screen systems (18), repeat
rates of less than 1% are reported with CR systems (19). This is primarily attri-
buted to fewer repeat examinations because of under and overexposure.
Digital radiography is usually an integral part of an information system
called picture archiving and communications system (PACS). At present, the
majority of radiographic images are acquired in a digital format, and it is esti-
mated that by the end of this decade, over half of all radiology departments in
medical centers will be completely or almost completely digital (20).
Major advantages of PACS include elimination of lost studies, marked
reduction in lm costs, fewer repeat examinations, improved image quality,
rapid availability of images, rapid viewing of large studies such as MDCT or
magnetic resonance imaging (MRI), and improved efciency of radiology
personnel and clinicians.
336 Barack et al.
C. Interpretation
Accurate interpretation of the portable chest radiograph demands a careful, sys-
tematic evaluation, which requires discipline and practice. The following
approach is intended as a guide, although individual approaches may vary.
The initial observation should include the technical quality of the radiograph
for exposure factors and patient positioning, and insure that the lung bases and
apices have been included on the lm. The radiograph should then be inspected
for any tubes, lines, or catheters, and any malposition, complication, or change
from the prior radiograph noted. Examination of the osseous structures, the
mediastinum (including the vascular pedicle), the heart, the pulmonary vessels,
and the lungs should follow. An attempt should be made to form a preliminary
impression and a relevant differential diagnosis.
There are ve additional factors to consider when interpreting a portable
chest radiograph. First, patients are expected to inspire to the same lung
volumes on successive radiographs in the absence of intra-abdominal changes,
cardio-respiratory problems, or cardiac decompensation (10). Then, lower lung
volumes are only because of a decrease in lung, chest wall, or abdominal
compliance.
Second, the heart size does not change when the patient goes from the erect
to the supine position. The apparent increase in heart size on a supine lm is
because of magnication. As the heart is an anterior structure within the chest,
it will be magnied on an AP view when the cassette is placed behind the
patient, as opposed to a postero-anterior (PA) view when the cassette is closer
to the heart. The upper limit of normal for the cardiothoracic ratio in supine
patients with a normal heart is 0.57 to 0.58 (10,21).
Third, the vascular pedicle extends from thoracic inlet to the superior
aspect of the heart. The right border is formed by the right brachiocephalic
vein superiorly and the superior vena cava inferiorly. The left border is usually
formed by the left subclavian artery above the aortic arch. Therefore, the right
side of the vascular pedicle is venous and the left side arterial. As veins are
more compliant than arteries, changes in intravascular volume will be reected
by a larger change in the right side of the vascular pedicle than the left side.
The width of the vascular pedicle is measured horizontally from where the
right mainstem bronchus crosses the superior vena cava to a perpendicular
line drawn from where the left subclavian artery arises from the aorta (22).
Fourth, the size of the azygous vein changes linearly with the width of the
vascular pedicle, and both closely correlate with changes in intravascular systemic
blood volume (10). Assuming the supine position, rotation to the right, uid over-
load, acute right heart failure, and pericardial tamponade (Fig. 3) will increase the
width of the vascular pedicle, whereas increasing ventilatory pressure, decreasing
intravascular volume, and rotation to the left will decrease the width of the vas-
cular pedicle (10, 22). The vascular pedicle width (VPW) varies according
to the patients body habitus. The normal VPW is 48 + 5.0 mm on a 72-in.
Figure 3 Pericardial tamponade. Antero-posterior radiograph in a patient with pericar-

dial tamponade demonstrates enlarged heart (reecting pericardial effusion), wide vascu-
lar pedicle (reecting decreased systemic venous return and increased systemic blood
volume), and markedly decreased pulmonary vascularity (reecting decreased cardiac
output). Subsegmental atelectasis is present in the right lower lung zone.
PA radiograph. On an AP supine radiograph obtained at 40 in., there is an

approximate 5% geometric increase in the VPW over the value obtained in a
72-in. PA radiograph (22).
Finally in the supine patient, the effect of gravity on vessels in the upper
and lower lobes is equal. Therefore, the ratio of the diameter of the pulmonary
veins in the upper lobes to the lower lobes is 1:1 and not 1:2 as is seen in the
upright chest radiograph.
D. Limitations of the Portable Chest Radiograph

The limitations of the portable chest radiograph are listed in Table 2.
Table 2 Limitations of the Portable Chest Radiograph
Lack of lobar localization

Inability to determine content of vascular lumen (e.g., clot) or integrity of vessel wall
(e.g., intimal ap)
Limited ability to differentiate parenchymal opacities
Limited ability to characterize and localize pleural effusions and pneumothorax
Limited ability to diagnose lung abscess and empyema, and to differentiate lung abscess
from empyema
338 Barack et al.
III. Devices Used in the ICU

A. Proper Position, Malposition, and Complications
Familiarity with the proper position, malposition, and potential complications of
the multiple devices used in patients admitted to the ICU is essential as malposi-
tions and complications associated with both insertion and use of such devices
may be life-threatening. The optimal positions of commonly utilized medical
devices are summarized in Table 3.
Tubes
Endotracheal Tube
The tip of the endotracheal tube should ideally be located 3 7 cm above the
carina with the head in neutral position (23,24). This position allows for
maximal excursion of the endotracheal tube within the airway, with changing
head position preventing the possibility of subsequent malposition. Neutral
head position is best insured with the patient supine, and this is the recommended
position for the immediate post-intubation radiograph.
When the carina is not identied on the immediate post-intubation radio-
graph, a repeat radiograph is recommended. The method of estimating the
location of the carina from the projection of the tip of the endotracheal tube in
Table 3 Optimal Position of Commonly Utilized Medical Devices in Critically Ill

Patients
Device Radiographic location
Endotracheal tube 3 7 cm above the carina

Changes with head position
Flexion Mean: 5.5 mm inward;
maximum range: 23 mm in/19 mm out
Extension Mean: 6.3 mm outward;
maximum range: 21 mm in/33 mm out
Lateral movement Mean: insignicant; maximum range:
right19 mm in/17 mm out
left22 mm in/18 mm out
Cuff ination Cuff diameter: tube diameter ,1.5: 1
Tracheostomy tube Tip 1/2 to 1/3 of the distance between
tracheal stoma and carina
Nasoenteric tube 10 cm from the gastroesophageal junction
Chest tube Proximal port .3 cm from the lateral chest wall
Central venous catheter Superior vena cava
Pulmonary artery catheter 2 3 cm distal to bifurcation of main pulmonary artery
Less than 2 cm lateral to the hilum
Not beyond the proximal interlobar artery
relation to the thoracic vertebral bodies is unreliable (24), as is the practice of

estimating the location of the carina by comparing its location on a previous
portable radiograph (23). The inward and outward movement of the tip of the
endotracheal tube with head exion, extension, and rotation has been reported
and is shown in Table 3 (25,26).
The portable chest radiograph is an unreliable predicator of unsafe endo-
tracheal tube cuff pressures (27). When balloon overination results in the cuff
diameter/tracheal lumen ratio of greater than 1.5, tracheal damage is likely to
occur (28). Excessive cuff pressure requirements (19%) and inability to seal the
airway (11%) are the frequent adverse consequences of endotracheal tube intuba-
tion (29). Complications from overination of the endotracheal tube are unusual as
most endotracheal tubes used today have high-volume, low-pressure cuffs (23).
Tube malposition occurs in 10% to 20% of intubations (29 33). The right
mainstem bronchus is the most common location of malposition because the left
mainstem bronchus arises from the carina at a more acute angle (Fig. 4).
Complications of persistent right mainstem bronchus intubation are inadequate
ventilation, left lung collapse with gradual leftward mediastinal shift, and
increasing risk of a right-sided pneumothorax (23).
Malposition of an endotracheal tube into the esophagus is uncommon and
may not be immediately apparent clinically. Such malposition is difcult to
detect on routine portable AP radiographs because the tube may be projected
over the trachea. Radiographic ndings seen in decreasing frequency on
routine portable AP lms include projection of the tube lateral to the trachea,
Figure 4 Endotracheal tube in the bronchus intermedius. Antero-posterior radiograph

demonstrates the tip of the endotracheal tube (arrowhead) extending into the bronchus
intermedius well below the carina (arrow).
340 Barack et al.
gastric distention, esophageal air, and deviation of the trachea by the balloon cuff.
Examination of the patient in a 258 right posterior oblique projection accurately
demonstrates the endotracheal tube position in 92% of radiographs (34).
Tracheal rupture is a serious, less common complication of endotracheal
tube intubation. Early radiographic ndings are orientation of the distal portion
of the tube to the right of the tracheal air column with overdistention of
the tube cuff towards the endotracheal tube tip and subsequent development of
pneumomediastinum, subcutaneous air, and pneumothorax (35).
Aspiration pneumonitis has been reported to occur in 8% of endotracheal
tube intubations (29,36). It is usually easily identied on the post-intubation
radiograph and frequently becomes more apparent on subsequent examinations.
Fillings, teeth, and dentures may also be aspirated or swallowed secondary to
dislodgment during intubation (37).
Tracheostomy Tube
Tracheostomy is performed on patients with upper airway obstruction or on
patients who require long-term ventilator support (23). The tracheostomy tube
is inserted between the rst and second tracheal ring or second and third tracheal
ring. The tip of the tracheostomy tube should be one-half to two-thirds of the
distance between the tracheal stoma and the carina at the level of the third thor-
acic vertebral body. Neither head exion nor extension affects the location of
the tip of the tracheostomy tube (12). The lumen of the tube should be one-
half to two-thirds the diameter of the trachea, and the cuff should not distend
the tracheal wall. A post-tracheostomy lm showing the internal and external
ends of the tube in the same plane may indicate that the tracheostomy tube has
not distended properly into the trachea, which can be conrmed on a portable
lateral radiograph (38).
A small amount of subcutaneous air in the neck and a small pneumo-
mediastinum are frequent ndings on the post-tracheostomy radiograph (39).
However, fascial plane disruption may result in a large amount of air in the
neck, with a larger pneumomediastinum and a pneumothorax frequently seen
following injuries to the lung apex. Pneumothorax may also be seen in tracheal
perforation. Late complications include tracheo-innominate artery stulas or
tracheoesophageal stula (40 42). The stulas are usually secondary to pro-
longed hyperination of the cuff and occur at the level of the cuff with erosion
anteriorly (tracheo-innominate artery) or posteriorly (tracheoesophageal). In tra-
cheoesophageal stulas above the cuff, gastric contents may accumulate in the
upper trachea. With stulas below the cuff, aspiration of gastric contents into
the lungs is a routine occurrence. Other late complications include stricture, tra-
cheomalacia, and tracheal stenosis.
Nasoenteric Tubes
Nasoenteric tubes are commonly used in ICU patients for gastric decompression
or feeding. Sideholes are present along the distal 10 cm of nasogastric tubes.
Therefore, the tip of the tube should extend at least 10 cm beyond the-
gastroesophageal junction. A more proximal position may result in obstruction
of the distal esophagus with applied suction or may allow contents entering the
distal portion of the tube to exit into the distal esophagus and predispose to
aspiration (23). Incorrect placement was seen in 4.4% of nasogastric tube
insertions in one series of 340 patients (43). Similarly, the most proximal sidehole
of the tube should be placed inferior to the anastomotic site in postoperative
esophagectomy patients.
Feeding tubes are used for nutritional support as an alternative to intrave-
nous feeding. Ideally, they should be positioned distal to the gastric pylorus to
reduce gastroesophageal reux. A post-placement radiograph should always be
obtained, as malpositions are common following insertion (44,45) (Fig. 5).
Malpositioned tubes may be coiled in the pharynx, esophagus, or stomach. In
intubated or obtunded patients, or patients lacking a gag reex, tubes may pass
into the trachea and preferentially enter the right mainstem bronchus.
Immediate complications of nasoenteric tube insertion include traumatic
esophageal perforation with pneumomediastinum or, more rarely, bronchial
perforation and pneumothorax (23). Parietal pleural perforation may also result
from malposition of the tube within the esophagus and should be suspected
when a pleural effusion, pneumomediastinum, mediastinal widening, or
mediastinal air uid levels appear rapidly after beginning tube feedings (36,46).
Figure 5 Feeding tube in the left lower lobe bronchus. Antero-posterior post-insertion
radiograph demonstrates the tip of the feeding tube in the left lower lobe bronchus (arrow-
head). The tube was subsequently removed and reinserted without complication.
342 Barack et al.
Late complications of insertion include esophagitis and esophageal

stricture formation. Late complications of unrecognized malposition include
inadvertent infusion of feeding solutions into the lungs and entry into the visceral
pleural space causing a pneumothorax. If the feeding tube is noted within the
tracheobronchial tree, an immediate post-removal radiograph should be obtained
to exclude the possibility of a pneumothorax (47,48).
Sengstaken Blakemore (SB) tubes are used as a temporary measure to
control esophageal variceal bleeding. These tubes contain two inatable bal-
loons. The larger, more distal spherical balloon is inated only in the stomach
to a volume of 250 cc. After ination, it is retracted back to the gastric cardia,
insuring correct positioning of the more proximal cylindrical balloon within
the esophagus, which is inated to a pressure of 15 to 30 mmHg (49,50). The
incidence of esophageal perforation following placement of SB tubes is reported
as high as 15% to 30%, and a post-placement radiograph of the chest and upper
abdomen is essential to ensure proper positioning (51). Esophageal perforation
may result from either ination of the distal balloon within the esophagus
or retraction of the inated distal balloon into the esophagus by an agitated
patient or unintentionally by the physician. Findings suggesting perforation of
the esophagus on the post-placement radiograph include mediastinal widening,
pneumomediastinum, and subcutaneous emphysema.
Thoracostomy Tube
Thoracostomy tubes are commonly used in ICU patients to drain pleural uid col-
lections and pneumothoraces. In the supine patient, mobile pleural uid collects
posteriorly and pleural air anteriorly. A pleural drainage tube for mobile uid col-
lection should therefore be positioned postero-inferiorly through the sixth to
eighth intercostal space at the level of the mid axillary line, preferably guided
by ultrasonic localization (52). If there is uncertainty about the potential for
injury to abdominal structures, placement through the fth intercostal space pro-
vides a margin of safety. A thoracostomy tube for pneumothorax treatment
should be positioned near the lung apex at the level of the anterior axillary line
and be directed antero-superiorly. Loculated pleural uid collections, as seen
in empyemas, require precise localization with ultrasonography or computerized
axial tomography (23). Rapid improvement is expected on post-drainage radio-
graphs if the tube is properly positioned. Conversely, a persistent pneumothorax
or undrained pleural uid collection suggests a malpositioned or malfunctioning
tube. A malfunctioning tube may result from kinking or plugging by debris (12).
Improper tube position may be appreciated on the AP radiograph.
However, a lateral radiograph or computerized axial tomography may be
required to determine the exact tube position (15,53). A malpositioned tube
may be in the subcutaneous or extrapleural soft tissues, within an interlobar
ssure (Fig. 6) or within the lung parenchyma (Fig. 7).
The proximal side hole of the chest tube should be medial to the inner
margin of the ribs and can be identied along the radio-opaque edge of the
Figure 6 Tip of right chest tube in the right major ssure. Sagittal computed tomogra-
phy reformation (mediastinal window) in a patient with a non-draining chest tube demon-
strates the tip of the right chest tube within the right major ssure (arrow). A large
empyema (E) is present in the right posterior hemithorax, causing anterior displacement
and compression of the right lower lobe (arrowheads).
tube (54). Malposition of the tube in the subcutaneous tissue may result in
subcutaneous emphysema and occurs more frequently in patients with excessive
subcutaneous fat (44). Malpositioned tubes within an interlobar ssure are associ-
ated with a 29% rate of unsatisfactory drainage (55). Malposition within the lung
parenchyma may be associated with bronchopleural stula, pulmonary lacera-
tion, and hematoma (36). Pulmonary laceration and hematoma are particularly
seen in patients with pleural adhesions or decreased lung compliance.
Complications of drainage tubes because of tube insertion include bleeding
secondary to laceration of an intercostal artery, laceration of the diaphragm with
associated splenic, hepatic, and gastric injuries, and mediastinal and parenchymal
lung injury (56). A delayed complication of thoracostomy tubes is unilateral
pulmonary edema on the side of tube placement due to rapid pulmonary
344 Barack et al.
Figure 7 Tip of the right chest tube in the right middle lobe. Axial computed tomogra-
phy image (mediastinal window) in a patient with a right empyema and a non-functioning
right chest tube demonstrates the tip of the right chest tube within the lateral segment of
the right middle lobe (arrow). The right major ssure is clearly visualized (arrowhead).
The tube was removed without complications.
re-expansion after removal of a large amount of uid or air from the pleural space
(36). A less common delayed complication is pulmonary infarction due to suc-
tioning of lung tissue by the chest tube even at low suction pressures (57).
After a thoracostomy tube has been removed, a tubular structure containing air
or uid may persist in the position previously occupied by the thoracostomy
tube. Because of a local inammatory response and associated pleural thickening,
free communication between this tubular tract and the rest of the pleural space
may not exist and the isolated tube tract may simulate a localized pneumothorax
or abscess on the chest radiograph (52). Such tracts usually decrease and
disappear within a few days and rarely become infected.
Catheters
Central Venous Catheters
Approximately 38% of patients in the ICU have central venous catheters (58).
The most proximal port is 5 cm from the tip. The ideal location of the tip of
the central venous catheter is the superior vena cava at or slightly above the
entrance of the azygous vein. This location assures that all ports of the central
venous catheter are beyond the most proximal venous valves. Below the
azygous vein, the superior vena cava becomes intrapericardial, and caval perfor-
ation is associated with a pericardial effusion and eventual tamponade (11).
If the catheter is to be used to monitor central venous pressure, the tip of the
pressure monitoring port must be proximal to any competent valve. The last valve
in the subclavian vein occurs 2 cm distal to the junction of the subclavian and
internal jugular veins at about the level of the rst anterior rib (59). Therefore,
the tip of the pressure monitoring port should be visualized medial to the anterior
portion of the rst rib at the level of the rst anterior intercostal space. The
right and left brachiocephalic veins join to form the superior vena cava at this
level (23). Neither the brachiocephalic veins nor the superior vena cava
contains valves.
Malposition of central venous lines occurs in approximately one-third of
catheter insertions, and most malpositions are not clinically suspected (60).
The subclavian vein is the most often used site for central venous placement.
A retrospective study of 500 subclavian vein catheterizations revealed only
68% to be properly positioned, with 21.4% to be positioned in the right atrium
and 0.4% in the right ventricle. Cardiac malposition was more common with a
right-sided approach (30.1%) than a left-sided approach (61). The average
safe insertion length of the central venous catheter from either the right or left
subclavian or internal jugular vein is 16.5 cm in most adults, and no central
venous catheter should be inserted greater than 20 cm from these access
sites (62).
The most common venous malposition of the subclavian venous catheter is
the ipsilateral internal jugular vein (Fig. 8), which is reported to occur in 15% of
such central venous insertions (36). A malpositioned catheter in this location
may produce abnormal sensations in the ear or severe headache (63). Retrograde
infusion of intravenous uids into the internal jugular vein in this malposition
may result in thrombosis, erosion of the vein, and pooling of infused solutions
near the brain in the supine patient (64).
Another common venous malposition occurs when the catheter enters
the contralateral brachiocephalic vein (Fig. 9). This is more common when the
junction of the brachiocephalic veins is higher and at a less acute angle and
more common with left-sided insertions. Malposition of the catheter in the
azygous vein occurs occasionally. Rarely, a catheter may be positioned in the
right superior intercostal vein, the internal mammary vein, the pericardiophrenic
vein, the left superior intercostal vein, and an inferior thyroid vein.
The coronary sinus, which drains the cardiac veins and empties into the
posterior aspect of the right atrium near the entry of the inferior vena cava,
may be entered by a malpositioned catheter in the right atrium, as may the
hepatic veins and inferior vena cava (23).
Catheter malposition within the superior vena cava may occur in left-sided
catheter insertions when the catheter tip abuts perpendicularly upon the right
lateral wall of the superior vena cava. This position causes the tip to repeatedly
hit the lateral caval wall with inspiratory movements or with head and neck
movements with left internal jugular catheters. This can eventually damage the
caval epithelium and cause caval rupture hours or days after insertion (65).
346 Barack et al.
Figure 8 Tip of right subclavian venous catheter in the right internal jugular vein.
Antero-posterior radiograph obtained after insertion demonstrates right subclavian
venous catheter extending into the right internal jugular vein (arrow).
A gentle curve of the tip of the catheter may be seen when the catheter tip is in
contact with the lateral caval wall, and the catheter should promptly be
repositioned (66).
A simulated catheter malposition occurs in the presence of a persistent
left superior vena cava, which is the most common anomaly of systemic
venous drainage (Fig. 10). This anomaly occurs in 0.3% of the normal population
and 4.3% of patients with congenital heart lesions when the left anterior and
common cardinal veins fail to regress normally (67,68). This vein courses
inferiorly along the left side of the mediastinum and drains into the coronary
sinus. Most patients with this anomaly also have a right-sided superior vena
cava, which may be smaller than normal (23).
The most common complication of central venous line placement is
pneumothorax, which has been reported in 5.6% of patients with any central
line placement (3) and 6% of subclavian venous insertions (69). The latter is
explained by the 1 cm or less proximity of the apical pleura to the subclavian
vein. Therefore, a radiograph is recommended following any unsuccessful
subclavian venous insertion to exclude a pneumothorax before contralateral
Figure 9 Tip of right subclavian catheter in the left brachiocephalic vein. Antero-
posterior radiograph demonstrates the right subclavian catheter extending into the left
brachiocephalic vein (arrow).
insertion is attempted. A pneumothorax may become apparent hours to days after

central line insertion because of slow accumulation of pleural air (70,71). The
typical location of the pneumothorax in the upright lm is apicolateral.
However, the most non-dependent aspect of the thorax in a supine patient is
the anteromedial aspect of the thorax and the subpulmonic recesses. Air collec-
tions in these regions should be carefully looked for on both the immediate and
subsequent delayed post-insertion radiographs (52).
Puncture of the subclavian artery is not uncommon and usually produces
bright red, pulsatile backow of blood. However, arterial blood may be non-
pulsatile and dark, stimulating venous blood in hypoxic and hypotensive patients
(72). Extrapleural collections of blood may be suggested by a mass in the soft
tissues of the neck or a focal mediastinal bulge on the unsuccessful post-insertion
radiograph following an arterial puncture (44). Unrecognized vessel perforation
is one of the most dangerous complications of catheter insertion. Radiographic
ndings of this potentially fatal complication include identication of the
catheter or parts of catheter lateral to the normal mediastinal shadow outside
the anatomic position of the vena cava coursing in an abnormal direction (44)
and the rapid accumulation of pleural or mediastinal uid after catheter insertion
348 Barack et al.
Figure 10 Pulmonary artery catheter in the left superior vena cava. Antero-posterior
radiograph demonstrates the tip of the pulmonary artery catheter in the right interlobar pul-
monary artery. The catheter was inserted into the right subclavian vein and coursed
through the right brachiocephalic vein, a left superior vena cava (arrow), the coronary
sinus, right atrium, right ventricle, and main pulmonary artery to reach the nal location.
(23). A correct AP projection is mandatory as normal anatomic landmarks

may be lost in a rotated lm.
Fragmentation of the catheter at the time of insertion is an uncommon but
potentially serious complication. The fragmented catheter may embolize to the
superior vena cava, right atrium, right ventricle, or pulmonary artery. Sepsis,
pulmonary embolization, arrhythmias, or perforation may follow. Most frag-
ments can be successfully removed percutaneously (23). Knotting of the catheter,
which may result in vessel thrombosis, arrhythmia, and vascular injury, is rare
and usually successfully treated percutaneously (52).
Mediastinal enlargement appearing days or weeks after catheter insertion
may indicate superior vena cava thrombosis (44). However, patients with
superior vena caval thrombosis may have a normal radiograph even in the pre-
sence of sepsis, superior vena cava syndrome, and loss of central venous
access (73). When superior vena caval thrombosis is suspected, contrast compu-
terized tomography should be performed with bilateral arm injections to
determine if thrombosis exists and to delineate the extent of the obstruction.
In general, the type and frequency of the malposition and complication are
strictly dependent on the site of catheter insertion (44). The more curves a cath-
eter has to pass through to reach the superior vena cava, the greater the risk for
malposition and vascular damage. Because the left brachiocephalic vein has a
more acute and curvaceous course than the right before joining to form the
superior vena cava, closer scrutiny for malpositions and complications is required
when left-sided insertions are reported. On the right side, the internal jugular
insertion is safer than the subclavian insertion because of the higher rate of
pneumothorax with the latter approach.
Peripherally Inserted Central Catheters

Peripherally inserted central catheters (PICC) are small caliber catheters inserted
in the supercial veins of the arm. The major advantage of PICCs is that they can
be left in place for weeks to complete a course of intravenous therapy (12) or
for total parenteral nutrition (74,75). The ideal location of the tip of the PICC
catheter is the superior vena cava (76,77).
Successful bedside insertion rates by specially trained nurses have been
reported at 82.6% (327 insertions), with rates as high as 98.2% (869 insertions)
(78). Radiologists use a venographic uoroscopic direct puncture and sheath
technique for difcult insertions, PICC salvage, and PICC exchanges. The
service length for a given PICC is 28.1 days (78,79). Peripheral malpositions,
including coiling, are easily recognized clinically or by the absence of visualiza-
tion of the PICC catheter on the post-insertion chest radiograph. The central
malpositions are similar to those associated with central venous catheters.
Pulmonary Artery Catheters (Swan Ganz Catheters)

Pulmonary artery catheters are 2 3 lumen catheters used to measure right-sided
cardiovascular parameters and infer left sided parameters. The smallest of the
lumens is situated close to the catheter tip and connected to an inatable
balloon. When the balloon is temporarily inated, the catheter oats peripherally
and wedges. When the balloon is deated, the catheter tip recoils proximally.
The ideal position of the catheter tip is 2 to 3 cm distal to the bifurcation of
the main pulmonary artery, less than 2 cm lateral to the hilum, and not beyond
the proximal interlobar artery (12).
Approximately 24% of Swan Ganz catheters are malpositioned on the
post-insertion radiograph (59). Catheter migration after initial placement is
common probably because of catheter warming (80). This is due to increased
catheter compliance, shortening of the loop of the catheter through the right
ventricle, and subsequent distal malposition. The incidence of pneumothorax is
the same as with any upper chest central line placement (5 6%). Rapid
insertion of the catheter in a patient with right ventricular dilatation may result
in intracardiac knotting (39).
The most common complication of distal malposition is pulmonary infarc-
tion. This can result from peripheral position of the catheter tip obstructing ow,
persistent balloon ination in a peripheral pulmonary artery, or clot formation
around the catheter (52). The size of the infarct depends upon the size and
distribution of the occluded vessel. Radiographically, one may see a patchy
area of consolidation peripheral to the Swan Ganz catheter or a more classic,
350 Barack et al.
wedge-shaped pleural-based opacity (53) (Fig. 11) or no evidence of infarction.

Other rare complications include thromboembolism, hemorrhage, and pulmonary
arterial pseudoaneurysm formation, which can present as a new solitary pulmon-
ary nodule months after discharge (81,82).
Dialysis Catheters
The types of dialysis catheters used in the ICU are acute dialysis catheters (83),
cuffed tunneled catheters (84), and subcutaneous vascular port catheters.
Acute dialysis catheters are primarily used as short-term access in patients
with malfunction of permanent access or in bed-bound patients with acute
renal failure.
Most acute dialysis catheters are non-cuffed, non-tunneled, dual lumen
catheters composed of polyurethane. Polyurethane is quite sturdy, permitting a
larger internal lumen for a given outer diameter, and is less thrombogenic than
silicone (83).
Bedside ultrasound to localize the vein reduces the number of needle passes,
failed placements, and insertion-related complications and increases the success
rate of inexperienced operators to 95% (85). In addition, 28% to 35% of dialysis
patients have demonstrated signicant vein abnormalities, such as total occlusion,
non-occlusive thrombus, stenosis, and anatomic variation (86,87).
The tip of acute dialysis catheters placed in the internal jugular vein should
lie in the superior vena cava. Catheters placed in the femoral vein should be long
enough to insure that their tip lies in the inferior vena cava to prevent excessive
recirculation. Right internal jugular placement provides superior blood ow and
is less likely to malfunction than left internal jugular or femoral placement. High
complications rates temper subclavian vein placement of these catheters (83).
Complications of acute dialysis catheters are both immediate insertion
related and delayed (83) and similar to those associated with central venous
catheter placement. Because acute dialysis catheters are relatively stiff, they
are associated with a higher incidence of vessel and right atrial perforation
than cuffed, tunneled catheters. Late complications include central vein steno-
sis/thrombosis, vessel perforation, cardiac tamponade, hemomediastinum, and
infection. Stenosis/thrombosis has been reported in up to 28% of subclavian
dialysis catheters with infection, increasing the risk in such insertions (88).
Because of the high rate of complication, subclavian insertions should be
avoided and are contraindicated in patients expected to receive permanent
access on the same side. In contrast, internal jugular insertions are associated
with only a 2% to 3% thrombosis/stenosis rate (89 92).
Because cuffed, tunneled catheters are softer and less rigid than acute dialy-
sis catheters, vessel and atrial perforation is not a major complication. The
optional location of the tip of cuffed, tunneled catheters is the mid right atrium
(84), as opposed to the superior vena cava. The right internal jugular site is the
preferred site for catheter placement because it has a lower rate of complication
during and after insertion than other locations (93).
Figure 11 Pulmonary infarction (the Hamptons hump). (A) Antero-posterior arrow

radiograph demonstrates the tip of the pulmonary artery catheter extending more than
2 cm beyond the origin of the right interlobar artery into the right mid lung zone
(arrow). (B) Radiograph obtained four days later demonstrates a homogeneous pleural
based opacity with convex medial contour towards hilum (the Hamptons hump) in a sub-
segment of the lateral segment of the right middle lobe (arrowheads). The pulmonary
artery catheter has been withdrawn, and the tip is now in good position (arrow).
352 Barack et al.
The value of the post-insertion chest radiograph has been questioned when
a cuffed tunneled catheter is inserted under uoroscopy (94). Only seven pro-
cedural complications were identied in 937 consecutive central venous access
procedures, and all were identied at uoroscopy. No procedural complication
was detected on post-procedure chest radiography that was not detected on
uoroscopy, and only one malpositioned catheter was detected by chest radio-
graphy. The most common acute complications are inadvertent carotid artery
puncture, air embolism, hemothorax, and pneumothorax (95 97). Late malfunc-
tions of cuffed, tunneled catheters are primarily associated with thrombosis,
which may be either extrinsic or intrinsic (96).
Other Devices
Intra-Aortic Balloon Pump
The intra-aortic counterpulsation balloon pump is used in patients in cardiogenic
shock or with severe ventricular dysfunction and patients who undergo high-risk
cardiac surgical procedures (53). The pump is composed of a catheter surrounded
by a long balloon. The catheter is placed in the descending thoracic aorta with its
tip immediately distal to the origin of the left subclavian artery (23). The balloon
is inated during systole, increasing coronary artery perfusion, and forcibly
deated during diastole, facilitating aortic blood ow and decreasing ventricular
afterload (98). On the portable chest X-ray, the radio-opaque tip is identied
within the aortic knob (Fig. 12). The inated balloon is a long radiolucent
Figure 12 Intra-aortic counterpulsation balloon pump. Antero-posterior radiograph

demonstrates the radio-opaque tip of the intra-aortic counterpulsation balloon pump in
the inferior aspect of the aortic arch (arrow), in good position.
tubular structure extending the length of the descending thoracic aorta (11), and
the deated balloon cannot be visualized on the radiograph.
The location of the tip should be noted on daily radiographs to avoid
complications. Advancement of the catheter too far may result in cerebral embo-
lization or occlusion or dissection of the left subclavian or vertebral artery. If the
balloon tip is positioned more distally in the descending thoracic aorta, counter-
pulsation is less effective, and potential occlusion of the major branches of the
abdominal aorta, including the renal arteries, may occur (99). Dissection of the
aorta at the time of insertion is a rarely reported complication (100). This is
accompanied by clinical symptoms and may be suggested by loss of denition
of the descending aorta or widening of the para-aortic line on the portable radio-
graph. The diagnosis may be conrmed by either contrast enhanced computer-
ized tomography (CT) or magnetic resonance angiography (MRA).
Transvenous Pacing Devices

Transvenous endocardial pacing is used for patients with heart block or bradyar-
rhythmias. The pacer is usually inserted into the subclavian or internal jugular
vein and manipulated into the apex of the right ventricle under uoroscopy
where it is wedged into the trabeculae. Malposition is the most common compli-
cation (53). On the post-insertion radiograph, the catheter tip should be
projected over the right ventricle with the lead wire directed downward and
toward the left. Superior deection of the tip of the wire is suggestive of coronary
sinus malposition and should be conrmed on a portable lateral view, which will
demonstrate a posterior course of the pacing wire, as opposed to an anterior pos-
ition in the apex of the right ventricle (12,23).
Over time, approximately 20% of pacing wires will change position,
migrating to the right atrium, pulmonary artery, or coronary sinus. Cardiac
perforation has been reported in 5% to 7% of placements (101). This compli-
cation should be suspected when the tip of the pacing wire projects beyond the
border of the myocardium into the epicardial fat or when the cardiac silhouette
rapidly enlarges after pacing wire placement. The former is best appreciated
on a portable lateral radiograph. Cardiac perforation can be conrmed with
multidetector CT.
Failure to pace properly may be the result of pacing wire malposition, lead
fracture, or myocardial perforation. Lead fracture has been reported in 3% of
cases and may be the result of a sharp bend in the wire or may be caused by
compression of the lead between the clavicle and the rst rib, called pinch-off
syndrome, which has also been documented with cardiac pacemaker leads
(102). Lead fracture may be difcult to visualize on a portable radiograph and
is sometimes more apparent on uoroscopy.
Automatic Implantable Cardioverter-Debrillator Devices

Automatic implantable cardioverter-debrillator devices are used in patients to
control potentially fatal arrhythmias. These devices may be inserted via
354 Barack et al.
thoracotomy or transvenously and may consist of additional wire leads in the

superior vena cava or wires or patches placed over the pericardium (103,104). A
post-insertion radiograph should be obtained to check for lead placement and
possible complications, including pneumothorax. Knowledge of the exact type
of device is mandatory as many combinations of transvenous sensing and de-
brillator catheters have additional subcutaneous and pericardial patches (12,53).
Temporary Epicardial Wire Electrodes
Temporary epicardial pacemaker electrodes are commonly attached to the right
atrium and right ventricle following cardiac surgery. These electrodes are com-
monly secured to the right atrium and right ventricle and are usually removed
prior to discharge (53). Reported rare complications associated with removal
of such electrodes include pneumothorax and hemorrhage (100). Another rare
complication associated with removal is rupture of the epicardial wire electrode
at the point of electrode attachment to the right atrium and ventricle (Fig. 13). The
retained epicardial electrode must be removed surgically under general
anesthesia.
B. Complications of Mechanical Ventilation (Barotrauma)
The term barotrauma refers to the adverse consequences of mechanical ventila-

tion in the lungs. The major adverse consequences of positive pressure ventilation
Figure 13 Retained epicardial pacing electrode. Overpenetrated antero-posterior radio-

graph status post-CABG demonstrates a retained right epicardial electrode (arrow), which
fractured from the pacing wire during attempted removal.
have been divided into three categories: (1) physiologic effects on the heart and
pulmonary vasculature, (2) direct lung injury, and (3) air leak phenomena (105).
Positive pressure ventilation (PPV) may affect cardiopulmonary function,
particularly when used with positive end-expiratory pressure (PEEP). Cardiac
output may be diminished by PPV with PEEP in patients with suboptimal
volume status, resulting in further reduction in cardiac blood volume. The dimin-
ished cardiac output and reduction in cardiac blood volume may be manifest
radiographically as a decrease in heart size.
However, the decrease in heart size does not necessarily indicate an
improvement in the cardiovascular status of the patient. Conversely, the increase
in heart size, which may occur when PPV and PEEP are discontinued or
decreased to lower levels, does not necessarily indicate cardiac disease or uid
overload. Therefore, changes in heart size in patients receiving mechanical ven-
tilation should be carefully interpreted along with information on current venti-
lator settings and changes in settings from the time of the previous radiograph.
The effect of PPV and PEEP on pulmonary vasculature is complex, and
large radiographically visible pulmonary vessels can either increase or decrease
in size depending on lung volume, ination pressures, and pre-existing pulmonary
vascular tone (105). Because the air to tissue ratio increases when the patient
is placed on PPV or PEEP, the post-PPV or -PEEP chest radiograph may
show signicant apparent improvement despite the presence of severe residual
underlying disease (106 110).
Additional lung injury, manifest as non-cardiogenic pulmonary edema or
diffuse alveolar damage, may occur with PPV and PEEP. Experimentally, the
threshold peak inspiratory pressures (PIP) that cause lung injury appear to be
approximately 25 to 30 cmH2O (111), that is, the same PIP at which perivascular
interstitial emphysema (PIE) occurs (110). Further increases in PIP produce more
severe lung injury (112). On the basis of the experimental studies, it has been rec-
ommended that maximum transalveolar pressure should not exceed 30 to
35 cmH2O, which usually corresponds to 45 cmH2O end-inspiratory pressure
(110).
The other major complication of mechanical ventilation is air-leak
phenomena, PIE and its sequellae, commonly recognized manifestations of baro-
trauma and oxygen toxicity. The incidence of barotrauma in the general
population of ventilator dependent patients varies from 4% to 15% (113,114)
and rises to as high as 60% when underlying lung disease, such as pneumonia
or acute respiratory distress syndrom (ARDS), is present (115).
Although PIE may be seen in any age, the majority of severe cases have
been reported in infants, children, and adults under 40, suggesting that peri-
vascular and connective tissues in older adults are less easily dissected than in
younger patients (105). Ventilator-related risk factors for development of air
leak include large tidal volumes, PIP greater than 40 cmH2O, plateau pressures
greater than 35 cmH2O, high levels and long-term duration of PEEP, and
higher minute ventilation. A PIP of 40 cmH2O or greater is usually required to
356 Barack et al.
produce radiographically visible PIE (116,117). While many studies imply an

association of high-peak airway pressures and the development of PIE, other
studies suggest that alveolar overdistension or volutrauma may be a more import-
ant factor than PIP in the development of PIE (118 120).
PIE occurs when pressure in the air spaces exceeds tension in the adjacent
perivascular connective tissues and interlobular septa, resulting in alveolar rupture
and air entering the adjacent interstitium producing interstitial emphysema (121).
Air then dissects centrally to the hilum along the vascular sheaths and interlobular
septa, resulting in pneumomediastinum. As there is no gradient between alveoli
and bronchi, peribronchial air dissection does not occur. PIE has not been ident-
ied in alveolar septal walls or in the bronchial walls (105).
The progressive dissection of air from the interlobular septa into the con-
tiguous subpleural connective tissues results in subpleural air cyst formation.
The subpleural air collections can become much larger than the air collections
within the lung interior because of the larger potential size of the subpleural
space. Air cysts greater than 1 cm in diameter are sometimes called pneumato-
celes. Extensive PIE can signicantly impact cardiopulmonary function by pro-
ducing progressive collapse of pulmonary vessels, resulting in increased vascular
resistance and vascular shunting (122). Veins are more compressed than arteries
because of their thinner walls and less distending pressure.
The preferential location of PIE in larger interlobular septa, hilar regions,
and the mediastinum and the greater compression of central veins than arteries
result in decreased pulmonary venous return to the heart. This may explain the
atypical acute pulmonary edema pattern that may be observed in cases of
severe PIE (105). In this setting, the peripheral portions of the lungs demonstrate
marked increased radio-opacity, resulting in a reversal of the usual central
perihilar or bat wing pattern of acute pulmonary edema (Fig. 14). Rapid clearing
of the peripheral edema is observed when the acute air-block is removed (123)
by either decreasing the pressures used during PPV or by sudden decompres-
sion of air into the soft tissues of the neck, pleural space, mediastinum, or
abdomen (124).
The diminished functioning lung volume, a consequence of the part of the
lung volume occupied by PIE, is usually only clinically signicant in patients
whose pulmonary function is already severely compromised or in cases of
severe, extensive PIE. However, the stiffening of the lungs or diminished lung
compliance that occurs because of the distended interstitial air pockets may
have far more clinical signicance. The increased lung stiffness or diminished
lung compliance increases resistance to both inspiration and expiration (121).
Thus, a repetitive cycle may be created in which PIE increases airway resistance,
requiring higher peak pressures, which forces more air into the interstitial com-
partments, mediastinum, and soft tissues, further increasing airway resistance.
This results in progressively stiffer and severely hyperexpanded lungs (105).
The radiographic recognition of PIE usually occurs before its clinical
recognition because deterioration of gas exchange, lung compliance, and
Figure 14 Pulmonary edema in pulmonary interstitial emphysema. Antero-posterior

radiograph in a patient with severe pulmonary interstitial emphysema demonstrates bilat-
erally symmetric peripheral airspace disease consistent with acute pulmonary edema. The
peripheral distribution is the reversal of the usual central perihilar or bat-wing pattern seen
in acute cardiogenic pulmonary edema.
vascular resistance may be due to causes other than barotrauma (124). The early
stage of PIE prior to formation of large air cysts is easier to recognize in the pre-
sence of signicant concomitant alveolar consolidation. A mottled increase in
radiolucency of the lung anteriorly and medially around the heart and surface
of the diaphragm is the rst chest radiograph abnormality detected in adults
(105) (Fig. 15). This corresponds to the hyperinated pulmonary lobules and
interstitial and subpleural linear and cystic air collections demonstrated on CT
scans (125,126). These lucencies are extremely difcult to recognize in patients
with underlying emphysema or subcutaneous emphysema in the absence of
consolidated lung. However, without CT correlation, the plain lm diagnosis
of PIE is difcult at this stage because of similar ndings in microabscesses,
hyperinated acini, honeycombing, and emphysema.
Streaky, non-branching, xed caliber radiolucencies radiating from the hila
to the lung periphery corresponding to air in the perivascular connective tissue
also may be identied as an early radiographic sign of PIE and are more
suggestive of its diagnosis. The non-branching, xed caliber appearance of
these lucencies distinguishes them from the branching, peripherally tapering
appearance of air bronchograms. Rarely, perivascular halos, which are pathogno-
monic for PIE, may be seen (127). Other, more disorganized streaky
358 Barack et al.
Figure 15 Early pulmonary interstitial emphysema. Antero-posterior radiograph in a

patient, who subsequently developed a tension pneumothorax as a manifestation of baro-
trauma, demonstrates a mottled increased radiolucency medially around the heart and hila
(arrowheads). This nding, although non-specic, has been described as one of the earliest
plain lm ndings in pulmonary interstitial emphysema.
radiolucencies probably correspond to air within the interlobular septa (105). The
subpleural cystic air collections that can be seen on CT scans in patients with
ARDS may originate in the interlobular septa.
The presence of subpleural air cysts is the most frequent plain lm radio-
graphic nding in PIE and often precedes a pneumothorax (124,128). Subpleural
air cysts may become large and may be effectively treated by percutaneous cath-
eter drainage (129). They also may become secondarily infected, demonstrating
progressive thickening of the cyst wall or an air uid level, with subsequent
rupture producing either a pneumothorax or tension pneumothorax (105,128).
Pneumomediastinum has been found in 37% of patients with PIE (117) and
has been reported to precede the appearance of pneumothorax in 50% of patients
with ARDS (115). On a supine radiograph, pneumomediastinum may be difcult
to distinguish from pneumopericardium, anteromedial pneumothorax, and an
optical illusion, the Mach effect, which is an apparent line of contrasting density
bordering a soft tissue shadow (130). All of these entities may produce a sharp
outline of the heart on the supine radiograph. The radiographic differentiation of
these four entities is mandatory given the potentially life threatening complication
of evolution to a tension pneumothorax in mechanically ventilated patients.
Pneumopericardium is more commonly observed in the pediatric than the

adult population as a complication of barotrauma (131). It is easily diagnosed
when the superior pericardial reection around the great vessels is outlined by
air and may also be diagnosed when the main pulmonary artery and undivided
right pulmonary artery are outlined by air, which is uncommonly seen (105).
When pneumopericardium cannot be differentiated from pneumomediastinum
on the supine radiograph, pericardial air may be distinguished from mediastinal
air by a change in the position of pericardial air on radiographs exposed in
different body positions (132). Pericardial air will rise on upright or decubitus
lms, whereas mediastinal air will remain xed.
Pneumomediastinum can be differentiated from the Mach effect by
visualizing air in regions of the mediastinum not enhanced by the Mach effect.
Similarly, pneumomediastinum may be distinguished from pneumothorax
when air outlines the borders of anatomical structures not normally visible in
the presence of a pneumothorax (133,134). Rarely, decubitus lms or CT may
be required to differentiate these two conditions (105).
Pathognomonic signs of pneumomediastinum include air dissecting along
the medial border of the superior vena cava, left subclavian artery, common
carotid artery and right innominate artery, and the continuation of these air dis-
sections into the neck (105). Air also can be seen around the central pulmonary
arteries, the ring around the artery sign (135) (Fig. 16), and the ascending aorta
and dissecting the pericardial fat pads.
Because of the continuity of the right and left sides of the mediastinum, air
interposed between the heart and diaphragm may permit identication of the
central portion of the diaphragm in continuity with its lateral portions, the
continuous diaphragm sign (133). This will not be seen in the presence of a
subpulmonic pneumothorax, as the latter will not cross the midline. A pneumo-
pericardium may permit visualization of the central portion of the diaphragm.
However, as pneumopericardium is almost always associated with the presence
of pericardial uid, the uid will silhouette the central portion of the diaphragm
on an upright lm, allowing it to be differentiated from a pneumomediastinum.
The incidence of pneumothorax in all patients on mechanical ventilation
varies from 4% to 15% (105). The incidence in patients with ARDS varies
from 25% to 87% (105,136). The incidence of pneumothorax in patients with
PIE has been reported to be 77% (117). In mechanically ventilated patients, pneu-
mothoraces tend to increase rapidly and progress to tension in 60% to 90% of
cases (107,137,138). Pneumothorax is associated with increased mortality in
patients with ARDS (121,136).
The radiographic diagnosis of pneumothorax in a supine radiograph
requires a detailed knowledge of anatomy of pleural recesses and mediastinal
structures (139,140) and a thorough understanding of gravity and lung compli-
ance, which determine the movement of pleural air (141,142). Therefore,
pleural air is usually not visualized in the apicolateral position as occurs in an
upright lm, but more likely occurs anteromedially, as this represents the
360 Barack et al.
Figure 16 Pneumomediastinumthe ring around the artery sign. (A) Close-up postero-
anterior view of the left hilum demonstrates the left main pulmonary artery outlined by air
(arrows), the ring around the artery sign, and a small amount of air inferiorly within the
mediastinum (arrowhead). (B) Lateral radiograph demonstrates air surrounding the left
main pulmonary artery (arrows).
highest position in the hemithorax of supine patients. The size of a pneumothorax

in an ICU patient has poor correlation with its clinical signicance. Therefore, it
is not relevant that the size of a pneumothorax cannot be calculated on a supine
radiograph.
In a study of supine radiographs in 88 ICU patients, anteromedial and sub-
pulmonic pneumothoraces predominated, with apicolateral pneumothoraces
being relatively uncommon (140). Because air preferentially accumulates in
front of the lung surrounding the anterior mediastinal structures in the supine pos-
ition (143), sharp visualization of the superior vena cava and azygous vein on the
right and left subclavian artery as it curves over the apex of the left lung are the
initial signs of anteromedial pneumothorax. The only signs of a small pneu-
mothorax conned to the medial pleural space may be sharp visualization of
the left superior intercostal vein and the superior pulmonary veins (105). Bilateral
anteromedial pneumothoraces result in a sharply outlined anterior junction line
(144). Infrahilar anteromedial air collections result in sharp delineation of the
heart border, a lucent cardiophrenic sulcus, a well-outlined pericardial fat pad
which may simulate a mass or segmental collapse (145), and sharp delineation
of the inferior vena cava as it enters the right atrium (105).
A hyperlucent upper quadrant of the abdomen, with visualization of the
superior aspect of the diaphragm to the midline, is seen in subpulmonic pneu-
mothorax (146). Other signs of subpulmonic pneumothorax include abnormal
deepening and lucency of the lateral costophrenic angle, the deep sulcus sign
(147), a sharp depressed ipsilateral hemidiaphragm, and visualization of the
inferior edge of a collapsed or consolidated lower lobe (Fig. 17). When subpul-
monic air occupies the anterior subpulmonic space, the superior position of
the anterior diaphragmatic sulcus in relation to the more inferior position of the
posterior diaphragmatic sulcus is easily appreciated (148). Occasionally, one
may see both the anterior surface and the dome of the diaphragm sharply outlined
by air, the double-diaphragm sign (145) (Fig. 18).
The presence of an apicolateral pneumothorax in a supine radiograph
implies that a large volume of pleural air has already accumulated and has dis-
placed the visceral pleura medially (143). The rst sign of an apicolateral pneu-
mothorax in the presence of a small amount of air is lack of contact of the minor
ssure with the chest wall (105). As air progressively accumulates, the visceral
pleura is further displaced medially and eventually becomes tangential to the
X-ray beam, and the pneumothorax becomes easier to diagnose (143).
The concomitant presence of subpleural consolidated lung and apicolateral
pleural uid may mask the presence of an apicolateral pneumothorax (105).
In this instance, the pleural line, which requires contiguous normal aerated
lung and pleural air to be seen, is obscured by the subpleurally contiguous con-
solidated lung, and the pleural surface of the consolidated lung is silhouetted by
the pleural uid.
In posteromedial pneumothorax, the posterior mediastinal structures, para-
spinal line, descending aorta, and the costovertebral sulcus may become outlined
362 Barack et al.
Figure 17 Subpulmonic pneumothorax in supine radiographthe deep sulcus sign.

Antero-posterior supine radiograph in a patient with barotrauma demonstrates a deep-
ened, lucent left costophrenic angle (arrows), the deep sulcus sign, consistent with a
subpulmonic pneumothorax. The inferior edge of the left lower lobe is visualized (small
arrowheads).
by air, and the medial surface of the lower lobe may be displaced laterally
(Fig. 19). Posteromedial pneumothorax is associated with lower lobe volume
loss and parenchymal disease, both of which are common in the ICU patient.
As air preferentially surrounds the surface of abnormal lung in lobar collapse,
the air in a posteromedial pneumothorax continues to surround the abnormal
collapsed lower lobe and fails to rise to less dependent pleural surfaces, even
when it is not loculated.
When the pressure in the pleural space exceeds atmospheric pressure, a
pneumothorax is considered to be under tension. Both a simple and a tension
pneumothorax may result in lung collapse and mediastinal shift (149).
However, both lung collapse and mediastinal shift may be absent in a tension
pneumothorax because of pleural adhesions and because of the stiffness of the
lungs in ARDS (150). Hence, a residual small loculated collection of air can
cause a tension pneumothorax even when the ipsilateral hemithorax is drained
by a thoracostomy tube.
Figure 18 Pneumothorax in supine radiographthe double diaphragm sign. Antero-

posterior radiograph in a patient with barotrauma and a right pneumothorax demonstrates
the anterior aspect (black arrows) and the dome (black arrowheads) of the right hemi-
diaphragm, the double diaphragm sign. The presence of air in the apicolateral region
(white arrows) indicates that a large amount of air has already accumulated in the right
hemithorax, displacing the visceral pleural medially.
Simple pneumothoraces in mechanically ventilated patients progress to

tension in 60% to 90% of cases (105). Signs of tension include inversion of
the diaphragm, contralateral shift of the heart and mediastinum, contralateral
displacement of the anterior junction line, and displacement of the azygoesopha-
geal recess. Flattening of the heart border and other vascular structures including
the superior and inferior vena cava, the radiographic reection of diminished
venous return to the right heart (140), has been considered to be the most specic
sign of tension and correlates with the clinical ndings (105) (Fig. 20).
As a result of pneumomediastinum and pneumothorax, air may be found
within the interlobar ssures, subcutaneously, and in the peritoneal and retroper-
itoneal spaces. Air may collect within the interlobar ssures with or without
pleural uid, resulting in apparent air uid levels or pneumatoceles (126,151).
Subcutaneous emphysema is a benign complication of pneumomediastinum
and pneumothorax, which resolves as the latter improves (105). Pneumoperito-
neum and pneumoretroperitoneum may occur with either pneumomediastinum
(152) or pneumothorax (153). In patients with other signs of air leak, pneumoper-
itoneum may be assumed to be due to the air leak (124) rather than a perforated
viscus (154).
364 Barack et al.
Figure 19 Left posteromedial hydropneumothorax. (A) Antero-posterior radiograph in

a patient with a left chest tube for treatment of a pneumothorax complicating barotrauma
demonstrates a persistent left posteromedial pneumothorax (P). The medial aspect of the
consolidated left lower lobe (arrowheads) is displaced laterally. (B) Axial computed tom-
ography image (lung window) demonstrates left posteromedial hydropneumothorax (P)
displacing the consolidated left lower lobe laterally.
IV. Common Thoracic Abnormalities Encountered

in the ICU
Common thoracic abnormalities encountered in the ICU include those of cardio-

vascular, pulmonary, and pleural origin and those associated with trauma. Chest
radiographic abnormalities may be present at the time of ICU admission or may
Figure 20 Tension pneumothorax. Antero-posterior radiograph in a patient with baro-

trauma and clinical signs and symptoms of tension demonstrates a large left pneumothorax
with shift of the heart and mediastinum to the right and inversion of the left hemidiaph-
ragm. There is incomplete collapse of the left upper lobe (white arrows) and the left
lower lobe (white arrowheads), reecting the decreased lung compliance seen in baro-
trauma. Flattening of the left heart border (black arrowheads) is demonstrated. This is con-
sidered to be the most specic radiographic sign of tension and correlates with the clinical
ndings.
develop during the course of the ICU stay. The etiology of any abnormality
should be interpreted in conjunction with the clinical history and the prior radio-
graphs (11).
A. Atelectasis
Atelectasis by denition means incomplete expansion, and it is used to describe
any condition where lung volume loss occurs. Atelectasis is a common radio-
graphic nding in ICU patients for a multitude of reasons including, but not
limited to, an impaired cough reex, supine position, and hypoventilation.
Most frequently, atelectasis is observed in the left lower lobe where the heart
causes mass effect on the left lower lobe bronchus in the supine position (155).
The radiographic appearance of atelectasis varies from undetectable in the
mildest cases to complete opacication of the affected hemithorax in cases of
total lung collapse. The latter represents the most extreme form of atelectasis.
Atelectasis can be described as subsegmental, segmental, or lobar, each of
which has characteristic radiographic ndings.
Subsegmental atelectasis manifests as linear, plate-like opacities, predomi-
nantly at the lung bases or lower lung zones. Segmental atelectasis has a
366 Barack et al.
triangular or wedge-shaped appearance. Lobar atelectasis, or lobar collapse, is a

common occurrence in the ICU due to mucous plugging. The appearance of lobar
collapse depends on the affected lobe. The hallmark of lobar collapse is the dis-
placement of the interlobar ssures. Important secondary signs of collapse
include displacement of the hila, shift of the mediastinum, compensatory hyper-
ination of the normal lung, and elevation of the hemidiaphragm (156 159).
Right upper lobe collapse causes a homogeneous opacity extending from
an elevated hilum towards the apex. There is mild shift of the trachea to the
right, a preserved tracheal stripe, and elevation of the minor ssure. Right
middle lobe collapse can be subtle in the anterior posterior projection, resulting
in a vague opacity in the right lower lung zone, which silhouettes the right heart
border. In addition, there is inferomedial displacement of the minor ssure.
Right lower lobe collapse causes a homogeneous opacity in the right retrocardiac
region, which silhouettes the medial aspect of the right hemidiaphragm. The right
hilum appears small secondary to incorporation of the right interlobar artery
within the collapsed lobe. The right heart border is preserved because the right
middle lobe is normally aerated (158).
Left upper lobe collapse is recognized by a poorly dened left perihilar
opacity, the veil sign, extending upward along the medial aspect of the left
upper lung zone. Because the left upper lobe collapses antero-medially, the
aortic arch and left hilum are visualized through the opacity since they are
middle mediastinal structures. Occasionally, herniation of the superior segment
of the left lower lobe between the collapsed upper lobe and the mediastinum
produces an air crescent that highlights the aortic arch, the Luftsichel sign.
Left lower lobe collapse results in a sharply marginated retrocardiac opacity,
which silhouettes the medial aspect of the left hemidiaphragm (158).
Atelectasis is difcult to differentiate from pneumonia. Although the signs
of volume loss discussed earlier are suggestive of atelectasis, they are often not
apparent, or they may co-exist with pneumonia. In the ICU, changing appearance
and rapid clearing of parenchymal opacities followed by their recurrence
suggests the diagnosis of atelectasis (12).
B. Pneumonia
Types
Pneumonia is a frequent cause of ICU admission, as well as a complication of
hospitalization (nosocomial). The radiological ndings in pneumonia are
varied. The unilateral and focal airspace opacication of pneumonia may be
difcult to distinguish from atelectasis, pulmonary edema, or hemorrhage.
These entities comprise the differential of airspace disease. In one study, the
accuracy of bedside radiography in diagnosing pneumonia in ventilated patients
was only 50% (160). Generally, pulmonary opacities due to infection appear later
and resolve slower than do those seen with aspiration or atelectasis (161).
Lobar pneumonia classically appears as a homogeneous, conuent opacity,

delineated by ssures, which may be expansile. However, the lobar stage is
less commonly seen today due to the early institution of antimicrobial therapy.
Air bronchograms are a common nding. The expansile pattern is most fre-
quently observed in pneumonia due to Klebsiella (162).
Bronchopneumonia, or lobular pneumonia, presents with multiple small,
indistinct nodules, corresponding to peribronchiolar inammation. Subsequently,
these nodules enlarge to form multifocal, ill-dened opacities (163). As the
pattern coalesces, it may difcult to differentiate from lobar pneumonia.
The incidence of nosocomial bacterial infections of the respiratory tract in
inpatients is 0.5% to 5% (164 167). Over 15% of patients hospitalized for
treatment of pneumonia acquire bronchopulmonary superinfections, some of
which prove fatal (168). A multilobar pattern of involvement is frequently
seen, particularly in patients with an altered immune system.
Aspiration
Predisposing factors for aspiration pneumonitis include vomiting, gastroesopha-
geal reux, altered mental status, intubation, and tracheostomy. Radiographic nd-
ings develop within 24 hours and commonly occur in the dependent portions of the
lung with a basilar distribution. However, in the supine patient, airspace opacities
may be seen in the upper lung zones due to aspiration into the posterior segments
of the upper lobes or the superior segments of the lower lobes. The radiographic
ndings of simple toxic aspiration begin to clear within 48 hours (169). Persist-
ence of opacities beyond this period suggests a complicating infection.
Complications of Pneumonia
Two of the more serious complications of pneumonia include lung abscess and
empyema. These entities may be difcult to diagnose and to differentiate by
plain chest radiography.
Lung Abscess
Lung abscess most frequently develops in pneumonias associated with infectious
aspiration, bronchial obstruction, and an immunocompromised state. A lung
abscess may also be a complication of septic emboli. Lung abscesses typically
develop one to two weeks after pneumonia. Approximately 20% are not apparent
on plain chest radiographs because a cavity is not evident (170).
The plain lm radiographic ndings of lung abscess are single or multiple
cavities, isolated or within areas of consolidation. In one study, 88% of cavities
demonstrated a smooth internal margin, and 72% of cavities contained air uid
levels (170). The wall thickness varies from 4 to 15 mm (171). Abscesses demon-
strating irregular margins and thick walls may simulate cavitary tumors radiogra-
phically and cannot be differentiated by MDCT. Lung abscesses may result in
formation of pneumatoceles, which can either resolve or persist indenitely
368 Barack et al.
(170). A rare and more serious complication of lung abscess is the development
of a bronchopleural stula (Fig. 21).
Empyema
Empyema is a collection of pus in the pleural space. Empyemas often progress
from parapneumonic effusions or may occur as a consequence of hematogenous
spread of infection, penetrating trauma, or any procedure disrupting the pleural
space.
Neither the plain lm nor the CT ndings of empyema are specic.
Thickening and enhancement of the visceral and parietal pleura, the split
pleura sign, may be seen in empyema and parapneumonic and malignant
pleural effusion (170). An air uid level may be present in an infection resulting
from a gas-forming organism. CT may be necessary to demonstrate the location
of the air uid level. In an empyema or a bronchopleural stula, the air uid
level will be in the pleural space. In contrast, the air uid level will be in the
lung parenchyma in a lung abscess. CT may also be useful in delineating
pleural uid loculations prior to drainage. Rare and more serious complications
of empyema include the spontaneous drainage through the chest wall (empyema
necessitans) or the development of a bronchopleural stula, which occurs
secondary to erosion into the tracheobronchial tree.
Figure 21 Bronchopleural stula. Axial computed tomography image (lung window)

demonstrates a bronchopleural stula (white arrow) between a subsegmental branch of
the superior segmental bronchus of the right lower lobe and the pleural space. A hydro-
pneumothorax is clearly visible.
C. Pulmonary Edema
Introduction
Pulmonary edema represents an abnormal increase in extravascular water in the
lung and is commonly classied into two major categories, hydrostatic edema
and increased permeability edema (172). However, it is not possible to group
all patients into these two categories because many patients have pulmonary
edema resulting from both hydrostatic and increased permeability mechanisms.
Therefore, a third category, mixed pulmonary edema, has been created (173).
The three major categories of pulmonary edema and their subcategories
with disease groups are shown in Table 4.
Hydrostatic pulmonary edema can be further divided into cardiogenic
edema, decreased capillary osmotic pressure edema, and edema associated
with disease of the pulmonary veins. The subcategory of cardiogenic edema
includes patients whose edema results from left ventricular failure and mitral
valve disease. The subcategory of decreased capillary osmotic pressure edema
includes patients whose edema results from renal failure, uid overload, or
hypoproteinemia (cirrhosis and nephrotic syndrome). A decrease in serum
osmotic pressure can contribute to hydrostatic edema. Edema secondary to
disease of the pulmonary veins occurs with pulmonary veno-occlusive disease
and brosing mediastinitis. This form of edema is primarily hydrostatic but is
incompletely understood. Although this subclassication is useful for further
grouping patients according to the pathophysiology of the edema, these three
subcategories of hydrostatic pulmonary edema cannot be differentiated
radiographically.
Similarly, the category of increased permeability edema may be subdivided
into two subcategories, increased permeability edema with diffuse alveolar
damage (DAD) and increased permeability edema without DAD. The subcate-
gory of increased permeability edema with DAD includes ARDS. The subcate-
gory of increased permeability edema without DAD includes the edema
associated with cytokine administration, high altitude, and heroin overdose.
These two subcategories of increased permeability edema also cannot be
radiographically differentiated from each other.
Acute lung injury (ALI) refers to the spectrum of acute diffuse pulmonary
injury characterized by acute onset, gas exchange abnormalities (PaO2/
FIO2 , 300 mmHg regardless of PEEP level), and diffuse bilateral opacities in
the absence of an elevated pulmonary wedge pressure (172). For practical
purposes and discussion, increased permeability edema and the edema associated
with ALI are synonymous with non-cardiogenic pulmonary edema.
Hydrostatic Pulmonary Edema

The most common causes of hydrostatic pulmonary edema are cardiogenic
edema as a result of left ventricular failure and mitral valve disease.
370
Table 4 Pulmonary Edema in ICU Patient
Increased permeability edema

Increased hydrostatic pressure (acute lung injury, non-cardiogenic edema)
Decreased capillary With diffuse Without diffuse

Cardiogenic Other osmotic pressure alveolar damage alveolar damage Mixed edema
Left ventricular Disease of Hypervolemia Acute respiratory Cytokines Acute and chronic
failure pulmonary (uid overload) distress syndrome pulmonary emboli
veins
Mitral valve disease Renal failure High altitude Near drowning
Asymmetric Hypoproteinemia Heroin overdose Neurogenic
(cirrhosis, nephrotic
syndrome)
COPD Reperfusion
Patient position Lung transplantation
Mitral regurgitation Reexpansion
Post pneumonectomy
or lung reduction
Air embolism
Barack et al.
Cardiogenic Pulmonary Edema

The radiographic appearance of cardiogenic edema is dependent upon the sever-
ity of pulmonary venous hypertension, reected by the pulmonary capillary
wedge pressure (PCWP) (174). There is often a time lag between the elevation
of PCWP and the subsequent development of radiographic ndings (175). In
addition, the radiologic ndings due to an elevated PCWP usually persist after
the PCWP has returned to normal (176). These two observations may explain
the all too common discrepancy between the radiographic appearance and the
clinical appearance of the patient at any given moment.
On the portable supine radiograph, the earliest radiographic sign of pul-
monary venous hypertension is redistribution of blood ow from the lower to
the upper lung zones, manifested by an increase in caliber of the upper lobe
vessels. This nding is difcult to appreciate in the supine position (177,178)
and begins to become apparent when the PCWP exceeds 12 mmHg. Assessment
of redistribution is performed by comparing the number and size of the upper
lobe vessels to lower zone vessels at an equal distance from the hila (179,180).
A more objective assessment is direct measurement of the diameter of the
bronchi and adjacent pulmonary arteries visualized end-on. In the supine pos-
ition, the ratio of the artery to bronchus is 1.01 (SD: 0.31). With redistribution
in left ventricular failure, the ratio is 1.49 (SD: 0.31) (181).
Interstitial pulmonary edema develops when the PCWP exceeds 17 to
20 mmHg (179). This results in uid accumulation in the perivascular, inter-
lobular septal, and parenchymal interstitial tissue (173). Radiographic ndings
of interstitial pulmonary edema are loss of the normal sharp denition of subseg-
mental and segmental pulmonary vessels, thickening of the interlobular septa and
of the ssures, peribronchial cufng, pleural effusions, and perihilar and lower
lobe haziness. Interlobular septal thickening is manifested on the chest radio-
graph as Kerley B lines, which are short lines perpendicular to the pleura.
Kerley B lines are most commonly observed in the costophrenic angles (Fig. 22).
Alveolar edema, or uid accumulation in air spaces, results from a further
increase in pulmonary venous pressure and occurs when PCWP exceeds
25 mmHg. Characteristic radiographic ndings are bilaterally symmetric
patchy or conuent opacities with a perihilar and lower lung zone distribution.
Air bronchograms can be seen in 10% to 30% of patients (182). An uncommon
manifestation of cardiogenic pulmonary edema is the bat-wing or buttery
pattern (Fig. 23). This appearance has been reported in 5% of the patients with
edema of varying etiology and is most commonly associated with acute left
ventricular failure (173). Radiographically, there is perihilar opacication with
sparring of the lung periphery.
Asymmetric cardiogenic edema may be caused by chronic obstructive
pulmonary disease (COPD), patient positioning, and mitral regurgitation.
COPD results in asymmetric pulmonary edema depending on the degree of the
underlying lung parenchymal disease. The edema is more evident in regions
372 Barack et al.
Figure 22 Cardiogenic pulmonary edemaKerley B lines. Close-up postero-anterior

view of the right lower lung zone in a patient with cardiomegaly, pulmonary venous hyper-
tension, and interstitial edema clearly demonstrates the Kerley B lines (arrowheads).
containing normal lung parenchyma in comparison to diseased lung parenchyma

(173). Asymmetric cardiogenic edema is a manifestation of the effect of gravity
on the patient and reects patient position with increased edema in the dependent
portion of the lungs. Mitral regurgitation causes a right upper lobe distribution of
abnormalities because of reux directed towards the right upper lobe pulmonary
vein (183,184).
The CT ndings of cardiogenic edema substantiate the ndings seen on plain
lms. Additionally, enlarged mediastinal lymph nodes have been reported in 80%
of patients and inhomogeneous attenuation of the mediastinal fat in 59% (185).
Increased Permeability Pulmonary Edema

(Non-cardiogenic Pulmonary Edema)
ARDS
ARDS is a clinical entity characterized by acute onset, gas exchange abnormal-
ities (PaO2/FIO2 , 200 mmHg regardless of PEEP level), and diffuse bilateral
opacities compatible with pulmonary edema in the absence of an elevated
pulmonary capillary wedge pressure (172). ARDS represents the most severe
end of the spectrum of ALI or increased permeability edema with DAD.
Figure 23 Acute cardiogenic pulmonary edemathe bat-wing pattern. Antero-

posterior radiograph in a patient with acute myocardial infarction demonstrates bilateral peri-
hilar air space opacication, the bat-wing pattern, characteristic of acute pulmonary edema.
Pathologically, ARDS is characterized by three phases. The rst, or exuda-

tive, phase begins with the leakage of protein-rich uid into the alveoli and is
associated with hemorrhage, type II pneumocytes proliferation, and hyaline
membrane formation. The second, or proliferative, phase of ARDS is character-
ized by formation of brinous exudate with thickening of the alveolar septa. The
third, or reparative, phase demonstrates either complete recovery, varying
degrees of brosis resulting from ARDS, and brosis and lung cysts as a
complication of mechanical ventilation (173).
Radiographically, the earliest ndings in the exudative phase are seen
approximately 12 hours after the clinical onset of respiratory failure. These
ndings consist of patchy, ill-dened opacities throughout both lungs (172).
Interstitial edema may be present and tends to have a more peripheral distribution
than the edema of cardiac origin (186 188). The heart size is usually normal.
Within 24 72 hours, there is rapid coalescence of the patchy opacities into
diffuse air space consolidation, which extends from the lung apex to the lung
base and into the lung periphery. Air bronchograms are frequent, but pleural
effusions are typically not seen (189).
The proliferative and brotic phases become apparent in approximately
one week. The lungs remain diffusely abnormal but begin to demonstrate a
374 Barack et al.
reticular pattern (186,188) corresponding to the beginning of interstitial and

airspace brosis. Similar to the exudative phase, pleural effusions are not typi-
cally seen. The appearance of pleural effusion at any time on a supine radiograph
suggests a complication such as pneumonia or pulmonary infarction. In general,
the radiograph shows improvement within 10 to 14 days in those patients who
survive. Failure to improve suggests a superimposed process (pneumonia) and
has a poor prognosis (190 191).
The radiologic course of ARDS differs from other entities, causing diffuse,
coalescent opacities. Specically, the radiologic abnormalities persist for a
longer period of time or may resolve completely. In those cases with incomplete
resolution, scarring, atelectasis, or traction bronchiectasis may be evident. Com-
plications of mechanical ventilation such as subpleural and intrapulmonary lung
cysts are also frequently noted.
As emphasized in the section on Complications of Mechanical Ventilation
(Barotrauma), mechanical ventilation and PEEP can dramatically alter the
radiographic appearance of diffuse airspace consolidation. This is particularly
true during the exudative phase of ARDS. Diffuse airspace consolidation may
clear within minutes following the institution of PEEP despite obvious clinical
deterioration.
The CT ndings in ARDS have been extensively described and depend on
the stage of ARDS (192 195). They are non-specic and cannot be used to
differentiate ARDS from severe cardiogenic pulmonary edema or widespread
bacterial pneumonia.
Radiographic Distinction Between Cardiogenic and

Non-Cardiogenic Edema
Although non-cardiogenic and cardiogenic pulmonary edema have distinct radio-
graphic characteristics, it may be difcult to distinguish between them. In the
ICU patient, this distinction is even more difcult because of the patients
supine position. In the supine patient, the distribution of blood ow has not
been found to be helpful in distinguishing between these two entities
(177,178). The distinction is further complicated in the ICU patient with
ARDS who may have associated cardiac failure and uid overload and therefore
demonstrate the characteristics of mixed edema. In general, it is easier to dis-
tinguish between cardiogenic and non-cardiogenic edema in patients who have
mild to moderate edema, on erect radiographs, and when serial radiographs are
available for comparison. The most useful radiographic ndings in distinguishing
between cardiogenic and non-cardiogenic edema are shown in Table 5.
While no one nding allows reliable distinction between cardiogenic and
non-cardiogenic edema, and imaging ndings overlap, the combination of nd-
ings as summarized in Table 5 has been reported to correctly distinguish
the two in most instances. Cardiogenic or hydrostatic pulmonary edema has
been accurately diagnosed radiographically in 80% to 90% of patients and
permeability pulmonary edema in 60% to 90% (178,182,196).
Table 5 Useful Radiographic Findings in Differentiating Non-cardiogenic or Increased

Permeability Edema in ARDS from Cardiogenic or Increased Hydrostatic Edema in the
ICU Patient
Non-cardiogenic Cardiogenic
Finding edema edema
Heart size Normal Commonly enlarged

Cardiothoracic ratio .0.58
on supine radiograph
Septal lines Uncommon Common
Vascular pedicle Normal Widened in chronic congestive
heart failure with left
ventricular failure;
normal in acute left
ventricular failure
Air bronchograms Common Uncommon
Pleural effusion Uncommonly seen in Small effusions
supine radiograph common
Distribution of edema Peripheral or diffuse Perihilar or diffuse
(perihilar more common in
renal failure, bat-wing in
acute left ventricular failure)
Upper lobe vessels Normal Increased caliber
on erect lm Ratio upper: Ratio upper:lower .1:2
lower 1:2
Thickening of Absent Present
interlobar ssures
Source: From Ref. 172.
D. Pleural Effusion
Pleural effusion represents an abnormal increase in the quantity of uid in the

pleural space. Normally, the pleural space contains 5 to 15 mL of uid. On the
basis of the composition of the uid, pleural effusions are classied as transudates
or exudates. Transudative pleural effusions are non-inammatory and develop as
a result of increased venous pressure and decreased plasma oncotic pressure, as in
congestive heart failure, liver cirrhosis, or nephrotic syndrome. Exudative pleural
effusions are caused by pleural inammation or impaired lymphatic drainage and
are most commonly seen with infection and neoplasm (197).
Pleural effusions are a common occurrence in the medical ICU and are
found in up to 62% of patients (198). Common causes of pleural effusion
in the ICU are cardiac failure, pneumonia, and ascites. Less common causes
include pulmonary embolism, postpericardiotomy syndrome (Dresslers
syndrome), post-thoracic and abdominal surgery, and intra-abdominal processes
such as pancreatitis or subdiaphragmatic abscess.
376 Barack et al.
Initially, free pleural uid accumulates in the most dependent aspect of the
thorax, which is located postero-inferiorly in the supine patient. With increasing
volume, pleural uid tracks into the subpulmonic space and along the convexity
of the lung postero-laterally, resulting in silhouetting of the diaphragm, obliteration
of the lateral costophrenic angle, and formation of a lateral meniscus sign. Fluid
tracks into the posterior superior aspect of the major ssures and the lateral
aspect of the minor ssure as it accumulates. With a large pleural effusion, an
apical cap develops. In one study, the reported sensitivity and specicity for detect-
ing pleural effusions on supine radiographs were 67% and 70%, respectively (199).
In a prospective study, the minimum amount of pleural uid necessary
for radiographic visualization on the supine radiograph was 175 mL (200).
Radiographic signs associated with the presence of small pleural effusions
(175 525 mL) included increased homogeneous density in the lower lung
zone in 91% of cases and silhouetting of the diaphragm in 45% of cases. Moder-
ate pleural effusions (.525 mL) demonstrated increased homogenous density in
the lower lung zones in all cases, silhouetting of the diaphragm in 71% of cases,
and blunting of the lateral costophrenic angle in 25% of cases. Large pleural
effusions showed increased homogenous density in the lower lung zones in all
cases, silhouetting of the diaphragm in 68% of cases, blunting of the lateral cost-
ophrenic angle in 41% of cases, and apical capping in 54% of cases. Thickening
of the minor ssure was present in 48% of cases (200).
Other radiographic ndings of free pleural uid on supine radiographs
include apparent elevation of the hemidiaphragm, which is caused by subpulmo-
nic uid accumulation, and decreased visualization of the pulmonary vessels
below the apparent dome of the hemidiaphragm because of the added density
of subpulmonic uid. However, these ndings are difcult to appreciate due to
the frequent presence of bilateral effusions silhouetting both hemidiaphragms
and simulating low lung volumes and insufcient penetration of the upper
abdominal structures on the supine radiograph (200). A subpulmonic pleural
effusion is characterized by apparent lateral displacement of the dome of the
hemidiaphragm and a more acute lateral costophrenic angle.
On the supine radiograph, pleural effusion may be difcult to recognize
because of overlying breast tissue, an enlarged heart, or a prominent epicardial
fat pad. Pleural effusion may be simulated by any condition resulting in an
increase or decrease in the density of a hemithorax. Conditions resulting in
increased density, such as chest wall or pleural masses, may simulate an ipsilat-
eral pleural effusion. Conditions causing decreased density, such as prior mas-
tectomy, unilateral emphysema, or bullous disease, and anterior pneumothorax
or atrophy of the pectoralis muscle may simulate a contralateral effusion (201).
The lateral decubitus lm is the most sensitive technique for detecting free
pleural uid, with a sensitivity threshold of 5 mL (202). It may be performed to
identify the presence of pleural uid and to distinguish free from loculated pleural
effusions. Bilateral decubitus views are preferred as the contralateral decubitus
view often yields additional important information.
An unusual appearance of pleural effusion results from localization of uid

within an interlobar ssure. This has been termed a pseudotumor because it may
simulate a lung mass. A localized interlobar pleural effusion is most commonly
seen in heart failure. It appears spontaneously and is usually self-limited. It is
more common in the right hemithorax and more often in the minor than the
major ssure. On the frontal chest radiograph, the pseudotumor has a character-
istic elliptical conguration and is sharply circumscribed. If the X-ray beam is not
tangential to the entire margin of the pseudotumor, a portion of the margin may
appear indistinct, the incomplete border sign (203,204). CT demonstrates the
uid attenuation of the pseudotumor as well as its mobile nature, which
changes with different patient position (Fig. 24). Pleural uid may loculate
against the chest wall or within the ssures (Fig. 25). Such loculations may
occur following episodes of pleuritis, including empyema and hemothorax.
E. Embolic Disease
Pulmonary Embolism
Pulmonary embolism results from occlusion of a pulmonary arterial branch by an
embolus. The most frequent source of emboli is the deep systemic veins. Predis-
posing factors for the development of pulmonary embolism include prolonged
immobilization, surgery, trauma, and hypercoagulable states.
Figure 24 Pseudotumor in the major ssure. Postero-anterior radiograph demonstrates

a homogeneous opacity (arrows) in the right lower lung zone with blunting of the right
costophrenic angle. The superior margin of the opacity is indistinct, the incomplete
border sign.
378 Barack et al.
Figure 25 Pseudotumor in the minor and upper portion of the major ssures. (A)
Antero-posterior radiograph demonstrates a sharply circumscribed, homogeneous, hori-
zontally oriented opacity in the right mid lung zone lateral to the right hilum (arrowheads)
and a second obliquely oriented opacity extending medial and superior to the right hilum
(arrows). Subpulmonic uid spills into the lateral costophrenic sulcus and extends upward
along the lateral chest wall. (B) Lateral view of chest demonstrates the horizontally
oriented opacity to lie in the minor ssure (arrowheads) and the obliquely oriented
opacity to lie in the superior aspect of the right major ssure (arrows). Subpulmonic
uid spills into the anterior and posterior costophrenic sulci and extends upward along
the anterior and posterior chest wall.
The chest radiograph is reported to be 33% sensitive and 59% specic for
diagnosing pulmonary embolism (205). In a large prospective study, the most
common radiographic abnormalities associated with pulmonary embolism were
cardiomegaly (27%), pleural effusion (23%), elevated hemidiaphragm
(20%), pulmonary artery enlargement (19%), atelectasis (18%), and parenchymal
opacity (17%). The chest radiograph was normal in 24% of the cases (206).
Radiographic ndings classically associated with pulmonary embolism are
hyperlucency of the lung distal to the occluded arterial segment because of
regional oligemia, the Westermarks sign (207), a prominent central pulmonary
artery, the Fleischner sign, and an abrupt cutoff of the interlobar pulmonary
artery, the knuckle sign (208) (Fig. 26). These ndings are highly specic but
insensitive for the diagnosis of pulmonary embolism (209). Pulmonary embolism
can also cause segmental and lobar consolidation secondary to edema and
hemorrhage, mimicking the radiologic appearance of pneumonia (210).
Pulmonary infarction is uncommon and is present in only 15% of pulmo-
nary embolisms because of the dual pulmonary arterial and bronchial blood
supply of the lung. Pulmonary infarction is more likely to occur in patients
with heart failure or severe hypotension. The characteristic radiographic appear-
ance of infarction is a homogeneous pleural-based opacity with convex medial
contour towards the hilum, the Hamptons hump (211).
At institutions where MDCT is available, CT pulmonary angiography is
the study of choice for suspected pulmonary embolism. Studies have shown
that MDCT pulmonary angiography is highly sensitive and specic for the
diagnosis of pulmonary embolism. Discrepancies with conventional angiogra-
phy are mainly at the subsegmental level, where even angiographers have
poor inter-observer variability (212 214). The newer scanners have thinner
slice acquisition, higher resolution, faster scanning times, and less motion arti-
fact, resulting in superior visualization of the peripheral pulmonary arteries. All
these factors improve detection of pulmonary emboli. In a prospective study
comparing the effect of different slice thickness on study interpretation,
MDCT at 1.25 mm collimation has been found to signicantly improve visu-
alization of the segmental and subsegmental arteries. In this study, 96% of
lobar, 90% of segmental, and 75% of subsegmental arteries were well visual-
ized using 1.25 mm slice thickness. Inter-observer agreement for detection of
pulmonary embolism at the segmental and subsegmental level was also
improved when these parameters were used (215). Animal studies have demon-
strated no difference in the sensitivity of MDCT at 1.25 mm collimation and
pulmonary angiography in the detection of subsegmental emboli. Specicity
was reported to be slightly higher for angiography, 88% versus 81% for
MDCT (216).
Clinical studies have shown that a negative MDCT pulmonary angiogram
has high negative predictive value for subsequent development of a pulmonary
embolism. The risk of pulmonary embolism in patients with a negative MDCT
pulmonary angiography is less than 1% at six months (217) and nine months
380 Barack et al.
Figure 26 Pulmonary embolism without infarctionthe Westermark sign. (A) Postero-

anterior radiograph with magnication of the right hilum and right lower lung zone (insert)
demonstrates normal tapering of the right interlobar artery with normal branching vessels
in the right lower lung zone. (B) Postero-anterior radiograph with magnication of the
right hilum and right lower lung zone (insert) obtained two years later demonstrates an
enlarged, bulging right main pulmonary artery, the Fleischner sign. The right interlobar
pulmonary artery demonstrates an abrupt cutoff, the knuckle sign. Increased lucency is
seen in the right lower lung zone, the Westermark sign, in comparison to the left due to
decreased blood ow. (C) Axial computed tomography pulmonary angiogram demon-
strates a large acute clot (C) expanding the right main pulmonary artery.
after the study (218), and less than 2% at 12-month follow-up (219). Addition-
ally, a large prospective study has shown that MDCT reveals additional poten-
tially signicant ndings in up to 76% of patients, with 47% of these ndings
not suspected on chest radiography (218).
On CT, a pulmonary embolus appears as a central or eccentric lling defect

within an opacied pulmonary arterial branch, which may partially or totally
occlude the vessel. Expansion of the lumen may be seen in acute thromboembo-
lism. Additional ndings include consolidation, pleural-based opacities, and
pleural effusion (220).
Eighty percent of pulmonary emboli are associated with deep venous
thrombosis. A positive Duplex ultrasound identies patients at risk for develop-
ment of pulmonary emboli. However, a negative Duplex ultrasound evaluation
does not exclude the diagnosis of pulmonary embolism (221).
Pulmonary angiography has been the gold standard in the diagnosis of pul-
monary embolism (222,223). Its role today is still being dened. It will continue
to be the gold standard in those institutions where MDCT is not available.
However, in those institutions in which MDCT is available, the use of pulmonary
angiography should be reserved for the small subset of patients with a high
clinical probability of pulmonary embolism, a negative CT pulmonary angio-
gram, and a negative lower extremity Duplex ultrasound examination and the
rare case of acute massive embolism where thromboembolic therapy is being
considered (224 227). Selective pulmonary angiography may also be performed
in those patients with compromised renal function and an increased risk of
nephropathy from the amount of contrast necessary for a CT pulmonary angio-
gram (228), recognizing that the interpretation of a digital substraction image
is more susceptible to motion artifact.
Septic Emboli
Septic emboli result from the hematogenous dissemination of infection to the lung
parenchyma. Septic emboli occur more often in people under the age of 40 (229).
The most common sources of emboli are the heart in association with tricuspid
endocarditis, or a ventricular septal defect, or the peripheral veins secondary to
septic thrombophlebitis or an indwelling vascular catheter (230,231). Common
predisposing factors associated with septic emboli include drug addiction, alcohol-
ism, infections in immunocompromised patients, congenital heart disease, and
dermal infections (232). The most common causative organisms are Staphy-
lococcus aureus, Streptococcus viridans, and Streptococcus pneumoniae (229).
On chest radiography, the diagnosis of septic emboli should be considered
when multifocal, peripheral, ill-dened, round or wedge-shaped opacities are
seen in the setting of a febrile patient (233). The opacities have a basilar predilec-
tion reecting greater blood ow. They may be uniform or variable in size,
reecting recurrent showers of emboli. Cavitation is frequent and early. The cav-
ities are usually thin-walled and occasionally contain a central loose body, the
target sign (234), which may simulate a fungus ball (234,235). The central
loose body represents a piece of necrotic lung. Air uid levels are conspicuously
absent in many of the cavities. Invasive aspergillosis is the major radiologic
differential diagnosis in the immunocompromised patient (233).
382 Barack et al.
CT demonstrates abnormalities before radiography in 30% of cases and

more extensive disease in 44% of cases (236). On CT examination, subpleural
peripheral nodules and wedge-shaped peripheral lesions less than 3 cm in size
are the most common nding (237). Cavitation is a common feature and is
seen in varying degrees in 50% of the lesions imaged. A feeding vessel
leading into the parenchymal opacity may be identied, suggesting a hemato-
genous origin (236).
F. Pulmonary Hemorrhage
Pulmonary hemorrhage can result from trauma, bleeding diatheses, infections,
pulmonary embolism, penicillamine therapy, and autoimmune diseases.
The most common autoimmune disease causing pulmonary hemorrhage is Good-
pastures syndrome (238).
Radiographically, pulmonary hemorrhage manifests as coalescent airspace
opacication. Unless there is recurrent bleeding, the airspace opacication clears
rapidly within two to three days. The chest radiograph ndings are nonspecic
and indistinguishable from pulmonary edema and pneumonia, and denitive
diagnosis is based on clinical, laboratory, and bronchoscopic ndings, if
necessary.
G. Dissecting Hematoma of the Aorta
Dissecting hematoma of the aorta results from an intimal tear followed by

formation of an intramural hematoma and the formation of a false lumen.
Aortic dissection has a male predominance. Predisposing factors include
hypertension, connective tissue disease, congenital valve abnormalities, and
coarctation (239). Rapid diagnosis of dissecting hematoma is paramount
because mortality in untreated patients is 50% in the rst 48 hours (240).
Therefore, suspected acute aortic dissection is one of the indications for emergent
radiographic evaluation.
MDCT, if available, is the imaging modality of choice when renal
function is normal. In those patients with compromised renal function, MRI is
the imaging modality of choice. The sensitivity and specicity of CT and MRI
for diagnosing dissecting hematoma are similar (241,242). The radiographic nd-
ings of dissecting hematoma on MRI and CT are an intimal ap and a false lumen
(Fig. 27). Aortography has long been considered to be the gold standard in
evaluation of aortic dissection (243). Aortography has essentially been replaced
by CT and MRI in the evaluation of dissecting hematoma when these imaging
modalities are available.
Two important entities that should be distinguished from dissecting hema-
toma are aortic intramural hematoma and penetrating atherosclerotic ulcer.
Aortic intramural hematoma represents blood within the aortic wall without
intimal disruption. Aortic intramural hematoma occurs primarily in elderly
hypertensive patients. Recognition is important because it may progress to
Figure 27 Stanford A aortic dissection (dissecting hematoma). Sagittal reconstruction

(mediastinal window) demonstrates Stanford A dissecting hematoma with intimal tear
extending into the proximal left subclavian artery (white arrow) and into the descending
thoracic aorta (white arrowheads). Abbreviations: A, aortic root; RV, right ventricle;
PA, main pulmonary artery; LA, left atrium; T, true lumen; F, false lumen.
dissecting hematoma. Diagnosis is made by non-contrast CT examination where

it appears as a hyperdense, crescentic collection located eccentrically within the
aortic wall, with or without signicant narrowing of the lumen (244).
Penetrating atherosclerotic ulcer of the aorta is a lesion caused by ulcera-
tion of an atheromatous plaque with disruption of the internal elastic lamina. It
is found in elderly, hypertensive patients. The most common sites of penetrating
atherosclerotic ulcer are the mid to distal descending thoracic aorta and the upper
abdominal aorta. The ulcer may remain stable or progress to form an aortic pseu-
doaneurysm in 25% to 35% of cases (245,246) or may rupture. CT and MRI
demonstrate a localized outpouching of the aortic wall corresponding to the
ulcer and thickening of the wall secondary to subintimal bleeding (247,248).
H. Traumatic Abnormalities
Aortic Injury
Rupture of the thoracic aorta is a common cause of death following blunt chest
trauma. Death occurs at the accident site in more than 80% of aortic ruptures.
Patients (60 70%) who reach the hospital survive if prompt treatment is insti-
tuted. Ninety percent of aortic injuries occur at the level of the ligamentum arter-
iosum. Less common sites of involvement are the aortic root and the
diaphragmatic hiatus (249).
384 Barack et al.
Types of aortic injury seen in patients who survive include dissecting

hematoma, pseudoaneurysm, and partially contained rupture. The radiographic
ndings in aortic injury are due to the presence of mediastinal hematoma.
The radiographic ndings of dissecting hematoma have been previously
described. The ndings of a pseudoaneurysm at the level of the ligamentum arter-
iosum on the frontal radiograph are a smooth convex lateral soft tissue opacity
silhouetting the distal aortic arch and proximal descending aorta (Fig. 28). A
negative frontal chest radiograph has a 98% negative predictive value in exclud-
ing aortic injury (250). On non-contrast CT, an acute pseudoaneurysm appears as
a homogeneous, hyperdense (.40 HU) collection within the wall of the aorta,
which does not enhance following the administration of intravenous contrast.
On MR, this collection follows intensity characteristics of acute blood. On
occasion, the distal extension of the pseudoaneurysm may result in marked
narrowing of the aortic lumen, which is termed pseudocoarctation. In a partially
contained rupture, there is active extravasation of contrast.
Esophageal Rupture
Esophageal rupture can be secondary to Boerhaaves syndrome, iatrogenic
causes, or blunt trauma. Boerhaaves syndrome is spontaneous perforation
of the thoracic esophagus due to sudden increase in intraluminal pressure.
Figure 28 Traumatic pseudoaneurysm of the aortic arch. Postero-anterior radiograph in

a patient two hours after blunt chest trauma demonstrates smooth convex lateral soft tissue
opacity silhouetting the distal aortic arch and proximal descending aorta (white and black
arrows). A traumatic pseudoaneurysm distal to the origin of left subclavian artery was
demonstrated at computed tomography aortography.
Iatrogenic esophageal perforation can occur during lung and cardiovascular

surgery, placement of an intraluminal tube for esophageal variceal bleeding, eso-
phageal dilatation, and endoscopy. Traumatic esophageal injuries are rare. When
present, they occur in the cervical and upper thoracic esophagus or just above the
gastroesophageal junction. The mortality rate for esophageal tears complicated by
mediastinitis is 38% (251). The prognosis is markedly improved if the injury is
identied within 24 hours of occurrence (252). Therefore, in suspected cases of
esophageal perforation, diagnostic imaging should be performed promptly.
An esophageal perforation may rst be evaluated with a portable
esophagram as in the section on Oral Contrast. If clinically indicated, CT with
oral and intravenous contrast may be performed. Extravasation of contrast from
the lumen of the esophagus is diagnostic of perforation. However, this may not
be seen. Other suggestive ndings are extraluminal periesophageal air, pleural
effusion, and periesophageal inammation with wall thickening (252,253).
Other
Any sequellae of trauma can result in admission to the ICU if severe enough to
require mechanical ventilation and close hemodynamic monitoring. Such
conditions include severe mediastinal and parenchymal hematoma, cardiac
contusion, osseous injuries (Fig. 29), and tracheobronchial rupture. The latter
should be suspected in a trauma patient following chest tube placement for a
pneumothorax with a persistent air leak.
V. Computerized Tomography
A. Indications
CT is the imaging modality of choice for the evaluation of pulmonary embolism,

acute aortic syndromes (dissecting hematoma, intramural hematoma, and pene-
trating aortic ulcer), and traumatic aortic injury. Other clinical situations where
CT may be indicated include: (1) persistent pulmonary opacities unresponsive
to treatment, (2) further denition of pneumothorax or pleural effusion for inter-
ventional treatment, (3) prior to bronchoscopy or biopsy, (4) differentiating lung
abscess from empyema for interventional treatment, (5) assessment of patency of
bypass grafts and dening anatomy prior to further thoracic surgery, (6) sus-
pected esophageal rupture, (7) differentiating unilateral bullous disease or
emphysema from pneumothorax, and (8) suspected bronchopleural stula.
B. Technique
The VA Greater Los Angeles Health Care System Hospital at West Los Angeles
performs all thoracic imaging on a 16-row multidetector Toshiba scanner.
Routinely, axial images are acquired with 1-mm collimation. The 1-mm axial
386 Barack et al.
Figure 29 Traumatic left apical pneumothorax. Postero-anterior radiograph demon-

strates fracture of the left clavicle (black arrowhead), fracture of the left third rib
(arrow), and traumatic left apical pneumothorax (white arrowheads) in an elderly
patient post-fall.
images with 3-mm axial reformations and 1.5-mm coronal and sagittal reforma-
tions are then sent to PACS for interpretation.
Nonionic contrast is administered when indicated and when renal function
permits (creatinine within 0 45 days of exam 1.5). When serum creatinine is
between 1.5 and 2.0 and in patients susceptible to developing nephrotoxicity,
contrast is administered depending upon the clinical situation and only after
close clinical consultation. Such patients are adequately hydrated and given
oral N-acetylcysteine 600 mg twice daily the day before and the day of the
exam. Isoosmolar nonionic contrast is used exclusively in this subset of patients,
and the volume of contrast administered is carefully tailored to the clinical
situation. The creatinine is closely monitored after the examination. Intravenous
contrast is not administered to patients whose serum creatinine is 2.0 unless
the patient is on dialysis. In these patients, contrast is administered on the day
of dialysis prior to the procedure. If the patient is on glucophage, the drug is
withheld on the day of the examination and for two days thereafter.
Radiation exposure is always a concern in these patients. Assuming a 2-mm
helical mode continuous axial acquisition is obtained on an average patient over a
30-cm length (lung apex to lung base) with a 16-row multidetector CT using the
following parameters (120 kVp, 360 mA, 0.5 sec gantry rotation, 1.25 effective
pitch, 16:1 acquisition, and 40% automatic dose reduction), the effective dose
equivalent given to the patient roughly corresponds to an effective dose of 120
PA chest radiographs obtained in a 70-kg patient using a technique of 110 kVp
and 3 mAs. It should be noted that variation among CT scanners, scanning
parameters, and actual patient size may result in considerable variation in this
number (Christopher H. Cagnon, Ph.D., Radiology Physicist, Department of
Radiological Science, David Geffen School of Medicine at U.C.L.A.).
C. Problems Unique to the ICU Patient
CT scanning of the ICU patient presents unique problems, including an inability

to suspend respiration (produces motion artifact), renal disease limiting contrast
utilization, and the risks associated with transporting the mechanically ventilated
patient. The newer scanners have partially overcome the problem of motion
artifact. With a 16-row MDCT scanner and a pitch of 20, a 1-mm contiguous
axial acquisition can be obtained on a 30-cm average thorax (apex to base) in
approximately 9.5 seconds. This time can be reduced to approximately six
seconds with a 2-mm axial acquisition.
The problems associated with transporting ICU patients remain unaddressed.
When considering a CT examination on the ICU patient, one must weigh the poten-
tial of identifying a potentially treatable condition against the risks associated with
transferring a critically ill patient. These problems are partially reduced when
CT and MRI scanners are placed in closer proximity to the ICU. Future hospital
design may include a dedicated radiographic suite adjacent to the ICU.
VI. Image-Guided Interventional Procedures in the ICU

A. Venous Access
Central venous catheters are important in critically ill patients for purposes of
pressure monitoring, blood sampling, and administration of medications and
blood products. It has become increasingly apparent that both success and safety
are enhanced by central venous catheter placement under image guidance.
Venous Evaluation
Before catheter placement, it is important to delineate the anatomy and patency
of the venous structures involved. Venography used to be the gold standard to
determine the patency of a venous segment. As ultrasound and MR techniques
become more rened, they largely replaced the need for venography, which
can cause patient discomfort because of needle puncture and contrast adminis-
tration. MRA (most frequently using contrast-enhanced and time of ight
techniques) can reliably delineate the central venous anatomy. Ultrasound com-
plemented by spectral and color Doppler analysis allows convenient assessment
of the venous system. Although ultrasound cannot provide direct visualization of
388 Barack et al.
the central veins [i.e., central subclavian and innominate veins], patency can be
inferred by evaluation of the spectral waveform for the presence of respiratory
phasicity and cardiac periodicity.
Central Line Placement
Central line placement performed under ultrasound and uoroscopy guidance

offers signicant advantages over blind placement because the chance for
successful placement is increased. Venous systems that are inaccessible to
blind puncture due to anatomic variation or obesity can be easily accessed, and
the safety prole is increased with decreased incidence of pneumothorax or arter-
ial puncture.
For ultrasound-guided procedures, high-frequency linear array transducers
are optimal (i.e., 7.5 or 5 MHz). The target vessel should be evaluated carefully
in transverse and longitudinal planes for its suitability for catheterization and its
proximity to structures to be avoided. Once the vessel is accessed, uoroscopy is
generally used for proper catheter position. A steerable wire or contrast injection
may be required.
Peripheral Venous Access: PICCs and Arm Ports

PICC line placement has become a preferred alternative to central line placement
because of favorable long-term complication rates. In a series of 404 PICC
placements, Cardella et al. (254) reported a 2% infection rate and 1% throm-
bophlebitis rate. In patients with a paucity of obvious peripheral veins, image gui-
dance is often required for line placement. The procedure is typically performed
using ultrasound or venography guidance to obtain access. Micropuncture
needles and .018-inch wires should be utilized for initial punctures as these
smaller systems are less traumatic and result in the least amount of venospasm.
Tourniquets can be used to distend upper arm veins, and nitroglycerin may be
injected through a peripheral intravenous line in aliquots of 100 mcg to help
relieve spasm.
The use of a hydrophilic vascular wire allows greater torque control to
negotiate through venous varients and collaterals, markedly enhancing the
success rate for PICC line placement in the angiography suite. In the study by
Cardella et al. (254), a 98.7% success rate with 100% optimal placement was
demonstrated in the angiography suite compared with 74% and 92.8%, respect-
ively, with bedside placement.
Venous arm ports are smaller single-lumen versions of products that have
been available for placement on the chest wall. Port placement requires strict
adherence to sterile technique, although prophylactic antibiotics are not necess-
ary. The catheter is placed in a similar fashion as described with a PICC. A skin
incision is made, and a small subcutaneous pocket is created using blunt dissec-
tion for port placement. Finally, the port and the infusion catheter are attached
securely together.
Alternative Central Vein Access

When conventional means of access is not possible for central catheter place-
ment, interventional radiologists and clinicians should be aware of alternative
access. This usually occurs in patients who have had multiple prior catheteriza-
tions. Translumbar and transhepatic approaches are the two main alternative
options. After conrming normal anatomy, the translumbar approach can
usually be performed with uoroscopy guidance, although CT may be required
for initial access. Transhepatic access can be performed under ultrasound
guidance. The middle hepatic vein is the largest and the primary choice.
Other possible venous access includes hypertrophied collateral veins,
azygous and hemiazygous veins, and intercostal veins. Furthermore, access can
be achieved in occluded veins after they are recanalized with wire, angioplasty,
and stenting techniques.
Malpositioned Catheters
Interventional radiologists are frequently consulted to salvage a malpositioned

central line. For non-tunneled catheters, repositioning of the catheters is easily
done over a guidewire. However, when a cuffed or implanted catheter tip is mal-
positioned, various snares or catheter and wire techniques allow manipulation
and repositioning of the catheters. This is commonly done via a femoral
approach. Pigtail catheters, tip deecting wires, or Amplatz snares have been
used with great success to pull catheters back into position.
Occluded Catheters
Catheter occlusions secondary to brin sheath formation or improper ushing
resulting in intraluminal clots are often treated with instillation of thrombolytic
agents (i.e., TPA or urokinase). Fibrin sheaths and clots at the tips of the catheters
can be treated by disruption with a curled heavy-duty wire, or even with a large
angioplasty balloon. However, resistant pericatheter brin sheaths require a
different approach. A non-cuffed catheter can be easily replaced over a guide-
wire. For cuffed or implanted catheters, exchange over a wire is usually not
feasible. A technique of catheter brin sheath stripping has been described by
Mewissen et al. (255) with an Amplatz snare from the femoral venous approach
for failing hemodialysis catheters.
Fibrin sheath typically acts as a ball valve, allowing infusion but occluding
side holes upon negative pressure application, therefore hindering catheter
aspiration. Its diagnosis is conrmed by venography. Once the brin sheath is
demonstrated, a guidewire is passed through the distal port of the central line
and manipulated through an open Amplatz snare (15, 25, or 35 mm) deployed
from the femoral approach. The loop is then guided around the catheter and
rmly snugged against it. As the snare is withdrawn, the brin sheath is
removed. Thirty-day patency rates are on the order of 30% to 50%. Repeated
stripping is necessary in maintaining long-term patency.
390 Barack et al.
B. Percutaneous Catheter Drainage

Interventional radiological techniques are now routinely incorporated into the
management of critically ill patients with thoracic uid collections. The currently
available radiologically guided procedures to diagnose and treat chest diseases
are discussed in the following paragraphs, with emphasis on indications,
techniques, potential complications, and post-procedural patient management.
Radiologically Guided Thoracentesis

Large pleural effusions may be safely aspirated at the bedside without imaging
guidance. However, guided thoracentesis is necessary in cases when the
volume of the uid is small, the effusion is loculated, and the patient has severely
compromised respiratory status or is dependent on respirator. Guidance improves
the success rate for obtaining uid and minimizes the risks of pneumothorax and
hepatic or splenic injury. Ultrasound is the most versatile modality to guide initial
access. It allows characterization of uid by accessing the echogenicity of the
uid and the degree of septations. Its portability enables procedures to be done
at the bedside. If a collection is severely loculated or has a particularly difcult
access route, CT may be utilized.
Pleural Collections
The common etiologies for pleural collections that require drainage include com-
plicated parapneumonic effusions and post-traumatic or post-surgical collections.
It is important to identify the collections that would benet from drainage early in
the course, as delays in drainage allow the collection to organize and ultimately
diminish the speed and efcacy of percutaneous drainage.
Three stages have been described in the natural history of parapneumonic
effusions. Stage I (exudative) is characterized by a sterile exudative effusion with
a pH .7.30, glucose .60 mg/dL, and LDH ,500 U/L. In stage II (brinopuru-
lent), bacteria enter the pleural space, polymorphonuclear cells increase, and
glucose decreases to ,40 mg/dL, pH ,7.10, and LDH . 1000 U/L. As uid
protein increases, intrapleural brinolytic activity diminishes, and coagulation
of the uid can occur. Fibrin deposition and broblastic activity are initiated.
Stage III (organizational) is characterized by formation of a thick, brotic
pleural peel (256,257).
Radiologically placed tubes (8 14 French) are relatively small compared
with the traditional, surgically placed tubes (24 38 French). Percutaneous drai-
nage catheters can be inserted under ultrasound and uoroscopy guidance most
of the time. However, if the uid collection is loculated, poorly accessible
(e.g., overlying scapula, brachial plexus, or subclavian vessels) or close to vital
structures, CT guidance may be necessary for more precise localization. In
severely complex collections, multiple drainage catheters may be necessary.
The combination of a glide catheter and wires can be used to disturb septations
in a complex collection. Thrombolytics have been utilized to break up loculations

and to increase the uidity of the collection. Various protocols with different
agents (urokinase, TPA, streptokinase) have been reported, demonstrating an
overall promising result. Gentle irrigation of the catheters with normal saline
may be performed every eight hours to maintain tube patency. Catheter output
should be monitored closely. If the output decreases, and there is signicant
residual uid, catheter reposition or upsizing may be necessary. Guided catheter
drainage of pleural collection has a success rate in the 70% range with a low
incidence of complications (258 261).
Lung Abscess
Lung abscesses are usually treated conservatively with antibiotics, postural drai-
nage, and occasionally bronchoscopy to clear bronchial obstruction. In patients
failing conservative treatment, catheter drainage has been shown to be effective
and safe (262 265). CT is the modality of choice for initial needle access. The
patient should be placed with the normal lung in a non-dependent position to
avoid spillage of purulent material into the normal lung. Such a complication
can lead to overwhelming sepsis and death. An 8 to 14 French catheter with a
retention lock is commonly used. The catheter is ushed two to three times a
day and placed on water seal for suction. The catheter can be removed when
clinical parameters improve, drainage volume decreases, and the radiographic
abnormality resolves.
Mediastinal and Pericardial Abscess
Acute mediastinitis complicated by abscess formation can result from esophageal

perforation (penetrating or blunt trauma, instrumentation, anastomotic leak) or
from infections of the neck, spine, lung, or pleura. Median sternotomy, usually
performed for cardiac surgery, is the most common cause of mediastinitis
(255). Percutaneous drainage may be curative or may temporize and improve
the patients condition, allowing denite repair at a later date. CT guidance is
necessary on account of the proximity of the heart and great vessels. The
catheters are cared for in the same manner as a pleural chest catheter with
Pleuro-Vac suction.
Pericardial abscess or other uid collections can be effectively treated
by guided aspiration or percutaneous catheter drainage. Ultrasound, and less
commonly CT, is utilized. Both aspiration and drainage have been shown to
be safer when performed with guidance than when performed blindly (266).
Both subxyphoid and intercostal approaches are utilized.
C. Management of Pulmonary Embolism and

Deep Vein Thrombosis
Acute pulmonary embolism represents the most severe end of the spectrum of
the venous thromboembolic disease. The majority of the patients with
392 Barack et al.
thromboembolic disease are successfully treated with anticoagulation. However,

in those patients in whom anticoagulation is not feasible, fails, or results in com-
plications, placement of an IVC lter is required. IVC lters have been placed
prophylactically in patients who have not had, but are at increased risks for the
development of, a thromboembolic event. However, this practice remains
controversial.
All lters currently available in the United States can be inserted safely
and quickly using percutaneous methods, with an approximate recurrent PE
rate of 4% to 5% (267). IVC lters are placed via the femoral or jugular approach.
An IVC venogram is necessary to evaluate the size of the cava and to assess for
the level of renal veins and for possible venous variants. Ideally, lters should be
positioned below the level of the renal veins to avoid thrombus propagation
compromising renal function. Overall, IVC lters are safe and effective devices.
For patients who are hemodynamically unstable as a result of an acute
massive pulmonary embolism, an emergent pulmonary angiogram, and possibly
thrombolytic therapy, should be considered. Catheter-directed thrombolysis
technique is performed with multisidehole infusion catheters to deliver the
thrombolytic agent directly into the thrombus. More recently, endovascular
thromboembolectomy has been promoted with mechanical devices.
D. Conclusion
Radiology and interventional radiology have assumed a growing role in the

diagnosis and treatment of the critically ill patient. More sophisticated imaging
techniques have improved diagnostic capability. Accurate imaging is integral
for optimal patient management, and the most favorable clinical outcomes are
achieved through multidisciplinary efforts.
References
1. Marik PE, Janower ML. Impact of routine chest radiography on ICU management
decisions: an observational study. Am J Crit Care 1997; 6(2):95 98.
2. Strain DS, KinaseWitz GT, Vereen LE, George, RB. Value of routine daily chest
x-rays in the medical intensive care unit. Crit Care Med 1985; 13(7):534 536.
3. Bekemeyer WB, Crapo RO, Calhoon S, Cannon CY, Clayton PD. Efcacy of chest
radiography in a respiratory intensive care unit. A prospective study. Chest 1985;
88(5):691 696.
4. Hall JB, White SR, Karrison T. Efcacy of daily routine chest radiographs in intu-
bated, mechanically ventilated patients. Crit Care Med 1991; 19(5):689 693.
5. Henshke CI, Pasternack GS, Schroeder S, Hart KK, Herman PG. Bedside chest
radiography: diagnostic efcacy. Radiology 1983; 149(1):23 26.
6. Silverstein DS, Levingston DH, Elcavage J, Kovar L, Kelly KM. The utility of
routine daily chest radiography in the surgical intensive care unit. J Trauma
1993; 35(4):643 646.
7. Fong Y, Whalen G, Hariri RJ, Barie PS. Utility of routine chest radiographs in the
surgical intensive care unit. Arch Surg 1995; 130(7):764 768.
8. American College of Radiology ACR Appropriateness Criteria. Expert Panel on
Thoracic Imaging. Daily portable chest X-ray. Radiology 2000; 215(S):621 626.
9. Wandtke J. Bedside chest radiography. Radiology 1994; 190(1):1 10.
10. Milne EN. A physiological approach to reading critical care unit lms. J Thorac
Imaging 1986; 1(3):60 90.
11. Tocino I. Chest imaging in the intensive care unit. Eur J Radiol 1996; 23(1):
46 57.
12. Henschke CI, Yankelevitz DF, Wand A, Davis SD, Shiau M. Chest radiography in
the ICU. Clin Imaging 1997; 21(2):90 103.
13. Herman MW, Patrick J, Tabrisky J. A comparative study of scattered radiation
levels from 80-kVp and 240-kVp X rays in the surgical intensive care unit.
Radiology 1980; 137(2):552 553.
14. Boles JM, Boussert F, Manens JP, Le Cam B, Bellet M, Garre M. Measurement of
irradiation doses secondary to bedside radiographs in a medical intensive care unit.
Intensive Care Med 1987; 13(1):60 63.
15. Brandstetter RD, Garcia JC, Majumder M, Chisolm A. The benet of lateral
radiographs in an intensive care unit. Chest 1994; 105(2):560 562.
16. Schaefer-Prokop C, Uffmann M, Eisenhuber E, Prokop M. Digital radiography of
the chest: detector techniques and performance parameters. J Thorac Imaging
2003; 18(3):124 137.
17. Willis CE, Thompson SK, Sheppard J. Artifacts and misadventures in digital-
radiography. Applied Radiology 2004; 33(1):11 20.
18. Willis CE, Mercier J, Patel M. Modication of conventional quality assurance
procedures to accomodate computer radiography. Proceedings of the 13th Inter-
national Conference on Computer Applications in Radiology, Denver, Colarado,
June 7, 1996: 275 281.
19. Sagel SS, Jost RG, Glazer HS, Molina PL, Anderson DJ, Solomon SL, Schwarberg J.
Digital mobile radiography. J Thorac Imaging 1990; 5(1):3648.
20. Gamsu G, Perez E. Picture archiving and communication systems (PACS). J Thorac
Imaging 2003; 18(3):165 168.
21. van der Jagt EJ, Smits HJ. Cardiac size in the supine chestlm. Eur J Radiol 1992;
14(3):173 177.
22. Milne EN, Pistolesi M, Miniati M, Giuntini C. The vascular pedicle of the heart and
the vena azygos. Part I: The normal subject. Radiology 1984; 152(1):1 8.
23. Aronchick JM, Miller WT Jr. Tubes and lines in the intensive care setting. Semin
Roentgenol 1997; 32(2):102 116.
24. Goodman LR, Conrardy PA, Laing F, Singer MN. Radiographic evaluation of
endotracheal tube position. Am J Roentgenol 1976; 127(3):433 434.
25. Yap SJ, Morris RW, Pybus DA. Alterations in endotracheal tube position during
general anesthesia. Anaesth Intensive Care 1994; 22(5):586 588.
26. Conrardy PA, Goodman LR, Lainge F, Singer MM. Alteration of endotracheal tube
position. Flexion and extension of the neck. Crit Care Med 1976; 4(1):7 12.
27. Valentino J, Myers RK, Baker MD, Woodring JH. Utility of portable chest
radiographs as a predictor of endotracheal tube cuff pressure. Otolaryngol Head
Neck Surg 1999; 120(1):51 56.
394 Barack et al.
28. Khan F, Reddy NC, Khan A. Cuff/trachea ratio as an indicator of tracheal damage
[abstr]. Chest 1976; 70:431.
29. Stauffer JL, Olsen DE, Petty, TL. Complications and consequences of endotracheal
intubation and tracheotomy. A prospective study of 150 critically ill adult patients.
Am J Med 1981; 70(1):65 76.
30. Zwillich CW, Pierson DJ, Creagh CE, Sutton FD, Schatz E, Petty TL. Compli-
cations of assisted ventilation. A prospective study of 354 consecutive episodes.
Am J Med 1974; 57(2):161 170.
31. Lotano R, Gerber D, Aseron C, Santarelli R, Pratter M. Utility of postintubation
chest radiographs in the intensive care unit. Crit Care 2000; 4(1):50 53.
32. Brunel W, Coleman DL, Schwartz DE, Peper E, Cohen NH. Assessment of routine
chest roentgenograms and the physical examination to conrm endotracheal tube
position. Chest 1989; 96(5):1043 1045.
33. Schwartz DE, Lieberman JA, Cohen NH. Women are at greater risk than men for
malpositioning of the endotracheal tube after emergent intubation. Crit Care Med
1994; 22(7):1127 1131.
34. Smith GM, Reed JC, Choplin RH. Radiographic detection of esophageal malposi-
tioning of endotracheal tubes. Am J Roentgenol 1990; 154(1):23 26.
35. Rollins RJ, Tocino I. Early radiographic signs of tracheal rupture. Am J Roentgenol
1987; 148(4):695 698.
36. Wechsler RJ, Steiner RM, Kinori I. Monitoring the monitors: the radiology of
thoracic catheters, wires, and tubes. Semin Roentgenol 1988; 23(1):61 84.
37. Leeds WM, Morley TF, Zappasodi SJ, Giudice JC. Computed tomo-
graphy for diagnosis of tracheoesophageal stula. Crit Care Med 1986; 14(6):
591 592.
38. Goodman LR. Pulmonary support and monitoring apparatus. In: Goodman LR,
Putman CE, eds. Critical Care Imaging. Philadelphia: Saunders, 1992:35 39.
39. Pasternack G, OCain C. Thoracic complications of respiratory intensive care.
In: Herman P, ed. Iatrogenic Thoracic Complication. New York: Springer,
1983:59 77.
40. Heffner JE, Miller KS, Sahn SA. Tracheostomy in the intensive care unit. Part 1:
Indications, technique, management. Chest 1986; 90(2):269 274.
41. Heffner JE, Miller KS, Sahn SA. Tracheostomy in the intensive care unit. Part 2:
Complications. Chest 1986; 90(3):430 436.
42. Wechsler RJ, Steiner RM, Goodman LR, Teplick SK, Mapp E, Laufer I. Iatrogenic
esophageal-pleural stula: subtlety of diagnosis in the absence of mediastinitis.
Radiology 1982; 144(2):239 243.
43. Ghahremani GG, Gould RJ. Nasoenteric feeding tubes. Radiographic detection of
complications. Dig Dis Sci 1986; 31(6):574 585.
44. Maffessanti M, Berlot G, Bortolotto P. Chest roentgenology in the intensive care
unit: an overview. Eur Radiol 1998; 8(1):69 78.
45. Rassias AJ, Ball PA, Corwin HL. A prospective study of tracheopulmonary compli-
cations associated with the placement of narrow-bore enteral feeding tubes.
Crit Care Epub 1998; 2(1):25 28.
46. Miller KS, Tomlinson JR, Sahn SA. Pleuropulmonary complications of enteral tube
feedings. Two reports, review of the literature, and recommendations. Chest 1985;
88(2):230 233.
47. Aronchick JM, Epstein DM, Gefter WB, Miller WT. Pneumothorax as
a complication of placement of a nasoenteric tube. J Am Med Assoc 1984;
252(23):3287 3288.
48. Balogh GJ, Adler SJ, VanderWoude J, Glazer HS, Roper C, Weyman PJ. Pneu-
mothorax as a complication of feeding tube placement. Am J Roentgenol 1983;
141(6):1275 1277.
49. Stark P, Thordarson S, McKinney M. Manifestations of esophageal disease on plain
chest radiographs. Am J Roentgenol 1990; 155(4):729 734.
50. Teplick SK, Brandon JC, Shah SH. Acute gastrointestinal disorders. Goodman LR,
Putnam CE, eds. Critical Care Imaging. 3rd ed. Philadelphia: WB Saunders,
1992:314 318.
51. Wand A, Davis S, Yankelevitz DF. Esophageal perforation caused by Sengstaken
Blakemore tube. Complications Surg 1995; 14:79 80.
52. Zarshenas Z, Sparschu RA. Catheter placement and misplacement. Crit Care Clin
1994; 10(2):417 436.
53. Swensen SJ, Peters SG, LeRoy AJ, Gay PC, Sykes MW, Trastek VF. Radiology in
the intensive-care unit. Mayo Clin Proc 1991; 66(4):396 410.
54. Webb WR, Godwin JD. The obscured outer edge: a sign of improperly placed
pleural drainage tubes. Am J Roentgenol 1980; 134(5):1062 1064.
55. Maurer JR, Friedman PJ, Wing VW. Thoracostomy tube in an interlobar ssure:
radiologic recognition of a potential problem. Am J Roentgenol 1982;
139(6):1155 1161.
56. Miller KS, Sahn SA. Chest tubes. Indications, technique, management and compli-
cations. Chest 1987; 91(2):258 264.
57. Stahly TL, Tench WD. Lung entrapment and infarction by chest tube suction.
Radiology 1977; 122(2):307 309.
58. Wagner DP, Knaus WA, Draper EA. Identication of low-risk monitor admissions
to medical-surgical ICUs. Chest 1987; 92(3):423 428.
59. Henschke CI, Yankelevitz DF, Wand A, Davis SD, Shiau M. Accuracy and efcacy
of chest radiography in the intensive care unit. Radiol Clin North Am 1996;
34(1):21 31.
60. Langston CS. The aberrant central venous catheter and its complications.
Radiology 1971; 100(1):55 59.
61. Conces DJ Jr, Holden RW. Aberrant locations and complications in initial place-
ment of subclavian vein catheters. Arch Surg 1984; 119(3):293 295.
62. McGee WT, Ackerman BL, Rouben LR, Prasad VM, Bandi V, Mallory DL. Accu-
rate placement of central venous catheters: a prospective, randomized, multicenter
trial. Crit Care Med 1993; 21(8):1118 1123.
63. Gilner LI. The ear-gurgling sign. N Engl J Med 1977; 296(22):1301.
64. Langston CS. The aberrant central venous catheter and its complications.
Radiology 1971; 100(1):55 59.
65. Lumb PD. Complications of central venous catheters. Crit Care Med 1993;
21(8):1105 1106.
66. Tocino IM, Watanabe A. Impending catheter perforation of superior vena cava:
radiographic recognition. Am J Roentgenol 1986; 146(3):487 490.
67. Campbell M, Deurachar DC. The left-sided superior vena cava. Br Heart J
1954;16(4):423 439.
396 Barack et al.
68. Cha EM, Khoury GH. Persistent left superior vena cava. Radiologic and clinical
signicance. Radiology 1972; 103(2):375 381.
69. Gibson RN, Hennessy OF, Collier N, Hemingway AP. Major complications of
central venous catheterisation: a report of ve cases and a brief review of the litera-
ture. Clin Radiol 1985; 36(2):205 208.
70. Plaus WJ. Delayed pneumothorax after subclavian vein catheterization. J Parenter
Enteral Nutr 1990; 14(4):414 415.
71. Sivak SL. Late appearance of pneumothorax after subclavian venipuncture. Am J
Med 1986; 80(2):323 324.
72. Silverman SM, Olson KW. Avoiding unintentional arterial cannulation. Can J
Anaesth 1989; 36(6):734.
73. Wechsler RJ, Spirn PW, Conant EF, Steiner RM, Needleman L. Thrombosis and
infection caused by thoracic venous catheters: pathogenesis and imaging ndings.
Am J Roentgenol 1993; 160(3):467 471.
74. Alhimyary A, Fernandez C, Picard M, Tierno K, Pignatone N, Chan HS, Malt R,
Souba W. Safety and efcacy of total parenteral nutrition delivered via a peripher-
ally inserted central venous catheter. Nutr Clin Pract 1996; 11(5):199 203.
75. Cowl CT, Weinstock JV, Al-Jurf A, Ephgrave K, Murray JA, Dillon K. Compli-
cations and cost associated with parenteral nutrition delivered to hospitalized
patients through either subclavian or peripherally-inserted central catheters. Clin
Nutr 2000; 19(4):237 243.
76. Hadaway LC. An overview of vascular access devices inserted via the antecubital
area. J Intraven Nurs 1990; 13(5):297 306. Erratum in: J Intraven Nurs 1991;
14(2):135.
77. James L, Bledsoe L, Hadaway LC. A retrospective look at tip location and compli-
cations of peripherally inserted central catheter lines. J Intraven Nurs 1993;
16(2):104 109.
78. Cardella JF, Cardella K, Bacci N, Fox PS, Post JH. Cumulative experience with
1,273 peripherally inserted central catheters at a single institution. J Vasc Interv
Radiol 1996; 7(1):5 13.
79. Barber JM, Booth DM, King JA, Chakraverty S. A nurse led peripherally inserted
central catheter line insertion service is effective with radiological support.
Clin Radiol 2002; 57(5):352 354.
ing in the intensive care unit (Part 1). Chest 1984; 85(4):537 549.
ing in the intensive care unit (Part 2). Chest 1984; 85(5):656 668.
82. Sise MJ, Hollingsworth P, Brimm JE, Peters RM, Virgilio RW, Shackford SR.
Complications of the ow-directed pulmonary artery catheter: A prospective
analysis in 219 patients. Crit Care Med 1981; 9(4):315 318.
83. Oliver MJ. Acute dialysis catheters. Semin Dial 2001; 14(6):432 435.
84. Work J. Hemodialysis catheters and ports. Semin Nephrol 2002; 22(3):211 220.
85. Randolph AG, Cook DJ, Gonzales CA, Pribble CG. Ultrasound guidance for place-
ment of central venous catheters: a meta-analysis of the literature. Crit Care Med
1996; 24(12):2053 2058.
86. Forauer AR, Glockner JF. Importance of US ndings in access planning during
jugular vein hemodialysis catheter placements. J Vasc Interv Radiol 2000; 11(2
Pt 1):233 238.
87. Lin BS, Kong CW, Tarng DC, Huang TP, Tang GJ. Anatomical variation of the
internal jugular vein and its impact on temporary hemodialysis vascular access:
an ultrasonographic survey in uraemic patients. Nephrol Dial Transplant 1998;
13(1):134 138.
88. Hernandez D, Diaz F, Rufno M, Lorenzo V, Perez T, Rodriguez A, De Bonis E,
Losada M, Gonzalez-Posada JM, Torres A. Subclavian vascular access stenosis in
dialysis patients: natural history and risk factors. J Am Soc Nephrol 1998;
9(8):1507 1510.
89. Agraharkar M, Isaacson S, Mendelssohn D, Muralidharan J, Mustata S, Zevallos G,
Besley M, Uldall R. Percutaneously inserted silastic jugular hemodialysis catheters
seldom cause jugular vein thrombosis. ASAIO J 1995; 41(2):169 172.
90. Trerotola SO, Kuhn-Fulton J, Johnson MS, Shah H, Ambrosius WT, Kneebone PH.
Tunneled infusion catheters: increased incidence of symptomatic venous thrombo-
sis after subclavian versus internal jugular venous access. Radiology 2000;
217(1):89 93.
91. Vanherweghem JL, Cabolet P, Dhaene M, Goldman M, Stolear JC, Sabot JP,
Waterlot Y, Marchal M. Complications related to subclavian catheters for hemodia-
lysis. Report and review. Am J Nephrol 1986; 6(5):339 345.
92. Silberman H, Berne TV, Escandon R. Prospective evaluation of a double-lumen
subclavian dialysis catheter for acute vascular access. Am Surg 1992; 58(8):
443 445.
93. Schwab S, Besarab A, Beathard G, et al. NKF-DOQI clinical practice guidelines for
vascular access. Am J Kidney Dis 1997; 30(suppl):S150 S190.
94. Caridi JG, West JH, Stavropoulos SW, Hawkins IF Jr. Internal jugular and upper
extremity venous access in interventional radiology: is a postprocedural chest
radiograph necessary? Am J Roentgenol 2000; 174(2):363 366.
95. Lung GB, Trerotola SO, Scheel PF Jr, Savader SJ, Mitchell SE, Venbrux AC,
Osterman FA Jr. Outcome of tunneled hemodialysis catheters placed by radiol-
ogists. Radiology 1996; 198(2):467 472.
96. Beathard GA. Catheter thrombosis. Semin Dial 2001; 14(6):441 445.
97. Skolnick ML. The role of sonography in the placement and management of
jugular and subclavian central venous catheters. Am J Roentgenol 1994; 163(2):
291 295.
98. Dunkman WB, Leinbach RC, Buckley MJ, Mundth ED, Kantrowitz AR, Austen WG,
Sanders CA. Clinical and hemodynamic results of intraaortic balloon pumping and
surgery for cardiogenic shock. Circulation 1972; 46(3): 465477.
99. Hyson EA, Ravin CE, Kelley MJ, Curtis AM. Intraaortic counterpulsation balloon:
radiographic considerations. Am J Roentgenol 1977; 128(6):915 918.
100. Landay MJ, Mootz AR, Estrera AS. Apparatus seen on chest radiographs after
cardiac surgery in adults. Radiology 1990; 174(2):477 482.
101. Tegtmeyer CJ. Roentgenographic assessment of causes of cardiac pacemaker
failure and complications. CRC Crit Rev Diagn Imaging 1977; 9(1):1 50.
102. Hinke DH, Zandt-Stastny DA, Goodman LR, Quebbeman EJ, Krzywda EA, Andris
DA. Pinch-off syndrome: a complication of implantable subclavian venous access
devices. Radiology 1990; 177(2):353 356.
103. Daly BD, Cascade PN, Hummel JD, Kalbeisch S, Strickberger SA, Langberg J,
Morady F. Transvenous and subcutaneous implantable cardioverter debrillators:
radiographic assessment. Radiology 1994; 191(1):273 278.
398 Barack et al.
104. Munter DW, DeLacey WA. Automatic implantable cardioverter-debrillators.

Emerg Med Clin North Am 1994; 12(3):579 595.
105. Tocino I, Westcott JL. Barotrauma. Radiol Clin North Am 1996; 34(1):59 81.
106. McLoud TC, Barash PG, Ravin CE. PEEP: radiographic features and associated
complications. Am J Roentgenol 1977; 129(2):209 213.
107. Zimmerman JE, Goodman LR, Shahvari MB. Effect of mechanical ventilation and
positive end-expiratory pressure (PEEP) on chest radiograph. Am J Roentgenol
1979; 133(5):811 815.
108. Kolobow T, Moretti MP, Fumagalli R, Mascheroni D, Prato P, Chen V, Joris M.
Severe impairment in lung function induced by high airway pressure during
mechanical ventilation. An experimental study. Am Rev Respir Dis 1987; 135(2):
312315.
109. Parker JC, Hernandez LA, Peevy KJ. Mechanisms of ventilator-induced lung
injury. Crit Care Med 1993; 21(1):131 143.
110. Slutsky AS. Mechanical ventilation. American College of Chest Physicians
Consensus Conference. Chest 1993; 104(6):1833 1859. Erratum in: Chest 1994;
106(2):656.
111. Tsuno K, Prato P, Kolobow T. Acute lung injury from mechanical ventilation at
moderately high airway pressures. J Appl Physiol 1990; 69(3):956 961.
112. Webb HH, Tierney DF. Experimental pulmonary edema due to intermittent positive
pressure ventilation with high ination pressures. Protection by positive end-
expiratory pressure. Am Rev Respir Dis 1974; 110(5):556 565.
113. Westcott JL, Cole SR. Interstitial pulmonary emphysema in children and adults:
roentgenographic features. Radiology 1974; 111(2):367 378.
114. Zwillich CW, Pierson DJ, Creagh CE, Sutton FD, Schatz E, Petty TL. Compli-
cations of assisted ventilation. A prospective study of 354 consecutive episodes.
Am J Med 1974; 57(2):161 170.
115. Gammon RB, Shin MS, Buchalter SE. Pulmonary barotrauma in mechanical
ventilation. Patterns and risk factors. Chest 1992; 102(2):568 572.
116. Petersen GW, Baier H. Incidence of pulmonary barotrauma in a medical ICU. Crit
Care Med 1983; 11(2):67 69.
117. Woodring JH. Pulmonary interstitial emphysema in the adult respiratory distress
syndrome. Crit Care Med 1985; 13(10):786 791.
118. Dreyfuss D, Soler P, Basset G, Saumon G. High ination pressure pulmonary
edema. Respective effects of high airway pressure, high tidal volume, and positive
end-expiratory pressure. Am Rev Respir Dis 1988; 137(5):1159 1164.
119. Manning HL. Peak airway pressure: why the fuss? Chest 1994; 105(1):242 247.
120. Pierson DJ. Alveolar rupture during mechanical ventilation: Role of PEEP, peak
airway pressure, and distending volume. Respir Care 1988; 33:472.
121. Macklin MI, Macklin CC. Malignant interstitial emphysema of the lungs and med-
iastinum as important occult complication in many respiratory diseases and other
conditions: Interpretation of clinical literature in light of laboratory experiment.
Medicine. 1944; 23:281 358.
122. Haake R, Schlichtig R, Ulstad DR, Henschen RR. Barotrauma. Pathophysiology,
risk factors, and prevention. Chest 1987; 91(4):608 613.
123. Emery JL. Interstitial emphysema, pneumothorax, and air block in the newborn.
Lancet 1956; 1:405 409.
124. Rohlng BM, Webb WR, Schlobohm RM. Ventilator-related extra-alveolar air in
adults. Radiology 1976; 121(1):25 31.
125. Tagliabue M, Merlini L. Computed tomography in the diagnosis of pulmonary bar-
otrauma associated with adult respiratory distress syndrome. Radiol Med (Torino)
1994; 87(1 2):45 52.
126. Tocino I. Abnormal air and pleural uid collections. Goodman LR, Putman CE,
eds. Critical Care Imaging. Philadelphia: WB Saunders, 1992:137 160.
127. Ovenfor C. Pulmonary interstitial emphysema. Acta Radiol Supple (Stockh) 1964;
224:1.
128. Albelda SM, Gefter WB, Kelley MA, Epstein DM, Miller WT. Ventilator-induced
subpleural air cysts: Clinical, radiographic, and pathologic signicance. Am Rev
Respir Dis 1983; 127(3):360 365.
129. Chon KS, van Sonnenberg E, DAgostino HB, OLaoide RM, Colt HG, Hart E.
CT-guided catheter drainage of loculated thoracic air collections in mechanically
ventilated patients with acute respiratory distress syndrome. Am J Roentgenol
1999; 173(5):1345 1350.
130. Lane EJ, Proto AV, Phillips TW. Mach bands and density perception. Radiology
1976; 121(1):9 17.
131. Grosfeld JL, Kilman JW, Frye TR. Spontaneous pneumopericardium in the
newborn infant. J Pediatr 1970; 76(4):614 616.
132. Felson B. The mediastinum. Semin Roentgenol 1969; 4:40.
133. Levin B. The continuous diaphragm sign. A newly recognized sign of pneumome-
diastinum. Clin Radiol 1973; 24:337.
134. Crylak D, Milne EN, Imray TJ. Pneumomediastinum: a diagnostic problem.
Crit Rev Diagn Imaging 1984; 23(1):75 117.
135. Hammond DI. The ring-around-the-artery sign in pneumomediastinum.
Can Assoc Radiol J 1984; 35(1):88 89.
136. Gattinoni L, Bombino M, Pelosi P, Lissoni A, Pesenti A, Fumagalli R, Tagliabue M.
Lung structure and function in different stages of severe adult respiratory distress
syndrome. J Am Med Assoc 1994; 271(22):1772 1779.
137. de Latorre FJ, Tomasa A, Klamborg J, Leon C, Soler M, Rius J. Incidence of pneu-
mothorax and pneumomediastinum in patients with aspiration pneumonia requiring
ventilatory support. Chest 1977; 72(2):141 144.
138. Steier M, Ching N, Roberts EB, Nealon TF Jr. Pneumothorax complicating
continuous ventilator support. J Thorac Cardiovasc Surg 1974; 67(1):17 23.
139. Tocino IM. Pneumothorax in the supine patient: Radiographic anatomy. Radio-
graphics 1985; 5:557 586.
140. Tocino IM, Miller MH, Fairfax WR. Distribution of pneumothorax in the supine
and semirecumbent critically ill adult. Am J Roentgenol 1985; 144(5):901 905.
141. Kollef MH. Risk factors for the misdiagnosis of pneumothorax in the intensive care
unit. Crit Care Med 1991; 19(7):906 910.
142. Lams PM, Jolles H. The effect of lobar collapse on the distribution of free intra-
pleural air. Radiology 1982; 142(2):309 312.
143. Moskowitz PS, Griscom NT. The medial pneumothorax. Radiology 1976;
120(1):143 147.
144. Markowitz RI. The anterior junction line: A radiographic sign of bilateral pneu-
mothorax in neonates. Radiology 1988; 167(3):717 719.
400 Barack et al.
145. Ziter FM Jr, Westcott JL. Supine subpulmonary pneumothorax. Am J Roentgenol

1981; 137(4):699 701.
146. Kurlander GJ, Helmen CH. Subpulmonary pneumothorax. Am J Roentgenol
Radium Ther Nucl Med 1966; 96(4):1019 1021.
147. Gordon R. The deep sulcus sign. Radiology 1980; 136(1):25 27.
148. Rhea JT, vanSonnenberg E, McLoud TC. Basilar pneumothorax in the supine adult.
Radiology 1979; 133(3 Pt 1):593 595.
149. Fraser RG, Pare JAP. Roentgenologic signs in the diagnosis of chest disease. In:
Fraser RG, Pare JAP, eds. Diagnosis of Diseases of the Chest. An Integrated
Study Based on the Abnormal Roentgenogram. Philadelphia: Saunders,
1970:373 375.
150. Gobien RP, Reines HD, Schabel SI. Localized tension pneumothorax: unrecog-
nized form of barotrauma in adult respiratory distress syndrome. Radiology
1982; 142(1):15 19.
151. Spizarny DL, Goodman LR. Air in the minor ssure: a sign of right-sided pneu-
mothorax. Radiology 1986; 160(2):329 331.
152. Joannides M, Tsoulos GD. The etiology of interstitial and mediastinal emphysema.
Arch Surg 1930; 21:333.
153. Summers B. Pneumoperitoneum associated with articial ventilation. Br Med J
1979; 1(6177):1528 1530.
154. Golden GT, Chandler JG. Colonic ileus and cecal perforation in patients requiring
mechanical ventilatory support. Chest 1975; 68:661.
155. Shevland JE, Hireleman MR, Hoang KA, Kealey GP. Lobar collapse in the surgical
intensive care unit. Br J Radiol 1983; 56:531 534.
156. Proto AV. Conventional chest radiographs: anatomic understanding of newer
observations. Radiology 1992; 183:593 603.
157. Proto AV, Moser ES Jr. Upper lobe volume loss: divergent and parallel patterns of
vascular reorientation. Radiographics 1987; 7:875 887.
158. Woodring JH, Reed JC. Radiographic manifestations of lobar atelectasis. J Thor
Imaging 1996; 11:109 114.
159. Woodring JH, Reed JC. Types and mechanisms of pulmonary atelectasis. J Thor
Imaging 1995; 11:92 108.
160. Winer-Muram HT, Rubin SA, Ellis JV, Jennings SG, Arheart KL, Wunderink RG,
Leeper KV, Meduri GU. Pneumonia and ARDS in patients receiving mechanical
ventilation: diagnostic accuracy of chest radiography. Radiology 1993; 188: 479485.
161. Goodman LR, Putnam CE. Intensive Care Radiology. St. Louis: Mosby, 1983.
162. Muller NL, Fraser RS, Colman NC, Pare PD. Pulmonary infection. In: Muller NL,
Fraser RS, Colman NC, Pare PD, eds. Radiologic Diagnosis of Diseases of the
Chest. Philadelphia: Saunders, 2001:141 212.
163. Heinzman ER. The Lung: Radiologic-Pathologic Correlations. 2nd ed. St. Louis:
Mosby, 1984.
164. Ohi H, Yanagihara K, Miyazaki Y, Hirakata Y, Tomono K, Kadota J, Tashiro T,
Kohno S. Hospital-acquired pneumonia in general wards of a Japanese tertiary-
hospital. Respirology 2004; 9(1):120 124.
165. Eickhoff TC. Nosocomial infections. N Engl J Med 1982; 306:1545 1546.
166. Cross AS, Roup B. Role of respiratory assistance devices in endemic nosocomial
pneumonia. Am J Med 1981; 70(3):681 685.
167. Sanford JP, Pierce AK. Current infection problemsrespiratory. Proceedings of

the International Conference on Nosocomial Infections. Center for Disease
Control. American Hospital Association, Chicago, August 3 6, 1970; pp 77 81.
168. Tillotson JR, Finland M. Bacterial colonization and clinical superinfection of the
respiratory tract complicating antibiotic treatment of pneumonia. J Infect Dis
1969; 119(6):597 624.
169. DePaso WJ. Aspiration pneumonia. Clin Chest Med 1991; 12:269 284.
170. Groskin SA, Panicek DM, Ewing DK, Rivera F, Math K, Teixeira J, Heitzman ER,
Markarian B. Bacterial lung abscess: A review of the radiographic and clinical-
features of 50 cases. J Thorac Imag 1991; 6:67 72.
171. Fraser RS, Muller NL, Colman N, Pare PD. General features of pulmonary infec-
tion. In: Fraser RS, Muller NL, Colman N, Pare PD, eds. Diagnosis of Diseases
of the Chest. Philadelphia: Saunders, 1999:703 704.
172. Muller NL, Fraser RS, Colman NC, Pare PD. Pulmonary edema. In: Muller NL,
Fraser RS, Colman NC, Pare PD, eds. Radiologic Diagnosis of Diseases of the
Chest. Philadelphia: Saunders, 2001:432 452.
173. Gluecker T, Capasso P, Schnyder P, Gudinchet F, Schaller MD, Revelly JP,
Chiolero R, Vock P, Wicky S. Clinical and radiologic features of pulmonary
edema. Radiographics 1999; 19:1507 1531.
174. McHugh TJ, Forrester JS, Adler L, Zion D, Swan HJ. Pulmonary vascular conges-
tion in acute myocardial infarction: Hemodynamic and radiologic correlations.
Ann Intern Med 1972; 76:29 33.
175. Slutsky RA, Higgins CB. Intravascular and extravascular pulmonary uid volumes:
II. Response to rapid increases in left atrial pressure and the theoretical implications
for pulmonary radiographic and radionuclide imaging. Invest Radiol 1983; 18:
33 39.
176. Fleischner FG. The buttery pattern of acute pulmonary edema. Am J Cardiol 1967;
20:39 46.
177. Smith RC, Mann H, Greenspan RH, Pope CF, Sostman HD. Radiographic differen-
tiation between different etiologies of pulmonary edema. Invest Radiol 1987;
22(11):859 863.
178. Aberle DR, Wiener-Kronish JP, Webb WR, Matthay MA. Hydrostatic versus
increased permeability pulmonary edema: diagnosis based on radiographic criteria
in critically ill patients. Radiology 1988; 168(1):73 79.
179. Ravin CE. Pulmonary vascularity: Radiographic considerations. J Thor Imaging
1988; 3:1 13.
180. Morgan PW, Goodman LR. Pulmonary edema and adult respiratory distress syn-
drome. Radiol Clin North Am 1991; 29:943 963.
181. Woodring JH. Pulmonary artery bronchus ratios in patients with normal lungs,
pulmonary vascular plethora, and congestive heart failure. Radiology 1991;
179:115 122.
182. Milne EN, Pistolesi M, Miniati M, Giuntini C. The radiologic distinction of cardio-
genic and non-cardiogenic edema. Am J Roentgenol 1985; 144:879894.
183. Gurney JW, Goodman LR. Pulmonary edema localized in the right upper lobe
accompanying mitral regurgitation. Radiology 1989; 171:397 399.
184. Miyatake K, Nimura Y, Sakakibara H, Beppu S, Nimura Y. Localization and direc-
tion of mitral regurgitant ow in mitral orice studied with combined use of
402 Barack et al.
ultrasonic pulsed Doppler technique and two dimensional echocardiography. Br

Heart J 1982; 48:449 458.
185. Slanetz PJ, Truong M, Shepard JA, Trotman-Dickenson B, Drucker E, McLoud TC.
Mediastinal lymphadenopathy and hazy mediastinal fat: New CT ndings of
congestive heart failure. Am J Roentgenol 1998; 171:1307 1309.
186. Ostendorf P, Birzle H, Vogel WM, Hermayer C. Pulmonary radiographic abnorm-
alities in shock: Roentgen-clinical pathological correlation. Radiology 1975;
115:257 263.
187. Joffe N. The adult respiratory distress syndrome. Am J Roentgenol, Rad Ther Nuc
Med 1974; 122:719 732.
188. Dyck DR, Zylak CJ. Acute respiratory distress in adults. Radiology 1973; 106:
497 501.
189. Abel SJ, Finney SJ, Brett SJ, Keogh BF, Morgan CJ, Evans TW. Reduced mortality
in association with the acute respiratory distress syndrome (ARDS). Thorax 1998;
53(4):292 294.
190. Wheeler AP, Carroll FE, Bernard GR. Radiographic issues in adult respiratory
distress syndrome. New Horiz 1993; 1(4):471 477.
191. Ortiz Chinchilla D, Jam Gatell MR. Long term of quality of life and mortality in
acute respiratory distress syndrome (ARDS) patients. Enferm Intensiva 2003;
14(3):88 95.
192. Ketai LH, Godwin JD. A new view of pulmonary edema and acute respiratory
distress syndrome. J Thorac Imaging 1998; 13(3):147 171.
193. Goodman LR. Congestive heart failure and adult respiratory distress syndrome.
New insights using computed tomography. Radiol Clin North Am 1996;
34(1):33 46.
194. Tagliabue M, Casella TC, Zincone GE, Fumagalli R, Salvini E. CT and chest radi-
ography in the evaluation of adult respiratory distress syndrome. Acta Radiol 1994;
35(3):230 234.
195. Goodman LR, Fumagalli R, Tagliabue P, Tagliabue M, Ferrario M, Gattinoni L,
Pesenti A. Adult respiratory distress syndrome due to pulmonary and extra-
pulmonary causes: CT, clinical, and functional correlations. Radiology 1999;
213(2):545552.
196. Miniati M, Pistolesi M, Paoletti P, Giuntini C, Lebowitz MD, Taylor AE, Milne EN.
Objective radiographic criteria to differentiate cardiac, renal, and injury lung
edema. Invest Radiol 1988; 23(6):433 440.
197. Armstrong P, Wilson AG, Dee P. Imaging of Diseases of the Chest. St. Louis:
Mosby, 1990:638 650.
198. Mattison LE, Coppage L, Alderman DF, Herlong JO, Sahn SA. Pleural effusions in
the medical ICU: prevalence, causes, and clinical implications. Chest 1997;
111(4):1018 1023.
199. Ruskin JA, Gurney JW, Thorsen MK, Goodman LR. Detection of pleural effusions
on supine radiographs. Am J Roentgenol 1987; 148(4):681 683.
200. Woodring JH. Recognition of pleural effusion on supine radiographs: how much
uid is required. Am J Roentgenol 1984; 142:59 64.
201. Reeder MM, Felson B. Gamuts in radiology: comprehensive lists of roentgen
differential diagnosis. Cincinnati: Audiovisual Radiology of Cincinnati, 1975;
F-28.
202. Rudikoff JC. Early detection of pleural uid. Chest 1980; 77:109 111.
203. Weiss W, Boucot KR, Gefter WI. Localized interlobar effusion in congestive heart
failure. Ann Intern Med 1953; 38:1177 1186.
204. Feder BH, Wilk SP. Localized interlobar effusion in heart failure: phantom lung
tumor. Dis Chest 1956; 30:289 297.
205. Greenspan RH, Ravin CE, Polanski SM, McLoud TC. Accuracy of the chest radio-
graph in diagnosis of pulmonary embolism. Invest Radiol 1982; 17(6):539 543.
206. Elliott CG, Goldhaber SZ, Visani L, DeRosa M. Chest radiographs in acute
pulmonary embolism. Results from the International Cooperative Pulmonary
Embolism Registry. Chest 2000; 118(1):33 38.
207. Westermark N. On roentgen diagnosis of lung embolism. Acta Radiol 1938;
19:357 372.
208. Palla A, Donnamaria V, Petruzzelli S, Rossi G, Riccetti G, Giuntini C. Enlargement
of the right descending pulmonary artery in pulmonary embolism. Am J Roentgenol
1983; 141(3):513 517.
209. Worsley DF, Alavi A, Arondick JM, Chen JT, Greenspan RH, Ravin CE. Chest
radiographic ndings in patients with acute pulmonary embolism: observations
from the PIOPED Study. Radiology 1993; 189(1):133 136.
210. Jacoby CG, Mindell HJ. Lobar consolidation in pulmonary embolism. Radiology
1976; 118:287 290.
211. Hampton AO, Castleman B. Correlation of postmortem chest teleroentgenograms
with autopsy ndings, with special reference to pulmonary embolism and infarc-
tion. Am J Roentgenol 1940; 43:305 326.
212. Qanadli SD, Hajjam ME, Mesurolle I, Barre O, Bruckert F, Joseph T, Mignon F,
Vieillard-Baron A, Dubourg O, Lacombe P. Pulmonary embolism detection:
prospective evaluation of dual-section helical CT versus selective pulmonary
angiography in 157 patients. Radiology 2000; 217:447 455.
213. Goodman LR, Curtin JJ, Mewissen MW, Foley WD, Lipchik RJ, Crain MR, Sagar
KB, Collier BD. Detection of pulmonary embolism in patients with unresolved
clinical and scintigraphic diagnosis: helical CT versus angiography. Am J
Roentgenol 1995; 164:1369 1374.
214. van Rossum AB, Pattynama PM, Ton ER, Treurniet FE, Arndt JW, van Eck B, Kieft
GJ. Pulmonary embolism: validation of spiral CT angiography in 149 patients.
Radiology 1996; 201(2):467 470.
215. Patel S, Kazerooni EA, Cascade PN. Pulmonary embolism: optimization of small pul-
monary artery visualization at multi-detector row CT. Radiology 2003; 227:455460.
216. Baile EM, Gregory GK, Muller NL, DYachkova Y, Coche EE, Pare PD, Mayo JR.
Spiral computed tomography is comparable to angiography for the diagnosis of
pulmonary embolism. Am J Respir Crit Care Med 2000; 161:1010 1015.
217. Garg K, Sieler H, Welsh CH, Johnston RJ, Russ PD. Clinical validity of helical CT
being interpreted as negative for pulmonary embolism: implications for patient
treatment. Am J Roentgenol 1999; 172(6):1627 1631.
218. Kavanagh EC, OHare A, Hargaden G, Murray JG. Risk of pulmonary embolism
after negative MDCT pulmonary angiography ndings. Am J Roentgenol 2004;
182(2):499 504.
219. Tillie-Leblond I, Mastora I, Radenne F, Paillard S, Tonnel AB, Remy J, Remy-
Jardin M. Risk of pulmonary embolism after a negative spiral CT angiogram in
404 Barack et al.
patients with pulmonary disease:1-year clinical follow-up study. Radiology 2002;

223(2):461 467.
220. Greaves SM, Hart EM, Brown K, Young DA, Batra P, Aberle DR. Pulmonary
thromboembolism: spectrum of ndings on CT. Am J Roentgenol 1995; 165(6):
1359 1363.
221. Beecham RP, Dorfman GS, Cronan JJ, Spearman MP, Murphy TP, Scola FH. Is
bilateral lower extremity compression sonography useful and cost-effective in
the evaluation of suspected pulmonary embolism? Am J Roentgenol 1993;
161:1289 1292.
222. Dalen JE, Brooks HL, Johnson LW, Meister SG, Szucs MM, Dexter L. Pulmonary
angiography in acute pulmonary embolism: indications, techniques, and results in
367 patients. Am Heart J 1971; 81:175 185.
223. Noveline RA, Baltarowich OH, Athanasoulis CA, Waltman AC, Greeneld AJ,
McKusick KA. The clinical course of patient with suspected pulmonary embolism
and a negative pulmonary arteriogram. Radiology 1978; 126:561 567.
224. Musset D, Parent F, Meyer G, Maitre S, Girard P, Leroyer C, Revel MP, Carette
MF, Laurent M, Charbonnier B, Laurent F, Mal H, Nonent M, Lancar R, Grenier
P, Simonneau G. Diagnostic strategy for patients with suspected pulmonary
embolism: a prospective multicentre outcome study. Lancet 2002; 360:1914 1920.
225. Perrier A, Howarth N, Didier D, Maitre S, Girard P, Leroyer C, Revel MP, Carette
MF, Laurent M, Charbonnier B, Laurent F, Mal H, Nonent M, Lancar R, Grenier P,
Simonneau G. Performance of helical computed tomography in unselected
outpatients with suspected pulmonary embolism. Ann Intern Med 2001; 135:
88 97.
226. Goodman LR, Lipchik RJ, Kuzo RS, Liu Y, McAuliffe TL, OBrien DJ. Subsequent
pulmonary embolism: risk after a negative helical CT pulmonary angiogram
prospective comparison with scintigraphy. Radiology 2000; 215:535 542.
227. van Strijen MJL, de Monye W, Schiereck J, Kieft GJ, Prins MH, Huisman MV,
Pattynama PM. Single-detector helical computed tomography as the primary diag-
nostic test in suspected pulmonary embolism: a multicenter clinical management
study of 510 patients. Ann Intern Med 2003; 138:264 271.
228. Fedullo PF, Tapson V. The evaluation of pulmonary embolism. N Engl J Med 2003;
349:1247 1256.
229. Jaffe RB, Koschmann EB. Septic pulmonary emboli. Radiology 1970; 96:527 532.
230. Iwama T, Shigematsu S, Asami K, Kubo I, Kitazume H, Tanabe S, Matsunaga
Y. Tricuspid valve endocarditis with large vegetations in a non-drug addict
without underlying cardiac disease. Intern Med 1996; 35(3):203 206.
231. Clifford CP, Eykyn SJ, Oakley CM. Staphylococcal tricuspid valve endocarditis in
patients with structurally normal hearts and no evidence of narcotic abuse. QJM
1994; 87(12):755 757.
232. Roberts WC, Buchbinder NA. Right-sided valvular infctive endocarditis: A clinico-
pathologic study of twelve necropsy patients. Am J Med 1972; 53:7 19.
233. Fraser RS, Muller NL, Colman N, Pare PD. Bacteria other than mycobacteria.
In: Fraser RS, Muller NL, Colman N, Pare PD, eds. Fraser and Pares Diagnosis
of Diseases of the Chest 4th ed. Philadelphia: Saunders, 1999:734 797.
234. Zelefsky MN, Lutzker LG. The target sign: a new radiologic sign of septic pulmon-
ary emboli. Am J Roentgenol 1977; 129:453 455.
235. Silingardi V, Canossi GC, Torelli G, Russo G. The radiology target sign of septic
pulmonary embolism in a case of acute myelogenous leukemia. Respiration 1981;
42:61 66.
236. Kuhlman JE, Fishman EK, Teigen C. Pulmonary septic emboli: diagnosis with CT.
Radiology 1990; 174:211 213.
237. Iwasaki Y, Nagata K, Nakanishi M, Natuhara A, Harada H, Kubota Y, Yokomura I,
Hashimoto S, Nakagawa M. Spiral CT ndings in septic pulmonary emboli.
Eur J Radiol 2001; 37(3):190 194.
238. Schwartz EE, Teplick JG, Onesti G, Schwartz AB. Pulmonary hemorrhage in
renal disease: Goodpastures syndrome and other causes. Radiology 1977; 122:
39 46.
239. Crawford ES. The diagnosis and management of aortic dissection. J Am Med Assoc
1990; 264(19):2537 2541.
240. Svensson LG, Crawford ES, Hess KR, Coselli JS, Sa HJ. Dissection of the aorta
and dissecting aortic aneurysms: improving early and long-term surgical results.
Circulation 1990; 82(suppl 5):IV24 38.
241. Nienaber CA, von Kodolitsch Y, Nicolas V, Siglow V, Piepho A, Brockhoff C,
Koschyk DH, Spielmann RP. The diagnosis of thoracic aortic dissection by
noninvasive imaging procedures. N Engl J Med. 1993; 328:1 9.
242. Sommer T, Fehske W, Holzknecht N, Smekal AV, Keller E, Lutterbey G, Kreft B,
Kuhl C, Gieseke J, Abu-Ramadan D, Schild H. Aortic dissection: a comparative
study of diagnosis with spiral CT, multiplanar transesophageal echocardiography,
and MR imaging. Radiology 1996; 199(2):347 352.
243. Andresen J, Baekgaard N, Allermand H. Evaluation of patients with thoracic
aortic dissection by intraarterial digital substraction angiography. Vasa 1992;
21(2):167 170.
244. Ledbetter S, Stuk JL, Kaufman JA. Helical (spiral) CT in the evaluation of emer-
gent thoracic aortic syndromes. Traumatic aortic rupture, aortic aneurysm, aortic
dissection, intramural hematoma, and penetrating atherosclerotic ulcer. Radiol
Clin N Am 1999; 37(3):1 17.
245. Movsowitz HD, Lampert C, Jacobs LE, Kotler MN. Penetrating atherosclerotic
aortic ulcers. Am Heart J 1994; 128(6 Pt 1):1210 1217.
246. Harris JA, Bis KG, Glover JL, Bendick PJ, Shetty A, Brown OW. Penetrating ather-
osclerotic ulcers of the aorta. J Vasc Surg 1994; 19(1):90 98.
247. Welch TJ, Stanson AW, Sheedy PF, Johnson CM, McKusick MA. Radiologic
evaluation of penetrating aortic atherosclerotic ulcer. Radiographics 1990;
10:675 685.
248. Yucel EK, Steinberg FL, Egglin TK, Geller SC, Waltman AC. Athanasoulis CA.
Penetrating aortic ulcers: diagnosis with MR imaging. Radiology 1990;
177:799 781.
249. Groskin SA. Selected topics in chest trauma. Semin Ultrasound CT MRI 1996;
17:119 141.
250. Mirvis SE, Bidwell JK, Buddemeyer EU, Diaconis JN, Pais SO, Whitley JE,
Goldstein LD. Value of chest radiography in excluding traumatic aortic rupture.
Radiology 1987; 163:487 493.
251. Goldstein LA, Thompson WR. Esophageal perforation: A 15 year experience. Am J
Surg 1982; 143:495 503.
406 Barack et al.
252. Maroney MJ, Mirvis SE, Shanmuganathan K. Esophageal occlusion caused by thor-
acic spine fracture or dislocation: CT diagnosis. Am J Roentgenol 1996; 167:714
715.
253. Lee S, Mergo PJ, Ros PR. The leaking esophagus: CT patterns of esophageal
rupture, perforation, and stulization. Crit Rev Diagn Imaging 1996; 37(6):
461 490.
254. Cardella JF, Fox PS, Lawler JB. Interventional radiologic placement of peripherally
inserted central catheters. J Vasc Intervent Radiol 1993:4:653 660.
255. Mewissen MW, Crain MR, Paz-Fumagalli R, Beres RA, Wertz RA. Interventional
Procedures in the Intensive Care Unit Patient. Radiol Clin North Am 1996; 34:
157 176.
256. Andrews NC, Parker EF, Shaw RR, Wilson NJ, Webb WR. Management of nontu-
berculous empyema. Am Rev Respir Dis 1962; 85:935 936.
257. Sahn SA. Management of complicated parapneumonic effusions. Am Rev Respir
Dis 1993; 148:813 817.
258. van Sonnenberg E, Nakamonto SK, Mueller PR, Casola G, Neff CC, Friedman PJ,
Ferrucci JT Jr, Simeone JF. CT and ultrasound-guided catheter drainage of
empyemas after chest-tube failure. Radiology 1984; 151(2):349 353.
259. Silverman SB, Muller PR, Saini S, Hahn PF, Simeone JF, Forman BH, Steiner E,
Ferrucci JT. Thoracic empyema: Management with image guided catheter drai-
nage. Radiology 1988; 169(1):5 9.
260. Robinson LA, Mountol AL, Fleming WH, Alonso A, Galbraith TA. Intrapleural
brinolytic treatment of multiloculated thoracic empyemas. Ann Thorac Surg
1994; 57(4):803 814.
261. Pollak JS, Passik CS. Intrapleural urokinase in the treatment of loculated pleural
effusions. Chest 1994; 105:868 873.
262. Ha HK, Kang MW, Park JM, Yan WJ, Shinn KS, Bahk YW. Lung abscess:
Percutaneous catheter therapy. Acta Radiologica 1993; 34:362 365.
263. Yellin A, Yellio EO, Lieberman Y. Percutaneous tube drainge: The treatment of
choice for refractory lung abscess. Ann Thorac Surg 1985; 39:266 170.
264. Vainrub B, Musher DM, Guinn GA, Young EJ, Septimus EJ, Travis LL. Percutaneous
drainage of lung abscess. Am Rev Respir Dis 1978; 117(1):153160.
265. van Sonnenberg E, DAgostino HB, Casola G, Wittich GR, Varney RR, Harker
C. Lung abscess: CT-guided drainage. Radiology 1991; 178(2):347 351.
266. Taavitsainen M, Bondestam S, Mankinen P, Pitkaranta P, Tierala E. Ultrasound
guidance for pericardiocentesis. Acta Radiologica 1991; 32(1):9 11.
267. Girard MS, Kaufman JA, Lau TN, Morris CS, Oliva V, Parr C. Inferior vena cava
ltration. Society of Interventional Radiology 28th Annual Scientic Meeting
Workshop Book, 2003, pp. 349 353.
Index
ABG (arterial blood gases) Airway stents, insertion techniques,

oxygen saturation, 107 304 305
surgical ICU and, 107 Airway trauma, 33 34
spontaneous variability, 107 barotrauma, 33 34
Abscess, mediastinal and pericardial, 391 nosocomial sinusitis, 33 34
Acute asthma, 12 pneumonia, 33 34
Acute respiratory failure, management of, tracheal stenosis, 33 34
33 34 Alveolar gas equation, 2
Air-entrainment mask, oxygen American Society of Anesthesiologists,
administration devices, 9 difcult airway algorithm, 23
Airway assessment, 14 Anaerobic infections, oxygen therapy, 3
Airway management, 13 24 Analgesics, in ventilator weaning, 170,
airway assessment, 14 171
direct laryngoscopy visualization, 15 Aortic injury, 383 384
endotracheal intubation, 13 ARDS, 372 374
indications for, 13 14 Arm ports, 388
laryngeal visualization, 15 Arterial blood gas analysis, problems
Mallampati classication, 14 with, 106
nasotracheal vs. orotracheal intubation, Arterial blood gases. See ABG.
14 Arterial catheters
oxygen therapy and, 1 24 complications, 222 223
predicting difcult mask ventilation, 14 equipment, 220
predicting difcult tracheal intubation, ICU procedures and, 219 225
14 15 indications, 220
thyromental distance, 15 technical errors, 223 225
Airway pressure waveform, patient- techniques, 220
ventilator dyssynchrony, 130 Arterialized capillary, blood gases, 108
407
408 Index
Aspergillus avus, 295 Breath types, 87 94

Aspergillus fumigatus, 295 closed-loop control systems, 93 94
Aspergillus niger, 295 equipment and special features,
Aspergillus species, 295 95 97
Aspiration pneumonia, 367 machine breath, 87 91
Assist/control ventilator, 98 99 adapted support ventilation, 90
Atelectasis, 365 366 airway pressure-release ventilation,
insertion techniques, 298 90
bi-level ventilation, 89
BACK maneuver, laryngoscopy pressure control ventilation, 88 89
visualization, 16 pressure-regulated volume control
BAL uid as marker, 289, 292 vendilation, 90
Barium volume-assured pressure support
portable antero-posterior view, 333 ventilation, 90 91
tracheoesophageal stula, 333 volume-control ventilation, 87 88
Barotrauma, 34, 65, 354 363 spontaneous breath, 91
Bat-wing pattern in pulmonary pressure support breath, 91 92
edema, 373 proportional assist ventilation
Benzodiazepines, over-sedation breath, 93
with, 288 tube compensation breath, 92 93
Blood gas analyzers, cost of, 110 volume support breath, 92
Blood gas monitors, mechanical Bronchoalveolar lavage (BAL) as
ventilation and, 109 110 diagnostic procedure, 289
Blood gases, arterialized capillary, 108 Bronchopleural stula, insertion
Boerhaaves syndrome, 384 techniques, 297
Brachytherapy, 304 Bronchopneumonia, 367
Breath, denition of, 78 Bronchoscope
Breath patterns/modes, 94 99 exible beroptic, 286
assist/control ventilator, 98 99 ultrathin, 286
continuous positive airway Bronchoscopic biopsies, 290
pressure, 99 Bronchoscopic lung volume reduction,
intermittent mandatory ventilation, insertion techniques, 305
98 99 Bronchoscopy, 285 306
Breath phases, 78 79 contraindications, 286 287
cycle phase, 83 84 history of, 285
primary cycling mechanisms, 83 immunocompetent patients, 293
secondary cycling mechanisms, indications for exible, 287
84 indications for rigid, 286
expiratory phase, 8487 laser, 299
FiO2:PEEP relationship, 8586 therapeutic, indications and options,
open lung approach, 86 87 298 305
pressure volume curve, 86 Bronchospasm, 291
limit phase, volume control BURP maneuver, laryngoscopy
ventilation, 81 visualization, 16
pressure control ventilation, 81 83
trigger phase, 79 80 Capnogram
patient triggering, 79 80 airow obstruction and, 121
time triggering, 79 pulmonary embolism, 124
Index 409
time-based, 120 121 Chest tube thoracostomy

volume-based, 119120 anatomy, 262
Capnography, mechanical ventilation, complications, 270 271
117 127 contraindications, 266
Capnometers drain size, 266 267
conguration of, 118 equipment, 266
mainstream and sidestream, 118 ICU procedures and, 261 272
Capnometry indications, 262 266
microstream, 118 technique, 267 269
weaning from mechanical ventilation, tube management, 271 272
125 tube removal, 272
Carbon monoxide poisoning, oxygen Chronic obstructive pulmonary disease
therapy, 3 (COPD)
Carboxyhemoglobin, half-life of, 3 BiPAP
Cardiogenic edema, radiographic denition of, 36
ndings, 375 use of, 36
Cardiogenic pulmonary edema (CPAP), exacerbations of, acute, 35
38 41 intubation rate, 37 38
APACHE II scores, 39 length of stay, 37 38
benets of, 38 39 mortality, 37 38
comparison to NPPV, 40 41 randomized trials, 35 36
randomized trials, 39 41 Ciaglia PDT technique, 314, 317
performance in, 38 39 Closed-loop control systems, 93 94
positive intrathoracic pressure, Clostridium difcile infection,
benets of, 38 ventilator-dependency, 200
respiratory failure, reversibility, 38 CO2 detector, single-use, 21
Cardiogenic edema CO2 rebreathing technique, 125
vs. non-cardiogenic edema, 374 Coccidioidomycosis immitis, 294
vs. pulmonary edema, 371 375 Common thoracic abnormalities, in the
Catheter ICU, 364 385
fragmentation of, 348 Compliance, rate, oxygen, and
malpositioned, 389 pressure (CROP) index,
in ICU, 344 352 162 163
occluded, 389 Computed radiography (CR), 334
Catheter drainage, in the ICU, 387
percutaneous, 390391 indications, 385
Catheter insertion, mediastinal technique, 385 387
enlargement, 348 Coumadin, and bronchoscopy, 288
Central vein access, 389 Cricothyrotomy, 22
Central venous catheters Cryptococcus neoformans, 295
complications, 231 236 Cystic brosis, 50
general technique, 226 231 Cytomegalic virus (CMV), 295
ICU procedures and, 225 226,
344 349 Decompensated obstructive sleep
indications, 225 226 apnea, 50
ultrasound-guided insertions, 231 Deep vein thrombosis, management of,
Chest radiograph, portable, 331 337 391 392
Chest trauma, insertion techniques, 296 Denition of oxygen (O2), 1
410 Index
Delirium, from weaning, 171 175 ETT

recommendation to treat, 171 care of, 22 23
risk factors, 175 cricothyrotomy, 22
Demand oxygen conservers, 8 difcult airway and, 22
Dialysis catheters, in the ICU, laryngeal mask airway (LMA), 22
350 352 placement, conrmation of proper,
Difcult airway, ETT and, 22 20 21
Difcult airway algorithm, American Extubation, weaning vs., 165
Society of Anesthesiologists, 23 Eye irritation, 65
Diffuse alveolar damage (DAD), 369
Diffuse non-infectious pulmonary Fasciitis, oxygen therapy for, 3
disease, 296 Feeding tube. See nasoenteric tubes, 341
Digital radiography (DR), 334 Fick equation, 110
Dissecting hematoma, 383 FiO2:PEEP relationship, in breath phases,
of the aorta, 382 383 85 86
Double diaphragm sign, 363 Fleischner sign, 379
Dumon, 304 Flexible beroptic bronchoscopy,
Dynamic-Y, 304 sedation for, 288
Dyshemoglobinemias, in pulse Flexible bronchoscopy, insertion
oximetry, 113 techniques, 289
Flexible beroptic
Electrocautery, insertion techniques, ancillary equipment, 286
303 bronchoscope, 286
Embolic disease, 377 382 Flumazenil, 288
Empyema, 368 Fluorescence bronchoscopy, 291
End-expiratory pause maneuver, Foreign bodies, insertion techniques,
measurement of auto-PEEP, 297 298
129 Functional force reserve of respiratory
End-inspiratory pressure spike, active muscles, in weaning from
exhalation, 132 mechanical ventilation, 154
Endobronchial brushing, 290
Endobronchial ultrasound (EBUS), 290 Galvanic cell analyzers, 3
Endotracheal intubation Gas exchange, as weaning parameter, 161
airway management, 13 Gastric distension, 65
equipment for, 16 Gastric tonometer
etomidate, 17 conguration, 117
medications to facilitate, 17 mechanical ventilation, 116 117
propofol, 17
rocuronium, 17 Hamptons hump, 351
succinylcholine, 17 Heliox, 10 13
vecuronium, 17 acute asthma, 12
Endotracheal tube placement administration of, 12
in ICU, 338 340 stridot, 10
insertion techniques, 298 Hematoma of the aorta, dissecting,
Esophageal pressure, measurement of, 382 383
130 Hemodynamic instability, 52
with auto-PEEP, 131 Hemodynamics, of pulmonary artery
Esophageal rupture, 384 385 catheters, 242 250
Index 411
Hemoglobin oxygen saturation, asthmatics, 47 48

gold standard for measurement asphyxia in, 47
of, 107 chest/wall trauma, rib fractures,
Hemoptysis, insertion techniques, 297 48 49
Hemorrhage, pulmonary, 382 atelectasis in, 48
Heparin and bronchoscopy, 288 ineffective cough in, 48
Herpes viruses, 295 pneumonia in, 48
High-ow delivery system, 11 pulmonary contusion in, 48
oxygen administration devices, 6, COPD patients, and, 49 50
9 10 cystic brosis, 50
Hydropneumothorax, 364 decompensated obstructive sleep
Hydrostatic pulmonary edema, 369 apnea, 50
Hypercapnia in patients with head do not intubate status, 49 50
injury, 125 end-stage-disease patients, 49 50
Hypoxemia, 291 intubation relief, 49 50
oxygen therapy, 2, 3 kyphoscoliosis, 50
lung cancer patients, post-operative, 48
ICU procedures, 219 272 neuromuscular disease, 50
arterial catheters, 219 225 Pneumocystis carinii pneumonia, 50
catheters, 344 352 post-operative lung cancer
central line placement, 388 mucus plugging in, 48
central venous catheters, 225 236, post-pneumonia in, 48
344 349 pulmonary edema, non-cardiogenic
chest tube thoracostomy, 261 272 in, 48
devices used in, 338 364 respiratory distress syndrome, acute,
dialysis catheters, 350 352 49
endotracheal tube in, 338 340 rib fractures, chest/wall trauma,
image-guided interventional 48 49
procedures, 387 392 Indirect digital radiography, 334
implantable cardioverter-debrillator Insertion techniques
devices, 353 354 airway stents, 304 305
intra-aortic balloon pump, 352 353 atelectasis, 298
nasoenteric tubes, 340 342 bronchopleural stula, 297
peripherally inserted central catheters bronchoscopic lung volume reduction,
(PICC), 349 305
pulmonary artery catheters, 236 261, chest trauma, 296
349 350 complications, 291 292
pulmonary edema, 370 cryotherapy, 302 303
Swan-Ganz catheters, 349 350 electrocautery, 303
thoracostomy tube, 342 344 endotracheal tube placement, 298
tracheostomy tube, 340 extraanatomical airway tract, 305
transvenous pacing devices, 353 foreign bodies, 297 298
venous access, 387 389 hemoptysis, 297
venous evaluation, 387 388 lung cancer, 292
Image-guided interventional procedures, lung transplant rejection, 296 297
ICU, 387 392 neoplasm, debulking of obstructing,
Immunocompromised patients, 4550 299
acute respiratory distress syndrome, 49 photodynamic therapy (PDT), 303
412 Index
Insertion techniques (Contd.) volume control ventilation, 87 88

post-procedure care, 291 volume-assured pressure support
suspected pulmonary infections, ventilation (VAPS), 90 91
292 293 Mallampati class I airway, example of, 15
thermal injuries, 296 Mallampati classication, airway
tracheoesophageal stula, 296 management, 14
Inspiratory airways resistance estimation, Mask air-entrainment, 6
133 Mask non-rebreathing, 6
Intermittent mandatory ventilation Masks
(IMV), 98 99, 164 claustrophobia, 55
Intra-aortic balloon pump, ICU, 352 353 comparison of, 55
Intramural tonometry, as weaning compliance, 55
parameter, 163 face, 54
mouth breathing, 55
Jaw thrust maneuver, in laryngoscopic oronasal, 54
visualization, 16 patient cooperation, 55
simple, as oxygen administration
Kaposi sarcoma, 292 devices, 6
Knuckle sign, 379 ventilation, efcacy, 55
Kyphoscoliosis, 50 Mean airway pressure, determination of,
134
Laryngeal mask airway (LMA), 22, 24 Mechanical ventilation
Laryngoscopic visualization arterial blood gases, 105 108
BACK maneuver, 16 blood gas monitors, 109 110
BURP maneuver, 16 breath patterns, 78
jaw thrust maneuver, 16 capnography, 117 127
maneuvers to improve, 15 16 classication scheme, 77
Laser, bronchoscopy, 299 complications of, 354 363
Lidocaine, 288, 292 nancial impact of, 136
Lobar pneumonia, 367 gastric tonometry, 116 117
Lorazepam, in ventilator weaning, 170 goals and values matrix, 177
LTAC, prolonged mechanical ventilation lung mechanics and graphics, 127 135
and, 190 manufacturers terminology, 77
Lung abscess, 391 mixed venous blood gases, 110 112
pneumonia and, 367368 modes of, 77 100
Lung mechanics, mechanical ventilation, monitoring during, 105 136
127 135 physiologic parameters, 106
Lung transplant rejection, insertion point-of-care testing (POC), 108 109
techniques, 296 297 prolonged, 187 207
patient response, 176
Mach effect, 359 pulse oximetry, 112 116
Machine breath, 8791 sublingual PCO2, 116 117
adapted support ventilation, 90 weaning from, 153 178. See also
airway pressure-release ventilation, 90 Weaning, from mechanical
bi-level ventilation, 89 ventilation.
pressure control ventilation, 88 89 Midazolam, in ventilator weaning, 170
pressure-regulated volume control Mixed venous oxygen saturation,
vendilation, 90 measurement of, 111
Index 413
Monitoring, of oxygen therapy, 3 4 humidication, 60

Monitoring response to treatment limitations, 50 51, 54 56
complications, 64 65 mask leaks, 50, 53 54
COPD, 62 nosocomial pneumonia, 43
erosion of nasal cartilage or bone, 65 post-extubation period, 41 46
eye irritation, 65 potential of, 67
failure predictors, 63 pressure support to patient, 60
disease conditions, 63 upper airway obstruction, 42
late failure, 63 ventilators, types of, 55, 56 57
level of consciousness, 63 Nocardia, 292
severity of illness, 63 Non-invasive ventilation, 33 67
time frame for failure requiring Non-cardiogenic pulmonary edema,
intubation, 63 372 374
gastric distension, 65 Non-rebreathing mask, 8
nasal congestion, 65 Non-tuberculosis mycobacteria (NTM),
nosocomial pneumonia, 65 294
patient cooperation, 61 Nosocomial pneumonia, 65, 293
sepsis, 65
sinusitis, 65 Open lung approach, in breath phases,
skin necrosis, 65 86 87
success vs. failure, 61 64 Open surgical tracheostomy
Mycobacterial disease, 294 (OST), 313
Mycobacterium abscessus, 294 Opiates, over-sedation with, 288
Mycobacterium avium complex, 294 Optically coupled direct radiography,
Mycobacterium fortuitum, 294 334
Mycobacterium kansasii, 294 Optimal position of medical devices in,
Mycobacterium malmoense, 294 ICU, 338
Mycobacterium tuberculosis, 294 Orotracheal intubation, complications of,
Mycobacterium xenopi, 294 24
Orotracheal intubation
Naloxone, 288 preparation for, 16 18
Nasal cannula, as oxygen administration technique for, 18 20
devices, 6 Oximetric methodology, to measure
Nasal congestion, 65 mixed venous oxygen saturation,
Nasal O2 cannula, 5 7 111
Nasoenteric tubes, in the ICU, Oxygen administration devices, 5 8
340 342 air-entrainment mask, 9
Nasotracheal vs. orotracheal high-ow devices, 6, 9 10
intubation, 14 mask air-entrainment, 6
Neoplasm, debulking of, insertion mask non-rebreathing, 6
techniques, 299 mask simple, 6, 8
Neuromuscular disease, 50 nasal O2 cannula, 5 7
NICO, rebreathing cycle, 125 126 non-rebreathing mask, 8
NIMV Oxygen analyzers, 3
clinical aspects of, 50 61 galvanic cell analyzers, 3
co-morbidities, 52 53 polarographic, 3
efcacy, 67 zirconium analyzer, 3
helium-oxygen mix, 60 61 Oxygen cost of breathing (OCB), 163
414 Index
Oxygen therapy Pleural effusion, 375 377

airway management and, 1 24 Pneumocystis carinii pneumonia, 50, 293
anaerobic infections, 3 Pneumocytis jiroveci, 293
carbon monoxide poisoning, 3 Pneumomediastinum, 359, 360
fasciitis, 3 Pneumonia, 34, 65, 366 368
haldane effect, 4 complications, 367
hypoxemia, 2, 3 lung abscess, 367 368
indications for, 2 3 types of, 366
monitoring of, 3 4 Pneumonitis, 291
physiologic complications of, 4 5 Pneumopericardium, 359
technical hazards of, 5 Pneumothorax, 346, 359
Oxyhemoglobin dissociation curve, 113 apicolateral, 361
posteromedial, 361
PACS, advantages of, 335 simple, 363
Patient synchrony, 50 51 Polarographic oxygen analyzers, 3
Patient triggering Polyex, 304
ow trigger, in breath phases, 79 Portable antero-posterior view
mechanical ventilation, 79 80 barium, 333
pressure trigger, 79 interpretation of, 336 337
PDT oral contrast, 333
complication rates of, 321 portable chest radiography, 332
esophageal perforation, 323 radiation dose level, 333
OST vs., 324 special views, 333
technique, 317 321 Portable chest radiography, 331 337
Peak airway pressure (PAP), and limitations of, 337
equation of motion, 128 indications of, 332
Penumbra effect, in pulse oximetry, 113 Positive end-expiratory pressure (PEEP),
Percutaneous catheter drainage, 355
390 391 Positive pressure ventilation (PPV), 355
Percutaneous dilatational tracheostomy in relaxed patient, 131
(PDT), 313 Post-extubation period, 41 45
complications of, 322 Pre-bronchoscopy, laboratory tests, 288
in the ICU, contraindications to, 317 Pressure control ventilation, in breath
Percutaneous tracheostomy, 313 325 phases, 81 83
Pericardial tamponade, 337 Pressure trigger, in breath phases, 79
Peripheral venous access, 388 Pressure volume curve
Peripherally inserted central catheters in breath phases, 86
(PICC), ICU, 349 during mechanical ventilation, 135
Perivascular interstitial emphysema Primary cycling mechanisms, in breath
(PIE), 355 phases, 83
PETCO2, causes of increases/decreases, Prolonged mechanical ventilation,
122 187 204
Photocoagulation, laser, 302 chronic obstructive pulmonary disease,
Photodynamic therapy (PDT), insertion (COPD), 189
techniques, 303 denition of, 188 189
Photoevaporation, laser, 302 frequency of, 188
PICCS, 388 history of, 187
Pleural collections, 390 391 in ICU, 190
Index 415
indications for, 189 Radiology, in the intensive care unit,

LTAC or post-acute care 331 392
hospitals, 190 Rapid shallow breathing index (RSBI),
patient outcomes, 191192 161 162
pneumonia, 190 Reservoir O2 cannula (moustache and
site of care for patients with, pendant types), 8
190 191 Respiratory failure
surgical procedure, 190 acute, management of, 33 34
weaning patients, 204 207 airway trauma, 33 34
Propofol, in ventilator weaning, 170 barotrauma, 33 34
Pulmonary angiography, 381 intubation, 33 34
Pulmonary artery catheters nosocomial sinusitis, 33 34
complications of, 257 259 pneumonia, 33 34
controversy re usage, 259 261 tracheal stenosis, 33 34
equipment and techniques, 239 240 Respiratory system compliance (Crs), 133
ICU procedures and, 236 261, Respiratory therapy, duration of, 50 51
349 350 Rigid bronchoscope, insertion
indications/contraindications, techniques, 289
237 239 Ruptured thoracic aorta, 383
insertion of, 240 242
principles of, 261 Secondary cycling mechanisms,
specic clinical scenarios, 250 257 in breath phases, 84
trials of, 261 Sedation
waveforms, 242 250 benzodiazepines, 170 172
Pulmonary disease, diffuse cost implications for ventilated patient,
non-infectious, 296 172 174
Pulmonary edema, 269375 weaning and, 168 169, 170, 171
cardiogenic, 371 375 Sepsis, 65
in ICU, 370 Septic emboli, 381 382
increased permeability, 372 374 causative organisms, 381
non-cardiogenic, 372 374 diagnosis of, 381
pulmonary interstitial Sinusitis, 6
emphysema, 357 Skin necrosis, 65
Pulmonary embolism, 377 381 Spiration, 305
computerized tomography, Spontaneous breath
385 387 pressure support breath, 91 92
management of, 391 392 proportional assist ventilation breath, 93
radiographic ndings, 379 tube compensation breath, 92 93
Pulmonary hemorrhage, 382 volume support breath, 92
Pulmonary infarction, 351, 379 Stanford aortic dissection, 383
Pulmonary interstitial emphysema, 358 Subclavian artery, puncture of, 347
pulmonary edema in, 357 Subclavian venous catheter, venous
Pulmonary mechanics, 127 malposition of, 344
Pulse oximetry Sublingual PCO2, mechanical
limitations of, 112 ventilation, 116 117
impact on patient outcomes, 115 Superior vena cava thrombosis, 348
mechanical ventilation, 112116 Swan Ganz catheters, in the ICU,
PV curve, 134 349 350
416 Index
Tension pneumothorax, 362, 365 Ventilator-dependency

Thermal injuries, insertion techniques, assessment tools to predict outcome,
296 202 204
Thoracentesis, radiologically guided, 390 care factors, 201
Thoracostomy tube, in the ICU, 342 344 Clostridium difcile infection, 200
Time triggering, mechanical ventilation, iatrogenic factors, 199 201
79 mechanical factors, 196 199
TIPS long-term weaning protocol, 169 non-invasive mechanical ventilation,
Tracheal stenosis, 34 as adjunct to weaning, 206
Tracheoesophageal stula, 296, 323, 334 potential causes for, 195
Tracheo-innominate stula, 323 reasons for, 192 202
Tracheostomy respiratory muscle weakness, 205
complications of, 323 spontaneous breathing trials, 204
contraindications, 316 317 standardization protocols, 205
in the ICU, 340 systemic failures, 192 196
patient care, 324 325 tracheotomies, 199
percutaneous, 313 325 weaning from prolonged mechanical
placement, methods for, 314 ventilation, 197
timing of, 315 weaning protocols, 205
translaryngeal intubation, 315 Ventilatory support
ventilator-dependency, 199 denition of, 34
Transbronchial biopsy, 290 efcacy, 34 35
Transbronchial needle aspiration, 290 general guidelines, 35
Translaryngeal intubation, 315 316 rationale for use, 34 35
Translaryngeal tracheostomy method, Volume control ventilation, in breath
314 phases, 81
Transtracheal O2 catheter, 8 Volume-assured pressure support
Transvenous pacing devices, 353 ventilation (VAPS), 90 91
Treatment, monitoring response to, Volume-based capnography, and cardiac
61 67. See also Monitoring output, 125
response of treatment.
Wallstent, 304
Waveforms, pulmonary artery catheters
Ultraex, 304
and, 242 250
Ultrasound-guided insertions, central
Weaning
venous catheters and, 231
agitated patient, 170
Ultrathin, bronchoscopes, 286
analgesics, 170
selection of, 173
VE recovery time, as weaning parameter, best practices, 164
163 intermittent mandatory ventilation
Venous access, in the ICU, 387 389 (IMV), 164
Venous oxygenation, mixed, 110 delirium from, 174
Ventilation, non-invasive, critical care, environment, importance of, 168
33 68 extubation vs., 165
Ventilator breath, ne tuning of, 99 100 from mechanical ventilation,
Ventilator efciency, 5051 153 178
Ventilator modes, 78. See also Breath compliance, rate, oxygen, and
phases. pressure (CROP) index, 162 163
Index 417
criteria of success/failure, 154 160 parameters, 161 163

functional force reserve of preventing pain and discomfort,
respiratory muscles, 154 173
intra-mural tonometry, 163 prolonged ventilation, 167 168
load vs. capacity, 154 155 protocol recommendations, 168
oxygen cost of breathing (OCB), 163 TIPS, 169
parameters, 161 163 protocol-guided, 165 167
pathophysiology of failure, 156 157 recommendations for delirium, 171
rapid shallow breathing index sedation, 168 169, 170
(RSBI), 161 162 benzodiazepines, 170 172
respiratory muscle fatigue, 155 156 sedative reversal, 173 174
reversible factors, 158 160 Weaning failure, causes of, 159
rib cage-abdominal motion, 163 Westermark sign, 379 280
success prediction, 160 164
VE recovery time, 163 Zirconium analyzer, 3

Zab Mohsenifar, Guy W. Soo Hoo Lung Biology in Health & Disease Volume 213 Practical Pulmonary and Critical Care Medicine Respiratory Failure PDF

Uploaded by

Document Information

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

Zab Mohsenifar, Guy W. Soo Hoo Lung Biology in Health & Disease Volume 213 Practical Pulmonary and Critical Care Medicine Respiratory Failure PDF

Uploaded by

Copyright:

Available Formats

Practical Pulmonary

and Critical Care Medicine

1. Immunologic and Infectious Reactions in the Lung,

The opinions expressed in these volumes do not necessarily represent the

Guy W. Soo Hoo

New York London

Library of Congress Cataloging-in-Publication Data

Visit the Taylor & Francis Web site at

Whereas attendance at national or international meetings once offered the latest

on immediate management and diagnosis. For the practicing physician, this is

reached, assisted ventilation should be removed as soon as possible. These basic

Janet Au Division of Pulmonary and Critical Care, Olive View UCLA

C. Matilda Jude Department of Imaging, VA Greater Los Angeles Healthcare

Nader Kamangar Division of Pulmonary and Critical Care, Olive

Hsin-Yi Lee Department of Imaging, VA Greater Los Angeles Healthcare

Michael I. Lewis Division of Pulmonary and Critical Care Medicine,

Zab Mosenifar Division of Pulmonary and Critical Care Medicine,

Michael L. Nevins Division of Pulmonary and Critical Care and Department

Brian Richards Division of Pulmonary and Critical Care Medicine,

Antoinette R. Roth Department of Radiological Science, Geffen School of

Ammar Sakkour UCLA Santa Monica Specialties, Geffen School of

Irawan Susanto UCLA Santa Monica Specialties, Geffen School of Medicine

Introduction Claude Lenfant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

1. Oxygen Therapy and Airway Management . . . . . . . . . . . . . . . 1

2. Non-Invasive Ventilation in Critical Care . . . . . . . . . . . . . . . 33

V. Clinical Aspects of NIMV . . . . 50

3. Modes of Mechanical Ventilation . . . . . . . . . . . . . . . . . . . . . . 77

4. Monitoring During Mechanical Ventilation . . . . . . . . . . . . . 105

5. Weaning from Mechanical Ventilation . . . . . . . . . . . . . . . . . 153

6. Prolonged Mechanical Ventilation . . . . . . . . . . . . . . . . . . . . 187

7. Procedures in the Intensive Care Unit . . . . . . . . . . . . . . . . . 219

9. Percutaneous Tracheostomy . . . . . . . . . . . . . . . . . . . . . . . . . 313

10. Radiology in the Intensive Care Unit . . . . . . . . . . . . . . . . . . 331

STEVE Y. WONG DEAN R. HESS

Mission Hospital, Mission Viejo Department of Respiratory Care

Oxygen administration and airway management are two of the fundamental

II. Oxygen Therapy

care areas via a bulk gas delivery system. By convention, O2 is supplied at a

B. Indications for Oxygen Therapy

Cylinder Flow (L/min)

An unusual, but occasionally effective, use of O2 therapy is the

C. Monitoring Oxygen Therapy

O2 is a paramagnetic gas. A wheatstone bridge O2 analyzer uses the principle of

D. Physiologic Complications of Oxygen Therapy

Experimental evidence implicates the formation of reactive O2 species

E. Technical Hazards of Oxygen Therapy

F. Oxygen Administration Devices

Table 2 Oxygen Delivery Devices for Adult Applications

Nasal 1 6 L/min 24 40% FIO2 is reduced with

Figure 1 Nasal cannula and simple mask.

Table 3 Estimation of FIO2 from a Nasal Cannula

Cannula ow 2 L/min (33 mL/sec)

FIO2 171 mL (O2)/500 mL (tidal volume) 0.34. Similar calculations can

Table 4 Oxygen Delivery Systems Designed to Conserve O2 by Eliminating Waste

either by providing an O2 bolus at the beginning of inspiration or by admini-

Figure 2 Non-rebreathing mask.

Figure 3 The air-entrainment mask and its principle of operation.

FIO2 Minimum O2 Entrainment Total ow

0.24 4 1:25 104

Figure 4 High-ow O2 delivery system.

Figure 5 Patient interfaces used with high-ow O2 delivery systems.

Table 6 Air:O2 Ratios and Determination of Air and O2 Flows

Air:O2 (1.0 2 FIO2)/(FIO2 2 0.21)