Cardiopulmonary Critical Care Compress

Cover
Page a2
This page intentionally left blank.

Page i
CARDIOPULMONARY CRITICAL CARE

Page ii

Page iii
CARDIOPULMONARY CRITICAL CARE

Thomas L.Higgins
Chief, Critical Care Division
Baystate Medical Center, Springfield, MA, USA
Associate Professor of Medicine and Anesthesiology,
Tufts University, School of Medicine, Boston, MA, USA
Jay S.Steingrub
Director, Medical Intensive Care Unit
Baystate Medical Center, Springfield, MA, USA
Associate Professor of Medicine,
Tufts University, School of Medicine, Boston, MA, USA
Robert M.Kacmarek
Director, Respiratory Care
Massachusetts General Hospital, Boston, MA, USA
Associate Professor of Anesthesiology,
Harvard Medical School, Boston, MA, USA
James K.Stoller
Head, Section on Respiratory Therapy
Department of Pulmonary/Critical Care Medicine, The Cleveland Clinic Foundation, Cleveland, OH, USA
Vice Chairman, Division of Medicine Associate Chief of Staff
The Cleveland Clinic Foundation, Cleveland, OH, USA
Professor of Medicine,
CCF Health Science Center of Ohio State University, Columbus, OH, USA
Page iv
© BIOS Scientific Publishers Limited, 2002

First published 2002
This edition published in the Taylor & Francis eLibrary, 2005.
To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.
All rights reserved. No part of this book may be reproduced or transmitted, in any form or by
any means, without permission.
A CIP catalogue record for this book is available from the British Library.
ISBN 020342767X Master ebook ISBN
ISBN 0203446674 (OEB Format)

ISBN 1 859962 37 8 (Print Edition)
BIOS Scientific Publishers Ltd

9 Newtec Place, Magdalen Road, Oxford OX4 1RE, UK
Tel. 144 (0) 1865 726286. Fax 144 (0) 1865 246823
World Wide Web home page: http://www.bios.co.uk/
Distributed exclusively in the United States of America, its dependent territories, Canada, Mexico,
Central and South America, and the Carribean by SpringerVerlag New York Inc., 175 Fifth
Avenue, New York, USA, by arrangement with BIOS Scientific Publishers Ltd., 9 Newtec Place,
Magdalen Road, Oxford OX4 1RE, UK
Important Note from the Publisher
The information contained within this book was obtained by BIOS Scientific Publishers Ltd from
sources believed by us to be reliable. However, while every effort has been made to ensure its
accuracy, no responsibility for loss or injury whatsoever occasioned to any person acting or
refraining from action as a result of information contained herein can be accepted by the
authors or publishers.
The reader should remember that medicine is a constantly evolving science and while the authors
and publishers have ensured that all dosages, applications and practices are based on current
indications, there may be specific practices that differ between communities. You should always
follow the guidelines laid down by the manufacturers of specific products and the relevant
authorities in the country in which you are practising.
Production Editor: Aimie Haylings

Designed and typeset by J&L Composition Ltd, Filey, North Yorkshire, UK
Printed by Cromwell Press, Trowbridge, UK
Page v
Contents
Abbreviations xii
Contributors xv
Introduction xvi
Acknowledgements xviii
1. Respiratory physiology/pulmonary gas exchange 1

Robert M.Kacmarek
Introduction 2
Lung mechanics 2
The law of motion 6
AutoPEEP 7
relashionships 8
Evaluation of hypoxemia 10
2. Cardiac physiology 17
Jay S.Steingrub
Cardiac function 18
Preload 18
Afterload 20
Contractility 21
Heart rate 22
Diastolic function 22
Diagnosis and therapy of acute congestive heart failure 24
Therapy 27
Patient subsets 29
3. Oxygen transport and tissue oxygenation 35
William T.McGee and Paul Jodka
Introduction 36
Calculating oxygen transport 36
Balancing oxygen supply and demand 38
Clinical factors influencing oxygen supply 39
Determinants of oxygen consumption 40
Critical factors influencing cellular oxygen utilization 40
Evaluation of tissue oxygenation 41
Influence of loss of vasoregulation 43
Continuous mixed venous oxygen saturation in clinical medicine 43
Goaldirected therapy with DO2 and VO2 44
Page vi
Summary 46
Essential equations 46
4. Acidbase disorders 49
Ali AlKhafaji, Marcus J.Hampers, Howard L.Corwin
Introduction 50
Basic concepts and definitions 50
Approach to acidbase abnormalities 52
Metabolic acidosis 52
Anion gap metabolic acidosis 52
Normal anion gap metabolic acidosis 56
Respiratory acidosis 58
Metabolic alkalosis 59
Respiratory alkalosis 60
5. Bedside hemodynamic monitoring 63
Jay S.Steingrub
Introduction 64
Central venous monitoring 64
Pulmonary artery catheterization 66
Validity of measurements 70
Intrathoracic pressure changes during spontaneous respiration 72
Positive endexpiratory pressure (PEEP) 73
Ventricular compliance 73
Limitations of pulmonary artery wedge pressure 74
Thermodilution cardiac output 75
Derived hemodynamic parameters 75
Risk versus benefit 78
6. Shock in the intensive care unit 81
Jay S.Steingrub
Introduction 82
Pathophysiology 82
Specific shock syndromes 82
Hypovolemic shock 83
Distributive shock 86
Anaphylactic shock 91
Cardiogenic shock 94
Obstructive shock 96
Shock states 96
7. Fluid resuscitation in the ICU 99
Laurie A.Loiacono
Introduction 100
Definitions 100
Physiology of resuscitation 101
Page vii
Strategy 104
Red blood cells (RBCs) 105
Crystalloid products 105
Cellfree colloids 107
Summary and conclusions 108
Recommendations for fluid resuscitation in the ICU 109
8. Circulatory support 111
Thomas L.Higgins
Introduction 112
The sympathetic nervous system 112
Receptors and signal transduction 114
Monitoring 115
Catecholamines and sympathomimetics 116
Alterations in adrenergic response 121
Bypassing the adrenergic receptor 123
Vasopressin 123
Weaning pharmacologic support 124
Summary 124
9. Hypertensive urgencies and emergencies 127
Thomas L.Higgins
Introduction 128
Definitions 128
Physiology of blood pressure control: MAP=SVR×CO 130
Monitoring considerations 132
Agents to decrease blood pressure 132
Specific clinical considerations 141
Medicationrelated hypertension 144
Conclusions 144
10. Acute myocardial infarction 147
Marc Schweiger
Overview 148
Diagnosis, definitions and triage 148
Initial treatment of acute ischemic syndrome 151
Thrombolytic therapy 155
Primary angioplasty 156
Lipidlowering therapy 157
Complications of myocardial infarction 157
Subsequent treatment strategies 158
Summary 161
11. Cardiac rhythm disturbances 165
Magdy Migeed and Lawrence S.Rosenthal
Introduction 166
ECG analysis and interpretation 166
Page viii
Supraventricular tachycardias 166

Treatment of supraventricular arrhythmias 168
Atrial fibrillation 170
Therapy of AF 171
Wide complex tachycardias 172
Approach to the patient with cardiac arrest 174
Bradyarrhythmias 179
Summary 179
12. Pulmonary embolism in the critically ill patient 183
Jay S.Steingrub
Introduction 184
Risk factors 184
Pathophysiology of pulmonary embolism 186
Differential diagnosis in the ICU 186
Diagnostic tests 187
Treatment 191
Right ventricular function 196
Other clinical interventions 197
Risk reduction 197
13. Modes of mechanical ventilation 201
Robert M.Kacmarek
Introduction 202
Volume versus pressure targeting 202
Traditional modes of ventilation 206
Pressure control inverse ratio ventilation (PCIRV) 209
Airway pressure release ventilation (APRV) 211
Combined modes of ventilation 212
14. The acute respiratory distress syndrome 221
C.Allen Bashour and James K.Stoller
Introduction 222
Historical perspective 222
Pathogenesis 223
Clinical presentation 225
Treatment 225
Prognosis and outcome 231
Conclusions 231
15. Lungprotective ventilation strategies 235
Robert M.Kacmarek
Ventilatorinduced lung injury (VILI) 236
Lungprotective ventilatory strategies (LPVSs) 242
Setting of PEEP 248
Prone positioning 248
Summary 250
Page ix
16. Nosocomial pneumonia 253

Thomas L.Higgins
Introduction 254
Risk factors for developing nosocomial pneumonia 254
Pathogens and routes of entry 255
Diagnosis 255
Antibiotic therapy 258
Choice of antibiotic 260
Prevention of nosocomial pneumonia 264
Antimicrobial resistance 265
Summary 266
17. Obstructive lung disease: asthma and COPD 271
Robert M.Kacmarek
Introduction 272
AutoPEEP 272
Nonventilatory management 272
Intubation 278
Mechanical ventilation 278
Weaning from ventilatory support 281
18. Weaning from mechanical ventilation 287
James K.Stoller and Thomas L.Higgins
Introduction 288
Who requires gradual withdrawal of mechanical ventilation? 288
Sequelae of prolonged intubation and ventilation 289
Overview of the logic of weaning 289
Evaluation of breathing pattern 290
Work of breathing (WOB) 292
Tracheal occlusion pressure 293
Endurance measurements 293
Techniques of weaning 293
Studies comparing weaning modes 295
Nonrespiratory impediments to successful weaning 296
Respiratory impediments to weaning 299
Pharmacologic aids to weaning 300
Weaning failure 301
Longterm prognosis 301
19. Noninvasive positive pressure ventilation (NIPPV) 307
Robert M.Kacmarek
Introduction 308
NIPPV for acute respiratory failure 308
Indication for noninvasive positive pressure 308
When should NIPPV be initiated for acute respiratory failure? 317
Page x
Complications or problems with NIPPV 317

The application of NIPPV 318
Summary 324
20. Postoperative care of the cardiac surgical patient 329
Thomas L.Higgins
Introduction 330
Delivery of the patient to the ICU 330
Respiratory management 330
Cardiovascular support 333
Postoperative complications 336
Discharge criteria 340
Prognosis and outcome 341
Summary 342
21. Postoperative care of the thoracic surgical patient 345
Thomas L.Higgins
Introduction 346
Identifying the highrisk patient 346
Operating room events affecting ICU care 347
Immediate postoperative issues 349
Endotracheal extubation and airway concerns 350
Postoperative fluid management 351
Prolonged ventilator support 352
Specific postoperative complications 355
Pulmonary parenchymal complications 358
Pleural complications 359
Other considerations 360
22. Thoracic trauma 363
Imtiaz A.Munshi
Introduction 364
Anatomy 364
Pathophysiology 364
Initial evaluation—ABCs 365
Injuries to the chest wall 366
Injuries to the bony thorax 367
Injuries to the lung 369
Tracheobronchial injury 374
Blunt aortic injury 375
Blunt cardiac injury 377
23. Techniques of vascular access for invasive hemodynamic monitoring 381
William T.McGee, Jay S.Steingrub and Thomas L.Higgins
Introduction 382
Intraarterial blood pressure monitoring 382
Page xi
Central venous access 384

Femoral venous cannulation 389
Internal jugular vein 390
Subclavian vein cannulation 394
Introducer placement 398
Complications and their prevention 399
Index 405
Page xii
Abbreviations
ABG arterial blood gas
A/C assist/control
ACE angiotensin converting enzyme
ACEI angiotensinconverting enzyme inhibitor
ACLS advanced cardiac life support
ACT activated clotting time
ADH antidiuretic hormone
AF atrial fibrillation
AFl atrial flutter
AG anion gap
ALI acute lung injury
ALV adaptive lung ventilation
AMI acute myocardial infarction
AP accessory pathway
APACHE Acute Physiology and Chronic Health Evaluation
APD atrial premature depolarization
APRV airway pressure release ventilation
APS adaptive pressure support
aPTT activated partial thromboplastin time
AR adrenergic receptor
AR aortic regurgitation
ARDS adult/acute respiratory distress syndrome
ARF acute respiratory failure
AS aortic stenosis
ATC automatic tube compensation
AV atrioventricular
avDO2 arteriovenous O2 content difference
AVNRT atrioventricular nodal reentrant tachycardia
AVRT atrioventricular reentrant tachycardia
BAL bronchoalveolar lavage
BiPAP bilevel positive airway pressure
BOOP bronchiolitis obliterans organizing pneumonia
BUN blood urea nitrogen
CABG coronary artery bypass grafting
CAD coronary artery disease
cAMP 3′, 5′cyclic adenosine monophosphate
CAP communityacquired pneumonia
CAVH continuous arteriovenous hemofiltration
CHF congestive heart failure
CI cardiac index
CNS central nervous system
CO cardiac output
COMT catecholamineomethyl transferase
COP colloid osmotic pressure
COPD chronic obstructive pulmonary disease
CPAP continuous positive airway pressure
CPB cardiopulmonary bypass
CPK creatine phosphokinase
CPR cardiopulmonary resuscitation
CRRT continuous renal replacement therapy
CSM carotid sinus massage
CT computer tomography
CVC central venous catheter
CVP central venous pressure
CVVHD continuous venovenous hemofiltration with dialysis
CXR chest radiograph
DAG diacylglycerol
DCA dichloroacetate
DIC disseminated intravascular coagulation
DOPA dihydroxyphenylethylamine
2, 3DPG 2, 3diphosphoglycerate
DVT deep venous thrombosis
ECCO2R extracorporeal membrane carbon dioxide removal
Page xiii
ECG electrocardiogram/ electrocardigraphy

ECMO extracorporeal membrane oxygenation
EDV enddiastolic blood volume
EKG electrocardiograph
ELISA enzymelinked immune serum assay
EMG electromyogram
ERV expiratory reserve volume
ESBL extended spectrum βlactamase
ESV endsystolic blood volume
ETA endotracheal aspirate FAST focused abdominal sonogram for trauma
FEV1 forced expiratory volume in 1 s
FFP fresh frozen plasma
FiO2 fraction of inspired oxygen
FOB fiberoptic bronchoscopy
FRC functional residual capacity
FVC forced vital capacity
GI gastrointestinal
Gi Ginhibitory (protein)
Gs Gstimulatory (protein)
HAP hospitalacquired pneumonia
HALF hypertonic albuminated fluid
HFJV highfrequency jet ventilation
Hgb hemoglobin
HSS hypertonic saline solution
5HT 5hydroxytryptamine
HTN hypertension
IABC intraaortic balloon counterpulsation
IABP intraaortic balloon pump
IC inspiratory capacity
ICD internal cardiac defibrillator
ICU intensive care unit
IL interleukins
IJV internal jugular vein
IMV intermittent mandatory ventilation
INR international normalized ratio
IP3 inositol 1,4,5triphosphate
IPC intermittent pneumatic compression
IPG impedance plethysmograpy
IRV inspiratory reserve volume
LA left atrial
LAP left atrial pressure
LBBB left bundle branch block
LDUH lowdose unfractionated heparin
LMWH low molecular weight heparin
LPVS lungprotective ventilatory strategy
LR lactated Ringer’s solution
LV left ventricle/ ventricular
LVEDP left ventricular enddiastolic pressure
LVEDV left ventricular enddiastolic blood volume
LVSWI left ventricular stroke work index
MAO monoamine oxidase
MAP mean arterial pressure
MAT multifocal atrial tachycardia
MCT mediumchain triglyceride
MI myocardial infarction
MIC minimal inhibitory concentration
MICU medical intensive care unit
MIP maximal inspiratory pressure
MMV mandatory minute ventilation
MODS multiple organ dysfunction syndrome
MPAP mean pulmonary arterial pressure
MPM Mortality Probability Model
MR mitral regurgitation
MRI magnetic resonance imaging
MRSA methicillinresistant Staphylococcus aureus
MS mitral stenosis
MSOF multiple system organ failure
MV mechanical ventilator
MVV maximal voluntary ventilation
MW molecular weight
NE norepinephrine
NG nasogastric
NIF negative inspiratory force
NIPPV noninvasive positive pressure ventilation
NPV negative predictive value
NS normal saline solution
NSTEMI nonST elevation myocardial infarction
OHS openheart surgery
OR operating room
Page xiv
PA pulmonary artery
PA/C pressure assist/control
PAC pulmonary artery catheterization/catheter
PaCO2 arterial CO2 tension
PACU postanesthesia care units
PAI1 plasminogen activator inhibitor I
PAO pulmonary artery occlusion
PaO2 mixed venous O2 tension
PAOP pulmonary artery occlusion pressure
PAP pulmonary artery pressure
PAWP pulmonary artery wedge pressure
PBS protected brush specimens
PCIRV pressure control inverse ratio ventilation
PCWP pulmonary capillary wedge pressure
PDI phosphodiesterase inhibitor
PE pulmonary embolism
PEA pulseless electrical activity
PEEP positive endexpiratory pressure
PLV partial liquid ventilation
PMC point of maximum compliance change
Ppl intrathoracic/intrapleural pressure
PPV positive predictive value
PRVC pressureregulated volume control
PS pressure support
PSV pressure support ventilation
PVC premature ventricular contraction
PVR pulmonary vascular resistance
q every
RBBB right bundle branch block
RBC red blood cells
rhAPC recombinant activated protein C
RIJV right internal jugular vein
RPF renal plasma flow
RPP rate pressure product
RTA renal tubular acidosis
rtPA recombinant tissue type plasminogen activator
RV residual volume
RV right ventricle/ventricular
RVSWI right ventricular stroke work index
SaO2 arterial O2 saturation
SB sinus bradycardia
SCM sternocleidomastoid muscle
SDD selective digestive decontamination
SI stroke index
SILV synchronous independent lung ventilation
SIMV synchronized intermittent mandatory ventilation
SIRS systemic inflammatory response syndrome
SR sinus rhythm
ST sinus tachycardia
ST surface tension
STEMI ST elevation myocardial infarction
SV stroke volume
SVC superior vena cava
SVR systemic vascular resistance
SVT supraventricular tachycardia
TBW total body water
TCO thermodilution cardiac output
TdP Torsades de Pointes
TEE transesophageal echocardiogram
TFPI tissue factor protein inhibitor
THAM tromethamine
TLC total lung capacity
TNF tumor necrosis factor
TV tidal volume
UAG urinary anion gap
UIP upper inflection point
US compression ultrasonography
VAP ventilatorassociated pneumonia
VAPS volumeassured pressure support
VC vital capacity
VILI ventilatorinduced lung injury
VIP vasoactive infusion port
VO2 O2 consumption
VQ ventilationperfusion (ratio)
VRE vancomycinresistant enterococci
VS volume support
VT ventricular tachycardia
VTE venous thromboembolism
WCT wide complex tachycardias
WOB work of breathing
WPW WolfeParkinsonWhite (pattern)
Page xv
Contributors
Ali AlKhafaji MD
Critical Care Medicine, DartmouthHitchcock Medical Center, Lebanon, New Hampshire, USA
C.Allen Bashour MD FACS FCCP

Department of Cardiothoracic Anesthesiology, Division of Anesthesiology and Critical Care Medicine, The Cleveland Clinic Foundation, Cleveland, Ohio, USA
Howard L.Corwin MD FCCM FACP

Marcus J.Hampers MD
Thomas L.Higgins MD MBA FACP FCCM FACC

Critical Care Division, Baystate Medical Center, Springfield, Massachusetts, USA
Paul Jodka MD
Robert M.Kacmarek RRT PhD

Respiratory Therapy Department, Massachusetts General Hospital, Boston, Massachusetts, USA
Laurie A.Loiacono MD
William T.McGee MD MHA

Magdy Migeed MD
New York Heart Center, 1000 East Genesse St, Syracuse, New York, USA
Imtiaz A.Munshi MD
Trauma Division, Baystate Medical Center, Springfield, Massachusetts, USA
Lawrence S.Rosenthal MD PhD FACC

Division of Cardiology, University of Massachusetts Memorial Medical Center, Worcester, Massachusetts, USA
Marc Schweiger MD FACC

Division of Cardiology, Baystate Medical Center, Springfield, Massachusetts, USA
Jay S.Steingrub MD FACP FCCM

Medical Intensive Care Unit, Baystate Medical Center, Springfield, Massachusetts, USA
James K.Stoller MD FACP FCCM FCCP

Respiratory Therapy Department, Cleveland Clinic Foundation, Cleveland, Ohio, USA
Page xvi
Introduction
Does the world need another critical care textbook? As I look around my office, I count four hefty, comprehensive critical care tomes, five ICU handbooks, and more
than two dozen related texts on mechanical ventilation, nutritional support, arrhythmia management, monitoring and pharmacotherapy. My handheld organizer provides
me with instant antibiotic and drugdosage recommendations, and I can easily search the literature on computers in my office, at home, and at every nursing station.
With all of this information readily at hand, why would anyone want another critical care text?
The answer may be found in the concept of the evolutionary ‘niche’. The audience for this book is not necessarily the expert in critical care, who presumably also has
a wellstocked bookshelf in his or her office. This book is intended for clinicians with intermittent responsibilities for critically ill patients in a world of competing
demands. Our target audience includes the medical student on a finalyear elective, the house officer on a onemonth rotation in the ICU, the nurse or respiratory
therapist seeking more information, and the general internist, surgeon or anesthesiologist who occasionally cares for critically ill patients. Hospitalists (hospitalbased
doctors) may find this book to be particularly useful, especially if their hospital does not support fulltime critical care staff. Our goal was to create a work with more
depth than the typical pocket reference, while avoiding the bulk and expense of a more encyclopedic work. In order to achieve this goal, we have concentrated this
volume on cardiac and pulmonary issues in the ICU. While the intent is for this book to be part of a larger family of books that would, in total, approximate a standard
text this volume also stands on its own. The hope is that it will provide a logical approach and bestpractice suggestions for a variety of common critical care issues,
while remaining small enough to be read covertocover. The authors recognize that controversy exists in many treatment decisions, but have chosen to present a
consensus approach, buttressed by essential references. My feeling is that questions raised at 3 in the morning should be answered succinctly. I’m happier to engage in
a spirited discussion supported by 200 conflicting references when the patient has been stabilized.
Where possible, chapters have been arranged to present a brief overview of the disease process and epidemiology, diagnostic criteria, important differential
diagnostic considerations, and practical, evidencebased advice on patient management. We have only touched on prognostic information, enough to use in discussions
with family members. Tables have been used to facilitate quick reference, and the figures are similar to what we would handdraw on morning rounds to illuminate a
concept. Some chapters are intended to expose background information (Respiratory and Cardiac Physiology, Oxygen Delivery and Utilization, Clinical Shock States,
Respiratory Muscle Function, ARDS) while others are more practically oriented (Pressors and Inotropes, Hypertensive Emergencies, Cardiac Rhythm Disturbances,
Nosocomial Pneumonia, Weaning from Mechanical Ventilation, and Postoperative Care). In looking through other textbooks, we felt that the mechanics of line
placement were seldom detailed, and have included a richly illustrated chapter on Line Placement Techniques.
Page xvii
In order to keep the book to a reasonable size, we have deliberately omitted important aspects of critical care—specifically neurologic and neurosurgical
management, gastrointestinal bleeding, toxicity and poisoning, hematalogic and oncologic management, renal issues including dialysis, and infectious/immunologic
issues. Plans are in the works for companion volumes to address these topics.
Just as the practice of critical care is a multidisciplinary, team effort, so to is the process of bringing a textbook to life. The authors would like to recognize the efforts
of a number of individuals whose contributions were essential to this book. First and foremost, Ms. Suzanne Allen, administrative assistant for Baystate Medical
Center’s Critical Care Division, spent countless hours typing and retyping manuscripts, chased down letters of permission to reprint figures and tables, and kept the
project moving and organized. Mr. Jonathan Gregory initiated the project for BIOS Scientific Publishers Ltd, helped formulate the style of the book and provided
muchneeded encouragement during the battle to complete the manuscripts. Ms. Victoria Oddie and Dr. Katie Deaton, of the Editorial Department at BIOS, helped
tidy up the manuscript and fix the many overlooked details as the book wound its way to production. Aimie Haylings, our Production Editor, gracefully pushed the
book through its final months, and managed to turn indecipherable scrawls and eighthgeneration photocopies into art.
Special thanks is also due to our colleagues: attending physicians, nurses, respiratory therapists, fellows, residents and medical students who asked the right questions
at the right times, and who generously read manuscript drafts and offered valuable suggestions. Our spouses and families deserve thanks for their unconditional love,
patience, wit and support when another impossible deadline loomed. Above all, we thank our patients and their families, from whom we have learned so much to be
applied in the future.
Thomas L.Higgins MD
Page xviii
Acknowledgements
‘To Louis and Mary Higgins, who started me on this journey, and to Suzanne, my traveling companion during the best years’—TLH
‘To Milagros and my son, Oron, who define love and happiness as not a destination, but rather, a method of life’—JSS
‘To my wife, Jan, who has consistently encouraged my academic pursuits and to my children Robert, Julia, Katie, and Callie who make it all worthwhile’—RMK
‘To Terry, for her abiding support and love, and to Jake, who gives purpose, grounding, and love.’—JKS
Page 1
Chapter 1
Respiratory physiology/pulmonary gas exchange
Robert M.Kacmarek, PhD, RRT
Contents
1.1 Introduction
1.2 Lung mechanics
1.3 The law of motion
1.4 AutoPEEP
1.5 relationships
1.6 Evaluation of hypoxemia
Page 2
1.1 Introduction
Managing respiratory dysfunction is a fundamental aspect of critical care medicine. Although respiratory failure is a rare cause of death in the intensive care unit (ICU)1,
the vast majority of patients admitted to the ICU experience respiratory dysfunction2. This introductory chapter focuses on pulmonary mechanics as related to critical
care and pulmonary gas exchange, V/Q relationships, dead space, shunting, and evaluation of hypoxemia and hypercarbia. Such knowledge provides a framework that
clarifies the rationale behind the various approaches to mechanical ventilatory support.
1.2 Lung mechanics

The maximum gascontaining capacity of the lungs (total lung capacity, TLC) is divided into four basic volumes and four capacities comprising the four basic volumes
(Table 1.1 and Figure 1.1). Of these, the two that have the most relevance to critical care are the functional residual capacity (FRC) and the vital capacity (VC). The
FRC is critical to gas exchange; it is an established gas reservoir allowing exchange of oxygen and carbon dioxide on a continual basis during both inspiration and
expiration3. Loss of FRC results in intrapulmonary shunting and oxygenation
Figure 1.1 The normal spirogram with all four lung volumes and capacities illustrated. See Table 1.1 for definition.
deficits. Much of the management of the critically ill mechanically ventilated patients is focused on restoring and maintaining the FRC (Chapter 15).
Vital capacity is critical because it defines the ventilatory reserve of a patient3,4. The closer a patient’s TV is to his or her VC, the less capable the patient is of
responding to ventilatory stress and sustaining spontaneous ventilation. VC is decreased postoperatively, as a result of the prolonged effects of anesthesia. Neuro
muscular/neurologic insult, and changes in the elastic recoil of the lungs and thorax as well as a result of pneumonia, atelectasis or edema also cause decreases in VC.
Restoration of the VC is an essential aspect of recovery from ventilatory failure3.
Table 1.1 Volume subdivisions and capacities of the lung
Residual volume RV Gas remaining in the lung after maximum exhalation

Expiratory reserve volume ERV Gas that can be forcefully exhaled after passive exhalation
Tidal volume VT or TV Gas inspired and expired each breath
Inspiratory reserve volume IRV Gas that can be forcefully inhaled above normal VT
Functional residual capacity FRC RV+ERV
Inspiratory capacity IC VT +IRV
Vital capacity VC ERV+VT +IRV
Total lung capacity TLC RV+ERV+VT +IRV
Page 3
1.2.1 Pulmonary mechanics

Anatomically the lung and the thorax are configured to move in opposite directions (Figure 1.2). The normal elastic recoil of the lungs favors the smallest unstressed
volume while the rib cage and chest wall tend to expand to a volume of about 60–70% of total lung capacity after which the chest wall tends to contract5,6. Movement
of the lungs and thorax is controlled by the muscles of ventilation. Opposing this movement are the elastic properties of the lung and thorax and the resistance of gas
flow into the lung.
1.2.2 Elastic properties of the respiratory system

The lung and thorax (chest wall) can be viewed in terms of a set of springs. The recoil tendency of a spring can be expressed in terms of an unstressed or resting length
and a lengthtension relationship. Similarly, the relevant properties of expandable volumetric structures are the unstressed volume and the relationship between volume
and the transmural pressure required to achieve that volume7. Transmural pressure is expressed as the difference between the pressure inside and that outside the
structure, or distending pressure. For the respiratory system the transmural pressure is equal to alveolar pressure (Palv) minus body surface pressure (PBS). During
spontaneous breathing a negative transmural pressure (Palv–PBS) causes lung volume to become larger (distending pressure), whereas a positive transmural pressure
(recoil pressure) causes lung volume to become smaller. If lung volume is plotted against transmural pressure the classic sigmoid compliance curve of the respiratory
system is established (Figure 1.3). The slope of the linear section of this pressurevolume relationship (P–V curve) is compliance of the total respiratory system (CRS):
(1.1)
1.2.3 Compliance
Compliance is the inverse of elastance and used to estimate the ease of distensibility of the respiratory system. In Figure 1.3, the point where the P–V curve crosses
the volume line represents FRC; to the left of zero on the horizontal axis the transmural pressure is positive, and to the right negative8. It should be noted also that as the
elastic limits of the system are reached (TLC and RV) a greater pressure change is required for any volume change, whereas,
Figure 1.2 The lung and thorax may be conceptualized as two springs opposing the movement of each other, the thorax expanding to
60–70% of TLC and the lung contracting. Reproduced with permission from: Scanlan, C.L., Wilkins, R.L. and Stoller, J.K.
(1999) Egan’s Fundamentals of Respiratory Care, 7th Edn. St, Louis: MosbyYearbook
Page 4
Figure 1.3 The normal pressurevolume relationship for the total respiratory system. As noted beyond FRC (to the right of the vertical
volume line) the P–V relationship is linear until TLC (maximum volume) is approached.
between FRC and roughly 60–70% of TLC, the P–V relationship is generally linear5. Compliance is usually measured on this linear aspect of the curve and is
expressed in l cm–1 H2O; however, in mechanically ventilated patients it is frequently expressed as ml cm–1 H2O. Compliance is dependent on size: the greater the size
of the lungs the greater the compliance. However, compliance divided by lung volume is volume independent and referred to as nondimensional specific compliance5.
At FRC the transmural pressure is zero. It is at this level that the opposing forces of the lung and thorax are equal and opposite. The elastic properties of the lung and
thorax can be independently determined by measuring esophageal pressure (a reflection of pleural pressure, PPL) and comparing it with alveolar pressure8. During
spontaneous breathing esophageal pressure is reflective of the elastic recoil of the lungs and is used to calculate lung compliance (CL):
(1.2)
To calculate thoracic compliance (CTh) the effect of the ventilatory muscles must be eliminated (using paralysis or heavy sedation) and lung volume provided by
mechanical ventilation. In this setting esophageal pressure represents the elastic recoil of the thorax:
(1.3)
Change in bladder pressure can be used as a substitute for esophageal pressure for the calculation of thoracic compliance9. Bladder pressure has been clearly shown to
be reflective of abdominal/thoracic recoil pressure in critically ill patients receiving controlled mechanical ventilation. Compliance of the total respiratory system, the
lungs and the thorax are related by the following equation:
(1.4)
Figure 1.4 plots the compliance curves of the respiratory system, the lungs, and the thorax. As indicated, the thoracic and lung compliance curves cross the zero recoil
pressure line at different lung volume levels. Any change in either lung or thoracic compliance affects the compliance of the total respiratory system.
1.2.4 Surfactant and effect of surface tension

At the surface of any liquid the intermolecular attraction of the molecules in the liquid establishes a barrier or force, surface tension (ST), preventing disruption at the
surface. Surface tension established at the airliquid interface
Figure 1.4 The compliance curves of the thorax, lungs, and total respiratory system.
Page 5
within the alveoli is an important component of the elastic properties of lung. As noted in Figure 1.5 the pressurevolume relationship of the lung during inspiration and
expiration is different dependent upon whether the lung is fluid or air filled10. The fluidfilled lung can be distended with much less pressure than an airfilled lung. The
airfilled lung also demonstrates a marked hysteresis, that is, the inflation and deflation P–V relationship is different5,10. The reason for this difference is the presence of
surfactant, the phospholipid secreted by type 2 alveolar cells. Surfactant reduces surface tension and stabilizes lung volume. Without the presence of surfactant at end
exhalation the alveoli would collapse. The relationship between ST and the pressure within a sphere is defined by LaPlace’s Law:
(1.5)
where the relationship of the pressure (P) within a fluid sphere to the tension (ST) created at the surface of the sphere is dependent upon the radius (r) of the sphere5.
The overall
Figure 1.5 Inflation and deflation pressurevolume curves of the normal airfilled lung and a fluidfilled lung. See text for discussion.
quantity of surfactant within the lung during health is constant, decreasing surface tension during exhalation and increasing surface tension during inspiration. As a result,
any disruption in surfactant production results in an increase in elastic recoil of the lung or a decrease in lung compliance, decreasing the compliance of the total
respiratory system and increasing the likelihood of atelectasis.
1.2.5 Flow resistance

Airflow between atmosphere and the distal lung is dependent upon driving pressure and the resistance to gas flow. Resistance (R) to flow in any system is dependent
upon the structure of the system and is determined by the pressure gradient (ΔP) needed to maintain flow and the flow rate (V):
(1.6)
Resistance is specifically affected by a number of physical factors of the fluid and system: density (D) and viscosity (n) of the gas flowing, and radius (r) and length (l) of
the airway5. These factors relationship to resistance are illustrated in a modification of Poiseuille’s Law11:
(1.7)
where B is a constant, and by the Reynold’s Number (RN)5:
(1.8)
Specifically, as the radius of the airway decreases by 50% the resistance to flow increases 16fold and the less dense and more viscous a gas the more likely it is to
flow in a laminar manner11. Laminar and turbulent flow are very different. Laminar flow is smooth and uniform flow with the molecules in the center of the gas column
proceeding more rapidly than those at the periphery because they do not encounter the resistance of the sides of the vessel. Turbulent flow is rough and tumbling flow
where all molecules in the system encounter the sides of the container. The concern between these different flow patterns is
Page 6
the pressure to maintain flow. During laminar flow, R is directly related to the flow, whereas with turbulent flow R is directly related to the flow rate squared. As a result,
a much greater pressure gradient is required to maintain turbulent than laminar flow. The use of heliumoxygen mixtures is directly based on these physical properties.
The lowerdensity, higherviscosity helium generally results in less driving pressure to maintain the same flow12.
1.2.6 Normal and abnormal values

Table 1.2 lists normal values in adults for resistance and compliance5 and levels generally present during different clinical settings in the ICU. However, these are
estimates and are affected by the size of the patient. A 5 ft tall, 95 pound woman and 6 ft 6 in tall, 220 pound man will have very different normal compliance and
resistance.
1.2.7 Measurements during mechanical ventilation

Bedside estimations of both compliance and resistance in mechanically ventilated patients are possible. However, precision in these measurements is generally lacking.
The specific instrumentation available in the Pulmonary Function Laboratory is generally not available in the ICU. Ideally, calculation of compliance requires a precise
measurement of the static volume of gas maintained in the lung and the pressure required to maintain it. With most ventilators, the tidal volume displayed does not take
into consideration volume lost to gas compression in the ventilator tubing and humidifier13. As a result, delivered tidal volume must be corrected for system specific
compressed volume (CV), which with most ventilating circuits equals about 1–2 ml cm–1 H2O (PPLAT)14. This correction can be avoided if actual exhaled tidal volume
(Vet) is measured at the artificial airway. Static inspiratory plateau pressure PPLAT must be corrected for total positive endexpiratory pressure (PEEP) (applied plus
auto PEEP) to determine the actual pressure maintaining the static lung volume. Thus an estimate of static total respiratory system compliance (CRS) can be made using
the following equation:
(1.9)
where PEEPA is applied PEEP and PEEPI is autoPEEP.

Compliance of the thorax can be determined in a similar manner except that static end expiratory esophageal (eso) or bladder pressure replaces airway pressure:
(1.10)
To estimate airway resistance the pressure required to establish a constant flow is generally determined. This requires volume ventilation with a square wave flow be
delivered. As noted in Figure 1.6 the difference between peak and plateau pressure when VT is delivered with a constant gas flow establishes the pressure gradient
required to maintain flow:
(1.11)
1.3 The law of motion

From the previous discussion it becomes apparent that the pressure required to ventilate either spontaneously or mechanically is affected by compliance, resistance,
volume and flow. These variables are related by the law of motion, which states that the pressure needed to ventilate a patient is equal to volume (V) divided by
compliance plus resistance times flow:
(1.12)
Peak ventilating pressure is a result of the combined effects of resistance and compliance whereas PPLAT is solely a reflection of the pressure to overcome compliance
and the difference between P1P and PPLAT the pressure to overcome airways resistance. A change in compliance would change the slope of the airway pressure curve
from zero pressure to point A (Figure 1.6) and the magnitude of both A and B, whereas a change in resistance would
Page 7
Table 1.2 Normal and abnormal compliance and airway resistance levels
Total respiratory systema (RS) 0.08–0.1 l cm–1 H2O
Lunga 0.16–0.2 l cm–1 H2O
Thoraxa 0.16–0.2 l cm–1 H20

Specific compliancea 0.08 (no units)
RS medically ventilated 60–80 ml cm–1 H2O
RS ARDS <40 ml cm–1 H20
Airway resistancea (R) 0.6–2.4 cm H2O l–1 s–1 at a flow of 0.5 l s–1
R—mechanically ventilated normal <5 cm H2O difference PIP—P PLAT, constant flow
R—mechanically ventilated marked obstruction >10 cm H2O difference PIP—P PLAT, constant flow
a
Normal values from Comroe, J.H. (1974) Physiology of Respiration, 2nd Edn. Year Book Medical Publishers, Chicago, IL. PIP, peak inspiratory pressure; PPLAT, end inspiratory plateau
pressure.
Figure 1.6 An inspiratory positive pressure waveform with an end inspiratory pause. Gas flow is delivered with a constant flow. A, peak
airway pressure; B, end inspiratory plateau pressure.
change the magnitude of the difference between A and B and the magnitude of A. Similarly, a change in volume or flow would alter this relationship. It is essential to
realize that the pressure generated during ventilation is solely affected by these factors regardless of whether pressure or volume ventilation is used or assisted or
controlled ventilation is provided. The greater the volume, flow or airway resistance and the lower the compliance of the system the greater the transmural pressure
required to ventilate the system. In controlled ventilation the ventilator provides all of the effort (pressure) to ventilate; during spontaneous ventilation the patient
provides all of the effort. During all forms of assisted ventilation the effort is shared between the patient and the ventilator. A frequent observation during the transition
from assisted to controlled ventilation is an increase in airway pressure. The reason for this increase is that the negative intrathoracic pressure established during
assisted ventilation must now be provided by the mechanical ventilator. The actual pressure required to ventilate has not increased, it simply has shifted from the patient
to the ventilator.
1.4 AutoPEEP
AutoPEEP, also referred to as intrinsic PEEP, is a result of incomplete emptying of the lung at end expiration or air trapping. The term unidentified PEEP is also used
to refer to autoPEEP because autoPEEP is not identified on the pressure manometer of the ventilator unless an end expiratory hold is imposed15. AutoPEEP is
established as a result of the intrinsic properties of the lung, not as a result of setting PEEP on the ventilator. As illustrated in Figure 1.7, at normal end exhalation
alveolar pressure is greater than central airway pressure when autoPEEP is present as a result of dynamic flow limitation. But when an end expiratory hold is
established, alveolar, central airway and ventilator circuit pressures equilibrate, indicating the average ‘system’ end
Page 8
Figure 1.7 Relationship between alveolar, central airway and ventilator circuit pressure during (a) normal conditions and in the presence
of severe dynamic airway obstruction, (b) with the expiratory port open, and (c) with the expiratory port occluded. AutoPEEP
level during occlusion can be read on the pressure manometer . From: Pepe, P.E. and Marini, J.J. (1982) Occult positive end
expiratory pressure in mechanically ventilated patients with airflow obstruction: the autoPEEP effect. Reproduced with
permission from the American Review of Respiratory Disease
expiratory pressure. In addition to dynamic airflow limitation, large minute ventilation and inadequate expiratory time can cause autoPEEP.
As already illustrated, autoPEEP does affect ventilatory mechanics and must be considered during all measurements. Failure to do so results in a misleading
interpretation of P–V relationships (lower compliance). During assisted ventilation, autoPEEP also increases patient effort/work of breathing.
1.5 relationships
Ventilation ( relationship can result in hypoxemia or hypercarbia.
1.5.1 Ventilation
Distribution of ventilation is gravity dependent, a greater percentage of tidal volume is distributed to gravity dependent lung than nongravity dependent lung (Figure
1.8)16. As the lung is contained within the visceral pleural and separated from the parietal pleural by a thin layer of fluid, gravity can dramatically affect local pleural
pressure and as a result local FRC and distribution of VT. In general, lung units that are nongravity dependent have a more negative pleural pressure, greater FRC and
lower compliance than gravity dependent units. A lower percentage of tidal volume distributes to nongravity dependent than to gravity dependent units. That is, the
majority of the tidal volume distributes to gravity,
Page 9
Figure 1.8 Measurement of regional differences in ventilation with radioactive xenon. Ventilation decreases from the apex to the base
of the lung in the standing position. From: West, J. (1990) Respiratory Physiology: The Essentials, 4th Edn. Reproduced with
permission from Lippincott, Williams & Wilkins.
whereas the majority of FRC is distributed to nongravity dependent lung. The largest proportion of ventilation is distributed to gravity regardless of position.
1.5.2 Perfusion
Distribution of perfusion is similar to that of ventilation, although at the extremes of the lung (base and apex) the differences are greater than those observed with
ventilation17. John West has proposed a threezone model of pulmonary perfusion (Figure 1.9)17. In zone 1, the apex of the lung, alveolar pressure is generally greater
than arterial, which is greater than venous pressure. This results in essentially no pulmonary capillary flow during rest. In zone 2, the middle portion of the lung, perfusion
is intermittent during rest, as arterial pressure is greater than alveolar but alveolar pressure is greater than venous. In zone 3, perfusion is continuous (during inspiration
and expiration), as arterial pressure and venous pressure are greater than alveolar pressure.
Figure 1.9 John West's model of uneven distribution of blood flow, Pulmonary blood flow is based on vascular and alveolar pressures
affecting actual blood flow to capillaries. From: West, J.B., Dollery, CT. and Naimark, A. (1964) Distribution of blood flow in
isolated lung: relation to vascular and alveolar pressure. Reproduced with permission from The Journal of Applied
Physiology.
Page 10
West’s model also dictates that the gravitational distribution of perfusion is maintained regardless of position. However, recent animal data18,19 cast some doubt on
redistribution of ventilation based on gravity in other than the erect or supine position. However, no data are currently available on patients to contradict West’s theory
of matching. In addition, recent data demonstrating the efficacy of prone positioning on oxygenation in about 70% of adult respiratory distress syndrome (ARDS)
patients studied would lend support to West’s model20,21.
From the above discussion, it is understood that there are specific relationships in health between and that establish the normal variability between
relationships from the apex to the base of the lungs. The base is normally better perfused than the apex and the apex is normally better ventilated. At the apex
approaches infinity because of the lack of perfusion, whereas the is greater than about 3.0 in the base, with the overall maintained in health at 0.822.
1.5.3 Shunting and dead space

A other than 1.0 indicates a maldistribution of either ventilation or perfusion. Numerous pathophysiologic conditions can alter relationships (Table 1.3). If the
is greater than 1.0 dead space is present, whereas a less than 1.0 indicates the presence of intrapulmonary shunting. Figure 1.10 illustrates the extremes of
relationships23. A normal unit is one in which ventilation and perfusion are perfectly matched, whereas a silent unit has complete absence of either ventilation or
perfusion. A dead space unit occurs if ventilation is present but perfusion is absent, and a shunt unit occurs when ventilation is absent but perfusion is present. In
addition to these extreme conditions, a dead space effect is established whenever ventilation is in excess of perfusion or whenever the relationship exceeds 1.0. A
shunt effect occurs whenever the relationship is less than 1.0.
Table 1.3 Cause of altered relationships
Decreased (shunt)
• Anatomic distribution of venous blood to the left heart
• Arterial or ventricular septal defects
• Arteriovenous anastomosis
• Severe liver disease
• Vascular pulmonary tumors
• Congenital cardiac anomalies
• Atelectasis
• Consolidating pneumonia
• Pneumothorax
• Pleural effusion
• Retained secretions
• Bronchospasm
• Partial or complete airway obstruction
• Regional increases in fibrotic lung tissue
• Decreased tidal volume
• Mucosal edema
• Any condition in which ventilation is less than perfusion
Increased (dead space)
• Positive pressure ventilation
• PEEP/continuous positive airway pressure (CPAP)
• Mechanical dead space
• Pulmonary emboli
• Emphysema
• ARDS
• Rapid shallow breathing
• Decreased cardiac output
• Alveolar septal wall destruction
• Any condition in which ventilation is greater than perfusion
1.6 Evaluation of hypoxemia

Hypoxemia can be caused by a number of pathophysiologic abnormalities, most commonly alveolar hypoventilation, mismatch (low and true shunting.
Diffusion abnormalities do cause hypoxemia as a result of delays in passage of O2 molecules across the alveolar capillary membrane. However, as noted in Figure
1.11, hemoglobin molecules become fully saturated with oxygen in about 0.25 s whereas the normal transit time for hemoglobin through the pulmonary capillaries is
about 1.0 s16. As a result, the impact of a diffusion abnormality is usually limited. In
Page 11
Figure 1.10 The theoretical respiratory unit (a) normal ventilation, normal perfusion; (b) normalmal ventilation, no perfusion; (c) no
ventilation, normal perfusion; (d) no ventilation, no perfusion. From: Shapiro, B.A., Harrison, R.A. and Walton, Jr. (1982)
Clinical Application of Arterial Blood Gases, 3rd Edn. Year Book Medical Publishers.
patients with severe interstitial edema or gross pulmonary edema diffusion defects do contribute to hypoxemia but the precise level of contribution is difficult to
establish.
Alveolar hypoventilation always results in hypoxemia if severe. As noted from the alveolar gas equation,
(1.13)
where PB is the barometric pressure and R is the respiratory quotient and increases in alveolar PCO2 result in a decrease in alveolar PO2.
Figure 1.11 Oxygen equilibration time across the alveolarcapillary membrane. (a) Time course when alveolar PO2 is normal. (b) Time
course when alveolar PO2 is markedly decreased. From; West, J. (1990) Respiratory Physiology: The Essentials, 4th Edn.
Reproduced with permission from Lippincott, Williams & Wilkins.
The last part of Equation (1.13) is frequently simplified to
(1.14)
In other words, for approximately every 1 mmHg PA CO2 increase the alveolar PO2 decreases by 1.25 mmHg (R=0.8). Alveolar hypo or hyperventilation can
markedly alter the alveolar PO2 and thus the arterial PO2 (PaO2). Of all of the causes of hypoxemia this is the easiest to correct. Reestablishment of normal alveolar
ventilation returns PaO2 to its normal range.
Page 12
1.6.1 mismatch
A areas result in the same level of hypoxemia as observed in a similar level of true shunting,
Figure 1.12 Hypoxemia as a result of low alveolus is raised enough to make its capillary PO2 nearly normal, Note that the average
P aO2 of the mixture of blood from both capillaries is a result of the average of the two O2 contents, not PO2s. From: Pierson,
D.J. J. (1992) Respiratory failure: introduction and overview. In: Foundations of Respiratory Care (eds D.J, Pierson and
R.M.Kacmarek). Reproduced with permission from Churchill Livingstone,
differentiation between the two is important because of their varied response to hypoxemia24. As illustrated in Figure 1.12, the hypoxemia as a result of low areas,
Figure 1.13 Hypoxemia as a result of true shunting and the effect of supplemental oxygen. (a) Blood leaving the normal alveoli is fully
saturated; however, no oxygen enters the blood of the capillary associated with the unventilated alveoli. (b) The addition of
40% oxygen has essentially no effect on oxygenation, as little O2 can be added to the blood perf using the normal alveolus and
no oxygen can be added to the blood passing the unventilated alveolus. From: Pierson, D.J. (1992) Respiratory failure:
introduction and overview. in: Foundations of Respiratory Care (eds D.J.Pierson and R.M. Kacmarek), Reproduced with
permission from Churchill Livingstone.
Page 13
the airway is partially open even though the alveoli are grossly underventilated. As a result, the administration of an increased FIO2 allows the alveolar PO2 in the low
area to increase sufficiently to reverse the hypoxemic effect. It is for this reason that the hypoxemia in many chronic obstructive pulmonary disease (COPD)
patients can be easily reversed by low flow oxygen via a nasal cannula. The administration of oxygen confirms the cause of hypoxemia, for small increases in FIO2
generally do not improve PaO2 if true shunting is the cause of hypoxemia.
Frequently the blood flow to an area of low is also reduced as a result of hypoxic vasoconstriction. However, the reduction in blood flow often is inadequate to
match the reduction in ventilation. Certain vasoactive drugs reduce hypoxic vasoconstriction and cause significant worsening of arterial oxygenation. Inhaled nitric oxide
in COPD patients has been shown to increase hypoxemia, as have other systemic vasodilators (nitroprusside).
When low area may be greater than the replenishment of oxygen by ventilation, the absence of nitrogen can lead to instability. If the size of the alveoli continually
decreases and nitrogen is absent, closing volume will eventually be reached and atelectasis develops.
In critically ill patients, low areas are generally considered the primary cause of hypoxemia.
1.6.2 True shunting

An intrapulmonary shunt is referred to as a right to left shunt, because essentially blood leaving the right heart enters the left heart without participating in gas exchange,
or as defined above, an area of lung perfused but without ventilation. As noted in Figure 1.13 this results in venous blood mixing with oxygena ted blood and a
dramatic decrease in the PaO2 because the amount of oxygen carried by hemoglobin (1.34 ml O2 g−1 Hb at 100% saturation) is much larger than that dissolved in the
plasma (0.003 ml O2 ml−1 blood per mmHg PO2). Additional O2 can be added to plasma only in wellventilated lung, as hemoglobin is usually saturated or near
saturated in normal lung units. As a result, little O2 is available to increase O2 content in blood from shunted areas.
The majority of shunting is a result of unventilated lung units but high pulmonary artery pressure can cause the foramen ovale to open directly, shunting blood from the
right to left heart. In addition, congenital cardiac anomalies and the emptying of a few veins (pleural, bronchial, thebesian veins) into the left heart and the development
of A–V anastomoses from severe liver disease also increase the shunt fraction. As illustrated in Figure 1.13, simple oxygen therapy is not effective in reversing the
hypoxemia associated with true shunting. To reverse the hypoxemic effect of true shunting the cause must be reversed, by the resolution of pneumonia, the expansion
atelectasis areas or the closing of a patent foramen ovale. The specific ventilator therapy designed to address true shunting is PEEP. As discussed in Chapter 15, PEEP
keeps unstable lung units open and improves oxygenation.
1.6.3. Dead space

Dead space is caused by a ventilation/perfusion mismatch greater than 1.0. Anatomically, as the conducting airways and the upper respiratory tract do not participate in
gas exchange, the volume of gas in these areas is considered dead space. In the healthy adult, this volume About 25–40% of the tidal volume normally does not
participate in gas exchange. Any equals about 1 ml kg–1 of ideal body weight. pathophysiologic condition that results in increased ventilation as compared with
perfusion causes an increase in dead space (Table 1.3). The most obvious cause of an increase in dead space is pulmonary emboli, preventing
Page 14
all blood flow to an area. However, the common disease states seen in the ICU (i.e. COPD, ARDS) also result in increased dead space (COPD as a result of the loss
of the overall pulmonary capillary bed and ARDS because of occlusion of pulmonary capillaries). In addition, any time cardiac output decreases, dead space increases.
Of primary concern in the ICU is application of positive pressure ventilation and PEEP to patients with inadequate vascular tone or volume, or inadequate cardiac
output.
An often overlooked cause of increased dead space and as a result increased PaCO2 is autoPEEP. Frequently, autoPEEP is not recognized because it is not
observed on the ventilator unless an end expiratory pause is activated. However, simple observation of the expiratory flow waveforms can identify the presence but not
the magnitude of autoPEEP (Figure 1.14). Whenever end expiratory flow does not return to zero autoPEEP is present. A common reason for CO2 elevation in the
asthmatic, the post (or during) cardiac arrest patient, the postoperative patient, or the patient with COPD is overly aggressive delivery of mechanical ventilation. In
these settings a decrease in minute ventilation resulting in a decrease in autoPEEP may improve PaCO2 especially if cardiac output is depressed by the autoPEEP. In
all settings an increase in dead space results in an increase in PaCO2 if the minute ventilation remains constant.
1.6.4 Estimation of shunt and dead space

The classic method of measuring total shunt effect is the shunt equation:
(1.15)
where QS/QT is total shunt per cent, CaO2 arterial O2 content, mixed venous O2 content and CcO2 capillary Cv O2 content in wellventilated and perfused lung units.
However, to use this equation, the PaO2 must be greater than 150 mmHg to insure 100% saturation of hemoglobin in capillary blood, and access to blood from the
pulmonary artery must be available to calculate mixed venous O2 content. As a result, clinically this approach is rarely used. As noted in Table 1.4, three other
‘estimates’ of shunting are widely used. Of these, the PaO2/PA O2 ratio is the least affected by anything but oxygenation status. However, clinically and experimentally
the PaO2/FIO2 is the most commonly used index.
Estimation of dead space to tidal volume ratio (VD/VT) uses a modification of the Bohr equation:
(1.16)
where PECO2 is equal to the mixed exhaled PCO2. However, as noted above, whenever the minute ventilation is increased without an associated decrease in PaCO2,
dead space must be elevated.
Figure 1.14 Pressure and flow curves from a COPD patient with autoPEEP. Note the expiratory flow, an initial spike then marked
decrease and expiration ended before flow reached zero. Whenever flow fails to return to zero autoPEEP is present.
Page 15
Table 1.4 Estimation of intrapulmonary shunting and dead space
Index Normal value Abnormal value Limitation

QS/QT 2–5% >10% Invasive
P AO2–P aO2 7–14 mmHg room air 100–150 at FIO2 1.0 Varies with FIO2
31–56 mmHg 100% O2
PaO2/PAO2 >0.75 <0.75 –
PaO2/FIO2 450–500 <300–350 Varies with PaCO2 and FIO2
VD/VT 25–40% >40% –
References
1. Montgomery, A.B., Stager, M.A., Carrico, CJ. and Hudson, L.D. (1985) Causes of mortality in patients with adult respiratory distress syndrome. Am. Rev.
Respir. Dis. 132: 485–489.
2. Knaus, W.A., Draper, E.A., Wagner, D.P. and Zimmerman, J.E. (1985) Prognosis in acute organsystem failure. Ann. Surg. 202:685–693.
3. Alex, C.G. and Tobin, M.J. (1995) Assessment of pulmonary function in critically ill patients. In: Textbook of Critical Care, 3rd Edn (eds S.M.Shoemaker,
A.Grenvik, P.R.Holbrook and W.C. Shoemaker). W.B.Saunders, Philadelphia, PA, pp. 649–658.
4. Chevrolet, J.C. and Deleamont, P. (1991) Repeated vital capacity measurements as predictive parameters for mechanical ventilation need and weaning success in
the GuillainBarré syndrome. Am. Rev. Respir. Dis. 144:814–819.
5. Comroe, J.H. (1974) Mechanical factors in breathing. Physiology of Respiration, 2nd Edn. Yearbook Medical Publishers, Chicago, IL, pp. 94–141.
6. Scanlan, C.L., Wilkins, R.L. and Stoller, J.K. (1999) Egan’s Fundamentals of Respiratory Care, 7th Edn. St. Louis: MosbyYearbook.
7. Culver, B.H. (1992) Mechanics of ventilation. In: Foundations of Respiratory Care (eds D.J. Pierson and R.M.Kacmarek). Churchill Livingstone, New York, pp.
83–92.
8. Murry, J.F. (1974) Ventilation. The Normal Lung: The Basis for Diagnosis and Treatment of Pulmonary Disease. W.B.Saunders, New York, pp. 77–112.
9. Gattinoni, L., Pelosi, P., Suter, P.M., Pedoto, A., Vercesi, P. and Lissoni, A. (1998) Acute respiratory distress syndrome caused by pulmonary and
extrapulmonary disease: different syndromes? Am. J. Respir. Crit. Care Med. 158: 3–11.
10. Clements, J.A. and Tierney, D.F. (1964) Alveolar instability associated with altered surface tension. In: Handbook of Physiology, Section 3, Respiration, Vol. II
(eds W.O.Fenn and H.Rahr). American Physiological Society, Washington, DC, pp. 1565–1583.
11. Slonim, N.B. and Hamilton, L.H. (1971) Resistance to breathing. Respiratory Physiology. C.V.Mosby, St. Louis, MO, pp. 16–65.
12. Manthous, C.A., Morgan, S., Pohlman, A. and Hall, J.B. (1997) Heliox in the treatment of airflow obstruction: a critical review of the literature. Respir. Care
42:1034–1042.
13. Kacmarek, R.M. and Hess, D. (1994) Basic principles of ventilator machinery. In: Principles and Practice of Mechanical Ventilation (ed. M.J. Tobin). McGraw
Hill, New York, pp. 65–110.
14. Kacmarek, R.M., Hess, D. and Stroller, J.K. (1993) Airway pressure, flow and volume waveforms and long mechanisms during mechanical ventilation.
Monitoring in Respiratory Care. Mosby Year Book, St. Louis, MO, pp. 497–544.
15. Pepe, P.E. and Marini, J.J. (1982) Occult positive end expiratory pressure in mechanically ventilated patients with airflow obstruction: the autoPEEP effect. Am.
Rev. Respir. Dis. 126: 166–170.
16. West, J. (1990) Ventilation. Respiratory Physiology: The Essentials, 4th Edn. Williams and Wilkins, Baltimore, MD, pp. 11–20.
17. West, J.B., Dollery, C.T. and Naimark, A. (1964) Distribution of blood flow in isolated lung: relation to vascular and alveolar pressure. J. Appl. Physiol 19:713–
724.
18. Glenny, R.W., Lamm, W.J., Albert, R.K. and Robertson, H.T. (1991) Gravity is a minor
Page 16
determinant of pulmonary blood flow distribution. J. Appl. Phys. 71:620–629.

19. Glenny, R.W., Bernard, S., Robertson, H.T. and Hlastala, M.P. (1999) Gravity is an important but secondary determinant of regional pulmonary blood flow in
upright primates. J. Appl Phys. 86:623–632.
20. Jolliet, P., Bulpa, P., Ritz, M., Riccou, B., Lopez, J. and Chevrolet, J.C. (1997) Additive beneficial effects of the prone position, nitric oxide, and almitrive
bismesylate on gas exchange and oxygen transport in ARDS. Crit. Care Med. 25:786–794.
21. Chatte, G., Sab, J.M., Dubois, J.M., Sirodot, M., Gaussorgues, P. and Robert, D. (1997) Prone position in mechanically ventilated patients with severe acute
respiratory failure. Am. J. Respir. Crit. Care 155:473–480.
22. West, J.B. (1985) Ventilation/Blood Flow and Gas Exchange, 4th Edn. Blackwell, Oxford.
23. Shapiro, B.A., Harrison, R.A. and Walton, J.R. (1982) The Physiology of external respiration. Clinical Application of Arterial Blood Gases, 3rd Edn. Year
Book Medical Publishers, Chicago, IL, pp. 55–64.
24. Pierson, D.J. (1992) Respiratory failure: introduction and overview. In: Foundations of Respiratory Care (eds D.J.Pierson and R.M. Kacmarek). Churchill
Livingstone, New York, pp. 295–302.
Page 17
Chapter 2
Cardiac physiology
Jay S.Steingrub, MD
Contents
2.1 Cardiac function

2.2 Preload
2.3 Afterload
2.4 Contractility
2.5 Heart rate
2.6 Diastolic function
2.7 Diagnosis and therapy of acute congestive heart failure
2.8 Therapy
2.9 Patient subsets
Page 18
2.1 Cardiac function

Cardiac performance is regulated by the integration of four major determinants:
• preload: an estimate of ventricular enddiastolic volume, highly predictive of systolic function;
• afterload: the degree of ventricular wall tension;
• inotropy: the contractility of the ventricular muscle;
• heart rate.
Cardiac muscle has the inherent ability to adapt to varying demands and conditions. This adaptation to alterations and length of resting myocardial muscle fibers is
known as the FrankStarling’s mechanism or Starling’s law1. Ventricular contraction is facilitated by increasing fiber length during diastole, thereby permitting the
ventricle to increase, up to a limit, the quantity of blood ejected from the ventricle with each beat (stroke volume). However, beyond the limits of optimal fiber length,
contraction is impaired. The pressure in the atria is directly related to the filling of the ventricle in diastole. With normal hearts, as atrial pressure increases in response to
venous return, cardiac output increases. The calculation of cardiac output can be determined by multiplying the stroke volume by the heart rate. The term ‘cardiac
index’ is used, and refers to cardiac output divided by body surface area expressed in square meters (m2). Indexing cardiac output permits us to compare values
between large and small patients.
2.2 Preload
Preload refers to the initial muscle fiber length or stretch on the cardiac muscle before ventricular contraction. Increases in preload are associated with increases in both
the extent and velocity of muscle fiber shortening, which combine to produce an increase in stroke volume. In the normal heart, preload corresponds to enddiastolic
blood volume (EDV) in the ventricles, just before ventricular systole, and is directly related to venous return to the heart during diastole. As pressure is directly related
to blood volume, diastolic pressure is a surrogate measure of diastolic volume. There is a direct relationship between preload and the force of myocardial contraction
and stroke volume. Preload may be described by both right and left ventricular enddiastolic ventricular filling pressures and is clinically assessed by measurements of
both right and left atrial pressure or pulmonary capillary wedge pressure (PCWP). The terms generally referring to left ventricular preload include: left ventricular end
diastolic blood volume (LVEDV), left ventricular enddiastolic pressure (LVEDP), left atrial pressure (LAP), pulmonary capillary wedge pressure/pulmonary artery
wedge pressure (PAWP), or pulmonary artery occlusion pressure (PAOP).
2.2.1 FrankStarling curve

The relationship between preload and left ventricle (LV) function during systole is graphically depicted by the FrankStarling curve (Figure 2.1). This curve plots
preload (end
Figure 2.1 The FrankStarling curve plots preload (enddiastolic volume) against an indicator of left ventricular (LV) performance (such
as stroke volume or cardiac output) at a given level of contractility (inotropic state) and impedance. The three curves shown
here represent a normal state as well as mild to moderate and severe decompensation. Reproduced with permission from: The
Journal of Critical Illness (1996) 11:86–87.
Page 19
diastolic volume) against an indicator of LV systolic performance (cardiac output (CO), stroke volume, stroke work) at a given level of contractility (inotropic state)
and impedance (any load acting against LV contraction).
The Starling curve fosters our understanding of how the heart responds to acute injury and chronic decompensation, and how different disease states can cause acute
decompensation. The three curves in Figure 2.1 represent three clinical states. In the presence of systolic dysfunction or increased impedance, the LV performs on a
lower curve. Cardiogenic shock can develop when cardiac index falls below 2.2 l min–1 m–2 2. Pulmonary edema generally occurs when LVEDV rises to greater than
or equal to 25 mmHg. Left ventricular performance moves to a higher curve when contractility increases or impedance decreases. Diuretics will decrease LVEDV and
LV filling pressure, but may also decrease systolic performance (Figure 2.1, A), without shifting the curve. Vasodilators or inotropes increase stroke volume at any
preload level so that ventricular function shifts to a higher curve (Figure 2.1, B). Inotropes plus volume infusion increase both preload and stroke volume (Figure 2.1,
C). With severe LV dysfunction, fluids may increase systolic performance, but at the risk of pulmonary edema (Figure 2.1, D).
2.2.2 Pressurevolume relationships

The relationship between ventricular enddiastolic pressure and stroke work (the ventricular function curve) can be altered by pharmacologic or neurohumoral
influences3. A nonlinear relationship exists between stroke volume and enddiastolic pressure, and hearts with diminished contractility are less responsive to increases in
enddiastolic pressure or preload. In the failing heart, the ventricular function curve is shifted to the right because as atrial pressure rises, cardiac output does not
increase proportionately to the same level as it does in the normal heart, and levels off at a lower magnitude of cardiac output. For the same work (LV systolic
performance), the normal heart performs at a lower filling pressure than the failing heart.
2.2.3 Alterations of preload

Alterations of preload greatly alter the cardiac output in both the normal and failing hearts. Reduction of preload resulting in a decrease in cardiac output or stroke work
may be observed with hypovolemia, displacement of blood from the thorax (positive pressure ventilation) or cardiac compression (tamponade). Furthermore,
impairment of right ventricular contractility such as in right heart failure may lead to a decrease in left ventricular filling. Any alteration in contractility or preload may be
manifested by tachycardia as a compensation for a decrease in stroke volume.
Clinically, with chronic systolic dysfunction, the LV dilates to maintain a normal stroke volume (SV). At rest, a dilated LV usually performs near the peak of the
Starling curve, such that further increases in preload will have little effect on stroke volume. Therefore, heart rate will often increase to maintain cardiac output in patients
with LV dilatation. Any conditions that increase venous return and thus increase preload (enddiastolic volume) will increase heart rate and inotropy to compensate for
elevation in LVEDV.
2.2.4 Ventricular interdependence and LV preload

Ventricular interdependence describes the process by which alterations in ventricular contraction and volume modify the function of the other ventricle. This interaction
is the result of the close anatomic association between the ventricles, which are encircled by common muscle fibers, share a septal wall, and are enclosed within the
pericardium.
As right ventricular (RV) volume and pressure increase, the left ventricular pressurevolume curve shifts to the left and becomes steeper (Figure 2.2). Conversely, as
left ventricular volume and pressure increases, the right ventricular pressurevolume curve shifts to the left and becomes steeper. The mechanism for this interaction is
diastolic alteration in
Page 20
Figure 2.2 Increases in RV volume can cause a septal shift and changes in the pressurevolume relationship of the left ventricle as a
result of ventricular interdependence. Reproduced with permission from Blackwell Science, The Journal of Intensive Care
Medicine (1989) 4:86.
ventricular configuration caused by contralateral ventricular volume changes. Acute distention of one ventricle causes septal displacement and the distensibility of each
ventricle is altered. The greater the right ventricular volume, the less compliant is the left ventricle. Therefore, for the same left ventricular filling pressure, right ventricular
dilatation will decrease LVEDV and cardiac output.
Increases in lung volume and increases in alveolar pressure (Table 2.1) impede right ventricular outflow by increasing pulmonary vasculature resistance, which in the
volume resuscitated patient will cause the right ventricle to dilate, pushing the intraventricular septum into the left ventricle and further
Table 2.1 Ventricular interdependence
Right ventricular pressure overload state

• Acute pulmonary hypertension
• Pulmonary embolism
• Chronic pulmonary disease
• Mitral stenosis
Right ventricular volume overload states
• Atrial septal defect
• Tricuspid insufficiency
• Pulmonary insufficiency
decreasing left ventricular diastolic compliance. Clinically, this reduction in left ventricular stroke volume for a given pulmonary artery wedge pressure may be
misinterpreted as left ventricular failure when it merely represents a mechanically induced reduction of left ventricular diastolic compliance.
The reduction in venous return during positive pressure ventilation is amplified in the settings of hypovolemia and vasodilatory states, such as in shock, sepsis, and
adrenal insufficiency. Management consists of intravascular volume replacement, use of vasopressors, and perhaps smaller tidal volume ventilation.
2.3 Afterload
Although stroke volume reduction can result in arterial hypotension, arterial blood pressure can also be maintained with a compensatory increase in peripheral systemic
vascular resistance. Afterload is the tension created in the walls of the left ventricle as the muscle fibers shorten during contraction. Its determinants include the
peripheral vascular resistance and the resistance against which the ventricle contracts (the impedance of aortic wall and valve). Hypertension (increased peripheral
resistance), arteriosclerosis (stiff arterial walls) or aortic stenosis increase afterload. Both aortic pressure and systemic vascular resistance are critical determinants of the
amount of blood pumped by the ventricle. The total sum of the forces resisting left ventricular ejection is referred to as impedance, which includes the resistance of the
small arteries and arterioles, the compliance of the large arteries, blood viscosity, and the forces of inertia. The arterioles are the main resistant vessels and most
resistance to flow occurs in the peripheral region. Pulmonary vascular resistance (PVR) and systemic vascular resistance (SVR) reflect the afterload of the right and left
side of the heart, respectively. Afterload alterations can increase or decrease systolic LV performance when the inotropic state remains constant (Figure 2.3). Drugs
that decrease
Page 21
Figure 2.3 Changes in afterload (primarily those resulting f from impedance) can increase or decrease systolic—left ventricular
performance if the inoropic state remains constant Reproduced with permission from The Journal of Critical Illness, (1996)
11:86–87,
peripheral vascular resistance (decrease afterload) increase systolic performance without altering the inotropic state (A). Inotropic agents that increase contractility
increase systolic performance but do not affect impedance (B). Vasoconstrictors decrease systolic performance by increasing peripheral vascular resistance
(impedance) (C).
2.3.1 Stroke work

Stroke work is a measure of the work required of the left ventricle to develop blood pressure and eject blood with each heartbeat. Stroke work can be calculated for
either side of the heart and is commonly indexed and presented in hemodynamic profiles as LVSWI and RVSWI (stroke work index).
All of these myocardial and vascular factors are interdependent and alterations of one physiological parameter will affect the others. An increase in afterload via
elevation in blood pressure or systemic vascular resistance will lead to a compensatory increase in preload to maintain stroke volume in the normal functioning heart.
However, in the presence of heart dysfunction, an increase in afterload frequently decreases stroke volume and cardiac output. Conversely, the reduction in afterload
causes the impedance to ventricular ejection to decrease (decrease wall tension) with subsequent stroke volume increase.
2.4 Contractility
Contractility or the inotropic state of the heart is defined by the force and velocity of ventricular contraction, and is independent of preload, afterload and heart rate. The
term refers to the strength of the contraction and the amount of cardiac muscle shortening that occurs with a given fiber length (preload) against a given afterload. An
increase in fiber length signifies an increase in the stretch or preload. With normal contractility, increasing enddiastolic fiber length increases stroke volume. If
contractility increases, stroke volume increases for the same fiber length (preload). At any given level of enddiastolic volume and degree of peripheral vascular
resistance, as the inotropic state increases so does the stroke volume and, as a result, the ejection fraction also increases. As shown in Figure 2.1, there is a shift in the
ventricular function curve to the right with decreased contractility. Left ventricular stroke work does not increase to the same level as occurs with normal contractility as
enddiastolic fiber length increases (preload). In the presence of a reduction of contractility, preload must increase to maintain stroke volume. Ejection fraction, the ratio
of stroke volume to the LVEDV, is used as a clinical measure of contractility. With a normal contractile state, cardiac output is dependent more on preload and
afterload than on the inotropic state of the myocardium. Thus in normal subjects, augmentation of contractility does not necessarily elevate cardiac output. In contrast,
when the myocardial contractile state is depressed, augmentation of the contractile state will increase cardiac output. In the perioperative period, several factors may
predispose to myocardial depression, including anesthetic drugs, hypoxia, hypercapnia, acidbase disturbance, electrolyte abnormalities, surgical manipulation, drugs,
and preexisting heart disease. The terms contractility and myocardial performance often contribute to confusion when reviewing cardiac physiology. Myocardial
performance is a summation of preload, afterload, contractility and heart rate, and is usually
Page 22
measured as cardiac output. Contractility is only one component of myocardial performance, much as the engine is only one component of a sports car’s performance.
Although contractility may not be measurable directly, it is important conceptually because of its ability to be enhanced by inotropic agents. Although it may be
reasoned that increasing the inotropic state improves LV function, agents that increase contractility may also increase heart rate and myocardial oxygen consumption,
causing lifethreatening ventricular arrhythmias. Chapter 8 more fully discusses inotropic agents and their effect on myocardial performance.
2.5 Heart rate

Heart rate is a major determinant of cardiac output, one that is subject to extrinsic autonomic modulation. Alterations in heart rate within the physiologic range may
produce significant changes in cardiac output. Such changes may be important when cardiac dysfunction is impaired or stroke volume is limited. Tachycardia often
serves as a compensatory response when stroke volume decreases. A reduction in diastolic filling time and possibly reduction in cardiac output may occur with heart
rates above 160 beats min−1. An increase in heart rate raises myocardial oxygen consumption and may affect the management of patients with congestive heart failure
(CHF).
2.5.1 Pacemakers
Temporary pacing is indicated for actual or threatened bradycardia caused by acute myocardial infarction, conduction system disease, procedures associated with
important bradycardia or overdrive suppression of tachyarrhythmias4. There are no categorical rules regarding the necessity of temporary pacing. Table 2.2 classifies
situations in which temporary pacing may be useful or effective.
2.5.2 Transcutaneous temporary cardiac pacing

Transcutaneous pacing can be used as an alternative to transvenous pacing and avoids some of the possible complications of central venous cannulation5. The
application of electrical stimuli of long duration and high current can cause skeletal muscle stimulation and severe pain. Transcutaneous ventricular pacing can be
employed as an alternative to transvenous pacing for Class I, Class IIA, and IIB indications as noted above. When the patient is dependent on pacing to maintain an
appropriate heart rate or when it is anticipated that temporary pacing will be needed for greater than 24 h, insertion of a transvenous pacemaker is usually preferred.
Transcutaneous ventricular pacing can be useful in acute settings such as cardiac arrest to avoid delays and complications from central venous cannulation.
2.6 Diastolic function

Intrinsic LV compliance is defined as the change in intrachamber pressure that occurs as a volume of blood enters the left ventricle. In a compliant ventricle, as blood
volume in the chamber increases, the wall distends. This will allow accommodation of large volumes of blood, normally with minimal changes in diastolic pressure.
However, in a noncompliant ventricle, small changes in volume result in marked changes in pressure. Figure 2.4 demonstrates the relationship between left ventricular
pressure and volume during increases and decreases in LV compliance. The points on the curve reflect the volume of blood remaining in the LV at endsystole and the
amount entering the ventricle during diastole. Pulmonary edema can develop in the presence of decreased, normal, or increased compliance as depicted by the
pressurevolume curve. With decreased left ventricular compliance, a normal left ventricular volume may increase diastolic pressure enough to cause pulmonary edema
(A); with lower volume, pressure increases, but to a lesser extent (B). Volume overload states can cause pulmonary edema even with normal LV compliance (C).
Chronic valvular regurgitation (D) and LV dilatation (E) increase compliance, allowing large volumes of blood to be accommodated with small
Page 23
Table 2.2 Indications for transvenous temporary pacings
Class I—usually indicated and considered useful/effective

1. Bradycardia owing to conduction system disease (permanent pacing not immediately available)
A. Complete heart block with:
1. symptomatic bradycardia
2. congestive heart failure
B. Seconddegree atrioventricular (AV) block with symptomatic bradycardia
C. Atrial fibrillation or flutter with periods of symptomatic bradycardia or congestive heart failure attributable to bradycardia
D. Symptomatic bradycardia in the setting of electrolyte abnormalities or drug toxicity
E. Torsades de point
F. Heart rate <30 beats min–1 persisting for ≥10 min
2. Acute myocardial infarction:
A. Asystole
B. Complete heart block
C. Right bundle branch block with left anterior or left posterior hemiblock with acute anterior myocardial infarction
D. Type II seconddegree AV block with acute anterior myocardial infarction
E. Symptomatic bradycardia unresponsive to atropine
3. Procedures that may be associated with significant bradycardia
A. General anesthesia with complete heart block or Type II seconddegree AV block
4. Overdrive suppression of tachyarrhythmias
A. Symptomatic recurrent ventricular tachycardia refractory to drug therapy
B. Symptomatic recurrent supraventricular tachycardia refractory to drug therapy
Class II—acceptable, of uncertain efficacy, may be controversial
A. Weight of evidence in favor of usefulness
B. Not wellestablished evidence, possibly helpful but not harmful
Class IIA
1. Acute myocardial infarction
A. Left bundle branch block with an acute infarction
B. Right bundle branch block with left anterior or left posterior hemiblock with an acute inferior infarction
C. Type II seconddegree AV block within acute inferior infarction
D. Sinus bradycardia with hypotension, unresponsive to atropine
E. Recurrent sinus pauses unresponsive to atropine
F. Atrial or ventricular overdrive pacing for incessant ventricular tachycardia
Class IIB
1. Bradycardia owing to conduction system disease (permanent pacing not immediately available)
A. Asymptomatic complete heart block
B. Asymptomatic type II seconddegree AV block
A. Bradycardia with acute inferior infarction responding to atropine
B. Left bundle branch block with firstdegree heart block of unknown duration
C. Bifascicular block of unknown duration
3. Procedures associated with important bradycardia
A. General anesthesia with right bundle branch block with left anterior or left posterior hemiblock
B. Right heart catheterization in the presence of left bundle branch block
Class III—not indicated, may be harmful
1. Bradycardia owing to conduction system disease
A. Firstdegree AV block
B. Asymptomatic type I seconddegree AV block
C. Congenital thirddegree heart block without symptoms
D. Intermittent AV dissociation without bradycardia or heart block
A. New right bundle branch block
B. Preexisting bundle branch block
C. Asymptomatic type I seconddegree AV block or firstdegree AV block
D. Accelerated idioventricular rhythm causing AV dissociation
Page 24
Figure 2.4 Pulmonary edema may develop whether left ventriculir (LV) compliance is decreased, normal or increased—as indicated by
the pressurevolume curves depicted here With decreased LV compliance, even a normal LV volume may increase diastolic
pressure enough to cause pulmonary ede ma (A); when the volume is low, pressure also increases, but to a lesser extent (B).
Volume overload states may lead to pulmonary edema even if LV compliance is normal (C). Chronic valvular regurgitation (D)
and LV dilatation (E) increase cormpliance; this allows large volumes of blood to be accommodated with only small changes in
pressure. However, eventually the diastolic volume rises enough to increase the diastolic pressure above the threshold 1evel for
pulmonary edema (F). Reproduced with permission from The Journal Critical Illness, (1996) 11:86–87,
changes in pressure. Ultimately, diastolic volume rises enough to increase the diastolic pressure above the threshold level for pulmonary edema (F). Conditions such as
LV hypertrophy, myocardial ischemia, hypertrophic cardiomyopathy, and restrictive ventricular disease decrease compliance. With decreased compliance, diastolic
pressure increases at all volume levels.
For example, in hypertensive patients with concentric LV hypertrophy, LV compliance may decrease during diastole despite the ventricle’s small size and strong
contractions. Hence, small changes in LV filling may generate a marked increase in diastolic pressures and lead to pulmonary edema. Up to 40% of patients with
congestive heart failure have isolated diastolic LV dysfunction6. Clinically, one needs to evaluate for LV diastolic dysfunction so as to determine whether therapeutic
interventions that increase or decrease filling volume should be employed. With decreased diastolic compliance, volume reduction can lead to a significant decrease in
systolic performance, according to the FrankStarling mechanism. Volume support can quickly lead to pulmonary edema as a result of this increase in both diastolic
pressure and PAWP.
Conversely, chronic volume overload states (aortic or mitral regurgitation) and dilated cardiomyopathy can increase compliance. Hence, the diastolic volume may be
high enough to increase the diastolic pressure above the threshold level for pulmonary edema (commonly 25–30 mmHg) to develop.
Coronary artery disease, including ischemia and/or acute myocardial infarction, can alter chamber stiffness resulting in a change in compliance shifting the diastolic
pressurevolume relationship up and to the left, so that small changes in enddiastolic volume are translated into much larger increments in enddiastolic pressure.
Diastolic dysfunction is commonly a diagnosis of exclusion. The primary treatment of heart failure with normal systolic function but abnormal diastolic function is
diuretics. ACEinhibitors are useful. However, in the presence of heart failure, betablockers and calcium channel blockers must be added cautiously, although they
improve diastolic compliance.
2.7 Diagnosis and therapy of acute congestive heart failure

In the diagnosis of acute heart failure in ICU patients, either the systolic and/or the diastolic performance may be affected. In those situations where the diastolic
function of the ventricle in impaired, both diastolic and pulmonary venous pressure increase, leading to pulmonary edema. When systolic function is present, cardiac
output decreases and may result in only fatigue as the initial presentation. The patient’s
Page 25
medical history and physical examination may provide the only clues that point to possible left ventricular dysfunction7.
Physical findings are associated with varying degrees of diagnostic or prognostic significance, and must be interpreted within the specific clinical setting. Although
laboratory testing is required to support the diagnoses of LV dysfunction, the following should be considered:
• In patients with moderate to severe dysfunction, an S3 gallop reflects poor LV diastolic compliance.
• Pulsus alternans is found in patients with severe LV dysfunction, and reflects a poor prognosis.
• Marked pulmonary edema suggests severe LV dysfunction and is a poor prognostic sign, particularly in the setting of acute aortic or mitral regurgitation.
• Evidence of right ventricular dysfunction is common in patients with moderate or severe LV dysfunction. Its prognostic significance depends on the degree of the
dysfunction in both the right and left ventricles.
• Mild LV dysfunction in patients who have myocardial infarctions may signify poor prognosis if chronic and lifethreatening arrhythmias develop.
2.7.1 Pathophysiology of congestive heart failure

Although CHF is a constellation of clinical symptoms, the underlying pathophysiology is characterized by impaired cardiac function, with the failing myocardium unable
to adequately perfuse tissues to meet the metabolic demands. The fundamental defect leading to CHF is impaired myocardial contractility. Virtually any form of cardiac
disease can cause heart failure (Table 2.3).
With progressive LV failure, increased filling pressure leads to pulmonary vascular congestion and eventually pulmonary edema and respiratory failure. The failing
myocardium’s inability to meet the body’s metabolic needs
Table 2.3 Causes Heart failure
• Acute myocardial infarction

• Acute myocardial ischemia
• Uncontrolled hypertension
• Valvular disease
• Pericardial disease
• Pregnancy
• Anemia
• Adverse drug effects
• Hyperthyroidism or hypothyroidism
• Nutritional disorders (vitamin deficiency)
• Severe sepsis
leads to increased neurohumeral tone, with elevated circulatory levels of catecholamines. Normally, low cardiac output causes both cardiac and aortic baroreceptor
dysfunction, which will exert an inhibitory effect on sympathetic tone. Instead, there is increasing release of circulatory catecholamines, activation of the renin
angiotensin axis, leading to sodium and water retention, volume overload and progressive cardiac failure8. Elevated adrenergic tone causes vasoconstriction, and with
subsequent increase in myocardial work leads to both cardiac irritability and potentially fatal arrhythmias. With left ventricular failure present, right ventricular failure can
develop from increased pulmonary vascular resistance.
2.7.2 Compensatory mechanisms

In chronic congestive heart failure, three principal compensatory mechanisms are activated to maintain the heart’s normal function as a pump. Initially, the Frank
Starling mechanism operates to increase preload (enddiastolic volume), which maintains stroke volume. The heart dilates to maintain a nearnormal or normal stroke
volume despite a markedly reduced ejection fraction. Myocardial hypertrophy also occurs secondarily, which maintains wall stress as the heart dilates. The amount of
contractile tissue also increases during hypertrophy. Finally, augmentation of the sympathetic nervous system will occur, which increases heart rate and contractile
function while maintaining perfusion pressure to vital organs. These
Page 26
compensatory mechanisms augment cardiac function in the early phases of heart failure. However, elevated venous pressure causes pulmonary and circulatory
congestion. Cardiac dilatation causes high wall stress, which reduces performance. Increasing peripheral resistance may worsen afterload stress on an already poorly
contracting heart. The net result is the full expression of CHF, and at this point the patient may benefit from more aggressive management in an ICU setting.
2.7.3 Diagnosis of CHF (Table 2.4)

If LV dysfunction is suspected, the next steps are to define the extent of any diastolic or systolic impairment, diagnose any structural or functional abnormalities, identify
the underlying cause of the dysfunction, and determine the prognosis. Systolic as well as diastolic dysfunction must be evaluated. In the absence of systolic LV
dysfunction, abnormalities in diastolic performance, except possibly for infiltrative myocardial disease, carry a more favorable prognosis. It is important to remember
that CHF and LV dysfunction are clinical syndromes, not laboratory diagnoses. The severity or extent of the impairment cannot be determined by one measurement.
2.7.4 Electrocardiography
Specific electrocardiographic patterns associated with left ventricular hypertrophy, previous myocardial infarction and ventricular branch block suggest left ventricular
disease. A biphasic
Table 2.4 Etiology of left ventricular hemodynamic dysfunction
Condition Hemodynamic profile

Myocardial disease
• Myocardial infarction Decreased contractility
• Myocardial ischemia
• Myocarditis
• Dilated cardiomyopathy
Extracardiac disorder
• Hypovolemia Increased/decreased atrial preload
• Cardiac tamponade
• Mitral stenosis
Infiltrative disease
• Restrictive cardiomyopathy Decreased compliance
Pressure overload
• Hypertension Increased afterload
• Aortic stenosis
• Aortic sderosis
• Aortic coarctation
Volume overload states
• Renal failure Increased preload, high output states
• Ventral septal defect
• Anemia
• Arteriovenous fistula
Aortic regurgitation
• Acute Increased preload, increased impedance
• Chronic Increased preload, normal impedance
Mitral regurgitation
• Acute Increased preload, very low impedance
• Chronic Increased preload, low impedance
Page 27
P wave (Pmitrale) marks discoordinate atrial contraction, often a marker of valvular disease. Atrial fibrillation results in the loss of atrial ‘kick’, which assumes
increasing importance in determining preload as ventricular compliance decreases. However, electrocardiography alone cannot be used to assess LV function.
2.7.5 Echocardiography
Echocardiography can comprehensively evaluate LV structure and function. Doppler echocardiography assesses valvular stenosis or regurgitation, estimates the left to
right shunt ratios at the atrial or ventricular level, and measures the peak tricuspid Doppler velocity and thereby the pulmonary artery pressure9. Studies of global LV
function and regional wall motion can provide important prognostic information. Stroke volume is derived by subtracting the endsystolic volume (ESV) from end
diastolic volume (EDV), and ejection fraction calculated by dividing stroke volume by EDV. Substantial regional wall motion abnormalities, however, introduce
inaccuracy into these calculations. In patients with chronic mitral and aortic regurgitation, determinations of LV dimensions at endsystole can guide decisions regarding
the need for surgical intervention.
2.7.6 Radiologic evaluation

Roentgenography can provide supporting evidence for LV dysfunction. Venous markings in the upper lung field and Kerley Blines both suggest increased pulmonary
venous pressure. Assessment of the size of the left ventricle can also provide additional clues.
2.7.7 Hemodynamic evaluation

Pulmonary capillary wedge pressure recordings and cardiac output may be measured with a pulmonary artery catheter. Further discussions of hemodynamic monitoring
are found in Chapter 5.
2.8 Therapy
Current management techniques for heart failure include preventing volume overload, reducing myocardial work via afterload reduction though vasodilator use, and
augmentation of cardiac output by administration of positive inotropic agents that act directly on the myocardium to improve cardiac performance (see Chapter 8).
Titration of inotropic or vasodilator therapy for heart failure in the ICU setting depends on establishing clear therapeutic endpoints. Inotropic agents are normally used
to augment cardiac output, improve urine output and lower pulmonary capillary wedge pressure while avoiding sinus tachycardia or ventricular ectopy. Bedside hemo
dynamic measurements via pulmonary artery catheterizations are sometimes important in titrating inotropic therapy. However, in the setting of acute pulmonary edema,
appropriate treatment (see below) often provides rapid resolution of this hemodynamic disorder and obviates the need for invasive monitoring. Pulmonary artery
catheters are most useful when there is uncertainty regarding left ventricular filling pressure or cardiac output.
2.8.1 Diuretics
Diureticinduced reduction in rightsided pressures may improve LV compliance, thereby improving cardiac output and renal perfusion. A decrease in ventricular filling
pressures with diuretics will enhance cardiac performance in both tricuspid and mitral regurgitation. For those patients not previously receiving diuretics, lowdose
furosemide (20–40 mg) is initially recommended. The dose is increased by 50% to 100% when there is no response within 1 h after administration. Shortacting loop
diuretics (Bumetanide 1–5 mg IV) may also be employed for acute heart failure. Rarely larger doses (furosemide 250–500 mg; bumetanide ≥10 mg) are needed for
patients with reduced renal function. Continuous furosemide infusion (5–10 mg h−1) is effective to improve diuresis in critically ill patients who require a controllable
diuretic or who demonstrate diuretic resistance. Furosemide may not be well absorbed orally in patients with decompensated heart failure. Diuretics will often
dramatically relieve symptoms by lowering right
Page 28
and left atrial filling pressures. Improved diuretic efficacy (synergism) exists between various classes of diuretics. Loop diuretics are often used in conjunction with
thiazides, potassium sparing diuretics or metolazone in patients exhibiting resistance to initial therapy. The potassium sparing diuretic, spironolactone, has unique
properties that are beneficial in congestive heart failure, and has been shown to reduce morbidity and mortality for class III or IV heart failure10.
2.8.2 Digoxin
Clinical investigations have substantiated the efficacy of digoxin in the treatment of heart failure. Hemodynamic studies in both animals and humans have shown that
digoxin increases cardiac index and decreases left filling pressure, right atrial pressure and heart rate. Digoxin improves heart inotropy by inhibiting sodiumpotassium
ATPase, leading to an increase in intracellular calcium levels and improved inotropy. Digoxin produces a modest improvement in cardiac output in patients with right
heart failure secondary to pulmonary hypertension. Clinical trials support a longterm benefit of digoxin in preventing recurrent hospitalization, improving symptoms and
functional capacity in patients with CHF, but not in reducing mortality11. Although there is no strong evidence for the benefit of digoxin in the acute setting of CHF, the
drug is still recommended in the absence of abnormal atrioventricular conduction. A loading dose is not necessary in the treatment of heart failure. Treatment can begin
with a dose of 0.25 mg day−1. As a result of changes in lean body mass and renal function, the therapeutic range of digoxin in the elderly is lower. In patients over 70
years of age, digoxin concentrations from 0.5 to 1.3 mg ml−1 can be associated with higher toxicity. There are no data supporting the use of serum levels to guide
dosing. It remains uncertain whether high doses are more effective than lower ones in the management of CHF.
In critically ill patients, a loading dose of digoxin is generally applied for the treatment of acute atrial arrhythmias associated with a rapid ventricular response rate.
Rapid digitalization, independent of renal function, can be accomplished by giving 10–15 µg kg−1 ideal body weight (IBW) in divided doses. Typically, a regimen would
include 50% of the total loading dose administered immediately, followed by 25% 6 h later, and the final 25% 6 h after that. A typical regimen when initiating digoxin
for CHF may include a maintenance oral or parenteral therapy without a loading dose (0.125–0.25 mg day−1), depending on the extent of renal dysfunction.
2.8.3 Beta agonists and phosphodiesterase inhibitors

Beta adrenergic agonists and phosphodiesterase inhibitors are mainstays of treatment for heart failure in the ICU, and are discussed in detail in Chapter 8.
2.8.4 Multimodal therapy

In the presence of acute heart failure, diuresis should be initially attempted. In the presence of renal dysfunction or hypoperfusion, inotropes and vasodilators
(nitroglycerin, nitroprusside, ACEinhibitors) are frequently required. Diuresisinduced reductions in rightsided pressures improve LV compliance, increasing both
cardiac output and renal perfusion.
Inotropes and vasodilators may be used in the presence of a lowoutput state or when diuretics are found to be ineffective or exacerbate renal dysfunction.
Concomitantly, efforts should be attempted to initiate longterm pharmacologic management, which should include a diuretic, an angiotensin converting enzyme inhibitor,
a vasodilator, digoxin and betablockers10. ACEinhibitor administration is associated with a decrease in enddiastolic and endsystolic volume, an increase in chamber
stiffness, and perhaps a decrease in LV mass. In the setting of acute exacerbation of heart failure, shorteracting vasodilators such as nitrates rather than ACEinhibitors
are more appropriate initially. Although high levels of norepinephrine in
Page 29
CHF are linked with poor prognoses, betablockade is helpful. However, betablockers have no documented shortterm benefit and can be harmful during acute
exacerbation. Synergism between betaagonists and the phosphodiesterase inhibitors has been documented, and varying combinations can be utilized to achieve
desired amounts of inotropy and vasodilation.
2.9 Patient subsets
2.9.1 Right ventricular failure

Right ventricular (RV) failure is often a consequence of LV failure. Other mechanisms of RV failure include right ventricular myocardial infarction, and dilated
cardiomyopathy. RV pressure overload may be secondary to chronic LV failure, chronic mitral valve disease or ‘cor pulmonale’ from COPD or pulmonary emboli. RV
volume overload may be secondary to pulmonic or tricuspid regurgitation or a left to right intracardiac shunt.
Patients with acute right ventricular (RV) failure secondary to an RV infarction or acute pulmonary embolism may manifest signs of low cardiac output and elevated
jugular venous pressure, a murmur of tricuspid regurgitation, rightsided S3 gallop, and Kussmaul’s sign. Patients with chronic RV pressure overload have signs of
pulmonary hypertension including an RV S4 gallop, RV heave and an accentuated pulmonic component to the second heart sound. Patients with chronic RV failure
frequently present with edema, ascites, and pleural effusions. Diuretics are useful in decreasing the central venous pressure elevation and edema seen in RV failure.
Loop diuretics can improve cardiac output by decreasing the degree of tricuspid regurgitation, decreasing RV distention and thereby reducing a ventricular
interdependent restraint on LV filling. In addition, spironolactone can reduce edema from RV failure because activation of the reninangiotensinaldosterone system is
involved in sodium retention in these patients. Systemic arteriolar vasodilators used in LV failure therapies act as pulmonary vasodilators (i.e. ACEinhibitors, nitrates,
nitroprusside) and improve RV output by reducing RV afterload12.
Supplemental oxygen may also decrease pulmonary hypertension in these patients and should be employed acutely. Bronchodilators can improve RV function and
symptoms in cor pulmonale exacerbation as a result of its bronchodilatory, pulmonary vasodilatory, and inotropic effects.
2.9.2 Valvular heart disease

Several valvular lesions can lead to symptoms of heart failure. Although medical management can provide shortterm benefit, timing of surgical intervention is of
essential importance (Table 2.5).
2.9.3 Mitral regurgitation

The most common causes of acute mitral regurgitation (MR) requiring ICU admission are acute infective endocarditis, chordal rupture in patients with myxomatous
valves, papillary muscle rupture or dysfunction in the setting of inferior ischemia, and blunt or perforating chest trauma resulting in leaflet tear or chorda rupture.
Acute MR presents with signs of left atrial pressure overload and resulting pulmonary venous hypertension. Highpressure regurgitant flow from the left ventricle into
the noncompliant left atrium produces left atrial systolic V waves that are transmitted through the pulmonary venous system into the pulmonary capillaries, resulting in
pulmonary edema. Left ventricular ejection into the ‘lowpressure’ left atrium can lead to an acute reduction in cardiac output with subsequent hypotension and
cardiogenic shock. In chronic MR, chronic LV volume overload results in progressive dilation. As the LV may decompress itself into the lowpressure left atrium,
afterload is reduced and ejection fraction is maintained in the presence of impaired intrinsic contractility.
In patients with pulmonary edema and hemodynamic collapse, clinical features that are suggestive of the diagnosis of acute MR include:
Page 30
Table 2.5 Valvular surgery indications
Mitral regurgitation Onset of symptoms with moderate or severe MR

Or
Asymptomatic LV systolic dysfunction (LVESD*≥4.5 cm)
Mitral stenosis Symptoms (heart failure, thromboembolism, fatigue) refractory to medical therapy with moderate or severe MS
Aortic regurgitation Symptoms with moderate or severe AR
Or
Asymptomatic LV systolic dysfunction (LVESD≥5.5 cm)
Aortic stenosis Symptoms (angina, syncope or heart failure) with moderate or severe AS
*
LVESD, left ventricular endsystolic dimension.
• a hyperdynamic apical impulse, possibly along with an RV heave;

• a loud P2 caused by pulmonary hypertension;
• a holosystolic murmur heard best at the apex and radiating into the axilla;
• a third heart sound; with severe MR the murmur can be soft.
Electrocardiogram can show only sinus tachycardia although patients with ischemic MR may show evidence of inferior wall infarction. The chest Xray demonstrates
pulmonary edema.
Doppler echocardiography is essential for accurate and rapid evaluation of the cause and pathophysiology of MR in the hemodynamically unstable ICU patient.
Chordal disruption, valvular abnormalities including flail leaflets and vegetations caused by endocarditis and papillary muscle dysfunction and rupture can be diagnosed.
Color Doppler can be used to confirm the presence of MR to evaluate severity, and to identify eccentric regurgitant jets. The severity of left atrial pressure overload
can be inferred from the characteristics of the regurgitant jet.
In the unstable population, pulmonary artery catheterization may be helpful in confirming the significance of mitral regurgitant lesions. With severe acute MR, the
pulmonary capillary V wave to mean pressure ratio is usually greater than two. In addition, serial cardiac output measurements assist in the selection of medical therapy.
Therapy for acute MR and pulmonary edema includes left ventricular afterload reduction, best achieved using parenteral vasodilator therapy, usually with
nitroprusside and/or angiotensinconverting enzyme inhibitors13. Nitrates are helpful in the presence of myocardial ischemia leading to papillary muscle dysfunction.
Diuretics may be necessary for the treatment of pulmonary edema. Intraaortic balloon pumping is particularly useful to reduce left ventricular afterload and augment
forward flow. Inotropic support with dopamine/dobutamine may be required in patients with cardiogenic shock who are manifesting hypoperfusion. Patients who do
not respond to medical therapy and require intraaortic balloon pumping require emergent valve surgery. Mitral valve repair is preferable to mitral valve replacement,
because maintenance of subvalvular continuity improves postoperative LV function.
2.9.4 Mitral stenosis

Mitral stenosis (MS) is almost always rheumatic in origin. Other rare causes include left atrial myxoma, obstructing thrombus, malignant carcinoid, lupus, congenital
defects, rheumatoid arthritis, and micropolysaccharidoses.
Isolated MS is associated with an impairment of left ventricular diastolic filling, with elevated left atrial pressure and volume, leading to pulmonary hypertension. The
left ventricle is protected and contractility is normal.
Page 31
Longstanding, severe pulmonary hypertension can lead to pulmonary vascular fibrosis, which may become irreversible.
Clinical features of MS include dyspnea and decreased exercise tolerance as a result of an increased left atrial pressure and the inability to achieve a significant
increase in cardiac output with exertion. The development of atrial fibrillation with loss of the atrial kick and decreased filling time leads to a dramatic worsening of
symptoms. Atrial embolic events as well as hemoptysis are common presentations of MS. Physical findings in MS include:
• elevated jugular venous pressure in the presence of pulmonary hypertension;
• right ventricular heave may be present; normal apical impulse;
• accentuated P2 as a result of pulmonary hypertension;
• loud S1;
• diastolic rumbling murmur sometimes with presystolic accentuation as a result of atrial contraction.
In severe MS, a decrease in cardiac output can make the murmur inaudible. In these cases, the signs of right ventricular failure are very prominent. The
electrocardiogram (EKG) shows evidence of left atrial enlargement (P mitrale) and may have a rightward QRS complex with other signs of right ventricular
hypertrophy. Left atrial enlargement and prominent pulmonary vasculature can be seen on chest Xray.
Echocardiography allows for evaluation of mitral valve pathology and can evaluate left atrial size and the presence of a thrombus as well. Echocardiography can
reveal a thick and unusually calcified mitral valve with decreased leaflet mobility in addition to left atrial enlargement. The severity of MS can be determined by
planimetry of the valve orifice (Table 2.6).
Therapy includes controlling ventricular rate and maintenance on restoration of sinus rhythm. Digoxin and diuretics can provide substantial symptomatic relief along
with negative chronotropic agents to control heart rate and antiarrhythmic agents to maintain sinus
Table 2.6 Severity of mitral stenosis in relation to mitral valve area
Normal mitral valve area 4–6 cm2

Trivial MS >2.0 cm2
Mild MS 1.5–2.0 cm2
Moderate MS 1.0–1.5 cm2
Severe MS <1.0 cm2
MS, mitral stenosis.
rhythm. Chronic anticoagulation is usually required as a result of atrial fibrillation and mitral stenosis association with thromboembolism. Surgical therapy provides
definitive therapy for patients who are moderately symptomatic with severe MS. Surgical options include valve replacement, commissurotomy and balloon valvulotomy.
The choice of therapy depends on the degree of subvalvular complication and concomitant mitral regurgitation and other clinical characteristics.
2.9.5 Aortic stenosis

Aortic stenosis (AS) is the most common isolated valvular lesion and is most often caused by degeneration and calcification, or a congenital bicuspid valve in patients
under the age of 70, or a tricuspid valve in persons over the age of 70. Degenerative AS caused by ‘wear and tear’ is characterized by calcification of the aortic side of
the valve preventing the cusps from opening normally.
AS manifests as a fixed obstruction to left ventricular outflow leading to left ventricular pressure overload. This increase in wall stress and left ventricular hypertrophy
occurs as a compensatory response. Hypertrophy decreases left ventricular compliance, and increases left ventricular diastolic pressure. Although cardiac output at rest
is usually normal in these patients, because of the outflow track obstruction there is limitation of stroke volume augmentation during exertion. Increased left ventricular
mass and wall stress increases myocardial oxygen consumption. A decrease in the aortic pressure distal to the obstruction and an increased left ventricular diastolic
pressure causes a decrease in coronary
Page 32
perfusion pressure and myocardial oxygen supply. This can result in subendocardial ischemia, further decreasing left ventricular compliance and systolic performance. In
the later stages of AS, systolic function deteriorates and left ventricular dilation occurs.
Classical presenting symptoms of AS include angina, syncope and heart failure, usually exercisedinduced. Presentation of AS in the ICU include pulmonary edema,
angina, severe hypotension and low cardiac output states. Physical findings include:
• delayed and diminished carotid upstroke;
• sustained apical impulse;
• an ejection click, soft A2, loud S4;
• a harsh systolic ejection murmur at the base, peaking later with increased severity of AS.
EKG shows left ventricular hypertrophy with strain and may show interventricular conduction delays or atrioventricular blocks. Echocardiography can determine the
severity of stenosis by evaluating the degree of thickening, calcification, and excursion of aortic valve. Pulse Doppler can provide an accurate assessment of both mean
and peak gradients across the stenotic valve (Table 2.7). Doppler examination permits estimation of the transvalvular pressure gradient and calculation of the aortic
valve area. Although catheterization remains the standard for determining valve area, echocardiographically determined aortic valve area correlates well with
catheterization results. Aortic catheterization should be reserved for the symptomatic patient being
Table 2.7 Severity of aortic stenosis in relation to aortic valve area
Normal aortic valve area 3–4 cm2

Trivial AS >1.3 cm2
Mild AS 1.1–1.3 cm2
Moderate AS 0.7–1 .0 cm2
Severe AS ≤0.7 cm
2
Critical AS ≤0.5 cm
2
AS, aortic stenosis.
considered for valvular surgery. In addition, catheterization is performed as part of the workup to provide information about concomitant coronary artery stenosis.
Management of patients with AS involves maintaining an adequate preload to sustain cardiac output. Prompt fluid resuscitation should be instituted for hypovolemia.
Preload and afterload reduction are contraindicated in the patient with known AS. Excess fluid administration, however, may result in pulmonary edema as a result of
impaired left ventricular diastolic compliance. As atrial systole makes an important contribution to fill in the noncompliance left ventricle in AS, efforts should be used to
maintain a normal sinus rhythm. When symptoms develop in a patient with moderate to severe AS, valve replacement is required.
2.9.6 Aortic regurgitation

The most common causes of acute aortic regurgitation are acute infective endocarditis, acute aortic dissection, spontaneous chordal rupture, or tears of myxomatous
valve leaflets, and valve destruction from blunt or perforating chest trauma.
Hemodynamically significant acute aortic regurgitation (AR) presents with manifestations of left ventricular diastolic pressure overload as a result of a highpressure
regurgitant flow into a noncompliant, nondilated LV. This increased diastolic pressure is transmitted to the left atrium and the pulmonary venous system. Although left
ventricular preload is increased, the ventricle is able to dilate enough to maintain a sufficient stroke volume to compensate for the regurgitate flow. Compensatory
tachycardia commonly occurs to maintain cardiac output. Rapidly declining aortic diastolic pressure along with increasing left ventricular enddiastolic pressure reduces
coronary perfusion pressure, which, when combined with decreased diastolic filling time owing to tachycardia, presents the potential for subendocardial ischemia.
Page 33
Clinical features of AR include:

• wide pulse pressure with a ‘water hammer’ pulse;
• inferior and lateral displacement of the apical impulse;
• a soft A2;
• a third heart sound;
• a highpitched blowing decrescendo early diastolic murmur along with a systolic murmur caused by increased flow across the aortic valve;
• an apical diastolic murmur.
In patients with acute AR, many of the above signs may be greatly attenuated or absent. Doppler echocardiograph is essential for accurate and rapid evaluation of AR
in the hemodynamically unstable patient. Valvular vegetations or prolapse caused by leaflet tears can be diagnosed, as can aortic annular dilation. Transesophageal echo
is highly useful in cases of suspected ascending aortic dissection or aneurysms, and in evaluation of the aortic route in patients with endocarditis and suspected aortic
ring abscesses.
Management of patients with acute AR and pulmonary edema includes reducing elevated left ventricular preload with diuretics and nitrate. Afterload reduction to
improve flow and reduce regurgitant fraction with ACEinhibitors, hydralazine or nitroprusside is also indicated14. Intraaortic balloon pumping worsens AR and is
contraindicated. Bradycardia increases regurgitant fraction and should be avoided. Atrial pacing may be useful to increase heart rate in normotensive patients, and
inotropes may be useful if concurrent hypotension or a low output state are present. AR frequently responds to medical therapy and aortic valve surgery can be
undertaken electively. However, patients with ascending aortic dissection with acute AR should be treated acutely with antihypertensives and betablockers to decrease
aortic wall stress. Pulmonary edema as a result of infective endocarditis and severe AR is also an indication of surgical intervention particularly when the vegetation is
large (>1 cm), when the infection involves a prosthetic valve or if the aortic root is involved.
References
1. Ross, J., Jr (1979) Afterload mismatch and preload reserve: a conceptual framework for the analysis of ventricular function. Prog. Cardiovasc. Dis. 14:255–264.
2. Hands, M.E., Rutherford, J.D., Muller, J.E., et al. (1989) The inhospital development of cardiogenic shock after myocardial infarction: incidence, predictors of
occurrence, outcome and prognostic factors. J. Am. Coll. Cardiol. 14: 40–46.
3. Schaffer, J., Tews, A., Langes, K., et al. (1987) Relationship between myocardial norepinephrine content and left ventricular function. An endomyocardial biopsy
study. Eur. Heart J. 8: 748.
4. Gregoratos, G., Cheitlan, M.D., Conill A., et al. (1998) ACC/AHA guidelines for implantation of cardiac pacemakers and antiarrhythmia devices. J. Am. Coll.
Cardiol. 31:1110–1114.
5. Chen, L.K., Teerlink, J.R. and Goldschlager, N. (1997) Pacemaker emergencies. In: Cardiac Intensive Care (ed. D.L.Brown). Philadelphia, PA: W.B.Saunders,
pp. 405–426.
6. Vasan, R.S., Benjamin, E.J. and Levy, D. (1995) Prevalence, clinical features, and prognosis of diastolic heart failure: An epidemiologic perspective. J. Am. Coll.
Cardiol. 26: 1565–1574.
7. Wilson, J.R., Rayos, G., Yeoh, T.K., et al. (1995) Dissociation between exertional symptoms and circulatory function in patients with heart failure. Circulation
92:47–53.
8. Francis, G.S., Goldsmith, S.R., Levine, T.B., et al. (1984) The neurohumoral axis in congestive heart failure. Ann. Intern. Med. 101:370–377.
9. Seward, J.B., Khandheria, B.K., Oh, J.K., et al. (1988) Transesophageal echocardiology: Technique, anatomic correlations, implementation and clinical applications.
Mayo Clin. Proc. 63: 649.
10. Kimmelstiel, C., Udelson, J., Smith, J., et al. (1998) Current concepts in the treatment of patients with heart failure. ACC Curr. J. Rev. 7: 39–43.
11. The Digitalis Investigation Group (1997) The effect of digoxin on mortality and morbidity in patients with heart failure. N. Engl J. Med. 336: 525–533.
Page 34
12. Dittrich, H.C., Chow, L.C. and Nicol, P.H. (1989) Early improvement in left ventricular diastolic function after relief of chronic right ventricular pressure overload.
Circulation 80: 823–830.
13. DePace, N.L., Nestico, P.F. and Morganroth, J. (1985) Acute severe mitral regurgitation: Pathophysiology, clinical recognition, and management. Am. J. Med.
78:293.
14. Miller, R.R., Vismara, L.A., DeMara, A.N., et al. (1976) Afterload reduction therapy with nitroprusside in severe aortic regurgitation: Improved cardiac
performance and reduced regurgitant volume. Am. J. Cardiol. 38:564.
Page 35
Chapter 3
Oxygen transport and tissue oxygenation
William T.McGee, MD, MHA and
Paul Jodka, MD
Contents
3.1 Introduction
3.2 Calculating oxygen transport
3.3 Balancing oxygen supply and demand
3.4 Clinical factors influencing oxygen supply
3.5 Determinants of oxygen consumption
3.6 Critical factors influencing cellular oxygen utilization
3.7 Evaluation of tissue oxygenation
3.8 Influence of loss of vasoregulation
3.9 Continuous mixed venous oxygen saturation in clinical medicine
3.10 Goaldirected therapy with DO2 and VO2
3.11 Summary
Page 36
3.1 Introduction
Physicians frequently evaluate indices of organ function as an endpoint of therapy. In the angina patient, relief of pain and resolution of ischemic electrocardiographic
changes are viewed as success of antianginal therapy. Similarly, in renal disease, maintenance of adequate urine output, a normal creatinine and blood urea nitrogen
(BUN) are useful global physiologic measures of renal health. Normal organ function implies at a minimum adequate blood flow and oxygen (O2) delivery. When
disease intervenes, however, organ function may deteriorate. As illness progresses and subsequent organ function deteriorates to a critical stage, a variety of
technologies are instituted in the critical care unit to prevent immediate death. Nonetheless, many patients continue to deteriorate and can die as a result of the gradual
decline of all organ function. Multisystem organ failure in the critically ill results from failure of adequate oxygen and other nutrients to be delivered to tissues. Methods
of determining adequate tissue oxygenation are crucial to the practice of medicine within the intensive care environment. It is critical to differentiate between organ
dysfunction as a result of primary disease and organ dysfunction as a result of failure of the oxygen transport system.
Since the mid1970s, it has been common to utilize invasive hemodynamic monitoring to delineate pulmonary artery pressures, pulmonary capillary wedge pressure,
and cardiac output. However, these measurements merely define the determinants and indices of cardiac function, but do not provide information on the distribution of
blood flow or the adequacy of peripheral tissue oxygenation and utilization. Critical illness ultimately results in the breakdown of the normal peripheral mechanisms of
blood flow distribution, oxygen delivery, and oxygen utilization.
Data obtained during pulmonary artery catheterization allow evaluation of the oxygen delivery system and the global utilization of oxygen by the body. Evaluation of
mixed venous oxygen and its general relationship to tissue oxygenation can at least assure us that bulk O2 delivery, which we have significant ability to manipulate in the
ICU, is not the cause of organ dysfunction and ultimately death.
3.2 Calculating oxygen transport
3.2.1 Oxygen delivery

The adequacy of tissue oxygenation depends upon the balance between the oxygen transported to cells and the oxygen requirement determined by their metabolic
activity. Bulk oxygen transport also known as oxygen delivery and frequently abbreviated as DO2 is simply the amount of oxygen carried within the blood multiplied by
the blood flow or cardiac output1:
Arterial oxygen content (CaO2) is the amount of oxygen carried in 100 ml of blood, which includes oxygen chemically bound to hemoglobin as well as a small portion
of oxygen dissolved in the plasma.
Quantitatively,
PaO2×0.003 represents the dissolved fraction of oxygen in plasma and is usually of insignificant quantity.
In a normal patient (with a hemoglobin content of 15 g and an arterial oxygen saturation of 99%)
Then
per 100 ml×10 dl l−1 (correction factor for the difference in units between CO and CaO2).
Thus, a normal DO2 is 1000 ml O2 min−1 1. DO2 may be indexed by body surface area and
Page 37
reported as DO2I. Normal DO2I in a 1.73 m2 individual would be 575–600 ml O2 min−1 m−2.
DO2 is the amount of oxygen that normally leaves the heart each minute and is delivered to the tissues. The distribution of this oxygen is determined in a complex
manner by autoregulation of the arterioles and capillary beds in the periphery. Tissues that require large amounts of oxygen will open capillary beds, thus increasing the
amount of oxygen delivered to the tissues. Similarly, tissues that have low metabolic activity close capillary beds, thus preventing wasted blood flow.
When supply is deficient, oxygen utilization is limited by decreased availability. Metabo lism becomes anaerobic, lactic acid is produced, and a metabolic acidosis
will result. Prolonged declines in O2 delivery are associated with lactic acidosis, organ dysfunction and ultimately death.
3.2.2 Oxygen consumption (VO2)

Oxygen is utilized by the mitochondria to produce energy for the cells. The quantity of oxygen extracted from the capillaries as blood passes to the venous side is
defined as oxygen consumption. Oxygen consumption varies as the metabolic needs of the tissues change. Oxygen consumption can be measured directly by
determining the difference between the quantity of inspired and expired oxygen. This may be done rather easily under laboratory conditions when the patient is
breathing room air, or in the operating room with a closedcircuit anesthetic delivery system. Under clinical conditions in the ICU, direct oxygen consumption
measurements are difficult, require specialized equipment, and are fraught with multiple sources of error2,3; indirect measurements are more practical.
Oxygen consumption (VO2) may be calculated indirectly by the Fick equation, which states that oxygen consumption equals the difference between arterial oxygen
delivery and venous oxygen return. The quantity of oxygen returned in venous blood to the right side of the heart can be calculated in a manner similar to the calculation
for DO2. Venous oxygen return equals venous return (VR)×venous oxygen content (CvO2). The calculation for CvO2 utilizes the same general formula as for
calculating CaO2, i.e.
The values for SvO2 and PvO2 are measured from mixed venous blood drawn from the pulmonary artery using the pulmonary artery catheter. The PvO2 and SvO2 in a
normal healthy human are approximately 40 mmHg and 75%, respectively:
which equals 15 ml of oxygen per 100 ml of blood. Venous return in steadystate conditions is equal to the cardiac output, a normal venous return is 5 1 min−1, and
therefore
As the normal oxygen delivery is 1000 ml min–1 and a normal venous oxygen return is 750 ml min–1, the difference between these two, 250 ml min–1, is the amount of
oxygen extracted, i.e. the oxygen consumption (Figure 3.1). The process of consuming oxygen is assumed to occur completely at the capillary bed. Oxygen
consumption equals oxygen delivery minus oxygen return or
As CO=VR in the steady state,
This is just a restatement of the Fick equation. Determining oxygen consumption in this way requires accurate measurement of cardiac output, hemoglobin, PaO2,
SaO2, PvO2, and SvO2, all of which are clinically available with a pulmonary artery catheter and a laboratory or bedside blood gas machine. Utilization of the Fick
equation and C(avO2) can provide important information about the balance
Page 38
Figure 3.1 Normal oxygen transport, giving normal values for oxygen consumption and venous oxygen content.
between oxygen delivery and oxygen consumption3. Like DO2, VO2 may be indexed by body surface area and reported as VO2I.
3.2.3 Oxygen extraction

As a tissue increases its metabolic activity, oxygen delivery is increased by an increase in flow. This may occur globally (i.e. by cardiac output) or locally (i.e. by an
increase in flow (recruitment of capillary beds through autoregulation)). The increase in blood flow and therefore oxygen delivery is sometimes unable to meet the
increased need secondary to higher metabolic activity. In these instances, oxygen extraction at the capillary level can increase. This results in a decrease in the oxygen
saturation of the venous blood (SvO2) leaving the capillaries. If this decrease occurs globally, or if it occurs regionally in a large volume of capillary beds, venous
saturation of blood in the pulmonary artery (SvO2) will decrease. An increasing C(avO2) or a decreasing SvO2 or PvO2 implies that peripheral oxygen utilization has
exceeded the body’s capacity to supply oxygen by increasing flow4. However, this ability to extract oxygen at the capillary bed is limited. The oxygen extraction ratio
(O2 ER) is a convenient measure of global oxygen extraction. It is calculated by dividing the VO2 by the DO2. Normal values are 24–28%.
3.3 Balancing oxygen supply and demand

An oxygen supply/demand imbalance may occur for four basic reasons:
(i) a critical deficiency in one or more of the components of oxygen delivery (i.e. acute pump failure, hypoxemia or acute anemia);
(ii) an increase in metabolic demand that exceeds oxygen transport reserves;
(iii) blockage of normal cellular oxygen metabolism;
(iv) adequacy of total oxygen supply, but inadequate distribution to regional vascular beds.
In these circumstances, cells develop anaerobic metabolism in an attempt to manufacture energy in the absence of oxygen. This is a situation that cannot be maintained.
If the metabolic need of the tissues does not decrease or if the oxygen delivery cannot increase, the cells supplied by these capillary beds will undergo ischemic death.
A common example of this phenomenon occurs within muscle vascular beds during exercise. Although blood flow to exercising muscle increases tremendously, the
metabolic activity of muscle cells demands more oxygen than can be supplied through increased blood flow alone. Oxygen extraction in exercis
Page 39
ing muscle therefore increases tremendously. Nonetheless, oxygen demand is still greater than the additional amount supplied by increased oxygen extraction.
Therefore, exercising muscle enters an anaerobic phase and lactic acid is produced. This is a physiologic condition that can be maintained only for short periods of time,
after which the muscle loses function. In this case, the organ dysfunction generally is beneficial, in that less exercise can be performed, the metabolic demands fall
accordingly, and the supply/demand balance is restored.
In critically ill patients with diseases such as sepsis, adult respiratory distress syndrome (ARDS) or major trauma, oxygen supply/ demand may not be autoregulated
efficiently, and may result in the gradual onset of multisystem organ failure and death. The mechanisms by which this occur are not simple and remain poorly
understood1,4.
3.4 Clinical factors influencing oxygen supply

Adequate oxygen delivery implies a normal or compensated cardiopulmonary system. Table 3.1 depicts the effects of separately altering the components of oxygen
transport. These calculations are oversimplistic, inasmuch as normal compensatory mechanisms adjust for individual deficiencies. Increases in heart rate and contractility
or decreases in afterload will increase cardiac output when changes in hemoglobin or oxygen saturation decrease oxygen transport. Normally, cardiac output will begin
to rise if hemoglobin drops below 10 g dl–1, and is doubled when hemoglobin drops to 5 g dl−1. The mechanism of increased flow is in part due to an increase in
contractility, but is primarily caused by increases in heart rate and a drop in afterload. This is the result of alterations in blood viscosity, which occur in acute anemia. In
chronic anemia, cardiac output returns toward baseline, and tissue oxygenation is maintained by an increase in red blood cell 2,3diphosphoglycerate (2,3DPG). This
change in 2,3DPG facilitates the release of oxygen from hemoglobin at the capillary level5 (Figure 3.2).
3.4.1 Oxygen saturation

Increasing arterial oxygen saturation increases oxygen transport in a linear fashion. Oxygenation is usually thought to be sufficient if PaO2
Table 3.1 Normal oxygen supply and demand and effects of low cardiac output, anemia and arterial desaturation on mixed venous (PvO2) and (SvO2)
DO2 variable Normal Low cardiac output=2.5 l min−1 Anemia Hb=7.5 g dl−1 Arterial desaturation SaO2 =.75 (PaO2 =40)
Arterial O2 content (CaO2) (ml dl−1) 20 20 10 15
O2 transport (CaO2) (ml O2 min−1) 1000 500 500 750
minus O2 consumption (VO2 min−1) −250 −250 −250 −250
equals venous return (ml O2 min−1) 750 250 250 500
divided by CO (dl min−1) ÷50 ÷25 ÷50 ÷50
equals venous O2 content (CvO2) (ml dl−1) 15 10 5 10
(av)O2 content (C(av)O2) (ml dl–1) 5 10 5 5
Mixed venous O2 saturation (SvO2) 0.75 0.50 0.50 0.50
Mixed venous PO2 (PvO2) (torr) 40 27 27 27

−1 −1 −1
Hb=15 g dl ; SaO2 =1.0; cardiac output=5 l min ; VO2 =250 ml min . Other components of O2 delivery are assumed to remain constant for this illustration.
Page 40
Figure 3.2 Effect of anemia on cardiac index and 2, 3DPG. Reproduced with permission from The New Journal of Medicine (1972)
286:412.
equals 60 torr because the oxyhemoglobin dissociation curve is horizontal beyond that point. However, a PaO2 of 60 mmHg is still 7% less than complete hemoglobin
saturation, and in patients with marginal oxygen transport, an increase of 7% may represent a crucial increase in oxygen supply. A PaO2 above 100 mmHg increases
the amount of dissolved oxygen in plasma but results in a quantitatively negligible increase in transported oxygen in most circumstances. However, the availability of
dissolved oxygen becomes increasingly important in severe anemia, as might occur acutely with massive bleeding or in a Jehovah’s Witness who refuses to accept
transfusion. If the hemoglobin is 3 g dl−1, then raising the PaO2 from 100 mmHg to 500 mmHg will increase oxygen delivery by 33%. If blood transfusions are delayed,
this therapeutic maneuver may be life saving.
3.4.2 Hemoglobin concentration

An obvious clinical maneuver to increase oxygen transport is to raise serum hemoglobin levels. Practically, this is limited by rheologic considerations of blood viscosity
and cell deformability. Raising hemoglobin levels above 10–11% generally offers no clearcut increase in total cellular oxygen delivery and may compromise oxygen
delivery when hemoglobin rises to polycythemic levels (>18 g %) owing to increased viscosity and sludging. If serum hemoglobin levels are less than 10%, significant
increases in oxygen transport can be obtained by transfusion; whether this improves outcome remains controversial.
3.5 Determinants of oxygen consumption

Oxygen consumption is stable in patients at rest but often is increased markedly in the ICU population as a result of fever, stress, cytokine release or other alterations of
basal metabolic rate. Oxygen consumption varies in proportion to body temperature, with a 10–13% increase per degree centigrade elevation above normal6.
Likewise, VO2 decreases approximately 7% per degree centigrade below normal7. Shivering may increase VO2 by over 100%8.
The typical metabolic response to illness and injury results in significant elevations of VO2 above normal. Moderate surgical trauma produces a 10–30% increase in
VO2, a 50% increase may accompany severe infections, and as much as a 100% increase is seen with large fullthickness burns9. Typical ICU procedures such as
chest physiotherapy have been shown to increase VO2 by 35%10.
A commonly overlooked source of increased oxygen consumption involves the work of breathing. Normally, spontaneous ventilation at rest accounts for 1–3% of
resting oxygen consumption. Critically ill patients with decreased pulmonary compliance or tachypnea may utilize as much as 25–40% of oxygen transport to maintain
the energy required for spontaneous ventilation11 (Figure 3.3). Intubation and mechanical ventilation can be important tools in balancing oxygen supply and demand in
the critical patient.
3.6 Critical factors influencing cellular oxygen utilization

The most commonly recognized deficiency of cellular oxygen utilization accompanies cyanide poisoning, which uncouples oxidative phosphorylation. Impairment of
oxygen utilization is also seen in shock, trauma, ARDS, sepsis, and multiple organ failure.
Page 41
Figure 3.3 Changes in oxygen consumption (VO2) in varying clinical disease states; percentages are averages, as wide ranges may be
seen for each pathologic state.
Oxygen diffuses from the capillaries into the interstitium along a pressure gradient that decreases in direct proportion to the distance traveled. The farther oxygen has to
travel, the less available it is to cells5. When the number of open capillary beds is inappropriately low for the metabolic need, areas distant from the capillary membrane
may become hypoxic. As blood flows along the capillary, a progressive drop in capillary PO2 results.
The administration of large volumes of crystalloid solutions to adequately resuscitate critically ill patients can result in tissue edema. This edema may cause decreased
tissue PO2, which probably is the result of increased intercapillary distance. Interstitial edema, when accompanied by increased tissue pressure, may also impede
capillary blood flow. Capillary occlusion may alter capillary pattern and density, increase intercapillary distances, and influence microcirculatory shunting of blood.
Mediators of the inflammatory cascade could also initiate or aggravate this pathophysiology. Metabolic uncoupling may occur at the cellular level if endogenously
produced mediators disrupt oxidative phosphorylation.
3.7 Evaluation of tissue oxygenation

In healthy humans at rest, the quantity of oxygen delivered to all tissues is in excess of that required by metabolic demand. With stress and disease, however, oxygen
consumption may increase markedly whereas supply is limited. An essential aspect of care of the critically ill is to define the point at which tissue oxygenation fails, so
that therapy may be instituted to optimize oxygen transport.
Hemodynamic monitoring focuses on the central circulation as a means of deriving information on the adequacy of the oxygen transport system. The component parts
of this system (lung, heart, hemoglobin, vascular conduits, and cellular oxygen consumption) must be individually evaluated.
3.7.1 Mixed venous oxygen tension (Pv02)

There is currently no ‘gold standard’ for evaluation of global or regional tissue oxygenation. Mixed venous oxygen tension (PvO2) is the most reliable, single physiologic
indicator of the overall balance between oxygen supply and demand3. The driving force that transports oxygen out of the capillaries is the partial pressure gradient
between capillary PO2 and cellular PO2. Endcapillary oxygen tension reflects tissue oxygen tension and the balance between blood flow, capillary density, oxygen
content, and oxygen consumption.
PvO2 is the flowweighted average oxygen tension of the capillary effluent from the entire body, and therefore reflects wholebody supply/demand balance. Some
investigators prefer utilizing mixed venous oxygen saturation (SvO2), which when measured directly incorporates alterations in oxyhemoglobin
Page 42
dissociation and more accurately reflects the quantity of oxygen released from hemoglobin. The PvO2 and the SvO2 from a specific organ bed may be significantly
lower or higher than the global PvO2 or SvO2 measured in the pulmonary artery. For instance, the SvO2 from the heart normally is 30% whereas that from the skin is
90%. These differences are indicative of the different resting metabolic rates, proportional nonnutrient blood flow, and basal oxygen extraction between the heart and
the skin. Oxygen saturation in the superior vena cava is slightly lower than that in the inferior vena cava, which drains organs such as the spleen, kidney, and resting
mesentery with lower oxygen extraction. In shock states, inferior vena caval saturation decreases and superior vena caval saturation becomes higher than the SvO2,
emphasizing the effect of vascular redistribution12.
It is crucially important to recognize that the PvO2 is a mixture of venous effluent from all perfused beds. It may be normal in spite of small (in relation to the entire
vascular system) areas of severe ischemia. Also, PvO2 will not be affected by areas of complete nonperfusion. When flow is distributed according to metabolic need,
PvO2 and SvO2 correlate well with oxygen reserve and tissue oxygenation. If flow is maldistributed or if the microcirculation malfunctions, interpretation of PvO2 and
SvO2 as indicators of tissue oxygenation becomes problematic.
In flowdeficient states an increased oxygen extraction maintains tissue oxygenation until the oxygen reserve is exhausted. Which level of PvO2 is indicative of an
exhausted oxygen reserve and consequent tissue ischemia? All oxygenbound hemoglobin is not extractable. Available evidence indicates that 30–35% of such oxygen
is unavailable for tissue consumption13. In a group of patients with cardiopulmonary disease, Kasnitz found a threshold PvO2 of 28 mmHg below which lactic acidosis
occurred and survival decreased. In this study, PvO2 predicted lacticacidemia and death more accurately than did PaO2 or cardiac output14.
3.7.2 Lactic acid and PvO2

The presence of lactic acidosis in critically ill patients is considered a marker of poor prognosis15. In particular, the failure of serially measured lactate levels to decline
portends a bad outcome in a variety of clinical settings and evaluation of such trends in lactate clearance has been found to have greater prognostic utility than oxygen
derived variables16–18. Although lactic acidosis is typically a marker of anaerobic metabolism, not all instances of lactate elevation are necessarily associated with
tissue hypoperfusion and cellular hypoxia19. Lactate levels represent the balance of lactate production and utilization and may be altered in a variety of critical illnesses.
In sepsis, for example, increased glycolytic flux combined with impaired pyruvate utilization20 is probably one of several mechanisms giving rise to elevated lactate levels
in the absence of frank oxygen deprivation (Figure 3.4). For practical purposes, one must still first determine adequacy of resuscitation of patients with lactic acidosis
before ascribing this finding to complex metabolic alterations induced by a given disease process.
Figure 3.4 Lactic acidosis is most commonly the result of conversion of excess pyruvate to lactate in the face of anaerobic conditions
that foreclose metabolism via the Krebs cycle. Other causes of lactic acidosis include use of fructose rather than glucose as a
substrate, stimulation of phosphofructokinase (PFK)9, thiamine deficiency and cyanide toxicity.
Page 43
Simmons found the best predictor of anaerobiosis and hyperlacticacidemia to be a critical PvO2 of 27 mmHg21. Others have argued that a PvO2 of approximately
20 mmHg is the crucial level below which cellular function deteriorates, because in healthy maximally exercising humans, PvO2 reaches a plateau at 20 mmHg22.
However, the relevancy of these data to the stress of disease states is questionable. Maximal exercise is a physiologic state of progressive oxygen debt, anaerobiosis,
and increasing acidosis, which cannot be maintained for more than short periods of time. In states of depressed myocardial function, the correlation of PvO2 with lactic
acidosis has been better than the correlation of any other index, including cardiac output23.
The PvO2 and SvO2 are normally about 40 mmHg and 75%, respectively. These values indicate that under normal circumstances, 25% of the oxygen bound to
hemoglobin is consumed by normal metabolic activity. When the PvO2 drops to 28 mmHg, the saturation is 50%. Because of the steep slope of the oxyhemoglobin
dissociation curve at partial pressures of oxygen in this range, a mere 12 mmHg drop in PvO2 reflects a doubling of oxygen extraction. This additional oxygen
extraction, for practical purposes, reflects the remaining oxygen reserves below which metabolic derangement begins. Therefore, as the PvO2 decreases from 40
mmHg to 28 mmHg, progressive exhaustion of oxygen reserve occurs. PvO2 below 28 mmHg and SvO2 values below 50% imply oxygen deficits that must be
corrected if survival is expected.
A low or decreasing SvO2 is associated with a poor prognosis in respiratory failure, myocardial infarction, and traumatic shock10,14. Some evidence suggests that in
chronic hypoxemia or chronic lowflow states, unusually low PvO2 may be tolerated without gross metabolic dysfunction24, possibly as a result of enzymatic
adaptation, or enhanced microcirculatory regulation25.
3.8 Influence of loss of vasoregulation

Although the PvO2 always indicates oxygen supply/demand balance for perfused tissues, the use of PvO2 to mark a critical threshold below which anaerobic
metabolism is or is not found depends on intact vasoregulation, the absence of significant interstitial edema, normal capillarity, and the absence of microcirculatory
obstruction. Normal or high PvO2 may be misleading in disease states such as sepsis, cirrhosis of the liver, and ARDS, as well as other capillary leak syndromes that
may be associated with functional systemic shunts26. The mechanism of systemic shunting may be related to differential capillary bed flow rates, inappropriate capillary
density capillaryvenule back diffusion of oxygen, or cellular inability to consume oxygen. The shunting of capillary blood can have a profound effect on mixed venous
oxygen tension (Figure 3.5). Therefore, a normal, high or rising PvO2 may coexist with tissue oxygen deprivation and anaerobiosis in these pathologic states.
Diagnostically, an elevated or increasing PvO2 may be an early indication of disease characterized by functional systemic shunts.
Therapeutic maneuvers that increase DO2 may be judged according to their effect on VO2. A twophase relationship has been proposed, with an initial direct
relationship giving way to a plateau at the point of critical O2 delivery (Figure 3.6). The VO2 plateau despite increasing DO2 implies satisfaction of cellular oxygen
requirements. This will be discussed more fully in Section 3.10.
3.9 Continuous mixed venous oxygen saturation in clinical medicine

The capability for continuous monitoring of SvO2 has been incorporated into pulmonary artery catheters as an early indicator of significant clinical deterioration. In
patients being
Page 44
Figure 3.5 Twocompartment model of systemic circulation. Q1, cardiac output; Qs, blood flow shunted from arterial to venous circuit;
Qc, tissue capillary blood flow; CaO2, arterial oxygen content; CvO2, mixedvenous oxygen content; CcO2, oxygen content of
blood leaving the tissue compartment. As all oxygen consumption (VO2) occurs in the tissue compartment, the blood leaving
the shunt compartment has the composition of arterial blood and will have a profound effect on CvO2. From: Mill, M.J. (1982)
Tissue oxygenation in clinical medicine: an Historical Review, The International Anesthesia Research Society. Reproduced
with permission from Anesthesia and Analgesia, Lippincott, Williams & Wilkins.
weaned from intraaortic balloon pumps, changes in mixed venous oxygen saturation occur almost instantaneously with changes in the balloon inflation interval27. In
patients maintained on AV sequential pacemakers, changes to ventricular pacing that are accompanied by a significant decrease in cardiac output may be ascertained
immediately by a drop in the mixed venous oxygen saturation28. The SvO2 in hypovolemic shock will increase dramatically after a fluid bolus and thereafter will be
sustained when higher filling pressures are maintained29 (Figure 3.7). Therapeutic manipulations (e.g. ventilator changes) often not associated with hemodynamic
perturbations may be recognized within minutes when utilizing continuous SvO2 monitoring30. Cost savings
Figure 3.6 Beyond the inflection point A, oxygen consumption is independent of delivery. Increasing oxygen delivery beyond this would
not be expected to have clinical utility, Before reaching this plateau (during active resuscitation of shock patients) restoring
Oxygen delivery is a primary goal during resuscitation of critically ill patients.
Figure 3.7 Trauma victim with hypovolemic shock, showing the effect of successive fluid boluses (B) and ultimate maintenance of
normal SvO2 once volume resuscitation is complete (A). The increase in SvO2 from 50% to 75% implies that oxygen transport
doubled during this successful resuscitation.
may result from decreased utilization of arterial blood gas studies.
3.10 Goaldirected therapy with DO2 and VO2
Defining what constitutes adequate (or optimal) global DO2, the relationship of oxygen consumption (VO2) to delivery (DO2) in critical illness, and how to detect
inadequate DO2 (either globally or regionally), has been the
Page 45
subject of debate for a number of years. In most circumstances, physicians utilize trends in hemodynamic parameters (e.g. Blood Pressure (BP), Cardiac output/cardiac
index (CO/CI), PCWP), oxygen transport variables (DO2, arterial/mixed venous O2 contents), oxygen consumption variables (VO2, O2 ER), endorgan function, and
serial laboratory assessments (e.g. arterial lactate level) to arrive at a composite impression as to the adequacy of tissue oxygenation. Monitors of regional (e.g. gastric
tonometry) and cellular perfusion and oxygen utilization are being developed but have not yet found widespread application for reasons of technical feasibility and need
for further validation.
3.10.1. DO2/VO2 in critical illness

Available experimental data suggest that VO2 is normally maintained at a relatively stable and unchanging level over a wide range of normal or elevated DO2 values
(supplyindependent VO2). However, as DO2 is progressively reduced, a DO2 value is reached below which compensatory mechanisms fail and VO2 becomes
dependent on DO2 (supplydependent VO2) (Figure 3.6)31. This DO2 value is referred to as the critical (or threshold) O2delivery value. At this point, any further
reduction in DO2 results in an oxygen supplydemand mismatch with resultant widespread tissue hypoxia, conversion to anaerobic metabolism, and generation of lactic
acidosis. The point at which this occurs is difficult to predict in humans, as maintenance of ‘adequate’ global DO2 values may not preclude regional oxygen supply
demand imbalances.32 Few human studies have actually identified the critical DO231,33, and cellular dysoxia as well as lactate generation in critical illness probably
involve mechanisms beyond simple disturbances in bulk O2 flow19,20.
Certain disease states, such as ARDS and sepsis, may raise the critical value of DO2 and alter the DO2–VO2 relationship and cellular oxygen utilization, resulting in a
state of ‘pathological’ deliverydependent VO2 with the risk of concomitant ‘occult’ tissue hypoxia
Figure 3.8 The oxygen delivery (DO2) and oxygen consumption (VO2) relationship In normal (continuous line) and pathologic (dashed
line) conditions. Under normal conditions, as DO2 decreases, VO2 ramains relatively constant until the critical VO2 (dashed
vertical line), at which point VO2 decreases as DO2. In contrast, in pathologic conditions, pathologic dependence of VO2 and
DO2 is characterized by higher critical DO2 and a much wider range of dependence of VO2 on DO2.
in the face of normal global DO2 values (Figure 3.8)34,35. Furthermore, observational data36 suggest that survivors of surgical critical illness manifest significantly
elevated levels of DO2 and VO2. Several investigators have explored the effect of interventions aimed at achieving such ‘supranormal’ (or ‘survivor’) hemodynamic
patterns and DO2 values on outcome in a variety of clinical settings37–42 based on assumptions that critical illness induces pathologic supplydependent oxygen
consumption and that raising DO2 to supranormal values prevents or reverses occult tissue hypoxia. In highrisk surgical patients the results of such an approach
instituted preoperatively are associated with a survival advantage43,44. Interventions to raise DO2 included (depending on study design) the use of fluids, blood
products, and catecholamines to achieve predetermined hemodynamic and/or DO2 endpoints (most commonly CI >4.5 l m−2 Body surface area (BSA), DO2 >600
ml min−1 m−2).
The results of the trials of supranormal DO2 in mixed populations of critically ill patients have suggested either no benefit37,42 or even a higher mortality in the
treatment group as compared with the control group40.
Page 46
Metaanalysis of available trials involving supranormal DO2 confirms lack of efficacy of this intervention if there is a delay between the insult and institution of therapy,
but also concedes that there may be some benefit to a select group of patients in whom it is begun preoperatively45. It is of interest that these trials have suffered from
significant ‘crossover’ of patients between study groups where ‘control’ patients spontaneously achieved the study patients’ supranormal hemodynamic values, and
‘study’ patients were unable to reach the hemodynamic goals expected of them by study design. Such crossovers tend to obscure any real treatment effect (should it
exist) but may simply reflect a difference in cardiopulmonary reserve between survivors and nonsurvivors, as a better outcome is associated with the ability to sustain
an elevated CI/DO2 (regardless of assignment to study or control group). Further methodologic problems found in studies evaluating supranormal DO2 include
inconsistencies in study design involving the timing of interventions, lack of standardization of cointerventions, use of retrospective subgroup analysis as well as lack of
doubleblinded design39,46.
3.11 Summary
A primary goal of critical care medicine is to optimize oxygen delivery to cells and tissues. Based on available data from trials involving supranormal DO2 strategies,
there is insufficient evidence to support using predetermined supraphysiologic hemodynamic or oxygen transfer values as empiric endpoints of therapy. Mixed venous
oxygen remains a valuable indicator of the balance between global oxygen delivery and utilization in most clinical situations. In general, a low or decreasing SvO2
indicates an imbalance between oxygen consumption and oxygen delivery and warrants investigation. A stable SvO2 suggests that global oxygen transport is balanced
to oxygen consumption. However, no single laboratory or clinical data point can be taken in isolation as an indication that the oxygen supplydemand relationship is
ideal or adequate. Rather, a thorough assessment of the adequacy of perfusion and oxygen delivery requires integration of serial physical examination findings with
trends in hemodynamic parameters, oxygen consumption variables, endorgan function, and laboratory data such as lactate level.
Essential equations
Page 47
References
1. Finch, C.A. and Lenfant, C. (1972) Oxygen transport in man. N. Engl J. Med. 286:407.
2. Ultman, J.S. and Bursztein, A. (1981) Analysis of error in the determination of respiratory gas exchange at varying FiO2. J. Appl. Physiol. 50: 210–216.
3. Browning, J.A., Linbert, S.E., Turney, S.Z. and Chodoff, P. (1982) The effects of a fluctuating FiO2 on metabolic measurements in mechanically ventilated
patients. Crit. Care Med. 10: 82–85.
4. Kandel, G. and Aberman, A. (1983) Mixed venous oxygen saturation: its role in the assessment of the critically ill patient. Arch. Intern. Med. 143:1400–1402.
5. Woodson, R.D. (1979) Physiological significance of oxygen dissociation curve shifts. Crit. Care Med. 7:368.
6. Beisel, W., Wannemacher, R.W. and Neufield, H.A. (1980) Relation of fever to energy expenditure. In: Assessment of Energy Metabolism in Health and
Disease. Ross Laboratories, Columbus, OH, p. 144.
7. Harris, E.A., Scalye, E.R. and Squire, A.W. (1971) Oxygen consumption during cardiopulmonary bypass with moderate hypothermia in man. Br. J. Anaesth.
43:1113.
8. Rodriquez, J.L., Weissman, L., Damask, M.C., et al. (1983) Physiologic requirements during rewarming: suppression of the shivering response. Crit. Care Med.
11:490.
9. Kinney, J.M., Duke, J.H., Long, C.L., et al. (1970) Tissue fuel and weight loss after injury. J. Clin. Pathol 23(Suppl. 4):65–72.
10. Weissman, C., Kemper, M., Damask, M.C., et al. (1984) Effect of routine intensive care interactions on metabolic rate. Chest 86: 815–818.
11. Aubier, M., Viires, N., Syllie, G., et al. (1982) Respiratory muscle contribution to lactic acidosis in low cardiac output. Am. Rev. Respir. Dis. 126:648.
12. Lee, J., Wright, F., Barber, R., et al. (1972) Central venous oxygen saturation in shock. Anesthesiology 36:472.
13. BryanBrown, C., Back, S., Makabali, G., et al. (1973) Consumable oxygen availability of oxygen in relation to oxyhemoglobin dissociation. Crit. Care Med. 1:17.
14. Kasnitz, P., Druger, G.L., Torra, F. and Simmons, D.H. (1976) Mixed venous oxygen tension and hyperlactatemia. Survival in severe cardiopulmonary disease.
JAMA 236:570.
15. Cady, L.D., Jr, Weil, M.H., Afifi, A.A., et al. (1973) Quantitation of severity of critical illness with special reference to blood lactate. Crit. Care Med. 1:75.
16. Singarajah, C. and Carlson, R.W. (1998) A review of the role of blood lactate measurements in the ICU. J. Intensive Care Med. 13: 218–228.
17. Friedman, G., Berlot, G., Kahn, R.J. and Vincent, J.L. (1995) Combined measurements of blood lactate concentrations and gastric intramucosal pH in patients
with severe sepsis. Crit. Care Med. 23:1184–1193.
18. Bakker, J., Coffernils, M., Leon, M., Gris, P. and Vincent, J.L. (1991) Blood lactate levels are superior to oxygenderived variables in predicting outcome in
human septic shock. Chest 99:956–962.
19. Simmons, D.H., Alpas, A.P., Tashkin, D.P. and Coulson, A. (1978) Hyperlactatemia due to arterial hypoxemia or reduced cardiac output, or both. Am. Phys. Soc.
45(2):195.
20. Bevegard, S., Holmgren, A. and Johnsson, B. (1960) The effect of body position on the circulation at rest and during exercise, with special reference to the
influence on the stroke volume. Acta Physiol. Scand. 29:279.
21. De La Roche, A.G., Edmonds, J.F., Williams, W.G., et al. (1978) Importance of mixed venous oxygen saturation in the care of critically ill patients. Can. J. Surg.
21:227.
22. Hotchkiss, R.S. and Karl I.E. (1992) Reevaluation of the role of cellular hypoxia and bioenergetic failure in sepsis. JAMA 267:1503–1510.
23. Fink, M.P. (2001) Cytopathic hypoxia—mitochondrial dysfunction as mechanism contributing to organ dysfunction in sepsis. Crit. Care Clinics 17(1):219–237.
24. Schlichtig, R., Cowden, W.L. and Chaitman, B.R. (1986) Tolerance of unusually low mixed venous oxygen saturation: adaptations in the chronic low cardiac output
syndrome. Am. J. Med. 80:813–818.
25. Robin, E.D. (1980) Of men and mitochondria: coping with hypoxic dysoxia. Am. Rev. Respir. Dis. 122:617.
26. Danek, S.J., Lynch, J.P., Weg, J.G., et al. (1980) The dependence of oxygen uptake on oxygen delivery in the adult respiratory distress syndrome. Am. Rev.
Respir. Dis. 122:387.
27. Gore, J. (1984) Use of continuous monitoring of mixed venous saturation in the CCU. Chest 86:757.
28. Baele, P.L., McMichan, J.C., March, H.M., et al. (1982) Continuous monitoring of mixed venous oxygen saturation in critically ill patients. Anesth. Analg. 61:513.
29. McMichan, J.C. (1983) Continuous monitoring of mixed venous oxygen saturations. In:
Page 48
Continuous Measurement of Blood Oxygen Saturation in the High Risk Patient, Vol. 1 (ed. J.F. Schweiss). Beach International Inc., San Diego, CA: 1–29.
30. Nelson, L.D. (1987) Continuous venous oximetry in surgical patients. Ann. Surg. 203:329–333.
31. Ronco, J.J., Fenwick J.C., Tweeddale M.G., et al. (1993) Identification of the critical oxygen delivery for anaerobic metabolism in critically ill septic and nonseptic
humans. JAMA 270(14): 1724–1730.
32. Marik, P.E. (1993) Gastric intramucosal pH, a better predictor of multiorgan dysfunction syndrome and death than oxygenderived variables in patients with sepsis.
Chest 104:225–229.
33. Shibutani, K., Komatsu T., Kubal K., et al. (1983) Critical level of oxygen delivery in anesthetized man. Crit. Care Med. 11:640–643.
34. Astiz, M.E., Rackow E.C., Falk J.L., et al. (1987) Oxygen delivery and consumption in patients with hyperdynamic septic shock. Crit. Care Med. 15(1):26–28.
35. Danek, S.J., Lynch J.P., Weg J.G., and Dantzker D.R. (1980) The dependence of oxygen uptake on oxygen delivery in the adult respiratory distress syndrome.
Am. Rev. Resp. Dis. 122: 387–395.
36. Shoemaker, W.C., Montgomery, E.S., Kaplan, E. and Elwyn, D.H. (1973) Physiologic patterns in surviving and nonsurviving shock patients: use of sequential
cardiorespiratory variables in defining criteria for therapeutic goals and early warning of death. Arch. Surg. 106:630–636.
37. Alia, I., Esteban A., Gordo F., et al. (1999) A randomized and controlled trial of the effect of treatment aimed at maximizing oxygen delivery in patients with severe
sepsis or septic shock. Chest 115:453–461.
38. Hayes, M.A., Timmins, A.C., Yau, E.H.S., Palazzo, M., Watson, D. and Hinds, C.J. (1997) Oxygen transport patterns in patients with sepsis syndrome or septic
shock: Influence of treatment and relationship to outcome. Crit. Care Med. 25:926–936.
39. Tuchschmidt, J., Fried, J., Astiz, M. and Rackow, E. (1992) Elevation of cardiac output and oxygen delivery improves outcome in septic shock. Chest 102:216–
220.
40. Hayes, M.A., Timmins, A.C., Yau, E.H.S., Palazzo, M., Hinds, CJ. and Watson, D. (1994) Elevation of systemic oxygen delivery in the treatment of critically ill
patients. N. Engl. J. Med. 330:1717–1722.
41. Yu, M., Burchell, S., Hasaniya, N., et al. (1998) Relationship of mortality to increasing oxygen delivery in patients (50 years of age: A prospective, randomized trial.
Crit. Care Med. 26: 1011–1019.
42. Gattinoni, L., Brazzi L., Pelosi P., et al. (1995) A trial of goaloriented hemodynamic therapy in critically ill patients. N. Engl J. Med. 333: 1025–1032.
43. Boyd, O., Grounds, M.R. and Bennett, D.E. (1993) A randomized clinical trial of the effect of deliberate perioperative increase of oxygen delivery on mortality in
highrisk surgical patients. JAMA 270:2699–2707.
44. Shoemaker, W.C., Appel, P.L., Kram, H.B., Waxman, K. and Lee, T. (1988) Prospective trial of supranormal values of survivors as therapeutic goals in high
risk surgical patients. Chest 94:1176–1186.
45. Heyland, D., Cook, D., King, D., et al. (1996) Maximizing oxygen delivery in critically ill patients: a methodologic appraisal of the evidence. Crit. Care Med.
24:517–524.
46. Yu, M., Levy, M.M., Smith, P., Takiguchi, S.A., Miyasaki, A. and Myers, S.A. (1993) Effect of maximizing oxygen delivery on morbidity and mortality rates in
critically ill patients: a prospective, randomized, controlled study. Crit. Care Med. 21:830–838.
Page 49
Chapter 4
Acidbase disorders
Ali AlKhafaji, MD,
Marcus J.Hampers, MD and
Howard L.Corwin, MD
Contents
4.1 Introduction
4.2 Basic concepts and definitions
4.3 Approach to acidbase abnormalities
4.4 Metabolic acidosis
4.5 Anion gap metabolic acidosis
4.6 Normal anion gap metabolic acidosis
4.7 Respiratory acidosis
4.8 Metabolic alkalosis
4.9 Respiratory alkalosis
Page 50
4.1 Introduction
Acidbase abnormalities are common in the intensive care unit. Most metabolic and physiologic processes in the body require the pH to be within a narrow range of
7.35–7.45. To achieve a normal acidbase balance, the interaction between the respiratory, renal and buffering systems should be intact. When there is an acidbase
abnormality, the buffer system will be the first to respond by altering hydrogen ions. The lungs will remove the carbon dioxide and the kidneys will excrete acidic urine.
This is a very wellorchestrated process, the net result of which will be a normal acidbase balance.
4.2 Basic concepts and definitions
Normal arterial blood gas values are pH 7.35–7.45, PCO2 35–45 mmHg, and HCO3 22–26 mequiv l−1.
There are four primary acidbase disorders:
• metabolic acidosis: decreased pH and decreased HCO3;
• respiratory acidosis: decreased pH and increased PCO2;
• metabolic alkalosis: increased pH and increased HCO3;
• respiratory alkalosis: increased pH and decreased PCO2.
4.2.1 Compensatory response (Figure 4.1)

For each acidbase disorder, there is a compensatory response mediated by the kidneys or the lungs that tends to bring the pH back towards normal (Table 4.1). It is
important to note that compensation does not fully correct the primary acidbase disorder.
4.2.2 Buffer system

A buffer is any substance that can reversibly bind to H+ to neutralize the effect:
A free H+ combines with a buffer to form a weak acid that, depending on the concentration of H+, can remain as an unassociated complex or dissociate back to H+
and buffer. In plasma and interstitial fluid, bicarbonate
Figure 4.1 Acidbase nomogram can be used to predict pH changes with changes in bicarbonate or PCO2. Reprinted with permission
from Mark Graber, M.D., VA Medical Center, Northport, NY. In: The Kidney Kard 3rd Edition (1995).
Page 51
Table 4.1 Causes of respiratory alkalosis
Primary disorder pH Compensatory response

Metabolic acidosis ( HCO3) <7.35 PCO2
Metabolic alkalosis ( HCO3) >7.45 PCO2
Respiratory acidosis ( PCO2) <7.35 HCO3
Respiratory alkalosis ( PCO2) >7.45 HCO2
(HCO3), proteins and phosphate compounds are the major buffer system, whereas in the red blood cells, hemoglobin serves as the major buffer. An acidbase buffer
system works to prevent changes in pH when either acid or base is added. The ratio of acid to base, along with an equilibrium constant (pK), defines the free H+
concentration of the solution. This relationship is expressed by the HendersonHasselbalch equation: pH=pK+ log (base/acid).
The pK of a buffer system identifies the pH at which the concentration of acids and bases in that system is equal. In acidosis, the base member of the buffer pair will
accept an H+, altering the ratio of acidbase leading to an increase in pH. In alkalosis, on the other hand, an acid member of the buffer pair donates H+, altering the
acidbase ratio leading to a decrease in pH.
4.2.3 Role of the respiratory system

Carbon dioxide (CO2) is formed continually in the body by different intracellular metabolic processes. PCO2 plays an important role in the acidbase regulation.
Referring to the HendersonHasselbalch equation, an increase in PCO2 decreases the pH whereas a decrease of PCO2 increases the pH. By adjusting the PCO2 up or
down, the lungs can effectively regulate the pH.
4.2.4 Role of the renal system

The kidneys control acidbase balance by excreting either acidic or basic urine. There are three mechanisms by which the kidneys regulate acidbase balance:
reabsorption of filtered bicarbonate (HCO3−) ions, generation of new bicarbonate ions and secretion of hydrogen (H+) ions.
4.2.5. Anion gap

The anion gap (AG) represents the difference between the major plasma cations and anions, and reflects usually unmeasured anions such as sulfate. The normal value of
AG is 8–16 mequiv l−1;
The calculation of the AG is essential in the workup of acidbase disorders. Anion gap is decreased in hypoalbuminemia, multiple myeloma, and bromide intoxication.
The anion gap is increased in some types of metabolic acidosis (see below).
4.2.6 Urinary anion gap

The urinary anion gap (UAG) is expected to be negative in patients with metabolic acidosis. A positive UAG in a patient with acidosis reflects a problem with urinary
acidification ability of the kidneys. The UAG is calculated as follows:
4.2.7 Osmolal gap

The normal plasma osmolality ranges from 280 to 295 mOsm kg−1 The calculated plasma osmolality accounts for the most prevalent solutes contributing to the
osmolality:
where glucose and BUN are measured in mg dl−1. If the measured plasma osmolality exceeds the calculated plasma osmolality by more than 10 mOsm kg−1 then
unmeasured solutes are present. A high osmolal gap is seen in methanol and ethylene glycol intoxication.
Page 52
4.3 Approach to acidbase abnormalities

In analyzing acidbase abnormalities, the following approach is helpful:
• Determine the pH: acidosis when pH is less than 7.35 versus alkalosis when pH is greater than 7.45.
• Calculate the anion gap. If the anion gap is high, then this suggests certain types of metabolic acidosis (see Table 4.2).
• Determine the PCO2. PCO2 is high with respiratory acidosis or as compensation to metabolic alkalosis. PCO2 is low with respiratory alkalosis or as compensation to
metabolic acidosis.
• Determine the HCO3. HCO3 is high with metabolic alkalosis or as compensation to respiratory acidosis. HCO3 is low with metabolic acidosis or as compensation to
respiratory alkalosis.
• Assess compensatory response to every acidbase abnormality. If the appropriate compensatory response is not present, a mixed acidbase disorder is present.
• Remember that the compensatory response should not fully correct the primary acidbase disorder. Overcompensation suggests a mixed acidbase disorder.
4.4 Metabolic acidosis
Metabolic acidosis is characterized by low pH (<7.35) and low HCO3 (<22 mequiv l−1). Metabolic acidosis develops by the addition of acid, loss of a base, or the
failure of the kidneys to excrete sufficient net acid to replenish the HCO3 used to buffer acids. Metabolic acidosis can be either anion gap or nonanion gap. Causes of
metabolic acidosis are listed in Table 4.2.
4.4.1 Compensatory mechanism

The compensatory response to metabolic acidosis is hyperventilation, which results in a decrease in PCO2. For each 1 mequiv decrease in HCO3− the PCO2 will fall
by 1–1.5 mmHg.
Table 4.2 Causes of metabolic acidosis
Increased anion gap metabolic acidosis

Ketoacidosis including diabetic, alcoholic and starvation
Lactic acidosis
Uremia
Salicylate intoxication
Methanol intoxication
Ethylene glycol intoxication
Paraldehyde intoxication
Normal anion gap metabolic acidosis
Diarrhea
Urinary diversion
Ileostomy
Pancreatic fistula
Surgical drainage
Cholestyramine
Renal tubular acidosis
Normal saline infusion
Carbonic anhydrase inhibitors
Following respiratory alkalosis
Parenteral nutrition
4.5 Anion gap metabolic acidosis
4.5.1 Diabetic ketoacidosis (Figure 4.2)

Diabetic ketoacidosis is characterized by the presence of hyperglycemia, low insulin levels, hyperglucagonemia, high catecholamine levels and, subsequently, the
overproduction of ketoacids (acetoacetic acid and βhydroxybutaric acid) and metabolic acidosis. Patients may present with hyperventilation, vomiting, abdominal
pain and sometimes shock and coma. Hyponatremia, hyperkalemia, hyperosmolarity, ketonurea and elevated BUN and creatinine are often present. The treatment of
diabetic ketoacidosis consists of replacing insulin by i.v. infusion, hydration with normal saline, and correcting electrolyte abnormalities. Replacing insulin stops excessive
gluconeogenesis and also stops mobilization of fatty acids, which are the precursors of ketones. The need for bicarbonate in the treatment of metabolic acidosis in
diabetic ketoacidosis has been debated and, if used at all, should be reserved for patients with severe metabolic acidosis despite adequate hyperventilation.
Page 53
Figure 4.2 Diagnostic algorithm for metabolic acidosis with elevated anion gap. Adapted from Breyer, M.D. and Jacobson, H.R. (1989)
Approach to the patient with metabolic acidosis or metabolic alkalosis. In: Textbook of Internal Medicine (ed. W.Kelley), p.
926. With permission from Lippincott, Williams & Wilkins.
4.5.2 Alcoholic ketoacidosis

Alcoholic ketoacidosis occurs in alcoholics who usually have poor nutrition. This is probably due to insulin deficiency caused by hypoglycemia induced by fasting.
Patients may have had several episodes of vomiting and might also present with metabolic alkalosis. The diagnosis of alcoholic ketoacidosis is made when the patient
presents with anion gap metabolic acidosis, history of chronic alcohol use, ketonemia, high osmolar gap and variable plasma glucose levels1. The treatment for alcoholic
ketoacidosis consists of dextrose infusion and fluid deficit correction2. Treatment with bicarbonate is not required for this condition.
4.5.3 Starvation ketoacidosis

Starvation ketoacidosis occurs because of a switch to lipolysis and gluconeogenesis as the sole source of calories. In starvation, as in patients with alcoholic
ketoacidosis, patients have low insulin levels induced by hypoglycemia and the associated increase in glucagon, epinephrine, cortisol and growth hormone stimulate
lipolysis with release of fatty acids. Treatment will include proper nutrition and a dextrose infusion. Treatment with bicarbonate is not required for this condition.
4.5.4 Lactic acidosis

Lactic acid is produced by most tissues and carried to the liver to be oxidized and converted back to glucose via gluconeogenesis. Lactic acidosis can be divided into
type A and type B. The causes of each are listed in Table 4.3.
Type A lactic acidosis is generally due to tissue hypoperfusion from any cause such as seen in cardiogenic, hypovolemic, and septic shock. On the other hand, type
B is not associated with tissue hypoperfusion and is usually due to
Page 54
Table 4.3 Causes of lactic acidosis
Type A Type B
Shock Drugs and toxins
Severe hypoxia Epinephrine, norepinephrine
Convulsions Salicylates
Severe heart failure Ethanol, methanol and ethylene glycol
Severe anemia Biguanides
Vigorous exercise
Cyanide poisoning Diseases
Hypothermic shivering Diabetes
Renal failure, hepatic failure
Malignancies
Iron deficiency
Severe infection
impaired lactate clearance as a result of hepatic failure. Patients with type A lactic acidosis have a worse prognosis as compared with patients with type B. A distinctive
type of lactic acidosis called Dlactic acidosis occurs as a result of bacterial overgrowth in patients with small bowel resection or patients with jejunoileal bypass3.
Serum lactate can be used as a prognosticator in patients with shock: with serum lactate above 9 mmol l−1, the mortality is more than 75%4.
Treatment of lactic acidosis consists of treating the underlying problem. Patients with shock should receive adequate fluid resuscitation, antibiotics and vasopressor
support.
The use of bicarbonate therapy in lactic acidosis is controversial5,6. There is no evidence that the use of bicarbonate in the treatment of lactic acidosis improves
outcome. In fact, its use may have several side effects, which include hyperosmolality7, hypernatremia7, volume overload7, hypocalcemia8, worsening intracellular
acidosis9,10 and transient increase in intracranial pressure11.
Some argue against the use of bicarbonate regardless of the pH value12. Other buffer solutions that potentially can be used in the treatment of metabolic acidosis
include the following.
Carbicarb
This is a mixture of sodium bicarbonate and disodium carbonate in a 1:1 ratio. Carbicarb does not produce the significant increase in PCO2 seen with the use of sodium
bicarbonate and thus might not cause worsening intracellular acidosis. The risks of hypervolemia, hyperosmolality and hypernatremia are the same as when sodium
bicarbonate is used. Carbicarb has not been extensively evaluated for human use.
Tromethamine
Tromethamine (THAM) is an amino alcohol that buffers acids and CO2 by its amine (NH2). THAM has a buffering capacity without the generation of CO2 and thus
does not worsen intracellular acidosis. THAM toxicity causes hypoglycemia, respiratory depression and hyperkalemia. THAM has been used to treat severe acidosis
caused by sepsis, diabetic ketoacidosis and renal tubular acidosis13. THAM use remains unproven in the treatment of metabolic acidosis.
Dichloroacetate
Dichloroacetate (DCA) is a buffer compound that increases the activity of pyruvate dehydrogenase and thus promotes the clearance of lactate. A large, multicenter,
placebocontrolled trial in patients with lactic acidosis failed to show improving hemodynamics or outcome with the use of DCA14.
4.5.5 Renal failure

Renal failure is usually associated with metabolic acidosis as a result of the inability of the kidney to produce ammonia. There is
Page 55
retention of sulfate and phosphate, leading to an increased anion gap. The aim should be to keep the pH and the HCO3 level near normal to limit progressive bone
disease. The acidosis is usually nonprogressive and relatively asymptomatic. Oral sodium bicarbonate can be used as a treatment for the acidosis.
4.5.6 Salicylate intoxication

Acute salicylate intoxication causes two acidbase abnormalities, respiratory alkalosis and an anion gap metabolic acidosis. Respiratory alkalosis occurs as a result of
hyperventilation caused by direct effect of salicylate on the respiratory center. Toxicity is considered mild after acute ingestion of less than 150 mg kg−1, moderate after
ingestion of 150–300 mg kg−1, and severe after ingestion of 300 mg kg−1 or more. The diagnosis of salicylate intoxication is based upon obtaining a history, performing
a physical examination, and obtaining laboratory testing. Patients may present with hyperventilation, fever, nausea, vomiting, dehydration, tinnitus, vertigo, stupor and
coma. The laboratory tests might show hyperglycemia or hypoglycemia, hypokalemia, hypouricemia, coagulation abnormalities and ketosis. Salicylate levels above 100
mg dl−1 indicate severe intoxication. Treatment for salicylate intoxication consists of the following:
• supportive measures: these include cooling blankets for the fever, hydration, close monitoring of the pH, coagulation indices and replacement of potassium;
• elimination of salicylate:
(i) induction of emesis and administration of charcoal;
(ii) alkaline diuresis: indicated for salicylate levels above 40 mg dl−1 and achieved by the use of 100 mequiv of sodium bicarbonate in 1 l of dextrose water infused
at a rate of 10–15 ml kg−1 h−1; the goal of therapy is a urine pH of 7–8;
(iii) hemodialysis: indicated when the serum salicylate level is above 100 mg dl−1 6 h after ingestion, refractory acidosis, seizure, renal failure and pulmonary edema.
4.5.7 Methanol intoxication

Methanol is metabolized to formic acid via alcohol dehydrogenase. Formic acid is usually responsible for the anion gap metabolic acidosis. Methanol metabolism is
represented as follows:
Patients with methanol intoxication may present with blurred vision, lethargy, tachypnea, confusion, nausea, vomiting, depressed mental status, and, in severe
intoxication, convulsions and coma. The diagnosis is made by the history of ingestion, an anion gap metabolic acidosis, high osmolal gap, and the finding of a high blood
methanol level.
The treatment of methanol toxicity consists of:
• Gastric lavage and charcoal if the patient presents soon after ingestion.
• Increasing the metabolism of formic acid by using folinic acid 50 mg i.v. followed by folic acid 1 mg kg−1 i.v. every 4 h for six doses.
• Decreasing the metabolism of methanol to formic acid by the administration of ethanol. Ethanol will compete with methanol for alcohol dehydrogenase. A 10%
ethanol solution is administered at a rate of 8–10 ml kg−1 and should be titrated to achieve a blood alcohol level of 100–150 mg dl−1. Ethanol infusion should be
continued until the methanol level is below 10 mg dl−1 and there is resolution of the anion gap and the metabolic acidosis.
• Fomepizole (4methylpyrizole), an alcohol dehydrogenase antagonist, can be used instead of ethanol. Fomepizole 15 mg kg−1 i.v. is followed by 10 mg kg−1 i.v.
every 12 h for four doses. This is followed by 15 mg kg−1 i.v. every 12 h until the methanol level is less than 20 mg dl−1.
• Correction of metabolic acidosis using sodium bicarbonate 1–3 mequiv kg−l infused i.v. to achieve normal pH.
Page 56
• Hemodialysis is indicated when the methanol level is more than 50 mg dl−1 or in the presence of renal failure and persistent visual symptoms.
4.5.8 Ethylene glycol intoxication

Ethylene glycol is metabolized to glycoldehyde via alcohol dehydrogenase. Glycoldehyde is further metabolized to toxic compounds such as formic acid, oxalic acid,
glycine, and ketoaldehyde. Patients present with lethargy, coma, convulsions, tachypnea, tachycardia, and renal failure. Hyperkalemia, hypocalcemia, anion gap
metabolic acidosis, and a high osmolal gap are usually found. Diagnosis is made by a history of ingestion of ethylene glycol and the presence of oxalate crystals in urine.
Treatment consists of the following:
• Gastric lavage if the patient presents soon after ingestion.
• Fomepizole can be used, using the same dose as in methanol intoxication. It should be continued until the ethylene glycol level is less than 20 mg dl−1.
• Ethanol infusion to compete for alcohol dehydrogenase as used for methanol.
• Correcting the metabolic acidosis with sodium bicarbonate as with methanol toxicity.
• Pyridoxine 100 mg IV qd and thiamine 100 mg i.v. qd.
• Hemodialysis is indicated when the ethyl ene glycol level is more than 50 mg dl−1, and there is persistent acidosis and renal failure.
4.6 Normal anion gap metabolic acidosis (Figure 4.3)

It is best to classify normal anion gap metabolic acidosis into three groups.
4.6.1 Acidosis owing to bicarbonate loss

Examples include diarrhea, fistula drainage, and villous adenoma. These are characterized by a negative urinary anion gap. Treatment with fluid, bicarbonate, and
potassium is recommended as needed.
4.6.2 Addition of acid

In this group, the addition of acid (hydrochloric) precursors such as ammonium chloride and arginine hydrochloride produces acidosis. Administration of arginine,
histidine, or lysine in parenteral nutrition may result in nonanion gap acidosis.
4.6.3 Renal tubular acidosis (RTA)

Renal tubular acidosis is divided into the following.
Type 1 or distal RTA

This is characterized by the inability to maximally acidify the urine regardless of the level of serum HCO3. Distal RTA is a primary idiopathic disorder although it may
have causes as shown in Table 4.4.
The diagnosis is made when there is a nonanion gap metabolic acidosis with inability to acidify the urine and the presence of hypokalemia.
Treatment consists of sodium bicarbonate (1–3 mequiv kg−1 day−1).
Type II or proximal RTA

This is characterized by reduction of the renal capacity to reabsorb HCO3. It can be an isolated defect or part of Fanconi syndrome (glycosurea, aminoacidurea,
hypouricemia and hypophosphatemia). Secondary causes of proximal RTA are listed in Table 4.5.
Patients with proximal RTA rarely develop severe acidosis and plasma bicarbonate is usually maintained at levels higher than 15 mequiv l−1. Diagnosis is made as
follows. After administration of i.v. sodium bicarbonate to
Table 4.4 Causes of distal RTA
Amphotericin B
Lithium
Toluene
Analgesics use
Nephrocalcinosis
Chronic infection
Page 57
Figure 4.3 Diagnostic algorithm for hyperchloremic metabolic addosis. Adapted from Breyer, M.D. and Jacobson, H.R. (1989)
Approach to the patient with metabolic acidosis or metabolic alkalosis. In: Textbook of Internal Medicine (ed. W.Kelley).
With permission from Lippincott, Williams & Wilkins.
reach a normal plasma bicarbonate level, urine pH and fractional excretion of bicarbonate is measured. In proximal RTA, urine pH is greater than 7.0 and fractional
excretion of bicarbonate is greater than 15%. Treatment consists of bicarbonate therapy (10–15 mequiv kg−1 day−1), and potassium supplementation and thiazide
diuretics may be used.
Type IV RTA (hyporeninemic hypoaldosteronism)

Aldosterone promotes potassium and hydrogen secretion and sodium reabsorption in distal tubules. This disorder is of unknown etiology and is characterized by the
presence of hyperkalemia, low plasma aldosterone, low renin level and a normal gap metabolic
Table 4.5 Causes of proximal RTA
Carbonic anhydrase inhibitors

Lead
Outdated tetracycline
Multiple myeloma
Amyloidosis
SLE, Sjögren’s syndrome
Chronic active hepatitis
Page 58
Figure 4.4 Diagnostic algorithm for metabolic alkalosis. Adapted from Breyer, M.D. and Jacobson, H.R. (1989) Approach to the patient
with metabolic acidosis or metabolic alkalosis. In: Textbook of Internal Medicine (ed. W.Kelley). With permission from
Lippincott, Williams & Wilkins.
acidosis. Acidosis develops as a result of impaired acid excretion from decreased ammonia formation secondary to hyperkalemia and aldosterone deficiency. In many
patients, the hyperkalemia is mild and well tolerated. The liberal use of salt intake combined with diuretic, potassium restriction, sodium bicarbonate 1–3 mequiv kg−1
day−1 and fludrocortisone 0.1–0.5 mg p.o. day−1 is considered the treatment of choice.
4.7 Respiratory acidosis

Primary respiratory acidosis is a condition characterized by low pH (<7.35) and high PCO2. Respiratory acidosis can occur with any condition that causes decreased
alveolar ventilation. Respiratory acidosis can be acute or chronic and the causes are listed in Table 4.6.
Patients with acute respiratory acidosis present with decreased mentation, restlessness, agitation, headache, tachycardia, tremors, and, in severe cases, coma.

The compensatory response to respiratory acidosis is achieved through increased bicarbonate reabsorption and increased ammonia excretion by the kidney. The
compensation for respiratory acidosis is predictable as follows:
• acute respiratory acidosis: HCO3 will increase by 1 mequiv 1–1 for each 10 mmHg increase of PCO2 above 40 mmHg;
Page 59
Table 4.6 Causes of respiratory acidosis
Airway obstruction, acute or chronic

Respiratory center depression
Neuromuscular diseases
Mechanical underventilation
Circulatory collapse
Restrictive defects
• chronic respiratory acidosis: HCO3 will increase by 3.5 mequiv l−1 for each 10 mmHg increase of PCO2 above 40 mmHg.
The renal compensation for respiratory acidosis usually occurs over several days and in general, the compensated serum bicarbonate does not exceed 35 mequiv l−1.
Treatment for respiratory acidosis is directed towards correcting the primary cause and improving alveolar ventilation. The use of bicarbonate is rarely indicated.
4.8 Metabolic alkalosis (Figure 4.4)

Metabolic alkalosis is a condition characterized by high pH (>7.45) and a high HCO3 (>26). Because there are few symptoms associated with mild to moderate
metabolic alkalosis, treatment is generally not required urgently. However, when the plasma pH rises above 7.55, treatment should be initiated. Metabolic alkalosis
results from three general processes: (i) loss of hydrogen ions from either the stomach or kidney; (ii) exogenous bicarbonate administration; (iii) loss of extracellular
bicarbonate free fluid, socalled ‘contraction alkalosis’. Common causes are listed in Table 4.7.
Metabolic alkalosis is a relatively common problem in the ICU. Before initiating treatment, it is helpful to classify the metabolic alkalosis as either chloride resistant or
chloride sensitive. Chloride responsive alkaloses are those that result from the concomitant loss of chloride with hydrogen. This is observed most commonly with
gastrointestinal or renal losses of hydrogen ions (listed above).
Table 4.7 Causes of metabolic alkalosis
Gastrointestinal losses of hydrogen ions

Vomiting
Gastric suctioning
Renal hydrogen ion losses
Loop or thiazide diuretics
Mineral corticoid excess
Renal compensation for chronic respiratory acidosis
Hypercalcemia
Hypokalemia
Alkali loading
HCO3 administration
Excessive antacid use
Although the history is usually sufficient to suggest the cause of a metabolic alkalosis, and whether or not it will be chloride responsive, the diagnosis can be established
by measuring urinary chloride concentration. Generally, if the urine chloride is less than 20 mequiv l−1, the alkalosis is chloride responsive. Conversely, if the urine
chloride is greater than 20 mequiv l−1, the alkalosis is generally chloride resistant. Common causes are listed in Table 4.8.
Whenever possible, the underlying cause for the metabolic alkalosis should be addressed (e.g. continuous nasogastric suctioning, diuresis). The majority of chloride
responsive alkaloses may be treated with normal saline and potassium chloride. Such therapy corrects volume contraction (if present), and repletes chloride and
potassium stores. The rate of administration of normal saline should be based upon the clinical evaluation of the patient’s volume status. The administration of potassium
chloride should not exceed 20 mequiv h−1.
4.8.1 Correction of metabolic alkalosis

Therapy with acetazolamide (250 mg IV) should be considered in those patients who are volume replete and/or those patients in whom large sodium and potassium
loads are not tolerated. Acetazolamide is a carbonic anyhydrase inhibitor, which acts to block renal bicarbonate resorption. In severe cases of metabolic alkalosis,
which are unresponsive
Page 60
Table 4.8 Causes of metabolic alkalosis
Urine chloride <20 mEq l−1 Urine chloride >20 mEq l−1
Gastric suctioning Mineralocorticoid excess
Vomiting Hypokalemia
Diuretic use (recent) Diuretic use (ongoing)
Posthypercapnea Bartter syndrome
Exogenous alkali loading Exogenous alkali loading
to saline and potassium chloride administration, hydrochloric acid may be administered. The rate of acid administered depends on the calculated bicarbonate excess,
calculated as 0.5 (body weight) (observed HCO3—desired HCO3). In patients with renal failure and severe metabolic alkalosis, treatment may require either
hemodialysis or continuous venovenous hemofiltration with dialysis (CVVHD). Chloriderich dialysis solutions are typically used.
Chloride resistant alkaloses are less common in the ICU, and treatment is directed at correcting the underlying cause (e.g. treating primary hyperaldosteronism).
Potassium chloride treatment is helpful in those patients who are hypokalemic.

The compensatory response to metabolic alkalosis is variable and often not significant.
4.9 Respiratory alkalosis

The diagnosis of a pure respiratory alkalosis is made when the PaCO2 decreases and a concomitant decrease in the serum bicarbonate is observed. Respiratory
alkalosis is common in the ICU and is often due to iatrogenic overventilation. Acute respiratory alkalosis in which serum pH rises above 7.55 may result in life
threatening cardiovascular dysfunction, cardiac arrythymias, decreased cerebral blood flow, bronchospasm and posthyperventilation hypoxia. Common causes of
respiratory alkalosis are listed in Table 4.9.
The treatment of a respiratory alkalosis should be aimed at the underlying cause. In the patient who is being iatrogenically overventilated, minute ventilation should
be adjusted.

The compensatory response to respiratory alkalosis is achieved through increased HCO3 excretion by the kidney.
• Acute respiratory alkalosis: HCO3 will decrease by 2 mequiv l−1 for each 10 mmHg decrease of PCO2 below 40 mmHg.
• Chronic respiratory alkalosis: HCO3 will decrease by 5−7 mequiv l−1 for each 10 mmHg decrease of PCO2 below 40 mmHg.
Table 4.9 Causes of respiratory alkalosis
Mechanical overventilation
Hypermetabolic states
Fever
Sepsis
Thyrotoxicosis
Anxiety and fear
CNS lesions
Hepatic failure
Hormones
Epinephrine
Progesterone
Drugs
Salicylate intoxication
Shock
Interstitial lung disease
Page 61
4.9.2 Mixed acidbase disorders

Mixed acidbase disturbances are common in the critically ill patient. Although the diagnosis is formally made by analyzing the patient’s arterial blood gas and serum
electrolytes, the history remains an important diagnostic tool. There are few absolute rules in determining a mixed acidbase disturbance. It should be noted that
compensation, whether metabolic or respiratory, never overcorrects, and rarely restores a normal pH.
References
1. Wrenn, K.D., Slovis, C.M., Minion, G.E. and Rutkowski, R. (1991) The syndrome of alcoholic ketoacidosis. Am. J. Med. 91:119.
2. Miller, P.D., Heinig, R.E. and Waterhouse, C. (1987) Treatment of alcoholic acidosis. The role of dextrose and phosphorus. Arch. Intern. Med. 139:67.
3. Bustos, D., Ponse, S., Pernas, J.C., et al. (1994) Fecal lactate and the short bowel syndrome. Dig. Dis. Sci. 39:2315–2319.
4. Peretz, D.L., Scott, H.M. and Duff, J. (1965) The significance of lactic acidemia in the shock syndrome. Ann. N. Y. Acad. Sci. 119:1133–1141.
5. Stacpoole, P.W. (1986) Lactic acidosis: The case against bicarbonate therapy. Ann. Intern. Med. 105:276–279.
6. Narins, R.G. and Cohen, J.J. (1987) Bicarbonate therapy for organic acidosis: The case for its continued use. Ann. Intern. Med. 106:615–618.
7. Mattar, J.A., Well, M.H., Shubin, H., et al. (1974) Cardiac arrest in critically ill: II. 8. Hyperosmolal states following cardiac arrest. Am. J. Med. 56:162–168.
8. Cooper, D.J., Walley, K.R., Wiggs, B.R. and Russell, J.A. (1990) Bicarbonate does not improve hemodynamics in critically ill patients who have lactic acidosis: A
prospective, controlled study. Ann. Intern. Med. 112:492–498.
9. Bersin, R.M. and Arieff, A.I. (1988) Improved hemodynamic function during hypoxia with carbicarb: A new agent for the management of acidosis. Circulation
77:227–233.
10. Shapiro, J.I. (1990) Functional and metabolic responses of isolated hearts to acidosis: Effect of sodium bicarbonate and carbicarb. Am. J. Physiol 258:H1835.
11. Huseby, J.S. and Gumprecht, D.G. (1981) Hemodynamic effects of rapid bolus hypertonic sodium bicarbonate. Chest 79:552–554.
12. Forsythe, S.M. and Schmidt, G.A. (2000) Sodium bicarbonate for the treatment of lactic acidosis. Chest 117:260–267.
13. Nahas, G.G., Sutin, K.M. and Fermon, C. (1998) Guidelines for the treatment of acidemia with THAM. Drugs 55:191.
14. Stacpoole, P.W., Wright, E.C., Baumgartner, T.G., et al. (1992) A controlled clinical trial of dichloroacetate for the treatment of lactic acidosis in adults. N. Engl.
J. Med. 327:1564–1569.
Page 62

Page 63
Chapter 5
Bedside hemodynamic monitoring
Jay S.Steingrub, MD
Contents
5.1 Introduction
5.2 Central venous monitoring
5.3 Pulmonary artery catheterization
5.4 Validity of measurements
5.5 Intrathoracic pressure changes during spontaneous respiration
5.6 Positive endexpiratory pressure (PEEP)
5.7 Ventricular compliance
5.8 Limitations of pulmonary artery wedge pressure
5.9 Thermodilution in cardiac output
5.10 Derived hemodynamic parameters
5.11 Risk versus benefit
Page 64
5.1 Introduction
Bedside hemodynamic monitoring can provide beneficial clinical information when patients are carefully selected, if the resulting data are interpreted accurately, and if
the procedure is properly performed. When the above criteria are met, hemodynamic monitoring provides precise and quantitative information that can help confirm
what was clinically suspected, help select appropriate therapeutic interventions and assist in followingup patients’ response to treatment. With hemodynamic
monitoring, variables such as preload, afterload, contractility, and heart rate can be altered by various therapeutic interventions. The clinical value of invasive
hemodynamic monitoring has been generally accepted for numerous clinical conditions including shock associated with myocardial infarction, sepsis, major trauma,
acute respiratory failure, cardiogenic or noncardiogenic pulmonary edema, and management of perioperative patients (Table 5.1). The purpose of this chapter is to
review and illustrate the detailed physiologic information that can be obtained from both pulmonary artery catheterization (PAC) and central venous pressure (CVP)
monitoring. We will also consider limitations inherent in pulmonary artery monitoring, and address risk versus benefit.
5.2 Central venous monitoring

Monitoring central venous pressure or right arterial pressure measurements along with urine output, blood pressure and physical examination provided information for
the management of acutely ill patients before the introduction of pulmonary artery catheterization in the early 1970s. Normal right atrial pressure is approximately 0–6
mmHg. The pressure is low because pressure produced during ventricular systole is blocked by the closed tricuspid valve. It should be noted that the atria do not have
systolic or diastolic pressure. A right atrial pressure tracing has ‘a’, ‘c’ and v waves during each cardiac cycle (Figure 5.1). The ‘a’ wave is produced by atrial
contraction, the ‘c’ wave reflects initiation of isovolemic ventricular contraction, and the v wave representing right ventricular contraction. The ‘a‘wave follows a
simultaneously recorded electrocardiographic P wave by about 80 ms and usually precedes the wave. The ‘v’ waves coincide with the T wave of the EKG.
The CVP measurement is an accurate measure of right ventricular enddiastolic pressure, but when used to predict left ventricular enddiastolic pressure (LVEDP)
assumes that the right and left ventricular enddiastolic pressures are normally close and directional changes are similar. An increase in CVP is assumed to be followed
by an increase in left ventricular enddiastolic pressure. Widespread availability of the pulmonary artery catheter has prompted a reevaluation of CVP measurements.
Previously accepted as an adequate guideline for managing shock and hypovolemia, poor or absent correlation was observed when comparing pulmonary artery wedge
pressure (PAWP) with CVP in critically
Figure 5.1 Normal right atrial pressure waveform. The rhythm is sinus. The man right right atrial pressure is 6 mmHg, The ‘a’ (a) is the
dominant positive pressure wave. The right; atrial ‘c’ wave (c) occurs immediatety following the QRS complex. The rig right
atrial ‘v’ wave occurs on the downslope of the electrocardiographic T wave.
Page 65
Table 5.1 Pulmonary arterty catheter indications and findings
Disorder Hemodynamic profile Comments

Ischemic RV Increased RA, decreased SV, Steep y descent
dysfunction decreased CO, decreased AP, RA≥PCWP RV diastolic dip and plateau (square root sign)
Volume loading may unmask hemodynamic changes
Acute mitral Increased PCWP, prominent ‘v’ waves, sometimes reflecting ‘v’ waves may not always differentiate mitral regurgitation from ventricular
regurgitation onto the PA tracing as well septal rupture
Acute ventricular septal Oxygen stepup f from PA to RV RV forward output exceeds LV forward output
rupture Early recirculation on the thermodilution curve
Shock
Cardiac Increased PCWP, decreased SV decreased CO, decreased AP, increased SVR
Hypovolemic Decreased or lownormal PCWP, decreased SV, decreased Orthostatic tachycardia
CO, decreased AP, increased SVR
Early septic Increased PA, increased PVR, increased CO, decreased AP, In later stages, SVR is elevated and cardiac output is lowered
decreased SVR
Noncardiac pulmonary Decreased or normal PCWP normal or increased CO Normal heart size
edema
Massive pulmonary Decreased SV, decreased CO, decreased AP, increased PA, PCWP is normal despite elevated pulmonary artery systolic and diastolic
embolism increased PVR, normal PCWP pressures
Chronic precapillary Increased RA, increased RV, systolic pressure, increased PA, Leftsided pressures are often normal Pulmonary arterial and RV systolic
pulmonary hypertension increased PVR, normal PCWP pressures may reach systemic levels
Acute cardiac Increased RA, increased PCWP, RA=PCWP, decreased SV, Paradoxical pulse Blunted y descent Prominent x descent on RA tracing
tamponade decreased CO, decreased PA
Constrictive pericarditis Increased RA, increased PCWP, dip and plateau in RV Paradoxical pulse is rare Positive Kussmal’s sign is common May simulate
pressure, ‘M’ or ‘W’shaped jugular venous pressure with ischemic RV dysfunction or restrictive cardiomyopathy
preserved x and steep y descent
Restrictive Findings are similar to those described for constrictive Simulates constrictive pericarditis; however, PA systolic pressure is usually >50
cardiomyopathy pericarditis, but PAWP may be higher than RA; difference mmHg and diastolic plateau is less than onethird peak RV systolic pressure;
between PCWP and RA may be exaggerated by exercise other tests are often needed for differentiation from constrictive pericarditis
Tricuspid regurgitation Increased RA, increased RV enddiastolic pressure Blunted x descent, prominent V wave, steep y descent
Ventricularization of RA pressure
Page 66
ill patients1. The CVP will not reflect LVEDP when the pumping function of either ventricle is selectively depressed.
5.2.1 Conditions where CVP may not reflect LVEDP

Clinical entities that may contribute to a discrepancy between CVP and LVEDP readings include diseases that may cause or contribute to the development of acute
right ventricular decompensation such as pulmonary embolism, hypoxic pulmonary vasoconstriction, primary pulmonary hypertension, acute respiratory distress
syndrome, positive pressure ventilation, valvular stenosis, tricuspid insufficiency, septal defects, and septic shock. Any given value of PAWP can be associated with a
wide range of CVP values. Ischemia, heart disease, and cardiomyopathies may result in a higher left ventricular enddiastolic pressure (PAWP) for a given blood
volume on the left side of the heart, when compared with the right ventricular system. CVP measurements may also be unreliable in predicting the amount of fluid
treatment in liver disease, sepsis or burns. However, CVP measurements may be helpful in predicting transfusion requirements in healthy adults with hemorrhage
shock2. Generally speaking, if the CVP is known to correlate with the PAWP in an individual patient, than CVP determinations are useful. Furthermore, in the absence
of cardiopulmonary dysfunction, CVP still remains a reliable assessment of right and left heart volume status. However, in critical illness when right and left ventricular
function is so disparate that absolute CVP readings and changes are unreliable and mislead estimates of left heart filling, pulmonary artery catheters may be used to
provide further monitoring.
5.3 Pulmonary artery catheterization

The pulmonary artery catheter measures pulmonary artery pressure (systolic, diastolic), right atrial and ventricular filling pressures, and PAWP, and allows sampling of
true mixed venous blood and measurement of cardiac output. The objective of hemodynamic monitoring and its derived physiological parameters is assessment of left
and right ventricular function by quantitatively expressing both the output and the filling pressures of both ventricles.
5.3.1 Apparatus
The pulmonary artery catheter is a balloon tipped flow directed catheter that allows rapid access to the central venous circulation and the right heart to secondary and
tertiary divisions of the right or left pulmonary artery. Standard catheters are constructed from polyvinyl chloride and coated with heparin to reduce catheter
thrombogenicity. The length of the catheter is approximately 110 cm, with an external diameter of 7 or 7.5 French (1 French=0.0335 mm) in standard catheters. The
balloon at the tip when inflated guides the catheter from the greater intrathoracic veins through the right atrial and ventricular chambers into the pulmonary artery.
Balloon capacity varies according to catheter size, from 0.5 to 1.5 ml. Conventionally, the balloon is inflated with air, but filtered carbon dioxide can be used in
situations such as right to left intracardiac shunt where balloon rupture might cause a systemic embolus into the arterial system. For best results, the catheter should be
positioned in the large pulmonary artery from which it can consistently flow into wedge position. The goal is to obstruct the distal vessel passively, but not to distend it.
The most commonly used catheters are multilumen, incorporating one lumen to allow balloon inflation, a distal lumen that opens at the catheter tip for measurement of
pulmonary artery and PAWP and for sampling of pulmonary artery mixed venous blood, and a proximal lumen 30 cm from the catheter tip for CVP measurements
used for injection of a thermal indicator for thermal dilution cardiac output measurements, fluid and drug administration. A fourth lumen (VIP or vasoactive infusion port)
allows for additional fluid and drug administration. A wire terminates in a thermistor bead, 3.5–4.0 cm proximal to the
Page 67
tip of the catheter, and provides an electrical connection to the cardiac output computer. The thermistor measures pulmonary artery blood temperature and allows
thermodilution cardiac output measurements.
5.3.2 Insertion site

The right internal jugular vein is considered a preferred site because of its direct access to the right atrium and right ventricle. The left subclavian vein is preferred to
the left internal jugular vein route because of potential damage of the thoracic duct. The broad anatomic curve of the left subclavian vein allows easier catheter passage
into the right atrium. If the right subclavian vein anatomy has a sharp curve into the thoracic cavity before entering the superior vena cava, this may predispose to
technical difficulties and catheter kinking. The internal jugular approach is associated with both a more rapid placement and an appreciable lower incidence of
pneumothorax compared with the subclavian approach. Major disadvantages with the subclavian approach include the significant higher rate of iatrogenic
pneumothorax and difficulty in applying pressure for hemostasis in those patients encountering coagulopathy problems. The femoral vein approach may have some
enhanced risks of deep vein thrombosis and resultant embolism3.
5.3.3 Techniques of insertion

The Seldinger technique allows for venous puncture with a small needle, followed by a wire to secure the position and provide a path for a vein dilator and side arm
introducer. As a general guide, after achieving venous access, the right atrium should be reached about 15–20 cm from the jugular vein and 10 cm from the subclavian
vein (Table 5.2). The balloon when inflated to its recommended volume (usually 1.5 ml of air) protrudes over the catheter tip and decreases the risk of endomyocardial
stimulation and induction of arrhythmias during passage of the catheter tip through the right ventricle. Initially, the balloon remains deflated until the catheter tip is in the
intrathoracic location, with the correct location indicated by increased respirationrelated pressure variation. Once in the right atrium, the balloon is inflated to its
recommended volume. The catheter is then smoothly advanced under continuous electrocardiographic and pressure monitoring across the tricuspid value through the
right ventricle and pulmonary artery and into the pulmonary wedge position. This occurs when the balloon occludes the distal pulmonary artery segment. Characteristic
pressure changes accompany its passage across each chamber. Generally, a catheter should reach the pulmonary artery within 50–55 cm if advanced from the internal
jugular vein, and within 65–70 cm if inserted via the femoral vein. If substantially greater lengths are required to reach the pulmonary artery, catheter coiling in the right
atrium or ventricle should be suspected. Occasionally, the catheter tip reverses direction after crossing the pulmonary valve and comes to lie in the papillary muscles of
the right ventricle. This positioning results in a damped pressure that often simulates a pulmonary capillary wedge tracing. In the absence of flow, pressure rapidly
equilibrates and the pressure tracings from the distal orifice of the catheter will reflect left atrial pressure if certain conditions are fulfilled. Wedge position may be
assumed if: (i) balloon inflation and deflation are consistently accompanied by disappearance and reappearance of pulmonary artery pressure; (ii) catheter obstruction
and damping artifact are excluded; (iii) wedge tracings show an appropriate wave form; (iv) withdrawn blood has an oxygen saturation equal to or greater than that of
systemic arterial blood with the balloon inflated. The normal pressure ranges and waveforms recorded within each chamber and vessels have certain specific
characteristics (Figure 5.2).
5.3.4 Proper positioning

Proper positioning within a pulmonary artery should be confirmed by waveform analysis at least once daily, and if possible before each balloon inflation. When proper
positioning can
Page 68
Figure 5.2 Hemodynamic waveform and pressures.
be confirmed, repeat wedge measurements may be obtained by full balloon inflation. In other settings, the balloon should be slowly inflated (in increments of 0.2–0.5 ml
of air) while the pulmonary artery pressure is continuously monitored.
Wedge tracings obtained at substantially less than full inflation volumes suggest distal migration of the catheter tip. Such migration may result from a combination of
forward thrusting
Table 5.2 Catheter insertion sites
Location Distance from skin to vena cava/RA junction (cm)

Internal jugular 15–20
Superior vena cava 10–15
Femoral vein 30
Right antecubital fossa 40
Left antecubital fossa 50
of the catheter with each heart beat and loop shortening caused by catheter softening.
Poor wedging may also result from patient movement with resultant catheter displacement, or from the administration of PEEP or other forms of mechanical
ventilation. The term overwedge refers to a pressure tracing that continues to rise, eventually exceeding the limits of the scale. Balloon inflation causes the distal lumen to
impact the pulmonary artery intima and occlude the distal lumen. The distal lumen pressure rises because the catheter is flushed at a continuous rate of 3 ml h−1 with the
pressure generated to reach 300 mmHg to flush the catheter. In Figure 5.3, the tracing suggests that the catheter is positioned too distal into the pulmonary artery,
which increases the risk of perforation.
5.3.5 Pulmonary artery wedge pressure measurements

Under normal conditions, the PAWP reflects left atrial and left ventricular enddiastolic
Page 69
Figure 5.3 Overwedge tracing, Scale 0–25 mmHg.
pressure. Continuous PAWP monitoring is impractical because prolonged obstruction of the artery is likely to cause endothelial damage and hence the possibility of
pulmonary artery rupture, and/or infarction. We recommend keeping wedge time to a minimum of 8–15 s and suggest that pressures be recorded over no more than 2–
4 respiratory cycles.
The positive and negative swings in intrathoracic pressures associated with all forms of respiration directionally influence intraluminal pulmonary vascular pressures.
During spontaneous breathing, intraluminal pressure is lower in inspiration than in expiration. However, during mechanical ventilation, pressures on inspiration exceed
those of expiration. Intrathoracic pressures are closest to zero during endexpiration regardless of whether the patient is breathing spontaneously or being mechanically
ventilated. Thus, all pressures should be measured at endexpiration to minimize the influence of intrathoracic pressure swings. Carefully measured endexpiratory
pulmonary diastolic pressures can be substituted for mean PAWPs whenever wedge tracings are unobtainable, as endexpiratory pressures are often most accurate. In
the absence of elevated pulmonary vascular resistance, pulmonary artery diastolic pressure usually approximates the wedge pressure (less than 2–4 mmHg difference).
This relationship is essentially fixed through the physiologic range of pressure. For example, if the initial PAWP is 12 mmHg and the pulmonary artery diastolic pressure
is 15 mmHg, when left ventricular failure causes the PAWP to increase to 20 mmHg, the pulmonary artery diastolic pressure would then be approximately 23 mmHg.
This gradient will remain fixed unless lung disease becomes more severe or an acute process occurs during the pressure monitoring. The wedge pressure should not be
greater than the pulmonary artery diastolic pressure. In the presence of pulmonary hypertension as with pulmonary embolic disease, pulmonary fibrosis, ARDS, or
reactive pulmonary hypertension, pulmonary artery diastolic pressure (PADP) may markedly exceed mean PAWP and is an unreliable index of left ventricular function.
5.3.6 PAWP and left atrial pressure

PAWP tracing is very similar to the right atrial pressure wave form but reflects mechanical events of the left atrium. In the wedge position, blood flow is stopped and
central pulmonary pressure occluded so that the catheter tip senses pressure transmitted backward through the static column of blood from the pulmonary veins. The
major clinical value of the PAWP is that it usually approximates left
Page 70
atrial pressure (LAP), intracavitary LVEDP (filling pressure) and by extension, left ventricular enddiastolic volume (preload). However, this assumption is correct only
if the pulmonary vascular system distal to the catheter tip (the pulmonary capillaries and veins) remains freely patent, and provides a direct bloodfilled vascular
connection between the catheter tip and the left atrium, serving essentially as a direct extension of the catheter (Figure 5.4). In essence, vascular pressure equilibrates
along the continuous column of blood that has no flow. The PAWP is a useful approximation of LVEDP because during the diastole, the PAWP, pulmonary venous
pressure, left atrial and left ventricular pressure equalize in patients with a normal mitral valve and normal left ventricular function. This relationship is found in most
clinical settings, but there are important exceptions.
5.4 Validity of measurements

Because LVEDV is difficult to measure routinely at the bedside, PAWP measurements are considered a reasonably accurate estimate of LVEDP and thus ventricular
preload. LVEDP is a major determinant of LVEDV, with the preload directly determining the force of cardiac contraction for any given level of muscle contractility.
The above relationship implies the following assumptions: (i) LVEDP is directly
Figure 5.4 (a) Simplified representation of the pulmonary artery catheter in the ‘wedge’ or pulmonary artery occlusion (PAO) position.
With the balloon Inflated, no flow exists. (b) By analogy to a ‘closed pipe’ system, equal pressure readings are found for the
wedge pressure, pulmonary venous (PV) pressure, and left atrial (LA) pressure. Adapted from Sprung, C.L., Rakow, E.C and
Civetta, J.M. (1983) Direct measurements and derived calcutations using the pulmonary artery catheter. In; The Pulmonary
Artery Catheter: Methodology and Clinical Applications (ed. Sprung, C.L). Baltimore: University Park Press, Baltimore,
MD, pp, 105–I–40.
Page 71
related to LVEDV; (ii) LVEDP and LAP equalize at enddiastole; (iii) intrathoracic pressures do not influence the measurement of pressures within the pulmonary
vasculature; (iv) PAWP is accurately measured. Although these assumptions are valid in a healthy population, they may not be so in critically ill patients. The correlation
between pressure and volume can be altered by the PA catheter tip location, increased airway pressure (PEEP), hypovolemia, and alterations in ventricular
compliance4.
Two principal factors, pulmonary artery catheter position and alterations in pressure during the respiratory cycle, may affect pulmonary artery pressure values and in
particular PAWP, as indices of left ventricular filling pressure. As discussed, the validity of PAWP as a measure of pulmonary venous pressure and LAP assumes an
uninterrupted column of blood between the balloon and the pulmonary veins. West et al. showed that there exists a marked variation in blood flow within the lung as a
result of the interrelationship between alveolar and vascular pressures and the effect of gravity5. As shown in Figure 5.5, the conceptually upright lung is divided into
zones based on relationship among alveolar (PA), pulmonary artery (Pa), and pulmonary venous pressures (Pv). In the upper zone (zone 1) near the lung apex, where
alveolar pressure commonly exceeds pulmonary, arterial, and venous pressures, pulmonary capillaries are usually closed and no blood flow occurs. This effectively
precludes PAWP reflection of left atrial pressure. In the central lung areas (zone 2), alveolar flow is primarily determined by the balance between arterial and alveolar
pressures only, because in these zones, alveolar pressure commonly exceeds pulmonary venous pressure. Balloon inflation and catheter wedging will convert a zone 2
situation into a zone 1 situation by preventing blood flow. Thus, if the catheter is lodged in either zone 1 or zone 2, PAWP will not reflect mean LAP, but rather alveolar
pressure. In the lower zone (zone 3), the gravitydependent area of the lung, capillaries remain open, because both pulmonary artery and venous pressures exceed
alveolar pressures. As a result, there is free communication between the left atrium and pulmonary arteries. Only in zone 3 will a patent vascular channel act as an open
conduit between the catheter lumen and the left atrium, and blood flow here remains constant. Fortunately, most of the lung is in zone 3 when a patient is supine; flow
directed catheters will usually enter zone 3 because most blood is flowing to this area. Nonzone 3 catheter placement is most likely to occur in hypovolemic patients
and in those with elevated intraalveolar pressure. One should keep in mind, however, that these zones are not defined anatomically, but rather the division is
physiologic and the zone sizes may change. For example, during spontaneous breathing in the supine position, the majority of the alveolar capillary units normally will fall
in zone 3 throughout the respiratory cycle. The extent of zone 1 or 2 will increase when alveolar pressure rises relative to pulmonary venous pressures as during
hypovolemia or mechanical ventilation with PEEP. Because PEEP both augments alveolar pressure and reduces venous return, its application tends to diminish the zone
3 region. When vascular pressures decrease as in hemorrhagic shock, or PEEP elevation raises alveolar pressure, zone 3 may be converted into zone 1 or zone 2. In
zone 1 and 2, the PA catheter measures predominantly alveolar pressure with the balloon inflated instead of left atrial pressure. It may be necessary to refloat the
catheter under new physiologic conditions or to increase zone 3 via volume expansion to reobtain a true wedge position. Therefore, only if the catheter tip is located in
zone 3 will PAWP truly reflect left atrial pressure. A lateral chest radiograph may confirm the catheter tip location at or below the left atrium. What is important to
remember is that as long as the catheter tip remains below the left atrium, zone 3 conditions will exist, despite even high levels of PEEP. A sudden increase or decrease
in PEEP level should not influence the PAWP. The pulmonary catheter tip position may be less important in the patient with pulmonary parenchymal injury, as many of
these patients
Page 72
have noncompliant lungs, thereby reducing transmission of alveolar pressure to the pulmonary vasculature. For further confirmation of the correct tip position, pressure
and wave form data should be examined. In a nonzone 3 position, wedge pressure can exceed PADP, with marked respiratory variations being seen in the wave form
in the absence of cardiac influences, such as the ‘a’ or ‘v’ waves.
5.5 Intrathoracic pressure changes during spontaneous respiration

Although LVEDP (PAWP) is an intravascular pressure, the LVEDP is not the only determinant of LVEDV. Transmural pressure is defined as the net distending
pressure of the left ventricle and is a true estimation of LVEDP without the effects of pressures extrinsic to the heart (juxtacardiac pressure)6. When we refer to the
PAWP as an index of the left ventricular preload, in actuality, transmural pressure results from subtracting the juxtacardiac pressure from intracavitary pressure
(LVEDP). Juxtacardiac pressure, the pressure surrounding the heart, is approximated by measuring the intrathoracic pressure (intrapleural). During normal spontaneous
breathing, the cyclic changes in intrathoracic pressure are transmitted to the intravascular pressure tracing, manifesting a fluctuating baseline reflecting both inspiratory
troughs and expiratory peaks relating to the respiratory cycle. Large fluctuations in intrapleural pressure are attributed to alterations in ventilatory patterns. These
intrapleural swings are due to positive pressure ventilation, labored respiratory efforts, coughing and valsalva maneuvers. These fluctuations can cause either over or
underestimations of LVEDP (PAWP). During periods of respiratory distress such as in noncardiac or cardiac pulmonary edema, pressure swings between inspiration
and expiration can also result in corresponding alterations in wedge pressure measurements during different parts of the respiratory cycle. Pressure readings obtained at
end expiration provide a standard reference point for following hemodynamic changes
Figure 5.5 lung zones characterize the relationship among pulmonary alveolar pressure (PA), pulmonary artery pressure (Pa) and
pulmonary venous pressure (Pv), Wedge pressure reflects Pv only when Pv exceeds PA (zone 3).
Page 73
and for subsequent therapeutic interventions. Digital readout displays average measurements over time, making it difficult to measure endexpiratory pressures
accurately. PAWP measurements should therefore be taken at end expiration with graphic tracings preferred over digital readouts. In patients on ventilators, a course
of shortacting paralysis should be considered if endexpiratory pressure readings are thought to be artificially high because of marked respiratory variations7.
5.6 Positive endexpiratory pressure (PEEP)

As the major pressure surrounding the heart and intrathoracic vessels is a juxtacardiac pressure (approximated by intrathoracic/ intrapleural pressure (Ppl)), the effects
of PEEP on transmural pressure can affect the relationship of PAWP to LVEDV. PEEP can alter ventricular distensibility by increasing intrapleural pressure and
decreasing venous return. With an increase in Ppl, the increment may be transmitted to pulmonary vessels and heart, resulting in an elevated measured PAWP in the
presence of a normal intravascular volume (LVEDV). Transmural pressure, and not PAWP, reflects the patient’s intravascular volume status in such situations. PEEP
application during mechanical ventilation, resulting in a positive Ppl, may cause a decrease or no alteration in transmural pressure and a disproportionate elevation in
LVEDP (PAWP) relative to LVEDV. Transmural pressure can be obtained by insertion of a small esophageal balloon to measure midesophageal pressure.
Esophageal pressure is subtracted from the PAWP and the transmural pressure is determined. A more practical approach is to estimate Ppl and subtract it from the
measured PAWP, although estimates of Ppl are prone to error.
In clinical studies in which Ppl was measured in patients with acute respiratory distress syndrome, intrapleural pressure does not usually become positive with levels
of PEEP below 10 cm of water8. At PEEP levels of 10 cm of water or less, transmural pressure accurately reflects PAWP. At PEEP levels above 10 cm of water,
there tends to be an increase in disparity between PAWP, LAP, and LVEDP. The more compliant the patient’s lungs, the greater the effect of Ppl will be reflected on
the cardiovascular structures, and the higher will be the recorded mean PAWP as a result of perivascular and pericardial compression. The degree to which PEEP is
transmitted to the pleural space and heart varies with the individual. Therefore, simply substracting the PEEP level from the PAWP measurement may not help. Even
knowledge of the intrathoracic pressure itself may not help, as PEEP transmission varies within the thorax9.
It is more important to assess alterations and trends in PAWP relative to medical therapy rather than attaching excessive significance to isolated measurements of
PAWP. Temporary disconnection of PEEP from the ventilator is no longer recommended and is in fact discouraged, as hemodynamics may become destabilized. It
may also leave the patient at increased risk for severe hypoxemia, alveolar collapse and increased intrapulmonary shunt10.
5.7 Ventricular compliance

LVEDV is determined by both the transmural pressure and ventricular compliance. Compli0ance is defined as volume change per unit pressure change. In the steady
state, under conditions of an unaltering compliance, the left ventricular volumepressure relationship is curvilinear (Figure 5.6). With increasing compliance, LVEDV
can increase with minimal change in LVEDP (PAWP). With compliance decreasing, LVEDP (PAWP) increases with minimal changes in LVEDV. Different
compliance curves for various disease states with acute therapeutic interventions may alter left ventricular compliance. Factors that decrease ventricular compliance
include myocardial ischemia, left ventricular hypertrophy, hypertension, cardiac tamponade, aortic stenosis, right to left interventricular septal shifts, constrictive
cardiomyopathies, and inotropes. Ventricular compliance
Page 74
Figure 5.6 Ventricular compliance curve. The relationship between LVEDV and LVEDP is dependent on compliance. A measure of
LVEDP can represent different pathologic states of LVEDV, reflecting left ventricular preload.
increases are associated with vasodilators, congestive cardiomyopathy, left to right interventricular septal shifts, and mitral and aortic regurgitation.
Left ventricular compliance is affected not only by left ventricular filling, intrinsic stiffness properties and extrinsic therapeutic interventions, but also by right ventricular
diastolic volume (see Chapter 2). As a result of the two ventricles being physically coupled by the interventricular septum and pericardium, the enddiastolic volume
curve of either ventricle is dependent upon the diastolic volume of the other. Therefore, any increase in right ventricular volume will impose limitations on LVEDV,
which in turn will be reflected in an increase in LAP, LVEDP, or PAWP.
The clinical implications of changes in left ventricular compliance contribute to possible alterations in PAWP without parallel changes in LVEDV. Variations of
PAWP may reflect only a shift in the left ventricular pressurevolume relationship rather than a change in the left ventricular enddiastolic volume (preload). This is
important when interpreting a given PAWP at the bedside in relation to LVEDV. For example, in a known hypertensive patient with a decreased left ventricular
compliance, optimal preload will require a higher LVEDV than would be necessary in a normal state. A decrease in ventricular compliance always results in an increase
in LVEDP. This explains the development of hydrostatic pulmonary edema in patients with diastolic dysfunction and hypertension with normal LVEDV. Therefore, it is
essential to assess whether a change in PAWP indicates a change in left ventricular compliance or reflects an actual change in LVEDV. By serial measurements of
cardiac output and PAWP, a FrankStarling curve can be plotted that depicts a relationship between volume (cardiac output) and pressure (PAWP) in a given patient
(see Chapter 2, Figure 2.1). Based on the curve, cardiac output can be optimized by the infusions of fluids to increase preload, diuretics to decrease preload,
inotropes to alter cardiac output, or vasodilators to reduce pulmonary and systemic vascular resistance. Echocardiography is capable of estimating LVEDV and may
serve as a noninvasive technique of preload estimation if PAWP measurements are unreliable or lack correlation with clinical status.
5.8 Limitations of pulmonary artery wedge pressure

A pressure gradient normally should not develop between pulmonary veins and the left atrium. Occlusion of the pulmonary veins (tumor, vasculitis, fibrosis, atrial
myxoma) can cause pulmonary venous pressure to exceed LAP. PAWP closely correlates with LAP and LVEDP only over a range of approximately 5–25 mmHg.
Above 25 mmHg, the LAP tends to be higher than PAWP. The left atrial pressure in turn is essentially identical to LVEDP when the left ventricle has not been
chronically diseased. With chronic lung disease, left ventricular enddiastolic pressure is usually higher than the left atrial pressure. In this instance, use of the peak ‘a’
wave of the wedge pressure rather than the mean PAWP may reduce the disparity between wedge and LVEDP Table 5.3 describes conditions in which PAWP is
greater or less than LVEDP.
Valvular heart disease can alter the LAPLVEDP relationship. With mitral insufficiency, retrograde transmission of ventricular
Page 75
Table 5.3 Conditions in which PAWP is greater or less than LVEDP
PAWP>LVEDP
• Mitral stenosis
• Left atrial myxoma
• Pulmonary venous obstruction
• High intraalveolar pressure (continuous positive pressure ventilation)
PAWP<LVEDP
• Stiff left ventricle
• High (>25 mmHg) LVEDP
systole can cause LAP to exceed LVEDP. With mitral stenosis, LAP exceeds LVEDP. Aortic insufficiency causes a reverse pressure gradient with LVEDP greater
than LAP as a result of continued retrograde ventricular filling from the aorta. As discussed previously, an alteration in left ventricular compliance can cause a disparity
between LAP and LVEDP. Physiologically, LVEDP is rate dependent, with bradycardia increasing LVEDP and tachycardia decreasing LVEDP.
5.9 Thermodilution cardiac output

Thermodilution cardiac output (TCO) is widely used as a method for determining cardiac output in the ICU. Clinical studies verify that there is good correlation
between TCO determination and the Fick method performed randomly in the respiratory cycle. Determination of cardiac output via thermodilution requires measuring
the patient’s blood temperature and the temperature of a known volume of fluid (indicator) injected into the right atrium through the proximal port of a pulmonary artery
catheter with temperature change detected distally by a thermistor located 4 cm from the catheter tip. This technique needs a constant blood flow during the period of
indicator flows from the right atrium to the thermistor. A cardiac output microprocessor solves the indicator dilution equation (Stewart Hamilton equation). When the
injectate is injected rapidly as an indicator bolus into the circulation, the solution becomes thoroughly mixed with blood in the pulmonary artery and the temperature
change is sensed by the thermistor. The computer will then process the change in temperature and produce a record of the temperature change over time, resembling a
skewed bellshaped curve (Figure 5.7). The area under the curve (indicator dilution curve) represents the denominator of the Stewart Hamilton equation and thus
determines cardiac output. When cardiac output is low, more time is required for the temperature to return to baseline, producing a larger area under the curve. With
high cardiac output, the cooler injectate travels faster through the heart and the temperature returns to baseline faster. This produces a small area under the curve.
Certain conditions may prevent appropriate mixing or directional flow of the indicator solution. Intracardiac shunts, tricuspid insufficiency, cardiac arrhythmias, and
mechanical ventilation with PEEP can alter the accuracy of TCO by affecting beattobeat cardiac ejection. TCO appears to have an inaccuracy of approximately 12%
for single measurements and 8% for values averaged over three measurements. Because many of the derived hemodynamic indices and planned therapies are
dependent on accuracy of TCO measurements, one needs to recognize the limitations in TCO calculations. Room temperature injectate is probably as accurate as iced
injectate except during hypothermia. The average of three consecutive TCO measurements performed randomly during respirations will reflect a valid cardiac output
measurement. When multiple measurements are performed for each cardiac output determination, a change of at least 15% between determinations should occur
before clinical significance is attached11.
5.10 Derived hemodynamic parameters

Several important hemodynamic parameters may be derived using data obtained by
Page 76
Figure 5.7 Thermodilution cardiac output curve. Adapted with modification from Tobin, M.J. (ed.) (1998) Principles and Practice of
Intensive Care Monitoring. Reproduced with parmission from McGrawHill, New York.
pulmonary artery catheterization. Calculations of these parameters and their normal values are included in Table 5.4. Understanding these physiologic measurements
may assist in defining specific hemodynamic patterns for a variety of clinical situations. Disorders for which hemodynamic profiles have been established include the
following.
5.10.1 Acute mitral regurgitation

This presents as elevations in left atrial pressure and PAWP, large ‘v’ waves in PAWP tracings and, at times, inflated ‘v’ waves in pulmonary artery pressure tracings.
In this situation, the left ventricle will eject blood into a normalsized and relatively noncompliant left atrium causing a giant ‘v’ wave in the wedge pressure tracing. This
large ‘v’ wave elevates the mean pulmonary venous pressure, resulting in heart failure. Acute mitral insufficiency is one example in which the pulmonary enddiastolic
pressure is actually lower than the mean PAWP. The transient reversal of pulmonary blood flow that accompanies the left atrial ‘v’ wave can cause highly oxygenated
blood to enter the pulmonary artery and the mistaken diagnosis of a left to right intraventricular shunt. When severe, this syndrome often leads to profound
hemodynamic disturbance and shock.
5.10.2 Acute right ventricular infarctions

These are seen in patients with ischemic right ventricular dysfunction, and in as many as 40% of patients with inferior infarctions, although the clinical recognition is often
sketchy and inadequate. In patients with right ventricular dysfunction related to ischemia or to an increased right ventricular afterload, low cardiac output and shock may
develop. The treatment and prognosis of shock secondary to a right ventricular infarct is different from that of shock caused by severe left ventricular dysfunction. The
characteristic hemodynamic profile of acute right ventricular infarction or dysfunction includes an elevated right atrial pressure that equals or exceeds the PAWP, a
steep y descent, and a right ventricular pressure tracing that may show a diastolic dip and plateau (square root sign). These findings can be detected in approximately
50% of patients with acute ischemic right ventricular dysfunction. In other patients with this disorder, volume loading may be necessary to unmask hemodynamic
abnormalities.
5.10.3 Acute ventricular septal rupture

The diagnosis for acute ventricular septal rupture can be confirmed by the detection of oxygen saturation stepup between the right atrium and the right ventricle of
10%. With small defects, intracardiac pressures are usually normal, but with large defects and associated left to right shunt, the left atrial pressure and LVEDP are
usually elevated. Larger defects are associated more frequently with congestive heart failure, arrhythmias or pulmonary hypertension development.
5.10.4 Septic shock

The characteristic profile of septic shock is manifested by an elevated cardiac output and a low system vascular resistance. In early shock, the profile can reveal a low
PAWP, occasional increases in pulmonary vascular resistance, and a low, normal or hyperdynamic CO state, depending on the underlying volume status. In the later
stages of septic shock, it is not uncommon to observe an increase in
Page 77
Table 5.4 Cardiovascular and respiratory parameters (measured and calculated)
Abbreviation Formula Unit Normal

Volumerelated variables
Mean arterial pressure MAP DP+1/3pp mmHg 93
Central venous pressure CVP (measured) cmH2O 10
Stroke volume SV (CO/HR)×1000 ml min−1 50–70

Stroke index SI (CI/HR)×1000 ml min−1 m−2 35–50
Hemoglobin Hgb (CI/HR)×1000 9% 12
Mean pulmonary arterial pressure MPAP (measured) mmHg 15
Pulmonary artery wedge pressure PCWP (measured) mmHg 6−12
Flowrelated variables
Cardiac output CO (measured) l min−1 4–7
Cardiac index Cl CO/BSA ml min−1 m−2 2.5–4
Left ventricular stroke work index LVSWI (MAP−PCWP)×SVI×0.0136 gm m −2
per beat 40–70
Right ventricular stroke work index RVSWI (PAP−CVP)×SVI×0.0136 gm m −2
per beat 4–8
Resistance
Systemic vascular resistance index SVRI (MAP−CVP)/CI ×79.92 dynes cm−5 1500–2600
Pulmonary vascular resistance index PVRI (MPAP−PCWP)/CI ×79.92 dynes cm−5 45–225
Oxygen transport
Arterial pH PH (measured) – 7.4
Arterial CO2 tension PaCO2 (measured) mmHg 40
Mixed venous O2 tension PvO2 (measured) mmHg 40
Arterial O2 saturation SaO2 (measured) % 99
Arterial O2 content CaO2 1.36 (Hg) (SqO2)+0.003 (PaO2) ml O2 dl −1 20
Mixed venous O2 content CVO2 (Hg) (SvO2)+0.003 (PqO2) ml O2 dl−1 15
O2 availability O2 avail. CaO2×Cl×10 ml min−1 500–700
O2 consumption VO2 CaO2×Cl×10 ml min−1 200–250
O2 extraction O2 ext. (CaO2−CvO2)/CaO2 % 25–32%
Rate pressure product RPP HR×SBP – 12000

Page 78
systemic vascular resistance and a decrease in cardiac output.
5.10.5 Noncardiac pulmonary edema

Distinguishing noncardiac pulmonary edema from edema secondary to cardiac disorders is often difficult because in both instances patients are critically ill, have severe
arterial blood gas abnormalities and have chest Xray findings of generally interstitial and alveolar infiltrates. Hydrostatic pulmonary edema from congestive heart failure
usually is characterized by a high CVP and PAWP recording and a increase in systemic vascular resistance. Non decreased CO along with an associated cardiac
permeability pulmonary edema is frequently associated with a high CO and a PAWP typically low to normal, but the latter may be elevated because of fluid overload or
associated acute pulmonary hypertension. Distinguishing between the two types of edema is crucial because therapeutic strategies and overall prognosis differ
markedly.
5.10.6 Acute massive pulmonary embolism

This can produce a clinical picture simulating that of cardiogenic shock. Although pulmonary vascular resistance is markedly increased, the PAWP is usually low or
normal in these patients. The CVP, the right ventricular systolic and diastolic pressures, and PA systolic pressures are also elevated. The mean pulmonary artery
pressure is in the range of 20–40 mmHg with right ventricular and pulmonary artery systolic pressure exceeding 50 mmHg.
5.10.7 Acute cardiac tamponade

This produces diastolic compression, which impairs ventricular filling. The hemodynamic profile reveals a low stroke volume, a falling cardiac output with normal
systolic function, equalization of right and left side filling pressures, and a prominent systolic x descent and blunted y descent on right atrial pressure monitoring.
However, in postoperative cardiac surgical patients where the pericardium has been left open, classic equalization of pressures may not be present.
5.10.8 Tricuspid regurgitation

This is associated with an elevated CVP and right ventricular enddiastolic pressure. Characteristically, the mean CVP demonstrates a rise or no fall with inspiration, an
absent or diminished x descent and a prominent early ‘v’ wave. As the tricuspid regurgitation becomes more severe, the CVP pressure tracing will increasingly
resemble the right ventricular pressure tracing.
5.10.9 Constrictive pericarditis

This, like tamponade, produces diastolic equalization and elevation of right and leftsided filling pressures. A characteristic ‘m’ or ‘w’shaped CVP tracing can be
detected with a preserved systolic×descent and a prominent early diastolic y descent.
5.11 Risk versus benefit

The information provided by hemodynamic monitoring can help identify patients most responsive to treatment and monitor closely the patient’s response to therapy in
the critical care setting. Although the information obtained with pulmonary artery catheterization is not readily available by other means, it has never been clearly
demonstrated that PAC improves survival in critically ill patients. Welldesigned, randomized controlled trials are still needed to resolve this point. Dysrhythmias, such
as premature ventricular depolarization or conduction abnormalities, are the most common complications, occurring in 13–78% of patients12.
Although several organizations have published guidelines for the use of PAC, the recommendations are based on the experience and beliefs of the expert panels
rather than on a critical appraisal of the available evidence. Furthermore, the guidelines have never been prospectively evaluated. If we employ PAC without having
effective therapy for the underlying disease, other than with supportive
Page 79
care, the catheter should not be expected to affect outcome. Therefore, the need arises for more investigations into the mechanisms of critical illnesses and the clinical
applications of new therapies.
References
1. Swan, H.J.C. (1974) Central venous pressure monitoring is an outmoded procedure of limited practical value. In: Controversies in Internal Medicine (eds
F.J.Ingelfinger, R.V.Ebert, M. Finland, et al.). W.B.Saunders, Philadelphia, PA, pp. 185–193.
2. Northfield, T.C. and Smith, T. (1970) Central venous pressure in clinical management of acute gastrointestinal bleeding. Lancet 2: 584–586.
3. Merrer, J., De Jonghe, B., Golliott, F., et al. (2001) Complications of femoral and subclavian venous catheterization in critically ill patients. A randomized controlled
trial. JAMA 286:700–706.
4. Biondi, J.W., Schulman, D.S. and Matthay, R.A. (1988) Effects of mechanical ventilation on right and left ventricular function. Clin. Chest Med. 9:55.
5. West, J.B., Dollery, C.T. and Naimark, A. (1964) Distribution of pulmonary blood flow in isolated lung: Relation to vascular and alveolar pressures. J. Appl Physiol.
19:713–724.
6. Quinn, R. and Marini, J.J. (1983) Pulmonary artery occlusion pressure; clinical physiology, measurement, and interpretation. Am. Rev. Respir. Dis. 128:319–326.
7. Schuster, D.P. (1984) Accuracy of pulmonary artery and wedge pressures. Crit. Care Med. 12: 695–696.
8. Marini, J., O’Quinn, R., Culver, B.W. and Butler, J. (1982) Estimation of transmural cardiac pressures during ventilation with PEEP. J. Appl Physiol 53:384–391.
9. Jardin, F., Genevray, B., BrunNey, D. and Bourdarias, J.P. (1985) Influence of lung and chest wall compliances on transmission of airway pressure to the pleural
space in critically ill patients. Chest 86:653–658.
10. Kumar, A., Falke, K.J., Geffin, B., et al. (1970) Continuous positivepressure ventilation in acute respiratory failure. Effects on hemodynamics and lung function. N.
Engl. J. Med. 283: 1430.
11. Ellis, R., Gold, J., Rees, R., et al. (1972) Computerized monitoring of cardiac output by thermal dilution. JAMA 220:507–511.
12. Shah, K.B., Rao, T.L.K., Laughlin, S. and ElEtz, A.A. (1984) A review of pulmonary artery catheterization in 6,245 patients. Anesthesiology 61:271–275
Page 80

Page 81
Chapter 6
Shock in the intensive care unit
Jay S.Steingrub, MD
Contents
6.1 Introduction
6.2 Pathophysiology
6.3 Specific shock syndromes
6.4 Hypovolemic shock
6.5 Distributive shock
6.6 Anaphylactic shock
6.7 Cardiogenic shock
6.8 Obstructive shock
6.9 Shock states
Page 82
6.1 Introduction
Shock is common among critically ill patients; as many as 20–25% of all ICU admissions are shock related. Although the diagnosis of shock should be suspected in any
patients with signs of altered tissue perfusion, commonly the clinical diagnosis is often established when hypotension develops. Variables determining outcome for a
patient in shock include age, preexisting medical conditions, severity and type of injury, and the length of time that elapses before initiation of resuscitation that can limit
the effectiveness of compensatory responses. Traditionally, shock has been estimated by the oxygen retention of pulmonary artery blood (mixed venous O2) and the
degree of lactic acidosis. Because the partial pressure of mixed venous O2 may not reflect oxygen delivery in patients with sepsis or acute respiratory distress syndrome
(ARDS), oxygen delivery is better assessed by measuring cardiac output, hemoglobin concentration, and arterial and mixed venous blood gases.
Ineffective tissue perfusion causes cellular and tissue ischemia as a result of an imbalance in the supply and demand of oxygen and nutrients. This is clinically
recognized as the syndrome of shock. The progression of shock can be categorized into three stages. During the early stage, several compensatory responses directed
towards maintaining specific organ perfusion (heart and brain) are initiated; the effectiveness of these compensatory responses depends on the rate and degree of
intravascular volume depletion. During the second stage, both microvascular and cellular injuries are observed. In the final stage, decompensated shock develops,
manifested by irreversible organ damage.
6.2 Pathophysiology
Compensatory mechanisms preserve cerebral and coronary circulation at the expense of skin, skeletal muscle, renal, and splanchnic circulation. A decrease in arterial
pressure secondary to cardiac output depression stimulates a compensatory autonomic response. Baroreceptor reflexes are stimulated by highpressure stretch
receptors located in the carotid sinus, aortic arch, and splanchnic circulation1. Volume receptors located in the right atrium trigger reflex pathways in the setting of
hypovolemia. Sympathetic nervous system activation causes arteriole vasoconstriction and redistribution of blood flow away from skeletal muscles and splanchnic beds.
Sympathetic outflow will also augment heart rate and myocardial contractility and cause venoconstriction, thus increasing venous return. Increases in sympathetic
nervous system discharge cause adrenomedullary hormone release, catecholamine release, and activation of the reninangiotensin axis2. Both the reninangiotensin
system and vasopressin increase vasomotor tone, predominantly in the mesenteric bed. Angiotensin II increases aldosterone release and sympathetic outflow;
vasopressin stimulates both catecholamine release and myocardial contractility. Ongoing ischemia at the cellular level can facilitate both humeral and proinflammatory
reactions (cytokines, nitric oxide), which can further compromise circulatory abnormalities, leading to endorgan damage and death.
6.3 Specific shock syndromes

Unless the underlying cause is determined, shock can be viewed as a dynamic state with several major etiologies potentially leading to progressive endorgan damage.
To identify the underlying cause, one must first determine whether shock is central or peripheral in origin. Shock can be further categorized into four categories. Central
causes of shock resulting in impairment in cardiac emptying include cardiogenic shock (pump failure) and obstructive shock (impaired blood flow). Peripheral causes of
shock are attributed to alterations at the vascular blood level, and include hypovolemic shock (inadequate blood volume) and distributive shock (abnormal distribution
of available blood volume) (Table 6.1).
Page 83
Table 6.1 Mechanisms of shock
Central Peripheral
Cardiogenic shock Hypovolemic shock
Pump failure Inadequate blood volume
• Myocardial infarction • Hemorrhage/bleeding
• Cardiomyopathy • Capillary leak
• Arrhythmias • Burns
• Postcardiopulmonary bypass
• Myocarditis
• Intrinsic depression SIRS related
• Valvular dysfunction
Obstructive shock Distributive shock
Blood flow impairment Blood volume maldistribution
• Pulmonary hypertension • Sepsis/inflammatory
• Pulmonary/air embolism • Anaphylaxis
• Tension pneumothorax • Spinal cord injury
• Positive pressure ventilation • High spinal anesthetic
• Adrenal insufficiency
• Dissecting aneurysm
• Tamponade
• Ball valve thrombosis
• Mediastinal tumors
6.4 Hypovolemic shock
6.4.1 Clinical signs and symptoms

This category of shock is caused by visible or occult hemorrhage and nontraumatic diseases including nonhemorrhagic hypovolemic shock. Inadequate circulatory
blood volume can be secondary to massive sequestration of ‘thirdspace’ fluid in burns, soft tissue injuries, bowel obstruction, and acute inflammatory disorders such as
pancreatitis. An increase in capillary permeability resulting in hypovolemia is observed in both sepsis and anaphylactic shock; however, these clinical syndromes have
other additional mechanisms that explain shock. This shock state can occur abruptly or gradually through several stages, depending on the severity of hypovolemia and
the rate of its development. The diagnosis of hypovolemic shock relies on the identification of symptoms and signs of inadequacy of tissue perfusion in the appropriate
clinical setting (Table 6.2). Poor skin turgor, dry mucous membrane, hyperthermia, elevated hematocrit, low central venous pressures, or pulmonary artery
Table 6.2 Monitoring tissue perfusion
Clinical assessment
Invasive blood pressure monitoring
Cardiac output
Continuous pulse oximetry
Cardiac filling pressure
Lactate levels
Gastric tonometry
Oxygen delivery and extraction (from mixed venous blood gases)
Clinical
Pulse rate
Skin color
Temperature
Capillary refill
Urinary output
Anion gap
Mental status
Page 84
wedge pressure (PAWP), hypernatremia, and glucosuria are common findings. With an increase in blood or fluid loss, heightened sympathetic tone becomes apparent,
manifested by increasing heart rate and respiratory rates, decreased capillary refill and narrowing of pulse pressure. An altered mental status is usually a late clinical
manifestation (Table 6.3).
Blood volume is approximately 7% of the body weight in an adult (5 l in a 70 kg patient) and 8–9% in a child. Patients are usually asymptomatic with blood volume
deficits of less than 10% (approximately 500 ml). Sympathetic vasoconstriction and tachycardia will maintain both blood pressure and cardiac output while secretion of
reninangiotensin, vasopressin and aldosterone contributes toward compensation. Orthostatic blood pressure changes are usually observed when blood loss
approximates 15–20% of the vascular volume (750–1000 ml), resulting in a decrease in left ventricular preload with a subsequent decrease in stroke volume (SV) and
cardiac output (CO). During this stage, compensatory arteriolar and venous vasoconstriction results in an increase in diastolic pressure and a narrowing of pulse
pressure. Blood volume deficits of greater than 25% (approximately 1250 ml) or at least 1.5 l in a 70 kg patient will cause a decrease in systolic and diastolic pressure,
cardiac output and tissue perfusion. When significant hypovolemia is suspected because of traumatic or nontraumatic mechanisms, fluid therapy should be initiated
regardless of the systolic blood pressure.
6.4.2 Therapy
Therapy for hypovolemic shock includes rapid restoration of circulating volume with blood, crystalloid or colloid fluids as indicated (see Chapter 7). The initial
resuscitation fluids used in hypovolemic or hemorrhagic shock should be IV isotonic crystalloid solutions, either lactated Ringer’s solution or isotonic 0.9% saline. These
fluids provide transient expansion of the intravascular space as well as correction of electrolyte abnormalities.
Controversy persists regarding the use of
Table 6.3 Signs of hypoperf fusion
Organ Clinical signs

Neurologic Altered mental status (agitation, coma)
Metabolic Hyperthermia/hypothermia, respiratory alkalosis, metabolic acidosis
Pulmonary Cardiac Hypoxemia, tachypnea Tachycardia, arrhythmias, ischemia, hypotension
Renal Oliguria (output ≤0.5 cc kg−1 h−1)
colloid solutions (hetastarch, albumin) as an acute resuscitative fluid in hypovolemic shock. Although both colloid and crystalloid solutions are equally effective in
resuscitation, the volume of crystalloid infused is usually twice to four times the total volume of colloid required to achieve the target end point.
Blood loss of 15–30% of blood volume (70 ml kg−1) in a healthy, uncomplicated patient can be corrected with crystalloids alone. Usually, a loss of greater than
30% (2000 ml) of blood volume in the presence of hypovolemic shock will require both crystalloid and blood products to restore both intravascular volume and
oxygen carrying capacity. Generally, healthy wellresuscitated hemodynamically stable patients tolerate hematocrits of approximately 21–25%. Older patients with
underlying heart disease and patients with shock may require higher hematocrits. Practice guidelines suggest that replacement of red cells is usually indicated at
hemoglobin levels below 7 g dl−1 and unnecessary at hemoglobin levels above 10 g dl−1. At intermediate values, the need for transfusions should be evaluated on a
casebycase basis. A large study in a mixed group of ICU patients showed no benefit of transfusion to a hemoglobin level of 10 g dL−1 versus 7 g dl−1 13. In this trial,
the use of the lower hemoglobin level trigger for transfusion resulted in improved survival. When considering blood transfusions to improve oxygenation in critically ill
patients, stored blood may be less effective than fresh blood.
Page 85
6.4.3 Blood components in hemorrhagic shock

Rapid blood infusion can be accompanied by coagulopathy and usually attributed to the consumption of platelets. Platelet infusions are helpful if the platelet count is less
than 100,000. One pooled pack of platelets (eight units) increases platelet count by 80,000 cells mm−3. Factor depletion usually occurs later than platelet loss. In
general, one unit of fresh frozen plasma (FFP) is generally required for every 6 units of packed red blood cells (PRBC), but replacement can be guided by coagulation
monitoring. Cell saver devices should be considered when feasible. FFP infusion might be helpful if the prothrombin time (PT) and/or activated partial thromboplastin
time (APTT) are prolonged (>11/2 times control). Administration of component therapy prophylactically during resuscitation has no benefit, but exposes the patient to
the risks of hepatitis and other bloodborne pathogens.
6.4.4 Evaluation of response

The clinical response to fluid or blood administration should be evaluated by following systolic blood pressure, pulse rate, and signs of organ perfusion (capillary refill,
mental status, urine output >0.5 cc kg−1 h−1 in adults). Additional bolus infusions should be employed based on the assessment. Monitoring of therapy may be
enhanced by central venous catheterization, which assesses right ventricular filling pressures. Trends in the central venous pressure (CVP) measurements are more valid
than absolute numbers. Placement of a pulmonary artery catheter (PAC) should be considered if the CVP is persistently elevated in the presence of shock or
pulmonary edema, or if increasing vasopressor dosing is required. The PAC allows for assessment of left ventricular filling pressures, mixed venous blood samples, and
cardiac output (Table 6.4). Arterial monitoring allows for the accurate monitoring of arterial pressure and access for blood sampling.
Patients may require vasopressors to maintain blood pressure while fluid and blood resuscitation are in progress. Ongoing vasopressor requirements after fluid
resuscitation is complete may indicate ongoing concurrent losses, distributive shock, or complications such as pump failure, myocardial infarction or tamponade. The
presence of lactic acidosis suggests inadequate fluid resuscitation.
6.4.5 Oxygen and mechanical ventilation

Most patients in hypovolemic shock should receive supplemental oxygen and may require
Table 6.4 Pulmonary artery catheterization and shock diagnosis
Diagnosis PAWP CO
Cardiogenic shock
Acute ventricular septal defect or N
Acute mitral regurgitation
Right ventricular infarction or N
Extracardiac obstructive shock
Pericardial tamponade or
Massive pulmonary emboli or N
Hypovolemia
Distributive shock
Septic or N or N
Anaphylaxis or N or N
or indicates moderate to severe increase or decrease; or indicates mild to moderate increase or decrease; N, normal; PAWP, pulmonary artery wedge pressure; CO, cardiac
output.
Page 86
mechanical ventilation. In hypovolemic shock, positive pressure ventilation will frequently decrease preload, causing a decrease in cardiac output and systemic oxygen
delivery (DO2) (See Section 2.2.4 and Figure 2.1).
6.5 Distributive shock

In distributive shock, patients characteristically have both hypotension and maldistribution of blood flow. Causes of distributive shock includes anaphylaxis, sepsis,
spinal cord injury, drug actions, and endocrine emergencies such as thyroid storm. Patients with distributive shock frequently develop extravasation of intravascular
volume secondary to intracapillary leak. A decreased resistance to flow, a maldistribution of flow, and relative hypovolemia characterize the underlying
pathophysiology. Intravascular monitoring of these patients will show a decrease in mean arterial pressure, a decreased or normal PAWP, an increase in cardiac output
(post fluid resuscitation) and a reduction in systemic vascular resistance. Sepsis and anaphylaxis are the two most common causes of distributive shock. The incidence
of shock in patients with Gramnegative bacteremia is approximately 40%, with mortality rates ranging from 30 to 40%4. In 1991, experts in the discipline of critical
care medicine proposed new definitions for sepsis and the adverse sequelae of sepsis. Sepsis was defined as a continuum of injury response, ranging from sepsis to
septic shock, to multiorgan dysfunction syndrome5. The proposed definitions for the systemic inflammatory response syndrome (SIRS) and sepsis were evaluated in a
prospective trial. The mortality rate increased with increasing number of SIRS criteria or the patients developed more severe adverse sequelae. The mortality rate
increased progressively as the patients manifested two SIRS criteria (7%), three SIRS criteria (10%), four SIRS criteria (17%), sepsis (16%), severe sepsis (20%) and
septic shock (46%)6. The diagnosis of septic shock is defined as sepsis plus organ dysfunction and evidence of persistent hypotension requiring vasopressor support to
maintain blood pressure (Table 6.5). Both Gramnegative and Grampositive bacteria, viruses, and fungi can trigger the clinical syndrome, which typically occurs over a
period of hours or days.
Anaphylactic shock, another etiology of distributive shock, occurs more acutely than septic shock, has a shorter duration, and is produced when patients are
exposed to foreign substances or when substances enter the blood stream causing various endogenous mediators to be released. The cardiovascular abnormality of
distributive shock is treated by restoring intravascular volume, increasing systemic vascular resistance with vasopressors and treating the underlying etiology.
6.5.1 Diagnosis of distributive shock

Identification and eradicating the source of sepsis is the mainstay of therapy in septic shock. Most patients presenting with severe sepsis usually have a pulmonary,
genitourinary, primary blood stream, intraabdominal or intravenous catheter as a source of infection. In approximately 75% of patients with presumed sepsis, an
infectious etiology can be isolated, usually divided between Grampositive and Gramnegative organisms7.
Specimens for Gram staining and cultures should be obtained from multiple sites, ideally before antibiotics are administered. However, the need for aggressive
therapy may render it impossible to collect specimens immediately, and early antibiotic administration offers an outcome advantage. Blood cultures should be obtained
from two different sites. In those patients with an indwelling venous or arterial catheter, specimens should be collected through each port of the catheter. At least one
blood specimen should be collected through the catheter port to rule out catheterrelated bacteremia. A minimum of two and a maximum of three sets of blood cultures
should be obtained for each episode of suspected bacteremia. No interval is required between taking sets of blood cultures.
Sputum specimens can be obtained following instillation of 3% saline, or can be obtained through nasotracheal or endotracheal sampling. The specimens should
contain
Page 87
Table 6.5 ACCP/SCCM consensus conference definitions of sepsis, severe sepsis, and septic shock*
Systemic inflammatory response syndrome (SIRS). The systemic inflammatory response to a wide
variety of severe clinical insults, manifested by two or more of the following conditions:
• temperature >38°C or <36°C
• heart rate >90 beats min−1
• respiratory rate >20 breaths min−1 or PaCO2 <32 mmHg
• white blood cell count >12,000 mm−3, <4000 mm−3, or >10% immature (band) forms
Sepsis. The systemic inflammatory response to a documented infection. In association with infection,
manifestations of sepsis are the same as those previously defined for SIRS. It should be determined
whether they are a direct systemic response to the presence of an infectious process and represent an
acute alteration from baseline in the absence of other known causes for such abnormalities. The clinical
manifestations would include two or more of the following conditions as a result of a documented
infection:
• temperature >38°C or <36°C
• heart rate >90 beats min−1
• respiratory rate >20 breaths min−1 or PaCO2 <32 mmHg
• white blood cell count >12,000 mm−3, <4000 mm−3, or >10% immature (band) forms
Severe sepsis/SIRS. Sepsis (SIRS) associated with organ dysfunction, hypoperfusion, or hypotension.
Hypoperfusion and perfusion abnormalities may include, but are not limited to, lactic acidosis, oliguria, or
an acute alteration in mental status
Sepsis (SIRS)induced hypotension. A systolic blood pressure <90 mmHg or a reduction of ≥40
mmHg from baseline in the absence of other causes for hypotension
Septic shock/SIRS shock. A subset of severe sepsis (AIRS) and defined as sepsis (SIRS)induced
hypotension despite adequate fluid resuscitation along with the presence of perfusion abnormalities that
may include, but are not limited to, lactic acidosis, oliguria, or an acute alteration in mental status. Patients
receiving inotropic or vasopressor agents may no longer be hypotensive by the time they manifest
hypoperfusion abnormalities or organ dysfunction, yet they would still be considered to have septic
(SIRS) shock
Multiple organ dysfunction syndrome (MODS). Presence of altered organ function in an acutely ill *From Bone, R.C., Balk, R.A., Cerra, F.B., et al. (1992) American
patient such that homeostasis cannot be maintained without intervention College of Chest Physicians/Society of Critical Care Medicine
Consensus Conference: Definitions for sepsis and organ failure and
guidelines for the use of innovative therapies in sepsis. Reproduced
with permission from: Chest 101:1644–1655.
fewer than 10 squamous epithelial cells per highpower field and have more than 25 leukocytes per highpower field to differentiate contamination or colonization from
true infection (see Chapter 16).
Urine samples should not be obtained from a closed collection container. A syringe and small gauge needle can be used to obtain urine sample directly from the
catheter tubing. Although the presence of large numbers of bacteria in the urine indicates infection, it does not necessarily indicate that the infection arose in the urinary
tract.
The clinical situation will determine which other specimens to collect. A stool specimen for Clostridium difficile cytotoxic assay should be sent if diarrhea develops
in patients receiving antibiotics. If meningitis is suspected, lumbar puncture specimens should be sent for cell count, culture, glucose testing, and testing for bacterial
antigen levels. Fluid collections found on radiologic exams should be aspirated and sent for Gram staining and culture. Differentiation of infected material from
hematoma or inflammatory fluid is not possible on the basis of radiology alone. The pH of the fluid obtained may be a sensitive marker. Ultrasound is the modality of
first choice in the diagnosis of intraabdominal infection, but CT scan should be
Page 88
considered in the presence of diffuse postoperative ileus.
6.5.2 Antibiotic therapy

Antimicrobial therapy remains the primary treatment for patients with sepsis. Antibiotic choices are dictated by the initial source of infection, the patient’s underlying
immunologic status and/or whether the infection is nosocomial or community acquired. Retrospective data shows that early administration of appropriate antibiotics
reduces the mortality in patients with bloodstream infections with Gramnegative sepsis8. In the majority of cases, empiric therapy should begin as soon as the diagnosis
of sepsis is considered. Although initial empiric therapy may require more than one antibiotic to cover potential pathogens, once a pathogen is isolated, monotherapy is
adequate for most infections. Monotherapy with carbapenem antibiotics is as effective as combination therapy with a βlactam and an aminoglucoside for the empiric
treatment of nonneutropenic patients with severe sepsis. Monotherapy with a third or fourthgeneration cephalosporin is as effective as combination therapy with a β
lactam plus an aminoglycoside in nonneutropenic severe sepsis9. Fluoroquinolones are highly effective for the treatment of documented Gramnegative bacteremia, but
evidence for their use as a singleagent treatment of severe sepsis is lacking. Third and fourthgeneration cephalosporins and carbapenem antibiotics are equally
effective as empiric therapy in patients with severe sepsis. The indiscriminate use of vancomycin is to be avoided to limit the risks. Antifungal therapy is recommended
for patients with candidemia. It should not be used on a routine basis as empiric therapy in patients with severe sepsis. Double antibiotic therapy is commonly employed
in those patients with suspected pseudomonas infections, enterococcal infections, febrile neutropenic patients, and severe intraabdominal infections, but monotherapy
may be equally appropriate (see Chapter 16). Empiric therapy is recommended for those patients with culture negative sepsis. Changing antibiotic regimens is
appropriate in patients with ongoing culture negative sepsis who do not respond to the initial empiric regimen. Antibiotics are usually continued for up to 2 weeks,
except in cases of osteomyelitis and endocarditis.
6.5.3 Hemodynamic support

In distributive shock, secondary to sepsis, peripheral vasodilation is associated with hypotension and abnormal distribution of blood flow. Increased capillary
permeability causes tissue edema and relative intravascular volume depletion. Fluid retention and signs of organ hypoperfusion are common in patients with sepsis,
despite a characteristic high cardiac output. Myocardial contractility is depressed with global left and right ventricular dilatation. Studies suggest that lactic acidosis and
low oxygenation are strong predictors of multiorgan system dysfunction and death in septic patients10.
Aggressive fluid resuscitation is the mainstay of therapy for septic shock. Intravascular volume depletion and hypotension can often be reversed with fluid
administration alone. Ventricular performance improves with fluid resuscitation, with the ventricle accommodating an increase in volume without any increase in pressure
as a result of an increased ventricular compliance in severe sepsis. Fluid requirements during the initial resuscitation phase are large, with up to 10 l of crystalloid being
required during the first 24 h. Adequate fluid resuscitation is essential before initiating vasoactive therapy. Despite aggressive fluid resuscitation, a subset of patients with
septic shock remain hypotensive and progress to endorgan dysfunction. These patients may benefit from the use of vasopressor therapy (see Chapter 8). Although the
benefit of pulmonary artery catheterization (PAC) in the management of septic shock remains unproven, the clinical evaluation of intravascular volume of septic patients
is extremely difficult. Pulmonary artery catheterization may allow more appropriate employment of fluids and vasoactive drug therapy11.
Page 89
6.5.4 Adrenal insufficiency and septic shock

Acute adrenal insufficiency should be suspected in patients who fail to respond to vasopressors in septic shock. The incidence of adrenal insufficiency may be 20–25%
12. A lowdose ACTH (cortrosyn) stimulation test (1–2 µg) and a trial of stress doses of hydrocortisone (100 mg every 8 h) pending test results is suggested when
adrenal insufficiency is suspected in septic shock13. A rapid and dramatic improvement in hemodynamics following the infusion of hydrocortisone within a few hours of
administration can be observed. If the ACTH (cortrosyn) stimulation test cannot be performed immediately, dexamethasone (2 mg) should be given and the test
performed within 12 h. Corticosteroids should not be used in severe sepsis or septic shock at high doses (30 mg kg−1) and for a short course (1–2 days). However,
steroids may be used during refractory septic shock, but not during severe sepsis without shock or mild shock. They should be used at low doses (<100 mg
hydrocortisone three times a day) for 5–10 days and then with subsequent tapering of the dose according to the hemodynamic status. The results of a large trial suggest
that low dose steroids can reduce 28day mortality in septic shock14. Several factors may explain the recent positive effect of steroids during sepsis. These include the
treatment of relative adrenal insufficiency and the potentiation of adrenergic receptivity in addition to the antiinflammatory effect.
6.5.5 Complications of septic shock

Acute renal failure is a significant complication in patients with sepsis, with the mortality rate of patients who develop acute tubular necrosis greater than 50%15. Typical
findings include oliguria and azotemia. Urinary sodium excretion may be below 20 mequiv l−1 and urinary osmolality may be increased (>450 mOsm kg−1). It is
important that all patients with sepsis be aggressively resuscitated to prevent this complication. Although renal dose dopamine is widely used in preventing or
ameliorating acute renal failure in patients with sepsis, data show that lowdose dopamine does not improve the survival or prevent the need for dialysis in patients with
ischemic or toxic acute tubular necrosis16. In nonoliguric patients with sepsis, dopamine may enhance diuresis without an alteration in creatinine clearance. Continuous
renal replacement therapies (CRRT) have been successfully used in the management of patients who have sepsis and oliguria despite aggressive resuscitation. CRRT
allows more precise volume control, causes less hemodynamic instability, provides for greater continuous clearance and compared with intermittent hemodialysis
provides a more stable metabolic control17. There are no controlled studies comparing the morbidity or mortality of patients treated with CRRT versus conventional
intermittent hemodialysis.
Acute respiratory failure frequently precipitates intubation and the need for mechanical ventilation in sepsis. Tachypnea, associated with alveolar hyperventilation, is
frequently seen in early sepsis. Hypoxemia is commonly observed, but may be masked by hyperventilation. Ventilationperfusion mismatch and shunt are the usual
causes of hypoxemia. Other pulmonary manifestations include respiratory muscle dysfunction and bronchoconstriction. The most common cause of acute lung injury or
ARDS is sepsis, with a mortality rate reported to be as high as 40–60%18. ARDS is extensively reviewed in Chapter 14.
Central nervous system changes in severe sepsis include disorientation, seizures, lethargy or obtundation. Symptoms and signs of encephalopathy including nonfocal
neurologic findings may be seen. Evidence of polyneuropathy, including muscle weakness, wasting, and impaired deep tendon reflexes are part of the syndrome of
critical care polyneuropathy.
Hematologic failure is another hallmark of severe sepsis. Leukocytosis is the most common hematologic manifestation, but some patients may be leukopenic.
Thrombocytopenia and coagulation abnormalities are often seen. Sepsis is the most common cause of disseminated intravascular coagulation (DIC),
Page 90
occurring in 10–20% of patients with Gramnegative bacteremia and 70% of patient with septic shock19. DIC is the most important independent predictor of ARDS,
multiorgan system dysfunction and death. DIC and sepsis can present either as a hemorrhagic syndrome or, less commonly, as a microvascular thrombosis leading to
tissue ischemia and endorgan damage. The treatment of DIC remains controversial. There is no conclusive evidence that heparin treatment reduces the morbidity or
mortality associated with DIC.
Blood component therapies are potentially hazardous in the treatment of patients with DIC. Components such as fibrinogen may become a substrate for higher levels
of fibrin degradation products, which will further impair homeostasis. The administration of FFP and platelets should be restricted to patients with acute hemorrhage
plus a prolonged prothrombin time and/or a severe decrease in platelet count, respectively.
Transfusion of red blood cells in patients with sepsis may cause microcapillary occlusion and tissue ischemia. Transfusions of red blood cells have also been shown to
be immunosuppressive and to increase the risk of postoperative infections and organ failure in surgical patients. A recent study did not demonstrate any benefit of
increasing the transfusion threshold from 7 to 10 g dl−1 in the treatment of sepsis3.
Abnormal gastric emptying or adynamic ileus are the most common gastrointestinal problems. Stress ulceration was a common problem before employment of
prophylactic H2blockers or sucralfate. Transient extreme elevations in serum transaminase levels may follow an episode of severe shock or hypotension. Recent data
suggest that early enteral nutrition, particularly with immune enhancing diets, may improve the hormonal, metabolic and immunologic derangements that occur in critically
ill patients with sepsis20. In those patients who do not tolerate gastric tube feedings, postpyloric feedings should be considered.
A variety of cardiovascular manifestations are observed in severe sepsis, the most common being tachycardia. Frequently, ejection fractions are depressed, the right
and left ventricles are dilated, and the diastolic pressurevolume relationships are altered. Initially, before shock onset, the skin is warm, the pulse volume is increased,
and pulse pressure is increased. Cardiac output is usually normal or elevated, and systemic vascular resistance (SVR) is commonly decreased. Despite cardiac output
elevations, serum lactate levels are often elevated with progressive shock.
6.5.6 Sepsis trials

The systemic inflammatory response syndrome may result from a variety of physiologic insults such as infections, trauma, and pancreatitis. Sepsis is characterized by a
constellation of clinical signs including tachycardia, fever, hyperthermia and tachypnea. These clinical signs result from the release of inflammatory mediators including
cytokines, proteases, and eicosanoids, which are a component of the body’s normal physiologic response to the aforementioned insults. The excessive release of these
inflammatory mediators results in the development of diffuse capillary injury, intravascular coagulation with microvascular thrombosis, and multiorgan failure, a
condition frequently called severe SIRS or in the presence of infection, severe sepsis. Over the past few years a number of therapeutic interventions aimed at inhibiting
or reducing the extent of the body’s inflammatory response have undergone extensive phase III trials. These therapies have met with only limited success and none have
resulted in a statistically significant improvement in 28 day mortality rates. Antiendotoxin therapies have been investigated in large populations of adult patients with
presumed sepsis21,22. Although initial results were encouraging, particularly in patients with Gramnegative sepsis, larger clinical trials have shown no benefit. Several
clinical trials have been evaluated with monoclonal antiTNF (tumor necrosis factor) antibodies in severe sepsis populations23,24. No benefit has been found with any
antiTNF alpha therapy in patients in septic shock.
Page 91
Ibuprofen, a powerful antiinflammatory agent, failed to demonstrate any effect on mortality, shock or acute respiratory failure, in a large multicenter trial25. Additional
studies are needed to determine whether some patients, for example those with hypothermia, could benefit from the drug. More recently, investigations have looked at
the complex interaction between the inflammation and coagulation systems. Proinflammatory cytokines activate coagulation cascades, in particular via an effect upon
tissue factor, which is a key player in the coagulation cascade. They can also reduce fibrinolysis and profoundly reduce the levels of protein C and antithrombin III,
which are important anticoagulant agents26. Antithrombin III inhibits several coagulation factors of the extrinsic pathways such as factors IXa, XIa and XIIa, in addition
to factors Xa, IIa and plasmin. Activated protein C inhibits factors Va, VIIIa, and plasminogen activator inhibitor I (PAI1). The overall net effect during sepsis is a
marked procoagulant balance. In human studies, both antithrombin III and protein C levels are sharply decreased and mortality of septic patients is inversely correlated
with the levels of these two products. Recently, several trials have evaluated the compounds antithrombin III, protein C and tissue factor protein inhibitor in the
treatment of severe sepsis and septic shock. A large multicenter prospective trial has recently been completed and showed no significant improvement in survival with
the employment of antithrombin III in severe sepsis27. A recently completed phase III trial with recombinant Activated Protein C (rhAPC) showed a significant
reduction in 28 day mortality in severe sepsis patients. The data also revealed a 20% relative risk reduction when employing rhAPC compared with placebo28. Tissue
factor protein inhibitor (TFPI) is another endogenous anticoagulant that inhibits the extrinsic pathway of blood coagulation triggered by activation of tissue factor in
sepsis. TFPI complexes with tissue factor, factor VIIa and factor Xa to inhibit thrombin generation and fibrin formation. A phase III multicenter trial is currently under
way to evaluate the efficacy of this compound in severe sepsis29.
The definition of sepsis and septic shock used in most clinical trials has not included consideration of the length of time that the process has been present nor its
anatomic site. Even though recent clinical trials have targeted specific mediators including endotoxin, Il−1 or TNFalpha, which are postulated to have a pivotal role in
the inflammatory cascade leading to organ system failure and sepsis, the presence of excessive levels of these mediators was not required for entry into the study.
Therefore, the lack of a requirement to recruit demonstrably infected patients with elevated plasma levels of endotoxin or of cytokines may have adversely affected
outcome in trials examining immunomodulatory agents in sepsis. Although a metaanalysis combining most of the clinical trials using antiinflammatory agents has
suggested that benefit in survival could be achieved with such therapies, the magnitude of such effect was small30.
6.6 Anaphylactic shock

Anaphylaxis is a systemic form of immediate hypersensitivity that may progress to shock. It is a true medical emergency that is associated with a mortality rate
approaching 10%. Anaphylactic shock is classically triggered by antigenic stimuli that initiate a sequence of events mediated by IgE, which results in the release of
potent mast cells and basophilderived biochemical mediators. The parenteral route of drug administration is most commonly implicated in severe reactions, but oral,
topical and inhalational exposure can all cause anaphylaxis (Table 6.6). Anaphylactoid reactions present similarly to anaphylactic reactions, but unlike the latter are not
IgE mediated. Among the agents eliciting mediator release independent of IgE are iodinated contrast materials, dextran, and some local anesthetics.
Most cases of true anaphylaxis are related to the use of antibiotics, primarily penicillin and cephalosporins. An acute allergic reaction develops in as many as 40 out
of every 100,000
Page 92
Table 6.6 Etiologies of anaphylaxis
IgE mediated
Antibiotics
Foods
Insect venom
Foreign proteins (hormones, enzymes)
Industrial chemicals
Non IgE mediated (anaphylactoid)
Iodinated contrast materials
Dextran
Dialysis membranes
Blood products
patients treated with penicillin31. The administration of radiocontrast agents is the most common cause of serious anaphylactoid reactions, with a rate of 1 per 1000 to 1
per 14,000 procedures32. Classic antibodymediated anaphylaxis can also occur as a result of stings from bees, wasps, and hornets. Activation of the complement
system has been suggested as the cause of anaphylaxis associated with the use of dialysis membranes.
Mediators released with anaphylactic reactions increase microvascular permeability and bronchial reactivity, clinically manifested by airway obstruction (laryngeal
edema), impaired pulmonary gas exchange and hypovolemic shock. Activation of the complement system coagulation cascade via IgG and IgM antibody exposure to
antigens will produce additional adverse systemic effects.
6.6.1 Clinical presentation

The presentation of anaphylaxis is characteristically acute and will vary depending on the rate of absorption of the offending stimuli, portal of entry and the patient’s
degree of hypersensitivity. A reaction may occur within minutes and in most instances the signs and symptoms begin within 1 h of exposure. Occasionally, clinical
findings may appear many hours after exposure, and may be recurrent. Generally, the more rapid the onset and progression of features, the more severe the reaction.
Physical findings are highly variable and may include diffuse erythema and pruritis. If swelling involves the tongue or the lips, swallowing may be compromised and it
may become difficult to keep a patent airway. Laryngeal, epiglottic or airway edema may result in acute obstruction that is often characterized by respiratory stridor.
Involvement of the lower airway leading to shortness of breath, wheezing and chest tightness is also common. Patients with anaphylactic shock may be susceptible to
pneumothoraces and acute pulmonary emphysema produced by forced expiration against a closed edematous glottis. Signs of circulatory shock may characterize
severe cases of anaphylaxis and may include hypotension and tachycardia, respiratory arrest, oliguria, and altered mental status. If hypotension persists beyond the first
few minutes, established anaphylactic shock is present. Persistent hypotension during anaphylaxis is attributed to an increased vascular permeability with leakage of
plasma proteins, which results in hypovolemia and vasodilation. Increased microvascular permeability secondary to this type of distributive shock may cause
hemoconcentration. Supraventricular or ventricular cardiac arrhythmias may occur secondary to either myocardial ischemia or elevation of circulatory catecholamine
levels. Periorbital and perioral angioedema can be seen. Recent physical activity, ethanol ingestion and betaadrenergic blocker therapy appear to exacerbate
anaphylactic reactions. Crampy abdominal pain, nausea, vomiting, who have anaphylaxis. Conditions that may and diarrhea may also be present in patients mimic
anaphylaxis include vasovagal episodes, acute cardiac and pulmonary events, carcinoid syndrome attacks, monosodium glutamate toxicity, and drug overdose.
6.6.2 Management of anaphylaxis

The treatment of anaphylaxis should be priority based, and physicians should be aware of the proper steps in its management. Most patients with anaphylaxis respond
to initial therapy, although they may display prolonged persistent reactions despite appropriate therapy. A biphasic anaphylactic reaction may be observed with a
reoccurrence of anaphylactic
Page 93
symptoms within 12–24 h after an initial favorable response to therapy. Continuous monitoring with electrocardiogram is important, as dysrhythmias are characteristic
of these reactions, and may accompany therapy with bronchodilators, epinephrine, and other catecholamines.
The initial management of anaphylaxis should include the establishment of a patent airway, removal or reduction of systemic absorption of the toxin, pharmacological
support with epinephrine and the establishment of an appropriate circulating blood volume with fluid resuscitation. Retained stingers from bees, wasps, or yellow
jackets must be removed completely. A proximal constricting band can be applied to delay the absorption of venom if the site of evenomization is an extremity. It
should be tightened sufficiently to impair venous flow from the involved area without impeding arterial flow to the distal extremity. Local epinephrine injection may
produce localized vasoconstriction and retard system absorption. The employment of suctioning kits to remove the snake venom is controversial. These efforts are
probably ineffective if more than 30 min have elapsed since the evenomization and may produce further tissue injury or enhance venom absorption.
A plan to provide sequential therapy for patients with acute anaphylaxis should be instituted rapidly and systematically and include stabilizing the patient’s airway,
administering parenteral epinephrine, and initiating treatment to correct hypotension. With increasing bronchospasm or laryngeal edema, endotracheal intubation should
be recommended or if not possible, emergent cricothyroidotomy.
Parenteral epinephrine therapy is the standard treatment for anaphylactic crisis. Epinephrine inhibits further mediator release from basophils and mast cells, reducing
bronchoconstriction, increasing vascular tone and improving cardiac output. Two concentrations are available: a 1:1000 solution (1.0 mg ml−1) and a 1:10,000 solution
(0.1 mg ml−1). Epinephrine therapy should be monitored, as it may produce lifethreatening dysrhythmias and enhance ventricular irritability. Epinephrine should be
given parenterally for severe reactions in a dose of 3–5 ml of a 1:10,000 solution (0.3–0.5 mg). Although epinephrine can be instilled into the tracheobronchial tree via
an endotracheal tube or directly through the cricothyroid membrane, the absorption of epinephrine is less predictable by this route. Subcutaneous or intramuscular
epinephrine may be injected in a dose of 0.3–0.5 ml of a 1:1000 solution (0.3–0.5 mg). Repeated dose of epinephrine may be employed at 15 min intervals if
symptoms persist and if adverse effects have not been observed. A continuous infusion of epinephrine at 0.05–2.0 µg kg−l min−1 can be titrated to response.
Patients with a history of radiocontrast allergy should receive a prophylactic regimen before undergoing diagnostic radiographic testing33. The use of nonionic low
osmolality radiocontrast media may minimize the likelihood of an adverse reaction (Table 6.7).
Fluid therapy reverses the significant hypovolemia in anaphylaxis secondary to an increase in vascular permeability. In cases of mild anaphylaxis, intravenous fluids
may not be required, as abnormalities in vascular permeability remain minor. Isotonic crystalloid fluids or colloidal fluids are preferred, as they produce the most rapid
expansion in the intravascular space. When the patient remains hypotensive despite fluid replacement and epinephrine, pharmacologic support should be instituted (see
Chapter 8). Pure alphaadrenergic stimulation is to be avoided, as it may enhance
Table 6.7 Prophylactic therapy for radiocontrast sensitivity
Prednisone (50 mg) 13 h, 7 h, and 1 h before exposure

Diphenhydramine (50 mg) IM. or orally, 1 h before exposure
Emergent procedure
Hydrocortisone (200 mg) IV immediately and 4 h post procedure
Diphenhydramine (50 mg) before exposure
Page 94
mediator release. However, if circulatory shock persists, alphaadrenergic agents may need to be employed to maintain organ perfusion.
Beta agonists and oxygen therapy should be initiated if bronchospasm persists despite epinephrine therapy. Aminophylline is a bronchodilator that is useful in the
management of patients who continue to exhibit bronchospasm after initial treatment with epinephrine. Most bronchodilators have associated significant toxic effects
including lifethreatening dysrhythmias and seizures. Nebulized epinephrine can be used in mild laryngeal edema, but has no role in the treatment of severe edema,
stridor, and respiratory distress.
Corticosteroids are recommended for anaphylactic shock to delay recurrent episodes by inhibiting a secondary wave of mediator release. Methylprednisolone may
be given intravenously at doses of 60–125 mg at 4–6 h intervals for 24 h and then tapered to prevent relapse. Therapies for anaphylaxis may be less effective in
patients with asthma or those with a history of using betaadrenergic blocking agents. These conditions decrease the responsiveness of bronchial smooth muscle to
therapy and there is a potential for unopposed alpha adrenergic stimulation and severe hypertension in those patients given epinephrine while receiving betaadrenergic
block agents. In these situations, glucagon can be administered at a rate of 5–15 ml min−1.
Antihistamines are not very helpful in the initial management of anaphylaxis, but may be administered while the patient is being stabilized. Nonselective antihistamines
inhibit histamine (H1 and H2) receptors, thus blocking the systemic effects of H1 receptor mediated increase in vascular permeability and bronchoconstriction. These
agents may be tried in patients who fail to respond to the initial treatment of anaphylaxis. Diphenhydramine, an H1 antagonist, may be given parenterally in doses of 25–
50 mg at 4–6 h intervals. Theoretically, cimetidine, an H2 antagonist, may also be given intravenously (300 mg) and repeated at 6–8 h intervals in patients with
anaphylactic shock.
6.7 Cardiogenic shock

Cardiogenic shock includes conditions that cause failure of the heart as a pump and result in hypoperfusion of body organs. Impairment of diastolic function, systolic
function or combination of these impairments can cause the heart to fail as a pump. However, the majority of patients in cardiogenic shock have sustained marked
depression of ventricular systolic dysfunction, associated with a decrease in stroke volume and cardiac output. Cardiogenic shock is most commonly a consequence of
myocardial ischemia and/or infarction. Complications of an infarction include pump failure, ruptured ventricular wall or septum and disrupted papillary muscle. Other
potential causes include cardiomyopathy arrhythmias, and valvular heart disease. Pulmonary artery catheterization in the setting of cardiogenic shock will classically
show a low cardiac index (≤1.8 l min−1 m−1) and an elevated pulmonary artery wedge pressure (18–25 cm H2O). Approximately 10–15% of all patients with acute
myocardial infarction will develop cardiogenic shock within 1 week of the initiating event34. In about 50% of this subset, this occurs within the first 48 h of an infarction.
Clinically, patients commonly present with hypotension, tachypnea, a deceased pulse pressure, diaphoresis, cool extremities, and altered mental status. Cardiac
examination may reveal tachycardia with an S3 and/or S4 gallop.
The recognition of cardiogenic shock is frequently delayed because of confusion regarding the shock syndrome in general. Although the diagnosis of cardiogenic
shock is based on evidence of hypoperfusion of organs, some of the more established signs of shock such as hypotension and oliguria may not be present during its
early phases or in shock of mild to moderate severity. The clinical manifestations of hypotension and oliguria usually indicate advanced disease and/or decompensation.
Pharmacological interventions can be tailored by the evaluation of hemodynamic profiles via pulmonary artery catheterization.
Page 95
6.7.1 Clinical presentations of cardiogenic shock

Approximately 20% of postinfarction shock patients manifest low ventricular filling pressures (low PAWP). The treatment goal is to improve systemic hypoperfusion
and to normalize systemic blood pressure by augmenting stroke volume and cardiac output via fluid resuscitation with IV normal saline. Left ventricular filling pressures
(PAWP) should generally be augmented from 12 mmHg or less to target pressures of 15–18 mmHg. A Frank Starling curve can be helpful to guide therapy in an
individual patient. Fluid administration should be undertaken cautiously to avoid the development of acute volume overload or pulmonary edema. In those patients who
do not display adequate perfusion after fluid infusion therapy, pharmacological support is usually required (see Chapter 8).
The majority of patients with post infarction cardiogenic shock will present with a normal to high PAWP, mild to moderate systemic hypotension, low to normal
cardiac output and systemic hypoperfusion. Systemic hypoperfusion may occur in the presence of left ventricular filling pressures (PAWP) of 15 mmHg or more,
systemic systolic pressures of between 70 and 100 mmHg and decreased stroke volume and cardiac output. Inotropic therapy aimed at improving ventricular
contraction is the most direct and effective form of therapy in this subset of patients. Dobutamine is the preferred agent in these circumstances. In the presence of left
ventricular dysfunction, high left ventricular filling pressures, mild to moderate hypotension and reduced peripheral perfusion, dobutamine (2–15 µg kg−1 min−1)
improves ventricular performance, stroke volume and cardiac output, decreases ventricular filling pressures, normalizes systemic blood pressure, and improves
peripheral perfusion. Dobutamine can be titrated by increments of 2–3 µg kg−1 min−1 every 15−30 min until the clinical targets are met. Milrinone, an agent with both
inotropic and vasodilatory actions, can also reduce both left and rightsided filling pressures while improving cardiac output.
6.7.2 Severe cardiogenic shock

Severe cardiogenic shock presents with markedly elevated left ventricular filling pressure (high PAWP), decreased stroke volume and cardiac index (often <1.8 l min−1
m−2), marked hypotension (systolic pressure <70 mmHg), and evidence of organ hypoperfusion. The mortality rate in this group of patients can be greater than 90%.
Either dobutamine or milrinone may be inappropriate as first line pharmacologic therapy, as both drugs can cause marked hypotension owing to vasodilation. Initial
therapy for severe shock should include an agent that acts as a vasopressor. Dopamine is both an inotrope and vasopressor at higher doses and will generally elevate
blood pressure to levels acceptable for brain, kidney, and cardiac perfusion. Renal and cerebral perfusion require systolic blood pressures of greater than 75–80
mmHg, whereas coronary perfusion requires a systemic diastolic pressure of 60 mmHg or more. If escalating doses of dopamine do not increase blood pressure, or
unacceptable tachycardia occurs, employment of a more selective alphaadrenergic agonist such as norepinephrine is warranted despite the increased afterload effect
and increasing myocardial oxygen demand.
6.7.3 Intraaortic balloon counterpulsation

With failed medical therapy, intraaortic balloon counterpulsation (IABC) should be considered. In patients with an identifiable, reversible cause of cardiogenic shock,
IABC is a form of bridge therapy, providing hemodynamic support until more definitive intervention (angioplasty, bypass surgery) is pursued. IABC effectively
improves tissue and coronary perfusion through its effects on decreasing afterload, with subsequent improvement in LV performance and decreasing myocardial oxygen
consumption. IABC is specifically indicated in acute mitral regurgitation, or ventricular septal rupture. Hemodynamic stabilization should be followed by a diagnostic
cardiac catheterization, angioplasty and/or a surgical procedure.
Page 96
6.8 Obstructive shock

This category of shock involves impedance to the flow of blood resulting in a decrease in cardiac filling. The causes involve mechanisms other than primary myocardial
or valvular dysfunction. Etiologies include impaired diastolic filling (cardiac tamponade, constrictive pericarditis, pneumothorax), and an increase in right or left
ventricular afterload (massive pulmonary embolism, acute pulmonary hypertension, aortic dissection). The differential diagnosis is extremely important in this shock
state, as a physical intervention (surgery, pericardiocentesis, or chest tube insertion) may reverse the condition and can be life saving. Disorders that impede blood flow
are often overlooked and must always be considered in appropriate settings. Maintenance of intravascular volume is extremely important in patients with obstructive
shock. Fluid resuscitation may temporarily enhance the cardiac output and hypotension. Diuretics should be avoided in this type of shock. Inotropes and/or
vasopressors probably have a minimal, often temporary role in the management of a patient with obstructive shock.
6.8.1 Cardiac tamponade

Tamponade is caused by the accumulation of fluid in the pericardial space under abnormally increased pressure (Table 6.8). Although physicians tend to diagnose
tamponade in the presence of hypotension, a decrease in blood pressure is usually a preterminal development that is often preceded by an increase in the central venous
pressure and a decreasing cardiac output. Central venous pressures are always elevated, except in situations of ‘lowpressure’ tamponade, which is due to pericardial
effusion in the presence of intravascular volume depletion. Equalization of right and leftsided filling pressures is present in tamponade when the pericardial infusion is
not loculated. Echocardiography is the most appropriate technique for diagnosing pericardial effusion in the ICU setting, but computer tomograhpy and magnetic
resonance imaging (MRI) are appropriate diagnostic intervention when echocardiography is inconclusive. Tamponade is assumed to be present when there is a
moderate or large pericardial effusion in the presence of a central venous pressure greater than approximately 15 mmHg. Fluid resuscitation will not usually precipitate
pulmonary edema in these patients, as congestion is completely limited to the systemic venous compartment rather than in the lungs. Classically, chronic tamponade
does not reduce coronary blood flow or induce ischemic myocardial dysfunction. The myocardial function appears to be normal in this shock state. Vasodilators and
venodilators are not beneficial in the treatment of cardiac tamponade.
6.9 Shock states

The four categories of shock reviewed here are not mutually exclusive and in many cases, shock may have several causes as the clinical situation progresses. For
example, hypovolemic shock may precipitate ischemia, or septic shock may induce volume depletion through third spacing of capillary fluid. One needs to evaluate
each of the determinants of
Table 6.8 Etiology of pericardial tamponade
• Postoperative hemorrhage (cardiac surgery)

• Anticoagulation
• Esophageal perforation postsclerotherapy
• Mechanical ventilation induced barotrauma
• Perforation of heart by indwelling catheter
• Internal jugular or subclavian vein puncture
• Balloon angioplasty perforation of coronary vessels
• Myocardial infarction
• Uremia
• Idiopathic pericarditis
• Neoplastic pericarditis
• Traumatic pericarditis
Page 97
perfusion pressure when trying to identify precipitating or aggravating factors leading to shock.
References
1. Bond R.F. and Green, H.D. (1969) Cardiac output redistribution during bilateral common carotid occlusion. Am. J. Physiol. 216:393.
2. Davis, J.O. and Freeman, R.H. (1976) Mechanisms regulating rennin release. Physiol. Rev. 56: 1.
3. Hebert, P.C., Wells, G., Blanchman, M.A., et al. (1999) A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion
Requirements in Critical Care Investigators, Canadian Critical Care Trials Group. N. Engl J. Med. 340:409–417.
4. Abraham, F., Wunderink, R., Silverman, H., et al. (1995) Efficacy and safety of monoclonal antibody to human tumor necrosis factor alpha in patients with sepsis
syndrome. JAMA 273: 934–941.
5. Bone, R.C., Balk, R.A., Cerra, F.B., et al. (1992) American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: Definitions for
sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest 101:1644–1655.
6. RangelFrausto, M., Pittet, D., Costigan, M., et al. (1995) The natural history of the systemic inflammatory response syndrome (SIRS). JAMA 273:117–123.
7. Fisher, C.J., Dhainaut, J.F., Opal, S.M., et al. (1994) Recombinant human interleukin I receptor antagonist in the treatment of patients with sepsis syndrome. Results
from a randomized, doubleblind, placebocontrolled trial. JAMA 271:1836–1843.
8. Taylor, F.B., Change, A.C., Peer, G.T., et al. (1991) DEGRfactor Xa blocks disseminated intravascular coagulation initiated by Escherichia coli without preventing
shock or organ damage. Blood 78:364–8.
9. Extermann, M., Regamey, C., Humair, L., et al. (1995) Initial treatment of sepsis in nonneutropenic patients: ceftazidime alone versus ‘best guess’ combined
antibiotic therapy. Chemotherapy 41:306–315.
10. Tuchschmidt, J., Fried, J., Swinney, R. and Sharma, O.M. (1989) Early hemodynamic correlates of survival in patients with septic shock. Crit. Care Med.
17:719–723.
11. Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine (1999) Practice parameters for hemodynamic support of
sepsis in adult patients in sepsis. Crit. Care Med. 27:639–660.
12. Soni, A., Pepper, G.M., Wyrwinski, P.M., et al. (1995) Adrenal insufficiency occurring during septic shock: Incidence, outcome, and relationship to peripheral
cytokine levels. Am. J. Med. 98:266–271.
13. Rasmuson, S., Olsson, T. and Hagg, E. (1996) A low dose ACTH test to assess the function of the hypothalamicpituitaryadrenal axis. Clin. Endocrinol. 44:151–
156.
14. Annane, D. (2000) Effects of the combination of hydrocortisone (HC)fludrocortisone (FC) on mortality in septic shock. Crit. Care Med. 28(12 Suppl.):A46.
15. Brivet, F.G., Kleinknecht, D.J., Loirat, P. and Landais, P.J.M. (1996) Acute renal failure in intensive care units—causes, outcome and prognostic factors of
hospital mortality: a prospective, multicenter study Crit. Care Med. 24:192–198.
16. Olson, D., Pohlman, A. and Hall, J.B. (1996) Administration of lowdose dopamine to nonoliguric patients with sepsis syndrome does not raise intramural gastric
pH no improve creatinine clearance. Am. J. Respir. Crit. Care Med. 154: 1664–1670.
17. Davenport, A., Will, E.J. and Davidson, A.M. (1993) Improved cardiovascular stability during continuous modes of renal replacement therapy in critically ill
patients with acute hepatic and renal failure. Crit. Care Med. 21:328–338.
18. Kollef, M.H. and Schuster, D.P. (1995) The acute respiratory distress syndrome. N. Engl J. Med. 332:27–34.
19. Kreger, B.E., Craven, D.E. and McCabe, W.R. (1980) Gramnegative bacteremia. IV. Reevaluation of clinical features and treatment in 612 patients. Am. J.
Med. 68:344–355.
20. Bower, R.H., Cerra, F.B., Bershadsky, B., et al. (1995) Early enteral administration of a formula (Impact) supplemented with arginine, nucleotides, and fish oil in
intensive care unit patients: results of a multicenter, prospective, randomized, clinical trial. Crit. Care Med. 23: 436–449.
21. Calandra, T., Glauser, M.P., Schellekens, J., et al. (1988) Treatment of Gramnegative septic shock with human IgG antibody to Escherichia coli J5: a
prospective, doubleblind, randomized trial. J. Infect. Dis. 158:312–319.
22. McCloskey, R.V., Straube, R.C., Sanders, C., et al. (1994) Treatment of septic shock with human monoclonal antibody HA1A. A randomized, doubleblind,
placebocontrolled trial. CHESS Trial Study Group. Ann. Intern. Med. 121:1–5.
Page 98
23. Abraham, E., Wunderink, R., Silverman, H., et al. (1995) Monoclonal antibody to human tumor necrosis factor alpha (TNF Mab): efficacy and safety in patients
with the sepsis syndrome. JAMA 273:934–941.
24. Abraham, E., Anzueto, A., Gutierrez, G., et al. for the NORASEPT II Study Group (1998) Monoclonal antibody to human tumor necrosis factor alpha (TNF
MAb) in the treatment of patients with septic shock: a multicenter, placebo controlled, randomized, doubleblind clinical trial. Lancet 351:929–933.
25. Bernard, G.R., Wheeler, A.P., Russell, J.A., et al. (1997) The effects of Ibuprofen on the physiology and survival of patients with sepsis. N. Engl. J. Med.
336:912–918.
26. Vervloet, M.G., Thijs, L.G. and Hack, C.E. (1998) Derangements of coagulation and fibrinolysis in critically ill patients with sepsis and septic shock. Semin.
Thromb. Hemost. 24:33–44.
27. Riess, H. (2000) Antithrombin in severe sepsis. ‘New’ indication of an ‘old’ drug. Intensive Care Med. 22: S471–S481.
28. Bernard, G.R., Vincent, J.L., Laterre, P.F., et al. (2001) Recombinant human Protein C World wide Evaluation in Severe Sepsis (PROWESS) study group:
Efficacy and safety of recombinant human activated protein C for severe sepsis. N. Engl J. Med. 344:699–709.
29. Abraham, E. (2000) Coagulation abnormalities in acute lung injury and sepsis. Am. J. Respir. Cell Mol. Biol. 22:401–404.
30. Zeni, F., Freeman, B. and Natanson, C. (1997) Antiinflammatory therapies to treat sepsis and septic shock: a reassessment. Crit. Care Med. 25: 1095–1100.
31. Carlson, R.W., Bowles, A.L. and Haupt, M.T. (1986) Anaphylactic, anaphylactoid and related forms of shock. Crit. Care Med. 2:347–372.
32. Sim, T.C. (1992) Anaphylaxis: How to manage and prevent this medical emergency. Postgrad. Med. 92:277–296.
33. Lasser, E.C., Berry, C.C., Talner, L.B., et al. (1987) Pretreatment with corticosteroids to alleviate reactions to intravenous contrast material. N. Engl. J. Med.
317:845.
34. Bengston, J.R., Kaplan, A.J., Pieper, K.S., et al. (1992) Prognosis in cardiogenic shock after acute myocardial infarction in the interventional era. J. Am. Coll
Cardiol. 20:1482.
Page 99
Chapter 7
Fluid resuscitation in the ICU
Laurie A.Loiacono, MD
Contents
7.1 Introduction
7.2 Definitions
7.3 Physiology of resuscitation
7.4 Strategy
7.5 Red blood cells (RBCs)
7.6 Crystalloid products
7.7 Cellfree colloids
7.8 Summary and conclusions
7.9 Recommendations for fluid resuscitation in the ICU
Page 100
7.1 Introduction
Recognition and treatment of fluid imbalance is particularly important in the critically ill patient, especially as the normal homeostatic mechanisms (thirst, salt craving,
renal elimination) may be bypassed or inoperative. Unrecognized fluid imbalance in this patient population can lead to severe metabolic derangements and shock states.
Hypovolemia is a common cause of circulatory shock; patients can become relatively or absolutely volume depleted as the result of hemorrhage, fluid shifts (‘third
spacing’), obligate osmotic diuresis, sweating, therapeutic maneuvers or acute vasodilalation (i.e. SIRS or sepsis). The differential diagnosis of clinical shock states is
covered in Chapter 6; common clinical situations that often require aggressive fluid resuscitation are listed in Table 7.1. Timely recognition and treatment of fluid deficits
can affect ect patient survival and minimize associated morbidity such as development of acute renal failure, acute lung injury and adult respiratory distress syndrome,
myocardial infarction, mesenteric ischemia, stroke, and multisystem organ dysfunction. This chapter will highlight principles of resuscitation and the advantages and
disadvantages of various replacement fluids.
7.2 Definitions
Acute resuscitation—fluid administration that restores intravascular volume.
Albumin—a relatively low molecular weight serum protein present in plasma, which account s for approximately threequarters of normal plasma colloid osmotic
pressure. Albumin can be manufactured into a proteinbased solution and administered as a colloid volume expander.
Chronic resuscitation—fluid administration directed toward maximizing and monitoring endorgan perfusion.
Colloid—a solution composed of osmotically active molecules that are relatively impermeable to capillary membranes and therefore create an osmotic gradient (or
pressure) across a membrane.
Colloid osmotic pressure (COP)—the fluid diffusion pressure created across a capillary membrane by differences in protein concentration on either side of the
membrane.
Crystalloid—an electrolytebased solution of osmotically inactive particles that freely equilibrate across capillary membranes.
Donnan forces—electrostatic forces generated across a membrane by large, negatively charged osmotically active protein molecules that determine differential ion
concentrations and total water flux across that membrane between two compartments.
Hetastarch—a synthetic colloid composed of hydroxyethylsubstituted branched chain amylopectin molecules supplied in normal saline solution (typically c. 310
mOsm l−1) as Hespan™ or in balanced electrolyte solution as Hextend™.
Hydrostatic pressure—the pressure created across a membrane by fluid flow through a space.
Hypertonic—a solution that has an osmolarity higher than plasma (typically >300 mOsm l−1).
Hypotonic—a solution that has an osmolarity lower than plasma (typically <270–280 mOsm l−1). Hypotonic solutions include halfnormal saline solution.
Isotonic—a solution that has an osmolarity similar to plasma (typically 280–300 mOsm 1−1). Isotonic solutions include lactated Ringer’s and normal saline solutions.
Table 7.1 Common circulatory shock states requiring aggressive fluid resuscitation
Hypovolemic states Distributive shock states with ongoing fluid loss

Hemorrhage Pancreatitis
GI loss (diarrhea, vomiting) SIRS
Ascitic leak Sepsis
Anaphylaxis
Page 101
Osmolarity—a measure of the number of osmotically active particles in a solution.

Resuscitation—the restoration of adequate and oxygenated tissue perfusion in an effort to avoid tissue loss or injury.
Total osmotic pressure (TOP)—the fluid diffusion pressure created by the COP plus Donnan forces of the cation concentrations in solution.
7.3 Physiology of resuscitation

Although the pathophysiology of various shock states will vary significantly, the principles of physiologic support are remarkably consistent. Restoration of tissue
oxygenation and perfusion are paramount for tissue preservation and recovery from shock. Acute tubular necrosis (ATN) and acute lung injury can occur following
even transient periods of significant hypotension. In shock, circulating catecholamines constrict prearteriolar capillary sphincters, globally decreasing capillary perfusion
and potentially producing significant tissue hypoperfusion and injury. Although the insult may be brief, endorgan effects can be prolonged or irreversible, and are
probably related to these microcirculatory changes that occur to preserve macrovascular circulating blood flow.
7.3.1 Fluid compartments

Water comprises about 60% of body weight, but intravascular volume is only a small portion (about 10%) of total body fluid (Figure 7.1). Twothirds of total body
water (TBW) is intracellular and contains plasma and proteins responsible for intracellular metabolism. Onethird of TBW is extracellular in a 3:1, interstitial
(extravascular):intravascular distribution. Fluid moves between the intracellular and extracellular spaces across cell membranes and fluid moves between intravascular
and extravascular spaces across capillary membranes.
7.3.2 Semipermeable membranes; fluid dynamics

Healthy cell membranes are semipermeable; therefore fluid shifts between the intracellular
Figure 7.1 Major compartmental distribution of total body fluid.
and extracellular compartments occur only with free water, by osmosis, and in response to differences in transmembrane solute and ion concentration gradients.
Electrolytes, most solutes and proteins do not freely cross the cell membrane unless the cell is injured or dying. Cations cross only through ATPdriven Na+−K+
pumps. Therefore, differences between intracellular and extracellular protein and total electrolyte concentrations (osmolarity) will determine fluid flux across cell
membranes, facilitating cellular hydration, dehydration or swelling.
Healthy capillary membranes are freely permeable to water and electrolytes but not proteins. Fluid flux between the intravascular and extravascular (interstitial)
space is therefore more dependent on transcapillary protein concentration gradients (i.e. COP) and transcapillary electrolyte concentrations are less significant, under
normal physiologic conditions. In addition, because there is a ‘driving’ flow (or hydrostatic pressure) through capillaries (Pc), transcapillary fluid dynamics are more
complex and other forces need to be considered.
The differences in permeability between cell and capillary membranes dictate in large part how different types of resuscitative fluids redistribute after administration
(Figure 7.2). Infused isotonic fluids such as normal saline and Ringer’s lactate are initially contained within the intravascular space but over time will equilibrate with the
extracellular space
Page 102
Figure 7.2 Redistribution of resuscitative fluids. Differences in permeability between cell and capillary membranes. ECC, extracellular
compartment; TBW, total body water.
(intravascular and interstitial water) across the electrolytepermeable capillary membranes. The intracellular space is less affected because cell membranes are not freely
permeable to electrolytes. Hypotonic solutions, on the other hand, rapidly distribute to a volume approximating total body water because both cell and capillary
membranes are permeable to free water. Dextrosecontaining solutions such as D5W have sufficient osmolarity to avoid hemolysis of red blood cells when initially
infused, but act as hypotonic solutions once the osmotic component (glucose) is metabolized. Replenishing lost intravascular fluid (plasma and/or whole blood) requires
roughly four times the desired replacement volume as crystalloid, and larger amounts of hypotonic solutions. Compartmental redistribution of fluid begins within minutes.
7.3.3 Starling’s principle

In 1896, Starling first described factors that influence the transfer of fluid across capillary membranes1. The Starling equation describes four forces that govern
transcapillary flow:
• Qf is total flow across the capillary membrane;

• Kf is the filtration coefficient (the rate of filtration min−1 mmHg−1 per 100 g tissue; varies from one tissue to another);
• Pc is capillary hydrostatic pressure (tends to force fluid out of the vascular space, varies from one end of the capillary to the other; mean Pc is c. 17 mmHg (see
Figure 7.4 (below));
• Pi is interstitial hydrostatic pressure (tends to draw fluid out of the vascular space, is usually negative; c. −6.3 mmHg);
• σ is the protein osmotic reflection coefficient (a measure of the ability of the capillary membrane to block protein movement; normal average is 0.9 for systemic
capillaries and 0.7 for pulmonary capillaries but may decrease as low as 0.4 in leaky capillary states);
Page 103
• COPc is the capillary, or plasma, osmotic pressure (tends to retain fluid in the capillary, the only Starling force that returns intravascular fluid, normally c. 28 mmHg
with 19 mmHg from protein and 9 mmHg from cations in solution);
• COPi is the interstitial osmotic pressure (tends to retain fluid in the interstitial space, normally c. 5 mmHg).
Figure 7.3 is a diagrammatic representation of capillary anatomy with transmembrane pressures and flows indicated.
7.3.4 Acute versus chronic resuscitation

It is generally understood that an ‘absolute’ and/or ‘relative’ blood volume deficit is the primary defect in shock states. Because of the catecholamine, cytokine and
complement release that occurs early in shock, it is helpful to separate acute from chronic resuscitative needs. We will exclude cardiogenic shock states from this
discussion, as they are to be discussed in Chapters 6 and 8.
The primary need in early hypovolemic and distributive shock is for rapid intravascular volume expansion. Aggressive fluid therapy with the goal of restoring
hemodynamic stability will minimize subsequent organ dysfunction. The composition of the fluid therapy is less important than the physical need to support right heart
filling (preload) leading to support of cardiac output and restoration of tissue oxygenation or perfusion. Generally, blood loss (i.e. hemorrhage) is replaced with blood
when possible, plasma loss (i.e. ascites) is replaced with plasma or colloid, and electrolyte loss (i.e. vomiting or diarrhea) is replaced with crystalloid.
Over the subsequent several hours to days, a postshock reactive systemic inflammatory response ensues and capillary membranes become leaky to fluid and
protein. Now the ‘intravascular availability’ of the fluid therapy becomes more important and more consideration should be given to the resuscitative fluid composition.
It is in this later phase
Figure 7.3 Diagrammatic representation of capillary anatomy with transmembrane pressures and flow represented.
Page 104
of shock and/or shock recovery where ‘chronic resuscitation’ is necessary, not only to meet basic maintenance fluid needs but also to replace the insensible
intravascular fluid losses that occur to the extravascular space as a result of fluid equilibration across leaky capillary membranes. In this chronic phase of resuscitation
colloids tend to have a physiologic advantage over more rapidly equilibrating edemagenic crystalloids.
Defining ‘the optimum resuscitative fluid’ is difficult and controversial. When compared with crystalloid, are colloids or hypertonic fluids more effective in maintaining
the integrity of intravascular volume, minimizing edema formation, better preserving interstitial osmolarity, and decreasing subsequent morbidity and mortality? Are there
advantages if colloids are given earlier and more consistently throughout the resuscitative period from shock states? Are there advantages or disadvantages to colloids
and hypertonic solutions other than osmotic support? As none of these questions have been conclusively answered, clinical practice tends to be variable.
7.4 Strategy
7.4.1 Initiation of circulatory support

Resuscitation begins with the ‘ABCs’: Airway, Breathing, and Circulation. Circulation implies the process of evaluating, establishing and maintaining adequate perfusion
via cardiovascular support. Intravenous access needs to be established in anticipation of large fluid volume requirements. Patients with significant traumatic injuries or
expected fluid shifts as the result of major thoracic and abdominal procedures should have at least two largebore (18 gauge or larger) peripheral lines. Alternatively,
central access can be established via the subclavian, internal jugular or femoral veins (see Chapter 5) with a wide variety of catheters ranging from single, double and
triplelumen CVP lines to largebore (usually 8 or 9 French) Cordis transducers, combined introducertriple lumen catheters, or Quintontype catheters normally used
for dialysis.
7.4.2 Endpoints of fluid therapy

Clinical criteria used to define the goals of fluid therapy are broadly defined because of the difficulties in standardizing and controlling study populations. Most clinicians
would consider a patient to be adequately resuscitated when the following criteria are met:
• PaO2 of 60–65 mmHg or more;
• pH greater than 7.25;
• Hematocrit (Hct) greater than 24%, higher with myocardial ischemia or organ failure;
• mean arterial pressure (MAP) ≥60 mmHg;
• urine output of 0.5 cm3 kg−1 or more in adults (in the absence of acute or chronic renal failure);
• wellperfused, warm extremities;
• oxygen delivery greater than 150 ml min−1 m−2.
It is possible, however, to meet some or all of these goals without adequate fluid resuscitation. An alphaadrenergic agent such as norepinephrine or phenylephrine (see
Chapter 8) will vasocontrict peripheral vessels and raise blood pressure, even if filling pressures are inadequate. Although this therapy would accomplish the goal of a
MAP greater than 60 mmHg, it does so at the expense of peripheral perfusion, and may ultimately result in lactic acidosis and organ failure.
Absolute central venous pressure or pulmonary artery occlusion pressure goals can be misleading because of variations in individual patient physiology. Although
conclusions about enddiastolic volumes can be drawn from measured pressures, such conclusions may be erroneous if pressurevolume relationships (i.e. compliance)
are altered. In general, low filling pressures (i.e. CVP<5, PCWP<8) suggest inadequate preload in patients with clinical symptoms of shock. Fluid challenges are helpful
in assessing volume replacement needs. To detect changes, fluid boluses (preferably crystalloid) have to be of sufficient volume (>500 ml in
Page 105
adults), rapidly infused (within 10 min) and carefully monitored (with the clinician at bedside). Physical examination, central venous and/or PCWP, cardiac output, urine
output, and arterial pressure give clues to the adequacy of fluid resuscitation. Following a fluid bolus, small (<3 mmHg), transient changes in filling pressures with
improved cardiac output or blood pressure indicate ongoing fluid needs. Sustained or large (>7 mmHg) increases in filling pressures without improved output mark the
point at which circulatory overload becomes a risk.
Fluid resuscitation can be tailored to the particular patient’s physiology based on the generation of a Starling curve. Clinically, the relationship between filling
pressures and cardiac output is plotted. (In the absence of a PA catheter, CVP can be plotted against arterial blood pressure.) Figure 7.4 displays such a curve based
on data from a patient before and after adequate fluid resuscitation. Pressor agents should be withheld until preload has been adequately addressed. The risks of over
resuscitation include hydrostatic pulmonary edema and diffuse tissue edema. In most cases, however, even gross fluid overload is a temporary problem, as much of the
fluid will diffuse into the interstitial space within hours, or can be managed via diuresis or dialysis. The risks of inadequate fluid resuscitation (organ failure) tend to be
more permanent and more devastating.
Figure 7.4 A curve based on data from a patient before and after adequate fluid resuscitation.
7.5 Red blood cells (RBCs)

Whole blood is the ideal fluid for replacement of intravascular volume deficits, but because of high demand for specific component therapy and technical difficulty in
longterm storage of wholeblood preparations, RBCs are most commonly provided as citrate anticoagulated packed RBCs. RBCs should be used to (i) replace blood
loss secondary to hemorrhage and (ii) individually maximize oxygencarrying capacity in critically ill patients. Whereas a hemoglobin of 7–8 g dl−1 is generally adequate
for stable preoperative patients1, this level may be insufficient in the critically ill patient with circulatory compromise and high oxygen consumption.
When rapid, highvolume (>100 ml min−1) citrated RBC infusions are required (i.e. massive gastrointestinal (GI) bleed or traumarelated hemorrhage), acute
hypocalcemia can occur as the result of citrate intoxication2. Clinical symptoms include skeletal muscle hyperexcitability, STT segment changes, depressed cardiac
output, hypotension, electromechanical dissociation, and ventricular fibrillation. Intravenous calcium replacement will correct these clinical symptoms, and should be
guided by serum ionized calcium levels in these typically hypoproteinemic peri or postresuscitative patients.
Other transfusionrelated risks are listed in Table 7.2.
7.6 Crystalloid products

Crystalloid solutions are electrolytebased solutions of osmotically inactive particles that freely equilibrate across capillary, but not cell, membranes. They can be mixed
and administered with or without dextrose solution. Addition of 5% dextrose to isotonic solutions will tend to increase the solution’s osmolarity by approximately 250
mOsm l−1. Dextrosecontaining solutions are seldom used for volume resuscitation, because (i) dextrose freely equilibrates between the intracellular and extracellular
spaces and is therefore of no value in terms of osmotic support, and (ii) patients in stressed
Page 106
Table 7.2 Most common acute blood transfusion risks
Hemolytic transfusion reactions (immediate)

Delayed transfusion reactions (3–21 days posttransfusion)
Cytokinemediated febrile reactions
Allergic reactions
Graft vs. host disease
Posttransfusion purpura
Metabolic disturbances (hypokalemia, hypothermia)
Immunosuppression
Infection (hepititis, HIV, EBV/CMV)
states will be glucose intolerant and can become hyperglycemic. Crystalloid solutions typically used for volume resuscitation are either isotonic or hypertonic (see
definitions).
7.6.1 Isotonic solutions (NS and LR)

The two most commonly used isotonic solutions are 0.9% normal saline solution (NS) and lactated Ringer’s solution (LR).
• 0.9% normal saline solution contains 9 g l−1 NaCl. It has an osmolarity of 308 mOsm l−1 (154 mequiv l−1 Na and 154 mequiv l−1 Cl). There is a risk of
hyperchloremic acidosis with rapid and/or prolonged administration of isotonic saline3–5.
• Ringer’s lactate contains 6 g of sodium per liter. It has an osmolarity of 273 mOsm l−1 (130 mequiv l−1 Na+, 109 mequiv l−1 Cl−, 3 mequiv l−1 Ca2+, 4 mequiv l−1
K+, and 28 mequiv l−1 lactate). There is less risk of hyperchloremic acidosis than with normal saline, and the potential advantage of a buffer effect on any ongoing
acidosis once the lactate in solution is converted to acetate in patients with preserved liver function.
Advantages of isotonic solutions

• Do not promote intracellularextracellular shifts.
• Effective immediate volume expansion in acute resuscitation.
• Readily available in most situations.
• Low cost.
Disadvantages of isotonic solutions

• Freely equilibrate between the intravascular and interstitial spaces over time leading to diminished intravascular volume. This occurs particularly rapidly in
hypoproteinemic patients. The intravascular ‘availability’ of these solutions as a result of rapid equilibration is, therefore, only a few hours.
• Larger volumes, compared with colloids, are required to achieve the same physiologic endpoints. Typically, crystalloid at 2–4 times the volume of colloid or
Hetastarch is required for a comparable physiologic effect during resuscitation6.
• Resuscitation with crystalloids has been shown to produce a rapid and sustained drop in plasma COP that is significantly more severe than the postresuscitative drop
in COP seen with colloid or HS7.
7.6.2 Hypertonic solutions—risks and benefits

Hypertonic solutions (HS) facilitate the redistribution of fluid between the intracellular and extracellular compartments. HS infusions tend to rapidly ‘recruit’ intracellular
water into the extracellular or intravascular space, maintaining relatively consistent serum sodium levels while simultaneously volume expanding the intravascular space.
Smaller volumes of HS are therefore effective in meeting resuscitative goals when compared with isotonic fluid options. The most commonly used hypertonic solutions
are hypertonic saline solution (HSS) and hypertonic albuminated fluid (HALF). Hypertonic solutions are well studied in burn resuscitation8,9 and their use in trauma
resuscitation and following neurotrauma or neurosurgery shows promise in terms of decreasing postresuscitative morbidity10.
• Hypertonic saline (HS) is a 3% sodium chloride solution that contains 513 mequiv l−1 Na, 513 mequiv l−1 Cl buffered to a pH of 5.0 using hydrochloric acid. It has
an osmolarity of 1027 mOsm l−1
• Hypertonic albuminated fluid (HALF)11 contains 240 mequiv l−1 Na, 120 mequiv
Page 107
l−1 Cl, 120 mequiv l−1 lactate, plus 1.5 g l−1 albumin.
Advantages of hypertonic solutions

• Effective immediate volume expansion in acute resuscitation.
• Lower volumes of resuscitation required to reach physiologic endpoints than with isotonic fluids11.
• Less cellular edema postresuscitation, therefore less risk to injured nerve tissue in brain and spinal cord10.
Disadvantages of hypertonic solutions

• Risk of severe hypernatremia, especially with HSS.
• Risk of hyperchloremic acidosis.
• Risk of cellular dehydration and secondary injury.
• More expensive than isotonic fluids.
• Requires central venous access for administration, because of intolerance of peripheral veins to the chemical stress of hypertonic solutions.
7.7 Cellfree colloids

Colloid solutions contain large molecules that are relatively impermeable to capillary membranes and therefore have a significant intravascular ‘availability’ post
administration.
7.7.1 Albumin
Albumin (5% or 25%) solutions

These solutions are prepared from heattreated donor plasma and have less than 0.5% risk of adverse reaction (i.e. urticaria, fever, chills). In normal plasma, albumin
provides about 80% of the plasma COP. Approximately 40% of the body’s albumin pool is intravascular, and 60% is extravascular in tissue stores (about onethird of
the extravascular tissue stores is in the skin)12. However, in situations of physiologic stress and critical illness, endogenous albumin levels drop precipitously, owing to
capillary leak and/or increased breakdown in the catabolic state13.
The 5% albumin solution contains 50 g l−1 albumin mixed in NSS and has a COP of about 20 mmHg. The 25% albumin solution contains 240 g l−1 albumin in
buffered diluent containing 130–160 mequiv l−1 Na and has a COP of about 100 mmHg.
Advantages of albumin solutions

• Lower volumes of resuscitation than isotonic solution are needed to reach physiologic endpoints14.
• Physiologic support of COP14.
• Intravascular effects can last up to 24 h14 depending on patient’s preperfusion status15.
Disadvantage of albumin solutions

• More expensive than crystalloids.
7.7.2 Hetastarch
Hetastarch is a synthetic colloid solution composed of hydroxyethylsubstituted branchedchain amylopectin molecules (each with an average molecular weight of
6.9×104, similar to that of an albumin molecule). The 6% solution of Hetastarch in NSS contains 60 g l−1 colloid, 154 mequiv l−1 Na, and 154 mequiv l−1 Cl. It has an
osmolarity of approximately 310 mOsm l−1 and a COP of approximately 30 mmHg. Hetastarch molecules are either degraded in the liver or excreted in the stool and
urine.
Advantages of Hetastarch
• Physiologic support of COP.
• Intravascular effects can last 24–36 h.
• Nontoxic and nonantigenic.
• Useful in Jehovah’s Witnesses, who will not accept albumin.
Disadvantages of Hetastarch
• May impair coagulation when given in high volumes (i.e. >1.5 l per resuscitation episode or 20 ml kg−1 day−1).
Page 108
• More expensive than crystalloids.

• Risk of hyperchloremic acidosis when administered in large volumes, as a result of the solution containing normal saline4.
7.7.3 Hetastarch with electrolytes (Hextend)

A 6% Hetastarch, formulated in a physiologically balanced crystalloid carrier similar to lactated Ringer’s solution, is marketed under the brand name Hextend. The
crystalloid carrier more closely resembles the ionic composition of normal plasma. It contains physiologic levels of sodium (143 mequiv l−1), chloride (124 mequiv 1−1)
and calcium (5 mequiv l−1), and slightly lower than normal physiologic levels of potassium (3 mequiv l−1) and magnesium (0.9 mequiv l−1). Hextend also contains a
subphysiologic level of dextrose (0.99 g 1−1) and 28 mequiv l−1 lactate, which provides a source of bicarbonate in patients with normal lactate metabolism. In contrast
to 6% Hetastarch with saline, the Hetastarch in electrolyte solution is associated with less intraoperative calcium administration, and lower mean estimated blood loss16.
The additional buffering capability may also reduce or eliminate the prolongation of thromboelastogram clot time seen with large volumes of Hetastarchsaline
administration16.
7.7.4 Dextrans
The dextrans are solutions of high molecular weight polysaccharides that are manufactured from products of bacteriamediated, enzymatic digestion of sucrose
containing media. Two dextran products are available for use in the United States.
Dextran 70 is a 6% solution of dextran in sodium chloride. It has a weightaverage molecular weight (MW) of 7×104, is hypertonic when compared with blood, and
retains approximately 20–25 ml of water per gram infused. Because of its high MW, intravascular retention is excellent and generally lasts longer than 24 h before
degradation in the viscera (to carbon dioxide and water) or excretion into the urine or gut.
Dextran 40 is a 10% solution of dextran in sodium chloride. It has a weightaverage MW of 4×104, is hypertonic and has an equivalent COP in vitro to 17%
albumin solution. However, because of its relatively low MW, dextran 40 molecules easily migrate across capillary membranes and are rapidly cleared via the kidneys.
Sixty per cent of infused dextran 40 can be excreted in the urine within 6 h (in patients with normal renal function) and nearly 70% is excreted within 24 h following
infusion17. Therefore, intravascular availability of dextran 40 is poor.
Advantages of dextrans
• Physiologic support of COP.
Disadvantages of dextrans
• Possibility of serious anaphylactoid reactions (incidence of about 5.3%).
• Rheologic effects can predispose to bleeding and coagulopathy (dextrans at large doses can coat platelets).
• Dextran 40 can contribute to acute renal failure in hypovolemic patients because the high rate of renal clearance in low flow states can result in increased tubular
viscosity and obstruction17.
7.8 Summary and conclusions

• Colloids and hypertonic solutions are more expensive, but also more efficient than crystalloids in correction of the absolute and relative volume deficit demonstrated
early in shock states.
• Colloids and hypertonic solutions are less edemagenic than crystalloids when used early in resuscitation. This effect may minimize risks of postresuscitative morbidity
including wound healing and skin complications, noncardiogenic pulmonary edema, and severely low plasma COP.
Page 109
• Risks of colloids and hypertonic solutions are product specific—hemostasis is a concern with large volumes of Hetastarch, and the dextrans have a significant
incidence of anaphylactoid reactions.
• Multivariate analyses have failed to show significant differences in efficacy of one colloid over another in the acute resuscitative setting.
• Isotonic crystalloids are the most costeffective acute resuscitative fluid. However, in early shock, resuscitation with blood or colloids, although more expensive, helps
preserve plasma COP and may be needed to maintain oxygencarrying capacity.
7.9 Recommendations for fluid resuscitation in the ICU

• Evaluate the ‘ABCs’ and establish adequate IV access early.
• ALWAYS use only colloid, blood or isotonic fluids (NOT hypotonic fluids) for acute resuscitation from hypovolemic states.
• Resuscitate with colloids and/or blood whenever possible instead of isotonic crystalloids (particularly early in shock). This strategy acts to preserve plasma COP,
maintain intravascular volume for longer periods, support oxygencarrying capacity and minimize postresuscitative edema.
• Remember to focus on resuscitative endpoints from the initiation of therapy, to minimize subsequent morbidity and mortality.
• Recognize that hypertonic solutions appear to be a good option for efficient shortterm intravascular volume expansion but need to be administered via central venous
lines.
• Monitor for infusionrelated metabolic complications (i.e. hyperchloremic acidosis with salinecontaining solutions, hypocalcemia with citrated blood products).
• Individualize therapy to individual patient physiology when possible (i.e. bedside Starling curve).
References
1. Starling, E.H. (1896) On the absorption of fluids from the connective tissue spaces. J. Physiology 9:312.
2. Civetta, J.M. (1979) A new look at the ‘Starling’ Equation. Crit. Care Med. 7:84.
3. Lucas, C.E. (1977) Resuscitation of the injured patient: the three phases of treatment. Surg. Clin. N. Am. 57:3.
4. Tullis, J.L. (1977) Albumin. JAMA 237:355.
5. Rackow, E.C., Fein, I.A. and Leppo, J. (1977) Colloid oncotic pressure as a prognostic indicator of pulmonary edema and mortality in the critically ill patient. Chest
72:709.
6. Rackow, E.C., Falk, J.L. and Fein, I.A. (1983) Fluid resuscitation in circulatory shock: a comparison of the cardiorespiratory effects of albumin, Hetastarch, and
saline solutions in patients with hypovolemic and septic shock. Crit. Care Med. 11:839.
7. Rackow, E.C., Fein, I.A. and Siegel, J. (1982) The relationship of colloid osmotic pulmonary artery wedge pressure gradient to pulmonary edema and mortality in
critically ill patients. Chest 82:433.
8. Lowe, R.J., Moss, G.S., Jilek, J., et al. (1977) Crystalloids vs. colloids in the etiology of pulmonary failure after trauma. A randomized trial in man. Surgery 81:676.
9. Weaver, D.W., Ledgerwood, A.M., Lucas, C.E., et al. (1978) Pulmonary effects of albumin resuscitation for severe hypovolemic shock. Surgery 1978; 113:387.
10. Prough, D.S. (2000) Acidosis associated with perioperative saline administration. Dilution or delusion? Anesthesiology 93:1167–1169.
11. Waters, J.H. and Bernstein, C.A. (2000) Dilutional acidosis following Hetastarch or albumin in healthy volunteers. Anesthesiology 93:1184–1187.
12. Dil. Acidosis. Surg. Gynecol Obstet.
13. Jarvela, K., Honkonen, S.E., Javela, T., Koobi, T. and Kaukinen, S. (2000) The comparison of hypertonic saline (7.5%) and normal saline (0.9%) for initial fluid
administration before spinal anesthesia. Anesth. Analg. 91(6): 1461–1465.
14. Jelenko, C., Wheeler, M.L., Callaway, B.D., Divilio, L.T., Bucklen, K.R. and Holdrege, T.D. (1978) Shock and resuscitation II: Volume repletion with
minimal edema using the ‘HALFD’ regimen. JACEP 7 (9): 326–333.
15. Jelenko, C. (1979) Fluid therapy and the HALFD method. J. Trauma 19(Suppl.):11.
Page 110
16. Gan, T.J., et al. (1999) Hextend, a physiologically balanced plasma expander for large volume use in major surgery: a randomized phase III clinical trial. Anesth.
Analg. 88(5): 992–998.
17. Sumas, M.E., Legos, J.J., Nathan, D., Lamperti, A.A., Tuma, R.F. and Young, W.F. (2001) Tonicity of resuscitative fluids influence outcome after spinal cord
injury. Neurosurgery 48(1):167–172; discussion 172–173.
Page 111
Chapter 8
Circulatory support
Thomas L.Higgins, MD
Contents
8.1 Introduction
8.2 The sympathetic nervous system
8.3 Receptors and signal transduction
8.4 Monitoring
8.5 Catecholamines and sympathomimetics
8.6 Alterations in adrenergic response
8.7 Bypassing the adrenergic receptor
8.8 Vasopressin
8.9 Weaning pharmacologic support
8.10 Summary
Page 112
8.1 Introduction
Choosing the optimal pressor or inotrope in the hemodynamically unstable patient requires us to understand a number of issues, including which drugs affect specific
adrenergic receptors and whether a drug is a full or partial agonist (Table 8.1). Catecholamines and sympathomimetics (drugs that mimic the effects of endogenous
catecholamines) have traditionally been first line pharmacotherapy for shock states, but noncatecholamine agents such as milrinone may be particularly useful when
catecholamine response is altered1. Catecholamines are not the only endogenous substances responsible for blood pressure homeostasis. Vasopressin may have a role
in treating septic shock and vasodilation following cardiopulmonary bypass.
8.2 The sympathetic nervous system

Homeostatic control of blood pressure is mediated via the sympathetic nervous system. Sympathetic nerves originate in the spinal cord. Preganglionic fibers generally
exit to synapse with postganglionic neurons in the lateral or paravertebral ganglia. The preganglionic fibers are cholenergic and secrete acetylcholine at their nerve
endings. Postganglionic fibers, which terminate in vessels and visceral organs, primarily are adrenergic and secrete norepinephrine. Tyrosine is the basic building block
for catecholamines (Figure 8.1). Through a series of reactions, tyrosine is hydrolyzed to dihydroxyphenylethylamine (DOPA) by the enzyme tyrosine hydroxylase and
then decarboxylated to dopamine, which is transported and stored at nerve endings2. Dopamine is subsequently converted to norepinephrine and stored in vesicles in
adrenergic nerves (see Figure 9.4). Sympathetic nerve stimulation causes norepinephrine vesicles to fuse to the membrane and be released into the synaptic cleft.
Norepinephrine then crosses the synaptic cleft and binds at the alpha1 receptor, activating a process that culminates in increased intracellular calcium and ultimately
endorgan effects such as vasoconstriction. Simultaneously, some norepinephrine also binds at presynaptic alpha2 receptors, inhibiting further norepinephrine release
and ongoing sympathetic stimulation. Adrenergic receptor physiology provides insight into the employment of exogenous agonists in promoting vasoconstriction and use
of alpha2 agonists to lower blood pressure. Chapter 9 has additional details on alphaadrenergic agonists and antagonists used for blood pressure control.
Table 8.1 Selectivity of catecholamines for adrenergic receptors at usual doses
Agent Alpha1 Alpha2 Beta1 Beta2 Dopamine1 Dopamine2

Phenylephrine ++ 0 0 0 0 0
Norepinephrine +++ ++ +++ + 0 0
Epinephrine +++ +++ +++ ++ 0 0
Isoproterenol 0 0 +++ +++ 0 0
Dopamine 0 to +++ 0 to + 0 to +++ 0 to ++ +++ ++
Dobutamine 0 to + 0 +++ ++ 0 0
Fenaldopam 0 0 0 0 +++ 0
Dopexamine 0 0 + ++ +++ ++
+ to +++ indicate relative amounts of stimulation; these are not necessarily to scale between drugs. Some agents, particularly dopamine, have dosedependent effects with alpha increasing
at higher doses. 0, no effect.
Page 113
Figure 8.1 Synthetic pathway of catecholamines. Tyrosine from dietary sources is eventually synthesized into the endogenous
catecholamines dopamine, norepinephrine and epinephrine, Substrate depletion impairs the body’s ability to generate an
appropriate catecholamine response. From; Chernow, B., Rainey, T.G. and Lake, C.R. (1982) Endogenous and exogenous
catecholamines in critical care medicine. Crit. Care Med. 10:409–416. Reproduced with permission from Lippincott, Williams
& Wilkins.
8.2.1
Adrenergic receptors
Properties of adrenergic receptors can be determined by radiolabeled binding, the effect of agonists and antagonists, and at a structural level. Although much more
detail can be demonstrated experimentally, Table 8.2 gives a clinically relevant summary for understanding adrenergic receptors. The alpha1 adrenergic receptor
controls blood pressure via vasoconstriction of arteries and arterioles. Alphaadrenergic receptors also exist in the heart and are implicated in the genesis of cardiac
arrhythmias; alphaadrenergic receptor blockade has been shown to increase the threshold for epinephrineinduced arrhythmias3. The alpha2 receptors on the pre
synaptic and postsynaptic membranes inhibit norepinephrine release and promote vasoconstriction, respectively. Beta1 receptors mediate increases in heart rate,
contractility, and conduction velocity. Beta2 receptors mediate inotropy in the heart, and promote vasodilation in the lungs and peripheral vasculature. The dopamine 1
(DA1) receptor produces both general and selective (renal and splanchnic) vasodilation, as well as specific effects on the renal proximal tubule and cortical collecting
duct. DA2 receptors, located on presynaptic
Table 8.2 Properties of adrenergic receptors
Receptor Primary Location Action

alpha1 Postsynaptic Vasoconstriction
alpha2 Presynaptic Inhibits NE release

Postsynaptic Vaso and venoconstriction
beta1 Heart Chronotropy and inotropy
beta2 Lungs, vessels Broncho and vasodilation

Myocardium Inotropy
DA1 Postsynaptic smooth muscle Peripheral vasodilation
Proximal tubule, cortical collecting duct Increased RBF, natriuresis, diuresis
DA2 Presynaptic nerve terminals Vasodilation only if baseline
Adrenal cortex norepinephrine levels elevated
Glomerulus Decreased renal blood flow (RBF), glomerular filtration rate (GFR) and water excretion
Page 114
nerve terminals, in the glomerulus and in the adrenal cortex, mediate peripheral vasodilation only if baseline norepinephrine levels are high at baseline. In contrast to DA
1 activation, DA2 stimulation decreases renal blood flow, glomerular filtration rate and sodium and water excretion4. A patient’s response to a given sympathomimetic
agent depends on the dosage given as well as patient factors, so responses are not always predictable. In general, however, agonists produce relatively predictable
results at relevant doses in most patients. Phenylephrine, for example, provides only alphaadrenergic stimulation at usual doses, whereas isoproterenol provides only
beta stimulation. Norepinephrine, epinephrine, dopamine and dobutamine have combined effects, which vary by infusion rate (Table 8.1). Fenoldopam, a relatively new
agent, is highly selective for the DA1 receptor.
8.3 Receptors and signal transduction

With the notable exceptions of steroids and thyroid stimulating hormone, most drugs and endogenous hormones are hydrophilic and cannot easily cross cell
membranes. A drug or a hormone must bind to a receptor on an extracellular surface. Signal transduction then generates a response from the cell by activating or
inhibiting a secondary messenger.
Binding of a ligand to the extracellular surface of a receptor causes a change in membrane proteins resulting in an intracellular change (Figure 8.2). For example,
binding of an alpha1 agonist such as phenylephrine causes a change in the Gq protein activating phospholipase C to increase the intracellular concentrations of inositol
1,4,5triphosphate (IP3) and diacetylglycerol (DAG). Alpha2 agonists cause activation of an inhibitory G protein, inhibiting adenylyl cyclase and decreasing cyclic
AMP concentrations within the cell. Conversely, betaadrenergic agonists such as isoproterenol act through the Gs protein to stimulate adenylyl cyclase and increase
cyclic AMP within the cell. Increased levels of cyclic AMP within the cell ultimately cause acceleration of heart rate or augmented cardiac contraction. Figure 8.3
describes cyclic AMP metabolism by the enzyme phosphodiesterase F3. Two mechanisms thus exist for raising cyclic AMP levels within cells: stimulating the beta
adrenergic
Figure 8.2 Signal transduction. Drugs or hormones (ligands) bind to extracellular adrenergic receptors (AR); an intermediary step of
signal transduction, mediated by G proteins, is necessary to produce an action within the cell, IP3 (inositol 1,4,5triphosphate)
mobilizes calcium from intracellular stores, DAG (diacylglycerol) increases affinity of protein kinase C for intracellular calcium.
cAMP (3',5'cyclic adenosine monophosphate) is increased by the action of the Gstimulatory (Gs) protein, and decreased by
the Ginhibitory (Gi) protein.
Page 115
Figure 8.3 Generation of cyclic AMP. Intracellular increases in 3',5'cyclic adenosine monophosphate (cAMP) result in increased
inotropy and chronotropy. cAMP levels are increased through binding of adrenergic ligands and subsequent signal transduction,
cAMP is broken down by phosphodiesterase F3, which is inhibited by the phosphodiesterase inhibitors amrinone and milrinone.
This explains why phosphodiesterase inhibitors are effective in the presence of betaadrenergic blockade, and also accounts for
observed synergism in the clinical setting.
receptor and preventing the breakdown of cyclic AMP with a phosphodiesterase inhibitor such as milrinone. Inhibition of the phosphodiesterase pathway is particularly
useful when receptor–effector coupling has been altered by sepsis1 or other clinical conditions, or when access to the receptor is competitively blocked by agents (such
as betaadrenergic blockers) that prevent normal ligandreceptor binding.
8.4 Monitoring
Treatment with vasopressors or inotropes is justified when there is concern about the need to improve organ perfusion. Because our ability to monitor tissue perfusion is
limited, we rely on nonspecific findings such as oliguria, diminished arterial blood pressure and lactate generation. Depending on clinical circumstances, bedside
examination, central venous pressure or PAC monitoring may help to establish the patient’s volume status; echocardiography can demonstrate ventricular contractility
and degree of ventricular filling. It is generally inappropriate to raise mean arterial pressure using pressor agents in a hypovolemic patient, unless such therapy is
temporary while fluid resuscitation is under way. In a patient with hypoperfusion, the first question should be ‘Is the volume status adequate?’ (Figure 8.4).
Hypovolemia should prompt fluid resuscitation, normally with crystalloid, but with blood if the hemoglobin level is already low (<7 g %) or is expected to drop further
with aggressive volume repletion. Colloid solutions are rarely indicated for acute hypovolemia. Chapter 7 contains a more complete discussion of fluid choices in
resuscitation.
Once volume status has been repleted, cardiac output should be reassessed. Inadequate cardiac output commonly reflects depressed ventricular function, resulting
from ischemia, infarction, or sepsisrelated myocardial depression. Other causes of acute decreases in cardiac output, such as cardiac tamponade, hypertrophic
cardiomyopathy or pulmonary embolism, should be considered in the differential diagnosis. Inotropic support is contraindicated in the presence of hypertrophic
cardiomyopathy, as increasing contractility with inotropes may worsen the obstruction. New presentation hypertrophic cardiomyopathy, which can be difficult to
Page 116
Figure 8.4 Algorithm for resuscitation. The patient’s volume status must always be considered first, and then decisions made on the
addition of inotropic or vasoconstrictive agents.
diagnose, requires a different therapeutic approach, including slowing the heart rate with a betaadrenergic blocker, raising afterload, and providing adequate volume
repletion. Myocardial failure secondary to most other causes will generally require inotropic support.
The final question to ask from a monitoring standpoint is whether or not vascular tone is adequate. A pulmonary artery catheter will allow calculation of systemic
vascular resistance (SVR), which can guide titration of vasopressor support. SVR, however, is a calculated variable dependent on the vascular pressure gradient (MAP
minus CVP) and the cardiac output. Under conditions of low cardiac output and low perfusion pressure, systemic vascular resistance might well be normal or low,
prompting efforts to increase afterload, when in fact the underlying problem is inadequate cardiac output. Vasodilation may be apparent by clinical examination,
although the classic picture of ‘warm shock’ can be modified when vascular disease is also present. Vasodilation responds to pressors, both alphaadrenergic agonists
such as norepinephrine and phenylephrine, and nonadrenergic agents such as vasopressin, discussed in section 8.8.
Inotropes and pressors are preferably administered through central lines because of the risk of IV infiltration, tissue necrosis and skin sloughing. If it is necessary to
give an agent peripherally, the site should be carefully monitored. Phentolamine (5 mg in 10 cm3 normal saline) can be injected into areas of extravasation to limit tissue
damage should infiltration occur with vasopressors.
8.5 Catecholamines and sympathomimetics
8.5.1 Phenylephrine (neosynephrine)

This is an alpha1 adrenergic agonist at conventional clinical doses, but activates the alpha2 and betaadrenergic receptors at 10fold higher dose. Bolus doses of
50–100 µg IV increase systemic vascular resistance and typically improve perfusion pressure transiently. Infusions can be initiated at 0.5–1 µg kg−1 min−1 and titrated
to clinical effect. Phenylephrine is usually mixed as 10–20 mg in 250 ml of D5W, but more concentrated mixtures may be used. Increased systemic vascular resistance
from phenylephrine may cause reflex bradycardia, and this action of phenylephrine was once used to terminate supraventricular arrhythmias before the availability of
Page 117
more specific antiarrhythmics. Phenylephrine vasoconstricts renal, cutaneous and splanchnic vessels, but usually increases coronary perfusion as the result of improved
diastolic perfusion pressure. However, phenylephrine is capable of constricting coronary vessels. Because phenylephrine has mostly alpha1 effects, it is a useful agent
in patients who develop tachyarrhythmias with norepinephrine or dopamine. Phenylephrine is compatible with a wide range of solutions and drugs, but incompatible
with amphotericin B, azathiapine, gancyclovir, human insulin, pentamidine, and propofol. The bottom line is that phenylephrine is not as potent as norepinephrine or
epinephrine, but can be extremely useful in the management of septic shock because it can selectively reverse vasodilating hypertension without causing undue cardiac
stimulation.
8.5.2 Norepinephrine
Norepinephrine (Levophed) is primarily an alpha agonist, but stimulates beta receptors at doses as low as 2 µg min−1. Norepinephrine is indicated for septic shock or
severe vasodilation, once volume status is repleted. The standard mix for norepinephrine is 8 mg in 250 ml of D5W, but it is not uncommon to concentrate it further in
patients requiring very high doses (see Table 8.3). The typical starting dose of norepinephrine is 0.1 µg kg−1 min−1, increasing by 0.1 µg kg−1 min−1 every 5–10 min
based on hemodynamic response. There is no upper limit to the dose of norepinephrine that can be used for short periods of time, but patients requiring very high doses
(>2 µg kg−1 min−1) for prolonged periods often have a refractory condition with poor likelihood of survival. Norepinephrine administration is associated with a poor
outcome in patients with high degrees of organ failure or a delay of more than 1 day from ICU admission to norepinephrine therapy5. Norepinephrine infusion must be
weaned slowly over 6–24 h to avoid abrupt hypotension. Norepinephrine is compatible with most antiarrhythmics, other pressors and inotropes, heparin, and sedative
agents such as midazolam and morphine sulfate. It is incompatible with aminophylline, amphotericin B, azathiaprine, furosemide, gancyclovir, insulin, lidocaine,
barbiturates and sodium bicarbonate.
Norepinephrine’s renal vasoconstricting properties have raised concern about its use in oliguric patients, yet the balance of evidence suggests that improvement in
renal and cardiac perfusion with norepinephrine may outweigh the renal vasoconstrictive effects. Desjars et al.6 evaluated norepinephrine at doses of 0.5–1 µg kg−1
min−1 in hyperdynamic septic patients who remained hypotensive despite volume expansion and dopamine therapy. Infusion of norepinephrine increased mean arterial
pressure and systemic vascular resistance, decreased heart rate, and resulted in either unchanged or increased cardiac output in most patients. Urine flow increased to
more than 0.5 ml min−1 when a critical renal perfusion pressure was reached as long as there was no preexisting renal damage. In a subsequent study7 the same authors
showed that norepinephrine in doses up to 1.5 µg kg−1 min−1 does not produce deleterious effects on the kidneys as measured by serial analysis of creatinine, osmolar
and free water clearance.
8.5.3 Dopamine
Dopamine is the immediate precursor in the metabolic pathway leading to the production of norepinephrine. Dopamine produces selective vasodilation in the renal and
splanchnic circulations, but does not cross the bloodbrain barrier. Its hemodynamic effects are variable, depending on dose utilized. Dopaminergic receptors are
considered to be activated at doses less than 5 µg kg−1 min−1 but perhaps more accurately the ‘dopaminergic’ range should be considered to encompass 0.3–5 µg
kg−1 min−1, overlapping into the betaadrenergic range. The betaadrenergic range is classically described as 5–10 µg kg−1 min−1, but tachycardia and arrhythmias may
occur with lower doses, and an inotropic effect can be seen with as little as 2 µg kg−1
Page 118
Table 8.3 Typical starting and highend doses
Agent and preparation Typical starting doses (80 kg patient) Highend doses (80 kg patient)*
Drug Typical mixture Concentration µg kg–1 min−1 µg min–1 cm3 h–1 µg kg–1 min–1 µg min−1 cm3 h–1
Norepinephrine 8 mg/250 ml D5W 32 µg cm –3 0.1 8 15 1.6 128 240
Epinephrine 4 mg/250 ml D5W 10 µg cm –3 0.05 4 15 1 80 300
Phenylephrine 20 mg/250 ml D5W 80 µg cm–3 0.5 40 30 5 400 300
Dopamine (renal) 400 mg/250 ml D5W 1600 µg cm–3 0.5 40 1.5 5 400 15
Dopamine (beta) 800 mg/250 ml D5W 3200 µg cm –3 3 240 4.5 10 800 15
Dopamine (alpha) 800 mg/250 ml D5W 3200 µg cm –3 10 800 15 40 3200 60
Dobutamine 500 mg/250 ml D5W 2000 µg cm–3 5 400 12 40 3200 96
Isoproterenol 2 mg/500 ml D5W 4 µg cm–3 0.025 2 30 0.25 20 300
Fenoldopam 10 mg/250 ml NS or D5W 40 µg ml –1 0.025 2 3 0.3 24 40
Milrinone 50 mg/250 ml NS or D5W 200 µg ml –1 0.375 30 9 0.750 60 18
*Drugs should generally be concentrated when used at these doses to reduce fluid load.
Page 119
min−1. Vasoconstrictive effects as a result of alpha receptor activation generally occur at doses greater than 10 µg kg−1 min−1, but in septic patients much higher doses
may in fact be necessary to achieve vasoconstriction. Clinically, some patients may demonstrate blood pressure changes even at very low doses of dopamine.
Dopamine increases renal sodium excretion independently of its effects on renal blood flow, probably through inhibition of tubular solute reabsorption. Schaer et al.8
randomized animals to receive dopamine 4 µg kg−1 min−1 or placebo while at the same time receiving escalating doses of norepinephrine. Renal blood flow increased
gradually with escalating doses of norepinephrine alone, reaching a 10% increase over baseline at the highest dose. However, when dopamine was combined with
norepinephrine, renal blood flow increased approximately 40%. Renal vascular resistance increased with norepinephrine alone, but decreased with the addition of
dopamine. Dopamine has been associated with a functional improvement in acute oliguria post cardiopulmonary bypass9, decreased renal impairment with orthotopic
liver transplantation10 and a dosedependent increase in urine output in oliguric uvolemic surgical intensive care patients11. Despite these observations, wellconducted
outcome studies showing dopamine’s advantage as a renal protective agent are still lacking, and no survival advantage has been documented12. Besides producing
alpha and beta1adrenergic activity, dopamine stimulates dopaminergic receptors on renal and splanchnic vessels. At high doses, it can precipitate myocardial
ischemia and arrhythmias. Dopamine can be used independently to manage septic shock when the goal is to attain significant cardiac stimulation and moderate
vasoconstriction. There are no data to confirm that lowdose dopamine clinically improves renal or splanchnic blood flow independent of other hemodynamic effects.
The use of lowdose dopamine does not appear to improve survival or obviate the need for dialysis.12
8.5.4 Fenoldopam
Fenoldopam is a selective dopamine1 (DA1) agonist with no alpha or betaadrenergic effects. An increase in heart rate may be seen as a reflex response to
dopaminergic mediated vasodilation and decreased blood pressure. Administration may cause dilation of the renal afferent and efferent arterioles and an increase in
glomerular filtration rate, and an increased excretion of water and electrolytes. The drug is currently marketed as an alternative to IV. nitroprusside for acute blood
pressure control. A starting dose of 0.1 µg kg−1 min−1 can be increased in 0.1 µg kg−1 min−1 increments every 20 min to a maximum dose of 1.7 µg kg−1 min−1.
Fenoldopam increases sodium excretion in both normal and hypertensive individuals by increasing renal blood flow, inhibiting tubular reaborption of sodium, and
possibly increasing glomerular filtration rate. Activation of the DA1 receptor by either dopamine or fenoldopam can increase intraoccular pressure13; a relative
contraindication in patients with glaucoma. Outcome studies regarding renal protective effects are under way. In normotensive volunteers given doses ranging from 0.03
to 0.3 µg kg−1 min−1, fenoldopam increased renal plasma flow (RPF) in a dosedependent manner. At the lowest dose, this increase in RPF occurred in the absence of
changes in mean arterial pressure or heart rate14,15. Fenoldopam has also been shown to reverse some of the adverse renal effects of PEEP16.
Stimulation of the DA1 (but not DA2) receptor reduces gastric acid secretion and may reduce gastric ulcer formation. In an animal study, administration of
fenoldopam decreased the incidence of gastric lesions, ulcer area, and mortality rate17.
Fenoldopam is conjugated by liver enzymes before renal excretion, but is not dependent on the cytochrome P450 system. There are no toxic metabolites or known
metabolic drug interactions, and no dosage adjustments are needed for renal or liver disease. Although not officially sanctioned for renal protective use, data suggest
that doses of fenoldopam as low
Page 120
as 0.03–0.05 µg kg−1 min−1 (lower than the dose recommended for blood pressure reduction) may be effective.
8.5.5 Epinephrine
Epinephrine is an extremely potent alpha and beta1 agonist with considerable beta2 activity useful for treatment of bronchospasm, anaphylactic shock, septic or
cardiogenic shock, asystole, or refractory hypotension. Reflex vagal effects are less intensive than with predominantly alpha agents. Epinephrine is a potent inotrope and
increases stroke volume, coronary blood flow and heart rate, but may also be arrhythmogenic. Treatment of asystole requires the largest dose (1 mg or 10 cm3 of a
1/10,000 solution) given by rapid IV infusion and repeated as necessary. Highdose bolus epinephrine (0.2 mg kg−1) has been promoted as treatment for refractory
asystole, but is not currently recommended by Advanced Cardiac Life Support (ACLS) guidelines. An advantage of epinephrine in patients lacking vascular access is
that the drug can be given via the endotracheal route using approximately twice the i.v. dose diluted in normal saline or distilled water.
More conventionally, continuous infusions of epinephrine are used for treatment of refractory hypotension and septic or cardiogenic shock. Initial IV infusion rates for
septic or cardiogenic shock begin at 1 µg min−1 (0.01 µg kg−1 min−1), but refractory hypotension can require a starting dose of 0.05 µg kg−1 min−1 (4 µg min−1) and
may go 10–20fold higher (see Table 8.3). Selective activation of the beta receptor occurs at lower doses, but receptor selectively is lost and then reversed in favor of
alpha receptors at high doses. The danger of excessive infusion rates or rapid infusion is a precipitous rise in blood pressure, which could theoretically result in cerebral
hemorrhage. Epinephrine is incompatible with alkaline solutions and becomes unstable if the pH is higher than 5.5. Like norepinephrine, epinephrine is compatible with
most antiarrhythmics (except lidocaine), and most sedatives and neuromuscular blocking agents (exceptions being barbiturates and propofol). Epinephrine is
incompatible with aminophyline, amphotericin B, some antibiotics and sodium bicarbonate. Like other catecholamines, epinephrine has a short halflife, with metabolism
by catecholamineomethyl trans ferase (COMT) and monoamine oxidase (MAO). Caution should be exercised when administering catecholamines to patients who
have previously received MAO inhibitors.
8.5.6 Dobutamine
Dobutamine (Dobutrex) possesses only small amounts of alphareceptor activity, and produces mostly betareceptor stimulation of the heart. Compared with
dopamine, dobutamine increases cardiac output with less tachycardia and smaller increases in arterial and pulmonary pressures. It may be better tolerated than
dopamine in the setting of heart failure, because of its pulmonary vasodilating effects. The expected hemodynamic responses of dobutamine can be altered by
preoperative betaadrenergic blockade18. Sepsis may also alter the expected response, if alphaadrenergic (vasoconstrictive) responses do not balance the beta2
vasodilatory response. Usual hemodynamic actions of dobutamine at infusion rates of 2–10 µg kg−1 min−1 include an increase in cardiac output and stroke volume, little
or no change in heart rate, and a decrease in pulmonary capillary wedge pressure and systemic vascular resistance. Dobutamine may produce substantial systemic and
pulmonary vasodilation in the clinical setting, an effect that can be used to advantage in patients with pulmonary hypertension or right ventricular failure. Dobutamine is
the preferred agent in the setting of hypoxemic respiratory failure and normotensive left ventricular failure, especially postmyocardial infarction, because additional
vasoconstriction is not necessary in these situations and may actually worsen right ventricular function. Synergistic effects have been noted when combining dobutamine
and phosphodiesterase inhibitors19. Dobutamine may also be given concurrently with norepinephrine if the expected decrease in SVR results in unacceptable
hypotension.
Page 121
Combination therapy with an αagonist significantly improves mean pressure, cardiac index and left ventricular strokework index in dobutamineresistant septic
shock20.
8.5.7 Isoproteronol
Isoproterenol (Isuprel) acts almost exclusively on the betaadrenergic system. By increasing cutaneous and muscular vasodilation, the drug may redistribute blood to
nonessential areas, resulting in significant hypotension. Isoproterenol was once utilized in therapy of cardiac failure because of its effects on inotropic activity and its
potential for increasing cardiac output by augmenting preload and reducing afterload. With the availability of phosphodiesterase inhibitors, isoproteronol is now rarely
used other than for management of bradyarrhythmias. Isoproteronol is generally mixed 1–4 mg in 250 ml of D5W or normal saline and titrated in the range of 0.05–0.5
µg kg−1 min−1.
8.5.8 Dopexamine
Dopexamine is principally a dopaminergic and beta2 agonist21; it is not yet available in the United States. Hemodynamic effects at doses of 1–6 µg kg−1 min−1 include
decreased mean arterial pressure and systemic vascular resistance; increased heart rate by reflex action; improved organ blood flow to cardiac, hepatic, splanchnic, and
renal beds; and increased inotropy via beta2 cardiac stimulation22. Dopexamine at lower doses (0.5–2.0 µg kg−1 min−1) may have utility in promoting gut mucosal
blood flow in highrisk patients23, via direct splanchnic vasodilation.
8.6 Alterations in adrenergic response

A variety of factors influence the expected response of adrenergic receptors to endogenous and exogenous catecholamines (Table 8.4). Mild acidosis increases
catecholamine synthesis and secretion, but severe acidosis (pH<7.10) is associated with decreased cardiac contractility, diminished cardiac output, and a
Table 8.4 Alterations in catecholamine function
Acidbase status Downregulation (stimulation)

Age Electrolyte disorder
Alphablockers Hemodilution
Anesthetics or opioids Hypothermia
Betablockers Individual variation
Calcium channel blockers Sepsis
Congestive heart failure Substrate depletion
Denervation Surgical stress
Diabetes mellitus Upregulation (blockade)
systemic vasoconstrictive response. Acidosis also may produce difficulty with defibrillation24. Alkalosis decreases catecholamine secretion, inhibits sympathetic nervous
system activity, and results in decreased vasoconstrictive ability, accentuated vasodilation, and a propensity to ventricular dysrhythmias24. In untreated shock states,
acidosis is more common, but alkalosis can result from overaggressive buffer therapy. Many clinicians choose to treat metabolic acidosis when arterial pH values are
substantially below 7.25 but tolerate mild degrees of acidosis rather than risk overcorrection and alkalosis.
8.6.1. Receptor upregulation and downregulation

Tachyphylaxis, or a reduction in the expected response to an agonist with repeated use, is familiar to anyone who has overused overthecounter nasal decongestants.
Tachyphylaxis appears to be mediated either by uncoupling of receptors from G proteins, or a net internalization of hormonereceptor complexes within the cell25, a
process called ‘downregulation’ (Figure 8.5). Upregulation, an increase in the number of receptors per cell, occurs following chronic receptor blockade (Figure 8.6).
This phenomenon becomes especially important if the chronic blockade is suddenly terminated,
Page 122
Figure 8.5 Downregulation (internalization and inactivation of adrenergic receptors) occurs after chronic stimulation. Higher doses of
agonist agents are required to obtain the desired effect
Figure 8.6 Upregulation occurs following chronic blockade. New receptors are generated; if the blocking agent is suddenly withdrawn
(as might happen with cessation of chronic antihypertensive betaadrenergic blocking agents), an exaggerated clinical response
will be seen with subsequent stimulation.
as the more numerous and now unoccupied receptors can generate an exaggerated clinical response. For example, when propranolol therapy is withdrawn
perioperatively, an increased response to isoproterenol can be seen during the next 12–48 h; clinical withdrawal of beta blockade can precipitate undesirable
tachycardia.
8.6.2 Electrolyte disorders and adrenergic response

Adrenergic receptor action is a calciumdependent process, and the level of serum electrolytes modulates the activity of the endogenous sympathetic system and
exogenous pharmacologic agents. Within a relevant range, left ventricular contractility varies directly with bloodionized calcium levels26.
Normal circulating calcium concentrations are needed for adequate myocardial performance and maintenance of vascular tone. Patients with Gramnegative
infection, however, may develop hypocalcemia as a result of parathyroid gland dysfunction and/or vitamin D insufficiency27. Concerns have been raised, however, over
calcium administration in sepsis, because increased mortality has been noted in calciumtreated animals with endotoxic shock28.
Hypophosphatemia is also common in critically ill patients and contributes to respiratory insufficiency, erythrocyte and leukocyte dysfunction, metabolic acidosis, and
central nervous system dysfunction. Maintenance of normal levels of 2,3diphosphoglycerate requires normal phosphate metabolism.
Page 123
Correction of hypophosphatemia improves myocardial function possibly via improving intracellular availability of ATP29.
Hypomagnesemia is another common finding in the critically ill. Magnesium has a permissive action on parathyroid hormone (PTH) release, and thus,
hypomagnesemia may contribute to hypocalcemia.
8.7 Bypassing the adrenergic receptor
8.7.1 Phosphodiesterase inhibitors

Phosphodiesterase inhibitors (PDIs) prevent the breakdown of cAMP and increase intracellular cAMP levels (see Figure 8.3). Selective phosphodiesterase inhibitors
include amrinone, milrinone, enoximone and a number of experimental agents; all raise intracellular cAMP levels by inhibiting phosphodiesterase F3. Because the PDIs
act intracellularly and bypass both the adrenergic receptor and the calcium channel, they can reverse the cardiopressant effects of drugs such as verapamil and
propranolol30. Administration of the PDI amrinone significantly improves cardiac index, pulmonary capillary wedge pressure and right atrial pressure, and decreases
systemic vascular resistance and pulmonary vascular resistance in low output states31. Amrinone may also have advantages over the alphaadrenergic agents in
restoring intracellular calcium levels toward normal32. Dobutamine and amrinone at equipotent doses produce similar effects on cardiac index, pulmonary capillary
wedge pressure, mean right atrial pressure, heart rate, systemic vascular resistance, and mean arterial pressure33. The addition of amrinone to 15 µg kg−1 min−1 of
dobutamine has an additional synergistic effect, resulting in continued increase in ejection fraction34. Combination therapy with dopamine and amrinone has also been
described in the setting of refractory lowoutput syndrome after open heart surgery35.
8.7.2 Milrinone
Milrinone is a chemical relative of amrinone with higher potency and fewer side effects. Milrinone is more cost effective and is largely supplanting amrinone, although
both drugs remain available in the United States. Milrinone is given as a 50–75 µg kg−1 i.v. bolus, followed by a 0.3–0.75 µg kg−1 min−1 infusion. Hypotension,
tachycardia, atrial fibrillation, and atrial bigeminy can occur, as with amrinone, but milrinone is associated with a lower incidence of fever and thrombocytopenia. In
patients with congestive heart failure, both milrinone and dobutamine improve cardiac index, but dobutamine increases myocardial oxygen consumption, whereas
milrinone does not36.
8.8 Vasopressin
The body’s response to stress includes release of epinephrine, cortisol, corticotropin, renin and vasopressin. Vasopressin is released in response to decreased blood
volume, increased plasma osmolality, and pain. Endogenous release of vasopressin corresponds with the return of spontaneous circulation after cardiopulmonary
resuscitation37, leading to speculation that exogenous vasopressin might be useful in shock. Experimentally, vasopressin increases vital organ flow more than
epinephrine38. Synergistic effects of vasopressin plus epinephrine have been noted in a porcine cardiopulmonary resuscitation model39. Vasopressin has also been
shown to reverse intractable hypotension in the late phase of hemorrhagic shock in dogs40.
Clinical studies in humans are limited41. Increases in mean arterial pressure and systemic vascular resistance are observed when vasopressin is infused in septic, but
not cardiogenic shock, patients42. In septic trauma patients, infusion of vasopressin at 0.04 U min−1 increased blood pressure and systemic vascular resistance and
increased ability to wean from pressors43. Vasopressin has also been used in a bolus dose of 40 U (c. 0.5 U kg−1) to restore spontaneous cardiovascular function in
patients failing IV epinephrine and defibrillation44.
As discussed earlier, responsiveness to beta stimulation is reduced in sepsis and
Page 124
severe heart failure. Although milrinone may improve inotropy and reduce afterload in this situation, it is often associated with unacceptable hypotension. One recent
approach has been to combine vasopressin with milrinone, as the afterload increasing effects of vasopressin are offset by milrinone and the hypotensive effects of the
milrinone may be specifically inhibited by vasopressin45.
Largescale trials are necessary before the routine use of vasopressin can be recommended, and it is not currently FDAapproved for this indication. However,
synthetic vasopressin (DDAVP, desmopressin) is readily available for other indications and the literature supports initial doses in the range of 0.01–0.07 U min−1, with
upward titration as needed. Appropriate monitoring should be carried out to ensure that the increase in afterload does not adversely decrease cardiac output. In
addition to the expected effect of increased afterload, there is the possibility of coronary and mesenteric vasoconstriction, which argues for continuous EKG monitoring.
8.9 Weaning pharmacologic support

The rapidity with which pharmacologic support can be weaned from the patient in shock must be individualized, and depends on the length of inotropic therapy, patient
characteristics and the halflives of the agents used. For catecholamines, which have short halflives (minutes), titration can be rapid, with decreases of 10–25% on a
halfhour or hourly basis, relying on clinical monitoring to determine the end point of weaning. The PDIs have longer halflives (50 min for milrinone, >200 min for
amrinone) and can be decreased by 50–100% acutely, but require a longer period of observation (6–8 h) before further weaning is attempted.
8.9.1 Removal of monitoring and ICU discharge

Despite the extremely high mortality rate with circulatory shock, some patients demonstrate hemodynamic improvement, which may either be dramatic or gradual over
a period of a week or more. Surviving patients often have multiorgan system dysfunction or continued respiratory support needs. The pulmonary artery catheter should
be removed when the cardiac index has been stable and the information gained from the catheter no longer outweighs the risk of leaving it in place. Most patients can
be weaned from inotropic support without a pulmonary artery catheter in place, although continued central venous monitoring and arterial monitoring are usually helpful.
Removal of central venous lines often depends more on vascular access requirements for medication and total parenteral nutrition rather than the need for CVP
readings.
The patient is ready for ICU discharge when hemodynamics have been stable for an extended period of time (12–24 h) and when the likelihood of emergent return
to the ICU from the ‘floor’ is small. Beware that the phosphodiesterase inhibitors have a longer ‘tailend’ effect; 36–48 h of postinfusion monitoring may be more
appropriate in these patients. Discharge criteria also depend on the hospital’s facilities and nursing staffing ratios, including availability of monitoring on a ward or
intermediate level unit.
8.10 Summary
Pharmacologic support is most effective when intravascular volume is repleted and the patient’s acidbase status is normal. Catecholamines are chosen on the basis of
inotropic and peripheral vasoconstrictive effects; combination therapy may be needed to address both inotropic and pressor needs. Catecholamine response may be
altered by the patient’s premorbid condition, acidosis, prolonged drug therapy (downregulation), and hypothermia. Although the catecholamines and phosphodiesterase
inhibitors are at present the approved therapeutic options, vasopressin is emerging as a valuable adjunct, particularly in sepsis.
Page 125
References
1. Matsuda, N., Hattori, Y., Akaishi, Y. et al. (2000) Impairment of cardiac βadrenoceptor cellular signaling by decreased expression of Gsa in septic rabbits.
Anesthesiology 93:1465–1473.
2. Karamian, A. (1984) Physiology of the nervous system: The autonomic nervous system. In: Physiology for the Anesthesiologist, 2nd Edn (eds N.Goudsozin and
A.Karamian). Appleton Century Crofts, Norwalk, CT; p. 328.
3. Maze, M., Haywood, E. and Gaba, D.M. (1985) Alpha1 adrenergic blockade raises epinephrine arrhythmic threshold in halothane anesthetized dogs in a dose
dependent fashion. Anesthesiology 63:611.
4. Carey, R.M., Siragy, H.M., Ragsdale, N.V., et al. (1990) Dopamine1 and dopamine2 mechanisms in the control of renal function. Am. J. Hypertension 3:595–
635.
5. Abid, O., Akca, S., HajiMichael, R, et al. (2000) Strong vasopressor support may be futile in the intensive care unit patient with multiple organ failure. Crit. Care
Med. 28:947–949.
6. Desjars, P., Pinaud, M., Potel, G., et al. (1987) A reappraisal of norepinephrine therapy in human septic shock. Crit. Care Med. 15:134–137.
7. Desjars, P., Pinaud, M., Bugnon, D. and Tasseau, F. (1989) Norepinephrine therapy has no deleterious renal effects in human septic shock. Crit. Care Med.
17:426–429.
8. Schaer, G.L., Fink, M.P. and Parrillo, J.E. (1985) Norepinephrine alone versus norepinephrine plus lowdose dopamine enhanced renal blood flow with combination
pressor therapy. Crit. Care Med. 13:492–496.
9. Davis, R.F., Lappas, D.G., Kirklin, J.K., et al. (1982) Acute oliguria after cardiopulmonary bypass renal functional improvement with lowdose dopamine infusion.
Crit. Care Med. 10: 852–856.
10. Polson, R.J., Park, G.R., Lindop, M.J., et al. (1987) The prevention of renal impairment in patients undergoing orthotopic liver grafting by infusion of low dose
dopamine. Anaesthesia 42: 15–19.
11. Flancbaum, L., Choban, P.S. and Dasta, J.F. (1994) Quantitative effects of lowdose dopamine on urine output in oliguric surgical intensive care unit patients. Crit.
Care Med. 22:61–66.
12. Chertow, G.M., Sayegh, M.H., Allgren, R.L., et al. (1996) Is the administration of dopamine associated with adverse or favorable outcomes in acute renal failure?
Am. J. Med. 101:49–53.
13. Brath, P.C., MacGregor, D.A., Ford, J.G. and Prielipp, R.C. (2000) Dopamine and intraocular pressure in critically ill patients. Anesthesiology 93:1398–1400.
14. Mathur, V.S., Swan, S.K., Lambrecht, L.J., et al. (1999) The effects of fenoldopam, a selective dopamine receptor agonist, on systemic and renal hemodynamics in
normotensive subjects. Crit. Care Med. 27: 1832–1837.
15. Murphy, M.B., McCoy, C.E., Weber, R.R., et al. (1987) Augmentation of renal blood flow and sodium excretion in hypertensive patients during blood pressure
reduction by intravenous administration of the dopamine1 agonist fenoldopam. Circulation 78:1312–1318.
16. Poinsot, O., Romand, J.A., Favre, H. and Suter, P.M. (1993) Fenoldopam improves renal hemodynamics impaired by positive endexpiratory pressure.
Anesthesiology 79: 680–684.
17. Desai, J.K., Goyal, R.K. and Parmar, N.S. (1999) Characterization of dopamine receptor subtypes involved in experimentally induced gastric and duodenal ulcers in
rates. J. Pharmacol. 51:187–192.
18. Tarnow, J. and Komar, K. (1988) Altered hemodynamic response to dobutamine in relation to the degree of preoperative βadrenoreceptor blockade.
Anesthesiology 68:912.
19. Uretsky, B.F., Lawless, C.E., Verbalis, J.G., et al. (1987) Combined therapy with dobutamine and amrinone in severe heart failure. Chest 92: 657–662.
20. Martin, C., Vivland, X., Arnaud, S., et al. (1999) Effects of norepinephrine plus dobutamine or norepinephrine alone on left ventricular performance of septic shock
patients. Crit. Care Med. 27:1708–1713.
21. Vincent, J.L., Reuse, C. and Kahn, R.J. (1989) Administration of dopexamine, a new adrenergic agent, in cardiorespiratory failure. Chest 96: 1233.
22. Poelaert, J.I.T., Mungroop, H.E., Koolen, J.J. and Van den Berg, P.C.M. (1989) Hemodynamic effects of dopexamine in patients following coronary artery
bypass surgery. J. Cardiothorac. Anesth. 3:441–443.
23. Byers, R.J., Eddleston, J.M., Pearson, R.C., et al. (1999) Dopexamine reduces the incidence of acute inflammation in the gut mucosa after abdominal surgery in
highrisk patients. Crit. Care Med. 27:1787–1793.
24. Barton, M., Lake, C.R., Rainey, T.G. and Chernow, B. (1982) Is catecholamine release pH mediated? Crit. Care Med. 10:751–753.
25. Lefkowitz, R.J., Caron, M.G. and Stiles, G.L. (1984) Mechanisms of membranereceptor regulation. N. Engl. J. Med. 310:1570–1576.
Page 126
26. Lang, R.M., Fellner, S.K., Neumann, A., et al. (1988) Left ventricular contractility varies directly with blood ionized calcium. Ann. Intern. Med. 108:524–529.
27. Zaloga, G.P. and Chernow, B. (1987) The multifactorial basis for hypocalcemia during sepsis. Ann. Intern. Med. 107:36.
28 Malcolm, D.S., Zaloga, G.P. and Holaday, J.W. (1989) Calcium administration increases the mortality of endotoxic shock in rats. Crit. Care Med. 17:900.
29. O’Connor, L.R., Wheeler, W.S. and Bethune, J.E. (1977) Effect of hypophosphatemia on myocardial performance in man. N. Engl J. Med. 297:901.
30. Makela, V.H.M. and Kapur, P.A. (1987) Amrinone and verapamilpropranolol induced cardiac depression during isoflurance anesthesia in dogs. Anesthesiology
66:792.
31. Goenen, M., Pedemonte, O., Baele, P. and Col, J. (1985) Amrinone in the management of low cardiac output after open heart surgery. Am. J. Cardiol. 56:33B.
32. Auffermann, W., Stefenelli, T., Wu, S.T., et al. (1991) Influence of positive inotropic agents on intracellular calcium transients. Part I. Normal rat heart. Am. Heart
J. 118:1219–1227.
33. Benotti, J.R., McCue, J.E. and Alpert, J.S. (1985) Comparative vasoactive therapy for heart failure. Am. J. Cardiol. 56:1B.
34. Sundram, P., Reddy, H.K., McElroy, P.A., et al. (1990) Myocardial energetics and efficiency in patients with idiopathic cardiomyopathy: response to dobutamine
and amrinone. Am. Heart J. 119:891.
35. Olsen, K.H., Kluger, J. and Fieldman, A. (1988) Combination high dose amrinone and dopamine in the management of moribund cardiogenic shock after open
heart surgery. Chest 94:503.
36. Grose, R., Strain, J., Greenberg, M. and Lejemtel, T.H. (1986) Systemic and coronary effects of intravenous milrinone and dobutamine in congestive heart
failure. J. Am. Coll. Cardiol. 7: 1107.
37. Lindner, K.H., Strohmenger, H.U., Ensinger, H., et al. (1992) Stress hormone response during and after cardiopulmonary resuscitation. Anesthesiology 77:662–
668.
38. Eichinger, M.R. and Walker, B.R. (1994) Enhanced pulmonary arterial dilatation to arginine vasopressin in the chronically hypoxic rat. Am. J. Physiol 267:H2413–
H2429.
39. Mulligan, K.A., McKnite, S.H., Lindner, K.H., et al. (1997) Synergistic effects of vasopressin plus epinephrine during cardiopulmonary resuscitation.
Resuscitation 35:265–271.
40. Morales, D., Madigan, J., Cullinane, S., et al. (1999) Reversal by vasopressin of intractable hypotension in the late phase of hemorrhage shock. Circulation
100:226–229.
41. Mets, B., Michler, R.E., Delphin, E.D., et al. (1998) Refractory vasodilation after cardiopulmonary bypass for heart transplantation in recipients on combined
amiodarone and angiotensinconverting enzyme inhibitor therapy: A role for vasopressin administration. J. Cardiothorac. Vasc. Anes. 12:326–329.
42. Landry, D.W., Levin, H.R., Gallant, E.M., et al. (1997) Vasopressin deficiency contributes to the vasodilation of septic shock. Circulation 95: 1122–1125.
43. Malay, M.B., Ashton, R.C., Landry, D.W., et al. (1999) Lowdose vasopressin in the treatment of vasodilatory septic shock. J. Trauma Infect Crit. Care 47:699–
705.
44 Lindner, K.H., Prengel, A.W., Brinkman, A., et al. (1996) Vasopressin administration in refractory cardiac arrest. Ann. Intern. Med. 124: 1061–1064.
45. Gold, J., Cullinane, S., Chen, J., et al. (2000) Vasopressin in the treatment of milrinoneinduced hypotension in severe heart failure. Am. J. Cardiol. 85:506–509.
Page 127
Chapter 9
Hypertensive urgencies and emergencies
Contents
9.1 Introduction
9.2 Definitions
9.3 Physiology of blood pressure control: MAP=SVR×CO
9.4 Monitoring considerations
9.5 Agents to decrease blood pressure
9.6 Specific clinical considerations
9.7 Medicationrelated hypertension
9.8 Conclusions
Page 128
9.1 Introduction
Hypertensive urgencies and emergencies can account for ICU admission or prolonged critical care lengthofstay. Interventions to control blood pressure elevations in
the ICU range from simple sedation and analgesia to multidrug IV regimens. Most blood pressure elevation can be controlled over the course of several hours with
oral agents, given if necessary via orogastric or feeding tube. Hypertensive emergencies require immediate aggressive intervention while avoiding overtreatment and
consequent hypotension. For malignant hypertension, it is recommended to lower the MAP by about 25% over the first 2–4 h, while keeping the diastolic blood
pressure above 100 mmHg. Initial therapy should be instituted while searching for an underlying cause (Tables 9.1–9.3). Although there are no studies showing an
outcome benefit with a particular treatment regimen, an understanding of the pathophysiology causing the hypertension can help rationalize the choice of
antihypertensives.
9.2 Definitions
Hypertension is frequently defined as systolic blood pressure in excess of 160 mmHg, and/or diastolic pressure above 90 mmHg. Perioperative hypertension is defined
as an increase of more than 20% over preoperative levels1. The importance of any given blood pressure reading; however, will vary by age, gender, and clinical
circumstances.
Hypertensive crisis is a severe elevation of blood pressure and may be urgent or emergent. Accelerated hypertension, also called hypertensive urgency, is defined by
diastolic blood pressures greater than 120 mmHg without evidence of endorgan injury. Malignant hypertension or hypertensive emergency is accelerated hypertension
with evidence of endorgan damage, typically a necrotizing vasculitis that results in smallvessel occlusion. The presence of retinopathy helps distinguish malignant from
accelerated hypertension. Patients with either accelerated or malignant
Table 9.1 Medical causes of hypertension
Essential hypertension
Renal disease
• Acute or chronic renal failure
• Renovascular hypertension
Endocrinerelated disorders
• Hypoglycemia
• Hyperthyroidism
• Pheochromocytoma (diagnosed or latent)
• Catecholaminesecreting tumors
• Hyperparathyroidism
• Acromegaly
• Myxedema
• Cushing’s syndrome
• Carcinoid syndrome
• Reninsecreting tumors
Myocardial ischemia or infarction
Medication related
• Rebound: clonidine, beta blockers
• Monoamine oxidase inhibitors
• Tricyclic antidepressant
Other
• Obesity
• Advanced age
• Pregnancy (preeclampsia or eclampsia)

• Oral contraceptives
• Alcohol or opioid withdrawal
• Polyarteritis nodosa
• Hypercalcemia
• Increased intracranial pressure
hypertension may have nonspecific symptoms such as headache and blurred vision, and cardiac symptoms are common. Nausea, vomiting and focal neurologic deficits
are uncommon. Hypertensive encephalopathy is a central nervous system vasculopathy usually seen in patients with chronic hypertension. Headache, nausea, vomiting,
blurred vision and confusion are common findings with hypertensive encephalopathy; focal neurologic deficits may also occur.
Medical emergencies commonly associated with hypertension include acute renal failure, pulmonary edema, eclampsia, cerebral ischemia, cerebral hemorrhage,
angina, pheochromocytoma, monoamine oxidase (MAO) inhibitor crisis, cocaine overdose, and antihypertensive withdrawal syndrome. Postoperative hypertension
occurs in 3–6% of
Page 129
Table 9.2 Surgical and anesthetic events associated with hypertension
Surgical causes Anesthetic causes

Cardiac surgery Intravenous agents
• Myocardial revascularization • Opioids (meperidine or pethidine)
• Valve repair or replacement • Ketamine
• Heart transplantation • Neuroleptic agents
Vascular surgery • Opioid antagonists
• Aortic aneurysm resection Anesthetic emergence
• Aortic coarctation • Metabolism of intravenous agents
• Carotid endarterectomy • Rebound from induced hypotension
Neurosurgical procedures Regional anesthesia
• Neurovascular procedures • Systemic absorption of epinephrine
• Increased intracranial pressure • Tourniquet pain
• Autonomic hyperreflexia Endotracheal intubation
Abdominal surgery
• Endocrine tumors Residual intraoperative medications
• Liver transplantation • Vasopressor effect
• Renal surgery and transplantation • Anticholinergics
• Urologic surgery Preoperative medications
Obstetric surgery • Monoamine oxidase inhibitors
• Eclampsia • Tricyclic antidepressants
• Preeclampsia • Illicit drugs
• Hypertension of pregnancy • Nonprescription drugs
Radical neck dissection Discontinuation of preoperative medication
• Beta blocker therapy
• Clonidine
Table 9.3 Quick differential checklist for hypertension in the ICU
• Hypoxemia
• Hypercarbia
• Acidosis
• Pain
• Anxiety
• Shivering
• Hypervolemia
• Hypovolemia
• Ventilator dyssynchrony or stimulation from endotracheal tube
• Distended bladder, stomach, or intestine
patients following noncardiac surgery1,2, and up to 50% following myocardial revascularization3. Postoperative hypertension is a significant risk factor for myocardial
ischemia, heart failure and cerebrovascular events. Pain, bladder distention, hypothermia with shivering, and intolerance of endotracheal intubation can trigger
hypertension as patients emerge from anesthesia. Patients may also experience respiratory difficulties related to airway obstruction or prior sedation and neuromuscular
blockade, resulting in hypercapnia, hypoxemia, or acidosis, which further increases sympathetic tone. The increased sympathetic nervous system activity results in both
alpha1mediated vasoconstriction increasing vascular resistance and venous return, and
Page 130
beta1receptormediated increases in the heart rate.

Hypotension is subnormal arterial blood pressure, but the precise level at which a patient might be considered hypotensive is situationspecific. Mean arterial
pressures of 60 mmHg or lower can be seen normally in healthy, young individuals who are asleep or sedated, but pressures this low raise concern in most critically ill
patients. In individuals with chronic hypertension, hypertrophy of the arterial musculature limits the ability of vascular beds to vasodilate when MAP drops.
Hypertensive individuals have a rightward shift in their cerebral autoregulatory curve (Figure 9.1), and may develop symptoms of ischemia even at pressures more than
adequate for normotensive patients. Tissue perfusion depends on the resistance of the vascular bed, so a MAP of 60 mmHg may produce sufficient flow in the
presence of vasodilation, whereas a MAP of 100 mmHg could be inadequate in the presence of severe vasoconstriction. Endorgan assessment (cerebral function,
myocardial function, urine output, peripheral skin temperature) tempers the clinical interpretation of hypotension. Common causes of low blood pressure are
hypovolemia, decreased venous and/or arterial tone, and decreased cardiac output. Chapter 8 more fully describes shock states and the interactions between blood
pressure, myocardial function, and systemic vascular resistance.
9.3 Physiology of blood pressure control: MAP=SVR×CO

Systemic vascular resistance is determined primarily by the state of vascular smooth muscle tone, which in turn depends on input from the adrenergic and renin
angiotensin systems as well as local control from vascular endothelium (Figure 9.2). Cardiac output is determined by heart rate and stroke volume. Stroke volume in
turn is determined by cardiac preload, afterload and myocardial contractility. The circulatory system has highly sophisticated mechanisms for maintaining mean arterial
Figure 9.1 Cerebral autoregulation. HTN, hypertension.
pressure in the normal range. These control mechanisms (Table 9.4) may be categorized as shortterm, intermediate, or longterm. Rapidly acting control mechanisms
(seconds to minutes) rely on baroreceptor and chemoreceptor feedback to regulate central nervous system responses. Baroreceptors are located in the walls of large
arteries in the thoracic and cervical regions, most notably in the aorta and the internal carotid artery. As blood pressure increases, stretch of nerve endings signals for
inhibition of the medullary vasoconstrictive center, enhancement of vagal tone, and ultimately peripheral vasodilation, diminished heart rate, and diminished force of
myocardial contraction. The baroreceptor response adapts to persistent pressure changes over a 24–48 h period, limiting its utility in chronic hypertension4. Baroreflex
failure rarely occurs, but may present as severe labile hypertension and/or hypotension with elevated plasma catecholamine levels5.
Chemoreceptors respond to decreased oxygen supply and accumulation of excess hydrogen ion6. Unlike the baroreceptors, which are active in the normal blood
pressure range, chemoreceptors do not respond until
Page 131
Table 9.4 Blood pressure adaptation and control mechanisms
Adaptation Onset Peak effect Duration Feedback

Short term
Baroreceptor <15 s 30 s 2–4 days Pressure
Chemoreceptor <15 s 60 s Prolonged PO2, PCO2, pH
Central nervous system <15 s Minutes Prolonged Ischemia
Intermediate
Stress relaxation 30 s Days Prolonged Pressure
Capillary fluid shift 8 min Days Prolonged Pressure
Reninangiotension 2 min 45 min Prolonged Pressure
Long Term
Blood volume 5–120 min Days Prolonged Renal
PCO2 , partial pressure of carbon dioxide, PO2 , partial pressure of oxygen
the mean arterial pressure falls well below 80 mmHg, arterial PO2 falls below normal, or arterial CO2 or hydrogen ion concentration rises significantly. Under these
conditions, nerve transmission to the medullary center results in sympathetic excitation and vagal inhibition7. In addition, stretch receptors located in the systemic and
pulmonary arteries recognize increased venous return, and will cause reflex vasodilation8.
Intermediate mechanisms of blood pressure control involve hormonal mechanisms active over minutes to hours. Acute hypotension stimulates adrenal release of the
endogenous vasoconstrictor norepinephrine, and triggers renin release leading to afferent renal arteriolar
Figure 9.2 Blood pressure control. SV=stroke volume, SVR=systemic vascular resistance, AT=angiotensin, ACE=angiotensin
converting enzyme.
Page 132
vasodilation and efferent vasoconstriction. Persistent low mean arterial pressure, however, results in increased levels of angiotensin, which ultimately constricts afferent
arterioles and decreases glomerular blood flow. Angiotensin I is inactive, but is converted in the lungs to active angiotensin II. Inhibitors of angiotensinconverting enzyme
(ACE) are therefore useful antihypertensives. Hypotension also stimulates hypothalamic secretion of vasopressin, which has direct vasoconstrictive effects, and also
plays a role in renal tubular fluid reabsorption. Further intermediate regulation occurs as a result of relaxation of capacitance vessels9.
Longterm regulation (hours to days) of blood pressure involves the reninangiotensin system, and adrenal cortical release of aldosterone10. Structural changes
associated with chronic hypertension include cardiac hypertrophy, diminished left ventricular distensibility, diastolic dysfunction, and arterial muscular hypertrophy.
9.4 Monitoring considerations

Mean arterial pressure (MAP) is typically the same in all large arterial vessels in a supine patient, with only a small gradient between the aorta and a radial arterial line.
Gradients may increase following hypothermia and rewarming, as in cardiac surgery11. Systolic and diastolic pressures, however, vary by location, and systolic pressure
tends to rise in the periphery owing to the influence of reflected waves (Figure 9.3). Automated sphygmomanometry (e.g. Dynamap) and manual cuff pressures are
inaccurate when cuff width is less than twothirds of arm circumference. Use of a narrow cuff, and presence of atherosclerotic vessels both tend to overestimate true
blood pressure.
Arterial cannulation is indicated with hemodynamic instability or shock, malignant hypertension, concurrent oxygenation problems, or when frequent blood sampling
is required. Complications of arterial line placement include local hemorrhage, infection, and thrombosis. Large catheter size relative to vessel size12, low cardiac output,
preexisting
Figure 9.3 Influences of reflected waves.
vascular disease and vasopressors all increase the risk of arterial thrombosis.
9.5 Agents to decrease blood pressure

Figure 9.2 demonstrates the points at which pharmacologic agents act to influence blood pressure. Antihypertensive agents can be categorized by their mechanisms of
action:
• direct vasodilation (e.g. sodium nitroprusside, hydralazine);
• autonomic ganglia blockade (e.g. trimethaphan);
• alphaadrenergic blockade (e.g. phentolamine);
• alpha2adrenergic stimulation (e.g. clonidine);
• beta blockade (e.g. esmolol, propranolol, metoprolol);
• combined alpha and beta blockade (e.g. labetalol);
• calcium channel blockade (e.g. nicardipine, isradipine, nifedipine);
• ACE inhibition (e.g. captopril, enalaprilat);
• angiotersin III receptor antagonists (e.g. Iosartan);
• diuresis (e.g. furosemide, bumetanide, thiazide diuretics);
• sedation and/or analgesia.
The wide variety of effective agents makes it possible to individualize therapy based on
Page 133
urgency, routes of access, and presence of contraindications to specific drugs (Table 9.5). Direct vasodilators, particularly sodium nitroprusside, tend to be the most
rapidly acting and effective agents, but require close titration and have undesirable side effects, including tachyphylaxis. The more indirect the mechanism of blood
pressure reduction, the longer it may take to achieve a satisfactory result in a hypertensive emergency. A typical approach is to use sodium nitroprusside, hydralazine or
fenoldopam for immediate control of acute hypertension, while also starting a longeracting agent, one that can be given orally and outside a closely monitored setting.
Patients may not respond to a single agent, so combinations of agents with differing mechanisms of action are frequently necessary. Betaadrenergic antagonists or
combined alpha and beta blockers prevent and treat the reflex tachycardia that can occur following peripheral vasodilation. Labetalol and beta blockers, however, are
relatively contraindicated in patients with bronchospastic disease or highdegree atrio ventricular conduction block. Nitroglycerine is useful in the presence of coronary
insufficiency, and loop diuretics and nitroglycerine together are indicated when pulmonary edema is also present. Afterload reduction with an ACE inhibitor may be
appropriate with heart failure or when the renin system is activated, but ACE inhibitors should be avoided in patients with bilateral renal artery stenoses or renal artery
stenosis with a solitary kidney. Centrally acting agents, particularly clonidine, are helpful in the presence of excess sympathetic discharge, as may be seen with head
injury or substance withdrawal. However, these agents also tend to cloud mental status and amplify the effects of opioids. Table 9.6 outlines considerations for therapy
in specific disease states.
9.5.1 Direct vasodilators

Nitroprusside (Nipride) acts directly on arterioles and venules by decreasing the tone of vascular smooth muscle, resulting in reduced afterload and less increase in
preload compared with other vasodilators. Nitroprusside has an immediate onset of action and effects dissipate in 2–3 min. The solution, typically mixed as 50 mg of
nitroprusside in 250 cm3 of D5W, must be protected from light; deeply discolored solutions lack potency and should be discarded. Nitroprusside should be given by
infusion pump. The usual starting dose is 0.1–0.3 µg kg−1 min−1, but doses may be rapidly titrated based on patient response to a maximum acute dose of 10 µg kg−l
min−1 for 10 min or less. Sustained infusions should not exceed 3 µg kg−1 min−1, particularly with hepatic or renal dysfunction. Extracardiac effects of sodium
nitroprusside include increased renal blood flow (as a result of increased cardiac output) and reduced pulmonary artery pressure and pulmonary vascular resistance,
which may alter regional blood flow in the lung, worsening the ventilationperfusion ratio, and increasing dead space ventilation. Nitroprusside combines with
hemoglobin to produce cyanomethemoglobin and cyanide ions. Cyanide combines with thiosulfate to produce thiocyanate, a reaction catalyzed by the liver enzyme
rhodanase. Thiocyanate is then eliminated by the kidneys. Excess cyanide inhibits oxidative metabolism, which will present clinically as metabolic acidosis, increased
mixed venous oxygen, and a reduced arteriovenous oxygen difference. Thiocyanate toxicity can also occur, especially with renal failure, but also if nitroprusside
infusions are maintained for more than 72 h. Thiocyanate toxicity can present as delirium, headache, nausea, abdominal pain, muscle spasms or restlessness. Although
thiocyanate levels can be measured, results are typically not available in a clinically relevant time frame. Increasing nitroprusside dose requirement, mental confusion and
lactic acidosis should raise suspicion of toxicity. Administration of thiosulfate (150 mg kg−1) provides sulfate donors facilitating the conversion of cyanide to
thiocyanate13. Amyl nitrate (inhaled) or sodium nitrate (5 mg kg−1 intravenously) provides nitrate ions to convert hemoglobin to methemoglobin, which binds the
cyanide molecule. Hydroxycobalamin
Page 134
Table 9.5 Agents for control of hypertension in the ICU
Mechanism Drug Dosage Onset Duration Indications Contraindications Side effects Comments
of action of action
Vascular Sodium IV: 0.3–10 Immediate 2–3 min Postoperative hypertension Pregnancy Hypotension, nausea, and Invasive pressure
smooth nitroprusside µg kg−1 Hypertensive crisis cyanide toxicity monitoring
muscle General purpose (increased with renal and recommended Shield
relaxation min−1 hepatic insufficiency) from light
Usually ≤3
µg kg−1
min−1 Mix 50
mg per 100
ml
Hydralazine IV: 2.5–20 15–30 min IV.: 3–20 Postoperative hypertension Severe coronary Fluid retention, lupuslike Tolerance may
mg bolus min Oral: disease, lupus, aortic rash, headache, positive develop Ideal for
i.m.: 20–40 2–4 h stenosis Coombs’ test eclampsia
mg q 4–6 h
Oral: 10–50
mg q 6 h
Nitroglycerin IV: 5–100 µg 1–2 min 3–5 min Myocardial ischemia or Septal hypertrophy, Bradycardia, Rapid tachyphylaxis
min−1 infarction hemoglobinopathies tachycardia, Venodilation at low
Sublingual methemoglobinemia, doses
0.15–0.5 mg flushing, vomiting,
headache
Diazoxide IV: 50–100 1–5 min 6–12 h Postoperative hypertension Diabetes, Hypotension, Potentiated by beta
mg q 5–10 hyperuricemia, CHF tachycardia, fluid blockers and
min, up to retention, myocardial diuretics
600 mg ischemia, hyperglycemia,
Infusion: 30 nausea, vomiting, heart
mg min−1 failure
Nitric oxide Inhaled 15– Immediate Seconds Experimental
30 ppm
Autonomic Trimethaphan IV: 1–5 mg 2–5 min 4–12 h Postoperative hypertension CAD, prostatism, Ileus, urinary retention, Invasive pressure
ganglia min−1 titrate cerebrovascular hypotension, blurred monitoring
blockers to effect. insufficiency, diabetes vision, dry mouth, angina, recommended
Mix 50 mg mellitus, glaucoma pupillary dilation
per 500cm3
of normal
saline (NS)
Alpha Phentolamine IV: 5–20 mg Immediate 2–10 min Pheochromocytoma CAD Tachycardiac, vomiting,
blockers q 10 min angina, hypotension,
miosis, dry mouth
Combined Labetalol IV: 10–80 5–10 min 3–6 h General purpose Heart block, asthma, Hypotension, heart Increases digoxin
alpha and beta mg q 10 min bradycardia, heart failure, heart block, level, prolongs
blockers to 300 mg failure, hypertension dizziness, nausea, neuromuscular
max. secondary to decreased phlebitis, bronchospasm, blockade of
Infusion: 0.5– cardiac output cardiac arrest if given succinylcholine
2 mg min−1 with verapamil
Beta blocker Esmolol IV: load with <5 min 10–20 min CAD w/o left ventricle Asthma, heart block, Heart failure, Elevates digoxin
500 µg kg−1 dysfunction, aortic sinus bradycardia, heart bronchospasm, heart level, prolongs
over 1–2 min dissection w/o heart failure, cardiogenic block, hypotension, neuromuscular
and infusion failure, postoperative shock, pregnancy phlebitis, dizziness, blockade with
50–200 µg hypertension related to urinary retention succinylcholine, risk
increased sympathetic of fatal cardiac
kg−1 min−1 tone, hyperthyroidism arrest if combined
with verapamil
Page 135
Mechnism of Drug Dosage Onset Duration Idication Contraindications Side effects Comments
action of action
Calcium Isradipine Oral: 2.5–20 mg b.i.d. 2–3 h 12+ h Afterload reduction Tight aortic stenosis Headache, dizziness, Diuretic effects
antagonist edema
Nifedipine SI: 10–20 mg q 15 1–5 20–30 min Has been used for Heart block, IV beta Headache, Watch for overshoot
min* min 30–60 min acute BP control blockers w/o pacer tachycardia, flushing, Rapid uncontrolled BP
Oral:10–20 mg t.i.d. 5–10 backup dizziness, hypotension, reduction can precipitate
min phlebitis strokes
or infarction
Nicardipine IV: 5–15 mg h–1 1–5 3–6 h Myocardial Severe aortic stenosis, Hypotension,
min ischemia or hypersensitivity tachycardia, nausea,
infarction minimal myocardial
depression
Nimodipine PO: 60 mg q 4 h 1–2 h 4–6 h Subarachnoid Pregnancy, hepatic Edema, hypotension, Can administer through
hemorrhage impairment headache, ileus, NG tube
nausea, abnormal liver
function
Centrally acting Clonidine IV: 0.1–0.2 mg, PO 30 6–8 h Rebound hyper Heart block, Drowsiness, sedation, Reduce dose in elderly
0.2 mg q 6 h min–1 tension following bradycardia, sick sinus dry mouth, augments cerebrovascular or
transdermal patch h discontinuation of syndrome opioids, nightmares, cardiovascular disease;
max. 2.4 mg day–1 clonidine restlessness rebound phenomena
Methyldopa IV: 250–200 mg q 6 h up to 4 12–24 h Postoperative Pheochromocytoma, Drowsiness,
h hypertension hepatic disease, MAO hypotension, fever,
inhibitors positive Coombs’ test,
chronic hepatitis,
lupuslike syndrome
Converting Enalaprilat IV: titrated 0.625–1.25 10–15 6–24 h Hypertension Pregnancy, renal failure, Cough and
enzyme mg q 6–8 h min related to renal bilateral renal stenosis angioedema, fever
inhibitors artery stenosis urticaria hypotension,
pancytopenia
Dopaminergic Fenoldopam IV: 0.1 µg kg−1 min−1 4 min 20 min Postoperative Glaucoma Increased intraoccular
stimulation titrated up at 0.1 µg hypertension, pressure, headache,
hypertensive crisis flushing, nausea
kg–1 increments q 20 asymptomatic EKG
min Maximum dose STcnanges
1.7 µg kg–1 min–1
*
Not FDA approved. CAD, coronary artery disease; CHF, congestive heart failure; MAO, monoamine oxidase; NG, nasogastric.
Page 136
Table 9.6 Considerations for specific Hypertensive situations
Clinical situations Treatment goals Preferred agents

Acute aortic dissection Reduce pressure gradient across lesion; reduce inotropy Nitroprusside plus beta blocker; or labetalol
Acute left ventricular dysfunction or CHF Afterload reduction while preserving inotropy. Consider milrinone or ACE inhibitor, nitroprusside or nitroglycerin Beta
dobutamine to provide both simultaneously blockers may paradoxically improve output by
slowing rate
Catecholamineinduced crisis (MAO inhibition, Block alpha and beta simultaneously (see separate entry for Labetalol, nitroprusside+beta blocker, phentolamine
cocaine overdose, clonidine or opiate pheochromocytoma)
withdrawal)
Eclampsia and severe preeclampsia Avoid nitroprusside, trimethaphan (fetal CNS toxicity); enalaprilat Magnesium sulfate if seizures, hydralazine,
and nimodipine relatively contraindicated labetalol, calcium channel blockers
Hypertensive encephalopathy Avoid cerebral vasodilation with intracranial bleeding Nicardipine
Intracranial hemorrhage, subarachnoid Goal: MAP 80% of initial value, keep DBP<140 Avoid sedative Labetalol; calcium channel blockers, esp.
hemorrhage, Increased ICP effects (methyldopa, clonidine) Treat underlying causes of increased nimodipine
intracranial pressure
Ischemic stroke Not clear if benefit to intervention Treat only if BP>180/105 Sodium nitroprusside Nicardipine Nimodipine
Myocardial ischemia Reduce preload and heart rate; control pulmonary hypertension Nitroglycerin, beta blockers and morphine
Perioperative hypertension Avoid stress on fresh suture lines Avoid hypotension (risk of labetalol, nitroprusside Sedatives or analgesics
thrombosis) Consider possible physical causes and treat pain
appropriately
Pheochromocytoma Unopposed beta blockade can precipitate crisis; block alpha first or Preoperative alpha blockade: phenoxybenzamine;
simultaneously; add beta blockade to prevent tachycardia acute treatment: phentolamine or Iabetalol
Postoperative cardiac surgery Avoid vasodilators with subaortic stenosis or tight aortic stenosis Labetalol, esmolol, nitroprusside, nicardipine,
nitroglycerine
Thyroid storm Control tachycardia while dealing with underlying hormone excess Propylthiouracil (PTU) beta blockers
combines with cyanide to produce nontoxic cyanocobalamin. Hemodialysis can also eliminate thiocyanate.
Because of concerns about tachyphylaxis and toxicity, other antihypertensive agents should be added while acute hypertension is
Page 137
controlled with nitroprusside. Diuretics potentiate nitroprusside’s effects by helping to eliminate pooled venous blood. Beta blockers counteract reflex tachycardia.
When using nitroprusside for medical management of acute aortic dissection, beta blockade will also decrease shear forces across the dissection by reducing the
pressure gradient between the heart and dilated peripheral vasculature.
Nitroglycerin relaxes smooth muscle in venules and arterioles. It is not a particularly potent antihypertensive in normal circumstances, but is often effective with acute
volume overload. Reduction of venous return is a major cause of decreased blood pressure with nitroglycerin. Side effects associated with nitroglycerin include
headache owing to meningeal vessel dilation, and methemoglobinemia, because nitroglycerin oxidizes hemoglobin. Nitroglycerin can be given sublingually (0.15–0.6 mg
per dose), transdermally (2% ointment) or as a sustainedaction patch, but the preferred method of administration in the ICU is IV infusion, typically a mix of 50 mg in
250 cm3 of D5W initiated at 10 µg min−1 and titrated to patient response.
Hydralazine (Apresoline) is a directacting vasodilator with an onset of action of 3–20 min following IV administration, and a duration of action of 2–6 h. It is the
drug of choice in eclampsia as it preserves myometrial blood flow. Hydralazine is relatively contraindicated as the sole antihypertensive agent in patients with severe
coronary disease, because of the potential for precipitating reflex tachycardia. In addition, hydralazine has been reported to cause a lupuslike rash, fluid retention and
headache. Typical IV doses of hydralazine are 5–10 mg q 4–6 h individualized to the patient’s response. Oral doses are 10–50 mg p.o. q 6 h.
Fenoldopam is a vasodilator that acts by selective stimulation of dopamine DA1 receptors. The drug has no effect at betaadrenergic or dopaminergic DA2
receptors. Fenoldopam can be safely given via a peripheral (as opposed to central) IV line. Its onset of action is rapid (within 4 min) and duration of action short. In
contrast to nitroprusside, it induces diuresis and natriuresis14. Other advantages include lack of lightinduced loss of potency, and absence of toxic metabolites15.
Fenoldopam and nitroprusside both cause rapid and significant reduction in systolic blood pressure and systemic vascular resistance. In postoperative cardiac surgery
patients, fenoldopam also results in increased cardiac index and stroke volume16.
Other direct vasodilators include minoxidil and diazoxide. Minoxidil therapy can be initiated with 2.5–5 mg orally, and dose doubled every 6 h until blood pressure
control is obtained. The effective dose can then be continued daily or b.i.d. Because of minoxidil’s potent vasodilating effects, sympathetic nervous activation is likely
and the drug’s vasodilating effect may be antagonized by endogenous release of norepinephrine and angiotensin II. Fluid retention is a complication of minoxidil therapy
requiring addition of a thiazide and/or loop diuretic.
Diazoxide is a direct arteriolar dilator with little effect on venous capacitance. Diazoxide has a long serum halflife (c. 30 h), which is further increased in patients with
impaired renal function. Because the drug is bound to albumin, it must be given as a rapid IV bolus, typically 50–150 mg of diazoxide over 30 s. Diazoxide also causes
sympathetic stimulation and fluid retention so its effects will be potentiated by sympatholytic agents and diuretics. Undesirable side effects of diazoxide include reduction
in cerebral blood flow17, increased blood glucose levels as a result of an inhibition of insulin release, and hyperuricemia. In hypertensive emergencies, the maximum
effect of diazoxide is reached in 5 min and dissipates over the next 12 h.
9.5.2 Calcium channel blockers

Calcium channel antagonists reduce muscle contractility and thus inotropy and systemic vascular resistance by limiting calcium influx though slow channels in
myocardium and smooth muscle cells. Nicardipine (Cardene) is a shortacting calcium channel antagonist. The initial rate of infusion is 5 mg h−1 up to a
Page 138
maximum dose of 15 mg h–1. Nicardipine's onset of action is 1–5 min and its duration of action is 3–6 h. Because of its coronary vasodilatory effect, nicardipine is
particularly useful for blood pressure control with concurrent myocardial ischemia or infarction. It has the potential side effects of tachycardia, nausea, and myocardial
depression, and is contraindicated in patients with tight aortic stenosis and with hypersensitivity reactions. Nicardipine is effective as sodium nitroprusside following
coronary bypass surgery, achieves a therapeutic response sooner, and requires less frequent dosage adjustment during maintenance therapy, with fewer episodes of
severe hypotension compared with sodium nitroprusside18.
Nifedipine (Adalat, Procardia) is a potent vasodilator and afterload reducing agent with mild inotrophic and chronotropic effects. A sublingual dose of 10–20 mg is
effective in the management of acute hypertension19, but more recent reports20 recommend against this practice, because of the potential for overshoot and
hypotension. Nicardipine is preferable to nifedipine because of a slower and more prolonged decrease in blood pressure. Relative contraindications to nifedipine
include hypersensitivity, severe congestive heart failure, tight aortic stenosis, and renal insufficiency. Side effects include hypotension, tachycardia, syncope, platelet
dysfunction, hepatotoxicity, and headache. In patients receiving digoxin, nifedipine increases serum digoxin levels.
Isradipine (Dynacirc) is a calcium channel antagonist similar to nifedipine that has been evaluated for sublingual use. At low doses, it produces vasodilation without
affecting myocardial contractility. With normal ventricular function, isradipine's overloadreducing properties frequently lead to an increase in cardiac output. Blood
pressure reduction occurs some 2–3 h following an oral dose of 2.5–20 mg, and the duration of action is 12 h or more, allowing twicedaily administration. Sublingual
doses of 1.25–5 mg have an onset in 30 min and a duration of 2 h21. Intravenous isradipine via infusion (mean dose 18.6 µg min−1) is effective and well tolerated in
patients developing hypertension post coronary artery bypass grafting (CABG)22.
Diltiazem (Cardizem) and Verapamil (Calan) relax smooth muscle and thus have antihypertensive effects, but have relatively greater effects on cardiac conduction,
particularly in the setting of combined therapy with betablocking agents or digitalis. Nimodipine has been shown to improve outcome in patients with aneurysmal
subarachnoid hemorrhage23.
9.5.3 Adrenergic blocking agents

Esmolol (Brevibloc) is a shortacting cardioselective beta blocker indicated for control of supraventricular tachycardia. Esmolol has no reported antihypertensive effects
in the absence of tachycardia or high catecholamine activity and is not recommended for first line treatment of hypertensive crisis. Numerous reports, however, attest to
its utility when hypertension is associated with a hyperdynamic state. It has an onset of action of 3–5 min with a duration of action of 10–20 min. Esmolol is metabolized
by red blood cell esterase and has an extremely short beta halflife (around 9 min). Doserelated hypotension occurs in 20–50% of patients when esmolol is given for
heart rate control. The initial infusion dose of esmolol is 500 µg kg−1 over 1 min, followed by 50 µg kg–1 min–l, and then titrated in the range of 50–200 µg kg−1 min−1
to the desired endpoint. Additional loading doses of 100 µg kg–1 over 1 min may be necessary. Although cardioselective, esmolol is still relatively contraindicated in
patients with asthma, severe diabetes, highdegree heart block, sinus bradycardia, heart failure, and pregnancy. Like other betablocking agents, esmolol reduces
systolic blood pressure by decreasing cardiac output. In severe hypertensive episodes where the elevated blood pressure results from high systemic vascular resistance,
a decrease in cardiac output may significantly compromise organ perfusion, so the drug should be avoided when cardiac output is borderline or inadequate. This danger
is somewhat offset by the very short halflife of esmolol. Betaadrenergic
Page 139
blockers with longer halflives include metoprolol, atenolol, and propranolol.

Labetalol (Normodyne, Trandate) is a combined alpha and betaadrenergic blocker, with more beta than alpha activity (a 7:1 relative activity ratio is commonly
quoted). It is particularly useful in the management of acute hypertension, particularly that caused by catecholamine excess. Full effects occur within 5–10 min of IV
injection. Because of the betaadrenergic blocking effect, labetalol is unlikely to precipitate tachycardia when given for severe hypertension. Labetalol’s metabolism is
primarily hepatic, with less than 5% excreted unchanged via the kidneys. Bolus doses of 5–10 mg IV may be repeated or doubled at 10 min intervals until a maximum
dose of 80 mg is reached. However, much higher doses (>500 mg) have been necessary in some patients24. It may also be administered as an infusion of 0.5–2 mg
min−1. Because of its betablocking properties, labetalol is relatively contraindicated in patients with congestive heart failure or bronchospastic disease. In patients with
pheochromocytomas, paradoxic responses have occurred because of the relatively greater betablocking potential. Labetalol may also block the tachycardia associated
with hypoglycemia and therefore, as with all betaadrenergic blocking agents, should be used cautiously in patients receiving IV insulin or oral hypoglycemic agents.
Side effects of labetalol include bradycardia, heart block, anaphylaxis, hypotension, and dizziness.
9.5.4 Alphaadrenergic agents (Table 9.7)

An understanding of the subtypes of the alpha receptor clarifies the role of selective and nonselective agents in blood pressure control (Figure 9.4). Alpha1 receptors,
located on the postsynaptic membrane, mediate vasoconstriction. Blockade of these alpha1 receptors
Figure 9.4 Sympathetic nerve terminal. NE, norepinephrine.
Table 9.7 Alphaadrenergic agents
Agonists Antagonists
alpha2 Guanfacine Yohimbine
Guanabenz
Clonidine
αMethylnorepinephrine
alpha1 and 2 Epinephrine Tolazoline
Phentolamine
Norepinephrine Phenoxybenzamine (irreversible binding)
alpha1 Phenylephrine Labetalol [alpha and beta]
Prazosin
Methoxamine Doxazosin
Terazosin
Page 140
is an effective treatment for hypertension. Alpha2 receptors, located on the presynaptic membrane, modulate the release of norepinephrine across the synaptic cleft.
Stimulation of alpha2 receptors thus reduces sympathetic stimulation and consequently blood pressure. Pressor agents such as epinephrine and norepinephrine
stimulate both forms of receptors (see Chapter 8) whereas nonselective alphaadrenergic blockers such as phentolamine and phenoxybenzamine block both types of
receptors. More selective agents can be categorized by their relative binding infinity at the alpha1 and alpha2 receptor (Table 9.5). Simulation of central alpha2
receptors diminishes sympathetic activity and drugs with alpha2 selectivity tend to be sedating.
Clonidine (Catapres) is a centrally acting alpha2receptor agonist that modulates tonic and reflex blood pressure control, and lessens sympathetic outflow25,26,
resulting in lowered heart rate and blood pressure. Clonidine is administered in a dose of 0.1–0.2 mg intravenously and may be also administered orally via a
nasogastric tube (0.1–0.3 mg twice daily) or by transdermal patch (3.5 cm2 or 0.1 mg day−1). The IV form has an onset of action of 30–60 min with a duration of
action of 6–8 h. Doses higher than 0.4 mg have little additional benefit. Clonidine is contraindicated in patients with bradycardia, sick sinus syndrome, or heart block.
The dose should be reduced in elderly patients and in patients with preexisting cerebral or cardiovascular disease, because of exaggerated bradycardiac responses and
potential for drowsiness and sedation. Clonidine is particularly useful in hypertension associated with substance withdrawal and with uncontrolled sympathetic
discharges in neurosurgical patients. Clonidine can significantly decrease cerebral blood flow and one death from progressive cerebral infarction has been reported27.
Dexmedetomidine (Precedex) is primarily marketed as a sedative agent. Like clonidine, dexmedetomidine acts at the alpha2 adrenergic receptor, and enhanced
sedation is noted when dexmedetomidine is coadministered with inhalation anesthetics, propofol, opioids, and benzodiazepines. Although hypotension is commonly
seen during sedative treatment with dexmedetomidine, hypertension may also be seen, particularly with rapid bolus administration.
Alphamethyldopa is a converted to an active metabolite that, like clonidine, acts as a potent alpha2 receptor agonist. Because of the requirement for a metabolite
to be formed, the onset of action may be delayed (4 h) and the action prolonged (up to 24 h). Alphamethyldopa tends to cause sedation, impair mental judgment, and
depress overall behavior. Roughly 25% of patients maintained on longterm methyldopa develop a positive direct Coombs test. In high doses, all of the alpha2 agonists
can stimulate peripheral vasoconstriction and cause hypertension. Thus, methyldopa should not be combined with clonidine, guanabenz, and guanfacine because of
additive effects.
Phentolamine is an alphaadrenergic blocker historically used in situations involving excess catecholamine levels such as pheochromocytoma and antihypertensive
withdrawal syndrome. Its use has largely been supplanted by labetalol or the combination of nitroprusside plus a beta blocker. Phentolamine may cause tachycardia,
angina, vomiting, and headache.
9.5.5 Angiotensinconverting enzyme (ACE) inhibitors

Enalaprilat is the IV form of enalapril with an onset of action in 10–15 min and a duration of action of 6–24 h. It is administered as a titrated dose of 0.625–1.25 mg
every 6 h. Enalaprilat is indicated in the management of hypertension when oral ACE therapy is impractical, and in particular when the hypertension is related to
unilateral renal artery stenosis. It is contraindicated in pregnancy, with a history of angioedema related to ACE inhibitors, and in renal failure as a result of bilateral renal
artery stenosis or unilateral stenosis of a solitary kidney. Enalaprilat has been reported to provide a more favorable profile than nitroprusside even in patients with no
demonstrable evidence of
Page 141
elevated plasma renin activity, suggesting that the mechanism for blood pressure control involves something more than the reninangiotensin system28.
A number of oral ACE inhibitors are available for clinical use, including captopril, enalopril, lisinopril, fosinopril, benazapril, and ramipril, the last three being prodrugs
that must be converted to their active form by a functioning liver. Pressure reduction with the ACE inhibitors may initially be dramatic in patients with high renin levels,
although subsequent doses tend to cause less hypotension. Volumedepleted patients typically manifest an exaggerated response. In patients with congestive heart
failure, cardiac output should increase with afterload reduction29. Side effects of the ACE inhibitors include excessive hypotension, angioedema, impaired renal function,
hypercalemia, and cough. In addition to ACE inhibitors, angiotensin receptor antagonists such as Iosartan are becoming available30 but clinical experience in the ICU
setting is sparse.
9.5.6 Adjunctive agents

In the critically ill patient, diuretics are generally avoided for the acute treatment of hypertension unless blood volume is elevated. Rapidly acting IV diuretics such as
furosemide (Lasix) or bumetanide (Bumex) are useful when there is overt congestive heart failure or pulmonary edema. Diuretics are generally needed with longterm
therapy to counteract renal sodium retention.
Sedative agents such as opioids, benzodiazepines, and propofol reduce blood pressure by controlling central stimulation, and may be useful adjuncts in blood
pressure control, particularly in mechanically ventilated or agitated patients.
9.5.7 Ganglionic blockers

Trimethapham is a ganglionic blocking agent historically used for the treatment of hypertension in the setting of acute aortic dissection. Trimethapham reduces left
ventricular ejection fraction and heart rate, but with the advent of betablocking drugs and combined alphabetablocking agents such as labetalol, trimethopram is now
rarely used. Side effects include urinary retention, paralytic ileus, midriasis, dry mouth, and development of tachyphylaxis.
9.6 Specific clinical considerations

An understanding of the pathophysiology in specific clinical conditions can guide the choice of agents (Table 9.6).
9.6.1 Acute aortic dissection

Medical management of acute aortic dissection involves rapid reduction of blood pressure to normal levels using IV sodium introprusside. The force and velocity of
ventricular contractions and pulsatile flow affect the shear stress on the aortic wall. Unless contraindicated, a betablocking agent should be added to decrease
myocardial contractility and the pressure on the aortic wall. If beta blockers are contraindicated, trimethaphan camsilate or urapidil can be substituted.
9.6.2 Excess circulatory catecholamines

Pheochromocytomas are the most common of the catecholaminesecreting tumors. Hypertension associated with these tumors may be silent or paradoxic in nature in
up to 40% of patients31. Postoperative hypertension may be the initial clinical presentation of these tumors32.
A hallmark feature of patients with these tumors is relative volume contraction. Hypotension is common following pheochromocytomas resection because of the
removal of the continuous source of catecholamines in the presence of the adrenergic blockade given for preoperative preparation. Such hypotension may necessitate
the administration of fluids to expand the intravascular volume as well as cautious administration of vasopressor agents. Excess circulatory catecholamines can also be
seen in rebound hypertension following clonidine withdrawal, following ingestion of street drugs (cocaine, amphetamines,
Page 142
phencyclidine, LSD and weight loss drugs) and as a result of the interaction of MAO inhibitors with dietary tyramine (cheese, wine, certain beers). Betablocking
agents should always be preceded by alpha blockade to avoid unopposed peripheral vasoconstriction33.
9.6.3 Autonomic hyperreflexia

Quadraplegic or paraplegic patients with spinal cord injuries above the inhibitory splanchnic outflow level (T4 to T7) have a potential for acute sympathetic
hyperactivity in response to stimuli below the level of the transsection. This generalized sympathetic activity results from reflex stimulation of the sympathetic neurons in
the anterolateral column of the spinal cord below the transection level without the inhibitory modulation that comes from the higher centers in the central nervous system.
This disorder occurs in 65–85% of quadraplegic and highlevel paraplegic patients, often in response to bladder catheterization34.
Hypertension in these patients is related to profound increases in systemic vascular resistance resulting from the sympathetically mediated vasoconstriction.
Autonomic hyperreflexia is commonly associated with a reflex bradycardia mediated by the baroreceptor reflex. In addition, there may be a second reflex arc involving
skeletal muscles that produce intense muscle spasm. Treatment of the hypertensive response in autonomic hyperreflexia is directed at reversing elevated systemic
vascular resistance with sodium nitroprusside or ganglionic blocking agents such as trimethaphan34,35.
9.6.4 Increased intracranial pressure

Marked elevation of intracranial pressure results in a reflex sympathetic response with peripheral vasoconstriction and hypertension. Through the baroreceptor
mediated response, this profound hypertension results in an accompanying reflex bradycardia. This combination of physiologic events is commonly referred to as the
Cushing reflex. In addition to spaceoccupying lesions, other causes of increased intracranial pressure include intracranial bleeding, cerebral edema, hypercapnia, and
hypoxemia. Management is directed at the underlying mechanism responsible for the intracranial hypertension and may warrant immediate surgical intervention.
Intracerebral pressure may be acutely lowered by reducing cerebral blood flow with deliberate hypocapnia via hyperventilation, and administering pharmacologic
agents that reduce blood pressure without causing an associated vasodilation. Initially, mannitol which induces an osmotic diuresis, may increase intracranial pressure as
a result of its effect on volume expansion. Pretreatment with furosemide helps prevent this effect. Pharmacologic blockade of alpha1 or beta1adrenergic receptors
reduces blood pressure with little effect on intracranial pressure, within the autoregulatory range36. Urapidil also blocks postsynaptic alpha1 receptors without
increasing cerebral blood flow. In the neurosurgical patient, early extubation should be considered, when possible, to avoid the stimulant effects of awakening while
intubated37.
Intracerebral hemorrhage causes a rise in intracranial pressure and thus higher mean arterial pressure for adequate perfusion. The resulting hypertension typically
resolves spontaneously within 48 h, and although rapid reduction in blood pressure may prevent further bleeding, there is a risk of cerebral hypoperfusion. Patients with
MAP lower than 145 mmHg and those whose MAP is controlled below 125 mmHg have a better outcome by retrospective analysis38. In these circumstances, blood
pressure should be reduced only to 80% of the pretreatment blood pressure level39.
9 6.5 Acute ischemic stroke

Autoregulation of cerebral blood flow is impaired following cerebral infarction, resulting in excess perfusion through damaged tissue, edema, and compression of normal
brain. On the other hand, local vasoconstriction may warrant a higher arterial pressure to perfuse jeopardized brain tissue in the periinfarction
Page 143
zone. There are no controlled outcome studies suggesting that shortterm lowering of blood pressure improves outcome in patients with acute ischemic stroke. In
patients with diffuse stenosis of cerebral arteries, neurologic improvement may be seen with blood pressure elevation40. The benefits of acute blood pressure control in
patients with acute stroke are questionable41.
9.6.6 Thyrotoxicosis or thyroid storm

The stress of critical illness may contribute to the release of high levels of T4 hormone from the thyroid gland into the circulation in patients with undiscovered
hyperthyroidism. Thyroid storm is often associated with elevated levels of the T4 hormone in comparison to the more frequent elevations of T3 seen in thyrotoxicosis42.
Patients with thyroid storm present in the perioperative setting with the signs of increased sympathetic tone and are commonly febrile, hypertensive, tachycardic,
agitated, and confused. Increased oxygen demands placed on the heart may precipitate high output failure or uncover underlying ischemic heart disease, resulting in the
acute onset of congestive heart failure. Treatment includes the management of the patient’s hyperthermia, the administration of betaadrenergic blockers and inhibitors of
the synthesis and release of thyroid hormones (propylthiouracil, methimazole, and sodium iodide). In severe circumstances, active removal of the thyroid hormone from
the circulation is indicated via plasmaphoresis or peritoneal dialysis43,44.
9.6.7 Eclampsia and preeclampsia

Eclampsia and preeclampsia occur in approximately 10% of all late pregnancies and cause up to 20% of all maternal deaths45. These disorders are related to an
immunologic reaction to the fetal tissue, resulting in placental vasculitis and subsequent ischemia, with release of uterine renin. Hydralazine is favored for control of blood
pressure owing to minimal effects on uterine perfusion. Labetalol and nifedipine are acceptable alternatives46 but diuretics, trimethaphan, nitroprusside and ACE
inhibitors should be avoided. The hypertensive response usually abates postpartum, but persistent postpartum hypertension can occur as a result of increased sensitivity
to endogenous or exogenous oxytocin and catecholamines. Oxytocin should be used sparingly, if at all, in the postpartum period to avoid prolonging the hypertension.
9.6.8 Acute renal ischemia

Acute renal ischemia can occur from the occlusion of renal arteries during major vascular surgery or following renal transplantation. The lack of blood flow and resulting
ischemia of the juxtaglomerular apparatus increases renin secretion and subsequent conversion of angiotensin I to angiotensin II. If there is significant myocardial
dysfunction and/or coronary artery disease, the increase in systemic afterload may be lifethreatening. Hypertension following renal transplantation has been associated
with stenosis of the renal artery anastomosis, rejection reactions, immunosuppressive therapy, and excessive renin release by the diseased kidneys47.
Immunosuppressive therapy with steroids increases the arteriolarsystemic vascular resistance to endogenous catecholamines along with excess renin released from the
diseased kidneys. Cyclosporine has been associated with an increase of sympathetic discharge47. Deterioration of renal function in the presence of elevated blood
pressure requires immediate intervention to reduce systemic vascular resistance without compromising renal blood flow. Labetalol, calcium channel blockers,
fenoldopam and nitroprusside with dosage adjustment are reasonable choices.
9.6.9 Volume overload

Between dialysis treatments, patients with chronic renal failure become hypervolemic as a result of inability to handle normal sodium and water excretion. These patients
frequently have an exaggerated response to vasopressor therapy or volume replacement, owing to the chronically elevated renin levels and enhanced
Page 144
vascular responsiveness. Hypertension can be managed with vasodilators or ACE inhibitors temporarily until dialysis or ultrafiltration can be provided. Severe
hypertension can precipitate acute left ventricular failure. Decreasing afterload will decrease the workload of the failing myocardium and improve cardiac function.
9.7 Medicationrelated hypertension

Ketamine causes stimulation of the sympathetic autonomic nervous system and inhibits the uptake of catecholamines by postganglionic neurons. Neuroleptics, such as
droperidol and haloperidol, block dopaminemediated peripheral vasodilation in the setting of sympathetic stimulation, vasopressor administration, or a catecholamine
secreting tumor. Acute clonidine withdrawal manifests with hypertension, starting within 18 h and lasting up to 3 days unless appropriate therapy, such as
transcutaneous clonidine, is reinstituted. Narcotic withdrawal may be a result of either a patient’s declining opioid blood level or the administration of narcotic antagonist
such as naloxone. Use of opioid agonistantagonists, such as pentazocine or nalbuphine, may also precipitate narcotic withdrawal and associated hypertension. Alcohol
or sedative withdrawal may occur as a result of either decreased chronic level of use or abrupt withdrawal as may occur in the critical care setting. Symptoms of
withdrawal from these agents include central nervous system excitation, diaphoresis, autonomic hyperactivity with associated hypertension, and tachycardia. Treatment
with beta blockers or clonidine can reduce the profound sympathetic stimulation associated with withdrawal syndromes.
Abrupt preoperative discontinuation of beta blockers has been associated with pronounced rebound tachycardia and hypertension48. Such rebound episodes may
be associated with the development of myocardial ischemia.
Monamine oxidase inhibitors are used as antidepressant agents because of their ability to block the oxidative deamination of catecholamine to vanillylmandelic acid.
Inhibition of monamine oxidase causes an accumulation of norepinephrine, dopamine, epinephrine, and 5hydroxytryptamine (5HT) in sympathetic tissues. Use of
indirectacting sympathomimetic agents, such as ephedrine, or the narcotic meperidine has been associated with exaggerated hypertensive responses secondary to the
triggered release of the catecholamines that have accumulated in the nerve terminals49.
9.8 Conclusions
Although hypertensive emergencies are rare, they require immediate treatment to limit the threat to the cardiovascular system and to prevent endorgan injury.
Aggressive IV therapy in an intensive care setting is indicated. At the same time, longeracting oral agents can be started to limit the amount of time the patient must be
on titratable IV drugs in a monitored setting. Postoperative hypertension, although not generally associated with endorgan damage, does increase stress on fresh suture
lines and is frequently treated with IV agents. Hypertensive urgencies, manifested by elevated blood pressure without endorgan damage, may be treated with oral or
IV agents, although outcome data showing the benefits of acute blood pressure reduction are not available. An accepted approach in these patients is to lower the
blood pressure more gradually over a 24–48 h period and to avoid precipitous declines in blood pressure.
References
1. Seltzer, J.L., Gerson, J.I. and Grogano, A.W. (1980) Hypertension in the perioperative period. N.Y. State J. Med. 80:29.
2. Gal, T.J. and Cooperman, L.H. (1975) Hypertension in the immediate postoperative period. Br. J. Anaesth. 47:70.
3. Estafanous, F.G., Tarazi, R.C., Viljoen, J.F. and Tawil, M.Y. (1973) Systemic hypertension following myocardial revascularization. Am. Heart J. 85:732.
Page 145
4. Kreiger, E.M. (1970) Time course of baroreceptor resetting in acute hypertension. Am. J. Physiol 8:486.
5. Robertson, D., Hollister, A.S., Biaggioni, L, Netterville, J.L., MosquedaGarcia, R. and Robertson, R.M. (1993) The diagnosis and treatment of baroreflex
failure. N. Engl. J. Med. 329:1449–1455.
6. Mancia, G., Lorenz, R.R. and Shepherd, J.T. (1976) Reflex control of circulation by heart and lungs. Int Rev. Physiol. 9:111.
7. Malliani, A. (1982) Cardiovascular sympathetic afferent fibers. Rev. Physiol. Biochem. Pharmacol. 194:10.
8. Mark, A.L. and Mancia, G. (1983) Cardiopulmonary baroreflexes in humans. In: Handbook of Physiology, Volume 3 (eds J.T.Shepherd and F.M.Abboud).
American Physiology Society, Bethesda, MD, p. 397.
9. Brenner, B.M. and Laragh, J.H. (1988) Advances in Atrial Peptide Research. Raven Press, New York.
10. Guyton, A.C., Hall, J.E. (1996) Dominant role of the kidneys in long term regulation of arterial pressure and in hypertension: The integrated system for pressure
control. In: Textbook of Medical Physiology, 9th Edn. W.B.Saunders, Philadelphia, p. 221–237.
11. Bazaral, M.G., Welch, M., Golding, L.A.R. and Badhwar, K. (1990) Comparison of brachial and radial arterial pressure monitoring in patients undergoing
coronary artery bypass surgery. Anesthesiology 73:38–45.
12. Bedford, R.F. (1977) Radial arterial function following percutaneous cannulation with 18 and 20gauge catheters. Anesthesiology 47:37–39.
13. Cole, P. (1978) The safe use of sodium nitroprusside. Anesthesia 33:473.
14. Post, J.B. and Frishman, W.H. (1998) Fenoldopam: A new dopamine agonist for the treatment of hypertensive urgencies and emergencies. J. Clin. Pharmacol.
38:2–13.
15. Brogden, R.N. and Markham, A. (1997) Fenoldopam: A review of its pharmacodynamic and pharmacokinetic properties and intravenous clinical potential in the
management of hypertensive urgencies and emergencies. Drugs 54(4):634–650.
16. Hill, A.J., Feneck, R.O. and Walesby, R.K. (1993) A comparison of fenoldopam and nitroprusside in the control of hypertension following coronary artery surgery.
J. Cardiothorac. Vasc. Anes. 7:279–284.
17. Pearson, R.M., Griffith, D.N.W., Wallard, M. and James, I.M. (1979) Comparison of effects of cerebral blood flow of rapid reduction in systemic arterial
pressure by diazoxide and lobetilol in hypertensive patients. Br. J. Clin. Pharm. 8:155S–198S.
18. Halpern, N.A., Goldberg, M., Neely, C., et al. (1992) Postoperative hypertension: A multicenter, prospective, randomized comparison between intravenous
nicardipine and sodium nitroprusside. Crit. Care Med. 20(12):1637–1643.
19. McAllister, R.G. (1986) Kinetics and dynamics of nifedipine after oral and subliminal doses. Am. J. Med. 81(Suppl. 6A):2–5.
20. Grossman, E., Messerli, F.H., Grodzicki, T., et al. (1996) Should a moratorium be placed on sublingual nifedipine capsules given for hypertensive emergencies and
pseudoemergencies? JAMA 276:1328–1331.
21. Saragoca, M.A., Portela, J.E., Plavnik, F., et al. (1992) Isradipine in the treatment of hypertensive crisis in ambulatory patients. J. Cardiovasc. Pharmacol. 19
(Suppl. 3):S76–S78.
22. Leslie, J., Brister, N., Levy, J.H., et al. (1994) Treatment of postoperative hypertension after coronary artery bypass surgery: Doubleblind comparison of
intravenous israpidine and sodium nitroprusside. Circulation 90(part 2): II256–II261.
23. Wong, M.C.W. and Haley, E.C., Jr (1990) Calcium antagonists: stroke therapy coming of age. Stroke 21:494–501.
24. Atkin, S.H., Jaher, M.A., Beaty, P., et al. (1992) Oral labetalol versus oral clonidine in the emerging treatment of severe hypertension. Am. J. Med. Sci. 303:9–15.
25. Ghigrione, M., Calvillo, O. and Quintin, L. (1987) Anesthesia and hypertension: the effect of clonidine on perioperative hemodynamics and isoflurane
requirements. Anesthesiology 67: 3.
26. Flacke, J.W., Bloor, B.C., Flacke, W.E., et al. (1987) Reduced narcotic requirement by clonidine with improved hemodynamic and adrenergic stability in patients
undergoing coronary artery bypass surgery. Anesthesiology 67:11.
27. Spitalewitz, S., Porush, J.G. and Oguayha, C. (1983) Use of oral clonidine for rapid titration of blood pressure in severe hypertension. Chest 83(Suppl.):404–407.
28. Mirenda, J., Edwards, C. (1991) Use of intravenous enalaprilat for hypertension following myocardial revascularization. Anesth. Analg. 72: 667.
29. Turini, G.A., Bribic, M., Drunner, H.R., Waeber, B. and Gavras, H. (1979) Improvement of chronic congestive heart failure by oral captopril. Lancet 1:1213–
1215.
30. Goodfriend, T.L., Elliott, M.E. and Catt, K.J. (1996) Angiotensin receptors and their antagonists. N. Engl J. Med. 334:1649–1654.
Page 146
31. Cryer, P.E. (1987) Diseases of the sympathochromaffin system. In: Endocrinology and Metabolism, 2nd Edn (eds Felig, P., Baxter, J.D., Broadus, A.E. and
Frohman, L.A.). McGrawHill, New York, p. 651–692.
32. Roizen, M.E, Horrigan, R.W., Koike, M., et al. (1981) A prospective randomized trial of four anesthetic techniques for resection of pheochromocytomas.
Anesthesiology 57:A43.
33. Calhoun, D.A. and Opanl, S. (1990) Treatment of hypertensive crisis. N. Engl J. Med. 323: 1177–1183.
34. Abin, M.S. (1990) Anesthesia for spinal cord injury. Probl. Anesth. 4:138–143.
35. Schonwald, G., Fish, KJ. and Perkaash, I. (1981) Cardiovascular complications during anesthesia in chronic spinal cord injured patients. Anesthesiology 55:550.
36. Tietjen, C.S., Hurn, P.D., Ulatowski, J.A. and Kirsch, J.R. (1996) Treatment modalities for hypertensive patients with intracranial pathology: Options and risks.
Crit. Care Med. 24: 311–322.
37. Bruder, N. and Ravussin, R. (1999) Recovery from anesthesia and postoperative extubation of neurosurgical patients: A review. J. Neurosurg. Anesth. 11:282–
293.
38. Dandapani, B.K., Suzuki, S., Kelly, R.E., et al. (1995) Relation between blood pressure and outcome in intracerebral hemorrhage. Stroke 26: 21–24.
39. Lavin, P. (1986) Management of hypertension in patients with acute stroke. Arch. Intern. Med. 146:66–68.
40. Rordorf, G., Cramer, S.C., Efird, J.T., et al. (1997) Pharmacologic elevation of blood pressure in acute stroke: Clinical effects and safety. Stroke 28:2133–2138.
41. Powers, W.J. (1993) Acute hypertension after stroke. The scientific basis for treatment decisions. Neurology 43:461–467.
42. Stehling, L.C. (1974) Anesthetic management of the patient with hyperthroidism. Anesthesiology 4l:585.
43. Bennett, M.H. and Wainwright, A.P. (1989) Acute thyroid crisis on induction of anesthesia. Anaesthesia 44:28.
44. Lombardi, A., Chiovato, L. and Braverman, L.E. (1991) Thyroid storm. In: Intensive Care Medicine (eds J.M.Rippe, R.S.Irwin, J.S.Alpert and M.P.Fink). Little,
Brown, Boston, MA, p. 976.
45. Assche, F., Spitz, B. and Vansteelant, L. (1989) Severe systemic hypertension during pregnancy Am. J. Cardiol. 63:22C.
46. Rey, E., LeLorier, J., Burgess, E., et al. (1997) Report of the Canadian Hypertension Society Consensus Conference: 3. Pharmacologic treatment of hypertensive
disorders in pregnancy. Can. Med. Assoc. J. 157:1245–1254.
47. Stanek, B., Kovarik, J., RasoulRockenschaub, S. and Silberbauer, K. (1987) Reninangiotensinaldosterone system and vasopressor in cyclosporine treated
renal allograft recipients. Clin. Nephrol. 8:186–189.
48. Miller, R.R. (1975) Propranalol withdrawal rebound phenomenon: exacerbation of coronary events after abrupt cessation of antianginal therapy. N. Engl. J. Med.
293:416.
49. Stack, C.G., Rogers, P. and Linter, S.P. (1988) Monamine oxidase inhibitors and anaesthesia. Br. J. Anaesth. 60:222.
Page 147
Chapter 10
Acute myocardial infarction
Marc Schweiger, MD
Contents
10.1 Overview
10.2 Diagnosis, definitions and triage
10.3 Initial treatment of acute ischemic syndrome
10.4 Thrombolytic therapy
10.5 Primary angioplasty
10.6 Lipidlowering therapy
10.7 Complications of myocardial infarction
10.8 Subsequent treatment strategies
10.9 Summary
Page 148
10.1 Overview
Each year, in the United States approximately one million people sustain an acute myocardial infarction (AMI) with about onequarter being fatal1. Almost all of these
patients have underlying atherosclerosis with superimposed thrombus. The exception are patients with conditions in which myocardial oxygen demand exceeds the
available supply (i.e. severe anemia, thyrotoxicosis). Although the incidence of AMI and its mortality have been decreasing, AMI remains the most common cause of
death in Western society. The reason for the decreased incidence of acute myocardial infarction is presumably multifactorial. Awareness of risk factors for
atherosclerosis, which include hypertension, cigarette smoking, diabetes mellitus, hyperlipidemia and a family history of premature coronary disease, has led to more
aggressive treatment and prevention. Newer medical treatments and use of more aggressive forms of interventional therapy (percutaneous coronary intervention,
coronary bypass surgery) may be partially responsible.
Previously, myocardial infarctions were categorized as either Q wave or nonQ wave. However, with our present understanding of the pathophysiology, the more
appropriate terminology is ST elevation MI (STEMI—Q wave or transmural MI) and nonST elevation MI (NSTEMI—nonQ wave or nontransmural MI).
Although many of the treatment modalities are appropriate for both types of MI, certain treatments are appropriate for STEMI but inappropriate for NSTEMI and vice
versa.
The acute coronary syndromes are a continuum from unstable angina to NSTEMI to STEMI. Fissure or rupture of the atherosclerotic plaque exposes the
subendothelial matrix to substances that activate platelets and generate thrombin2. If the thrombus occludes the lumen (with platelets, thrombin and red blood cells), an
STEMI occurs unless the distal territory of the blood vessel has adequate collateral blood flow. Patients with a nonocclusive thrombus typically have more plateletrich
thrombi and are more likely to present with either ST segment depression or T wave inversion. Elevation of cardiac enzymes defines patients with ST depression or T
wave inversion who have an NSTEMI (Figure 10.1).
10.2 Diagnosis, definitions and triage

The history, physical examination, electrocardiogram and laboratory tests are the tools used to make the diagnosis and appropriately triage patients. The definition of a
myocardial infarction requires the presence of symptoms suggestive of ischemia and either EKG evidence or enzyme evidence. EKG evidence requires the
development of Q waves in two or more leads to retrospectively establish a diagnosis of STEMI3.
10.2.1 Classic history

The classic history of MI discomfort involves a discomfort usually in the chest, often associated with diaphoresis, which may radiate into the arms or back (see Table
10.1). The discomfort may be localized to only the arms, back or jaw or may radiate to these areas. The description of the discomfort is usually a pressure,
Table 10.1 Differential diagnosis of chest discomfort
Cardiac Noncardiac
• Angina • Esophageal disorders
• Acute MI
• Aortic dissection • Peptic ulcer
• Pericarditis • Costochondritis
• Myocarditis • Cervical disk
• Mitral valve prolapse • Hyperventilation
• Anxiety
• Pulmonary hypertension
• Pneumonia
Page 149
Figure 10.1 This shows the classification of patients who present with prolonged ischemic discomfort. Patients with ST segment
elevation or a new left bundle branch block on their eltctrocardiogram will will have, in the presence of+enzymes, an ST
elevation myocardial infarction (also known as a Q Wave Myocardial Infarction or Transmural Myocardial Infarction). Patients
without EKG changes or with ST depression or T wave inversion will have a non STelevation myocardial infarction (also
known as a NonQ Wave Myocardial Infarction or Subendocardial Myocardial Infarction) if their cardiac enzymes are +, and
unstable angina if their cardiac enzymes are negative.
burning or constricting sensation. It may be described as indigestion. Often patients will answer ‘no’ to the question of whether there is any pain but will answer in the
affirmative to the question of whether there is discomfort. There is often associated nausea and vomiting. There are, however, many variations of presenting symptoms
and some patients may not have any discomfort at all. In the intensive care unit setting patients may develop myocardial ischemia while intubated, sedated and
paralyzed. Residual anesthesia or analgesics may lessen or eliminate the discomfort. Questions regarding any similar symptoms recently, previous cardiac history and
presence of risk factors help in assessing the likelihood of ischemic pain when the presentation is not classic. Discomforts that last seconds are rarely ischemic. The
differential diagnosis of chest discomfort is shown in Table 10.1.
10.2.2 Electrocardiogram
The electrocardiogram of patients with acute ST segment elevation MI will show new ST segment elevation (>1 mm in at least two contiguous leads) (Figure 10.2) or
a new left bundle branch block. Patients presenting with a posterior wall myocardial infarction may present with a normal EKG. The electrocardiogram of patients with
unstable angina or nonST elevation MI will usually show ST segment depression or T wave inversion (Figure 10.3). The majority of patients presenting with ST
segment elevation will evolve a Q wave myocardial infarction but some will develop a nonQ wave myocardial infarction. Patients presenting with ischemic discomfort
at rest without ST segment elevation are likely to ultimately be diagnosed as unstable angina or nonQ wave myocardial infarction; a minority will develop a Q wave
myocardial infarction.
Page 150
Figure 10.2 EKG showing changes of an inferior wall ST elevation myocardial infarction. Significant (>3 mm) ST elevation is noted in
leads II, III and aVF.
10.2.3 Laboratory evidence

Laboratory evidence, particularly cardiac enzymes, is useful for diagnosis and prognosis. For patients presenting with rest chest pain without ST segment elevation,
positive cardiac enzymes confirm the diagnosis of myocardial infarction. Troponin I and T are sensitive and specific markers for myocardial necrosis but do not elevate
until 4–6 h following necrosis. Thus a negative early troponin requires an additional troponin 8–12 h after symptom onset. Troponins may remain elevated for up to 2
weeks and thus are useful in diagnosing recent MI. A significant number of patients presenting with ischemic pain without ST segment elevation are diagnosed as an
NSTEMI because their troponin is positive whereas their Creatine PhosphokinaseMyocardial Band (CKMB) is negative. In patients who present with unstable
angina/nonST segment elevation MI, troponins are also useful for risk stratification4, as a positive troponin predicts recurrent myocardial infarction and mortality.5 CK
MB, an isoenzyme of CK, although widely used, is not specific for myocardial necrosis, as some CKMB is present in skeletal muscle. The diaphragm and esophagus
are also sources of CKMB, which may elevate following thoracic surgical procedures (see Chapter 24). CKMB often does not elevate until more than 6 h after
myocardial necrosis. The timing of the peak CK is useful in assessing reperfusion in ST elevation MI (earlier peak with successful reperfusion). Myoglobin is a specific
but not sensitive marker of myocardial necrosis. It rises rapidly (>2 h) and returns to normal less than 24 h after myocardial necrosis. A negative myoglobin, 4–8 h
following symptoms, in a patient with a nondiagnostic EKG effectively rules out myocardial necrosis. Previously utilized laboratory tests, such as serum glutamic
oxaloacetic acid transferase (SGOT) and lactic dehydrogenase (LDH), have limited utility in the era of newer cardiac markers.
Page 151
Figure 10.3 EKG showing significant ST segment depression in leads V2 and V3.
10.2.4 Triage decision

When evaluating patients with symptoms of acute coronary syndrome, the initial triage decision is determined by an EKG, which should be performed within 10 min of
arrival. An EKG showing ST segment elevation or a new or presumably new left bundle branch block leads to an evaluation for reperfusion therapy. An
electrocardiogram suspicious for ischemia, or ongoing ischemic pain without electrocardiographic changes, should prompt admission to a monitored bed. Should the
pain appear classic and unrelenting, in the absence of EKG changes, the diagnosis of a posterior wall MI should be entertained and an echocardiogram may be useful in
this situation (posterior wall akinesis in the absence of a previous MI history). The diagnosis of a myocardial infarction requires the presence of elevated cardiac
enzymes. The electrocardiogram will distinguish a nonST elevation MI from an ST elevation MI and the enzymes will distinguish a nonST elevation MI from unstable
angina.
10.3 Initial treatment of acute ischemic syndrome
10.3.1 Routine requirements

All patients with acute ischemic syndromes should undergo continuous EKG monitoring for arrhythmia detection1,6. Intravenous access must be established. Oxygen
supplementation should be routine airway for O2 saturation of 95% or more.
10.3.2 Nitroglycerine
All patients with acute ischemic chest pain should receive one sublingual nitroglycerine (generally 1/150 grain) unless the systolic blood pressure is less than 90 mmHg.
Nitro glycerine should not be used in patients with right ventricular MI because these patients are extremely volume dependent and may become hypotensive with
vasodilation. Only one sublingual nitroglycerine should routinely be given to patients with ST segment elevation MI. Intravenous nitroglycerine is recommended for
patients with acute STEMI with persistent hypertension or congestive heart failure. In patients with STEMI, IV nitroglycerine should not be used in lieu of adequate
analgesia. In patients with NSTEMI, sublingual NTG can be given 5 min apart×3 followed by an IV nitroglycerine infusion for continued ischemic discomfort.
10.3.3 Morphine sulfate

This should be used to treat ischemic pain that does not respond promptly to sublingual nitroglycerine in patients with ST segment elevation MI. Morphine serves as an
anxiolytic and analgesic. It also decreases arterial pressure and venous return. Morphine should be given in 2–4 mg increments every 5 min until the pain is controlled
and the patient less anxious. The use of analgesia may make it difficult to assess pain relief in patients given thrombolytic therapy, but its use should not be
Page 152
withheld. Side effects of morphine include hypotension, nausea and respiratory depression. Although these must be considered, adequate analgesia is a cornerstone of
initial MI treatment. Patients experiencing an NSTEMI (or unstable angina—enzymes not yet back) should also receive IV morphine for pain not responding to nitrates,
pulmonary congestion, or significant agitation. In the ICU setting where respiration is monitored and often controlled, fentanyl 50–100 µg IV is less likely to cause
hypotension, as fentanyl does not cause histamine release.
10.3.4 Oral antiplatelet therapy

Aspirin has been shown to reduce mortality in patients with STEMI7 when used alone or in combination with streptokinase. It also has been shown to decrease
reocclusion and recurrent ischemic events following thrombolytic therapy. Aspirin works by inhibition of platelet activation via its inhibition of thromboxane A2. Aspirin
has also been shown to decrease risk in patients with unstable angina/nonST segment elevation MI8. It is recommended that patients presenting with acute ischemic
syndrome who have not been taking aspirin promptly receive a chewable aspirin, dose of 160 mg or 325 mg1,6. Patients can then be maintained on 75, 160 or 325 mg
of aspirin daily. Patients who have an allergy to aspirin may be given either ticlopidine (Ticlid) or clopidogrel (Plavix), which inhibits ADPmediated platelet activation.
Clopidogrel has fewer reported side effects than ticlopidine. Clopidogrel is usually given in a dose of 75 mg day−1; however, this dosage has delayed efficacy and a
loading dose of 300 mg is preferred in acute ischemic syndrome for aspirin intolerant patients.
10.3.5 Beta adrenergicblocker therapy

For patients with STEMI, intravenous followed by oral beta blocker therapy reduces mortality in patients not receiving thrombolytic therapy9, and reduces reinfarction
and recurrent ischemia in patients receiving thrombolytic therapy10. Intravenous beta blockers are indicated in patients with STEMI treated within 12 h, followed by
oral therapy, unless there are contraindications such as bradycardia (heart rate <60), systolic pressure <100 mmHg, bronchospastic lung disease, second or
thirddegree heart block, active congestive heart failure, or prolonged PR interval (>0.24 s). Patients are generally given 5 mg of metoprolol (lopressor) intravenously
every 5 min for three doses provided their blood pressure and heart rate are within parameters. They are then generally started on 25 mg of metoprolol orally, which is
increased as necessary to achieve an appropriate heart rate (and blood pressure) response. Betaadrenergic blockers with intrinsic sympathomimetic activity (i.e.
Pindolol) should be avoided. For patients admitted with acute ischemic syndrome without ST segment elevation, intravenous beta blockers should be given, in the
absence of contraindications, if ongoing chest pain is present, followed by oral administration. Table 10.2 shows properties of commonly used betaadrenergic blockers.
10.3.6 Calcium channel blocker therapy

The role of calcium channel antagonism in acute myocardial infarction has been somewhat controversial. It is not first line therapy in patients with STEMI or NSTEMI.
In patients with myocardial infarction, the use of nonhydropyridine calcium antagonists, verapamil or diltiazem, may be appropriate for patients, with ongoing ischemia
or rapid ventricular response to atrial fibrillation, who are intolerant of beta blockers secondary to bronchospastic lung disease. These agents should not be used with
ongoing congestive heart failure or significant heart block (second or thirddegree heart block or firstdegree heart block with PR interval >0.24 s). Diltiazem
(cardizem) may be added as standard therapy to patients with NSTEMI after 24 h and continued for 1 year in the absence of LV dysfunction or congestive heart
failure11. Immediaterelease dihydropyridine calcium channel blockers should never be given in the absence of a betaadrenergic blocker but can in certain instances be
given to
Page 153
Table 10.2 Betaadrenergic receptor antagonists
Agent Selectivity Agonist activity Dose (usual)

Acebutolol (Sectral) B1 Partial 200–600 mg twice daily
Atenolol (Tenormin) B1 No 50–200 mg daily
Metoprolol (Lopressor) B1 No 25–100 mg twice daily
Bisoprolol (Zebeta) B1 No 10 mg daily

Betaxolol (Kerlone) B1 No 10–20 mg daily
Esmolol (Breviblock) B1 No IV 50–300 mg kg−1 min−1

Propranalol (Inderal) None No 10–40 mg 4 times daily
Nadalol (Corgard) None No 40–240 mg daily
Timolol (Blocadren) None No 10 mg twice daily
Labetalol (Normodyne) Combined alpha and nonselective B No 200–600 mg daily
Pindolol (Visken) None Yes 10–30 mg 3 times daily
patients with ongoing ischemia postinfarction who are receiving a beta blocker.
10.3.7 Angiotensinconverting enzyme inhibitors (ACEI)

These have been shown to reduce mortality in patients with a recent myocardial infarction. The benefit is most pronounced in patients with left ventricular systolic
dysfunction12. Recent evidence suggests a benefit for patients with vascular disease and/or diabetes greater than 55 years of age13. Therefore in the absence of
contraindications, ACEI therapy may be initiated in patients with acute myocardial infarction (STEMI or NSTEMI) early in their hospital course, particularly when there
is known or suspected left ventricular dysfunction or continued hypertension.
10.3.8 Parenteral anticoagulants

These include unfractionated heparin, low molecular weight heparin, hirudin, and bivalirudin. Hirudin and bivalirudin are direct thrombin inhibitors. Bivalirudin has been
approved for use in percutaneous coronary intervention. Hirudin is utilized in patients with heparininduced thrombocytopenia. Unfractionated heparin or low molecular
weight heparin, when added to aspirin, reduces the risk of recurrent MI or mortality in patients with unstable angina/NSTEMI by approximately 50%14. Unfractionated
heparin inactivates thrombin through its action on antithrombin III and has a lesser effect on factor Xa and IXa. Unfractionated heparin does have limitations related to
its nonspecific binding to proteins, which leads to variable anticoagulation requiring frequent activated partial thromboplastin time (APTT) measurements and dose
adjustments. Low molecular weight heparin has a more predictable anticoagulant response and a longer plasma halflife than unfractionated heparin, and is therefore
easier to use than unfractionated heparin. Low molecular weight heparin has less of an effect on thrombin than unfractionated heparin, as the cross bridging between
antithrombin and thrombin requires pentasaccharide containing chains of at least 18 saccharide units. Heparininduced thrombocytopenia appears to be somewhat less
common with low molecular weight heparin than unfractionated heparin.
Multiple clinical trials suggest that low molecular weight heparin is at least as effective as unfractionated heparin in the treatment of unstable angina/NSTEMI.
Although some trials
Page 154
have suggested a clinical benefit of low molecular weight heparin over unfractionated heparin, other trials have revealed no benefit14. Trials comparing the various low
molecular weight heparins have not been performed. The safety and efficacy of low molecular weight heparins in the catheterization laboratory and in concert with
glycoprotein Ilb/IIIa inhibition has not been adequately studied.
Patients with unstable angina/NSTEMI should receive either IV unfractionated heparin or subcutaneous low molecular weight heparin (Table 10.3). Unfractionated
heparin is usually given as a weightadjusted regimen with an initial bolus of 60–70 U kg−1 (maximum 5000 U) with an initial infusion of 12–15 U kg−1 h−1 (maximum
1000 U). Each institution should establish its own nomogram to achieve target aPTT (1.5–2.5 times control). Measurements should be made 6 h after a dosage change
and used to adjust heparin dosing until the aPTT is therapeutic. When two consecutive aPTT results are therapeutic the next measurement can be made in 24 h unless a
change in clinical condition ensues. Daily hemoglobin, hematocrit and platelet count measurements are suggested with patients on heparin. Most trials of unfractionated
heparin in unstable angina/nonST segment elevation myocardial infarction continued therapy for 2–5 days (it is the current practice to discontinue heparin after a
percutaneous coronary intervention is performed). Mild thrombocytopenia occurs in 10–20% of patients receiving unfractionated heparin; severe thrombocytopenia
(platelet count <100000) occurs in 1–2% and usually occurs late (4–14 days after starting therapy). Fewer than 0.2% of patients experience heparininduced
thrombocytopenia with thrombosis6. Patients who develop heparininduced thrombocytopenia should be given hirudin, a direct acting thrombin inhibitor, given as a bolus
of 0.4 mg kg−1 IV over 15–20 s followed by an IV infusion of 0.15 mg kg−1 h−1 with a target aPTT of 1.5–2.5 times control.
Patients presenting with STEMI who are undergoing reperfusion therapy with alteplase or tenectoplase should receive a 60 U kg−1 bolus of heparin followed by a
maintenance infusion of 12 U kg−1 h−1 (maximum 4000 U bolus and 1000 U h−1 infusion for patients weighing >70 kg), adjusting the aPTT to 1.5–2.0 times control for
24–48 h. Most physicians stop heparin after a successful percutaneous coronary intervention. Heparin can be continued for patients at high risk for systemic or venous
thromboembolism.
Oral anticoagulation with warfarin is at present recommended only for patients with established indications such as atrial fibrillation, mechanical heart valves or left
ventricular thrombus. Recently presented studies suggest a benefit for longterm highdose coumadin in preventing late vessel occlusion but this needs confirmation in
larger trials.
10.3.9 Glycoprotein IIb/IIIa antagonists

The glycoprotein Ilb/IIIa receptor is the final common pathway for platelet aggregation. Platelet agonists activate the platelet and induce a conformational change in the
Ilb/IIIa receptor that increases its affinity to bind fibrinogen and von Willebrands factor. There are at present three intravenous glycoprotein Ilb/IIIa receptor antagonists
available for clinical use. The monoclonal antibody abciximab (ReoPro) has been extensively studied and has been clearly shown to reduce clinical events following
percutaneous coronary intervention. Its efficacy, however, in unstable angina/ NSTEMI has not been established. The small molecule glycoprotein Ilb/IIIa antagonists,
eptifibatide (integrelin) and tirofiban (aggrastat), have been shown in clinical trials to reduce recurrent events (predominantly recurrent myocardial infarction) in patients
presenting with unstable angina/nonQ wave myocardial infarction15,16. It is recommended that these agents be administered, in addition to aspirin and unfractionated
heparin, in patients with ongoing ischemia, highrisk features, or in whom a percutaneous coronary intervention is planned. Highrisk features include positive serum
markers, transient ST elevation or depression with chest pain, more than 20 min of rest pain, accelerating pain
Page 155
over the previous 24–48 h, age over 75 years, ischemiainduced congestive heart failure or mitral regurgitation, hypotension or sustained ventricular tachycardia6
(Table 10.3).
10.4 Thrombolytic therapy

Reperfusion therapy with a thrombolytic agent is utilized for the treatment of acute STEMI (or new or presumably new left bundle branch block) in the presence of
chest pain less than 12 h after onset of the most severe symptoms. Thrombolytic therapy is contraindicated in patients with NSTEMI. Thrombolytic therapy restores
patency in the infarct artery and therefore reduces mortality and improves left ventricular function. It has the greatest benefit when given early after symptom onset. It is
estimated that for every 1 h delay in administering thrombolytic therapy there is a loss of approximately 20 lives for every 1000 patients treated. Therefore it is
important to educate patients to come to hospital promptly with symptoms of acute myocardial infarction, and it is incumbent on hospital emergency rooms to have a
short ‘doortoneedle’ time.
Thrombolytic therapy has the advantage of universal hospital availability and thus, unlike primary angioplasty, can be offered expeditiously at all hospitals. There are,
however, both absolute and relative contraindications to patient treatment with thrombolytic therapy. Absolute contraindications for thrombolytic therapy include: active
internal bleeding, known intracranial neoplasm, and previous intracranial bleed. Relative contraindications include: elevated systolic blood pressure over 180 mmHg,
diastolic blood pressure over 110 mmHg (note: blood pressure is the first reliable blood pressure taken after arrival to the emergency room), previous cerebrovascular
event at any time, active peptic ulcer, pregnancy, noncompressible vascular puncture, prolonged (>10 min) CPR, recent (<3 weeks) major surgery, recent trauma (<4
weeks), recent internal bleeding (<4 weeks). Relative contraindications will vary depending on the availability and speed to alternative therapy (i.e. primary angioplasty)
as well as the perceived risk to the patient of their myocardial infarction (i.e. speedy access to primary angioplasty would increase the threshold to utilize thrombolytic
therapy in a patient with a relative contraindication; apparent large myocardial infarction early after symptom onset in a situation where transport for primary angioplasty
is not
Table 10.3 Nonparental antithrombotic therapy
Intravenous antiplatelet Dose

therapy
Eptifibatide (Integrilin) 180 µg kg−1 bolus followed by 2 µg kg−1 min−1 infusion
Tirofiban (Aggrastat) 0.4 µg kg−1 min−1 bolus over 30 min followed by 0.1 µg kg−1 min−1 infusion
Unfractionated heparin UA/NSTEMI: 60–70 U kg−1 (max. 5000 U) bolus followed by 12–15 U kg−1 h−1 infusion (titrated to aPTT 1.5–2.5 times control)
STEMI—with thrombolytic tenectepase (TNK): 60 U kg−1 bolus, infusion of 12 U kg−1 h−1 (max. 4000 U bolus, 1000 U h−1 infusion if
weight >70 kg), adjusted to aPTT between 1.5 and 2 times control
Lovenox (Enoxaparin) 1 mg kg−1 subcutaneously every 12 h
Dalteparin (Fragmin) 120 IU kg−1 subcutaneously every 12 h (maximum 10000 IU twice daily)
Page 156
feasible would decrease the threshold to utilize thrombolytic therapy).

Table 10.4 shows characteristics of currently available thrombolytic agents. Retaplase and tenectaplase are derived from tissue plasminogen activator. Retaplase is a
deletion mutant, tenectaplase a substitution mutant of the tPA molecule. With currently available agents, optimal flow (TIMI (thrombolysis in myocardial infarction) 3
flow) is established at 90 min in 35–60% of patients. Intracranial hemorrhage occurs in 0.5–1% of patients depending on the agent and patient population. Trials
comparing the mortality for the various thrombolytic agents have been performed; alteplase was shown to decrease mortality compared with streptokinase17, retaplase
and streptokinase were shown to be equivalent18, retaplase was not shown to be superior to alteplase19, and tenectaplase and alteplase were shown to be equivalent
with slightly less bleeding in the tenectaplasetreated patients20.
Recently, preliminary trials of halfdose thrombolytic and glycoprotein Ilb/IIIa inhibition therapy have been shown to achieve reperfusion more rapidly and improve
TIMI 3 flow rates without an increase in bleeding21. At present, there are numerous ongoing trials involving combinations of thrombolytic therapy with all of the
currently available glycoprotein Ilb/IIIa inhibitors.
10.5 Primary angioplasty

Primary angioplasty has been compared with thrombolytic therapy in many small randomized clinical trials. A metaanalysis suggests a mortality benefit for patients
treated with primary angioplasty22. The benefit of primary angioplasty appeared to attenuate over 6 months in the GUSTO III trial23. This may be related to restenosis
or reocclusion of the dilated infarct artery causing lesion and may be obviated with the use of coronary stents and perhaps glycoprotein Ilb/IIIa inhibitors. Primary
angioplasty is not routinely available at many hospitals. The benefit of primary angioplasty is probably obtained when patients can be taken to the catheterization
laboratory quickly (doortoballoon time <90 min); however, the average doortoballoon time in registry studies appears to be approximately 120 min24. This is of
concern, as during this ‘delay’ there is unlikely to be improvement in flow. Future studies are likely to compare primary angioplasty with combined lowdose
thrombolytic and glycoprotein Ilb/IIIa inhibition therapy (with or without early coronary angiography and percutaneous intervention). Primary angioplasty does appear
to be superior in patients less than 75 years old presenting with cardiogenic shock25. Primary angioplasty is also indicated in patients with
Table 10.4 Comparison of agents
Characteristics Streptokinase Alteplase (tPA) Retaplase (rPA) Tenectaplase (TNKtPA)

Immunogenicity Yes No No No
Fibrin specificity + ++ + +++
Plasma halflife ? 4–6 min 18 min 20 min
90 min TIMI 3 30–35% 45–60% 50–60% 50–60%
flow
Cost + +++ +++ +++
Dose 1.5×106 over 60 15 mg in to 50 mg 30 min, up to 35 mg over next 10+10 MU; double bolus 30 min 0.5 mg kg−1 IV bolus (not to exceed 50
min 60 min apart mg)
Page 157
contraindications to thrombolytic therapy. Patients with prior coronary artery bypass surgery may also benefit from primary angioplasty, as the benefit of thrombolytic
therapy in this subset of patients is not well established. However, there may also be a great deal of difficulty in performing angioplasty on acutely occluded saphenous
vein grafts. The use of stents in primary angioplasty has been shown to decrease the need for repeat procedure secondary to restenosis26.
10.6 Lipidlowering therapy

Guidelines have been established for the treatment of hypercholesterolemia in patients with known coronary artery disease. It is at present recommended that patients
be treated to obtain an LDL cholesterol less than 100 mg dL−1 27. The mortality benefit seen in previous clinical trials begins to appear shortly after the initiation of
therapy, suggesting a role for acutely treating patients with acute coronary syndromes to enhance plaque stabilization. The recently presented MIRACL trial28
suggested a benefit at 4 months for patients with acute coronary syndrome treated immediately with aggressive lipidlowering therapy.
10.7 Complications of myocardial infarction
10.7.1 Right ventricular infarction

This frequently accompanies inferior myocardial infarction. There is a characteristic hemodynamic pattern of elevated right sided filling pressures. Left ventricular filling
pressures are normal or only minimally elevated. Cardiac output may be depressed. An increase in right atrial pressure with inspiration (Kussmal’s sign) may be
present. Patients presenting with an inferior wall myocardial infarction with signs of decreased cardiac output (hypotension) who have elevated right atrial pressure
should be suspected of having a right ventricular infarction. The EKG will usually show an inferior wall MI and right precordial leads will show ST segment elevation.
Coronary angiography will generally show involvement of the proximal right coronary artery, although rarely there will be circumflex involvement. Firstline therapy
consists of aggressive fluid resuscitation guided by central hemodynamic monitoring. Intravenous fluid is administered until an increase in left ventricular filling pressure
occurs, as indicated by an increase in pulmonary capillary wedge pressure. When fluid resuscitation fails to improve forward flow, inotropic stimulation is required (see
Chapter 8). Dobutamine (usually 4–10 µg kg−1 min−1) is administered to augment cardiac output. Besides providing inotropic support, dobutamine decreases
pulmonary vascular resistance and reduces right ventricular afterload, thereby facilitating flow from the right heart to the left ventricle.
Right ventricular MI may be associated with highdegree atrioventricular (AV) conduction abnormalities; these can result in various forms of heart block, including
bradycardia and loss of AV synchrony. Management typically consists of temporary right ventricular pacing until the cardiac conduction system recovers function.
10.7.2 Free wall rupture

This usually leads to hemopericardium and death secondary to cardiac tamponade. It occurs most frequently 1–4 days following a transmural myocardial infarction, and
is more common in elderly patients, hypertensive patients and patients who have received corticosteroids. It usually involves the anterior and lateral walls of the left
ventricle and most commonly occurs at the junction of infarcted and normal muscle. Presenting symptoms include jugular venous distention, hypotension, paradoxical
pulse and distant heart sounds. The presentation is generally catastrophic with sudden death from pericardial tamponade. Occasionally the tear may be subacute with
the patient complaining of pain (generally pleuritic) with hypotension and often dramatic ST segment elevation on the
Page 158
electrocardiogram. Prompt recognition leading to surgical repair is mandatory for survival.
10.7.3 Intraventricular septal rupture

This usually occurs 1–4 days following a transmural myocardial infarction. It is more common in elderly patients and patients with hypertension. The size of the septal
defect determines the degree of shunting (left to right) and consequently survival. Anterior wall infarction is associated with apical septal rupture whereas inferior wall
infarction is associated with a basal septal rupture and has a higher mortality secondary to a more complicated surgical procedure. Intraventricular septal rupture
presents with a new loud systolic murmur that generally is holosystolic and can be heard maximally at the left lower sternal border or the apex. Patients frequently are
hypotensive and have signs and symptoms of congestive heart failure. Twodimensional echocardiography will be diagnostic, as will placement of a pulmonary artery
balloon catheter with oximetry. A stepup in oxygenation saturation from the right atrium to the pulmonary artery is considered a definitive sign. The treatment is
surgical. Stabilization with the use of an intraaortic balloon pump is usually necessary.
10.7.4 Rupture of papillary muscle

This usually occurs 1–4 days following a transmural myocardial infarction. Inferior wall infarction leads to rupture of the posteriomedial papillary muscle, whereas
patients with anterior wall infarction can have ruptures of the anterolateral papillary muscle. The presentation is similar to the rupture of an interventricular septum and
can be difficult to distinguish clinically. Twodimensional echocardiography with color flow Doppler imaging is useful in distinguishing these two entities.
10.8 Subsequent treatment strategies

The use of coronary angiography will differ by clinical presentation.
10.8.1 Unstable angina/NSTEMI

For unstable angina/NSTEMI, the necessity of a routine invasive strategy following admission has been evaluated in four clinical trials. The TIMI 3 trial, which evaluated
patients with unstable angina or nonQ wave myocardial infarction, demonstrated no mortality/reinfarction difference between a routine invasive strategy carried out
within 48 h versus a conservative strategy with catheterization for breakthrough ischemia or a strongly positive exercise test29. In this trial many patients in the
‘conservative’ group ultimately underwent angiography and there was less rehospitalization in the invasive strategy group. The conclusion was that either strategy was
acceptable and in stable patients there was no necessity to transfer patients to a hospital with the facilities to perform coronary intervention. The Vanquish trial of
patients with nonQ wave myocardial infarction suggested a mortality benefit for patients randomized to a conservative approach30. In this trial, all of the excess
mortality was secondary to a high surgical mortality among patients in the invasive group undergoing coronary artery bypass surgery. The FRISC II trial randomized
patients to a low molecular weight heparin (Fragmin) or unfractionated heparin with a second randomization to an invasive or conservative strategy. Patients undergoing
an invasive procedure did so 4 days following hospital admission and only 10% of the conservative strategy patients underwent coronary angiography. A clinical benefit
was observed for patients in the invasive strategy group31.
The recently completed TACTICS TIMI 18 trial randomized patients with unstable angina and nonQ wave myocardial infarction to an invasive or selective invasive
strategy4. All patients received the glycoprotein Ilb/IIIa inhibitor Tirofiban. Invasive strategy patients underwent coronary angiography 4–48 h following admission.
Sixty per cent of the invasive patients underwent revascularization during their initial hospitalization (41% percutaneous coronary intervention, 19% coronary bypass
surgery). Fiftyone per cent of the
Page 159
‘selective invasive’ patients underwent coronary angiography during their initial hospitalization. By 6 months, 61% of the ‘invasive strategy’ patients had undergone
revascularization, whereas 44% of the ‘selective invasive’ patients also had revascularization. The primary combined endpoint of death, myocardial infarction and
rehospitalization at 6 months was significantly lower in the invasive strategy patients. This improvement was magnified among patients with positive troponin and patients
with an intermediate or high TIMI risk score32. This would suggest that in the present era of coronary intervention, including the use of stents and glycoprotein Ilb/IIIa
inhibitors, early coronary angiography is indicated for patients with highrisk features.
10.8.2 Coronary angiography for STEMI

This may be divided into three time periods: rescue angioplasty for those who have an occluded infarctrelated artery following thrombolytic therapy, early angioplasty
for patients who have a patent (TIMI flow 2 or 3) infarct artery (adjunctive angioplasty), and angiography later in the hospitalization following thrombolytic therapy with
possible delayed angioplasty.
The utility of rescue angioplasty has been demonstrated in one randomized trial of patients with anterior wall myocardial infarction, which showed a decrease in 30
day mortality or congestive heart failure33. Recent trials have shown improved outcomes in patients undergoing rescue angioplasty. It is likely that patients will benefit
from rescue angioplasty if it is carried out as quickly as possible in patients experiencing a large myocardial infarction. Patients should be assessed within 1 h following
thrombolytic therapy for signs of reperfusion (pain relief, resolution or marked improvement in ST segments, reperfusion arrhythmia). Should there be a clinical
suspicion of thrombolytic therapy failure, clinical consideration should be given to rapid transfer to a catheterization laboratory, particularly if the amount of myocardium
in jeopardy appears to be large and a catheterization laboratory is readily available.
Previous studies (TIMI 2A, TAMI (thrombolysis and angioplasty in myocardial infarction), European Cooperative Study Group, ECSG)34–36 performed in the mid
1980s suggested that patients undergoing immediate percutaneous coronary intervention in association with thrombolytic therapy for acute myocardial infarction were at
increased risk for adverse outcomes. In a recent study of patients undergoing early angiography, however, appropriate intervention following halfdose thrombolytic
therapy showed no hazard and a shorter time to optimal flow37. Observational data from TIMI 10B and TIMI 1438 also suggest excellent outcomes with reduced
hospital length of stay for patients undergoing immediate catheterization and intervention, when appropriate, with TIMI 2 or 3 flow. These improved outcomes may be
due, in part, to the availability of stents and better pharmacologic agents. Future trials will be needed to demonstrate the utility of an early invasive strategy among
patients treated with pharmacologic therapy for acute myocardial infarction.
The use of a routine invasive strategy a few days following thrombolytic therapy has not demonstrated an advantage over a more conservative strategy. Previous
studies were carried out in an era of less advanced interventional and surgical techniques and may not be relevant in the current era. It should be emphasized that
patients with recurrent ischemia postmyocardial infarction (either spontaneous or provoked) should undergo coronary angiography.
10.8.3 Stress testing

In patients with unstable angina (NSTEMI) who are admitted with a definite diagnosis of acute coronary syndrome, triage is by their clinical presentation6. As
previously mentioned, patients with factors associated with increased risk should be considered for early invasive strategy. Stress testing is indicated for those patients
who stabilize clinically, appear to be at low risk, and have an ejection fraction greater than 40%. Patients with a ‘lowrisk’
Page 160
stress test result can be managed with medical therapy.

Following thrombolytic therapy, patients are considered for a submaximal stress test beginning on day 4 to assess future risk and the need for coronary angiography.
Patients with a markedly abnormal test should undergo coronary angiography. Patients with a mildly abnormal test should either undergo an exercise imaging study to
further risk stratify or undergo coronary angiography. Patients who undergo a submaximal test that does not suggest an increased risk should undergo a symptom
limited test 3–6 weeks following their myocardial infarction.
10.8.4 Echocardiography
The assessment of patients with acute MI and unstable angina is useful when clinical and ECG findings are nondiagnostic. Transient segmental wall motion
abnormalities during episodes of chest discomfort are diagnostic of ischemia. A fixed segmental wall motion abnormality can represent an acute myocardial infarction, a
remote myocardial infarction, or chronic ischemia (hibernation)39.
Echocardiography can be used to assess risk. Left ventricular dysfunction, as assessed by echocardiography, predicts early and late complications. It can be used, at
the bedside, to demonstrate sequential changes (improvement or deterioration) in left ventricular function39.
Echocardiography can be used to evaluate complications of myocardial infarction. These complications include acute mitral regurgitation, ventricular septal rupture,
infarct expansion and LV remodeling, free wall rupture, intracardiac thrombus, RV infarction (occurring in approximately a third of patients with inferior wall MI) and
pericardial effusion39.
Stress or dobutamine echocardiography is a useful tool to assess risk following myocardial infarction, particularly in those patients in whom baseline abnormalities are
expected to compromise interpretation39.
10.8.5 Indications for hemodynamic monitoring (pulmonary artery catheter)

Insertion of a flow directed catheter into the pulmonary artery may be indicated when there is uncertainty regarding the hemodynamics of a patient following myocardial
infarction. Examples include persistent hypotension unresponsive to fluid administration, cardiogenic shock, and suspected mechanical complications such as ventricular
septal defect, mitral regurgitation and pericardial tamponade, although these mechanical complications can often be evaluated effectively by physical exam and
echocardiography.
10.8.6 Indications for intraaortic balloon counterpulsation

The intraaortic balloon derives its efficacy primarily by reducing left ventricular afterload with rapid balloon deflation in end diastole. In contrast to other methods of
afterload reduction, which generally lower arterial pressure, the sequential inflation and deflation of the balloon helps maintain coronary artery perfusion pressure and
thus perfusion of the myocardium. The intraaortic balloon pump is recommended for: patients with cardiogenic shock not quickly reversed with pharmacologic therapy,
as a stabilizing measure before angiography and prompt revascularization, in patients with acute mitral regurgitation or ventricular septal defect complicating MI as a
stabilizing therapy before operative repair, patients with refractory ischemia postMI as a bridge to angiography and revascularization, and in patients with recurrent
ventricular arrhythmias postMI thought secondary to ischemia or hemodynamic instability. It is probably beneficial in patients with hemodynamic instability, poor LV
function or persistent ischemia in the presence of a large myocardial infarction1.
10.8.7 Standby temporary pacemaker

There are situations postMI where transcutaneous patches need to be placed to provide standby pacing capability in patients who are felt unlikely to progress to high
grade AV
Page 161
block, and who do not require immediate pacing. As these systems are potentially painful they should not be used if there is a high likelihood that pacing will be
required. Transcutaneous pacing does not entail the complications and risks of intravenous pacing, particularly for patients who have recently received thrombolytic
therapy. These systems are intended to be prophylactic and temporary, and can be used in an emergency situa tion as a bridge to temporary transvenous pacing.
Patients who are felt to have a high probability of requiring pacing (>30% risk of AV block) should receive a temporary pacing wire. Appropriate situations include
those patients with asymptomatic hypotension with sinus bradycardia that is unresponsive or poorly responsive to medical (i.e. atropine) treatment and patients with
newly acquired or ageindeterminate left bundle or right bundle branch block with a left fascicular hemiblock. Patients with a new right bundle branch block with a first
degree heart block should also be considered for application of transcutaneous patches.
10.8.8 Indications for temporary transvenous pacing

These include asystole, symptomatic bradycardia (sinus bradycardia or type 1 seconddegree AV block, associated with hypotension and unresponsive to atropine),
bilateral bundle branch block, new or age indeterminate bifascicular block with firstdegree AV block, Mobitz type II seconddegree AV block, or thirddegree AV
block with a wide QRS (generally seen in patients with anterior wall myocardial infarction)1.
10.8.9 Permanent pacing

Permanent pacing after myocardial infarction is related to the degree and type of AV block present after the myocardial infarction. It is recommended for patients with
persistent seconddegree AV block in the HisPurkinje system with bilateral bundle branch block or thirddegree AV block within or below the HisPurkinje system
following AMI. Permanent pacing is also recommended in patients with transient second or thirddegree infranodal AV block and associated bundle branch block. If
the site of block is uncertain, an electrophysiological study may be necessary. Permanent pacing is also recommended in patients with persistent symptomatic secondor
thirddegree AV block40.
10.8.10 Indications and role for cardiac surgery

A complete description of the indications and role for cardiac surgery in the setting of AMI (nonQ or Q) or unstable angina is beyond the scope of this chapter. In
patients with NSTEMI/unstable angina, there are many anatomic, clinical and ventricular function variables that are considered in the decision regarding
revascularization following an acute ischemic event. Among patients who have sustained an STEMI, emergent or urgent cardiac surgery is indicated for the repair of
postinfarction papillary muscle rupture leading to severe mitral insufficiency, for the repair of a postinfarction ventricular septal defect and for patients with post
infarction ventricular aneurism who are encountering intractable ventricular arrhythmias or pump failure.
10.9 Summary
The treatment of acute coronary syndromes is an everevolving field with numerous advances in diagnostic evaluation, medical therapy and interventional (both catheter
based and surgical) approaches occurring on a yearly basis. The cardiology community has put a tremendous amount of emphasis on randomized clinical trials that are
constantly being conducted and routinely change our approach to caring for different patient subsets. This chapter is meant as an overview and a beginning level of
understanding for physicians caring for patients with acute myocardial infarction. Continued attention to the cardiology literature is necessary to update the knowledge
obtained from this work.
Page 162
References
1. Ryan, T.J., Anderson, J.L., Antman, E.M., et al. (1996) ACC/AHA guidelines for the management of patients with acute myocardial infarction. J. Am. Coll
Cardiol. 1328–1428.
2. Falk, E., Shah, P.K. and Fuster, V. (1994) Coronary plaque disruption. Circulation 92: 657.
3. TunstallPedoe, H., Kuulasmaa, K., Amouyel, P., et al. (1994) Myocardial infarction and coronary deaths in the World Health Organization MONICA Project.
Circulation 90:583–612.
4. Cannon, C.P., Weintraub, W.S., Demopoulos, L.A. et al. (2001) Comparison of Early Invasive and Conservative Strategies in Patients with Unstable Coronary
Syndromes treated with the Glycoprotein Ilb/IIIa Inhibitor Tirofiban. N. Engl. J. Med. 344:1879–87.
5. Antman, E.M., Tansaijevic, M.J., Thompson, B., et al. (1996) Cardiacspecific troponin I levels to predict the risk of mortality in patients with acute coronary
syndromes. N. Engl J. Med. 335:1342–1349.
6. Braunwald, E., Antman, E.M., Beasley, J.W., et al. (2000) ACC/AHA guidelines for the management of patients with unstable angina and nonSTsegment
elevation myocardial infarction: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on the
Management of Patients with Unstable Angina). J. Am. Coll. Cardiol. 36:970–1062.
7. ISIS 2 (Second International Study of Infarct Survival) Collaborative Group (1988) Randomised trial of intravenous streptokinase, oral aspirin, both, or neither
among 17,187 cases of suspected acute myocardial infarction: ISIS2. Lancet 2:349–360.
8. Antiplatelet Trialists’ Collaboration (1994) Collaborative overview of randomised trials of antiplatelet therapy, 1: prevention of death, myocardial infarction, and
stroke by prolonged antiplatelet therapy in various categories of patients. Br. Med. J. 308:81–106.
9. First International Study of Infarct Survival Collaborative Group (1986) Randomised trial of intravenous atenolol among 16,027 cases of suspected acute
myocardial infarction. ISIS1. Lancet 57–66.
10. The TIMI Study Group (1989) Comparison of invasive and conservative strategies after treatment with intravenous tissue plasminogen activator in acute myocardial
infarction: results of the thrombolysis in myocardial infarction (TIMI) phase II trial. N. Engl J. Med. 320: 618–627.
11. The Multicenter Diltizem Postinfarction Trial Research Group (1988) The effect of diltizem on mortality and reinfarction after myocardial infarction. N. Engl. J.
Med. 319:385–392.
12. Rutherford, J.D., Pfeffer, M.A., Moye, L.A., et al. (1994) Effects of captotril on ischemic events after myocardial infarction: results of the Survival and Ventricular
Enlargement trial: SAVE Investigators. Circulation 90:1731–1738.
13. Yusuf, S., Sleight, P., Pogue, J. et al. (2000) Effects of an angiotensinconvertingenzyme inhibitor, ramipril, on cardiovascular events in highrisk patients. The Heart
Outcomes Prevention Evaluation Study Investigators. N. Engl. J. Med. 342:145–153.
14. Eikelboom, J.W., Anand, S.S., Malmberg, K., Weitz, J.I., Ginsberg, J.S. and Yusuf, S. (2000) Unfractionated heparin and lowmolecularweight heparin in
acute coronary syndrome without ST elevation: a metaanalysis. Lancet 355:1936–1942.
15. The PURSUIT Trial Investigators (1998) Inhibition of platelet glycoprotein Ilb/IIIa with eptifibatide in patients with acute coronary syndromes. Platelet glycoprotein
Ilb/IIIa in unstable angina: receptor suppression using integrilin therapy. N. Engl. J. Med. 339: 436–443.
16. Platelet Receptor Inhibition in Ischemic Syndrome Management in Patients Limited by Unstable Signs and Symptoms (PRISMPLUS) Study
Investigators (1998) Inhibition of the platelet glycoprotein Ilb/IIIa receptor with tirofiban in unstable angina and nonQ wave myocardial infarction. N. Engl. J. Med.
338: 1488–1497.
17. The GUSTO Investigators (1993) An international randomized trial comparing four thrombolytic strategies for acute myocardial infarction. N. Engl. J. Med.
329:1616–1622.
18. International Joint Efficacy Comparison of Thrombolytics (1995) Randomised, doubleblind comparison of reteplase doublebolus administration with streptokinase
in acute myocardial infarction (INJECT): Trial to investigate equivalence. Lancet 346:329–336.
19. The GUSTO III Investigators (1997) An international, multicenter, randomized comparison of reteplase and tissue plasminogen activator for acute myocardial
infarction. N. Engl. J. Med. 337:1118–1123.
20. The Assessment of the Safety and Efficacy of a New Thrombolytic Investigators (1999) Singlebolus tenecteplase compared with frontloaded alteplase in acute
myocardial infarction: the ASSENT2 doubleblind randomised trial. Lancet 354:716–722.
Page 163
21. Antman, E.M., Giugliano, R.P., Ribson, C.M., et al. for the TIMI 14 Investigators (1999) Abciximab facilitates the rate and extent of thrombolysis: results of the
Thrombolysis in Myocardial Infarction (TIMI) 14 trial. Circulation 99:2720–2732.
22 Weaver, W.E., Simes, R.J., Betriu, A., et al. (1997) Comparison of primary coronary angioplasty and intravenous thrombolytic therapy for acute mycoardial
infarction: a quantitative review. JAMA 278:2093–2098.
23. GUSTO II Angioplasty Substudy Investigators (1997) An international randomized trial of 1138 patients comparing primary coronary angioplasty versus tissue
plasminogen activator for acute myocardial infarction. N. Engl J. Med. 336:1621–1628.
24. Cannon, C.P., Gibson, C.M., Lambrew, C.T., et al. (2000) Relationship of symptomonsettoballoon time and doortoballoon time with mortality in patients
undergoing angioplasty for acute myocardial infarction. JAMA 283:2941–2947.
25. Hochman, J.S., Sleeper, L.A., Webb, J.G., et al. (1999) Early revascularization in acute myocardial infarction complicated by cardiogenic shock. N. Engl. J. Med.
341:625–634.
26. Grines, C.L., Cox, D.A., Stone, G.W., et al. (1999) Coronary angioplasty with or without stent implantation for acute myocardial infarction. N. Engl J. Med.
341:1949–1956.
27. Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (1993) Summary of the second report of the national cholesterol
education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel II). JAMA 269: 3015.
28. MIRACL Study (2000) Presented at the annual scientific sessions of the American Heart Association, New Orleans, LA, November 2000.
29. The TIMI IIIB Investigators (1994) Effects of tissue plasminogen activator and a comparison of early invasive and conservative strategies in unstable angina and
nonQwave myocardial infarction: Results of the TIMI IIIB Trial. Circulation 89:1545–1556.
30. Boden, W.E., O’Rourke, R.A., Crawford, M.H., et al. (1998) Outcomes in patients with acute nonQwave myocardial infarction randomly assigned to an invasive
as compared with a conservative strategy. N. Engl J. Med. 338: 1185–1192.
31. Fragmin and Fast Revascularization during Instability in Coronary Artery Disease Investigators (1999) Invasive compared with noninvasive treatment in
unstable coronaryartery disease: FRISC II prospective randomised multicentre study. Lancet 354:708–715.
32. Antman, E.M., Cohen, M., Bernink, P.J.L., et al. (2000) The TIMI risk score for unstable angina/nonST elevation MI. A method for prognostication and
theraputic decision making. JAMA 284:835–842.
33. Ellis, S.G., Ribeiro da Silva, E., et al. for the RESCUE Investigators (1994) Randomized comparison of rescue PCI with conservative management of patients with
early failure of thrombolysis for acute anterior myocardial infarction. Circulation 90:2280–2284.
34. The TIMI Research Group (1988) Immediate vs delayed catheterization and PCI following thrombolytic therapy for acute myocardial infarction. TIMI IIA results.
JAMA 260: 2849–2858.
35. Topol, E.J., Califf, R.M., George, K.B.S., et al. and the Thrombolysis and Angioplasty in Myocardial Infarction Study Group (1987) A randomized trial of
immediate versus delayed elective PCI after intravenous tissue plasminogen activator in acute myocardial infarction. N. Engl J. Med. 317:581–588.
36. Simoons, M.L., Beitriu, A., Col, J., et al. (1988) Thrombolysis with tissue plasminogen activator in acute myocardial infarction: no additional benefit from immediate
percutaneous coronary intervention. Lancet 1:197–203.
37. Ross, A.M., Coyne, K.S., Reiner, J.S., et al. for the PACT Investigators (1999) A randomized trial comparing primary PCI with a strategy of shortacting
thrombolysis and immediate planned rescue PCI in acute myocardial infarction: The PACT Trial. J. Am. Coll Cardiol. 34: 1954–1962.
38. Schweiger, M.J., Cannon, C.P., Murphy, S.A., et al. (2000) Percutaneous coronary intervention for TIMI 2 or 3 flow: the combined TIMI 10b/TIMI 14
experience. Circulation 102(11): 753.
39. Cheitlin, M.D., Alpert, J.S., Armstrong, J.E., et al. (1997) ACC/AHA guidelines for the clinical application of echocardiography: executive summary. A report of
the American College of Cardiology/American Heart Association task force on practice guidelines (committee on clinical application of echocardiography). J. Am. Coll
Cardiol. 29:862–870.
40. Cheitlin, M.D., Conill, A., Epstein, A.E., et al. (1998) ACC/AHA guidelines for implantation of cardiac pacemakers and antiarrthymia devices: a report of the
American College of Cardiology/American Heart Association task force on practice guidelines (committee on pacemaker implantation). J. Am. Coll Cardiol. 31: 1175–
1209.
Page 164

Page 165
Chapter 11
Cardiac rhythm disturbances
Magdy Migeed, MD and
Lawrence S.Rosenthal, MD, PhD
Contents
11.1 Introduction
11.2 ECG analysis and interpretation
11.3 Supraventricular tachycardias
11.4 Treatment of supraventricular arrhythmias
11.5 Atrial fibrillation
11.6 Therapy of AF
11.7 Wide complex tachycardias
11.8 Approach to the patient with cardiac arrest
11.9 Bradyarrhythmias
11.10 Summary
Page 166
11.1 Introduction
Imagine that you are the most senior physician in the hospital and you are asked at 3 a.m. to evaluate a patient with a wide complex rhythm. The task of making the
correct diagnosis and prescribing the best treatment can be a source of fear and apprehension, or can be relatively a straightforward decision, if you know the cues and
clues. This chapter will provide a concise approach to the diagnosis and acute management of cardiac arrhythmias.
11.2 ECG analysis and interpretation

The 12lead ECG analysis plays an integral part in the determination of the arrhythmia and includes:
• identification of P waves and their morphology then determination of atrial rhythm;
• analysis of QRS morphology;
• determination of A–V relationship.
11.2.1 Step I. Determine the atrial rhythm

Knowledge of normal ECG findings makes determination of abnormal findings easier. However, one should interpret ECGs in an algorithmic fashion to avoid missing
important findings. The first step in the diagnosis of an arrhythmia is to determine the atrial rhythm. Thus one must carefully inspect the ECG or rhythm strip for P waves.
A sinus origin is present if the P waves are upright in leads II, III, and aVF (ecg lead). By convention:
• sinus bradycardia (SB): sinus rate less than 60 beats min−1;
• sinus rhythm (SR): sinus rate between 60 and 100 beats min−1;
• sinus tachycardia (ST): sinus rate greater than 100 beats min−1.
11.2.2 Step II. Look at the QRS complexes

It now becomes important to differentiate supraventricular tachycardia (SVT) from ventricular tachycardia (VT). Most important in this determination is the evaluation
of the hemodynamic status of the patient. Regardless of the underlying mechanism of the arrhythmia, immediate direct electrical cardioversion should be considered in
patients with heart failure, acute ischemia, hypotension, hypoxia, or impending cardiovascular collapse. In addition to restoring sinus rhythm, the clinician should look for
underlying causes, as many arrhythmias are not primary abnormalities but secondary arrhythmias as a result of other systemic abnormalities such as cardiac ischemia,
hypoxia, or electrolyte abnormalities.
11.2.3 Step III. Determine the AV relationship

If P waves always precede QRS complexes by a fixed normal P–R interval (0.12–0.2 s), then AV conduction is normal. Other types of impaired AV conduction
include:
• Firstdegree AV delay: PR interval greater than 200 ms.
• Seconddegree heart block:
Type I (Wenckebach block): Progressive PR lengthening with intermittent dropped QRS complexes;
Type II: Constant PR intervals with dropped QRS complexes.
• Thirddegree heart block: AV dissociation with atrial rate greater than ventricular rate.
11.3 Supraventricular tachycardias
11.3.1 Evaluation of the R–P interval

When confronted with a narrow complex tachycardia, it is often helpful to determine the QRS to P wave relationship. However, at times P waves may be difficult to
locate. Thus, the use of vagal maneuvers (see below) or the use of an intravenous AV nodal blocking agent may be helpful in determining the etiology (and even to
terminate the tachycardia). Having determined the QRS to P relationship, SVTs can be classified as short R–P tachycardias or long R–P tachycardias (see Figure
11.1). Defining a tachycardia as a short R–P SVT narrows the list of potential arrhyth
Page 167
mias. Most SVTs will be short R–P tachycardias consistent with typical atrioventricular nodal reentrant tachycardia (AVNRT). One must still eliminate other
possibilities including atrial tachycardia, sinus tachycardia with a firstdegree AV delay, and AV reentry or pathway
Figure 11.1 Nomenclature for the classification of SVT. Having determined the QRS to P relationship, SVTs can be classified as short
R–P tachycardias or long R–P tachytardias. Defining a tachycardia as a short R–P SVT narrows the list of potential
arrhythmias, Most SVTs will be short R–P tachycardias consistent with typical atrioventricular nodal reentrant tachycardia
(AVNRT), One must still eliminate other possibilities including atrial tachycardia, sinus tachycardia with a firstdegree AV
delay, and AV reentry or pathway mediated tachycardias. Table 11.1 helps with the differential diagnosis of common SVTs.
mediated tachycardias. Table 11.1 helps with the differential diagnosis of common SVTs.
11.3.2 Diagnostic maneuvers for SVT

Carotid sinus massage (CSM) enhances vagal tone and slows the heart rate, and the response to CSM provides important information (Table 11.2). Carotid sinus
massage slows the impulse input in to the AV node. If conduction worsens, then this favors block at the level of the AV node or above. If conduction improves, then
the site of block is likely below the AV node, ie. HisPurkinje System. CSM should not be performed if a carotid bruit is auscultated. CSM is performed with the
patient supine and the head hyperextended and turned away from the side being tested. The carotid impulse can be felt at the angle of the jaw. A gentle touch to carotid
sinus to detect hypersensitive response is followed by firm pressure for 5 s first on one side then on the other side with gentle rotating massaging motion. Never test
both sides simultaneously. CSM carries the risk of cerebral emboli, particularly in older patients.
Table 11.1 Differential diagnosis of common SVTs
I. Regular supraventricular tachycardias

Short R–P tachycardias (R–P<P–R) Sinus tachycardia with firstdegree AV delay
Typical AV nodal reentrant tachycardia (common)
AV reentrant tachycardia
Atrial tachycardia
Long R–P tachycardias (R–P>P–R) Sinus tachycardia
Sinus nodal reentrant tachycardia
Atypical AV nodal reentrant tachycardia
AV reentrant tachycardia
Permanent junctional reciprocating tachycardia
II. Irregular supraventricular tachycardias
Atrial fibrillation
Atrial flutter
Table 11.2 Cardiac response to carotid sinus massage
• Sinus tachycardia Slowing then return to baseline after massage

• AV node or AV reentry tachycardia (WPW) Slow, terminate, or no change
• A fib or flutter Slowing ventricular rate during massage, atrial arrhythmia persists
• seconddegree AV block, Type I Worsens conduction or no change
• seconddegree AV block, Type II Improvement of conduction or no change
• VT (2:1 blocks) No change, rare conversion
Page 168
11.4 Treatment of supraventricular arrhythmias

The acute treatment of SVT (Table 11.3) is termination, but depends on patient hemo aimed at both arrhythmia diagnosis and dynamic stability. If clinically unstable
then Advanced Cardiac Life Support (ACLS) guidelines1 should be followed. In hemodynamically stable patients, it is reasonable to perform vagal maneuvers or
administer AV nodal blocking agents in an attempt to terminate their SVT. One should always listen for carotid bruits before carotid sinus massage. In addition, the
Valsalva maneuver, splashing cold water on the face, or gently rubbing the upper eyelid have been shown to stimulate the vagus nerve and terminate supraventricular
tachycardia. Intravenous use of adenosine, calcium channel blockers, beta blockers, and even digoxin can terminate SVT by causing block at the level of the AV node
(see Table 11.4). Antiarrhythmic agents can also be used for acute conversion and longterm suppression of SVTs and are listed in Table 11.5.
11.4.1 Sinus tachycardia

First determine whether sinus tachycardia is appropriate (i.e. hypovolemia, fever, anemia, hyperthyroidism, CHF). Treat the underlying condition. However, if the sinus
tachycardia is causing ischemia, then the preferred drug would be beta adrenergic blockers.
11.4.2 Sinus bradycardia

Treat only if symptomatic (withdrawal of AV nodal agents, temporary or permanent pacemaker, atropine 0.5–2 mg IV).
11.4.3 Atrial premature depolarization (APD)

Usually not clinically symptomatic and requires no therapy. If symptomatic, beta adrenergic blockers may be used.
11.4.4 Multifocal atrial tachycardia

This is often related to underlying lung disease. Treat the underlying cause and optimize cardiopulmonary status. Calcium channel blockers or beta blockers (unless
contraindicated) may be used for rate control. Cardioversion has no value.
11.4.5 Atrioventricular nodal reentrant tachycardia (AVNRT)

Acute treatment includes vagal maneuvers (CSM, Valsalva). If no response, IV adenosine, IV verapamil, IV diltiazem, or IV beta blockers may be used. If
hemodynamically unstable, electrical cardioversion is warranted. Chronic treat
Table 11.3 Acute and.chronic management of SVT
Acute
1. Hemodynamically unstable (CHF, hypoxia, hypotension, impending CV collapse)
• Follow ACLS guidelines
2. Hemodynamically stable
• Carotid sinus massage (see precautions in text)
• Adenosine IV push (doses of 6, repeat 6, then 12 mg if necessary)
• Calcium channel blockers (e.g. verapamil 5 mg IV push)
• Beta adrenergic blockers (e.g. metoprolol 5 mg IV push)
• Digoxin (loading dose necessary)
Chronic
1. Drug therapy
• Calcium channel blockers
• Beta adrenergic blockers
• Digoxin (not very effective)
• Antiarrhythmic agents
2. Catheter ablation
Page 169
Table 11.4 Intravenous drugs in the acute management of SVT
Drug Target IV dosage Comments

Adenosine (Adenocard) Adenosine receptors 6–24 mg Rapid onset
at the level of the AV node Short half life
and sinus node Reflex sinus tachycardia
Atenolol (Tenormin) Beta adrenergic receptors 5–10 mg Bradycardia
Hypotension
Diltiazem (Cardizem) Calcium channels 20–25 mg Hypotension
Esmolol (Brevibloc) Beta adrenergic receptors 5–10 mg bolus or infusion Short half life
Metoprolol (Lopressor) Beta receptors 5 mg Longer acting
Verapamil (Calan, Isoptin) Calcium channels 2.5–5 mg Hypotension
Digoxin (Lanoxin) + +
Na /K pump inhibitor 0.250 mg Toxicity
Long onset of action
Phenylephrine (Neosynephrine) Alpha1 adrenergic receptors (ref lex slowing with increased SVR) 100 µg May be hetpful in hypotension
Table 11.5 VaughnWilliams classification of antiarrhythmic agents and their site of action
Class Ia Quinidine, Procainamide, Disopyramide Na+ channels

Class Ib Lidocaine, Mexilitine
Class Ic Flecainide, Propafenone
Class II Beta blockers Beta receptors
Class III Amiodarone, Sotalol, Dofetilide K+ channels
Class IV Calcium channel blockers Ca2+ channels
ment: infrequent, welltolerated, brief episodes can be treated with vagal maneuvers alone. Patients with frequent symptomatic episodes should consider catheter
ablation (>95% success). AV nodal blocking agents such as beta blockers, verapamil, or diltiazem have been used and generally reduce the frequency of episodes.
11.4.6 Atrioventricular reentrant tachycardia (AVRT)

Occurs in patients with accessory pathways (APs). Baseline ECG may show the characteristic delta wave (manifest APs or WolfeParkinsonWhite (WPW) pattern) or
a nonpreexcited ECG as seen in patients with concealed APs (retrograde conduction only). The pathophysiology involves an abnormal muscular network of
specialized tissues connecting the atrium and the ventricle (accessory AV pathway). Manifest pathways are found in 0.2–0.4% of population, especially young males.
Most patients have no underlying heart disease.
• Asymptomatic patients usually require no therapy.
• Patients with syncope or presyncope require electrophysiology (EP) study and consideration for catheter ablation.
Page 170
Chronic management of SVTs includes oral use of beta blockers, calcium channel blockers, and even the use of antiarrhythmic agents (see Table 11.6). However,
given the small but everpresent risks of proarrhythmia and drugrelated side effects associated with the use of antiarrhythmic agents, most clinicians would favor
radiofrequency catheter ablation (>95% success) as an alternative to longterm antiarrhythmic use.
11.5 Atrial fibrillation

Atrial fibrillation (AF) is the most common arrhythmia. Its prevalence is 5% in patients of 65 years of age or older, and the prevalence increases with advancing age.
Data from the Framingham study have shown that AF is associated with a 1.5–1.9fold higher risk of death, presumably as a result of thromboembolic stroke. AF is
associated with several conditions that should be sought when taking the history (Table 11.7).
Overall, about 15–25% of all strokes in the USA (75 000 per year) can be attributed to AF. Known risk factors contributing to the development of AF are male
gender, valvular heart disease (rheumatic heart disease), CHF, hypertension and diabetes (Table 11.8). Additional risk factors, such as advancing age, prior history of
stroke, hypertension, and diabetes place patients with preexisting AF at even higher risk for further comorbidities. Patients with nonvalvular AF and with risk factors
have a fivefold increased risk for stroke. Patients with rheumatic heart disease have an even higher (17fold) risk for stroke. Large multicenter clinical trials have clearly
demon
Table 11.6 Oral drugs for long term suppression
Drug Target Dosage Comments

Verapamil Calcium channel 240–480 mg day −1 Well tolerated
Diltiazem Calcium channel 120–360 mg day−1 Well tolerated
Atenolol Beta receptors 25–200 mg day −1 Well tolerated
Metoprolol Beta receptors 25–200 mg day −1 Well tolerated
Nadolol Beta receptors 20–160 mg day−1 Well tolerated
Procainamide Ia antiarrhythmic agent 50 mg kg−1 day−1 Longterm side effects
Propafenone Ic antiarrhythmic agent 150–300 mg t.i.d. Caution in patients with reduced LVF
Flecainide Ic antiarrhythmic agent 50–150 mg b.i.d. Caution in patients with reduced LVF
Sotalol III antiarrhythmic agent 80–160 mg b.i.d. Caution in patients with renal dysfunction, CHF
Amiodarone III antiarrhythmic agent 200–400 mg day−1 Organ toxicity
Table 11.7 Common causes and associated conditions with AF
Advancing age Hypertension Diabetes

Cardiomyopathy Alcohol abuse Thyrotoxicosis
WPW and sicksinus syndromes Thrombosis; particularly pulmonary embolism Valvular heart disease (e.g. mitral stenosis)
Page 171
Table 11.8 Risk factors for stroke in nonrheumatic AF
Advancing age LV dysfunction

Previous TIA, CVA LA enlargement
Prosthetic valve Diabetes mellitus
Thyrotoxicosis Congestive heart failure
Hypertension
Generally all patients with chronic AF (paroxysmal, persistent, permanent) should receive anticoagulation (see Table 11.9) to an INR between 2 and 3 except patients under 65 years old
with no risk factors.
strated that anticoagulation with warfarin reduces the risk of stroke by 50–80%. Anticoagulation, however, carries the risk of bleeding, hemorrhagic stroke, and
intracranial bleeding, particularly in elderly patients prone to falls.
AF is classified into acute (<48 h) or chronic (>48 h). Chronic AF is further divided into three groups:
• paroxysmal: intermittent;
• persistent: patient in AF but can be converted to sinus rhythm by chemical or electrical cardioversion;
• permanent: accepted rhythm is AF, goals are anticoagulation and rate control.
The history may reveal symptoms such as palpitations, dizziness, or general malaise. The presence of heart failure suggests concomitant cardiac disease (e.g. valvular
heart disease). Persistent SVT can cause a tachycardiainduced cardiomyopathy with LV dysfunction and heart failure. On ECG AF should be distinguished from atrial
flutter (AFL) (uniform flutter waves) and from multifocal atrial tachycardia (MAT) (i.e. irregular rhythm with P waves of variable (at least three) morphologies).
However, most physicians would anticoagulate patients with AFL in a similar manner to those with AF.
Recommendations for chronic anticoagulation are controversial. Although the risk of bleeding can be minimized with tight control of the INR, the possibility of stroke
is frightening to most and can be devastating. Most clinicians generally recommend longterm anticoagulation in any patient with atrial fibrillation or atrial flutter who is
not at risk for falls or injury or who cannot determine when they are in atrial fibrillation (Table 11.9). Patients with PAF are instructed to call when they have AF lasting
more than 12 h, so that prompt cardioversion can be arranged.
11.6 Therapy of AF
Electrical cardioversion is indicated in patients who are clinically unstable (e.g. heart failure, chest pain, hypoxic, hypotensive, impending hemodynamic collapse). Acute
AF (<24–48 h) can be cardioverted electrically or chemically regardless of anticoagulation status. For AF (>48 h) cardioversion should be deferred until a minimum of
3 weeks of therapeutic anticoagulation (INR>2) unless a transesophageal echo (TEE) showed no thrombus. However, if there were any uncertainties about the exact
duration of AF, then most clinicians would defer to a more conservative approach, including therapy with warfarin. Treatment of chronic AF includes three goals:
control of ventricular rate, maintenance of sinus rhythm, and stroke prophylaxis.
11.6.1 Rate control

Rate control not only alleviates symptoms but also prevents tachycardiainduced cardiomyopathy. In patients who have already developed cardiomyopathy, rate
control can dramatically improve ventricular systolic function.
• Pharmacologic rate control can be achieved using three categories of drugs, alone or in combination:
Table 11.9 Anticoagulation of nonrheumatic AF
Age Risk factors* Recommend

<65 + Coumadin, INR 2–3
− ASA or nothing
65–75 + Coumadin, INR 2–3
− Coumadin or ASA
>75 + or − Coumadin, INR 2.0–2.5
*Risk factors include diabetes, hypertension, prior CVA/TIA.
Page 172
(i) beta adrenergic blockers;

(ii) calcium channel blockers;
(iii) digoxin.
• Nonpharmacologic control of ventricular rate using AV nodal catheter ablation (resulting in complete AV block) with implantation of a pacemaker is indicated if the
rate is not well controlled by drugs or the patient has intolerable side effects from these drugs. Improvements in LV size and function, quality of life, and exercise
tolerance with AV nodal ablation and permanent pacing have been demonstrated.
11.6.2
Rhythm control
The second goal is maintenance of sinus rhythm after conversion from AF to sinus rhythm (spontaneously or with cardioversion). Note that maintaining sinus rhythm has
not been shown to decrease the likelihood of thromboembolism or to increase survival. In fact, some drugs used to prevent AF recurrence can cause new arrhythmias
(proarrhythmic effects).
Pharmacologic control of rhythm can be achieved with Class Ia drugs, but these can be associated with Torsades de Pointes (TdP) and increased mortality. Monitor
QT interval and initiate these agents with continuous EKG monitoring. These agents also enhance AV nodal conduction, so rate control agents should be given before
their use. In patients with structurally normal hearts, Class Ic agents can safely be used. Amiodarone has been proven safe after MI and in systolic dysfunction.
Because of limitations of medical therapy (proarrhythmia, drug side effects, high recurrence rate of AF), there is interest in using nonpharmacologic control of rhythm.
These include:
• dual atrial pacing and implantable atrial defibrillators;
• atrial pacing versus ventricular pacing in patients with sick sinus syndrome (SSS) (reduces stroke, CHF and recurrence of AF);
• surgical maze procedure (by creating linear incisions in the atria that prevent atrial reentry and AF, but requires thoracotomy);
• AF ablation, under development.
11.6.3 Stroke prevention

Because up to 7% of AF patients undergoing cardioversion without anticoagulation experience clinical thromboembolism, patients with AF of longer than 48 h must
receive anticoagulation for a minimum of 3 weeks before cardioversion and 4 weeks after cardioversion. Alternatively, TEE may be performed and if no LA or LV
thrombus is found, safe cardioversion can be accomplished, followed by 3–4 weeks of anticoagulation. Table 11.8 lists factors that are known to increase stroke risk.
11.7 Wide complex tachycardias

The diagnosis and treatment of wide complex tachycardias (WCT) (Table 11.10.) can be crucial in the care of critically ill patients. When confronted with a WCT, the
most important piece of information that one can obtain is whether there is a history of coronary disease, prior heart attack, or heart failure. A ‘yes’ answer almost
certainly means ventricular tachycardia (VT). Unfortunately, the physical exam is of very little clinical utility, as most patients will be hemodynamically compromised.
However, there will be patients with wide complex VT who will have heart rates in the low 100s and have sufficient blood pressure to have walked into the emergency
room, especially those on antiarrhythmic agents such as amiodarone. Thus the need to identify VT is important because not all patients will present
Table 11.10 Differential diagnosis of a regular wide complex rhythm
• VT
• SVT with aberrant conduction
• WPW with antegrade conduction through the accessory pathway (rare)
• Ventricular pacing (usually obvious)
Page 173
pulseless and hypotensive. In addition, there may be patients with runs of nonsustained wide complex tachycardia, and the ability to distinguish VT from SVT will be
critical in the management of such patients. VT is more common in older patients, and in any patient with a history of prior MI, CHF, or reduced LV function from any
cause. Thus any person with a history of MI (recent or in the distant past) and reduced LV function, who presents with a wide complex rhythm should be treated as VT
until proven otherwise. One should look for signs of AV dissociation such as cannon A waves in the neck as well as signs of CHF, but the initial focus should be on
obtaining a 12lead EKG.
11.7.1 SVT with aberrancy

Distinguishing SVT with aberrancy from VT can be difficult. SVT with aberrancy generally occurs in the setting of atrial fibrillation and variable R–R intervals. Look for
long short sequences as the initiating event and the absence of a pause after the arrhythmia terminates. This is more suggestive of aberrant conduction (Figure 11.2).
After a long R–R interval (that leaves one bundle branch refractory) an early impulse from the atria is conducted aberrantly because the bundle branch has not
recovered yet and is still partially refractory. With VT, one often sees a HisPurkinje conduction system that is refractory from retrograde bombardment of ventricular
beats. Upon termination of the VT or nonsustained VT, the HisPurkinje system requires extra time to regain normal conduction from the AV node down towards the
ventricles (Figure 11.3). However, if a wide complex tachycardia is seen in a patient with LV dysfunction, it should be presumed to be VT until proven otherwise.
11.7.2 EKG diagnosis

The nomenclature adopted for the description of a wide complex rhythm begins with the examination of lead V1. If V1 is predominantly positive, it is described as a
right bundle pattern. If the polarity of V1 is negative then a left bundle pattern can be diagnosed. The inferior
Figure 11.2 SVT with aberrant conduction. Rhythm strip) on a 28yearold women with documented SVT. During her EP study, atrial
pacing produced a run of wide complex tachycardia. Note the lack of pause after wide complex tachycardia terminates.
Intracardiac electrograrm conf firmed the diagnosis of SVT.
Figure 11.3 Aberrant conduction consistent with VT. Rhythm strip on a 48yearold male who had suffered an anterior MI and had an
LVEF of 35%. Note the pause after termination of wide complex tachycardia. At EP study, the patient was readily inducible
for monomorphic VT with a similar configuration.
leads II, III, and AVf help orient the axis of the VT (Table 11.11).
Table 11.12 lists the ECG findings favoring the diagnosis of VT. After determining the wide complex tachycardia morphology (right bundle branch block (RBBB) or
left bundle branch block (LBBB)), focus on these criteria, which support or refute the diagnosis of VT. With an LBBB tachycardia, first examine lead V1 (or V2) and
measure the width of the R wave. A broad R wave (>30 ms) or finding of a notch on the downslope of the S wave is consistent with VT. Measurement of the onset of
Table 11.11 Nomenclature for the classification of VT
V1 positive Right bundle pattern
V1 negative Left bundle pattern
Inferior leads upright Inferior axis

Inferior leads negative Superior axis
Page 174
Table 11.12 ECG finding favoring the diagnosis of VT
• QRS>140 ms (RBBB)
(i) V1 monophasic R, qR, Rr’, RS
(ii) V6 R/S<1
• QRS>160 ms (LBBB)
(i) V1−2 R>30 ms
(ii) V1−2 notching on the downslope of S wave
(iii) V1−2 beginning of QRS to nadir of S>70 ms
(iv) V6 qR
• AV dissociation
• Positive concordance
• Change in axis (from baseline ECG)
• Axis–90±180
• Fusion beats
• Capture beats
the R wave to the nadir of the S wave (>70 ms) also adds to the mounting evidence pointing to VT (Figure 11.4). Finally, any Q wave in V6 is pathologic and should
raise the suspicion of VT. The criteria with RBBB tachycardias are a little more difficult, but making an assessment of the R to S ratio in V6 is helpful (R/S<1 is
consistent with VT). A comparison of ECG findings is seen in Figure 11.5. There will be exceptions, but utilizing this simple algorithm along with patient history should
allow the correct diagnosis of VT to be made in most cases.
11.7.3 Treatment of VT
Hemodynamically unstable VT requires immediate action. Remember the timehonored saying, ‘Shock early and often’. Studies have shown that the earlier one
delivers energy to the heart, the better the chance of success and patient survival. The year 2000 Advanced Cardiac Life Support (ACLS) guidelines have incorporated
IV amiodarone into the adult VT/Pulseless VT algorithm. Currently, IV amiodarone is FDA approved for the treatment and prophylaxis of persistent VF as well as
hemodynamically unstable VT in patients refractory to other antiarrhythmic agents.
Figure 11.4 ECG features of LBBB morphologic VT. A prolonged interval (>70 ms) from the onset of the R wave to the nadir of the S
wave or any Q wave In V6 raises the suspicion of VT, From Wellens, H., and Brugada, P. (1987) Diagnosis of ventricular
tachycardia from the 12lead electrocardiogram. In: Cardiology Clinics. Reproduced with permission from W.B. Saunders.
11.8 Approach to the patient with cardiac arrest

The ACLS guidelines have been recently updated; Figure 11.8 (at the end of this chapter) reproduces the main algorithm. The reader is encouraged to read and study
these guidelines in more detail1.
In the field setting ACLS begins with establishing unresponsiveness, and following ABCD:
A: open Airway by the look, listen, and feel method;

B: assess Breathing and give two slow breaths;
C: assess Circulation and start Chest compressions;
D: attach Defibrillator/monitor and give 200, 300, 360 joules.
If no pulse, continue CPR and assess rhythm.

In the intensive care unit, the situation is usually different (the airway is frequently secured with an endotracheal tube; breathing is monitored on the ventilator or with
endtidal CO2) and the ACLS flowchart may need to be modified. A quick check of monitoring electrodes should exclude mechanical disconnection as the cause of
‘flatline’ EKG. Pulseless conditions (pulseless VT, pulseless electrical activity (PEA)) require immediate institution of chest com
Page 175
Figure 11.5 Comparison of the bundle branch morphologic criteria in the diagnosis of VT. VT may be difficult to diagnose in the
presence of bundle branch block, but interpretation of the QRS complex in leads V1 and V6 along with patient history is often
helpful. From; Zipes, P. and Jalife, J. (1995) Recognition of VT. In: Cardiac Electrophysiology: from Cell to Bedside.
Reproduced with permission from W.B. Saunders,
pressions. Rhythms with perfusion pressure (often visible by the arterial line tracing) can usually be dealt with chemically or electrically.
11.8.1 VF/pulseless VT
Give up to three defibrillation shocks. If no response perform CPR for 1 min followed by secondary ABCD survey:
• Airway: place airway device if needed;
• Breathing: secure airway and start ventilation and oxygenation (100% FiO2 preferred);
• Circulation: gain IV access, identify the rhythm on monitor and administer agents to support blood pressure, such as vasopressin 40 U IV×1 or epinephrine 1 mg IV
q 3–5 min.
If no response:
• Defibrillate: 360 joules within 30–60 s.
If no response:
• consider antiarrhythmic agents (such as amiodarone, lidocaine; Table 11.13).
If no response:
• resume attempts to defibrillate.
11.8.2 Asystole or pulseless electrical activity (PEA)

Perform CPR for up to 3 min then proceed to secondary ABC survey as above. Administer epinephrine 1 mg IV q 3–5 min. If no response, administer atropine 1 mg
IV q 3–5 min up to 0.04 mg kg–1.
• For PEA look for a cause:
• Hypovolemia, Hypoxia, Hyper/hypocalcemia, Hypothermia, Hydrogen ion (acidosis);
• Tamponade, Tension pneumothorax, Thrombus (coronary), Thrombus (pulmonary).
• For asystole:
• consider transcutaneous pacing immediately.
11.8.3 Stable ventricular tachycardia

Occasionally, patients will present with hemodynamically stable VT. If the patient is hemodynamically unstable, or if ischemia is suspected, then cardiovert immediately.
If hemodynamically stable, one can use any of the following:
Page 176
Table 11.13 Drugs or the management of ventricular dysrhythmias
Drug initial bolus Secondary infusion(s) Maintenance infusion

Amiodarone 300 mg IVP (6 cc undiluted, then flush w/NS or 150 mg IV (3 cc diluted in 100 cc mg min−1 for 6 hours (360 mg) then 0.5 mg min−1 for 18 hours
D5W) D5W) (540 mg)
Lidocaine 100 mg IVP (1–1.5 mg kg−1) 50–100 mg IVP (0.75–1.5 mg kg−1) 1–4 mg min−1
Procainamide 20–30 mg min −1 −1
Max 17 mg kg n/a 2–6 mg min−1
Bretylium 5–10 mg kg−1 n/a 0.5–2.5 mg kg−1
Max 30 mg kg−1 over 24 h
• lidocaine (1 mg kg−1 IV push, followed by an IV infusion of 1–4 mg min−1) (Table 11.13);

• amiodarone (150 mg IV bolus over 10 min followed by a constant infusion of 1 mg min−1 for 6 h and 0.5 mg min−1 thereafter, can rebolus with 150 mg for
breakthrough VT);
• procainamide IV (18 mg kg−1 over 1 h, followed by a constant IV infusion of 2–6 mg min−1);
• if the VT does not convert chemically then consider synchronized cardioversion.
11.8.4 Drugs for the management of ventricular dysrhythmias

Table 11.13 provides a quick summary of the initial dose and maintenance infusion for the four commonly used ventricular antidysrhythmics, although most conditions
can be managed with either amiodarone or lidocaine. Amiodarone, because of its long halflife, requires a complicated initial dosing scheme. A bolus dose of 300 mg is
used to initiate therapy, and a supplemental dose of 150 mg may be given if the patient experiences breakthrough episodes of VT or VF. Once spontaneous circulation
is established, an initial maintenance infusion of 360 mg over 6 h (1 mg min−1) is given, followed by an additional 540 mg over 18 h (0.5 mg min−1). Amiodarone is
compatible with some other vasoactive agents (isoproterenol, norepinephrine, dopamine, lidocaine, phenylephrine, dobutamine, nitroglycerin) but compatability is
unknown with other agents. The major side effect of amiodarone is hypotension (10–30% of patients when given as a rapid infusion), asystole (1.2%), VT (1.1%),
cardiogenic shock (1%) and Torsade de Pointes (<1%)2. Pulmonary fibrosis and pneumonitis have been reported a few days following initiation of IV therapy, and can
occur any time thereafter3. Patients on chronic therapy should be monitored with pulmonary function tests and periodic chest radiographs4. After suppression of
arrhythmia by IV amiodarone, patients can be switched to an oral formulation, with the dose dependent on the duration of IV treatment.
Lidocaine has been the mainstay of treatment for many years. In contrast to amiodarone, it has a relatively short half life, and no oral equivalent. Serum levels are high
after an IV bolus, but drop during the first minutes of continuous infusion, so a second bolus (usually half the initial dose) may be necessary for breakthrough
arrhythmias. Lidocaine is metabolized by the liver, and has a tendency to accumulate in elderly patients. The major side effects are central nervous system effects, and
can range from tinnitus and confusion at low doses to seizures to CNS depression with severe overdose.
Procainamide is effective against atrial, AV nodal and ventricular arrhythmias, and slows conduction through accessory conduction
Page 177
pathways. It decreases automaticity and conduction in all cardiac tissue, both normal and ischemic. Its toxicity is first manifest on the heart (Torsades des Pointes; AV
nodal block, intraventricular conduction delay, myocardial depression) and then the peripheral circulation (vasodilation) and the CNS (excitatory syndrome, confusion,
seizures). The lupuslike syndrome with chronic administration is not seen with shortterm therapy.
Bretylium is notable for the hypotension it causes with bolus administration, and is now rarely used.
11.8.5 Pacemakers and internal cardiac defibrillators (ICDs)

Patients with pacemakers or ICDs should be treated according to ACLS protocols during episodes of cardiac arrest. If at all possible, avoid delivering shocks directly
over such devices as they may shunt energy away from the heart. One should consult a trained expert on the magnet function of such devices. In general, placing a
magnet over a pacemaker causes the pacemaker to pace in a preprogrammed, manufacturerspecific, paced mode. Most defibrillators are inhibited from delivering
therapy (overdrive pacing protocols and/or shocks) when a magnet is placed over the device. Pacing functions are maintained. Again, not all devices behave similarly
and consultation with an expert is advised.
11.8.6 Proarrhythmia
Proarrhythmia refers to the capacity of a drug (cardiac or noncardiac) to aggravate an existing arrhythmia or provoke a new arrhythmia at therapeutic or sub
therapeutic levels5. Risk fac tors promoting proarrhythmia include having a reduced left ventricular function (LVF), a prolonged QT interval at baseline, electrolyte
imbalances, prominent U waves, ischemia, presence of atrial arrhythmias (atrial fibrillation or flutter) or preexisting ventricular arrhythmias (nonsustained VT)6 (Table
11.14). Prevention begins with identifying those patients at risk for proarrhythmia, before drug initiation. Electrical instability occurs in the setting of
Table 11.14 Clinical risk factors associated with proarrhythmia
General
• History of VT or VF
• Poor LVEF (<30%)
• CHF
• Female gender
• Metabolic or electrolyte abnormalities
• Diuretic use
• Baseline QT prolongation
• Atrial fibrillation or flutter
Polymorphic VT/VF (Type IA drugs)
• Hypokalemia, hypomagnesemia
• Bradycardia or pauses
Incessant wide complex tachycardia (Type 1C drugs)
• Previous sustained VT
• Prior infarct or scar
hypokalemia, hypomagnesemia, bradycardia, or ischemia. Thus electrolytes should be monitored closely. With the exception of oral amiodarone (long halflife and low
incidence of proarrhythmia), antiarrhythmic agents should be started in the inpatient setting. Close monitoring of QT intervals is warranted. The management of
proarrhythmia includes eliminating the offending drug, close electrographic monitoring, correcting electrolyte abnormalities, acute treatment, and finally prevention of
further episodes. Treatment includes electrical cardioversion of unstable rhythms, overdrive pacing or increasing the heart rate with isoproterenol or atropine,
magnesium sulfate infusion (Table 11.15), and even antiarrhythmic agents that do not prolong the QT (lidocaine, beta blockers).
11.8.7 Polymorphic VT/VF (Figure 11.6)

Polymorphic VT/VF suggests ischemia and consideration towards immediate cardiac
Table 11.15 The use of IV magnesium for TdP
• µg sulfate 2 g bolus over 2 min

• Repeat if necessary
• Infusion of 1–2 mg min−1 to 20 mg min−1 for the next 24–48 h until QT shortens to <0.5 s
Page 178
Figure 11.6 Algorithm for treatment of persistent pulseless ventricular tachycardia (VT) and ventricular fibrillation (VF). From:
ACC/AHA Guidelines for the management of patients with acute myocardial infarction. J. Am. Coll Cardiol. (1999) 34:890–
911,
intervention should be considered. After restoration of sinus rhythm one should check the baseline QT interval. If QT interval is normal, consider ischemia or electrolyte
imbalance. If cardioversion fails to restore or maintain sinus rhythm, then consider using IV amiodarone or IV lidocaine in addition to cardioversion.
11.8.8 Torsade de Pointes (TdP)

TdP is a pausedependent form of polymorphic VT with ventricular rates 200–250 beats min−1. The most common form of TdP is iatrogenic or drug induced. Many
medications can cause widening of the QT interval, leading to excessive delay in repolarization and the development of early after depolarizations at the cellular level.
Table 11.16 lists drugs that have been associated with TdP. Changes in renal or hepatic function can also affect circulating drug concentrations, leading to prolongation
of the QT interval. Thus one needs to monitor these systems periodically. TdP has also been observed in patients who have undergone AV nodal ablation and
permanent pacemaker implantation and have a relatively slow lower rate (e.g. 60 beats min−1). For that reason these
Table 11.16 Drugs associated with TdP
Antiarrhythmics Antihistamines Antibiotics

Quinidine Terfenadine Pentamidine
Procainamide Astemizole Bactrium
Disopyramide Sparfloxacin
Sotalol Antipsychotics Erythromycin
Dofetilide Haloperidol Clarithromycin
Ibutalinde Thioridazine Ketoconazole
Amiodarone Other
Azimilide Antireflux Terodiline
Tedisimil Cisapride Grapefruit juice
Bepridil ASA
Page 179
patients are generally paced at relatively high baseline rates (85–95 beats min−1) for 1–3 months postAV node ablation. Additionally, TdP can be seen in patients with
CHB with a slow ventricular response (Figure 11.7). Correct electrolyte abnormalities and administer IV magnesium (Table 11.15). As TdP is pause dependent,
consider either IV isoproterenol or temporary overdrive pacing.
11.8.9 Premature ventricular contractions (PVCs)

Asymptomatic PVCs do not require any treatment, as the CAST I and II trials demonstrated7,8. Patients with depressed LV function (<40%) and frequent PVCs were
given Type Ic antiarrhythmic agents (flecainide and encainide, and later morizicine). PVCs became less frequent at the cost of a higher mortality rate than in patients
taking placebo, presumably as a result of ischemic mediated proarrhythmia. This trial underscores the importance of a placebo group. In patients with symptomatic
PVCs, a trial of a beta blocker or calcium channel blocker might be warranted.
11.9 Bradyarrhythmias
If there are symptoms (lightheadedness, dizziness, syncope or near syncope, hypotension,
Figure 11.7 Pausedependent Torsades de Pointes (TdP). An elderly woman awaiting pacemaker implantation for complete heart block
experienced an episode of TdP. Her cardiac catheterization was unremarkable, and she had preserved left ventricular function.
An EP study was also performed and was negative for inducible arrhythmias. TdP is treated with administration of IV
magnesium, correction of electrolyte abnormalities, and, if necessary, IV isoproterenol or temporary overdrive pacing.
hypoxia), associated with the bradycardia, consider these options:

• atropine 0.5–1.0 mg IV;
• dopamine 5–20 µg kg−1 min−1;
• epinephrine 2–10 µg min−1;
• transcutaneous pacing or temporary endocardial pacing if available.
If there are no symptoms then evaluate for underlying causes (i.e. an overdose of an AV nodal agent). In general, Type II seconddegree or thirddegree AV block
with heart rates less than 40 beats min−1 will require transvenous pacing (temporary or permanent).
11.9.1 Temporary pacing

External, transthoracic temporary pacing can be accomplished through most newer external defibrillators. Patches should be positioned in the anterior and lateral
positions or in the anterior and posterior positions. As this form of pacing can cause muscle stimulation, it should be reserved for urgent temporary pacing while a
temporary endocardial pacemaker is inserted. The preferred venous routes for temporary endocardial pacing leads are via the right internal jugular vein (IJ) or left
innominate vein followed by the left IJ and right subclavian vein. Fluoroscopy can make insertion easier (it is necessary when insertion is performed via the femoral
veins), but standard techniques can achieve equal success. Most temporary wires have inflatable balloon tips, which allow smooth passage into the RV. By connecting
the temporary endocardial pacing wire to an external pacemaker, the operator can visualize ventricular capture. Deflating the balloon and advancing a few millimeters
will ensure stable position and serve as a site to measure pacing threshold and sensing parameters. An upright chest Xray is a must, as it not only eliminates
pneumothorax, but also confirms lead position.
11.10 Summary
When faced with any arrhythmia, one should not panic. Knowledge of the underlying rhythm
Page 180
Figure 11.8 Advanced Cardiac Life Support (ACLS) guidelines have been recently updated; vasopressin 40 U IV as a single dose is
now recommended as an alternative to epinephrine for VF/VT refractory to initial defibrillation. From: ACC/AHA Guidelines
for the management of patients with acute myocardial infarction. J. Am. Coll. Cardiol. (1999) 34:890–911.
Page 181
and its cause should be sought. When faced with a wide complex rhythm, remember that VT is more common, especially in older patients and those with a history of
infarct or heart failure. A stepwise approach examining first for atrial rhythm, then assessing the QRS complex and finally the AV relationship, will generally disclose the
rhythm disturbance and point to the most appropriate therapy (see Figure 11.8).
References
1. The American Heart Association in collaboration with the International Liaison Committee on Resuscitation (ILCOR) (2000) Guidelines 2000 for
cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 102(Suppl.): 1136–1165.
2. Gonzalez, E.R., Dannewurf, B.S. and Ornato, J.P. (1998) Intravenous amiodarone for ventricular arrhythmias: overview and clinical use. Resuscitation 39:33–42.
3. Dusman, R.E., Stanton, M.S., Miles, W.M., et al. (1990) Clinical features of amiodarone induced pulmonary toxicity. Circulation 82: 51–59.
4. Hilleman, D., Miller, A.M., Parker, R., et al. (1998) Optimal management of amiodarone therapy: efficacy and side effects. Pharmacotherapy 18:138S–145S.
5. Kerin, N.Z. and Somberg, J. (1994) Proarrhythmia, risk factors, causes, treatment and controversies. Am. Heart J. 128:575–585.
6. Kerin, N.Z. and Somberg, J. (1994) Proarrhythmia: Definition, risk factors, causes, treatment and controversies. Am. Heart J. 128: 575–585.
7. The CAST investigators: Preliminary report (1989): Effect of encainide and flecainide on mortality in a randomized trial of arrhythmia suppression after myocardial
infarction. NEJM 321:406–412.
8. The CAST investigators: (1992) Effect of the antiarrhythmic agent moricizine on survival after myocardial infarction. NEJM 327: 227–33.
Page 182

Page 183
Chapter 12
Pulmonary embolism in the critically ill patient
Jay S.Steingrub, MD
Contents
12.1 Introduction
12.2 Risk factors
12.3 Pathophysiology of pulmonary embolism
12.4 Differential diagnosis in the ICU
12.5 Diagnostic tests
12.6 Treatment
12.7 Right ventricular function
12.8 Other clinical interventions
12.9 Risk reduction
Page 184
12.1 Introduction
Pulmonary embolism (PE) is the third most common cardiovascular disorder in the United States and results in approximately 50000 deaths per year1. It is potentially
lethal if not recognized early and aggressively treated. According to autopsy studies and chart reviews, major emboli remain undiagnosed in 60–84% of patients who
experience PE2. Approximately 11% of patients with pulmonary embolism die within 1 h after the embolic event3. Mortality rates in the majority of unsuspected cases
who survive longer than 1 h is 30%, but with correct diagnoses and therapy, the mortality rate is reduced to 8%4. Our clinical challenge is to improve our ability to
diagnose and manage PE more rapidly and more effectively.
12.2 Risk factors

A number of variables place patients at increased risk for venous thromboembolism and hence pulmonary emboli (Table 12.1).
• Surgery and trauma. Prolonged venous stasis carries the highest risk; causes include venous valvular insufficiency, right heart failure, and prolonged immobility. Type
and length of surgery are determinants of deep venous thrombosis (DVT). Abdominal operations requiring general anesthesia for greater than 2 h increase risk for
DVT. The risk associated with surgery is higher with pelvic and leg procedures, hip replacement, hip fracture repair or knee replacement5 (Table 12.2). Measures
such as perioperative prophylactic therapy with heparin, low molecular weight heparin, or mechanical techniques such as compression stockings or intermittent
pneumatic compression can lower the risk. The risk of DVT in trauma patients is related to prolonged immobilization, severe head trauma, paralysis, pelvic and
lower extremity fractures, venous trauma, shock and multiple transfusions, with the risk greatest in the first 2 weeks postinjury6. With spinal cord injury, the
incidence of venous thromboembolism (VTE) is approximately 4% and that of PE about 5%.
• Age. Age is an important risk factor for VTE, with the incidence increasing exponentially over the age of 40. The risk of VTE is 46% even for younger patients (<30
years) with multiple trauma. Compared with healthy adults, patients over 30 years of age with any heart disease have a three times higher risk of PE, with this risk
increasing with patients in heart failure and arrhythmias, especially atrial fibrillation7.
• Indwelling catheters. Patients with right heart failure and those with indwelling right heart catheters are at increased risk for right ventricular mural thrombi and
pulmonary emboli. Central venous catheters are a risk factor for superior vena cava and upper extremity thrombosis.
• Hypercoagulable states. Genetic predisposi tion to thromboembolism is recognized more frequently now. Inherited abnormalities most commonly associated with
VTE are the factor V Leiden mutation (resistance to activated protein C), and the prothrombin G20210A mutation8. A workup
Table 12.1 Clinical risk factors for DVT/PE
Age >40 years • Nephrotic syndrome

• Prolonged immobilization • Inflammatory bowel disease
• Cancer • Estrogen use
• Previous DVT or PE • Acquired or hereditary disorders
• Obesity • Indwelling femoral catheters
• Varicose veins • Fractures of pelvis, hip, knee
• CHF • Major surgery of the abdomen, pelvis, lower extremities
• Stroke
Page 185
Table 12.2 Levels of thromboembolism risk in surgical patients without prophylaxis*
Levels of risk examples Calf Proximal Clinical Fatal Successful prevention

DVT, % DVT, % PE, % PE, % strategies
Low risk 2 0.4 0.2 0.002 No specific measures Aggressive
Minor surgery in patients<40 years with no additional risk factors mobilization
Moderate risk 10–20 2–4 1–2 0.1–0.4 LDUH q 12 h, LMWH, ES or
Minor surgery in patients with additional risk factors; nonmajor surgery in patients IPC
aged 40–60 years with no additional risk factors; major surgery in patients<40 years
with no additional risk factors
High risk 20–40 4–8 2–4 0.4–1.0 LDUH q 8 h, LMWH or IPC
Nonmajor surgery in patients>60 years or with additional risk factors; major surgery in
patients>40 years or with additional risk factors
Highest risk 40–50 10–20 4–10 0.2–5 LMWH, oral anticoagulants,
Major surgery in patients>40 years plus prior VTE, cancer or molecular IPC/ES+ LDUH/LMWH or
hypercoagulable state, hip or knee arthroplasty, hip fractrue surgery; major trauma; ADH
spinal cord injury
*Modified from Gallus, A.S., Salzman, E.W. and Hirsh, J. (1994) Prevention of VTE. In: Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 3rd Edn (eds R.W.Colman, J.
Hirsh, V.J.Marder, et a/.). Reproduced with permission from Lippincott, Williams & Wilkins.
LDUH, lowdose unfractionated heparin; LMWH, low molecular weight heparin; ES, elastic stockings; IPC, intermittent pneumatic compression; ADH, adjusted dose heparin.
for heritable causes is not indicated after the first episode of VTE particularly in those patients who have clinical risk factors. Hypercoagulability has been
implicated as a risk factor and includes inherited abnormalities such as deficiency of antithrombin III, protein C, protein S, hyperhomocysteinemia, and the
presence of antiphospholipid syndrome9.
• Obesity. Although obesity alone does not warrant VTE prophylaxis, it increases a patient's potential risk when other clinical factors are present.
• Varicose veins. The impact of varicose veins on VTE risk remains controversial. Varicose veins may be an additive risk factor, when they develop in patients
younger than 60 years10.
• Malignancy. Although the association between malignancy and VTE is well known, most patients with pulmonary emboli do not have cancer. Hypercoagulable
states secondary to advanced cancer are associated with a higher risk of VTE. The incidence of VTE is 2–3 times higher among patients undergoing surgery for
malignant disease than those operated on for nonmalignant conditions with other risk factors being equal11.
• Estrogen therapy. Estrogen therapy for prostate cancer may increase the incidence of VTE among men. In women,
Page 186
estrogen replacement therapy during perimenopausal and postmenopausal periods can be a significant risk factor12. Currently, a woman’s risk of a VTErelated
death from an estrogenbased contraceptive is lower than her risk of death from pregnancy.
• History of thromboembolism. A history of thromboembolism is one of the most important risk factors of VTE. The incidence of acute venous thrombosis is nearly
eightfold higher among hospitalized patients with a history of VTE compared with those without a history. Therefore, when undergoing surgery or with periods of
prolonged immobility or when in a hospital for serious medical reasons, patients with a history of VTE must receive aggressive prophylactic therapy. The incidence
of VTE is 15% among patients who require absolute bed rest for less than 1 week, and is higher among those in bed for longer periods13. Recently, an association
has been established between VTE and prolonged sitting position as in long automobile or airplane trips14.
VTE prophylaxis may not be justified by bed rest alone. DVT typically develops in the calf veins and does not typically progress to proximal vein thrombosis, which is
more dangerous. Commonly, calf vein thromboses are asymptomatic and usually appear to resolve without clinical sequelae once ambulation is resumed. However,
prophylaxis is required if other risk factors, as discussed above are also present.
Postoperative deep venous thromboses are usually clinically silent until they are dislodged to form pulmonary emboli. As venous thrombi can develop despite optimal
prophylactic regimens, the question remains whether highrisk patients should undergo postoperative screening tests for DVT. The clinical experience suggests that
although venous ultrasonography can actually diagnose proximal vein thrombosis, it appears to be less useful for detecting calf vein thrombosis.
12.3 Pathophysiology of pulmonary embolism

Most pulmonary emboli originate from the deep veins of the lower extremities. Other sites include the iliac and pelvic veins, and less commonly, the inferior vena cava.
Two primary pathologic processes explain the typical clinical presentations of PE. Obstruction of the pulmonary circulation can result in circulatory compromise with the
extent of the compensation depending on the site and size of the pulmonary embolism, and the previous circulatory reserve of the right heart and pulmonary vasculature.
Patients with previously normal pulmonary artery pressures and right heart pressures can withstand moderate or large occlusions of pulmonary arteries caused by PE
and still maintain adequate cardiac output. However, in the presence of existing pulmonary hypertension, right heart failure, or extensive clot involvement of vessels,
severe elevation of pulmonary artery pressures may result in acute right heart failure and subsequent hemodynamic instability. Furthermore, pulmonary artery
obstruction by thromboembolism results in hypoxemia. The mechanism of PE associated hypoxemia can be attributed to a combination of ventilationperfusion
mismatch, atelectasis, redistribution of pulmonary blood flow, and increased blood transit time. Infrequently, acute pulmonary hypertension secondary to PE may lead
to the reopening of a patent foramen ovale present in a minority of the population, leading to an intraatrial right to left shunt and severe refractory hypoxemia.
12.4 Differential diagnosis in the ICU

Making a diagnosis of PE in an ICU is challenging because of obstacles in transporting the patient to radiology for a spiral CT or ventilation perfusion scan. Treatment
may be complicated by the underlying or associated illnesses in the critically ill patient. Physical examination may be nonspecific, as a result
Page 187
of underlying obstructive pulmonary disease, pneumonia, congestive heart failure or other conditions. Clinical symptoms and signs are nonspecific (Table 12.3).
Except for chest pain and dyspnea, no symptoms are consistently associated with PE. Common physical findings of tachypnea and tachycardia may be suggestive, but
again these findings are nonspecific. Although the presence of a pleural rub or hemoptysis may be useful diagnostically, these signs are nonspecific and frequently
absent. The differential diagnosis is extensive, including pneumonia, asthma, myocardial infarction, pleuritis, pericarditis, aortic dissection, cardiac tamponade, and
costochondritis. Fixed splitting of the second heart sound, which indicates right ventricular overload and a flow murmur over the lung fields, is of major importance.
Rales are heard in approximately 60% of the patients with angiographically proven PE. Chest roentgenographic results, V/Q scanning, and Ddimer assay are less likely
to be normal in the absence of a PE, and more difficult to interpret in a critically ill patient in whom PE develops. Unless they can be explained by a more likely
diagnosis, dyspnea, tachypnea, and pleural chest pain occurring together in the presence of risk factors for DVT justifies a diagnostic evaluation for PE. Syncope and
cyanosis suggest severe massive embolism as signs and symptoms of right heart failure develop.
12.5 Diagnostic tests
Figure 12.1 presents the American Thoracic Society (ATS) algorithm for the evaluation of suspected PE15.
12.5.1 Arterial blood gas analysis

Respiratory alkalosis as a result of hypoxemia is frequently observed in dyspneic patients with PE. Blood gas analysis will demonstrate an abnormal alveolar arterial
oxygen gradient in most patients with PE, with oxygen tension usually less than 80 mmHg. However, patients with PE, especially those with a previously normal
cardiopulmonary state, may have a normal alveolar arterial O2 gradient16.
12.5.2 Ddimer assay

Recent trials suggest that the Ddimer enzymelinked immune serum assay (ELISA) may have a role in excluding PE in symptomatic outpatients17. The assay detects
degradation products of crosslinked fibrin and is very sensitive to activation of the fibrinolytic system. Use of this assay may reduce the use of angiography by 30%,
and combining the assay with noninvasive tests for PE might improve the negative predictive values and avoid invasive tests. This test is nonspecific, with its negative
predictive value for PE is high, but its positive predictive value is low. The assay may be positive even after mild trauma associated with ecchymosis. At this
Table 12.3 Pulmonary embolism manifestations
Symptoms Signs ECG Chest Xray findings

Dyspnea Tachypnea RV strain pattern Focal oligemia
Cough Pulmonary rales Atrial fibrillation Westermark sign
Hemotysis Fever Pwave enlargement Wedgeshaped density (Hampton hump)
Syncopy Tachycardia Q wave deepening (lead III)
Chest pain (pleuritic) Phlebitis T wave inversion (lead V1, V5)
Edema Ventricular ectopy Enlarged descending right pulmonary artery
Gallop rhythm S wave enlargement (lead I)
Palla sign
Page 188
Figure 12.1 Diagnostic algorithm for patients with symptoms suggesting acute pulmonary embolism. As with the approach to acute deep
venous thrombosis, the recommended diagnostic approach allows for some flexibility depending upon the resources at a
particular institution. PE, pulmonary embolism; V/Q, ventilationperfusion; US, compression ultrasonography; IPG, impedance
plethysmography; CV, contrast venography; MRI, magnetic resonance imaging; DVT, deep venous thrombosis. Adapted from
Tapson, V.F., Carroll, B.A., Davidson, B.L., et al. al. (1999) The diagnostic approach to acute venous thromboembolism.
Clinical practice guidelines. Reproduced with permission from American Thoracic Society. Am. J. Respir. Crit. Care Med.
160:1043–1066.
time, the Ddimer assay may be most useful for ruling out DVT and PE in the emergency department because a negative result is highly reliable. However, routine use
as a diagnostic test for PE is not recommended. The combination of a normal dead space ventilation/tidal volume fraction and normal Ddimer concentration was 100%
sensitive in excluding PE in a recent trial18.
12.5.3 Chest roentgenography

Chest Xray abnormalities include pulmonary infiltrates, pleural effusions, elevated hemidiaphragm and atelectasis, but may be normal. Hampton’s hump, Palla sign or
the Westermark sign may suggest the diagnosis, but are infrequent and can in no way be taken as evidence that a PE is the cause of the patient’s dyspnea.
12.5.4 EKG
Sinus tachycardia is the most common electrocardiographic abnormality with acute PE. The ECG may reveal Ppulmonale in massive PE. Right heart strain may
manifest right bundle branch block. Signs of acute right ventricular strain or the socalled S1Q3 (large S wave in lead 1, Q wave in lead 3) pattern are infrequently
encountered, and their absence is not reliable evidence that a VE is absent.
Page 189
12.5.5 V/Q lung scanning

The lung scan is the most commonly used test to diagnose PE. The scanning results is graded on the basis of perfusion defects in relation to lung ventilation as follows.
• Normal perfusion scan. Perfusion defect is not observed and effectively rules out clinically significant PE. However, in those cases where there is an overwhelming
clinical suspicion of PE with high risk factors and the perfusion scan is normal, one must consider performing pulmonary angiography.
• Highprobability scan. The presence of more than one large perfusion defect (>75% of the segment) with normal ventilation or ventilatory defect significantly smaller
than perfusion defects, or perfusion defects in an area of absent or smaller infiltrates on chest Xray, strongly suggests PE in a patient with the appropriate clinical
profile. A pulmonary angiogram should be considered for patients with a highprobability V/Q scan and lowprobability profile, such as absent risk factors for PE.
However, a positive compression ultrasonography obtained before the pulmonary angiogram obviates the need for an angiogram for most patients with a high
probability V/Q scan.
• Indeterminate perfusion scans. Low and intermediateprobability scans are sometimes grouped together and called indeterminate or nondiagnostic scans,
implying the need for additional studies. Scanning for most patients with suspected PE suggests low or intermediate probability. A lowprobability scan does not rule
out PE and necessitates further studies, using compression ultrasonography. If the test results are positive, anticoagulation therapy should be initiated if no absolute
contraindications exist. With negative lower extremity studies, clinical observation is appropriate for patients with a lowprobability V/Q scan. The negative
predictive value of a lowprobability scan increases when there is no strong clinical impression for a PE.
Intermediateprobability V/Q scans require further evaluation via pulmonary angiogram or other studies to confirm or exclude PE.
The most useful data regarding the utility of V/Q scanning for PE was from the PIOPED trial, the Prospective Investigation of Pulmonary Embolism Diagnosis19. This
study compared the sensitivity and specificity of V/Q scanning with that of angiography. The researchers also estimated the clinical likelihood of PE before the results of
the scanning and angiography were known. Approximately 1000 patients in whom PE was suspected underwent both V/Q lung scanning and pulmonary angiography.
In this study, a minority of patients (13%) had a highprobability V/Q scan. The most frequently seen results were intermediateprobability (39%) and lowprobability
(34%) scans. The findings of the study revealed the following.
• A normal or nearnormal perfusion scan made PE very unlikely. A highprobability V/Q scan strongly supports a diagnosis of PE with probability increasing with a
high clinical suspicion or in the absence of underlying cardiopulmonary disease.
• V/Q scanning did not yield highprobability findings for most patients with angiographic proven pulmonary emboli. Approximately 41% of patients with angiogram
proven PE had highprobability scans.
• More than 50% of documented PEs were in patients with intermediate or lowprobability scans. In patients with low or intermediateprobability scans the likelihood
of PE ranged from 14 to 31%.
• For most patients with suspected PE, the diagnosis cannot be ruled out with scanning alone. Clinical assessments were more accurate in excluding than in identifying
PE.
In summary, the PIOPED findings suggest that a highprobability scan has excellent positive predictive value but identifies fewer than half of all pulmonary emboli. If
low and intermediateprobability scans are considered
Page 190
negative, a significant number of cases will not be diagnosed. If every patient with abnormal V/Q scans were treated, many patients would be unnecessarily treated. For
most patients with suspected PE, V/Q scans must be complemented by further testing to confirm or rule out the diagnosis of PE.
12.5.6 Pulmonary angiography

Pulmonary angiography has been considered to be the gold standard for the diagnosis of pulmonary embolism. Indications include high clinical suspicion of DVT with
nondiagnostic studies, in situations where anticoagulation is considered high risk, and before embolectomy or thrombolytic therapy. The test provides information
about the size of the clot in approximately 20–30% of patients with clinically suspected PE and a nondiagnostic scan, and identifies pulmonary artery pressure
measurements20. The procedure is considered safe and when performed by experienced clinicians carries an estimated mortality of approximately 0.3% and morbidity
of 1–4% (arrhythmias, acute right heart failure and right ventricular perforation). The procedure increases the patient risk of bleeding when standard anticoagulation is
given. It also prevents administration of a thrombolytic agent because of the potential for severe bleeding from the puncture site. A study obtained within 1–2 weeks of
acute symptoms will reliably diagnose PE. The angiogram has a high sensitivity for emboli in the large central pulmonary arteries and a much lower sensitivity for small
peripheral emboli. Risks associated with pulmonary angiography are considerably overrated. Renal failure (creatinine≥2 mg dl–1) is considered an absolute
contraindication, because of the risk of the contrast agent. Recent myocardial infarctions, current arrhythmias, and severe pulmonary hypertension are relative
contraindications. Interobserver agreement in the PIOPED study was only 66% for subsegmental PE diagnosis using angiography. Interobserver agreement among
radiologists interpreting angiograms deteriorates for subsegmental and smaller PE. For these reasons, newer noninvasive methods of diagnosing PE are being
employed.
12.5.7 Helical (spiral) CT

New generation scanners that employ rapid sequential scanning can reliably identify a clot in the second to fourth division of pulmonary vessels. Sensitivity and
specificity of a contrast enhanced helical (spiral) CT scan for the diagnosis of PE vary in the literature, with some studies reporting sensitivities ranging from 53% to
100% and specificities ranging from 81% to 100%21. A normal helical CT scan result when the index of suspicion is high should be followed by pulmonary angiography
and the careful evaluation of the distal pulmonary vascular bed. Although helical CT may be inadequate to detect subsegmental emboli, it is unclear how often
subsegmental emboli occur in the absence of larger emboli. Outcome studies are being carried out to evaluate the use of spiral CT in patients with PE. Current
limitations of helical CT angiography include the requirement of an intravenous contrast agent and the unavailability of equipment and personnel needed to perform and
interpret the study at every hospital. Performing CT contrast venography in tandem with CT angiography may improve the diagnostic yield. Helical CT is useful in
patients with baseline V/Q scan abnormalities and provides information about potentially other disease processes in the lungs.
12.5.8 Compression ultrasonography

Noninvasive lower extremity testing with compression ultrasonography should be considered to evaluate for evidence of DVT when lung testing is nondiagnostic.
Although this test does not rule out pulmonary embolus, a large majority of pulmonary emboli arise from the femoral vein, and the treatment for both DVT and PE will
be the same. Therefore, a diagnosis of DVT provides grounds for initiating therapy for pulmonary embolus without further investigation. Compression ultrasound has a
sensitivity of 89–98% for proximal DVT22. The test’s accuracy decreases in
Page 191
asymptomatic patients and in those with chronic DVT or pelvic thromboembolic disease. A normal examination does not rule out PE when clinical suspicion is
moderately high. In patients at high risk for DVT, including those with pelvic or hip trauma, and critical illness, compression ultrasonography is the most sensitive and
specific lower extremity study. Impedance plethysmography test is highly accurate for diagnosing proximal deep vein thrombosis, but it is not widely available. Forms of
venous obstruction and congestive heart failure can produce false positive results.
12.5.9 Magnetic resonance angiography

Magnetic resonance angiography is a fast and accurate modality for detecting PE and does not involve nephrotoxic contrast agents. It may also be a noninvasive
alternative for venography to detect deep vein thrombosis. It is a promising technique that appears to be highly sensitive to pulmonary embolus. This technique has the
added advantage of distinguishing between acute and chronic thrombosis and scanning the lungs and the legs at the same time. The sensitivity of the magnetic resonance
technique was 100% in one study and its specificity was 95%23.
12.6 Treatment
Providing effective therapy for PE in the ICU setting is more complicated than it would be outside the ICU. ICU patients are more likely to have a number of
complicating conditions, including stress gastritis, ulcers, thrombocytopenia, anemia, and coagulopathies, or have sustained trauma or undergone a recent surgical
procedure.
The decision on whether to treat suspected PE is usually based on the results of the V/Q scan. A highprobability scan is an indication for treatment. A non
diagnostic scan should be followed by pulmonary angiography if the patient is hypotensive or severely hypoxemic or by bilateral leg evaluation (i.e. compression
ultrasonography) if the patient’s condition is stable. In those patients who have a normal perfusion scan, prophylactic therapy may be given where risk factors for DVT
or PE are present.
Specific treatment issues in the management of the PE include:
• anticoagulation therapy;
• low molecular weight heparin (LMWH) in treatment and prophylaxis;
• thrombolytic therapy;
• Inferior vena cava (IVC) filter;
• management of hypotension with pharmacologic support;
• surgical embolectomy.
12.6.1 Heparin
Heparin remains the mainstay of therapy for PE and DVT involving the proximal veins of the thigh or pelvic veins. Heparin inhibits thrombin by activating antithrombin.
Therapeutic heparinization will prevent additional clot formation; the body’s intrinsic fibrinolytic system should subsequently clear the existing clot. Once the diagnosis of
a possible lifethreatening PE is considered, heparin should be administered if there are no contraindications, while the patient undergoes further noninvasive evaluations
(Table 12.4). Weightadjusted protocols have been employed to enhance rapid and consistent anticoagulation and to minimize the risk of recurrent PE and bleeding
(Table 12.5)24. Patients receive 80 U kg−1 of heparin as a bolus, followed by an infusion of 18 U kg−1 per hour. The dose is then titrated to maintain the activated
partial thromboplastin time (APTT) more than 1.5 times control.
The APTT is monitored every 6 h until two consecutive therapeutic results are obtained. Subsequently, the APTT can be checked every morning. Additionally, a
platelet count is obtained at baseline and every other day until heparin infusion is stopped.
Heparininduced thrombocytopenia may cause arterial thrombosis. In those patients without a history of hemorrhagic diathesis, hemorrhagic side effects are rare.
Heparin
Page 192
Table 12.4 Guidelines for anticoagulation: unfractionated heparin*
Indication Guidelines
VTE suspected • Obtain baseline APTT, PT, CBC count
• Check for contraindications to heparin therapy
• Order imaging study, consider giving heparin 5000 IU IV
VTE confirmed • Rebolus with heparin, 80 IU kg−1 IV; start maintenance infusion at 181 U kg−1 h−1
• Check APTT at 6 h to keep APTT in a range that corresponds to a therapeutic blood heparin level
• Check platelet count between days 3 and 5
• Start warfarin therapy on day 1 at 5 mg and adjust subsequent daily dose according to INR
• Stop heparin therapy after at least 4–5 days of combined therapy when INR is >2.0
• Anticoagulate with warfarin for 3 months at an INR of 2.5; range: 2.0–3.0
−1
*For subcutaneous treatment with unfractionated heparin, give 250 IU kg SC q 12 h to obtain a therapeutic APTT at 6–8 h. PT, prothrombin time.
Table 12.5 Body weightbased dosing of IV heparin*
APTT, S† Dose change, IU kg−1 h−1 Additional action Next APTT, h

<35 (<1.2×mean normal) +4 Rebolus with 80 IU kg −1 6
35–45 (1.2–1.5×mean normal) +2 Rebolus with 40 IU kg−1 6
46–70 (1.5–2.3×mean normal) 0 none 6§
71–90 (2.3–3.0×mean normal −2 none 6
>90 (>3.0 Q4×mean normal) −3 Stop infusion for 1 h 6
* −1 −1 −1
Initial dosing: loading, 80 IU kg ; maintenance infusion; 18 IU kg h (APTT in 6 h).
†
Therapeutic range in seconds.
§During the first 24 h, repeat APTT every 6 h. Thereafter, monitor APTT once every morning unless it is outside the therapeutic range.
should not be given for longer than 5 days. Failing to achieve adequate anticoagulation in the first 24 h of treatment significantly increases the risk of recurrent emboli.
Therapy commonly requires 5 days of heparin and a warfarininduced prolongation of the international normalized ratio (INR) before discontinuation of heparin.
Within the first 24 h of heparin administration, warfarin is started with an initial dose of 5 mg day−1. The therapeutic goal is to reach an INR of 2–3. Warfarin acts by
inhibiting the vitamin K dependent factors II, VII, IX and X.
It is important to understand that anticoagulation with heparin and warfarin does not prevent the development of persistent pulmonary hypertension, alleviate
embolusinduced hemodynamic instability, or avoid subsequent valvular damage to deep veins of the lower extremities. The optimum duration of anticoagulation therapy
for DVT and pulmonary emboli remains debatable, with studies showing that 3–6 months of warfarin
Page 193
Table 12.6 Duration of Therapy*
3–6 months • First event with reversible† or timelimited risk factor (patient may have underlying factor V Leiden or prothrombin 20210A)
≥6 months • Idiopathic VTE, first event
12 months to lifetime • First event, recurrent idiopathic VTE are a continuing risk factor‡
*
All recommendations are subject to modification by individual characteristics including patient preference, age, comorbidity, and likelihood of recurrence.
†Reversible or timelimited risk factors: surgery, trauma, immobilization, estrogen use.
‡Proper duration of therapy is unclear in first event with homozygous factor V Leiden, homocystinemia, deficiency of protein C or S, or multiple thrombophilias; and in recurrent events
with reversible risk factors.
alone results in an acceptable low (<5%) frequency of DVT recurrence25. Patients with recurrent DVTs or high risk factors such as CHF or hypercoagulable states
should receive anticoagulation therapy for a longer period of time (Table 12.6).
Table 12.7 Bridge therapy protocol
(1) Obtain INR 1 week before procedure

(a) INR 2–3: discontinue warfarin 5 days before surgery
(b) INR 3–4.5: discontinue warfarin 6 days before surgery
(c) Enoxaparin 1 mg kg−1 SC, q 12 h starting 36 h after discontinuing warfarin
(2) Discontinue enoxaparin 12–18 h before surgery
(3) Restart enoxaparin 12–24 h postsurgery or once hemostasis has been achieved
(4) Restart warfarin at the previous maintenance dose. Dosing can be initiated at any time following the surgical procedure, as enoxaparin will provide the necessary
anticoagulation during the crossover period
(5) INR 2–3: Stop enoxaparin for 2 consecutive days
INR, international normalized ratio. in: Merli G.J. (2000) Prophylaxis for deep venous thrombosis and pulmonary embolism in the surgical patient. Reproduced with permission from Clinical
Cornerstone. 2:22.
12.6.2 Low molecular weight heparin (LMWH)

LMWH is as effective as unfractionated heparin for the treatment of hemodynamically stable patients with PE. LMWH has been employed in hip and knee surgery and
highrisk abdominal surgery. LMWH binds specifically to antithrombin and has advantages over unfractionated heparin, including its administration of the drug
subcutaneously, greater bioavailability in tissue, and more predictable anticoagulation. LMWH use requires no monitoring of APTT. Total costs may thus be lower
when laboratory testing over the course of therapy is considered. However, monitoring may be necessary in patients who are morbidly obese, have renal insufficiency,
or who weigh less than 50 kg. The incidence of heparininduced thrombocytopenia is extremely low among patients who receive LMWH compared with
unfractionated heparin. The disadvantages of LMWH in certain patients are due to its longer halflife and lesser reversibility with protamine compared with
unfractionated heparin. Choosing to use LMWH or unfractionated heparin depends on the clinical setting and the risk factors. In those patients scheduled to undergo
invasive procedures, unfractionated heparin may be preferred over LMWH. Protocols for bridging between warfarin and enoxaparin are available (Table 12.7)26.
Page 194
Table 12.8 Regimens to prevent VTE
Method Description
LDUH Heparin 5000 IU SC, given q 8–12 h starting 1–2 h before operation
ADH Heparin SC, given q 8 h starting at approximately 3500 IU SC and adjusted by ±500 SC, per dose to maintain a midinterval APTT at high normal
values
LMWH and General surgery, moderate risk:
heparinoids Dalteparin, 2500 IU SC, 1–2 h before surgery and once daily postop
Enoxaparin, 20 mg IU SC, 1–2 h before surgery and once daily postop
Nadroparin, 2850 SC 2–4 h before surgery and once daily postop
Tinzaparin, 3500 IU SC 2 h before surgery and once daily postop
General surgery, high risk:
Dalteparin, 5000 IU SC 8–12 h before surgery and once daily postop
Danaparoid, 750 IU SC 1–4 h before surgery and q 12 h postop
Enoxaparin, 40 mg IU SC, 1–2 h preop and once daily postop
Enoxaparin, 30 mg IU SC, q 12 h starting 8–12 h postop
Orthopedic surgery
Dalteparin, 5500 SC 8–12 h preop and once daily starting 12–24 h postop
Dalteparin, 2500 SC 6–8 h postop; then 5000 U SC daily
Danaparoid, 750 IU SC 1–4 h preop and q 12 h postop
Enoxaparin, 30 mg IU SC q 12 h starting 12–24 h postop
Enoxaparin, 40 mg SC once daily starting 10–12 h preop
Nadroparin, 38 IU kg−1 SC 12 h preop, 12 h postop, and once daily on
postop days 1, 2 and 3; then increase to 57 IU kg−1 SC once daily
Tinzaparin, 75 IU kg−1 SC once daily starting 12–24 h postop
Tinzaparin, 4500 IU SC 12 h preop and once daily postop
Major trauma
Enoxaparin, 30 mg SC q 12 h starting 12–36 h postinjury if hemostatically stable
Acute spinal cord injury
Enoxaparin, 30 mg SC q 12 h
Medical conditions
Dalteparin, 2500 IU SC once daily
Danaparoid, 750 IU SC q 12 h
Enoxaparin, 40 mg SC once daily
Nadroparin, 2850 IU SC once daily
Perioperative warfarin Start daily dose with approximately 5–10 mg the day after surgery; adjust the dose for a target INR of 2.5 (range 2–3)
IPC/ES Begin immediately before operation, and continue until fully ambulatory
Postop, postoperative; LDUH, lowdose unfractionated heparin; ADH, adjusted dose heparin; IPC, intermittent pneumatic compression; ES, elastic/graduated compression.
12.6.3 Thrombolytic therapy

Thrombolytics have been utilized most commonly in patients with massive PE and severe hemodynamic compromise. Proven benefits of thrombolytics include
improvements in right ventricular hemodynamics and pulmonary perfusion with reversal of right heart failure27. For patients with massive pulmonary emboli, right heart
failure, or hemodynamic instability, thrombolytic therapy may be potentially life saving. No study to date has shown a reduction in mortality with thrombolytics. In
Page 195
the presence of severe hemodynamic compromise secondary to a PE, and echocardiographic determination of acute pulmonary hypertension, thrombolytic therapy can
be employed.
The clearest indication for thrombolytics is hemodynamic instability, in the absence of absolute contraindications. Contraindications to thrombolytic therapy include:
current internal bleeding, recent trauma, major surgery or head injury (within the previous 6 weeks), acute cerebrovascular hemorrhage or a cerebrovascular procedure
within the previous 2–3 months. Other relative contraindications include: cerebrovascular event within 6 months, a 10 day or shorter postpartum period, recent organ
biopsy or puncture of a noncompressible vessel, uncontrolled bleeding diathesis, recent serious internal trauma, pregnancy, cardiopulmonary resuscitation with rib
fractures, thoracentesis, paracentesis, lumbar puncture, or any condition that places the patient at an increased risk for bleeding. Most bleeding episodes occur in
patients undergoing invasive procedures. Major bleeding complications for PE are approximately 14% in patients undergoing pulmonary angiograms, but only 4% in
patients having V/Q scans28. If bleeding occurs during therapy, discontinuation of treatment usually is all that is required to control bleeding. However, with ongoing
bleeding, cryoprecipitate infusion targeted to attain a fibrinogen level of 100 mg dl–1 or more is indicated. If bleeding persists, the administration of fresh frozen plasma
should be considered. In the presence of ongoing lifethreatening hemorrhage following thrombolytic therapy, antifibrinolytic therapy with εaminocaproic acid is
indicated.
There is no evidence that any one thrombolytic agent is superior to another. Recommended drugs and dosages are: recombinant tissue type plasminogen activator
(rtPA), 100 mg infused over 2 h; urokinase, 4400 IU kg−1 given over 10 min as a loading dose, then 4400 IU kg−1 h−1 infused for 12 h; streptokinase, 250000 IU
loading dose given over 30 min and 100000 IU h−1 infused for 24 h given via a peripheral intravenous site is less cumbersome to employ than the 12–24 h infusion
regimen.
12.6.4 IVC filters

Indications for IVC interruption include:
• contraindication to anticoagulant use;
• coagulationinduced hemorrhage;
• hemodynamic instability with absolute contraindication for thrombolytic therapy;
• recurrent thromboemboli owing to anticoagulation failure;
• chronic thromboembolic pulmonary hypertension.
Patients who may survive an initial massive PE must be considered for an IVC filter placement because death in this population may be a result of a relatively
small reoccurrence. Employment of an IVC should be based on evidence that thromboemboli are originating from the deep vein of the thigh or pelvic system.
The IVC filter can be placed under fluoroscopic guidance in the ICU safely and cost effectively, circumventing the need to transport the critically ill patients to
the radiology suite or operating room. Controversy exists on whether to administer anticoagulation therapy to patients with IVC filters in place. Most
investigators recommend that if filters are inserted because of recurrent pulmonary emboli during anticoagulation, anticoagulation is usually continued to prevent
additional thrombi from forming on the filter. The use of anticoagulation with IVC filters may increase the patency rate of filters over time and facilitate venous
drainage. Risk of recurrent PE after IVC interruption is approximately 2–3%.
12.6.5 Surgery and rheolytic thrombectomy

Criteria for pulmonary embolectomy should include patients with severe hemodynamic instability despite heparin and resuscitative efforts, massive PE, and failure of
thrombolytic therapy in patients. Emergency surgical thrombectomy is difficult to accomplish unless a cardiac surgical team, including perfusionist, is immediately
available. The mortality rate for
Page 196
surgical intervention is reported to be from 10% to 95%29.

Rheolytic thrombectomy is one of the newest methods for treating massive pulmonary embolus30. It involves the use of a catheter that is designed to remove
intravascular thrombi from vessels. Direct highpressure saline jets located at the catheter tip cause mechanical thrombolysis. Dissolving microparticles are aspirated
through the same catheter. Larger studies are needed to determine whether this treatment would be an alternative to thrombolysis in a select patient population.
Intraembolic infusion of thrombolyic therapy has been employed in several institutions where local thrombolysis of massive pulmonary emboli without systemic
fibrinogenolysis may be achieved by delivering lowdose thrombolytic therapy (20 mg rtPA) to the site of embolization in the pulmonary artery by means of catheter
infusion31. Intraembolic lowdose thrombolysis has been used successfully in critically ill patients with PE in situations where systemic thrombolysis is considered
contraindicated, but its efficacy has yet to be demonstrated in randomized, controlled trials.
12.7 Right ventricular function

Right ventricular (RV) dysfunction secondary to right heart failure often causes death in patients with acute massive pulmonary emboli. Therefore, it is important to
evaluate the effects of the embolus on RV function. The hemodynamic response to acute PE depends upon embolic size, the presence of preexisting cardiopulmonary
disease and, to a lesser extent, the magnitude of the neurohumoral response to emboli. In the presence of underlying cardiopulmonary disease, a relatively small
embolus can have significant consequences. The latter explains why only 50% of fatal emboli are anatomically massive.
The hemodynamic results of acute emboli are a reduction in the crosssectional area of the pulmonary circulation resulting in increased resistance to blood flow
through the lungs. Physiologically, PE increases right ventricular work or afterload, best measured by assessing mean pulmonary artery pressure or calculating
pulmonary vascular resistance with a pulmonary artery catheter. This increase in mean pulmonary artery pressure is directly related to the extent of embolization. In the
absence of preexisting underlying cardiopulmonary disease, patients with emboli have marked levels of pulmonary hypertension when at least 50% of pulmonary
vasculature is occluded. Mean pulmonary artery pressure begins to rise at 25% obstruction of the vascular tree, and continues to rise with increased degrees of
obstruction to reach a maximum value of about 40 mmHg in the absence of preexisting pulmonary hypertension. Normally, the RV being a thinwalled chamber cannot
accommodate high pulmonary artery pressures (PAP) and increased RV work. Therefore, it predisposes to rapid contractile failure. Right ventricular dysfunction
becomes even more apparent when there is an elevation in the mean right atrial pressure. It is usually associated with a mean PAP of 30 mmHg and obstruction of at
least 35–40% of the pulmonary vascular tree. As right atrial pressure increases concomitantly with PAP, an initial compensatory cardiovascular response is an increase
in heart rate to maintain cardiac output. When the right atrial pressure rises by 10 mmHg and mean PAP is greater than 30–40 mmHg, right ventricular failure occurs
and cardiac output will decrease.
In summary, available data suggest that the normal right ventricle is unable to generate a mean PAP greater than 40 mmHg. With excessive resistance to blood flow
from PE, this causes the RV to dilate and become hypokinetic, causing a reduction in right ventricular stroke volume. This decline in right ventricular stroke volume
contributes to a decrease in LV preload and cardiac output when combined with RV dilation. Echocardiography can determine the degree of pulmonary hypertension
in
Page 197
patients with acute PE. Pulsed continuous wave and color Doppler techniques can estimate right ventricular systolic pressure and systolic PAP, demonstrate the
presence of tricuspid and pulmonary regurgitation, and determine flow velocities, thereby permitting estimation of right ventricular systolic pressures and systolic PAP32.
Transesophageal echocardiography has been employed to assess for central pulmonary artery thromboemboli in patients with risk factors for PE and right heart
strain33.
12.8 Other clinical interventions
12.8.1 Oxygen therapy and mechanical ventilation

Oxygen therapy is administered to augment oxygen saturation and reduce pulmonary vascular resistance caused by hypoxemia. Continuous positive airway pressure
may be used in patients with focal atelectasis. Intubation and mechanical ventilation may be required for hypoxemic respiratory failure.
12.8.2 Fluid resuscitation (See Chapter 7)

In the presence of PEinduced hypotension accompanied by impaired right ventricular contractility, initial volume expansion is indicated. However, excessive fluid
resuscitation may lead to severe RV dysfunction and high pulmonary artery pressures, subsequently causing right heart failure. The use of volume expansion and
vasoactive agents to manage the hemodynamic instability in critically ill patients with PE has not been studied in randomized, controlled trials. Clinical experience
suggests that pharmacologic hemodynamic support does improve blood pressure long enough to administer thrombolytic therapy or any other therapy in the acute
management of PE. Fluid resuscitation may increase myocardial wall stress, which may decrease the oxygen supply demand ratio, causing ischemia and/or right
ventricular infarction. Therefore, it is recommended that volume infusion should be titrated to maintain a systolic blood pressure of 90 mmHg.
12.8.3 Vasoactive agents (See Chapter 8)

Epinephrine or other alpha agents are indicated for therapy of massive PE to reverse significant hypotension from reduced cardiac output. These vasoactive agents
attempt to restore normal RV hemodynamics via increased RV coronary blood flow, improved RV contractility and reduced pulmonary vascular resistance.
Norepinephrine has been shown to produce consistent hemodynamic improvement and improve cardiac contractility through its beta receptor effect. It can enhance
coronary artery blood flow and right ventricular perfusion pressure. Dobutamine may be an adjunct to norepinephrine; it reduces both systemic vascular resistance and
pulmonary vascular resistance through vasodilation of respective vascular beds. Dobutamine has been shown to increase cardiac output, decrease pulmonary vascular
resistance and improve oxygenation in patients with PEinduced pulmonary hypertension34. Combination of norepinephrine and dobutamine may be the best current
regimen of vasoactive therapy of hemodynamically unstable patients with pulmonary embolus. Doses should be titrated to preserve organ perfusion.
12.9 Risk reduction
12.9.1 DVT prophylaxis

The incidence of DVT in general surgery is approximately 20–70%. Recommended prophylaxis is tailored to the risk classification (see Table 12.2). General surgery
patients at low risk do not require prophylaxis, except for early ambulation. Lowdose heparin (5000 IU SC every 12 h), LMWH, intermittent pneumatic compression
sleeves or gradient elastic stockings are to be given to moderaterisk surgical patients. In highrisk general surgery, lowdose heparin (5000 IU SC every 8 or 12 h) or
LMWH is recommended. Placement of a vena caval filter as prophylaxis for major surgery is recommended in patients with significant cardiopulmonary disease and in
patients who cannot receive any form of prophylaxis following highrisk surgical procedures. In very
Page 198
highrisk surgical patients, external pneumatic compression devices and lowdose heparin (5000 U every 8–12 h) or LMWH are recommended.
12.9.2 Bridge therapy

For patients on chronic anticoagulation scheduled for invasive surgical procedures, Table 12.7 lists the recommendations on the appropriate management of
anticoagulants in the perioperative period. A recent study suggests that bridge anticoagulation therapy should depend on the medical condition and the time elapsed
since the thromboembolic event35. Those with acute venous thromboembolism or acute arterial embolism less than 1 month after the event should receive bridge
therapy. Those with acute thromboembolism between 2 and 3 months after the event should be placed on anticoagulation, postoperatively. All other chronically
anticoagulated patients do not require bridge therapy.
12.9.3 Regional anesthesia

Epidural or spinal hematomas have been described following LMWH use. Regional anesthesia should be avoided in patients who are receiving drugs that affect
hemostasis, and those having multiple spinal attempts with hemorrhagic aspirates. When DVT/PE prophylaxis is initiated before surgery, spinal needle insertion should
be postponed for 10–12 h after the initial LMWH prophylactic dose is administered36. In those situations where an epidural anesthesia catheter is in place postsurgery
for 24–36 h, LMWH prophylaxis should be delayed for 12 h after catheter removal. Patients should be monitored for early signs of cord compression secondary to
epidural bleeding (numbness, weakness, bowel/bladder dysfunction).
12.9.4 Trauma
LMWH is the most efficacious option for most highrisk trauma patients. Current contraindications to early LMWH prophylaxis include intracranial bleeding,
incomplete spinal cord injury associated with perispinal hematoma, ongoing, uncontrolled bleeding, and uncorrected coagulopathy. Lacerations or contusions of the
lungs, liver, spleen or kidneys, or the presence of retroperitoneal hematoma associated with pelvic fractures do not by themselves contraindicate the use of LMWH
prophylaxis, as long as there is no active bleeding. Mechanical modalities such as elastic stockings and intermittent pneumatic compression can be used in the presence
of LMWH contraindications.
12.9.5 Acute spinal cord injury

The highest risk of DVT in hospitalized patients is for patients with acute spinal cord injuries. PE remains the third most common cause of death in this population.
LMWH is recommended for prophylaxis; lowdose unfractionated heparin, elastic stockings, and intermittent pneumatic compression are relatively ineffective when used
alone.
12.9.6 Nonsurgical patients

Although there are a limited number of trials addressing prophylaxis for the nonsurgical population, medical patients with risk factors for DVT (cancer, bed rest, heart
failure, severe lung disease) should receive LMWH or lowdose unfractionated heparin. For patients with ischemic stroke, LMWH or lowdose unfractionated heparin
is recommended.
References
1. Anderson, F.A., Wheeler, H.B. Goldberg, R.J., et al. (1991) A populationbased perspective of the hospital incidence and casefatality rates of deep vein
thrombosis and pulmonary embolism: The Worcester DVT study. Arch. Intern. Med. 151:933–938.
2. Dalen, J.E. and Alpert, J.S. (1975) Natural history of pulmonary embolism. Prog. Cardiovasc. Dis. 17:259–270.
3. Tapson, V.F. and Witty, L.A. (1995) Massive pulmonary embolism. Diagnostic and therapeutic strategies. Clin. Chest Med. 16:329–340.
4. Goldhaber, S.Z. (1998) Pulmonary thromboembolism. N. Engl J. Med. 339:93–104.
5. Gallus, A.S., Salzman, E.W. and Hirsh, J. (1994) Prevention of VTE. In: Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 3rd Edn
Page 199
(eds R.W. Colman, J. Hirsh, V.J. Marder, et al.). J.B. Lippincott, Philadelphia, PA, pp. 1331–1345.
6. Geerts, W.H., Code, K.I., Jay, R.M., et al. (1994) A prospective study of venous thromboembolism after major trauma. N. Engl J. Med. 331:1601–1604.
7. Coon, W.W. (1976) Risk factors in pulmonary embolism. Surg. Gynecol. Obstet. 143:385–390.
8. Ridker, P.M., Miletich, J.P., Stampfer, M.J., et al. (1995) Factor V Leiden and risks of recurrent idiopathic venous thromboembolism. Circulation 92:2800–2802.
9. Ridker, P.M., Hennekens, C.H., Selhub, J., et al. (1997) Interrelation of hyperhomocyst(e)emia, factor V Leiden, and risk of future venous thromboembolism.
Circulation 95: 1777–1782.
10. Nicolaides, A.N. and Irving, D. (1975) Clinical factors and the risk of deep venous thrombosis. In: Thromboembolism Etiology. Advances in Prevention and
Management (ed. A.N. Nicolaides). University Park Press, Baltimore, MD, pp. 193–204.
11. Rahr, H.B. and Sorenson, J.V. (1992) Venous thromboembolism and cancer. Blood Coagul. Fibrinolysis 3:451–460.
12. Devor, M., BarrettConnor, E., Renvall, M., et al. (1992) Estrogen replacement therapy and the risk of venous thrombosis. Am. J. Med. 92: 275–282.
13. Gibbs, N.M. (1957) Venous thrombosis of the lower limbs with particular reference to bedrest. Br. J. Surg. 45:209–236.
14. Ferrari, E., Chevallier, T., Chapelier, A. and Baudouy, M. (1999) Travel as a risk factor for venous thromboembolic disease: A casecontrolled study. Chest
115:440–444.
15. Tapson, V.F., Carroll, B.A., Davidson, B.L., et al. (1999) The diagnostic approach to acute venous thromboembolism. Clinical practice guideline. American
Thoracic Society. Am. J. Respir. Crit. Care Med. 160:1043–1066.
16. Tapson, V.F. and Fulkerson, W.J. (1994) Pulmonary embolism in the intensive care unit. J. Intensive Care Med. 9:119–131.
17. Van Beek, E.J., Schenk, B.E., Michel, B.C., et al. (1996) The role of plasma Ddimers concentration in the exclusion of pulmonary embolism. Br. J. Haematol
92:725–732.
18. Kline, J.A., Meek, S., Boudrow, D., et al. (1997) Use of the alveolar dead space fraction (VD/VT ) and plasma Ddimers to exclude acute pulmonary embolism in
ambulatory patients. Acad. Emerg. Med. 4:856–863.
19. PIOPED Investigators (1990) Value of the ventilation/perfusion scan in acute pulmonary embolism. Results of the prospective investigation of pulmonary embolism
diagnosis (PIOPED). JAMA 263:2753–2759.
20. Stein, P.D., Athanasoulis, C., Alavi, A., et al. (1992) Complications and validity of pulmonary angiography in acute pulmonary embolism. Circulation 85:462–468.
21. Rathbun, S.W., Raskob, G.E. and Whitsett, T.L. (2000) Sensitivity and specificity of helical computed tomography in the diagnosis of pulmonary embolism: a
systematic review. Ann. Intern. Med. 132:227–232.
22. Lensing, A.W., Prandoni, P., Brandjes, D., et al. (1989) Detection of deepvein thrombosis by realtime Bmode ultrasonography. N. Engl. J. Med. 320:342–345.
23. Meaney, J.F., Weg, J.G., Chenevert, T.L., et al. (1997) Diagnosis of pulmonary embolism with magnetic resonance angiography. N. Engl J. Med. 336:1422–
1427.
24. Raschke, R.A., Reilly, B.M., Guidry, J.R., et al. (1993) The weightbased heparin dosing nomogram compared with a ‘standard care’ nomogram. A randomized,
controlled trial. Ann. Intern. Med. 119:874–881.
25. Goldhaber, S.Z. (1999) Treatment of pulmonary thromboembolism. Intern. Med. 38:620–625.
26. Spandorfer, J., Lynch, S., Weitz, H., et al. (1999) Use of enoxaparin for the chronically anticoagulated patient before and after procedures. Am. J. Cardiol.
84:478–480.
27. Hyers, T.M., Agnelli, G., Hull, R.D., et al. (1998) Antithrombotic therapy for venous thromboembolic disease. Chest 114(Suppl.): 561S–578S.
28. Dalen, J.E., Alpert, J.S. and Hirsch, J. (1997) Thrombolytic therapy for pulmonary embolism: Is it safe? When is it indicated? Arch. Intern. Med. 157:2550–2556.
29. Dalen, J.E. and Hirsch, J. (eds) (1998) Fifth ACCP Consensus Conference on Antithrombotic Therapy. Chest 114(Suppl.):439S–769S.
30. Koning, R., Cribier, A., Gerber, L., et al. (1997) A new treatment for severe pulmonary embolism: percutaneous rheolytic thromboectomy. Circulation 96:2498–
2500.
31. Goldhaber, S.Z., Kessler, C.M., Heit, J., et al. (1992) Recombinant tissuetype plasminogen activator versus a novel dosing regimen of urokinase in acute
pulmonary embolism: a randomized, controlled, multicenter trial. J. Am. Coll Cardiol. 20:24–30.
32. Metz, D., Chapoutot, L., Ouzaan, J., et al. (1991) Doppler echocardiographic assessment of the severity of acute pulmonary embolism: A correlative angiographic
study in fortyeight adult patients. Am. J. Noninvasive Cardiol. 5: 223–228.
Page 200
33. Ritto, D., Sutherland, G.R., Samuel, L., et al. (1993) Role of transesophageal echocardiography in diagnosis and management of central pulmonary artery
thromboembolism. Am. J. Cardiol. 71:1115–1118.
34. Layish, D.T. and Tapson, V.F. (1997) Pharmacologic hemodynamic support in massive pulmonary embolism. Chest 111:218–224.
35. Kearon, C. and Hirsh, J. (1997) Management of anticoagulation before and after elective surgery. N. Engl. J. Med. 336:1506–1511.
36. Horlocker, T. and Heit, J. (1997) Low molecular weight heparin: biochemistry, pharmacology, perioperative prophylaxis regimens, and guidelines for regional
anesthetic management. Anesth. Analg. 85:874–885.
Page 201
Chapter 13
Modes of mechanical ventilation
Contents
13.1 Introduction
13.2 Volume versus pressure targeting
13.3 Traditional modes of ventilation
13.4 Pressure control inverse ratio ventilation (PCIRV)
13.5 Airway pressure release ventilation (APRV)
13.6 Combined modes of ventilation
Page 202
13.1 Introduction
The available modes on mechanical ventilators have markedly expanded over the last 20 years. Before 1980 ICU ventilators provided only volumetargeted modes.
Today’s newest mechanical ventilators not only incorporate modes that are volume targeted but also provide modes that are pressure targeted as well as a new group
of modes most of which are considered combined pressure and volumetargeted modes. Because with all of these modes feedback regarding the actual ventilatory
pattern, they result in automatic adjustment of the way gas is delivered, these modes may more precisely be called closedloop, computercontrolled modes of
ventilation. This chapter will address global differences between pressure and volumetargeted modes, all of the standard modes of ventilation and the newest
computercontrolled modes of ventilation. Where possible, indications for each mode will be highlighted; however, with many modes few or no data are available to
guide the choice of mode.
13.2 Volume versus pressure targeting
With all new generation ICU ventilators the clinician must decide not only on a specific mode but also whether gas delivery is to be pressure or volume targeted1. In
volumetargeted modes, tidal volume is constant,
Figure 13.1 Flow, pressure, and volume waveforms during constant flow, volume A/C Note the nearly square flow waveform,
representing a constant flow until the preset volume is delivered.
Page 203
whereas in pressuretargeted modes, maximum peak alveolar pressure is constant. Volumetargeted modes use a simple, constant flow, whereas pressuretargeted
modes require computercontrolled variable flow generators.
13.2.1 Volume ventilation

Figure 13.1 depicts the pressure, flow, and volume waveforms during constant flow, volume ventilation. With the start of each breath, airway pressure increases from
baseline in essentially a linear manner, as gas flow is constant. Volume is delivered in a similar linear manner and the total tidal volume is not delivered until the moment
inspiration ends. With this approach, VT, flow rate, and inspiratory time as well as I:E ratio can be adjusted over a wide range. What stays constant is VT. Peak airway
and peak alveolar pressure can vary considerably from breath to breath as impedance to ventilation or ventilatory drive changes. In volume ventilation some ventilators
allow a decelerating (‘reverse ramp’ waveform) flow to be selected (Figure 13.2). With a decelerating flow, flow linearly decelerates from a peak level to the terminal
level. Many ventilators end the breath when the flow is at a low predetermined level. When changing the inspiratory flow waveform from square to decelerating it is
important to consider how a particular model of ventilator responses to the change. Ventilators can either maintain peak flow, and allow the decelerating flow to
increase inspiratory time (all else constant), or the ventilator can maintain inspiratory time constant, causing peak flow to increase2. During volume ventilation inspiratory
time and peak flow are related as follows:
Figure 13.2 Flow, volume, and pressure waveforms during decelerating flow, volume A/C with an endinspiratory pause, The
decelerating (reverse ramp) flow waveform allows a more constant inspiratory pressure waveform.
Page 204
(1)
where is peak flow, TI is inspiratory time, and f is endinspiratory flow2.

As discussed in Chapter 1, peak pressure in constant flow, volume ventilation reflects the combined effects of compliance and resistance. In the passively ventilated
patient peak alveolar pressure can be determined by setting an inflation hold of 0.5 s or less.
Critical alarms and monitors during volume ventilation differ from those in pressure ventilation. With volume ventilation peak airway pressure is the most critical
variable to monitor, whereas with pressure ventilation tidal volume is the most critical monitored variable2. The importance of this different approach to monitoring is
illustrated during the expansion of a tension pneumothorax. With volume ventilation, peak pressure increases with every breath, causing the highpressure alarm to
sound and continue to sound with each breath. During pressure ventilation, the airway pressure does not vary but VT decreases with each breath. The lowVT alarm
begins to sound and continue with each breath. The maximum pressure of the tension pneumothorax is also limited in pressure ventilation to the set ventilatory pressure.
13.2.2 Pressure ventilation

Figure 13.3 depicts the airway pressure, flow, and volume waveforms during pressure ventilation. As noted, these differ considerably from those seen with
conventional volume ventilation (Figure 13.1). The primary reason for this difference is the pressure target. The ventilator delivers high flow to rapidly drive airway
pres
Figure 13.3 Flow, volume, and pressure waveform during pressure A/C where inspiratory flow justy returns to zero before the end of
inspiration.
Page 205
sure to the target level. However, once the target is reached flow must decrease to avoid pressure exceeding the target. As a result, flow may decrease in a linear or in
an exponential manner. Generally, the greater the impedance to ventilation the more likely flow will decelerate in an exponential manner3. As inspiratory time is set in
pressure assist/control ventilation, with very stiff lungs or chest wall, flow can decrease to zero before the end of the inspiratory phase, resulting in a plateau appearance
of the volume wave (Figure 13.4).
Figures 13.3 and 13.4 depict the key characteristics of pressure ventilation. Because flow decreases as inspiratory time progresses, the majority of the tidal volume
is delivered early in the inspiratory phase. If flow decreases to zero before the end of inspiration, peak airway pressure is equal to peak alveolar pressure2. In addition,
with pressure ventilation delivered flow varies with patient demand; thus, as demand increases, flow increases. This is contrary to what happens in volume ventilation,
where peak flow and waveform are constant regardless of patient demand.
As demonstrated in Figures 13.3 and 13.4, tidal volume during pressure control is affected by inspiratory time3. In patients breathing spontaneously, inspiratory time
is set to equal the patient’s spontaneous inspiratory time. However, during controlled ventilation in ARDS patients, inspiratory time should be set to ensure that flow
decreases to zero followed by a short 0.1–0.3 s inspiratory pause. Allowing flow to return to zero ensures that VT is maximized at the specific pressure setting. Figure
13.5 demonstrates that increasing inspiratory time increases VT to a maximum but
Figure 13.4 Flow, volume, and pressure waveform during pressure A/C ventilation when inspiratory flow returns to zero with an end
inspiratory pause caused by the lengthy inpiratory time. This is the same patient as in Figure 13.3 but with inspiratory time
lengthened.
Page 206
further increases past a short plateau result in a decreased tidal volume3. The actual VT, and inspiratory time ensuring maximum VT, is dependent upon compliance,
resistance and autoPEEP. The greater the airway resistance or autoPEEP and the lower the compliance, the lower the VT at a set pressure level and inspiratory time.
Whenever adjustments are made to the pressure control setting and VT decreases, autoPEEP should be considered. As the ventilator does not measure autoPEEP
with each breath it cannot compensate for the level of autoPEEP3. In volume ventilation autoPEEP increases peak airway and peak alveolar pressure. However, with
pressure ventilation, as maximum pressure is targeted, autoPEEP results in a decreased VT2. Airway pressure goes from the set baseline to the autoPEEP level
without any gas delivery. Essentially, autoPEEP decreases the pressure control level by the autoPEEP level, thus decreasing driving pressure and VT
Figure 13.5 Relationship of inspiratory time fraction to tidal volume (pressure target=20 cmH2O). Wheh inspiratory (RI) and expiratory
(RE) resistance are equal, optimal duration (D)=0.5. When RI>RE, more inspiratory time is required, and optimal D>0.5.
Conversely, when RE>RI, optimal D<0.5. From: Marini, J.J., Crooke, P.S, and Truitt, J.P., (1989) Determinants and limits of
pressurepreset ventilatlon: a mathematical model of pressure control. Reproduced with permission from the American
Physiological Society. J. Appl, Physiol. 67:1081–1092.
Few data are available to distinguish the use of pressure and volume ventilation during control ventilation. The decision to use one is based on the clinician’s desire to
primarily maintain peak alveolar pressure or VT constant. However, it is possible to maintain peak alveolar pressure constant with volume ventilation or tidal volume
constant with pressure ventilation but considerable monitoring and adjustment is required. In ARDS where current data indicate that a 6 ml kg−1 VT should be targeted,
either approach can be used4. If volume ventilation is used, a decelerating flow pattern does result in better distribution of ventilation5. Pressure ventilation, with its
variable flow delivery, better meets the changing demand of spontaneously breathing patients than volume ventilation6,7. Pressure ventilation is always preferable in the
spontaneously breathing patient, especially if demand is rapidly changing. Table 13.1 compares pressure and volume ventilation.
13.3 Traditional modes of ventilation

Modes of ventilation have traditionally been classified based on the level of patient interaction with the mechanical ventilator: control, assist/control (A/C), and assist
ventilation8. Synchronized intermittent mandatory ventilation (SIMV), developed in the 1970s, is also considered a traditional ventilatory mode. All
Table 13.1 Pressure vs. volume assist/control ventilation
Pressure Volume
Tidal volume Variable Constant
Peak alveolar pressure Constant Variable
Peak inspiratory pressure Constant Variable
Minimum rate Preset Preset
Inspiratory time Preset Preset
Peak flow Variable Constant
Flow pattern Variable Constant
Page 207
of these modes on the newest generation of mechanical ventilators can be set for pressure or volume ventilation.
13.3.1 Control
The control mode implies that the patient does not participate in the process of gas delivery. Today, full support is accomplished by providing appropriate sedation;
however, with some older ventilators the control mode prevented the ventilator from responding to the patient’s demand. On today’s ventilators control has been
replaced by assist/control1.
13.3.2 Assist/control
With assist/control the patient has the opportunity to trigger the mechanical breath. However, if the rate at which the patient triggers the ventilator falls below the set
backup rate, a control breath is delivered8. When the patient resumes inspiratory effort, he or she can again trigger the ventilator. Pressure control on the newer
ventilators is an assist/control mode (PA/C); the patient can trigger each breath as in volume assist/control. As pressure control was introduced in conjunction with
inverse ratio ventilation9,10 many have considered this mode only when inverse ratio was used. More recently, pressure control has been used on all types of patients
including those intubated11 and those receiving noninvasive positive pressure ventilation12.
With both control and assist/control VT or pressure level, and flow waveform and flow rate or inspiratory time must be set. In addition, the sensitivity must be set
appropriately to allow patient triggering without having to generate excess negative inspiratory pressures.
13.3.3 Assist
True assist ventilation implies that no backup rate is available. Volumeassisted ventilation is not available on any currently manufactured ventilators. Pressureassisted
ventilation is generally termed pressure support (PS). Although most ventilators do not provide a backup rate during pressure support, apnea beyond a set time period
converts pressure support to a backup ventilation mode, either pressure or volume targeted1. Pressure support is the mode of ventilation that coordinates best with
patient demand. Patients control not only initiation of inspiration but also termination of the breath.
With pressure support a breath is terminated by one of three mechanisms: (i) patient inspiratory flow decelerates to a minimal level; (ii) system pressure exceeds set
pressure; (iii) a specific time interval is exceeded. The terminal inspiratory flow cycling criterion is either a fixed flow (i.e. 5 l min−1) or a percentage of peak flow (i.e.
25% or 5%)8. Peak system pres sure, after the first 200–400 ms of inspiration, exceeding the set level by about 2–20 cmH2O (depending on the manufacturer) also
causes the breath to cycle to exhalation. With most manufacturers an inspiratory time exceeding 2–3 s results in termination of an assisted breath.
13.3.4 Inspiratory cycling criteria

As illustrated in Figure 13.6, problems with pressure support and patient synchrony can occur at the onset and termination of inspiration13. Initial flow can be either
excessive or inadequate and the criteria to terminate inspiration can be too low or too high a flow. Inspiratory cycling criteria are a potential problem in COPD
patients14,15. Many COPD patients choose to end inspiration when their peak inspiratory flow is high (Figure 13.7). If the flow is higher than the ventilator’s cycling
criteria, the patient must force airway pressure above the set level by activating accessory muscles of exhalation15. If this occurs, patient ventilatory drive is increased
and dyssynchrony is common. Some new generation ventilators (Puritan Bennett 840 and Hamilton Galileo) allow for adjustment of the inspiratory termination criteria1.
However, if the patient’s endinspiratory flow changes the adjustment may again be inadequate. Future generations of ventilators may automatically adjust cycling
criteria as the patient’s endinspiratory flow changes. This ongoing adjustment is currently being performed on some of
Page 208
Figure 13.6 Design characteristics of a pressuresupported breath. In this example, baseline pressure (i.e. PEEP) is set at 5 cmH2O and
pressure support is set at 15 cmH2O (PIP 20 cmH2O). The inspiratory pressure is triggered at point A by a patient effort
resulting in an airway pressure decrease. Demand valve sensitivity and responsiveness are characterized by the depth and
duration of this negative pressure. The rise to pressure (line 8) is provided by a fixed high initial flow delivery into the airway.
Note that if f flows exceed patient demand, initial pressure exceeds set level (B, 1), whereas if flows are less than patient
demand, a very slow (concave) rise to pressure can occur (B, 2), The plateau of pressure support (line C) is maintained by
servo control of f flow. A smooth plateau reflects appropriate responsiveness to patient demand; fluctuations would reflect less
responsiveness of the servo mechanisms. Termination of pressure support occurs at point D and should coincide with the end
of the spontaneous inspiratory effort. If termination is delayed, the patient actively exhales (bump in pressure above plateau)
(D, 1); if termination is premature, the patient will have continued inspiratory efforts (D, 2). From: MacIntyre, N., Nishimura,
M., Usuada, Y., et al. (1990) The Nagoya conference on system design and patientventilator interactions during pressure
support ventilation. Reproduced with permission from Chest 97:1463–1467,
the bilevel pressure ventilators used for noninvasive ventilation. If endinspiratory pressure exceeds set inspiratory pressure target, the set cycling flow criteria are too
low. Inspiratory
Figure 13.7 Esophageal pressure (Pes, continuous line) and estimated recoil pressure of the chest wall (Pescw, dashed line) tracings in a
patient receiving pressure support ventilation of 20 cmH2O. Pressure tracings have been superimposed so that Pescw is equal
to Pes at the onset of the rapid fall in Pes during late expiration. Times at which the Pes tracing is is higher than Pescw
represent lower bound expiratory effort, Note the presence of expiratory muscle activation during late inspiration. From: Jubran,
A., van de Graff, W. and Tobin, M.J, (1995) Variability of patientventilator interaction with pressure support ventilation in
patients with chtonic obstructive pulmonary disease. Reproduced with permission f from Am. J. Respir. Crit. Care Med.
152:129–136.
termination criteria should be slowly increased until a smooth transition to exhalation occurs.
The other option during endinspiratory dyssynchrony is to change modes from pressure support to pressure assist/control. As noted from Table 13.2, the only
difference between pressure support and pressure assist/control is the mechanism used to terminate inspiration, (flow (PS) vs time (PA/C)), otherwise gas delivery is
identical16. Thus, PA/C set at the same pressure level with inspiratory time set to equal the patient’s desired inspiratory time (≤1.0 5) results in better patient comfort
and synchrony than PS when the cycling criteria in PS do not match the patient’s endinspiratory flow.
Page 209
Table 13.2 Pressure A/C vs. PS
Pressure assist/control Pressure support

Target Pressure Pressure
Inspiratory time Set Patient determined
Rise time Variable Variable
End inspiration Set Variable
Peak f low Variable Variable
Rate Patient determined Patient determined
Backup rate Set Apnea ventilation
13.3.5 Rise time

The speed with which flow increases driving pressure from baseline to set level during PS or PA/C may not meet patient demand13. Many patients in severe respiratory
distress demand a rapid rise in pressure to peak levels, whereas other patients not markedly stressed prefer a more gradual rise in pressure17. Rise time allows
variability in slope of the flow delivery curve. Rise time adjustment is available during both PS and PA/C on most new generation mechanical ventilators. That is, the
clinician can match the flow delivery slope to the patient’s demand17–19. However, as with cycling criteria, rise time setting may need to be changed as patient demand
changes19. Again, it can be expected that on future generations of ICU ventilators adjustment of rise time will be automated. Proper setting results in a linear increase in
the airway pressure curve with pressure not exceeding set level at the onset of inspiration or the pressure curve being concave. Pressure overshoot indicates rise time is
too rapid, whereas a concave pressure curve indicates rise time is too slow.
13.3.6 Synchronized intermittent mandatory ventilation (SIMV)

As illustrated in Figure 13.8, SIMV is a combination of mechanical breaths and spontaneous unassisted breathing. The mechanical ventilation rate is set, with a window
of time based on set rate during which patients can trigger a mechanical breath. If the breath is not triggered during the time window a control breath is delivered.
Between time windows, the patient can breath spontaneously without a mechanical breath. As with the other traditional modes, the mandatory breath can be either
pressure or volume targeted.
Much concern has been raised regarding patient effort and synchrony during lowrate SIMV20–22. Because the respiratory center essentially is programmed by the
current breath, patient effort with low rate SIMV (≤6 min−1) is no different during the mechanical breaths and spontaneous unassisted breaths20,21 and patient effort
during the mechanical breath increases with decreasing SIMV rate (Figure 13.9)20,21. It is impossible to distinguish mandatory from spontaneous breaths when
observing the diaphragmatic electromyograph (EMG) and esophageal pressure tracings22. Leung et al. demonstrated that adding pressure support to the spontaneous
breaths can markedly modify workloads (Figure 13.10)21. However, the rationale of adding PS must be questioned, for it would appear that a more appropriate
application of ventilatory support would be the use of assist/control or pressure support without SIMV. In general, patient effort during SIMV is equivalent to
assist/control or pressure support if the mandatory rate accounts for 80% or more of the patient’s minute volume20.
13.4 Pressure control inverse ratio ventilation (PCIRV)

The use of the pressure assist/control mode in the late 1980s developed in association with
Page 210
Figure 13.8 Flow, pressure, and volume waveforms during volumetargeted SIMV without pressure support.
PCIRV9,10. With PCIRV (Figure 13.11) inspiration is lengthened until it exceeds expiration, in some applications establishing an I:E ratio up to 4:1. The goal with
PCIRV is to maintain a lengthy inspiratory phase to open collapsed lung and a short expiratory phase to avoid collapse of opened lung. In doing so, high levels of auto
PEEP are established. Proponents of PCIRV argue that high mean airway pressures can be maintained without high peak alveolar pressures23. The opponents of
PCIRV argue that the autoPEEP created is maldistributed to units least in need of PEEP and that the high mean airway pressures compromise hemodynamics24,25. A
number of controlled comparison of PCIRV with normal I:E ratio ventilation in both animal models25 and humans26–28 indicate no benefit from PCIRV in relation to
oxygenation but a marginal (about 5%) improvement in PCO2.
When PCIRV is used progression to an inverse ratio should be gradual. Once the patient is stable in PA/C, assure that the patient is not spontaneously breathing.
Fighting the long inspiratory times results in marked hemodynamic compromise and deteriorating gas exchange. Then slowly increase inspiratory time until the
oxygenation target is achieved. Generally, increases in inspiratory time of 0.1–0.25 s can be made without hemodynamic compromise; however, a positive fluid balance
is commonly needed if the I:E ratio exceeds 1.5:1 or 2:1. In general, the same targeted oxygenation status can be achieved with conven
Page 211
Figure 13.9 Inspiratory work per unit volume (work per liter, W L−1) done by the patient during assisted cycles (open bars) and
spontaneous cycles (solid bars). W L−1 increased monotonically with decreasing SIMV percentage for both types of breath. W
L−1 for spontaneous breaths tended to exceed that for machineassisted breaths, Asterisk indicates P<0.01; ns, not significant.
From: Marini, J.J., Smith, T.C. and Lamb, V.J. (1988) External work output and force generation during synchronized
intermittent mechanical ventilation. Effect of machine assistance on breathing effort. Reproduced with permission from Am.
Rev. Respir. Dis. 138:1169–1178.
tional I:E ratios (Chapter 15) and appropriately applied PEEP without the problems inherent in lengthy inspiratory times.
13.5 Airway pressure release ventilation (APRV)
APRV is very similar conceptually to PCIRV but differs in respect to the patient’s ability to breath spontaneously8. As noted in Figure 13.12, with APRV two levels of
continuous positive airway pressure (CPAP) are set, establishing an inverse I:E ratio as in PCIRV, but as the two airway pressure levels are established by setting
alternating levels of CPAP, the patient can breath spontaneously at each CPAP level. In the mechanical ventilators that offer
Figure 13.10 Cumulative pressuretime product of both triggering and nontriggering iggering efforts (PTP min−1) increased as the level
of pressure support (PS) alone was decreased, as intermittent mandatory ventilation (IMV) alone was decreased, and as IMV
was decreased in the presence of PS 10 cmH2O (P<0.0005 in each instance). At proportional levels of ventilator assistance,
PTP min−1 was not different during IMV alone and PS alone, but both were higher than that of combined IMV and PS 10
cmH2O. From; Leung, P., Jubran, A, and Tobin, M J. (1997) Comparison of assisted ventilator modes on triggering, patient
effort, and dyspnea. Reproduced with permission from Am. J. Respir, Crit. Care Med. 155:1940–1947.)
APRV, the inspiratory (high CPAP) level is set independent of the expiratory (low CPAP) level and the exhalation valve is active throughout the inspiratory and
expiratory phases, allowing spontaneous breathing regardless of CPAP level1. In some ventilators pressure support can be added to all spontaneous breaths1.
As with PCIRV, the use of APRV is highly controversial, with limited data in patients with ARDS or acute lung injury indicating a benefit to APRV versus
conventional ventilation29,30. However, because patients are allowed to breath spontaneously throughout the ventilatory cycles, much less sedation is required during
APRV than during conventional mechanical ventilation30.
Page 212
Figure 13.11 Flow, pressure, and volume waveforms during pressure control inverse ratio ventilation (I:E 3:1).
13.6 Combined modes of ventilation

A number of ventilator manufacturers have attempted to develop modes of ventilation that include the beneficial aspects of both pressure and volume ventilation31, that
is, they target both a maximum peak alveolar pressure (allow for variable flow delivery based on patient demand) and a tidal volume. Many of these modes can be
considered extensions of the mandatory minute ventilation (MMV) concept of ventilation initially described by Hewlett et al. in 197732. During MMV the ventilator
insures that minute volume remains at or above a set target. If the patient can maintain minute volume spontaneously no mechanical breaths are applied. However, if the
patient becomes apneic then controlled ventilation at preset levels is applied. Patients can also share the level of ventilation with the ventilator anywhere between these
two extremes. The intent is to allow the patient to wean themselves by applying the least amount of ventilatory support required to maintain the target minute volume.
Unfortunately, this approach does not work clinically. Patients can assume an inappropriate ventilatory pattern (rapid and shallow), exceeding the target, or simply
establish a level of comfort with the provided ventilatory support and not continue to assume a greater percentage of total support.
As listed in Table 13.3, these new modes of ventilatory support can be listed as modes that adjust breath delivery within a given breath or adjust breath delivery on
the subsequent
Page 213
Figure 13.12 Theoretic spirometric tracing depicting change in lung volume that would occur in a patient with SIMV compared with the
change that would occur with APRV. Inspiratory lung volume is the lung volume during spontaneous inspiration with CPAP
during APRV. Expiratory lung volume is the lung volume during release of Paw during APRV (i.e. lung volume after
mechanical expiration). Expiratory lung volume is similar to FRC during APRV, FRC is the passive expiratory lung volume
during APRV and is greater than FRC during ring Intermittent positive pressure ventilation (IPPV). (Used with permission,
from Stock, M.C., Downs, J.B. and Frolicher, D.A. (1987) Airway pressure release ventilation. Crit. Care Med. 15:462–466.)
Table 13.3 Computer controlled modes of ventilation
Withinbreath adjustment Betweenbreath adjustment

Automatic tube compensation Volume support
Volumeassured pressure support Pressureregulated volume control
Automode
Adaptive support ventilation
breath. It should be noted that few data are currently available to support the use of any of these modes of ventilation. As a result, their benefits and risks can be
discussed only theoretically. Forms of these modes are currently available on many of the newest generation of mechanical ventilators.
13.6.1 Volume support (VS)

This mode of ventilation can best be described as volumetargeted pressure support31. Target VT and maximum pressure support level are set. The ventilator
automatically adjusts the pressure support level up to the maximum pressure setting to insure that the volume target is meet. As noted in Figure 13.13, this is
accomplished on the Seimens Servo 300 by providing test breaths at 10 cmH2O pressure support, then calculating the pressure support level required to achieve the
targeted VT Following this, the ventilator delivers three test breaths at 75% of the calculated pressure support level. On subsequent breaths, the ventilator adjusts the
pressure support level a maximum of 3 cmH2O per breath to achieve and maintain the targeted tidal volume.
Volume support is intended to insure a specific tidal volume is delivered during changing impedance to ventilation and ventilatory drive,
Page 214
Figure 13.13 Pressure and flow waveforms illustrating volume support mode. See text for details.
or system leaks. This mode does accomplish this task. In small children and neonates with uncuffed endotracheal tubes and variable leak, volume support is able to
maintain a constant tidal volume33. However, changes in patient demand for ventilation based on developing sepsis, fever, hypoxemia, pain or anxiety may result in
diminished levels of pressure support31,34. If, because of heightened ventilatory demand, VT increases at a given pressure support level, the ventilator decreases
provided pressure support. As a result, in settings where patients may need more support, VS may provide less34. At this time, there are no published papers
establishing the benefits of VS. However, it may end up being most clinically indicated in children with uncuffed endotracheal tubes and varying leaks.
13.6.2 Pressureregulated volume control (PRVC)

PRVC is similar to VS except it is pressure A/C with a volume target28. A maximum pressure level and target VT are set. In addition, a respiratory rate and inspiratory
time are set. As with VS, a 10 cmH2O test breath is delivered then pressure control level is adjusted to 75% of maximum required. On subsequent breaths, pressure
control level is adjusted by up to 3 cmH2O. This mode may be the ideal mode for a lung protective ventilatory strategy in ARDS, as it targets both VT and peak
alveolar pressure. However, as with volume support, there are no data whatsoever to evaluate its effectiveness and to define indications for its use.
13.6.3 Automode
This is a recently introduced mode of ventilatory support on the Seimens Servo 300 ventilator. The mode functions by converting from a controlled mode of ventilatory
support to an assisted mode based on the patient’s spontaneous breathing capabilities31. If the patient triggers the ventilator on two consecutive breaths the ventilator
shifts to an assist mode and when the patient fails to trigger a breath for a specified time period the ventilator shifts to a control mode. The Servo 300 can be set in
adult, pediatric, and neonatal operating ranges. In the adult range, 12 s of apnea results in a shift to the control mode, whereas 8 s in the pediatric range and 5 s in the
neonatal range result in a shift to the control mode. In automode the Servo 300 can be set to shift
Page 215
between ‘volume control and volume support’, ‘pressure control and pressure support’, or ‘PRVC and volume support’.
It is difficult to conceptualize the benefits of automode over standard modes of ventilation, as each of the ‘control modes’ is actually an assist/control mode.
However, this mode clearly moves the operation of the ventilator further along the continuum of closedloop controlled ventilation. Unfortunately, as with VS and
PRVC, there are no data available to assist our application of this mode or to identify its indications.
13.6.4 Automatic tube compensation (ATC)

This approach to computerized control of ventilation is intended to provide sufficient pressure support during inspiration and sufficient decompression of the ventilator
circuit during exhalation to maintain the tracheal pressure at the PEEP level (constant)31. This mode is based on the work of Guttmann et al.35, who showed that
tracheal pressure change in intubated patients could be predicted by knowledge of the size of the endotracheal tube and continual measurement of inspiratory flow.
Ventilators incorporating this mode (Drager Evita 4, PuritanBennett 840) have programmed the resistive properties of varioussized endotracheal tubes and are able to
continuously measure flow1. As a result, they can vary the level of pressure support applied to maintain tracheal pressure according to the following formula:
(2)
where P is pressure and tube coeff is the resistance factor of the endotracheal tube. The proposed advantages of ATC are the elimination of the endotracheal tube as a
source of imposed work of breathing, improvement in patientventilator synchrony, and reduction of air trapping by decompression of the ventilator circuit reducing
expiratory resistance. However, the control algorithm prevents expiratory circuit pressure from falling below zero.
A number of groups have demonstrated the efficacy of ATC compared with pressure support36,37. ATC results in the use of lower pressure support levels in
postoperative patients without COPD but higher pressure support levels in patients with COPD than those set by clinicians.
Secretions and kinks or bends in the endotracheal tube decrease the efficiency of ATC.
Whether the endotracheal tube is a clinical concern for increased work of breathing in adults is hotly debated. Early reports indicated a high imposed endotracheal
tube workload38,39. However, more recent data suggest this is not the case. Strauss et al.40 reported that the endotracheal tube accounts for only 10% of the total
work of breathing, and Esteban et al.41 showed no difference between spontaneous breathing trials with 7 cmH2O PS versus a Tpiece. However, ATC appears to be
a better method of applying PS to a patient requiring ventilatory support. It eliminates the guesswork regarding the level of support required and varies the PS applied
based on patient demand not on a fixed VT target. It should be recalled that spontaneously breathing individuals vary VT on a breathtobreath basis based on an
integration of all signals received by the respiratory center.
13.6.5 Adaptive pressure support (APS)

This is a mode of ventilation available only on the Hamilton Galileo ventilator. APS is based on the minimal work of breathing concept developed by Otis et al. in
195042. Essentially, APS forces a ventilatory pattern that results in the least patient effort to overcome compliance and airways resistance31. As a result, it should be
considered a betweenbreath closedloop ventilation mode not intended to specifically wean patients but to minimize patient effort while maintaining oxygenation and
acidbase balance. This process can be expressed mathematically as
(3)
where RC is the respiratory time constant, B is the breathing rate, alveolar ventilation, and
Page 216
dead space31. APS is set by inputting the patient’s ideal body weight. Based on this, the ventilator calculates the patient ideal minute ventilation using 100 ml min−1
kg−1 for adults and 200 ml min−1 kg−1 for children. The clinician also selects the percentage of the ideal minute volume the mandatory mechanical breaths provide (20–
200% of ideal minute ventilation). Between the mandatory breaths the patient can breathe spontaneously. Once activated, the first few breaths are test breaths
measuring compliance, resistance and autoPEEP using the leastsquares fit technique43. The ventilator selects a respiratory frequency, inspiratory time, I:E ratio and
pressure limit, and measures these variables on a breathtobreath basis, altering them as required to meet the target ventilatory pattern. The ventilator adjusts
inspiratory time and I:E ratio to minimize autoPEEP, based on the ventilator’s calculation of the expiratory time constant. If the patient is not spontaneously breathing
the ventilator selects the respiratory rate, VT, and pressure limit required to deliver the tidal volume, inspiratory time, and I:E ratio. As the patient begins spontaneous
breathing the number of mandatory breaths decreases and the ventilator selects a PS setting that establishes a VT sufficient to insure alveolar ventilation based on a
dead space of 2.2 ml kg43.
A number of initial reports indicate that the APS can provide adequate ventilation in both anesthetized and spontaneously breathing patients, and matches the level of
gas exchange established by clinicianset parameters44–46. Clearly, sufficient data exist to support APS ability to automatically select ventilatory parameters and make
changes in response to patient’s effects and lung mechanics. However, whether this approach will effect the length of mechanical ventilation is still unknown. It is difficult
at this time to identify a clinical role for APS but conceptually it appears sound.
13.6.6 Volumeassured pressure support (VAPS)

The final mode discussed is VAPS (Bird 8400 Sti and TBird, Thermo Respiratory Group) or pressure augmentation (PA, Bear 1000 Thermo Respiratory Group)31.
This mode combines a minimum VT target with the variable inspiratory flow characteristic of pressure ventilation (Figure 13.14)47. During VAPS the clinician sets
minimum VT, squarewave flow and a pressure target or pressure support level. A VAPS breath may be either a control or a patienttriggered breath. Once triggered,
the ventilator attempts to reach the target pressure as quickly as possible. When the pressure is reached the ventilator determines the volume delivered by the ventilator,
compares this with the target volume and determines if the minimum volume can be reached. As noted in Figure 13.14, the overall configuration of the pressure and
flow waveforms during VAPS can vary markedly dependent on the patient demand. If patient’s demand is low, the breath is essentially a volumetargeted squarewave
flow breath, except for the potential initial boost of flow to rapidly establish target pressure31. If the patient’s demand is great, the breath becomes a pressure support
breath with the potential of exceeding the target VT if the patient demands a larger VT. In this case, inspiration ends like any other pressure support breath. Additionally,
anything between these two extremes is also possible.
Choosing the proper flow and pressure setting is critical to successful use of VAPS. If pressure is too high or the volume target is too low, all breaths tend to become
PS breaths, negating the volume guarantee. On the other hand, if the set flow is too high all breaths tend to become volume breaths and inspiratory time can be
prolonged.
Data from both Amato et al.47 and Haas et al.48 demonstrate less patient effort and better synchrony with VAPS as compared with volume ventilation. This
approach to volume targeting during pressure support does not appear to have the inherent problems seen with volume support, and although not necessarily ideal does
appear to be a very reasonable approach to providing pressure support with a volume guarantee.
Page 217
Figure 13.14 Representative tracings of esophageal pressure (Pes) and volume during standard volumeassisted ventilation (VAV) and
volumeassured pressure support ventilation (VAPSV) and the corresponding Campbell diagrams. The hatched areas represent
the time interval during which the work diagram ram was plotted volume versus esophageal pressure, Pes versus VT ). Area A
presents the work performed to Inflate the lung, and area B is the work performed to distend the chest wall. The use of
VAPSV resulted in a decreased work of breathing, an increased tidal volume, and a lower endexpiratory Pes, indicating less
intrinsic PEEP From: Amato, M.B.P., Barbos, C.S.V., Bonassa, J., et al. (1992) Volume assisted pressure support ventilation
(VAPSV): A new approach for reducing muscle workload during acute respiratory failure. Reproduced with permission from
Chest 102:1225–1234.
References
1. Chatburn, R.L. (1999) Mechanical ventilation. In: Respiratory Care Equipment (eds R.D. Branson, D.R.Hess and R.L.Chatburn). Lippincott, Williams and Wilkens,
New York, pp. 395–527.
2. Hess, D.R. and Kacmarek, R.M. (1996) Essentials of Mechanical Ventilation. McGrawHill, New York, pp. 17–25.
3. Marini, J.J., Crooke, P.S. and Truitt, J.P. (1989) Determinants and limits of pressurepreset ventilation: a mathematical model of pressure control. J. Appl. Physiol
67:1081–1092.
4. ARDS net (2000) 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. 342: 1301–1308.
5. Knelson, J., Howatt, W. and DeMuth, G. (1970) Effects of respiratory pattern on alveolar gas exchange. J. Appl Physiol. 29:328–331.
6. Cinnella, G., Conti, G., Lofasco, F., Lorino, H., Harf, A., Lemaire, F. and Brochard, L. (1996) Effect of assisted ventilation on the work of breathing: Volume
controlled versus pressure controlled ventilation. Am. J. Respir. Crit. Care 153:1028–1035.
Page 218
7. McIntyre, N.R., McConnell, R., Cheng, K.C. and Sane, A. (1997) Patientventilator flow dyssynchrony: flowlimited versus pressurelimited breaths. Crit. Care
Med. 25:1671–1677.
8. Kacmarek, R.M. and Hess, D. (1994) Basic principles of ventilator machinery. In: Principles and Practices of Mechanical Ventilation (ed. M.J. Tobin). McGraw
Hill, New York, pp. 65–110.
9. Tharratt, R.S., Allen, R.P. and Albertson, T.E. (1988) Pressure controlled inverse ratio ventilation in severe adult respiratory failure. Chest 94: 755–762.
10. Abraham, E. and Yosihara, G. (1989) Cardiorespiratory effects of pressure controlled inverse ratio ventilation in severe respiratory failure. Chest 96:1356–1359.
11. Amato, M.B.P., Barbas, C.S.V., Medeiros, D.M., et al. (1998) Effect of a protectiveventilation strategy on mortality in the acute respiratory distress syndrome.
N. Engl. J. Med. 338:347–354.
12. Calderini, E., Confalonieri, M., Puccio, P.G., Francavilla, N., Stella, L. and Gregoretti, C. (1999) Patientventilator asynchrony during noninvasive ventilation:
the role of expiratory trigger. Intensive Care Med. 25:662–667.
13. MacIntyre, N., Nishimura, M., Usuada, Y., et al. (1990) The Nagoya conference on system design and patientventilator interactions during pressure support
ventilation. Chest 97:1463–1467.
14. Jubran, A., van de Graff, W. and Tobin, M.J. (1995) Variability of patientventilator interaction with pressure support ventilation in patients with chronic obstructive
pulmonary disease. Am. J. Respir. Crit Care Med. 152:129–136.
15. Parthasarathy, S., Jubran, A. and Tobin, M.J. (1998) Cycling of inspiratory and expiratory muscle groups with the ventilator in airflow limitation. Am. J. Respir.
Crit. Care Med. 158: 1471–1478.
16. Williams, P., Muelver, M., Kratohvil, J., Ritz, R., Hess, D.R. and Kacmarek, R.M. (2000) Pressure support and pressure assist/control: are there differences?
An evaluation of the newest intensive care unit ventilators. Respir. Care 45: 1169–1181.
17. MacIntyre, N.R. and Ho, L. (1991) Effects of initial flow rate and breath termination criteria on pressure support ventilation. Chest 99: 134–138.
18. Bonmarchand, G., Chevron, V., Chopin, C., et al. (1996) Increased initial flow rate reduces inspiratory work of breathing during pressure support ventilation in
patients with exacerbation of chronic obstructive pulmonary disease. Intensive Care Med. 22:1147–1151.
19. Bonmarchand, G., Chevron, V., Menard, J., et al. (1999) Effects of pressure ramp slope values on the work of breathing during pressure support ventilation in
restrictive patients. Crit. Care Med. 27:715–718.
20. Marini, J.J., Smith, T.C. and Lamb, V.J. (1988) External work output and force generation during synchronized intermittent mechanical ventilation. Effect of
machine assistance on breathing effort. Am. Rev. Respir. Dis. 138: 1169–1178.
21. Leung, P., Jubran, A. and Tobin, M.J. (1997) Comparison of assisted ventilator modes on triggering, patient effort, and dyspnea. Am. J. Respir. Crit. Care Med.
155:1940–1947.
22. Imsand, C., Feihl, F., Perret, C. and Fitting, J.W. (1994) Regulation of inspiratory neuromuscular output during synchronized intermittent mandatory ventilation.
23. Lain, D.C., DiBenedetto, R., Morris, S.L., Van Nguyen, A., Sautlers, R. and Causey, D. (1989) Pressure control inverse ratio ventilation as a method to reduce
peak inspiratory pressure and provide adequate ventilation and oxygenation. Chest 95:1081–1088.
24. Kacmarek, R.M., Kirmse, M., Nishimura, M., Mang, H. and Kimball, W.R. (1995) The effects of applied versus autoPEEP on local lung unit pressure and
volume in a fourunit lung model. Chest 108:1073–1079.
25. Mang, H., Kacmarek, R.M., Ritz, R., Wilson, R.S and Kimball, W.R. (1995) Cardiorespiratory effects of volume and pressure controlled ventilation at various
I:E ratios in an acute lung injury mode. Am. J. Respir. Crit. Care Med. 151:731–736.
26. Bandolese, R., Broseghini, C., Polese, G., Bernasconi, M., Bandi, G., MilicEmili, J. and Rossi, A. (1993) Effects of intrinsic PEEP on pulmonary gas
exchange in mechanically ventilated patients. Eur. Respir. J. 6:358–363.
27. Mercat, A., Graini, L., Teboul, J.L., Lenique, F. and Richard, C. (1993) Cardiorespiratory effects of pressure controlled ventilation with and without inverse ratio
in the adult respiratory distress syndrome. Chest 104:871–875.
28. Lessard, M.R., Guerot, E., Lorino, H., Demaire, F. and Brochard, L. (1994) Effects of pressurecontrolled with different I:E ratios versus volumecontrolled
ventilation on respiratory mechanics, gas exchange, and hemodynamics in patients with adult respiratory distress syndrome. Anesthesiology 80:983–991.
29. Stock, M.C., Downs, J.B. and Frolicher, D.A. (1987) Airway pressure release ventilation. Crit. Care Med. 15:462–466.
30. Sydow, M., Burchardi, H., Ephraim, E., et al. (1994) Long term effects of two different ventilatory modes on oxygenation in acute lung
Page 219
injury. Comparison of airway pressure release ventilation and volumecontrolled inverse ratio ventilation. Am. J. Respir. Crit. Care Med. 149: 1550–1556.
31. Hess, D.R. and Branson, R.D. (2000) Ventilators and weaning modes. Respir. Care Clin. N. Am. 6:407–435.
32. Hewlett, A.M., Platt, A.S. and Terry, V.G. (1977) Mandatory minute volume. Anesthesia 32:163–169.
33. Keenan, H.T. and Martin, L.D. (1997) Volume support ventilation in infants and children: Analysis of a case series. Respir. Care 42:281–287.
34. Sottiaux, T.M. (2001) Patientventilator interactions during volume support ventilation: Asynchrony and tidal volume instability—a report of three cases. Respir. Care
46:255–262.
35. Guttmann, J., Eberhard, L., Fabry, B., et al. (1993) Continuous calculation of intratracheal pressure in tracheally intubated patients. Anesthesiology 79:503–513.
36. Fabry, B., Haberthur, C., Zappe, D., et al. (1997) Breathing pattern and additional work breathing in spontaneously breathing patients with diff erent ventilatory
demands during inspiratory pressure support and automatic tube compensation. Intensive Care Med. 23:545–552.
37. Guttmann, J., Bernhard, H., Mols, G., et al. (1997) Respiratory comfort of automatic tube compensation and inspiratory pressure support in conscious humans.
Intensive Care Med. 23: 1119–1124.
38. Shapiro, W., Wilson, R.K., Casar, G. et al. (1986) Work of breathing through different sized endotracheal tubes. Crit. Care Med. 14: 1028–1031.
39. Wright, P.E., Marini, J.J. and Bernard, G.R. (1989) In vitro versus in vivo comparison of endotracheal tube airflow resistance. Am. Rev. Respir. Dis. 140:10–16.
40. Strauss, C., Louis, B., Isabey, D., et al. (1998) Contribution of the endotracheal tube and the upper airway to breathing workload. Am. J. Respir. Crit. Care Med.
157:23–30.
41. Esteban, A., Alia, I, Gordo, F., et al. (1997) Extubation outcome after spontaneous breathing trials with Ttube or pressure support ventilation. Am. J. Respir. Crit.
Care Med. 156: 459–463.
42. Otis, A.B., Fenn, W.O. and Rahn, H. (1950) Mechanics of breathing in man. J. Appl. Physiol. 2:592–603.
43. Iotti, G.A., Brashi, A., Brunner, J.X., et al. (1998) Respiratory mechanics by least squares fitting in mechanically ventilated patients: Applications during paralysis
and during pressure support ventilation. Intensive Care Med. 21: 406–413.
44. Weiler, N., Eberle, B. and Heinrichs, W. (1998) Adaptive lung ventilation (AVL) during anesthesia for pulmonary surgery: Automatic response to transitions to and
from onelung ventilation. J. Clin. Monit. 14: 245–252.
45. Laubscher, T.P., Frutiger, A., Fanconi, S., et al. (1994) Automatic selection of tidal volume, respiratory frequency and minute ventilation in intubated ICU patients
as start up procedure for closedloop controlled ventilation. Int J. Clin. Monit. Comput. 11:19–30.
46. Campbell, R.S., Sinamban, R.P., Johannigman, J.A., et al. (1998) Clinical evaluation of a new closed loop ventilation mode: Adaptive support ventilation (abstract).
Respir. Care 43: 856.
47. Amato, M.B.P., Barbos, C.S.V., Bonassa, J., et al. (1992) Volume assisted pressure support ventilation (VAPSV): A new approach for reducing muscle workload
during acute respiratory failure. Chest 102:1225–1234.
48. Haas, C.F., Branson, R.D., Folk, L.M., et al. (1995) Patient determined inspiratory flow during assisted mechanical ventilation. Respir. Care 40:716–721.
Page 220
Page 221
Chapter 14
The acute respiratory distress syndrome
C.Allen Bashour, MD
and James K.Stoller, MD
Contents
14.1 Introduction
14.2 Historical perspective
14.3 Pathogenesis
14.4 Clinical presentation
14.5 Treatment
14.6 Prognosis and outcome
14.7 Conclusions
Page 222
14.1 Introduction
The acute respiratory distress syndrome (ARDS) is a common clinical condition that affects both medical and surgical patients. It carries a high mortality rate that has
changed only modestly since it was first described over 30 years ago. The true incidence of ARDS has been difficult to determine because the definition of the
syndrome has only recently been standardized. Promising advances in the treatment of ARDS continue to be evaluated but the benefit of these treatments has not been
clearly established, and management strategies have only recently affected mortality. This chapter provides an overview of the definitions, epidemiology, pathogenesis,
clinical features, established and evolving treatment, and prognosis in this important clinical syndrome.
14.2 Historical perspective
The acute respiratory distress syndrome was known as ‘sudden pulmonary collapse’ and ‘traumatic wet lung’ during World Wars I and II, respectively1. The first
description of ARDS by Ashbaugh et al. appeared in 19672. Twelve patients described in that report developed acute dyspnea, tachypnea, cyanosis refractory to
maximal oxygen therapy, decreased lung compliance, and diffuse bilateral pulmonary infiltrates on chest radiographs, and their mortality rate was 50%.
14.2.1 Definitions
In 1982, Petty provided a definition of the ‘adult respiratory distress syndrome’3. In 1988, Murray et al. developed a fourpart Lung Injury Score (Table 14.1) based
on the level of positive endexpiratory pressure (PEEP), the ratio of partial pressure of arterial oxygen (PaO2) to fraction of inspired oxygen (FiO2), the static lung
compliance, and the extent of pulmonary infiltrates observed on the chest radiograph, to quantify the severity of lung injury4. Although this Lung Injury Score cannot be
used to predict outcome during the first 3 days5,6, it has been shown that a score of 2.5 or higher on days 4–7 is associated with prolonged mechanical ventilation7.
It was not until 1994 that the AmericanEuropean Consensus Conference on ARDS recommended a new definition, and a new name: ‘acute respiratory distress
syndrome’, to emphasize the acute onset of the syndrome and its occurrence in children8. The three new definitional criteria are: a PaO2 to FiO2 ratio of 200 or less,
bilateral pulmonary infiltrates on the chest radiograph, and a pulmonary artery occlusion pressure of 18 mmHg or less. The third criterion excludes left ventricular failure
or a cardiogenic cause of pulmonary edema. Patients with less severe hypoxemia (PaO2 to FiO2 ratio ≤300) are considered to have acute lung injury (ALI), a milder
functional lung injury resulting from structural changes in the alveolarcapillary unit. The term ARDS is therefore used for the most severe form of ALI.
14.2.2 Epidemiology
The true incidence of ARDS has been difficult to determine because there have been various definitions of the syndrome since it was first described. The incidence has
been reported to
Table 14.1 Murray scale for acute lung injury
Points 0 1 2 3 4
Chest Xray (no. of quadrants) No infiltrate one two three four
PaO2/FiO2 ≥300 225–299 165–224 100–174 <100
PEEP ≤5 6–8 9–11 12–14 ≥15
Cs ≥80 60–79 40–59 20 ≤19
Final score is aggregate sum divided by number of components: 0 indicates no lung injury; 0.1–2.5 indicates mild to moderate lung injury; 2.5 indicates ARDS. From: Murray, J.F. et al.
(1988) An expanded definition of the adult respiratory distress syndrome. Reproduced with permission from Am, Rev. Respir, Dis. 138:720–723.
Page 223
be as low as 1.5–8.3 cases9–11 and as high as 75.0–88.6 cases12,13 per 100 000. A prospective epidemiological study using the 1994 consensus committee definition
is currently being undertaken to more accurately determine the incidence of ARDS.
14.3 Pathogenesis
The acute respiratory distress syndrome is a common endpoint response to either direct or indirect lung injury. Precipitating factors that directly injure the lung include
pneumonia, aspiration of gastric contents, pulmonary contusion, fat emboli, neardrowning, inhalational injury, and reperfusion pulmonary edema after lung
transplantation or pulmonary embolectomy. The acute respiratory distress syndrome may occur as a consequence of indirect lung injury in the setting of a systemic
process such as sepsis, severe trauma with shock and multiple transfusions, cardiopulmonary bypass, drug overdose, acute pancreatitis, and transfusion of blood
products14. Of all the ARDS precipitating factors, sepsis has the greatest potential to induce ARDS, and the presence of multiple predisposing factors significantly
increases the risk of developing the syndrome15 (Table 14.2).
14.3.1 Early ARDS

The result of direct or indirect lung injury is diffuse alveolar damage with capillary endothelial and alveolar epithelial cell necrosis, proteinacious edema accumulation,
and hyaline membrane formation. This process begins within the first 24 h and marks the beginning of the acute or exudative stage of ARDS, which continues for
approximately 1 week. During this stage, there is buildup of exudative fluid that moves from the intravascular to the interstitial space, and ultimately into the alveolar
unit. The ARDSinduced accumulation of interstitial pulmonary edema is difficult to distinguish radiographically from cardiogenic pulmonary edema and causes a
decrease in lung compliance. It is by attempting to maintain adequate oxygenation and ventilation in the setting of decreasing lung compliance and gas exchange that
secondary or ventilatorinduced lung injury can occur.
The alveolar epithelium and capillary endothelium together form a barrier that prevents crossing of edema fluid from the intravascular space into the alveolus. The
alveolar epithelium is composed of flat (type I) and cuboidal (type II) cells. The flat epithelial cells make up 90% of the total alveolar surface and are more susceptible
to ventilatorinduced
Table 14.2 Prevalence of ARDS with predisposing factors
Clinical condition n Prevalence with this factor alone (%) Prevalence with multiple risk factors
Sepsis syndrome 38 38 47
Gastric aspiration 23 30 31
Pulmonary contusion 29 17 38
Multiple transfusions 17 24 45
Multiple fractures 12 8 44
Neardrowning 3 67 –
Pancreatitis 1 100 –
Reprinted from Pepe et al. (1982) Clinical predictors of the adult respiratory distress syndrome. Am. J. Surg. 144:124–130. With permission from Excerpta Medica Inc.
Page 224
lung injury. The more resilient cuboidal cells produce surfactant and participate in ion transport. Injured cuboidal cells differentiate into flat type I alveolar cells and
produce either no surfactant or surfactant that has altered lipid content and increased minimal surface tension16,17. The alveolar epithelium is normally less permeable
than the capillary endothelium and contributes more to overall barrier function. A lung injury that involves the alveolar epithelium will therefore disproportionately
increase permeability and allow influx of proteinrich fluid into the alveolus. Barrier loss also allows for alveolar bacterial translocation similar to what can occur in the
intestine when gut mucosal integrity is lost. Patients with bacterial pneumonia and loss of the alveolarcapillary barrier may develop bacteremia and septic shock18. In
addition to barrier loss, alveolar epithelial cell injury results in abnormal fluid transport, and fluid accumulation in the alveolar space19,20.
The ARDSprecipitating factor may lead to neutrophil activation and sequestration. Eighty per cent of patients who develop ARDS have a transient decrease in
neutrophil count because neutrophils adhere to the vascular endothelium21. The subsequent release of mediators of the inflammatory response (elastase, proteolytic
enzymes, oxygen free radicals, leukotrienes, thromboxane, tumor necrosis factor, and platelet activating factor) directly damages the capillary endothelium,
compromises barrier function, and leads to increased interstitial edema.
During the acute stage of ARDS, several factors lead to reduced gas exchange and lung compliance. The normally small interstitial space between capillary and
alveolus widens as a result of edema formation, migrating neutrophil accumulation, and procollagen release from fibroblasts. This widened gap impairs gas exchange.
On the alveolar side, hyaline membranes form and line the denuded basement membrane, further blocking gas exchange. Alveolar macrophages release macrophage
inhibitory factor, tumor necrosis factoralpha, and interleukins (IL) 1, 6, 8, and 10. Tumor necrosis factoralpha and IL1 activate neutrophils, causing the release of
leukotrienes, oxidants, platelet activating factor, and proteases. These factors, as well as inactivated surfactant, necrotic or apoptotic type I cells, fibrin, cellular debris,
sloughed bronchial epithelium, red blood cells, and proteinrich edema fluid, flood the alveolus. Swollen and injured capillary endothelial cells and the increased number
of neutrophils and platelets in the capillaries impair capillary blood flow and further impede gas exchange.
Oxygen free radical production in ARDS occurs early during sustained exposure to high FiO2 levels (>0.60) and is generated by neutrophils and macrophages.
Naturally occurring antioxidants (glutathione, transferrin, and ceruloplasmin) found in the bronchoalveolar lavage of patients with ARDS offer a defense against oxygen
free radicalinduced lung injury. The proteinrich fluid itself present within the alveoli as a result of increased permeability may also limit oxidant injury. Plasma levels of
the antioxidant vitamin E are low in ARDS, suggesting increased consumption in the setting of increased oxygen free radical production and/or decreased vitamin E
absorption as a result of malnutrition22.
14.3.2 Late ARDS

The initial exudative stage lasts approximately 1 week and is followed either by recovery or progression to a stage characterized by fibroproliferation and subsequent
fibrosis. If progression to the fibroproliferative and fibrotic stages occurs, lung compliance continues to decrease and the risk of ventilatorinduced lung injury
(volutrauma and/or barotrauma) increases as attempts are made to ventilate stiffer lungs with higher pressures. This stage may begin as early as the fifth day when the
alveolar space becomes filled with mesenchymal cells. Overstretching of the alveolus stresses the epithelial cells and may cause cell disruption. During the
fibroproliferative stage, inflammatory cells accumulate as a result of chemoattractants released by neutrophils. Normally, inflammatory cells together with
Page 225
alveolar macrophages remove debris and begin repair, but in ARDS epithelial repair is disorganized and may lead to fibrosis and decreased lung compliance23.
Corticosteroids are thought to be useful during the fibroproliferative (inflammatory) stage, but true benefit has not been established.
14.4 Clinical presentation

Patients with ARDS experience an acute onset of severe dyspnea and respiratory distress and the manifestations of arterial hypoxemia refractory to oxygen therapy:
cyanosis, tachypnea, and tachycardia, in the absence of left ventricular failure (pulmonary artery occlusion pressure ≤18 mmHg). A predisposing factor such as sepsis,
hypotension, trauma, or multiorgan dysfunction may be present. By definition, the presentation cannot be attributed to congestive heart failure. The key laboratory
finding is a PaO2 to FiO2 ratio of 200 or less. The arterial pH relates to the adequacy of ventilation in the setting of decreasing lung compliance and to the presence or
absence of sepsis or hypotension that may contribute to metabolic acidemia. If multiorgan failure is present, laboratory investigations may reveal hematologic,
coagulation, renal, hepatobiliary, or electrolyte abnormalities. The chest radiograph will show characteristic diffuse bilateral pulmonary infiltrates.
14.5 Treatment
The treatment of ARDS begins with determining and treating the underlying cause. If ARDS is caused by sepsis, the focus should be identified and appropriately
treated. If a septic source is identified from blood, respiratory, or urine cultures, or by chest radiography or computed tomography, antibiotics should be directed
against the causative organism. Prevention, or early detection and culturedirected treatment of nosocomial infections is essential.
Ventilatory management has been a key aspect of ARDS treatment since the syndrome was first described in 1967. In the initial report, oxygenation was enhanced
dramatically in some of the 12 patients by applying an expiratory retard that created PEEP. Since that time, PEEP has been a mainstay of treatment, its benefit
mediated by recruiting and maintaining the patency of flooded alveoli. More recently, attention has focused on lowstretch ventilatory strategies as a method to optimize
alveolar recruitment while limiting the adverse effects of overdistended alveoli. Results of the NIH ARDS Network trial (comparing lowstretch ventilation (6 ml kg−1)
tidal volume versus traditional tidal volumes (12 ml kg−1)) show a 32% mortality reduction associated with the lowstretch approach24.
14.5.1 Recruitment
Optimal ventilatory management of ARDS requires an understanding of the state of the alveoli in ARDS lungs. Indeed, four populations of alveoli exist in ARDS
(Figure 14.1). The first population is composed of normal alveoli that expand at 20–30 cmH2O. These recruited alveoli participate in gas exchange and can be
damaged if overdistended or exposed to endinspiration plateau pressures greater than 35 cmH2O. The second population of alveoli is composed of atelectatic alveoli
that can be reexpanded and are therefore recruitable. These alveolar units can be enlisted to participate in gas exchange if increased PEEP is employed to hold open
the units and if sufficient inspiratory time is used. Positive endexpiratory pressure is used to restore functional residual capacity above the closing volume, thereby
reducing the risk of derecruitment or alveolar collapse during exhalation. Adhered walls of collapsed terminal bronchi typically require high inflation pressures to
separate but will remain open at much lower pressures once opened. The compliance in these alveolar units is normal. The third population is composed of alveolar
units that have poor compliance and require increased time and peak inflation pressure to open and keep expanded. Opening these alveoli requires exceeding the
critical opening pressure for an adequate time period. These alveolar units may become derecruited and thus excluded from gas exchange
Page 226
Figure 14.1 Four populations of alveoli.
if the critical opening pressure is not reached and maintained. The fourth population is made up of destroyed alveolar units that will not open at any pressure. Using high
PEEP in a futile attempt to reexpand these units will damage the remaining normal lung.
In the context of this spectrum of alveoli, it has been shown that reduced lung compliance in patients with ARDS is due to a decrease in the number of normally
functioning alveoli rather than to a generalized dysfunction that affects all alveolar units equally, as is suggested by the appearance of diffuse bilateral alveolar infiltrates
on chest radiography25. Computed tomographic scans of lungs in patients with ARDS have demonstrated consolidation and atelectasis in the dependent lung zones,
with other areas being relatively spared26,27.
14.5.2 Positive endexpiratory pressure

Positive endexpiratory pressure redistributes pulmonary edema from the alveolar to the interstitial space and increases functional residual capacity as a result of
recruitment of collapsed alveoli, thereby improving gas exchange28. During the initial exudative stage, hypoxemia rapidly progresses from mild and FiO2 responsive
(meaning that an adequate PaO2 can initially be maintained by increasing the FiO2) to severe in spite of a maximal FiO2. Maintaining adequate oxygenation during this
early stage depends on using incrementally higher levels of PEEP. In the later fibroproliferative and fibrotic stages, hypoxemia becomes refractory to both high FiO2 and
PEEP levels. Optimal PEEP is the level that allows for adequate tissue oxygenation with an FiO2 of 0.60 or less. The level of PEEP is therefore individualized for each
patient. Amato et al. have shown benefit by raising PEEP to a level above the lower inflection point on a pressurevolume curve determined for each patient29. In
general, PEEP should be employed early and increased to a level that maintains an adequate
Page 227
oxygen saturation (SpO2≥90%) at the lowest possible FiO2. Prophylactic PEEP (8 cmH2O) in patients at risk for ARDS has been shown not to confer benefit30.
The risk of pneumothorax increases with higher PEEP levels. Although prophylactic bilateral chest tubes have been advocated by some to prevent this complication,
we do not advocate this practice. Tube thoracostomy carries potential risks including bleeding from an intercostal artery or from the lung parenchyma itself, as well as
infection. Therefore, the ARDS patient is closely monitored and a chest tube is inserted only if a pneumothorax occurs.
Once optimal PEEP is reached, it is maintained and weaned gradually only after the patient stabilizes and begins to recover. An improvement in pulmonary
compliance (the change in volume for a unit pressure) is an early indicator of recovery and can be calculated daily to determine when recovery is beginning. The change
in volume (ΔV) for a unit pressure (ΔP) under conditions of no flow is termed the pulmonary static compliance (Cs) and can be calculated using the following formula:
Similarly, pulmonary dynamic compliance can be calculated substituting peak for plateau airway pressure in the formula above. A related early indicator of recovery is
decreasing peak inflation pressure. This is followed by improved oxygenation at progressively lower FiO2 and PEEP levels. The chest radiograph will improve in
appearance but this change usually lags by 1 week or more after the earlier indicators of recovery appear.
In ARDS survivors, maximal recovery typically occurs by 6 months after extubation and pulmonary function at 1 year following hospital discharge is normal or only
mildly impaired. Poorer outcomes are seen in patients who have a more severe ARDS course and a longer ventilation period31. Longterm pulmonary function sequelae
may include tracheal stenosis from longterm intubation and/ or tracheotomy, and mild impairment of diffusing capacity.
14.5.3 Secondary lung injury

The main objective in ARDS management is to adequately oxygenate and ventilate the patient while avoiding oxygen toxicity as a result of oxygen free radical
production at higher FiO2 levels, volutrauma caused by overstretching noncompliant alveolar units, and barotrauma from high inflation pressures. These three potential
lung injuries are collectively termed secondary or ventilatorinduced lung injury. At higher FiO2 levels (≥0.60), oxygen free radicals are produced to an extent that
causes lipid peroxidation, neutrophil recruitment, and protein degradation, causing alveolar epithelial and vascular endothelial cell damage and increased permeability.
Antioxidant use has been proposed to protect against secondary lung injury as a result of oxygen free radical production. In one trial using the antioxidants N
acetylcysteine and procysteine there was no difference in mortality among study group; however, the number of days of ALI and the incidence of new multiorgan
system failure were decreased32.
14.5.4 Volutrauma
Secondary lung injury caused by overstretching alveoli is referred to as volutrauma. Type II cuboidal cells are susceptible to this type of lung injury, which impairs their
ability to produce surfactant. Several recent studies have suggested that avoiding alveolar overdistention during ventilatory support enhances survival in patients with
ARDS. In the NIH ARDS Network trial, patients treated with lower tidal volumes of 6 ml kg−1 (a normal tidal volume for people breathing at rest) and with plateau
pressures less than 35 cmH2O experienced higher survival and fewer nonpulmonary organ failures than did patients treated with conventional tidal volumes24. The
formulae for
Page 228
calculating the predicted (ideal) body weight (kg), which should be used instead of the measured body weight, to determine the tidal volume setting, are:
Given this current evidence, lowstretch ventilation, often achieved using a pressurecontrolled ventilatory mode, has become the standard of care in patients with
ARDS. Of course, one consequence of the lowstretch approach in patients with stiff lungs is that minute ventilation decreases, which results in hypercapnea and
respiratory acidemia. This ‘permissive hypercapnea’ and the associated acidemia is accepted as a consequence of the benefits of avoiding volutrauma. Recognizing that
profound acidemia can have independent adverse effects (e.g. cardiac arrhythmias, refractoriness to inotropic agents), the NIH study accepted acidemia to a pH of
7.30. Lower pH values led to bicarbonate infusion in that trial. As a possible explanation of the survival benefit of a lowstretch ventilatory approach, recent evidence
has shown that cyclic opening and closing of alveoli can initiate a cascade of proinflammatory cytokine release, which may enter the circulation and cause other organ
systems to fail33. The enthusiasm for these ventilatory advances reflects the fact that the lowstretch approach is the first ARDS treatment of many studied to confer a
survival advantage24. Other proposed treatments, including highfrequency ventilation and raising systemic oxygen transport with inotropes, have not been shown to be
beneficial. Still, maintaining oxygen delivery, which is the product of cardiac output and arterial oxygen content, is sensible. Although there is no automatic transfusion
threshold, blood transfusion to maintain hemoglobin values at 8 mg dl−1 or more is often advocated to avoid a decreased arterial oxygen content and consequent
reduced systemic oxygen transport. The goal is to maintain oxygen delivery while avoiding ventilatorinduced lung injury caused by oxygen toxicity, volutrauma, or
barotrauma.
14.5.5 Barotrauma
Barotrauma or pressureinduced lung injury can cause pneumomediastinum, pneumoperitoneum, tension or simple pneumothorax, interstitial emphysema, tension gas
cyst, injury to small and terminal bronchi, lung edema, or subcutaneous emphysema, and becomes more likely with higher plateau pressures. This form of lung injury
more often appears gradually and may therefore not be obvious. If subcutaneous emphysema extends into the neck, it may compress and thereby completely obstruct
the airway. This is critical in both the intubated and nonintubated patient. Although the endotracheally intubated patient with subcutaneous emphysema has a
reasonably secure airway, it is important to consider airway patency before extubation. In ARDS patients, as lung compliance decreases, endinspiration plateau
pressures rapidly increase and, if not recognized, will lead to pressureinduced lung injury or barotrauma. Limiting inflation pressures can minimize barotrauma.
Although an absolutely safe inflation pressure has not been well defined, maintaining a plateau pressure of 35 cmH2O or less has been traditionally accepted as
desirable34. In fact, the recent NIH trial suggests that tidal volume may be more important than plateau pressure in determining patient outcome24.
14.5.6 Prone positioning

Prone positioning is being evaluated as a way of enhancing oxygenation in patients with ARDS. In available studies, patients were shown to have improved oxygenation
after prone positioning typically within 2 h, and the improvement was sustained at about 50% beneficial effect when the patient was returned to the supine position35–38.
The optimal duration of prone position is not defined, alternating between supine and prone position every 6–8 h is common practice. In one study in which 14 ARDS
patients were treated for at least 1 h in the prone position, the mortality of the prone
Page 229
positioned patients was 42% (versus 58% in those not turned prone)38. The responders (PaO2 to FiO2 ratio increase >10%) were treated for at least 6 h. Eleven were
primary responders and two of the other three improved with a second prone positioning. Although the mechanism by which prone positioning might confer benefit is
not known, the benefits may in part be due to the relatively nonuniform distribution of lung injury in ARDS. A benefit, however, has not been established.
Potential complications of prone positioning include hemodynamic instability, vena caval compression, decreased functional residual capacity, peripheral nerve injury,
joint dislocation, corneal damage, decubiti of the face and chest, accidental extubation or central venous or arterial line removal, endotracheal tube obstruction by
mucus, and aspiration of enteral nutrition formulas. A specific risk for cardiac surgery patients related to prone positioning is delay in emergency reopening of the
sternotomy if required for hemodynamic instability.
14.5.7 Steroids and nonsteroidal antiinflammatory agents

During the fibroproliferative stage of ARDS, pulmonary interstitial inflammation peaks. It is during this stage that glucocorticoid use has been proposed to attenuate the
inflammatory response and thereby improve lung compliance. However, large prospective trials have demonstrated that glucocorticoids administered early in the course
of established ARDS do not confer benefit39–41. Moreover, in early ARDS it is difficult to exclude infection as the precipitating cause, and glucocorticoids can increase
mortality if administered in the setting of unrecognized infection. More encouraging results have been reported in studies where glucocorticoids were administered later
in the course of ARDS42,43. The current recommendation is that highdose glucocorticoids can be used as rescue therapy in the fibroproliferative stage in patients who
have poor lung compliance and nonresolving disease as long as infection has been excluded either as a cause or complication of ARDS. An uncontrolled study
reported an 80% survival and an improved Lung Injury Score when a methylprednisolone 200 mg bolus was administered followed by 2–3 mg kg−1 day−1, and
tapered after extubation. It is important to point out, however, that 60% of the treated patients in this study developed infection42. Use of the nonsteroidal anti
inflammatory agents pentoxyfylline, a phosphodiesterase inhibitor with antiinflammatory effects on neutrophils and macrophages, prostacyclin (PgI2) which decreases
pulmonary and systemic pressures and intrapulmonary shunt, PgE1, and nitric oxide, which has antiinflammatory as well as vasodilating activity, in ARDS has been
proposed, but there is insufficient evidence of benefit to support their use44–46.
14.5.8 ECMO, ECO2R, surfactant, nitric oxide, and partial liquid ventilation
Other techniques that have been tried to improve outcome in ARDS include extracorporeal membrane oxygenation (ECMO), extracorporeal membrane carbon
dioxide removal (ECCO2R), exogenous surfactant use, nitric oxide, and partial liquid ventilation. Given the success of ECMO in infant respiratory distress syndrome,
the reversibility of alveolar damage in ARDS, and the presence of severe hypoxemia despite maximal ventilator support, ECMO was tried in adults with ARDS in the
late 1970s. The NIH ECMO trial that initially planned to recruit 300 patients to evaluate ECMO as a therapy for severe acute respiratory failure was terminated after
the first 90 patients because of poor survival (<10%)47. In this study, ECMO was intended for shortterm use to theoretically rest the lungs while recovery could begin.
However, the membrane oxygenators frequently failed, and bleeding and infection complicated the patient’s intensive care course. Extracorporeal membrane
oxygenation reversed the poor oxygenation and supported respiratory gas exchange in these patients, but it did not improve survival. Studies using ECCO2R have
shown no mortality improvement48.
Page 230
Because surfactant levels are low, and because alveolar collapse and reduced lung compliance are features of ARDS, alveolar surfactant replacement has been
proposed as a treatment for ARDS. However, in a trial using a 5 day course of nebulized synthetic surfactant in 725 patients, there was no significant effect on 30 day
survival or PaO2 to FiO2 ratio49. Thus, surfactant is not routinely used in adult patients with ARDS. Some have proposed that alternative surfactant preparations and
methods of delivery might be eff ective, and newer products and methods of delivery are currently being investigated. Nitric oxide (NO), a selective pulmonary
vasodilator with rapid onset and reversal, has appeal in ARDS patients because it has been shown to improve ventilationperfusion matching by dilating vessels only in
wellventilated lung regions. Although NO has been shown to reduce shunt fraction and improve oxygenation50, a multicenter trial using NO (1.25–80 ppm) in 177
ARDS patients failed to show a change in mortality or ventilator time51. Thus, despite its occasional use as shortterm salvage therapy to improve oxygenation, NO
cannot at present be endorsed in ARDS.
Partial liquid ventilation (PLV) using perfluorocarbons that have a low surface tension and high solubility, and that offer gas exchange and surfactant functions, has
been evaluated in patients with ARDS. As the lungs fill with perfluorocarbons, the closed alveoli open and exudates are flushed out. In animal studies, increased lung
compliance and oxygenation has been observed, and in 10 adult patients with ARDS on ECMO, partial liquid ventilation was well tolerated with decreased physiologic
shunting and increased static lung compliance. However, the overall survival of patients treated with PLV was 50%, equal to that of historic controls52. To date, benefit
has not been clearly established and research in this area is ongoing.
14.5.9 Fluid balance

The increase in pulmonary edema that occurs during the early exudative stage in ARDS peaks within 24–48 h and subsides by 3–5 days. This observation led to the
proposal that a decrease in intravascular volume by aggressive diuresis could decrease pulmonary interstitial edema and thereby improve oxygenation. The risk
associated with this strategy is that fluid restriction and aggressive diuresis causes reduced cardiac filling pressures and a low cardiac output. This increases the risk of
multiorgan failure and creates the need for inotropic support. Hypotension in spite of inotropic support, and a diuresisinduced prerenal state, can lead to acute renal
failure that converts single organ failure to multisystem failure with a commensurately increased overall mortality. It is important, therefore, to maintain normal
intravascular volume so as to optimize oxygen delivery and assure adequate tissue perfusion. Aggressive diuresis will not decrease pulmonary interstitial edema or
improve lung compliance, and may lead to renal insufficiency. Once intravascular volume is optimized, vasopressors may be used to increase cardiac output and restore
normal oxygen delivery53. Two prospective studies of highrisk surgical patients have shown no relationship between positive fluid balance and development of
ARDS54,55.
14.5.10 Nutritional support

Nutritional support should be initiated once the diagnosis of ARDS is established. The patient with ARDS will require prolonged intensive care and therefore there is no
reason to delay feeding. Enteral feeding is preferable to parenteral because it helps maintain normal intestinal mucosa integrity and maintains a barrier against Gram
negative bacterial translocation and subsequent sepsis. Additionally, the enteral route has fewer potential complications. Feeding via the enteral route avoids the risks of
central venous line insertion, which include hemothorax, pneumothorax, and thoracic duct injury, as well as the risks associated with indwelling central venous lines,
which include sepsis, catheter erosion into a vascular structure, and air embolism.
Difficulty can be encountered when placing and maintaining the correct position of a nasoenteric feeding tube. The preferred tip
Page 231
location is in the fourth portion of the duodenum. A nasogastric tube may be used but is not required if the feeding tube tip is in proper position and the head of the bed
is maintained at 30° at all times. Correct feeding tube position must be confirmed radiographically after placement, as the tube may be malpositioned in the trachea,
bronchus, or beyond. Finally, prevention of gastrointestinal bleeding and venous thromboembolism is critical in these patients.
In the majority of patients who survive ARDS, pulmonary function is normal or only slightly impaired within 6–12 months following extubation31. Until recently, the
mortality rate of ARDS has remained 40–60%56,57. However, two recent reports suggest a decreasing mortality rate of approximately 35%58,59. The outcome
improvement may be related to better critical care and more effective treatment for sepsis syndrome, as well as to the increased use of protective lung ventilation
strategies60. Features associated with poorer survival in ARDS include failure of the PaO2 to FiO2 ratio to improve over the first week of therapy and the presence of
liver disease, nonpulmonary organ dysfunction, sepsis, and advanced age5,6,61. In fact, only 16% of deaths from ARDS are caused by respiratory failure (hypoxemia,
hypercapnia, and respiratory acidosis)62. Most of the attributable mortality is due to multiorgan failure mediated by inflammatory cytokine release from the lungs and
spread via the circulation to other organs, and is not related to the length of ventilator time.
14.7 Conclusions
There is significant morbidity associated with ARDS and mortality has only recently been shown to be decreasing from the level first reported by Ashbaugh et al. in
19672. The definition of ARDS has now been clarified and this should make future studies more reliable. The etiology is often multifactorial and undetermined.
Management remains largely supportive (Table 14.3) but with newer strategies such as lowstretch pressurecontrolled ventilation, the goal of providing adequate
oxygenation and ventilation with the least amount of secondary lung injury is being achieved, and mortality is being reduced. The role of newer
Table 14.3 Summary of recommended management strategies for treating acute respiratory distress syndrome
• Identify underlying causes; if infection is identified, treat with appropriate antibiotics

• Increase PEEP to a level that allows for an FiO2 ≤0.60 (typically 12–18 cmH2O), to avoid oxygen free radical induced lung injury; once optimal PEEP is reached, wean
gradually (to avoid alveolar derecruitment) only when signs of recovery appear
• Use a lowstretch pressurecontrolled ventilatory (PCV) approach (6 ml kg−1 with endinspiratory plateau pressure <35 cmH O) to avoid volutrauma (kg is predicted or
2
ideal, not measured, body weight)
• Accept hypercapnea as long as the arterial pH can be maintained above 7.25
• Use recruitment maneuvers
• Provide adequate sedation, and neuromuscular blockade if necessary
• Maintain normal intravascular volume
• Although there is no automatic transfusion threshold level, maintain oxygen delivery by keeping hemoglobin levels ≥8 mg dl−1, and by using inotropic support if required
to normalize cardiac output
• Initiate prophylaxis against gastrointestinal bleeding and venous thromboembolism
• Initiate nutritional support (enteral is preferable to parenteral) once the diagnosis of ARDS is made
• Consider glucocorticoid use in the fibroproliferative stage only if infection has been excluded (methylprednisolone 2 mg kg−1 day−1)
Page 232
interventions is being defined. Acute respiratory distress syndrome survivors generally resume productive lives without pulmonary limitation.
References
1. Buford, T.H. and Burbank, B. (1945) Traumatic wet lung: Observations on certain physiologic fundamentals of thoracic trauma. J. Thorac. Surg. 14:415–424.
2. Ashbaugh, D.G., Bigelow, D.B., Petty, T.L. and Levine, B.E. (1967) Acute respiratory distress in adults. Lancet 2:319–323.
3. Petty, T.L. (1982) Adult respiratory distress syndrome: Definition and historical perspective. Clin. Chest Med. 3 (1):3–7.
4. Murray, J.F., Matthay, M.A., Luce, J.M. and Flick, M.R. (1988) An expanded definition of the adult respiratory distress syndrome. Am. Rev. Respir. Dis.
138:720–723.
5. Doyle, R.L., Szaflarski, N., Modin, G.W., et al. (1995) Identification of patients with acute lung injury: predictors of mortality. Am. J. Respir. Crit. Care Med.
152:1818–1824.
6. Zilberberg, M.D. and Eptein, S.K. (1998) Acute lung injury in the medical ICU: comorbid conditions, age, etiology, and hospital outcome. Am. J. Respir. Crit. Care
Med. 157:1159–1164.
7. Heffner, J.E., Brown, L.K., Barbieri, C.A., Harpel, K.S. and DeLeo, J. (1995) Prospective validation of an acute respiratory distress syndrome predictive score.
Am. J. Respir. Crit. Care Med. 152:1518–1526.
8. Bernard, G.R., Artigas, A., Brigham, K.L., et al. (1994) The AmericanEuropean Consensus Conference on ARDS. Definitions, mechanisms, relevant outcomes,
and clinical trial coordination. Am. J. Respir. Crit. Care Med. 149: 818–824.
9. Webster, N.R., Cohen, A.T. and Nunn, J.F. (1988) Adult respiratory distress syndrome—how many cases in the UK? Anaesthesia 43: 923–926.
10. Villar, J. and Slutsky, A.S. (1989) The incidence of the adult respiratory distress syndrome. Am. Rev. Respir. Dis. 140:814–816.
11. Thomsen, G.E. and Morris, A.H. (1995) Incidence of the adult respiratory distress syndrome in the state of Utah. Am. J. Respir. Crit. Care Med. 152:965–971.
12. Murray, J.F. (1977) Conference Report: Mechanisms of acute respiratory failure. Am. Rev. Respir. Dis. 115:1071–1078.
13. Lewandowski, K., Metz, J., Deutschmann, C., et al. (1995) Incidence, severity, and mortality of acute respiratory failure in Berlin, Germany. Am. J. Respir. Crit.
Care Med. 151(4):1121–1125.
14. Ware, L.B. and Matthay, M.A. (2000) The acute respiratory distress syndrome. N. Engl J. Med. 342:1334–1349.
15. Pepe, P.E., Potkin, R.T., Reus, D.H., Hudson, L.D. and Carrico, CJ. (1982) Clinical predictors of the adult respiratory distress syndrome. Am. J. Surg.
144:124–130.
16. Greene, K.E., Wright, J.R., Steinberg, K.P., et al. (1999) Serial changes in surfactantassociated proteins in lung and serum before and after onset of ARDS. Am.
J. Respir. Crit. Care Med. 160:1843–1850.
17. Gregory, T.J., Longmore, W.J., Moxley, M.A., et al. (1991) Surfactant chemical composition and biophysical activity in acute respiratory distress syndrome. J.
Clin. Invest 88:1976–1981.
18. Kurahashi, K., Kajikawa, O., Sawa, T., et al. (1999) Pathogenesis of septic shock in Pseudomonas aeruginosa pneumonia. J. Clin. Invest 104:743–750.
19. Modelska, K., Pittet, J.F., Folkesson, H.G., et al. (1999) Acidinduced lung injury. Protective effect of antiinterleukin8 pretreatment on alveolar epithelial barrier
function in rabbits. Am. J. Respir. Crit. Care Med. 160:1450–1456.
20. Sznajder, J.I. (1999) Strategies to increase alveolar epithelial fluid removal in the injured lung. Am. J. Respir. Crit. Care Med. 160:1441–1442.
21. Thommasen, H.V., Russell, J.A., Boyko, W.J. and Hogg, J.C. (1984) Transient leucopenia associated with adult respiratory distress syndrome. Lancet 1
(8381):809–812.
22. Richard, C., Lemonnier, F., Thibault, M., et al. (1990) Vitamin E deficiency and lipoperoxidation during adult respiratory distress syndrome. Crit. Care Med.
18:4–9.
23. Bitterman, P.B. (1992) Pathogenesis of fibrosis in acute lung injury. Am. J. Med. 92:39S–43S.
24. The Acute Respiratory Distress Syndrome Network (2000) 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. 342: 1301–1308.
25. Marini, J.J. (1990) Lung mechanics in the adult respiratory distress syndrome. Recent conceptual advances and implications for management. Clin. Chest Med.
11:673–690.
26. Goodman, L.R. (1996) Congestive heart failure and adult respiratory distress syndrome. New insights using computed tomography. Radiol. Clin. N. Am. 34:33–46.
Page 233
27. Gattinoni, L., Bombino, M., Pelosi, P., et al. (1994) Lung structure and function in different stages of severe adult respiratory distress syndrome. JAMA
271:1772–1779.
28. Falke, K.J., Pontoppidan, H., Kumar, A., et al. (1972) Ventilation with endexpiratory pressure in acute lung disease. J. Clin. Invest. 51: 2315–2323.
29. Amato, M.B.P., Barbas, C.S.V., Medeiros, D.M., et al. (1998) Effect of a protective ventilation strategy on mortality in the acute respiratory distress syndrome.
N. Engl. J. Med. 338: 347–354.
30. Pepe, P.E., Hudson, L.D. and Carrico, CJ. (1984) Early application of positive endexpiratory pressure in patients at risk for adult respiratorydistress syndrome. N.
Engl. J. Med. 311:281–286.
31. McHugh, L.G., Milberg, J.A., Whitcomb, M.E., et al. (1994) Recovery of function in survivors of the acute respiratory distress syndrome. Am. J. Respir. Crit.
Care Med. 150:90–94.
32. Bernard, G.R., Wheeler, A.P., Arons, M.M., et al. (1997) A trial of antioxidants Nacetylcysteine and procysteine in ARDS. The Antioxidant in ARDS Study
Group. Chest 112: 164–172.
33. Slutsky, A.S. and Tremblay, L.N. (1998) Multiple system organ failure: is mechanical ventilation a contributing factor? Am. J. Respir. Crit. Care Med. 157:1721–
1725.
34. Rappaport, S.H., Shpiner, R., Yoshihara, G., et al. (1994) Randomized, prospective trial of pressurelimited versus volumecontrolled ventilation in severe
respiratory failure. Crit. Care Med. 22:22–32.
35. Langer, M., Mascheroni, D., Marcolin, R. and Gattinoni, L. (1988) The prone position in ARDS patients. A clinical study. Chest 94: 103–107.
36. Pappert, D., Rossaint, R., Slama, K., et al. (1994) Influence of positioning on ventilationperfusion relationships in severe adult respiratory distress syndrome. Chest
106: 1511–1516.
37. Chatte, G., Sab, J.M., Dubois, J.M., et al. (1997) Prone position in mechanically ventilated patients with severe acute respiratory failure. Am. J. Respir. Crit. Care
Med. 155:473–478.
38. Flaatten, H., Aardal, S. and Hevroy, O. (1998) Improved oxygenation using the prone position in patients with ARDS. Acta Anaesthesiol. Scand. 42:329–334.
39. Bernard, G.R., Luce, J.M., Sprung, C.L., et al. (1987) Highdose corticosteroids in patients with the adult respiratory distress syndrome. N. Engl. J. Med.
317:1565–1570.
40. Luce, J.M., Montgomery, A.B., Marks, J.D., et al. (1988) Ineffectiveness of highdose methylprednisolone in preventing parenchymal lung injury and improving
mortality in patients with septic shock. Am. Rev. Respir. Dis. 138:62–68.
41. Sprung, C.L., Caralis, P.V., Marcial, E.H., et al. (1984) The effects of highdose corticosteroids in patients with septic shock. A prospective, controlled study. N.
Engl. J. Med. 311:1137–1143.
42. Meduri, G.U., Belenchia, J.M., Estes, R.J., et al. (1991) Fibroproliferative phase of ARDS. Clinical findings and effects of corticosteroids. Chest 100:943–952.
43. Meduri, G.U., Chinn, A.J., Leeper, K.V., et al. (1994) Corticosteroid rescue treatment of progressive fibroproliferation in late ARDS. Patterns of response and
predictors of outcome. Chest 105:1516–1527.
44. Walmrath, D., Schneider, T., Schermuly, R., et al. (1996) Direct comparison of inhaled nitric oxide and aerosolized prostacyclin in acute respiratory distress
syndrome. Am. J. Respir. Crit. Care Med. 153:991–996.
45. Zwissler, B., Kemming, G., Habler, O., et al. (1996) Inhaled prostacyclin (PgI2) versus inhaled nitric oxide in adult respiratory distress syndrome. Am. J. Respir.
Crit. Care Med. 154: 1671–1677.
46. Bone, R.C., Slotman, G., Maunder, R., et al. (1989) Randomized doubleblind, multicenter study of prostaglandin E1 in patients with the adult respiratory distress
syndrome. Prostaglandin E1 Study Group. Chest 96(1): 114–119.
47. Zapol, W.M., Snider, M.T., Hill, D., et al. (1979) Extracorporeal membrane oxygenation in severe acute respiratory failure. A randomized progressive study. JAMA
242:2193–2196.
48. Morris, A.H., Wallace, C.J., Menlove, R.L., et al. (1994) Randomized clinical trial of pressurecontrolled inverse ratio ventilation and extracorporeal CO2 removal
for adult respiratory distress syndrome. Am. J. Respir. Crit. Care Med. 149:295–305.
49. Anzueto, A., Baughman, R.P., Guntupalli, K.K., et al. (1996) Aerosolized surfactant in adults with sepsisinduced acute respiratory distress syndrome. Exosurf
Acute Respiratory Distress Syndrome Sepsis Study Group. N. Engl. J. Med. 334:1417–1421.
50. Zapol, W.M., Falke, K.J., Hurford, W.E., et al. (1994) Inhaling nitric oxide: a selective pulmonary vasodilator and bronchodilator. Chest 105:87S–91S.
51. Dellinger, R.P., Zimmerman, J.L., Taylor, R.W., et al. (1998) Effects of inhaled nitric oxide in patients with acute respiratory distress
Page 234
syndrome: Results of a randomized phase II trial. Inhaled Nitric Oxide in ARDS Study Group. Crit. Care Med. 26:15–23.
52. Hirschl, R.B., Pranikoff, T., Wise, C., et al. (1996) Initial experience with partial liquid ventilation in adult patients with the acute respiratory distress syndrome.
JAMA 275: 383–389.
53. Matthay, M.A. and Broaddus, V.C. (1994) Fluid and hemodynamic management in acute lung injury. Semin. Respir. Crit. Care Med. 15:271–288.
54. Bishop, M.H., Jorgens, J., Shoemaker, W.C., et al. (1991) The relationship between ARDS, pul monary infiltration, fluid balance, and hemodynamics in critically
ill surgical patients. Am. Surg. 57:785–792.
55. Svensson, L.G., Hess, K.R., Coselli, J.S., et al. (1991) A prospective study of respiratory failure after highrisk surgery on the thoracoabdominal aorta. J. Vasc.
Surg. 14:271–282.
56. Fowler, A.A., Hamman, R.F., Good, J.T., et al. (1983) Adult respiratory distress syndrome: risk with common predispositions. Ann. Intern. Med. 98:593–597.
57. Suchyta, M.R., Clemmer, T.P., Elliott, C.G., et al. (1992) The adult respiratory distress syndrome. A report of survival and modifying factors. Chest 101:1074–
1079.
58. Milberg, J.A., Davis, D.R., Steinberg, K.P., et al. (1995) Improved survival of patients with acute respiratory distress syndrome (ARDS): 1983–1993. JAMA
273:306–309.
59. Abel, S.J., Finney, S.T., Brett, S.J., et al. (1998) Reduced mortality in association with the acute respiratory distress syndrome (ARDS). Thorax 53:292–294.
60. Luhr, O.R., Antonsen, K., Karlsson, M., et al. (1999) Incidence and mortality after acute respiratory failure and acute respiratory distresssyndrome in Sweden,
Denmark, and Iceland. The ARF Study Group. Am. J. Respir. Crit. Care Med. 159:1849–1861.
61. Monchi, M., Bellenfant, F., Cariou, A., et al. (1998) Early predictive factors of survival in the acute respiratory distress syndrome. A multivariate analysis. Am. J.
Respir. Crit. Care Med. 158:1076–1081.
62. Montgomery, A.B., Stager, M.A., Carrico, C.J. and Hudson, L.D. (1985) Causes of mortality in patients with the adult respiratory distress syndrome. Am. Rev.
Respir. Dis. 132:485–489.
Page 235
Chapter 15
Lungprotective ventilation strategies
Contents
15.1 Ventilatorinduced lung injury (VILI)

15.2 Lungprotective ventilatory strategies (LPVSs)
15.3 Setting of PEEP
15.4 Prone positioning
15.5 Summary
Page 236
Over the last 10 years much of the emphasis regarding mechanical ventilation of critically ill patients with acute lung injury (ALI) or acute respiratory distress
syndrome (ARDS) has focused on lung protection, specifically, the establishment of ventilatory settings that avoid the extension of lung injury by the use of lung
protective ventilatory strategies (LPVSs). There clearly is a large quantity of data indicating that specific ventilatory strategies induce injury and affect outcome in animal
models1–10. In addition, two recent clinical trials have established a link between outcome (death and ventilationfree days) and ventilatory strategy11,12. This chapter
will address the concept of ventilatorinduced lung injury and lungprotective ventilatory strategies.
15.1 Ventilatorinduced lung injury (VILI)

Numerous terms are currently used to address specific types or causes of VILI (Table 15.1). Barotrauma has been used for years to define the form of VILI manifest
by the movement of air out of the lung into a body space or tissue (Figure 15.1)13,14. Generally, this occurs as a result of rupture of the lung parenchyma and dissection
of air via fascial planes into the pleural, mediastinal, or pericardial space14. Although not precisely linked to high airway pressure, barotrauma occurs most commonly in
settings where severe lung disease exists along with mechanical ventilation using high alveolar pressures and large tidal volumes13.
15.1.1 Volutrauma
VILI that results in an increase in the permeability of the alveolar capillary membrane, the development of pulmonary edema, the accumulation of neutrophils and
protein within
Table 15.1 Terms used to refer to specifictypcs of VILI
Barotrauma—Rupture of lung parenchyma resulting in gas outside the lung

Volutrauma—Injury similar to ARDS caused by resulting in gas outside the lung overdistention
Atelectrauma—Injury similar to ARDS caused by repetitive opening and closing of unstable lung units
Biotrauma—Pulmonary inflammatory mediator activation by overdistention or inadequate PEEP
Figure 15.1 Mechanism of alveolar rupture during mechanical ventilation. Pressures between adjacent alveoli equalize rapidly, but
especially in the presence of high alveolar volume, increased alveolar pressure in comparison with that in the adjacent
bronchovascular sheath establishes a pressure gradient that may result In rupture of the alveolar watl, allowing passage of air
into the interstitial tissue of the bronchovascular sheath. From: Maunder; R.J., Pierson, D.J. and Hudson, L.D. (1984)
Subcutaneous and mediastenal emphysema: pathophysiology, diagnosis, and management Reproduced with permission from
The American Medical Association. Arch. Intern, Med. 144:1447–1453,
Page 237
the lung parenchyma, the disruption of surfactant production, the development of a hyaline membrane and a decrease in compliance of the respiratory system is called
volutrauma15. The term volutrauma is used because the primary mechanism for the development of VILI proposed by Dreyfuss and Saumon is overdistention of local
lung units by high alveolar volumes associated with high alveolar pressure15. Webb and Tierney first demonstrated in an intact animal the effects of large tidal volumes
(VT) at high distending pressure on the development of lung injury4. They ventilated healthy rats with peak airway pressures of 45 cmH2O and zero positive end
expiratory pressure (PEEP) (VT of 40 mL kg−1) for 60 min, resulting in gross hemorragic edema and decreased compliance. As noted in Figure 15.2, Dreyfuss et al.
observed marked alternations at the alveolar capillary membrane after only 20 min of ventilation in rats ventilated at peak alveolar pressures of 45 cmH2O (VT 40 mL
kg−1) and no PEEP5. In fact, they demonstrated that injury was observed in minutes after the onset of the injurious ventilatory pattern. Similar injury was observed by
Parker et al. in isolated perfused dog lungs3. Increasing peak alveolar pressure beyond 30 cmH2O resulted in an exponential increase in capillary filtration pressure.
Similar data are now available in a wide variety of animal models confirming the relationship of large distending lung volumes and VILI15.
The single pressure most responsible for the development of VILI appears to be the transpulmonary pressure gradient10. In a now classic experiment, Dreyfuss et
al. demonstrated that rats with their chest walls strapped were protected from VILI compared with those ventilated with the same peak alveolar pressure but without
chest wall binding10. In fact, injury was similar regardless of the use of positive or negative pressure ventilation provided high transpulmonary pressures were
established. As a result, higher ventilating pressures
Figure 15.2 Effect of ventilation at a peak airway pressure of 45 cmH2O for 5–20 min in closedchest rats. Pulmonary edema was
assessed by measuring the extravascular lung water content (QwI/BW) and changes in permeability by determining the
bloodless dry lung weight (DLW/BW), and the distribution space of 125Ilabeled albumin (Alb. Space) in lungs. Control rats (C)
were ventilated at a peak airway pressure of 7 cmH2O. Pulmonary edema developed rapidly (5 min) and was associated with
changes in permeability. All the indices increased markedly after 20 min of mechanical ventilation (P<0.01 versus other
groups). From: Dreyfuss, D., Basset, G., Soler, P. and Saumon, G. (1985) Intermittent positivepressure hyperventilation with
high Inflation pressures produces pulmonary microvascular injury in rats. Reproduced with permission from Am, Rev. Respir.
Dis. 132:880–884.
Page 238
can be used in the presence of a decreased chest wall compliance without the development of VILI, as the decreased chest wall compliance results in lower tidal
volumes at a given peak alveolar pressure.
In addition to causing injury in healthy animal models, inappropriate ventilatory patterns can extend injury in animals presenting with lung injury16. Again in rats,
Dreyfuss et al. demonstrated increased extravascular lung water and distribution of albumin within the lung parenchyma within 2 min of ventilation at 45 cmH2O peak
alveolar pressure16.
15.1.2 PEEP effects

As noted in Figure 15.3, a PEEP of 10 cmH2O in spite of a peak airway pressure of 45 cmH2O decreased the level of VILI compared with similar animals with the
same peak pressure but zero PEEP4. Only mild interstitial edema developed in the animal with 10 cmH2O PEEP. Similar findings have been reported by Corbridge et
al. in dogs with hydrochloric acidinduced lung injury17. Dogs ventilated with 12.5 cmH2O PEEP and a VT of 15 mL kg–1 after lung injury had much lower wet lung
weight and dry lung weight to body weight ratios at autopsy than dogs ventilated with 3.2 cmH2O PEEP and 30 mL kg–1, in spite of the fact that peak alveolar
pressure (33 cmH2O) was the same during the 5 h of ventilation.
Other groups have also shown in various animal models that PEEP has a protective effect on the development or extension of lung injury. Dreyfuss and Saumon, in
healthy rats, showed that the beneficial effects of PEEP on reducing the extent of lung injury could be minimized by the administration of dopamine16. They argued that
part of the reduction of edema formation with PEEP was a direct result of cardiac output reduction and decreased pulmonary perfusion. Recent data from Broccard et
al. also point to a relationship between pulmonary vascular flow and VILI18. This group studied the effect of three levels of constant intravenous infusion and two levels
of endinspiratory plateau pressure (P plat) at zero PEEP. Isolated rabbit lungs receiving 900 mL min–1 of perfusion at a Pplat of 30 cmH2O demonstrated the greatest
decrease in lung compliance, the highest level of alveolar hemorrhage, and the greatest weight gain as compared with animals receiving 300 mL min−1 perfusion at Pplat
of 30 cmH2O or 500 mL min−1 perfusion at Pplat of 15 cmH2O. These data support the suggestion that PEEP protects the lung by decreasing vascular flow.
Although vascular flow may affect the level of VILI, the data from Muscedere et al. support the independent effect of PEEP on the development of VILI19. This
group studied the effect of PEEP in an ex vivo rat lung model on the development of VILI. Four groups were compared (Figure 15.4): PEEP of 0 cmH2O (peak
pressure 30 cmH2O), PEEP less than Pflle (4 cmH2O, peak pressure 26 cmH2O), PEEP greater than Pflle (15 cmH2O, peak pressure
Figure 15.3 Comparison of left lungs excised from rats ventilated with peak pressure of 14 cmH2O, zero positive endexpiratory
pressure (PEEP); peak pressure of 45 cmH2O, 10 cmH2O PEEP; and peak pressure 45 cmH2O, zero PEEP (left to right).
The perivascular groove is distended with edema in the lungs from rats ventilated with peak pressure of 45 cmH2O and 10
cmH2O PEEP. The lung ventilated at 45 cmH2O and zero PEEP is grossly hemorrhaged. From: Webb, H.H. and Tierney, D.
(1974) Experimental pulmonary edema due to intermittent positive pressure ventilation, with high inflation pressures, protection
by positive endexpiratory pressure. Reproduced with permission from Am. Rev. Respir. Dis. 110:556–565.
Page 239
Figure 15.4 Composite pressurevolume curves before (circles) and after (squares) ex vivo ventilation with different levels of positive
ertdexpiratory pressure (PEEP), From: Muscedere, J.F., Mullen, B.M., Gan, K. and Slutsky, A.S. (1994) Tidal ventilation at
low airway pressures can augment lung injury. Reproduced with permission from Am. J Respir. Crit. Care Med. 149:1327–
1334,
32 cmH2O), and PEEP of 4 cmH2O (control, no ventilation), where Pflle is the inflection point on the inspiratory pressurevolume (P–V) curve of the lung. Following a
2 h ventilation period, animals in the PEEP greater than Pflle group demonstrated no change in compliance (similar to control animals), whereas the PEEP of 0 cmH2O
and PEEP less than Pflle groups showed a marked decrease in compliance.
15.1.3 Mechanism for PEEP protection

Although the precise mechanics by which PEEP attenuates VILI is still unclear, the most plausible cause was described by Mead et al. in 1970 using mechanical and
mathematical models20. Using the formula
where Peff is the effective stress pressure across the alveolar wall, Pappl is applied alveolar pressure, V is final alveolar volume and Vo is initial alveolar volume, they
argued that stress pressure much greater than the pressure applied to adjacent alveoli would be experienced by collapsed alveoli. They predicted, on the basis of the
above formula and the model in Figure 15.5, that an alveolar pressure of 35 cmH2O would result in stress pressure across the wall of a collapsed alveolus greater than
150 cmH2O. As noted in Figure 15.5, as the central collapsed alveolus is expanded the pressure within this alveolus is increased but when
Page 240
Figure 15.5 Crosssectional configuration and pressure within central unit (Pc) of a balloon model of the lung at several central unit
balloon volumes, while pressure in surrounding units (Palv) is held constant at 20 20 cmH2O. From; Mead, J., Takishima, T. and
Leith, D. (1 970) Stress distribution in lungs: a model of pulmonary elasticity. Reproduced with permission from J. Appl.
Physiol. 28:596–608
totally collapsed a large stress pressure gradient exists across adjacent alveolar walls. It is the repetitive opening and closing of these unstable lung units with each
breath that contributes to VILI. This type of VILI is referred to as atelectrauma. PEEP is thus believed to attenuate lung injury by preventing the recruitment and
derecruitment of these unstable lung units. Thus, the most appropriate PEEP level is one that maintains lung units open at end exhalation.
15.1.4 Translocation of cells

In addition to the injury already described, data from Fu et al. (Figure 15.6) clearly indicate that high lung volumes and low PEEP levels in animals result in fractures of
the capillary walls.21 If this is true in humans, cells and other substances can migrate from the
Figure 15.6 Scanning electron micrograph showing disruption of the bloodgas barrier with a red blood cell protruding from the fracture
of a rabbit lung perfused at 20 cmH2O transpulmonary pressure and 52.5 cmH2O transmural pressure. From: Fu, Z., Costello,
M.L., Tsukimoto, K., et al. (1992) High lung volume increases stress failure in pulmonary capillaries. Reproduced with
permission from J. Appl. Physiol. 73:123–133.
lung into systemic circulation. In fact, two studies in large animal models have clearly demonstrated that animals with their lung inoculated with bacteria and then
ventilated with a high VT and low PEEP developed bacteremia whereas those animals ventilated with a low VT and high PEEP did not develop bacteremia22,23.
15.1.5 Inflammatory mediator production

Cytokines and other mediators direct recruitment and activation of neutrophils at the alveolar level. These molecules are associated with inflammation and severe
disruption of normal lung function24. Increasing data are now available to support the hypothesis that the mechanical stress of inappropriate application of mechanical
ventilation activates inflammatory mediators in the Iung25–30. Tremblay et al. in an ex vivo healthy and injured rat lung model, demonstrated that both pro and
antiinflammatory mediators are activated by ventilatory patterns associated with high peak alveolar pressure causing overdistention and zero PEEP (Figure 15.7)29. In
those animals in
Page 241
Figure 15.7 Effect of ventilation strategy on absolute lung lavage cytokine concentration for animals previously lung injured with
lipopolyaccharide injection. (C, control; MVHP, moderate volume, high PEEP; MVZP, moderate volume, zero PEEP; HVZP,
high volume, zero PEEP: TNFα, tumor necrosis factor alpha; IL1β, interleukinlβ; MIP2 IFNγ, immune interferon.) The
pattern of lavage cytokines seen in response to ventilatioin strategy was similar to the salinetreated groups except for MIP2,
in which the control group (C) has comparable levels with the MVZP group (both increased significantly versus the MVHP
group), *P<0.05 versus C, MVHP, MVHP; ‡P<0.05 versus C, MVHP; ζP<0.05 versus C; ¶P<0.05 vs MVMP. From:
Tremblay, L., Valenza, F., Riberiro, S., et al. (1997) Injurious ventilatory strategies increases cytokines and cfos mRNA
expression in an isolated rat lung model. J. Clin. Invest. 99:944–952.
which PEEP was set greater than Pflle and overdistention was avoided, minimal mediator activation was observed. Similar data were reported by Bethmann et al. in an
ex vivo perfused mouse lung model26. Hyperventilation at 2.5 times normal transpulmonary pressure using either positive or negative pressure ventilation resulted in a
1.75fold increased expression of tumor necrosis factor alpha (TNFα) and interleukin6 messenger RNA (mRNA). Imai et al.28 and Takata et al.30 noted greater
TNFα mRNA expression with conventional ventilation at low peak pressures (30 cmH2O) but low PEEP (5 cmH2O) as compared with highfrequency oscillation at a
mean airway pressure of 15 cmH2O (same as with conventional ventilation). Ranieri et al. showed the same effect of ventilatory pattern on pulmonary lavage and
serum TNFα level in ARDS patients31. In a randomized comparison, patients ventilated with PEEP set greater than Pflle and peak alveolar pressure kept below the
upper inflection point of the pressurevolume curve had a decrease in lung lavage and serum TNFα after 36 h. Patients randomized to PEEP based on oxygenation and
VT set to produce eucapnia increased TNFα levels after 36 h. These data support the contention that inappropriate ventilatory patterns result in increased expression of
pro and antiinflammatory mediators leading to increasing lung inflammation and dysfunction.
Page 242
15.1.6 Multiorgan system failure

Both Drefuss and Saumon32 and Slutsky and Tremblay33 have proposed that VILI may cause or extend the level of multiple system organ failure (MSOF). As noted in
the previous discussion, an inappropriate ventilatory pattern can induce both mechanical failure of the lung and the release of inflammatory mediators. These
inflammatory mediators can move from the lung into the systemic circulation via fractures in pulmonary capillaries and affect organ systems external to the lung. Although
this is speculative, the evidence in animal models clearly indicates it is plausible. In fact, the lung may be the engine in many settings that drives the extent of MSOF33.
15.2 Lungprotective ventilatory strategies (LPVSs)

As already defined, an LPVS primarily consists of a ventilatory strategy that avoids overdistention and repetitive collapse of unstable lung units. However, an
increasingly important aspect of an LPVS primarily studied in laboratory models is the need to recruit the lung, that is to open collapsed, atelectatic lung units before
applying an LPVS.
15.2.1 P–V curve

Much of the approach to an LPVS is based conceptually on the implications of the inflation and deflation P–V curves of the respiratory system. As noted in Figure
15.8, two points on the inspiratory P–V curve are commonly referred to during LPVS; the Pflle and the upper inflection point (UIP). The Pflle is believed to be the
airway pressure where significant recruitment begins and the UIP the airway pressure where the rate of recruitment decreases and overdistention of select lung units
begins. It should, however, be emphasized that some ARDS patients do not demonstrate a Pflle because of markedly nonhomogeneous distribution of lung disease.
Generally, LPVSs have focused on ventilating between these two pressures, that is, setting PEEP above Pflle and maintaining peak alveolar pressure (endinspiratory
plateau pressure) below the UIP. The point of maximum compliance change (PMC) on the deflation limb has been referred to as the pressure where derecruitment
begins. However, the inflation and deflation curves are different, primarily as a result of hysteresis. This means that the lung volume maintained at a given pressure is
greater on the deflation than on the inflation limb. If the lung is fully recruited (sufficient airway pressure applied to open collapsed lung units) and PEEP is applied
without ventilator circuit discontinuation, the volume maintained at end exhalation is greater following recruitment than without a lung recruitment maneuver (RM) even if
the same PEEP level is applied.
Figure 15.8 Ideal pressurevolume curve of the respiratory system in ARDS, determined with a peak pressure of 45 cmH2O. Pflex,
lower inflection point; UIP, upper inflection point; PMC, point of maximum curvature on the deflation limb.
Page 243
15.2.2 Clinical data

There are five prospective, randomized clinical trails comparing conventional ventilatory strategies with LPVSs11,12,34–36. In each of these trials a low tidal volume was
compared with a larger tidal volume. However, in all trials the strategy for setting PEEP was the same in the control and treatment groups, except for the Amato et al.11
trial (Table 15.2). That trial used the results of a prerandomization P–V curve to identify PEEP level (Pflle+2 cmH2O) and determine maximum endinspiratory plateau
pressure. In addition, the Amato et al. trial was the only trial that used recruitment maneuvers: 35–40 cmH2O CPAP for 30–40 s before setting PEEP and whenever
the patient was disconnected from the ventilator. All other approaches to ventilator management were similar across both groups in all five studies.
Table 15.3 lists the mortality and number of subjects from each trial. Only two of the five trials were positive, with lower mortality in the treatment (LPVS) group
than in the control group. The other three studies demonstrated similar mortality, although a strong trend for better outcome in the control group existed in the Brochard
et al. trial36.
To understand why these differences existed and derive a recommended approach to ventilating ARDS patients from these data, careful comparison of actual data
for VT, plateau pressure (Pplat) and PEEP in each trial group (Figure 15.9) must be performed and compared with the actual protocols for each trial (Table 15.3). As
noted in Table 15.3, by design large differences in VT and as a result Pplat between groups should have occurred in each trial; however, actual data in Figure 15.9
would indicate that only in the Amato et al.11 and NIH (ARDSnet)12 trials were protocolized differences in VT and Pplat between groups maintained. In all three of the
other trials only small difference existed in VT and PPlat between control and LPVS groups. As a result, the VT difference signal may have been insufficient in the
Brochard et al.36, Stewart et al.35, and Brower et al.34 trials to produce an effect on outcome, whereas in the NIH12 and Amato et al.11 trials much larger difference
in VT and Pplat were maintained. In fact, in these trials over the first 7 days VT values of 6 and 12 mL kg−1 were maintained in the two groups. However, the question
to be asked is why the NIH12 trial required about 425 patients in each group to demonstrate a significant difference in mortality between groups when the Amato et
al.11 trial was able to demonstrate a difference in mortality with only 29 patients in the LPVS group and only 24 patients in the control group. The answer to this may
be in the approach to setting PEEP.
As noted in Table 15.3 and Figure 15.9, the Amato et al.11 trial was the only trial that used a different approach to setting PEEP in each group, and maintained the
difference throughout the trial. In the Amato LPVS group, initial PEEP was about 16 cmH2O and tended to decrease over time. In the Amato control group, and in the
LPVS and control groups of the other four trials (Figure 15.9), a low PEEP level (8–10 cmH2O) was used throughout the first week of the trial. It appears reasonable
to argue that the much smaller number of patients needed in the Amato trial versus the NIH trial to achieve significance was a result of the high PEEP used in the Amato
LPVS group. Two additional factors were different in the Amato LPVS group compared with all other
Table 15.2 LPVS trials outcome mortality
Amato11 Stewart35 Brochard36 Brower34 NIH12

LPVS 38% (29) * 46% (60) 47% (58) 50% (26) 31% (429)†
Control 71% (24) 48% (60) 38% (58) 46% (26) 40% (432)
†
*P<0.002; P=0.0054.
Values are percentages of patient dying, with number of patients in that group given in parentheses.
Page 244
Figure 15.9 LPVS trials with actual VT , plateau pressure and PEEP levels in the LPVS groups and control groups.
groups in the five trials: lung recruitment and the use of pressure control ventilation. On the basis of other available data, mode of ventilation is unlikely to have had an
impact on mortality; however, as discussed in the next section, lung recruitment and high PEEP tend to be most effective when used together.
Much criticism regarding the high mortality in the Amato11 control group (71%) has been raised. However, as noted in Table 15.4, although the patients were
younger in the Amato trial than in any of the other trials they also had a greater number of organ system failures (3.6) compared with the other trials. As a result, on the
basis of historical data, one would expect a high rate of mortality in the Amato control groups.
15.2.3 Ventilation guidelines (Table 15.5)

On the basis of the above data VT in ALI or ARDS should be maintained at about 6 mL kg−1 (8–4 mL kg−1) with endinspiratory plateau
Page 245
Table 15.3 LPVS trials; protocols

LPVS Control LPVS Control LPVS Control LPVS Control LPVS Control
VT (mL kg−1) <6 12 <8 10–15 7 10–15 8 10–12 6 12
Pressure Limit PIP 40 No limit P1P 30 PIP 50 P plat25 PIP 60 P plat 30 P plat 45 P plat 30 P plat 50
(cmH2O)
PEEP P flex+ 2 cm Based on oxygen Both based on Both based on Both set by F1O2/PEEP Both set by F1O2/PEEP
H20 ation oxygenation oxygenation Table Table
Table 15.4 LPVS trials; number of organ system failures and age

LPVS Control LPVS Control LPVS Control LPVS Control LPVS Control
Age 33±13 36±14 59±17 58±19 57±15 56±15 50±14 47±17 51±17 52±18
No. 3.6±1.3 3.6±1.5 1.4±1.0 1.3±1.2 1.3±0.6 1.3±0.7 NA NA 2.8±1.1 2.8±1.0
No. is total number of organ system failures. Age in years.
Page 246
Table 15.5 Ventilatory settings in ARDS
VT 6 mL kg−1 (range 4–8 mL kg−1)

Rate 25–35 min−1 (limit rate based on development of auto PEEP)
Inspiratory time <1.0 s controlled and
Mode assisted ventilation Pressure or volume assist/control (my preference pressure A/C)
PEEP 15 cmH2O (10–20 cmH2O) (set to prevent derecruitment after a lung recruitment maneuver)
FIO2 <0.6 if possible (set to achieve target PaO2 55–80 mmHg)

Recruitment at initiation of ventilation and whenever lung derecruited.
pressure less than 30 cmH2O. To maintain an adequate level of ventilation, respiratory rate must be increased, frequently into the upper twenties per minute (25–35
min−1). Respiratory rate is generally limited to the highest rate that does not cause autoPEEP. Mode of ventilation is based on clinician preference. With either pressure
or volume ventilation a VT of 6 mL kg−1 can be targeted. There are no convincing data supporting extending inspiratory time to establish an inverse I:E ratio37.
Particularly with high rates inspiratory time needs to be short (≤1.5 s) during controlled ventilation and even shorter (<1.0 s) during assisted ventilation. Sufficient PEEP
should be set to prevent endexpiratory collapse of unstable lung units. Generally, in adults with ARDS this is about 15 cmH2O, ranging in most patients between 10
and 20 cmH2O (see later sections). FIO2 is set to maintain the target PaO2 (55–80 mmH2O) after PEEP is set. Ideally, FIO2 should be less than 0.6. I would also
recommend the use of recruitment maneuvers at the onset of an LPVS and whenever derecruitment occurs (see later section). It should be remembered that the NIH
trial determined that a 6 mL kg−1 VT is better than a 12 mL kg−1 VT but did not test any other aspect of ventilatory support.
15.2.4 Permissive hypercapnia

Because ALI or ARDS patients are ventilated with small VT values, and as their overall dead space to tidal volume ratio is commonly elevated, maintaining a normal
PaCO2 in spite of increased respiratory rates is difficult. In some patients it becomes impossible to maintain the PaCO2 without increasing VT and Pplat. As a result,
permissive hypercapnia becomes the only option. That is, permissive hypercapnia should not be the desired outcome but the only acceptable outcome considering the
pathophysiology of the patient and the need for an LPVS38.
Generally, problems with permissive hypercapnia are not related to the PaCO2, but to pH and the effect pH has on other organ system function. That is, an
otherwise healthy chest trauma patient may easily tolerate a pH of 7.2 as a result of permissive hypercapnia, whereas a patient with sepsis and poor myocardial function
may not tolerate a pH of 7.3. Clearly, patients with sepsis and metabolic acidosis, increased intracranial pressure, myocardial dysfunction, and renal failure may not
tolerate the acidosis associated with the rapid development of permissive hypercapnia38. However, most patients will tolerate moderate permission hypercapnia
(PaCO2 50–60 mmHg) if it is allowed to develop slowly enough to insure the kidneys can compensate for the respiratory acidosis39. A PaCO2 rise of 1 mmHg per
hour or two in the appropriate patient is usually well tolerated40.
15.2.5 Lung recruitment

Lung recruitment maneuvers are the application of an airway pressure higher than that used to ventilate patients for a short period of time, to open collapsed lung. Lung
recruitment is not new and has been frequently proposed as part of the initial establishment of highfrequency oscillation41. The use of lung recruitment maneuvers (RM)
during conventional ventilation can be credited to Amato et al. in their LPVS trial11.
Page 247
Lungs are recruited because the collapsed lung requires greater pressure to ventilate, increases the risk of VILI (repetitive opening and closing), requires a high FIO2,
depresses surfactant production, and increases the likelihood of nosocomial pneumonia32,33,41. Numerous animal studies illustrate the ability of an RM to markedly
elevate PaO2 and to sustain the elevated PaO2 level over time with appropriate PEEP (Figure 15.10)42–46. In otherwise healthy, anesthetized, supine individuals 40
cmH2O CPAP sustained for 7–15 s is required to open atelectasis areas47, whereas, in alert and healthy individuals, Greaves et al. indicated that 30 cmH2O
transpulmonary pressure is required to reverse atelectasis48. Upwards to 60 cmH2O have been required in animal models of ARDS to recruit atelectatic lung42,43. In
patients, Gattinoni et al. reported needing 46 cmH2O peak airway pressure to fully recruit lung49, and in a single patient case report Medoff et al. reported the need
for 60 cmH2O (PEEP 40 cmH2O, pressure control 20 cmH2O, I:E 1:1,
Figure 15.10 Arterial oxygen tension (PaO2) median ±95% confidence interval for the recruitment maneuver group and control group in
lavage injured rabbits. RM consists of 30 crnH2O CPAP for 30 s. PEEP set above Pflex in both groups, *P <0.05. From;
Rimensberger, P.C., Cox, P.N., Frndova, H., et. al. (1999) The open lung during small tidal volume ventilation: concepts of
recruitment and optimal PEEP. Crit. Care Med 27:1946–1052.
rate 10 min–1 for 2 min) to recruit the lung of a young septic patient50.
The greatest concerns during a recruitment maneuver are barotrauma, hemodynamic compromise, and arterial desaturation. As a result, very careful monitoring of
the cardiopulmonary system during and after the recruitment maneuver is required (Table 15.6). Any time a patient becomes unstable or meets monitoring thresholds
the recruitment maneuver should be immediately discontinued.
On the basis of available patient data, recruitment maneuvers performed to a maximum pressure of 40 cmH2O appear to be safe11,47–51 However, patients selected
for RM should not have an existing pneumothorax or a high probability of developing a pneumothorax (severe COPD, pulmonary cysts or blebs), and should be
hemodynamically stable with adequate fluid balance. Ideally, RM should be performed early in the course of ALI or ARDS. Although data are not yet available, it
seems reasonable to assume that fibrotic lung in late ARDS is not as recruitable as atelectatic lung in early ARDS.
Recruitment maneuvers should always be performed with the mechanical ventilator, as disconnection of a manual ventilator and reconnection to the ICU ventilator
can result in derecruitment. In addition, in CPAP or pressure ventilation the target pressure is maintained, as additional volume is provided to accommodate the
recruited lung volume. I would recommend first trying an RM at 30 cmH2O CPAP for 30–40 s. If this is well tolerated, elevate the CPAP level to 35 then
Table 15.6 Monitoring during recruitment maneuvers47
Abort RM if:
• Mean arterial pressure <60 mmHg or decreased by >20 mmHg
• SpO2 <85%
• Heart rate >140 min–1 or <60 min–1
• New arrthythmias develop
Page 248
40 cmH2O. As indicated, few patient data are available at greater than 40 cmH2O RM pressure49,50. However, in many animal models, up to 60 cmH2O is required to
recruit the lung42,43,46 and multiple RM are needed to maximize the recruiting effect43,46,51. Finally, before any recruitment maneuver the patient should be stabilized on
100% oxygen.
15.3 Setting of PEEP

In patients with ARDS PEEP should always be set at the minimal level maintaining the PaO2 following recruitment. To determine this level, first maximally recruit the
lung. Then set PEEP at a level higher than anticipated (20 cmH2O), after which decrease the FIO2 until the SpO2 is 90–95%. Following stabilization slowly decrease
the PEEP in 2 cmH2O increment (every 15–20 min) until the patient desaturates (SpO2 < 90%). The PEEP level just preceding desaturation is the ideal PEEP level: the
lowest PEEP maintaining the oxygenation response of the RM. The lung should then again be recruited and PEEP set at the identified level. To insure maximum
recruitment is maintained, PEEP should be decreased from a level higher than needed rather than increased from a level lower than required. Whatever the method
used for setting PEEP, if the oxygenation benefit of the recruitment maneuver is lost over time the PEEP level is inadequate. Rerecruit the lung and set the PEEP at a
high level.
During weaning of FIO2 and PEEP, FIO2 should always be weaned before PEEP and PEEP not weaned until the FIO2 is less than 0.5. If during the weaning of
PEEP oxygenation decreases, the PEEP should always be restored. A decreased oxygenation during PEEP decrease always reflects derecruitment and requires
reestablishment of PEEP.
15.4 Prone positioning
Prone positioning recruits lung of ALI or ARDS patients, resulting in a PaO2 increase of 20 mmHg or more in 75% or more of patients52–57. Although the precise
mechanism for improvement in PaO2 is still controversial, the overall benefits of prone position are essentially a result of improved ventilationperfusion matching in the
prone versus the supine position58–61. A more uniform transpulmonary pressure gradient exists in the prone versus the supine position, resulting in a greater distribution
of ventilation to dependent lung in the prone position and better matching with perfusion59,61. Table 15.7 summarizes the mechanisms that contribute to better
distribution of ventilation to dependent lung in the prone position.
15.4.1 Clinical response

The most detailed clinical study of prone positioning is that of Chatte et al.52, who presented a case series of 32 consecutive patients with ALI or ARDS and
PaO2/FIO2 ratios of less than 150 mmHg. A total of 25 of 32 patients (78%) responded to prone positioning. However, as noted in Figure 15.11, the benefit of
prone positioning on oxygenation did not persist in all patients. Chatte et al. categorized some patients as persistent responders (oxygenation benefit persisted after
return to the supine position), nonpersistent responders (oxygenation benefit lost after return to the supine position), and nonresponders to prone positioning. No data
Table 15.7 Potential mechanisms for improved PaO2 in the prone position
Ventilation effects
• More homogeneous pleural pressure gradient
• Smaller percentage of lung dependent (triangular shape of lung)
• Less gravitational effect from heart or great vessels
• Altered shape of the thorax or decreased thoracic compliance
• Increased functional residual capacity (FRC)
• Mobilization of secretions Perfusion effect
• Less dependent perfusion in prone position Overall effect: improved ventilationperfusion ratio( )
From: Kacmarek, R.M. and Schwartz, D. (2000) Lung
recruitment. Respir. Clin. N. Am. 6:597–623. Reproduced with permission from W.B. Saunders.
Page 249
Figure 15.11 Evolution of PaO2/FIO2 before, during, and:after the first 4 h of a prone trial in different groups, Sbf, 1 h before prone;
PH1, first hour during prone; PH4, fourth hour during prone; Saft, 1 h after returning to supine; NR, nonresponders; RNP,
responders nonpersistent; RP, responders persistent. Results among the four times (ANOVA) are indicated with larger asterisk
than results between PH1, PH4, and Saft to Sbf (ttest), *P <0.05; **P <0.01; ***P < 0.001. From: Chatte, G., Sab, J.M.,
Dubois, J.M., et al. (199?) Prone position in mechanically ventliated patients with severe acute respiratory failure. Reproduced
with permission from Am, J. Respir. Crit. Care Med 155:473–478.
are available to prospectively differentiate patients into the three categories but it seems most likely that patients with welldefined dependent atelectasis or consolidation
in the supine position respond well to prone positioning whereas those with a more generalized pattern of atelectasis or consolidation are least likely to respond.
However, data from Pelosi et al.57 indicate that patients whose chest wall compliance decreases with prone positioning are most likely to respond to prone positioning.
In some patients a large quantity of secretions is removed when placed prone58. As a result, their oxygenation remains improved when returned to the supine position.
15.4.2 Risks of prone positioning

Placing critically ill patients in the prone position is not without risks (Table 15.8). The most
Table 15.8 Risks associated with prone positioning
• Hemodynamic instability
• Cardiac arrhythmias
• Arterial desaturation
• Vascular line occlusion or loss
• Inadvertent endotracheal extubation
• Dependent edema (facial edema)
• Apical atelectasis
• Skin breakdown
severe risk is the loss of the artificial airway or vascular lines. Although these are significant risks, they can be avoided by appropriate care during positioning.
Hemodynamic instability and desaturation are also potential problems. However, they generally can be avoided by appropriate fluid balance before positioning. Skin
breakdown and facial edema are the most distressing concerns of family, and family members should be alerted to these expected sequelae before patients are
positioned prone.
15.4.3 Prone positioning and lung protection

At least one animal study indicates that prone positioning results in less VILI than supine positioning62. In addition, prone positioning results in less need for FIO2 and
PEEP, and lung recruitment maneuvers in animal models work better in the prone than the supine position63. The prone position should be an integral part of an LPVS.
Prone positioning should be considered in any ALI or ARDS patient requiring greater than 60% O2 after RMs and appropriate adjustment of PEEP. As with other
LPVSs, the earlier in the course of the disease they are implemented the greater the likelihood for success.
15.4.4 How long to position prone

Few data on how long to maintain patients prone are available. However, if prone positioning continues to demonstrate a positive response, why move the patient back
to the supine position except for nursing care? Figure 15.12, from Fridrich et al.64, illustrates the continued benefit of 20 h per day of prone positioning. This appears
to be the ideal length of
Page 250
Figure 15.12 Course of PaO2/F1O2 ratios during four consecutive 24 h periods of prone positioning. In each period patients were prone
(P) 20 h and supine (S) 4 h. From: Fridrich, P., Krafft, P., Hochleuthner, H., et al. (1996) The effects of iongterm prone
positioning in patients with traumainduced adult respiratory distress syndrome. Anesth. Analg. 83:1206–1211.
time to maintain patients prone. The 4 h supine positioning period is generally sufficient for nursing care.
15.5 Summary
It is increasingly clear that the approach to mechanical ventilation does affect outcome in ARDS. From the onset of mechanical ventilation an LPVS should be
employed. This strategy is characterized by low VT, high PEEP following lung recruitment, and prone positioning. The critical concepts of this approach are to avoid
overdistention, opening the lung and keeping it open.
References
1. Egan, A. (1982) Lung inflation, lung solute permeability, and alveolar edema. J. Appl. Physiol 53:121–125.
2. Hernandez, L.A., Coker, P.J., May, S., et al. (1990) Mechanical ventilation increases microvascular permeability in oleic acidinjured lungs. J. Appl Physiol
69:2057–2061.
3. Parker, J.C., Hernandez, L.A., Longenecker, G.L., et al. (1990) Lung edema caused by high peak inspiratory pressures in dogs. Am. Rev. Respir. Dis. 142:321–
328.
4. Webb, H.H. and Tierney, D. (1974) Experimental pulmonary edema due to intermittent positive pressure ventilation, with high inflation pressures, protection by
positive endexpiratory pressure. Am. Rev. Respir. Dis. 110: 556–565.
5. Dreyfuss, D., Basset, G., Soler, P. and Saumon, G. (1985) Intermittent positivepressure hyperventilation with high inflation pressures produces pulmonary
microvascular injury in rats. Am. Rev. Respir. Dis. 132:880–884.
6. Kolobow, T., Moretti, M.P., Fumagalli, R., et al. (1987) Severe impairment in lung function induced by high peak airway pressure during mechanical ventilation.
Am. Rev. Respir. Dis. 135:312–315.
7. Tsuno, K., Miura, K., Takeya, M., et al. (1991) Histopathologic pulmonary changes from mechanical ventilation at high peak airway pressures. Am. Rev. Respir.
Dis. 143:1115–1120.
8. Wyszogrodski, I., KyeiAboagye, K., Taeusch, H.W. and Avery M. (1975) Surfactant inactivation by hyperventilation: conservation by endexpiratory pressure. J.
Appl. Physiol 38: 461–466.
9. Albert, R.K., Lakshminarayan, S., Hilderbandt, J., et al. (1979) Increased surface tension favors pulmonary edema formation in anesthetized dogs’ lung. J. Clin.
Invest 63: 1015–1018.
10. Dreyfuss, D., Soler, P., Basset, G. and Saumon, G. (1988) High inflation pressure pulmonary edema: respective effects of high airway pressure, high tidal volume,
and positive endexpiratory pressure. Am. Rev. Respir. Dis. 137: 1159–1164.
11. Amato, M.B.P., Barbas, C.S.V., Medeiros, D.M., et al. (1998) Effect of protectiveventilation strategy on mortality in the acute respiratory distress syndrome. N.
Engl. J. Med. 338:347–354.
12. ARDSnet (2000) Ventilation with lower tidal volume as compared with traditional tidal volume for acute lung injury and the acute respiratory distress syndrome. N.
Engl J. Med. 342: 1301–1308.
13. Pierson, D.J. (1995) Barotrauma and bronchopleural fistula. In: Principles and Practice of Mechanical Ventilation (ed. M.J.Tobin). McGrawHill, New York; pp.
813–836.
14. Maunder, R.J., Pierson, D.J. and Hudson, L.D. (1984) Subcutaneous and mediastenal emphysema: pathophysiology, diagnosis, and management. Arch. Intern.
Med. 144:1447–1453.
15. Dreyfuss, D. and Saumon, G. (1998) Ventilator induced lung injury: Lessons from experimental studies. Am. J. Respir. Crit. Care Med. 157: 294–323.
Page 251
16. Dreyfuss, D., Soler, P. and Saumon, G. (1995) Mechanical ventilationinduced pulmonary edema: interaction with previous lung alterations. Am. J. Respir. Crit.
Care Med. 151: 1568–1575.
17. Corbridge, T.C., Wood, D.H., Crawford, G.P. (1990) Adverse effects of large tidal volume and low PEEP in caine acid aspiration. Am Rev. Respir. Dis. 142:311–
315.
18. Broccard, A.F., Hotchkiss, J.R., Kuwayama, N., et al. (1998) Consequences of vascular flow on lung injury induced by mechanical ventilation. Am. J. Respir.
Crit. Care Med. 157: 1935–1942.
19. Muscedere, J.F., Mullen, B.M., Gan, K. and Slutsky, A.S. (1994) Tidal ventilation at low airway pressures can augment lung injury. Am. J. Respir. Crit. Care
Med. 149:1327–1334.
20. Mead, J., Takishima, T. and Leith, D. (1970) Stress distribution in lungs: a model of pulmonary elasticity. J. Appl. Physiol 28: 596–608.
21. Fu, Z., Costello, M.L., Tsukimoto, K., et al. (1992) High lung volume increases stress failure in pulmonary capillaries. J. Appl. Physiol. 73: 123–133.
22. Nahum, A., Hoyt, J., Schmitz, L., et al. (1997) Effect of mechanical ventilation strategy on dissemination of intratracheally instilled Escherichia coli in dogs. Crit.
Care Med. 25: 1733–1743.
23. Verbrugge, S.J.C., Sorm, V., Veen, A., et al. (1998) Lung overinflation without positive endexpiratory pressure promotes bacteremia after experimental Klebsiella
pneumoniae inoculation. Intensive Care Med. 24:172–177.
24. Christman, J.W. and Blackwell, T.S. (1998) Mechanical stress and cytokine production: implications for mechanical ventilation. Intensive Care Med. 24:884–885.
25. Rimensberger, P.C., Fedorko, L., Cutz, E. and Bohn, D.J. (1998) Attenuation of ventilatorinduced acute lung injury in an animal model by inhibition of neutrophil
adhesion by leumedine (NPC 15669). Crit. Care Med. 26: 3548–3555.
26. Bethmann, A.N., Brasch, F., Nusing, R., et al. (1998) Hyperventilation induces release of cytokines from perfused mouse lung. Am. J. Respir. Crit. Care Med.
157:263–272.
27. Wirtz, H.R.W. and Dobbs, L.G. (1990) Calcium mobilization and exocytosis after one mechanical stretch of lung epithelial cells. Science 250: 1266–1269.
28. Imai, Y., Kawano, T., Miyaska K., et al. (1994) Inflammatory chemical mediators during conventional ventilation and during high frequency oscillatory ventilation.
Am. J. Respir. Crit. Care Med. 150:1550–1554.
29. Tremblay, L., Valenza, F., Riberiro, S., et al. (1997) Injurious ventilatory strategies increases cytokines and cfos mRNA expression in an isolated rat lung model. J.
Clin. Invest. 99:944–952.
30. Takata, M., Abe, J., Tanaka, H., et al. (1997) Intraalveolar expression of tumor necrosis factor alpha gene during conventional and highfrequency ventilation. Am. J.
31. Ranieri, V.M., Tortorella, D., DeTullio, R., et al. (1999) Limitation of mechanical lung stress decreases BAL cytokines in patients with ARDS. JAMA 282:54–61.
32. Dreyfuss, D. and Saumon, G. (1998) From ventilatorinduced lung injury to multiple organ dysfunction? Intensive Care Med. 24:102–104.
33. Slutsky, A. and Tremblay, L. (1998) Multiple system organ failure: is mechanical ventilation a contributing factor? Am. J. Respir. Crit. Care Med. 157:1721–1725.
34. Brower, R.G., Shanholtz, C.B., Fessler, H.E., et al. (1999) Prospective, randomized, controlled trial comparing traditional versus reduced tidal volume ventilation in
acute respiratory distress syndrome patients. Crit. Care Med. 27: 1492–1496.
35. Stewart, T.E., Meade, M.O., Cook, D.J., et al. (1998) Evaluation of a ventilator strategy to prevent barotrauma in patients at high risk for acute respiratory distress
syndrome. N. Engl. J. Med. 338:355–361.
36. Brochard, L., RoudotThoravol, F., Roupie, E., et al. (1998) Tidal volume reduction for prevention of ventilatorinduced lung injury in acute respiratory distress
syndrome. Am. J. Respir. Crit. Care Med. 158:1831–1838.
37. Froese, A.B. (1997) High frequency oscillatory ventilation for adult respiratory distress syndrome: Let’s get it right. Crit. Care Med. 25: 906–908.
38. Hickling, K.G., Walsh, J., Henderson, S., et al. (1994) Low mortality rate in adult respiratory distress syndrome using lowvolume, pressurelimited ventilation with
permissive hypercapnia. A prospective study. Crit. Care Med. 22:1568.
39. Kacmarek, R.M. and Hickling, K.G. (1993) Permissive hypercapnia. Respir. Care 38:373.
40. Feihl, F. and Perret, C. (1994) Permissive hypercapnia. How permissive should we be? Am. J. Respir. Crit. Care Med. 150:1722.
41. Froese, A.B., McCulloch, P.R., Sugiura, M., et al. (1993) Optimizing alveolar expansion prolongs the effectiveness of exogenous surfactant therapy in the adult
rabbit. Am. Rev. Respir. Dis. 148:569–577.
Page 252
42. Sjostrand, U.H., LichtwarckAschoff, M., Nielsen, J.B., et al. (1995) Different ventilatory approaches to keep the lung open. Intensive Care Med. 21:310–318.
43. Fujino, Y., Goddon, S., Dolhnikoff, M., et al. (1999) Repetitive high pressure recruitment maneuvers (RM) required to maximally recruit lung in ARDS sheep
model. Am. J. Respir. Crit. Care Med. 159:A479.
44. Rimensberger, P.C., Cox, P.N., Frndova, H., et al. (1999) The open lung during small tidal volume ventilation: concepts of recruitment and optimal PEEP. Crit.
Care Med. 27:1946–1952.
45. Rimensberger, P.C., Pristine, G., Mullen, J.B.M., et al. (1999) Lung recruitment during small tidal volume ventilation allows minimal PEEP without augmenting
lung injury. Crit. Care Med. 27:1940–1945.
46. Van der Kloot, T.E., Blanch, L. and Youngblood, A.M. (2000) Recruitment maneuvers in three experimental models of acute lung injury. Am. J. Respir. Crit.
Care Med. 161:1485–1494.
47. Rothen, J.U., Neuman, P., Berglund, J.E., et al. (1999) Dynamics of reexpansion of atelectasis during general anesthesia. Br. J. Anaesth. 82: 551–556.
48. Greaves, I.A., Hildebrandt, J. and Hopppin, F.G. (1985) Micromechanics of the lung. In: Handbook of Physiology (eds P.T.MacKlem and J. Mead). American
Physiologic Society, Bethesda, MD, pp. 217–231.
49. Gattinoni, L., Pelosi, P., Crotti, S., et al. (1995) Effects of positive endexpiratory pressure on regional distribution of tidal volume and recruitment in adult
respiratory distress syndrome. Am. J. Respir. Crit. Care Med. 151: 1807–1995.
50. Medoff, B.D., Harris, R.S., Kesselman, H., et al. (2000) Use of recruitment maneuvers and high position endexpiratory pressure in a patient with acute
respiratory distress syndrome. Crit. Care Med. 18:1210–1216.
51. Lapinsky, S.F., Aubin, M., Mehta, S., et al. (1994) Safety and efficacy of a sustained inflation for alveolar recruitment in adults with respiratory failure. Intensive
Care Med. 25:1297–1301.
52. Chatte, G., Sab, J.M., Dubois, J.M., et al. (1997) Prone position in mechanically ventilated patients with severe acute respiratory failure. Am. J. Respir. Crit. Care
Med. 155:473–478.
53. Blanch, L., Mancebo, J., Perez, M., et al. (1997) Shortterm effects of prone position in critically ill patients with acute respiratory distress syndrome. Intensive
Care Med. 23: 1033–1039.
54. Jolliet, P., Bulpa, P., Ritz, M., et al. (1997) Additional beneficial effects of prone position, nitric oxide, and almitrine bimesylate on gas exchange and oxygen
transport in the acute respiratory distress syndrome. Crit. Care Med. 25: 786–794.
55. Langer, M., Mascheroni, D., Marcolin, R., et al. (1988) The prone position in ARDS patients. A clinical study. Chest 94:103–107.
56. Mure, M., Martling, C.R. and Lindahl, S.G. (1997) Dramatic effect on oxygenation in patients with severe acute lung insufficiency treated in the prone position.
Crit. Care Med. 25: 1539–1544.
57. Pelosi, P., Tubiolo, D., Mascheroni, D., et al. (1998) Effects of the prone position on respiratory mechanics and gas exchange during acute lung injury. Am. J.
Respir. Crit. Care Med. 157: 387–393.
58. Kacmarek, R.M. and Schwartz, D. (2000) Lung recruitment. Respir. Clin. N. Am. 6:597–623.
59. Gattinoni, L., Pelosi, P., Vitale, G., et al (1991) Body position changes redistribute lung computedtomographic density in patients with acute respiratory failure.
60. Amis, T.C., Jones, H.A. and Hughes, J.M. (1984) Effect of posture on interregional distribution of pulmonary ventilation in man. Respir. Physiol 56:145–167.
61. Albert, R.K., Leasa, D., Sanderson, M., et al. (1987) The prone position improves arterial oxygenation and reduces shunt in oleicacidinduced acute lung injury.
Am. Rev. Respir. Dis. 135:628–633.
62. Broccard, A.F., Shapiro, R.S., Schitz, L.L., et al. (1997) Influence of prone position on the extent and distribution of lung injury in a high tidal volume oleic acid
model of acute respiratory distress syndrome. Crit. Care Med. 25: 16–27.
63. Caker, N., Van der Kloot, T., Youngblood, M., et al. (2000) Oxygenation response to a recruitment maneuver during supine and prone positions in an oleic acid
induced lung injury model. Am. J. Respir. Crit. Care 161:1949–1956.
64. Fridrich, P., Krafft, P., Hochleuthner, H., et al. (1996) The effects of longterm prone positioning in patients with traumainduced adult respiratory distress
syndrome. Anesth. Analg. 83: 1206–1211.
Page 253
Chapter 16
Nosocomial pneumonia
Contents
16.1 Introduction
16.2 Risk factors for developing nosocomial pneumonia
16.3 Pathogens and routes of entry
16.4 Diagnosis
16.5 Antibiotic therapy
16.6 Choice of antibiotic
16.7 Prevention of nosocomial pneumonia
16.8 Antimicrobial resistance
16.9 Summary
Page 254
16.1 Introduction
Nosocomial or hospitalacquired pneumonias are infections of the lung parenchyma occurring more than 48 h after hospital admission and excluding infections
incubating at the time of admission1. Environmental sources (air, water, food, fomites), medical devices (endotracheal tubes, suction catheters, bronchoscopes,
respiratory therapy equipment), other patients, and hospital staff can serve as vectors for nosocomial infection2. Pneumonia is the most common nosocomial infection in
the ICU and the second most common hospitalacquired infection. Occurrence ranges from 0.4% in unselected hospitalized patients to 23% in the ICU3. The risk of
ventilatorassociated pneumonia (VAP; a subset of nosocomial pneumonia) is estimated to be 1% per day of mechanical ventilation4. More than half of patients with
ventilatorassociated pneumonia die during the same hospitalization, although not all deaths are directly related to pneumonia. The attributable mortality (percentage of
deaths that would not have occurred in the absence of infection) is lower. In a casecontrol study of 41 patients who developed nosocomial pneumonia, 44% of the
deaths were attributed to infection5.
ICU patients with any nosocomial infection have 2½ times the death rate of uninfected patients. Even after adjustment is made for confounding factors such as organ
system dysfunction, APACHE II score and intensity of therapy, the mortality risk is at least doubled6. Nosocomial infection has its greatest mortality impact in younger,
less severely ill patients, where failure to accurately diagnose and treat infection plays a larger role than in elderly patients with multiple mortality risks6. Nosocomial
pneumonia also prolongs hospital stay by an average of 7–9 days, and increases hospital expenses7.
16.2 Risk factors for developing nosocomial pneumonia
The risk of developing nosocomial pneumonia depends on the host, the environment and medical interventions (Table 16.1). Neurologic failure on the third ICU day5,
witnessed aspiration, exposure to paralytic agents8, supine position (head of bed <30°)9, sedative use10, nasogastric tubes, reintubation, blood transfusion11 and patient
transport from the ICU12 increase risk. Subsequent mortality, once pneumonia occurs, is related to APACHE II score, the number of dysfunctional organs, nosocomial
bacteremia, the presence of an underlying
Table 16.1 Risk factors for development of nosocomial pneumonia
Patient related
Alcoholism
Advanced age
Azotemia
Cigarette smoking
CNS dysfunction
Coma
COPD
Diabetes mellitus
Malnutrition
Number of organ dysfunctions
Physiologic derangement (APACHE score)
Underlying fatal illness
Environment
Condensation in ventilator tubing
Contaminated respiratory equipment
Head of bed in horizontal (<30°) position
Pooling of secretions over endotracheal tube cuff
Poor handwashing practice
Transfer from another ICU
Intervention
Antacids
Corticosteroids
Cytotoxic agents
H2 blockers or gastric overgrowth
Presence of endotracheal or nasogastric tube
Prolonged and inappropriate use of antibiotics
Prolonged or complicated surgery
Sedatives and neuromuscular blocking drugs
Thoracoabdominal procedures
CNS, central nervous system; COPD, chronic obstructive pulmonary disease; H2, histamine type 2. From: Higgins, T.L. (1999) Nosocomial pneumonia: intensive care unit perspective.
Reproduced with permission from Curr. Treatment Options Infect Dis. 1:159–175.
Page 255
fatal illness, and admission from another ICU13.
16.3 Pathogens and routes of entry

Organisms typically associated with communityacquired pneumonia (CAP) plus organisms in the hospital environment likely to colonize the upper airway are
responsible for most nosocomial infections. About half of nosocomial pneumonia isolates in mechanically ventilated patients are polymicrobial14,15. Viral etiologies must
also be considered. Pathogens frequently associated with nosocomial pneumonia (Table 16.2) include S. pneumoniae and Haemophilus influenzae with early onset
and aerobic Gramnegative bacilli or Staphylococcus aureus as hospital stay lengthens. A growing percentage of nosocomial organisms have become resistant to
antibiotics (see Section 16.8.1).
16.4 Diagnosis
16.4.1 Colonization
In clinically unstable patients who might also be febrile with an elevated white blood cell count, it is difficult not to react to a positive culture report from the
microbiology laboratory,
Table 16.2 Common pathogens currently associated with nosocomial pneumonia*
Pathogen Frequency, Source of organism

%
Early onset bacterial pneumonia
S. pneumoniae 5–20 Endogenous, other patients, respiratory
droplet
H. influenzae <5–15
Lateonset bacterial pneumonia
Aerobic Gram >20–60 Endogenous, other patients,
negative bacilli environment, enteral feeding, health
P. aeruginosa care workers, equipment or devices
Enterobacter spp.
Acinetobacter
spp.
K. pneumoniae
S. marcescens
E. coli
Grampositive cocci 20–40 Endogenous, healthcare workers,
S. aureus environment
Early and lateonset pneumonia
Anaerobic bacteria 0–35 Endogenous, potable water, showers,
L pneumophila 0–10 faucets, cooling towers
M. tuberculosis <1 Endogenous, other patients or staff
Viruses
Influenza A and B <1 Other patients or staff
Respiratory <1 Other patients or staff, fomites
syncytial virus
Fungi or Protozoa
Aspergillus <1 Air, construction, endogenous, other
Candida spp. <1 patients or staff
P. carinii <1 Endogenous, other patients or staff *Crude rates of pneumonia may vary by hospital, patient population, and method of diagnosis.
Adapted from: Craven, D.E. and Steger, K.A. (1995) Epidemioiogy of nosocomial pneumonia: New
perspectives on an old disease. Reproduced with permission from Chest 108(Suppl.): 1S–16S.
Page 256
yet it is critical to distinguish colonization of the airway from true pulmonary infection, particularly with conditions that mimic pneumonia on chest Xray (Table 16.3).
Organisms such as Enterobacter, Citrobacter, Flavobacterium, Pseudomonas cepacia, and Stentrophomonas (formerly Xanthomonas) maltophilia are unusual
causes of nosocomial pneumonia, but often appear as colonizers and prompt inappropriate antibiotic ‘coverage’. Staphylococcus aureus, whether methicillin sensitive
or multiresistant, is a common colonizer but only occasionally the cause of ventilatorassociated pneumonia. Colonization alone should not be treated, as inappropriate
use of powerful antimicrobials alters the microflora environment and encourages the emergence of highly resistant strains.
16.4.2 Clinical diagnosis

Although invasive strategies offer advantages in establishing the microbiologic cause of nosocomial pneumonia16, routine clinical practice is more likely to rely on non
invasive diagnosis. Clinical diagnosis of nosocomial pneumonia requires a chest radiographic abnormality that is new, progressive or persistent for more than 24 h, with
evidence of infec
Table 16.3 Noninfectious findings on chest Xray that may mimic nosocomail infection
ARDS (often associated with aspiration, pancreatitis, systemic inflammatory response syndrome, multisystem organ failure)
Atelectasis
Bronchiolitis obliterans organizing pneumonia (BOOP)
Cardiac pulmonary edema
Collagenvascular disorders
Contusion (trauma)
Drug or radiationinduced lung disease
Fat embolization
Neoplasm
Neurogenic pulmonary edema
Pulmonary hemorrhage
Pulmonary thromboembolism with infarction
From: Higgins, T.L. (1999) Nosocomial pneumonia: intensive care unit perspective. Reproduced with permission from Curr. Treatment Options Infect Dis. 1:159–175..
tion. Evidence should include at least two of the following: purulent sputum, temperature below 36°C or above 38°C, and a white cell count less than 5000 mm−3, or
10 000 mm−3 or more. The specificity of these findings for nosocomial pneumonia is low. Pneumonia is more likely when the chest radiograph shows rapid, progressive
cavitation, an air space process abutting a fissure, an air bronchogram, or infiltrate next to an empyema. Pneumonia is less likely if rapid improvement occurs with serial
examination, as can be seen with atelectasis, aspiration, hemorrhage, or congestive heart failure.
16.4.3 Microbiology laboratory results

Microbiology results help determine if infection explains new clinical signs and symptoms; identify the pathogen; and help to define prognosis. Semiquantitative reports
are generated by a serial plating technique (Figure 16.1). Growth in the first quadrant (1+) will almost always occur, but growth in the fourth quadrant (4+) suggests a
significant number of organisms. Unfortunately, processing time and inherent unreliability of bacteriologic sampling mandate a strategy of early empiric treatment. Blood
cultures are specific but not sensitive, identifying the pathogen in less than 20% of all patients with hospitalacquired pneumonia (HAP)17. If blood cultures are positive,
however, they define the organism and identify patients at risk for a complicated course.
Sputum examination and culture have limited value in diagnosing nosocomial pneumonia. In intubated patients, endotracheal aspirates (ETAs) are easily obtained, but
nonquantitative ETA specimens do not reliably distinguish infection from colonization. A quantitative ETA of more than 1 million colony forming units (c.f.u.) per ml is
highly suggestive of infection, but most laboratories do not perform quantitative cultures on ETAs. Elastin fibers, which can be identified using a 40% KOH stain,
suggest necrotizing pneumonia18. A negative sputum culture excludes certain pathogens, allowing the clinician to
Page 257
Figure 16.1 Semiquantitative cultures. Microbiology laboratories frequently report semiquantitative results based on organism growth
after incubation. The initial specimen is plated in the first quadrant of an agar plate. After flaming the loop, a specimen is drawn
from the first quadrant and plated in the second quadrant. The process is repeated from quadrant 2 to 3 and then from 3 to 4
After the growth is read out as scant (1+), light (2+), moderate (3+) or heavy (4+). Although quantitative cultures, this method
provides a way to communicate relative abundance to a clinically significant degree. More than one organism may be identified;
the relative abundance can help sort out clolonizers from true pathogens.
narrow the initially broad empiric antibiotic choices. The antimicrobial sensitivity of isolated organisms can be correctly defined 87% of the time by the endotracheal
aspirate19.
16.4.4 Fiberoptic bronchoscopy and bronchoalveolar lavage

Recent work has confirmed the value of an invasive diagnostic strategy in reducing mortality, organ failure and antibiotic use16. Fiberoptic bronchoscopy (FOB) allows
acquisition of protected brush specimens (PBSs) and/or bronchoalveolar lavage (BAL) specimens. Bacterial growth can then be quantified, typically using diagnostic
thresholds of 1000 c.f.u. ml−1 for bronchial brushings and 10000 c.f.u. ml−1 for BAL fluid cultures. These arbitrary thresholds may be too sensitive early in the course
of an illness, if the patient has already received antibiotic therapy, or as a result of variation in sampling techniques20. The additional risk and cost of FOB is offset in
part by improved outcome and lower antibiotic costs16,21. A recent study suggests that BAL with 3% infected cells is the only test with sufficient predictive value to
guide initial therapy
Page 258
while awaiting culture results22. Perhaps the most important contribution of a negative BAL or FOB is to redirect the search for infection in the febrile patient to
extrapulmonary sites. FOB and/or BAL should be strongly considered in the patient who fails to respond to therapy or where bronchial obstruction may be present.
16.4.5 Other invasive tests

Thoracentesis is not sensitive, but is highly specific in identifying infection. Thoracentesis is indicated if a parapneumonic effusion is present, particularly when the
effusion is large or the patient is seriously ill. Laboratory examination of pleural fluid should include protein, Lactate Dehydrogenase (LDH) and glucose levels, cell
count and differential, pH, Gram’s stain and acidfast stain, and bacterial and fungal cultures. Fluoroscopically guided transthoracic needle aspiration is highly specific
but subject to sampling false negatives and risks pneumothorax or hemorrhage. Open lung biopsy has a diagnostic yield of 69%, at the cost of additional morbidity and
mortality. Transtracheal aspiration of irrigated secretions using a small angiocath through the cricothyroid membrane is highly sensitive and specific, but has a high rate of
false positive in patients with COPD, and is not useful if the patient is already intubated.
16.4.6 Serology
Serologic studies are not routinely needed, but can be useful for retrospective confirmation of bacterial, viral and legionella infections. An enzymelinked immune serum
assay (ELISA) is available for rapid detection of legionella serogroup 1 antigen in the urine23.
16.5 Antibiotic therapy
Delay in administration of antibiotics increases mortality24 so therapy cannot be withheld while studies are pending. Empiric therapy should be initiated as soon as
possible, with only minimal delay to obtain pretherapy sputum samples and/or blood cultures. This chapter focuses on treatment of nosocomial pneumonia only;
treatment for communityacquired pneumonia requires a different strategy25.
16.5.1 ATS consensus statement

The American Thoracic Society (ATS)1 produced a consensus statement on hospitalacquired pneumonia in 1995. Although this is still a useful general framework for
empiric antibiotic therapy, changes in antibiotic resistance patterns, and the availability of newer agents such as the fluroquinolones modify these recommendations.
Treatment decisions are based on the severity of the patient’s illness, the presence of risk factors for a specific pathogen, and the time of onset of clinical infection
(before versus after day 5)1 (Figure 16.2). The ATS guidelines target empiric therapy based on individual patient risk factors. Highrisk patients (as defined in Table
16.4) receive broad spectrum coverage (accepting the risk of promoting antimicrobial resistance) whereas lowrisk patients receive more focused treatment. Empiric
antibiotic choices are
Table 16.4 Criteria for severe hospitalacquired pneumonia
Admission to the intensive care unit

Respiratory failure, defined as the need for mechanical ventilation or the need for >35% oxygen to maintain an arterial oxygen saturation >90%
Rapid radiographic progression, multilobar pneumonia, or cavitation of a lung infiltrate
Evidence of severe sepsis with hypotension and/or endorgan dysfunction:
• Shock (systolic blood pressure <90 mmHg or diastolic blood pressure <60 mmHg)
• Requirement for vasopressors for >4 h
• Urine output <20 ml h−1 or total urine output <80 ml in 4 h (unless another explanation is available)
• Acute renal failure requiring dialysis
Adapted from: American Thoracic Society (1995) American Thoracic Society Consensus Statement: Hospitalacquired pneumonia in adults: diagnosis, assessment of severity, initial
antimicrobial therapy, and preventative strategies. Reproduced with permission from Am. J. Respir. Crit. Care Med 153:1711–1725.
Page 259
Figure 16.2 Severity of illness. ‘Early onset’ is before hospital day 5. Adapted from: ATS guidelines American Thoracic Society (1995)
American Thoracic Society Consensus Statement: Hospitalacquired pneumonia in adults: diagnosis, assessment of severity,
initial antimicrobial therapy, and preventative strategies. Reproduced with permission from Am, J. Respir. Crit. Care Med.
153:1711–1725).
reassessed once culture and sensitivity information becomes available, generally within 48 h.
16.5.2 Severity of illness

Table 16.5 describes the expected microbiology and initial antibiotic choices for patients having mildtomoderate nosocomial pneumonia in the absence of significant
risk factors. Those criteria are also useful for the patient with earlyonset severe pneumonia unless immunosuppression is present. Table 16.6 addresses the organisms
and antibiotic choices in patients with host or therapeutic factors increasing risk for specific pathogens. The ATS guidelines consider patients admitted to the ICU or
those with severe organ dysfunction to automatically meet the criteria for severe hospitalacquired pneumonia. However, intensive care patients with no risk factors and
earlyonset (before day 5) pneumonia can be treated under the mildtomoderate guidelines in Table 16.5.
16.5.3 Risks for specific pathogens

Risk factors for Staphylococcus aureus infection include head trauma, diabetes mellitus, chronic renal failure, age less than 25 years, intravenous drug abuse, and
recent influenza infection26. Methicillinresistant Staphylococcus aureus (MRSA) is more likely with prior steroid administration, mechanical ventilation for more than 6
days, advancing age, preexisting COPD, and prior antibiotic therapy. Bacteremia, shock and subsequent mortality are all higher in ventilated patients infected with
MRSA than with sensitive strains27.
Page 260
Table 16.5 Antibiotic choices for patients with pneumonia
Core organisms Core antibiotics

Enteric Gramnegative bacilli Cephalosporin
(Nonpseudomonal) Enterobacter species Secondgeneration or nonpseudomonal thirdgeneration (e.g. cefotaxime or ceftraxone)
Escherichia coli
Klebsiella spp.
Proteus spp. OR
Serratia marcescens βlactamβlactamase inhibitor combination
Hemophilus influenzae OR
Methicillinsensitive Staphylococcus aureus Newer fluoroquinolone
Streptococcus pneumoniae
Patients have (a) mildtomoderate hospitalacquired pneumonia, no unusual risk factors, onset any time; or (b) earlyonset, severe hospitalacquired pneumonia (excluding patients with
immunosuppression). Adapted from: Higgins T.L. (1999) Nosocomial Pneumonia: Intensive Care Unit Perspective. Current Treatment in Infectious Disease. 1:159–175.
Table 16.6 patients with mildtomoderate hospitalacquired pneumonia with risk factors, onset any time*
Core organisms Plus Core antibiotics Plus

Anaerobes Clindamycin
(recent abdominal surgery, witnessed aspiration) or βlactamβlactamase inhibitor (alone)
Staphylococcus aureus ± Vancomycin
(coma, head trauma, diabetes mellitus, renal failure) (until methicillinresistant Staphylococcus aureus is ruled out)
Legionella Erythromycin ± rifampint
(highdose steroids)
Pseudomonas aeruginosa Treat as severe hospitalacquired
(prolonged ICU stay, steroids, antibiotics structural lung disease) pneumonia (Table 16.7)
*Excludes patients with immunosuppression.
†Rifampin may be added if Legionella sp. is documented. Adapted from: Higgins T.L. (1999) Nosocomial Pneumonia: Intensive Care Unit Perspective. Current Treatment in Infectious
Disease. 1:159–175.
Risk of Pseudomonas aeruginosa infection increases with prolonged hospital or ICU stay, prior antibiotics or steroids, the presence of structural lung disease such as
bronchiectasis or cystic fibrosis, malnutrition, and mechanical ventilation28. Legionella infection is likely with prior steroids, neutropenia, chronic renal failure,
chemotherapy, or active cancer.
Anaerobic infection is more likely with recent thoracoabdominal surgery, a witnessed aspiration, an obstructing foreign body, or poor dentition.
16.6 Choice of antibiotic

No single antibiotic or drug combination is appropriate for all patients. ‘Core’ antibiotics as defined by the ATS guidelines provide reasonable general empiric
coverage, and additional choices or substitutions can be based on the presence of risk factors and severity of illness1.
16.6.1 Treatment by risk category

Lowrisk patients should receive empiric coverage with one of the following: a second or nonpseudomonal thirdgeneration cephalosporin,
Page 261
or a βlactamβlactamase inhibitor combination. For patients with penicillin allergy, a fluroquinolone or the combination of clindamycin plus aztreonam would be
appropriate. The patient with mildtomoderate hospitalacquired pneumonia with risk factors (Table 16.6) should receive the ‘core’ antibiotics, plus additional risk
specific coverage: clindamycin or a βlactamβlactamase inhibitor for anaerobes, vancomycin for MRSA, and erythromycin or fluroquinolone with or without
rifampicin for Legionella. Patients at risk for Pseudomonas should be treated as having severe hospitalacquired pneumonia. Anaerobes isolated by protected
specimen brush usually (63%) produce βlactamase and thus are not reliably sensitive29.
The pathogens in patients with severe nosocomial pneumonia are the core organisms plus Pseudomonas aeruginosa, Acinetobacter, and MRSA. Earlyonset,
severe pneumonia in a patient without risk factors is likely to be caused by ‘core’ organisms and can be treated following the recommendations in Table 16.5. Late
onset severe nosocomial pneumonia or severe nosocomial pneumonia in the patient with risk factors requires aggressive empiric therapy. Recommendations include an
aminoglycoside or a fluroquinolone, plus one of the following: an antipseudomonal penicillin, a cephalosporin such as ceftriaxone or cefoperaxone, a βlactam β
lactamase inhibitor combination, a carbapenem, or aztreonam if only Gramnegative coverage is needed. If there is concern for MRSA, vancomycin should be added
(Table 16.6). Vancomycin should not be used for routine prophylaxis, treatment of colonization, empiric treatment of febrile or neutropenic patients, as firstline therapy
for Clostridium difficile colitis, or as a topical irrigant. Its use should be restricted to treatment of βlactamresistant Grampositive organisms, serious βlactam
allergy, and failed metronidazole therapy for C. difficile. Vancomycinresistant enterococci are an increasing problem, and may be associated with use of not only
vancomycin, but potent βlactam antibiotics as well30.
16.6.2 Combination antibiotic therapy versus monotherapy

Antibiotic monotherapy is acceptable for patients with nonpseudomonal mildtomoderate nosocomial pneumonia. Whether monotherapy is adequate for severely ill
patients is less clear. Monotherapy with meropenem versus the combination of ceftazidime plus tobramycin resulted in better clinical response and bacterial
eradication31. Overall success rate in a randomized trial of 177 patients with nosocomial pneumonia was only marginally improved by adding the aminoglycoside
netilmycin to imipenem. Even for patients with Pseudomonas infection, monotherapy was equally efficacious32. This lack of synergism may reflect a limitation of
aminoglycosides owing to poor penetration into the lung or inactivity at an acid pH. Monotherapy with thirdgeneration cephalosporins, however, is associated with
emergence of antimicrobial resistance33.
Combination therapy can be justified by reports of synergistic activity against certain pathogens, especially Pseudomonas aeruginosa, the need for broader
spectrum coverage than one agent can provide, and concerns about the emergence of resistance during monotherapy. Combination therapy generally employs
antibiotics with differing mechanisms of action, such as a βlactam agent plus an aminoglycoside or fluroquinolone. Table 16.7 indicates possible combinations for
severe nosocomial pneumonia. Combination drug therapy may be needed to cover both ‘core’ organisms and specific risk factors with lowergrade pneumonia,
particularly if the MRSA is a likely pathogen.
16.6.3 Antibiotic dosage and administration (Table 16.8)

Bacteriocidal antibiotics are either concentration dependent or time dependent. The aminoglycosides and quinolones are concentration dependent and most effective
when high peak concentrations are attained34. These drugs also have a postantibiotic effect, meaning they will suppress bacterial growth even when the serum
concentration falls below the minimal inhibitory concentration (MIC) of the infectious
Page 262
Table 16.7 Patients with severe hospitalacquired pneumonia with risk factors, early onset or patients with severe HAP, late onset*
Core organisms, Plus Therapy

P. aeruginosa Aminoglycoside or newer fluoroquinolone
Acinetobacter spp. plus one of the following:
antipseudomonal penicillin
βlactamβlactamase inhibitor
Ceftazidime, cefoperazone or cefepime
Imipenem or meropenem
Aztreonam†
Consider MRSA ± Vancomycin
*Excludes patients with immunosuppression.
†Aztreonam efficacy is limited to enteric Gramnegative bacilli and should not be used in combination with an aminoglycoside if Grampositive or Haemophilus influenzae infection is of
concern. Adapted from: Higgins T.L. (1999) Nosocomial Pneumonia: Intensive Care Unit Perspective. Current Treatment in Infectious Disease. 1:159–175.
agent. Concentrationdependent agents can usually be administered as oncedaily therapy, which reduces toxicity and costs without compromising the bacteriologic
cure35.
Penicillins, cephalosporins, aztreonam and vancomycin are timedependent antibiotics, and rely on keeping the serum level above the MIC threshold for the longest
possible time. Multiple daily doses are typically required to achieve adequate therapeutic effect with timedependent drugs. The carbapenems, unlike most βlactam
antibiotics, have a significant postantibiotic effect, allowing less frequent dosing intervals. Table 16.8 provides the specific drugs and usual dosing intervals for agents
commonly used in empiric therapy for nosocomial pneumonia.
16.6.4 Duration of therapy

Optimal duration of antibiotic therapy is not well defined and should be individualized based on the organism, severity of illness, and the patient’s response to therapy.
Infections with lowergrade organisms (Table 16.5) are likely to respond adequately to a 7–10 day course of therapy36. Patients who are malnourished or debilitated;
those with multilobar involvement, or those infected with Acinetobacter or Pseudomonas species or a necrotizing Gramnegative bacillus usually require 14–21 days
of therapy1. Whereas initial therapy should be provided with intravenous antibiotics, substitution of oral agents is appropriate and economical37, once the patient
demonstrates clinical improvement and the ability to tolerate enteral administration.
16.6.5 Assessing therapeutic response

Radiographic improvement of pneumonia is unlikely to occur in the first 72 h, so the initial antibiotic choice should be maintained unless there is rapid deterioration.
Serial cultures are appropriate to document antimicrobial eradication and to identify onset of superinfection. Protected specimen brush sampling yielding more than 103
c.f.u. ml−1 72 h after initiation of therapy indicates microbiologic failure to eradicate and better than 50% risk of clinical treatment failure38. Further diagnostic workup
including FOB, and, if necessary, open lung biopsy may identify unusual organisms, airway obstruction, or noninfectious processes when pneumonia fails to resolve.
Host defense problems may also contribute to nonresolution of infection, particularly in the elderly and patients with chronic obstructive pulmonary disease or diabetes
mellitus39.
16.6.6 Fungal pneumonia

Candida albicans is a frequent colonizer in the ICU setting, but is rarely the cause of pneumonia. Aspiration of Candida from the oropharynx will seldom cause a
primary pneumonia40. More commonly, hematogenously
Page 263
Table 16.8 Selected antibiotic choices for nosocomial pneumonia
Antibiotics by class Usual IV dose (adjust as needed for age and renal Notes
function)
Secondgeneration
cephalosporins
Cefoxitin (Mefoxin) 1 g q 8 h to 2 g q 4 h in vitro induction of βlactamase
Cefuroxime (Zinacef, Kefurox, 0.75 –1
gq8h
Ceftin)
Nonpseudomonal third
generation
Cefotaxime (Claforan) 1 g q 8 h to 2 g q 4 h (maximum 12 g day−1)
Ceftriaxone (Rocephin) 500 mg q 12 h to 2 g q day (q day gives better tissue levels) ‘Pseudocholelithiasis’ with TPN
Antipseudomonal
cephalosporins
Ceftazidime (Fortaz, Ceftaz, 1–2 g q 8–12 h selection of Vancomycin Resistant Enterococcus (VRE); ESBLs
Tazidime)
Cefoperazone (Cefobid) 2 g q 12 h to 4 g q 6 h (max. dose 8 g day−1) Disulfiramlike reactions q 6 h dosage for Pseudomonas ‘fourth’
generation; less resistance
Cefepime (Maxipime) 1–2 g q 12 h
βlactamβlactamase inhibitor
drugs
TicarcillinClavulanate (Timentin) 3.1 g q 4–6 h Ticarcillin interferes with platelets; Clavulanate may cause diarrhea
Piperacillintazobactam (Zosyn) 3.375 g q 6 h
Carbapenems and monobactams
ImipenemCilastin (Primaxin) 500 mg q 6 h; decrease in elderly and renal failure Avoid with Penicillin (PCN) allergy, seizures
Meropenem (Merrem) 1gq8h False elevation of serum and urine creatinine
Aztreonam (Azactam) 1 g q 8 h to 2 g q 6 h Gramnegative bacilli (GNB) coverage only; may crossreact with
ceftazidime allergy
Aminoglycosides
Gentamicin (Garamycin) 2 mg kg−1 load then 1.7 mg kg−1 Potential ATN, renal failure, deafness, vertigo, augmentation of
neuromuscular blockade
Tobramycin (Nebcin) −1
q 8 h or 5.1 mg kg once daily; check trough levels and
adjust for renal failure
Fluroquinolones
Ciprofloxacin (Cipro IV) 200–400 mg q 12 h Anaphylaxis; CNS toxicity enhanced Grampositive cocci (GPC) and
Levofloxacin (Levaquin IV) 500 mg q day atypical activity
Glycopeptides
Vancomycin (Vancocin) 15 mg kg−1 q 12 h; adjust for renal dysfunction using peak “red man” syndrome; with rapid infusion
and trough levels
Streptogramins
Quinupristin+dalfopristin (Synercid) 7.5 mg kg−1 q 8 h Reserve for highly resistant organisms including MRSA and some E.
faecium
Oxazolidinones
Linezolid (Zyvox) 600 mg q 12 h Reserve for MRSA, VRE, monitor platelets
Adapted from: Higgins T.L. (1999) Nosocomial Pneumonia: Intensive Care Unit Perspective. Current Treatment in Infectious Disease. 1:159–175.
Page 264
disseminated candidiasis will produce a miliary pattern on chest Xray, along with other organ involvement. Patients who are immunocompromised are at higher risk for
fungal infections, but even in this population, diagnosis should not be based on aspirated secretions alone, as benign airway colonization is far more likely than actual
pneumonia. Use of antifungal therapy with oral or tracheobronchial colonization is associated with emergence of resistant organisms. Recent practice guidelines
recommend histopathological confirmation of the diagnosis of Candida pneumonia before treatment41. Secondary pneumonia related to hematogenous spread should
be treated as disseminated candidiasis.
16.7 Prevention of nosocomial pneumonia

Only a few strategies to prevent nosocomial pneumonia (vaccination, isolation of patients with multiple resistant pathogens, and hand washing) are considered ‘probably
effective’1. Vaccination is effective in preventing pneumonia caused by Haemophilus influenzae, pneumococcus, and influenza virus42. Although of limited use for the
hospitalized patient, vaccination is indicated for those with chronic lung disease or otherwise at risk for hospitalization. Handwashing, although effective, is
underutilized43. The use of boxed, clean, nonsterile gloves over unwashed hands drastically reduces fingertip colony counts, and there is no additional benefit to prior
antiseptic handwashing or the use of individually packaged sterile gloves44. Nutritional support, control of gastric pH and volume, respiratory equipment protocols,
aspiration of subglottic secretions, and lateral rotation bed therapy are considered of ‘promising efficacy’ and are easily implemented. Biologic response modifiers,
monoclonal antibodies to specific bacterial antigens and manipulation of exogenous sources of bacteria (e.g. manipulation of the bacterial biofilm on endotracheal tubes)
are unproven regimens still being evaluated.
16.7.1 Airway issues

Tracheal and gastric intubation bypasses normal defense mechanisms, increasing the likelihood of colonization and aspiration. Routine ventilator tubing changes should
occur no more frequently than every 48 h45. Reintubation and nasal intubation (risk of maxillary sinusitis) should be avoided46. Microaspiration of secretions occurs
commonly, even in healthy patients, and aspiration risk is a function of the amount of time a patient is kept supine. Maintaining the head of the bed at an angle of 30° is
effective, nocost prophylaxis9,47. Patients maintained in a semirecumbent (versus supine) position for the first 24 h in the ICU have a lower rate of ventilator
associated pneumonia, and lower ICU mortality48. Specialized endotracheal tubes with an extra lumen above the cuff allow continuous or intermittent aspiration of
subglottic secretions. Removing contaminated secretions reduces the incidence of early, but not necessarily lateonset nosocomial pneumonia with more virulent
organisms49.
Specialized ICU beds that provide frequent changes in body position reduce the incidence of early ICU pneumonia in selected patients, but are without significant
effect on ventilation or ICU days, or hospital mortality50,51. Ventilatorassociated sinusitis is a major unrecognized cause of persistent fever, and frequently requires
drainage in addition to appropriate antibiotic therapy52.
16.7.2 Gastric acidity

Normal gastric flora is altered by an alkaline stomach pH53. Pneumonia incidence doubles in patients receiving stressprophylaxis antacids or histamine type 2
antagonists compared with sucralfate, which does not usually alter stomach acidity54. Gastric overdistention can precipitate aspiration53. Attention to gastric residual
volumes, cautious use of opioids, administration of motility agents such as metoclopramide, and use of postpyloric feeding tubes has experiential if not scientific
support.
Page 265
16.7.3 Selective digestive decontamination

The benefits of topical and systemic antibiotic prophylaxis in critically ill patients are contested. Early studies suggested that selective digestive decontamination (SDD)
reduced overall colonization and infection rates without affecting antibiotic resistance55,56. Later doubleblind, randomized trials, however, failed to document a survival
advantage, decreased nosocomial infection rate, reduced mechanical ventilation, or shorter ICU length of stay57,58. A recent metaanalysis concludes that antibiotic
prophylaxis with combined topical and systemic drugs reduces the rate of respiratory tract infections and overall mortality59. Institution of SDD may be effective at
controlling outbreaks of nosocomial pneumonia when traditional infection control measures have failed60.
Prophylactic treatment with aerosolized antibiotics leads to emergence of resistant organisms, and is no longer recommended except for the outpatient cystic fibrosis
population61. Immunoglobulin has been shown to reduce nosocomial infection, including pneumonia62, but further studies are needed to demonstrate costeffectiveness.
16.8 Antimicrobial resistance

The issue of antibiotic resistance should be of great concern to critical care practitioners given that the emergence of resistant organisms in recent years is occurring
more rapidly than the pharmaceutical industry can introduce new agents. It is particularly disturbing to note that enterococci (both Enterococcus faecium and
Enterococcus faecalis) resistant to the relatively new antibiotic, quinupristindalfopristin (Synercid) have been isolated from livestock where a related antibiotic,
virginiamycin, has been used as a growth promoter63.
16.8.1 Mechanisms of resistance

Bacteria defend themselves through a variety of strategies, including production of βlactamases and cephalosporinases, alterations of binding sites, changes in porin
size and perme ability, and production of antibioticmodifying enzymes. Resistance may be transmitted by plasmids or incorporated into the chromosomal structure of
the bacteria. The reader is referred to an excellent recent review for a complete description of mechanisms and potential therapeutic strategies64. Antibiotics (penicillins,
cephalosporins) containing a βlactam ring are hydrolyzed by βlactamases; this is the most prevalent mechanism of antibiotic resistance. βlactamases may be inducible
(i.e. encoded by chromosomes) or constitutive (chromosomally or plasmamediated). Transferable resistance affects βlactamβlactamase combination agents,
penicillins, cephalosporins, and monobactams. Plasmids coding for mutant βlactamases have been recovered from isolates of Klebsiella, E. coli, and other
Enterobacteriaceae. These extended spectrum βlactamases (ESBLs) cause nosocomial outbreaks. Emergence of resistant strains is encouraged by prolonged and/or
inappropriate use of antibiotics65. Organisms producing ESBLs can appear to be sensitive on routine susceptibility tests, but can be resistant to therapy in the clinical
setting.
Penicillin resistance in S. pneumoniae has risen from less than 5% in the 1980s to more than 30% at present66. Only half of US isolates of pneumococcus are now
fully susceptible to penicillin, resistance to cephalosporin varies from 4% to 92%, and resistance to macrolides ranges from 4% to 14%66. At this writing, all S.
pneumoniae isolates have thus far remained sensitive to vancomycin, although it may only be a matter of time until a case of vancomycinresistant pneumococcus is
reported. Vancomycinresistant strains of Staphylococcus aureus have recently been identified. New agents available to counter these threats include oxazolidinone
(Linezolid), streptogramins, such as quinupristindalfopristin (Synercid) and nextgeneration macrolides, fluoroquinolones and glycopeptides currently in development.
16.8.2 Rotation of routine antibiotic choices (crop rotation)

Crop rotation, or scheduled changes in the selection of empiric antibiotic therapy, may
Page 266
diminish the emergence of resistant organisms. In cardiac surgery patients given ceftazidime as usual empiric therapy for 6 months, changing empiric therapy to
ciprofloxacin for the next 6 months reduced resistance and the overall rate of ventilatorassociated pneumonia67. Further study is needed in different ICU patient
populations, and using different pairs of antibiotics. Alternatives to thirdgeneration cephalosporins include the βlactamβlactamase inhibitor combinations,
carbapenems, fluroquinolones, or fourthgeneration cephalospo rins which are more resistant to, and have lower affinity for common βlactamases than the third
generation drugs.
16.9 Summary
Treatment of nosocomial pneumonia requires a thoughtful approach, balancing the need to adequately cover potential pathogens against a risk of fostering antibiotic
resistance with inappropriately broad therapy. Prompt administration of antibiotics improves outcome. The spectrum of organisms to be covered is a function of the
patient’s risk factors, severity of illness, and length of hospitalization. With mildtomoderate nosocomial pneumonia and no risk factors, monotherapy with the ATS
defined ‘core’ antibiotics (a secondgeneration or nonpseudomonal thirdgeneration cephalosporin, a βlactamase inhibitor combination, or a fluoroquinolone should
suffice. Patients with specific risk factors should receive the ‘core’ antibiotics plus appropriate additional coverage for specific risks including recent abdominal surgery,
witnessed aspiration of gastric contents, coma, head trauma, diabetes mellitus, renal failure, chronic highdose corticosteroid therapy, prior antibiotic treatment,
structural lung disease, and prolonged ICU stay. Patients who require intubation and mechanical ventilation need more aggressive coverage, generally consisting of the
‘core’ antibiotics plus supplemental therapy for virulent pathogens such as Acinetobacter species and Pseudomonas species. BAL is the only test recognized as
having sufficient sensitivity to guide initial antibiotic choices. Initial multidrug coverage can then be focused based on microbiology results. Invasive diagnostic methods
have been shown to reduce mortality and inappropriate antibiotic use, and should also be considered in patients who fail to respond to initial empiric management.
Emergence of resistant organisms is of concern, but definitive recommendations regarding crop rotation or other strategies to counter resistance are not yet available.
References
1. American Thoracic Society (1995) American Thoracic Society Consensus Statement: Hospitalacquired pneumonia in adults: diagnosis, assessment of severity, initial
antimicrobial therapy, and preventative strategies. Am. J. Respir. Crit. Care Med. 153:1711–1725.
2. Craven, D.E. and Steger, K.A. (1995) Epidemiology of nosocomial pneumonia: New perspectives on an old disease. Chest 108(Suppl.): 1S–16S.
3. Tablan, O.C., Anderson, L.J., Arden, N.H., et al. (1994) Guideline for prevention of nosocomial pneumonia. Infect Control Hosp. Epidemiol 15:588.
4. Fagon, J.Y., Chastre, J., Domart, Y., et al. (1989) Nosocomial pneumonia in patients receiving continuous mechanical ventilation. Am. Rev. Respir. Dis. 139:877.
5. Girou, E., Stephan, F., Novara, A., et al. (1998) Risk factors and outcome of nosocomial infections: Results of a matched casecontrol study of ICU patients. Am. J.
6. BuenoCavanillas, A., DelgadoRodriguez, M., LopezLuque, A., SchaffinoCano, S. and GalvezVargas, R. (1994) Influence of nosocomial infection on
mortality rate in an intensive care unit. Crit. Care Med. 22:55–60.
7. Boyce, J.M., PotterBynoe, G., Dziobek, L. and Solomon, S.L. (1991) Nosocomial pneumonia in Medicare patients. Hospital costs and reimbursement patterns
under the prospective payment system. Arch. Intern. Med. 151:1109–1114.
8. Cook, D.J., Walter, S.D., Cook, R.J., et al (1998) Incidence of and risk factors for ventilatorassociated pneumonia in critically ill patients. Ann. Intern. Med.
129:433–440.
9. Drakylovic, M.B., Torres, A., Bauer, T, et al. (1999) Supine body position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: A
randomized trial. Lancet 354: 1851–1858.
Page 267
10. FernandezCrehuet, R., DiazMolina, C., de Irala, J., et al. (1997) Nosocomial infection in an intensivecare unit: Identification of risk factors. Infect Control
Hosp. Epidemiol 18:825–830.
11. LealNoval, S.R., MarquezVacaro, J.A., GarciaCuret, A., et al. (2000) Nosocomial pneumonia in patients undergoing heart surgery. Crit. Care Med. 28:935–
940.
12. Kollef, M.H., Von Harz, B., Prentice, D., et al. (1997) Patient transport from intensive care increases the risk of developing ventilatorassociated pneumonia. Chest
112:765–773.
13. Fagon, J.Y., Chastre, J., Vuagnat, A., et al. (1996) Nosocomial pneumonia and mortality among patients in intensive care units. JAMA 175:866–869.
14. Rouby, J.J., DeLassale, E.M., Poete, P., et al. (1992) Nosocomial bronchopneumonia in the critically ill: Histologic and bacteriologic aspects. Am. Rev. Respir. Dis.
146:1059–1066.
15. Fagon, J.Y., Chastre, J., Hance, A.J., et al. (1988) 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. 138: 110–116.
16. Fagon, J.Y, Chastre, J., Wolff, M., et al. (2000) Invasive and noninvasive strategies for management of suspected ventilatorassociated pneumonia. Ann. Intern.
Med. 132:621–630.
17. Bryan, C.S. and Reynolds, K.L. (1984) Bacteremic nosocomial pneumonia. Am. Rev. Respir. Dis. 129:688–691.
18. Shales, D.M., Ledeman, M.M., Chmielewski, R., et al. (1983) Elastin fibers in the sputum of patients with necrotizing pneumonia. Chest 83: 885–889.
19. Kirkland, S.H., Corkey, D.E., Winterbauer, R.H., et al. (1997) The diagnosis of ventilatorassociated pneumonia. A comparison of histologic, microbiologic and
clinical criteria. Chest 112:445–457.
20. Niederman, M.S., Torres, A. and Summer, W. (1994) Invasive diagnostic testing is not needed routinely to manage suspected ventilatorassociated pneumonia. Am.
J. Respir. Crit. Care Med. 150:565–569.
21. Croce, M.A., Fabian, T.C., Shaw, B., et al. (1994) Analysis of charges associated with diagnosis of nosocomial pneumonia: Can routine bronchoscopy be justified?
J. Trauma 37: 721–727.
22. Veber, B., Souweine, B., Gachot, B., et al. (2000) Comparison of direct examination of three types of bronchoscopy specimens used to diagnose nosocomial
pneumonia. Crit. Care Med. 28:962–968.
23. Schluger, N.W. and Rom, W.N. (1995) The polymerase chain reaction in the diagnosis and evaluation of pulmonary infections. Am. J. Respir. Crit. Care Med.
142:11–16.
24. Meehan, T.P., Fine, M.J., Krumholz, H.M., et al. (1997) Quality of care, process and outcomes in elderly patients with pneumonia. JAMA 278: 2080–2084.
25. Heffelfinger, J.D., Dowell, S.F., Jorgensen, J.H., et al. (2000) Management of communityacquired pneumonia in the era of pneumococcal resistance. A report
from the DrugResistant Streptococcus pneumoniae Therapeutic Working Group. Arch. Intern. Med. 160: 1399–1408.
26. Rello, J., Quintara, E., Ausina, V., et al. (1990) Risk factors for Staphylococcus aureus nosocomial pneumonia in critically ill patients. Am. Rev. Respir. Dis.
142:1320–1324.
27. Rello, J., Torres, A., Ricart, M., et al. (1994) Ventilatorassociated pneumonia by Staphylococcus aureus. Comparison of methicillinresistant and methicillin
sensitive episodes. Am. J. Respir. Crit. Care Med. 150:1545–1549.
28. Rello, J., Austina, V. and Ricart, M. (1993) Impact of previous antimicrobial therapy on the etiology and outcome of ventilatorassociated pneumonia. Chest
104:1230–1235.
29. Robert, R., Grollier, G., Dore, P., et al. (1999) Nosocomial pneumonia with isolation of anaerobic bacteria in ICU patients: Therapeutic considerations and outcome.
J. Crit. Care 14: 114–119.
30. Donskey, C.J., Schreiber, J.R., Jacobs, M.R., et al. (1999) A polyclonal outbreak of predominantly VanB vancomycinresistant enterococci in Northeast Ohio.
Clin. Infect Dis. 29: 573–579.
31. Sieger, B., Berman, S.J., Geckler, R.W., et al. (1997) Empiric treatment of hospitalacquired lower respiratory tract infections with meropenem or ceftazidime with
tobramycin: A randomized study. Crit. Care Med. 25:1663–1670.
32. Cometta, A., Baumgartner, J.D., Lew, D., et al. (1994) Prospective randomized comparison of imipenem monotherapy with imipenem plus netilmicin for treatment
of severe infections in nonneutropenic patients. Antimicrob. Agents Chemother. 38:1309–1313.
33. Chow, J.W., Fine, M.J., Shlaes, D.M., et al. (1991) Enterobacter bacteremia: Clinical features and emergence of antibiotic resistance during therapy. Ann. Intern.
Med. 115:585–590.
34. Craig, W. (`1993) Pharmacodynamics of antimicrobial agents as a basis for determining dosage regimens. Eur. J. Clin. Microbiol. Infect. Dis. 12: s6–s8.
Page 268
35. Hatala, R., Dinh, T. and Cook, D.J. (1996) Oncedaily aminoglycoside dosing in immunocompetent adults: A metaanalysis. Ann. Intern. Med. 124:717–725.
36. Schleupner, C.J. and Cobb, D.K. (1992) A study of the etiologies and treatment of nosocomial pneumonia in a communitybased teaching hospital. Infect Control
Epidemiol. 13: 515–525.
37. Paladino, J.A., Sperry, H.E., Backes, J.M., et al. (1991) Clinical and economic evaluation of oral ciprofloxacin after an abbreviated course of intravenous
antibiotics. Am. J. Med. 91: 462–470.
38. Montravers, P., Fagon, J.Y., Chastre, J., et al. (1993) Followup protected specimen brushes to assess treatment in nosocomial pneumonia. Am. Rev. Respir. Dis.
147:38–44.
39. Augustine, G., Fein, A.M., Feinsilver, S.H., et al. (1992) When pneumonia fails to resolve: Risk factors and diagnostic options. J. Crit. Illness 7:213–229.
40. ElEbiary, M., Torres, A., Fabrega, N., et al. (1997) Significance of the isolation of Candida species from respiratory samples in critically ill, nonneutropenic
patients. Am. J. Respir. Crit. Care Med. 156:583–590.
41. Rex, J.H., Walsh, T.J., Sobel, J.D., et al. (2000) Practice guidelines for the treatment of candidiasis. Clin. Infect. Dis. 30:662–676.
42. Gross, P.A., Hermogenes, A.W., Sacks, H.S., et al. (1995) The efficacy of influenza vaccine in elderly persons: a metaanalysis and review of the literature. Ann.
Intern. Med. 123:518–527.
43. Albert, R.K. and Condie, F. (1981) Handwashing patterns in medical intensive care units. N. Engl. J. Med. 304:1465–1466.
44. Rossoff, L.J., Borenstein, M. and Isenberg, H.D. (1995) Is hand washing really needed in an intensive care unit? Crit. Care Med. 23: 1211–1216.
45. Craven, D.E., Connolly, M.G., Lichtenberg, D.A., et al. (1982) Contamination of mechanical ventilators with tubing changes every 24 or 48 hours. N. Engl J.
Med. 306:1505–1509.
46. Rouby, J.J., Laurent, P., Gosnach, M., et al. (1994) Risk factors and clinical relevance of nosocomial maxillary sinusitis in the critically ill. Am. J. Respir. Crit.
Care Med. 150:776–783.
47. Torres, A., SerraBatlles, J., Ros, E., et al. (1992) Pulmonary aspiration of gastric contents in patients receiving mechanical ventilation: The effect of body position.
Ann. Intern. Med. 116:540–543.
48. Kollef, M.H. (1993) Ventilatorassociated pneumonia. A multivariate analysis. JAMA 270: 1965–1970.
49. Valles, J., Artigas, A., Rello, J., et al. (1995) Continuous aspiration of subglottic secretions in preventing ventilatorassociated pneumonia. Ann. Intern. Med.
122:179–186.
50. Fink, M.P., Helsmoortel, C.M., Stein, K.L., et al. (1990) The efficacy of an oscillating bed in the prevention of lower respiratory tract infection in critically ill
victims of blunt trauma. Chest 97:132–137.
51. DeBoisblanc, B.P., Castro, M., Everret, B., et al. (1993) Effect of airsupported, continuous postural oscillation on the risk of early ICU pneumonia in nontraumatic
critical illness. Chest 103:1543–1547.
52. Souweine, B., Mom, T., Traore, O., et al. (2000) Ventilatorassociated sinusitis. Microbiological results of sinus aspirates in patients on antibiotics. Anesthesiology
93:1255–1260.
53. Donowitz, L.G., Page, M.C., Mileur, G.L., et al. (1986) Alteration of normal gastric flora in critically ill patients receiving antacid and cimetidine therapy. Infect.
Control Hosp. Epidemiol. 7:23–36.
54. Driks, M.R., Craven, D.E., Celli, B.R., et al. (1987) Nosocomial pneumonia in intubated patients given sucralfate as compared with antacids or histamine type 2
blockers. N. Engl. J. Med. 317:1376–1382.
55. Reidy, J.J. and Ramsey, G. (1990) Clinical trials of selective decontamination of the digestive tract: Review. Crit. Care Med. 18: 1449–1456.
56. Cockerill, F.R., Muller, S.R., Anhalt, J.P., et al. (1992) Prevention of infection in critically ill patients by selective decontamination of the digestive tract. Ann.
Intern. Med. 117: 545–553.
57. Gastinne, H., Wolff, M., Delatour, F., et al. (1992) A controlled trial in intensive care units of selective decontamination of the digestive tract with nonabsorbable
antibiotics. N. Engl. J. Med. 326:594–599.
58. Ferrer, M., Torres, A., Gonzalez, J., et al. (1994) Utility of selective digestive decontamination in mechanically ventilated patients. Ann. Intern. Med. 120:389–
395.
59. D’ Amico, R., Pifferi, S., Leonetti, C., et al. (1998) Effectiveness of antibiotic prophylaxis in critically ill adult patients: Systematic review of randomized controlled
trials. Br. Med. J. 316: 1275–1285.
60. Nouira, S., Elatrous, S., Boukef, R., et al. (1998) Effectiveness of selective digestive tract decontamination to control an outbreak of nosocomial pneumonia caused
by Pseudomonas aeruginosa in mechanically ventilated patients. Clin. Int Care 9:180–184.
Page 269
61. Ramsey, B.W., Pepe, M.S., Quan, J.M., et al. (1999) Intermittent administration of inhaled tobramycin in patients with cystic fibrosis. N. Engl J. Med. 340:23–30.
62. The Intravenous Immunoglobulin Collaborative Study Group (1992) Prophylactic intravenous administration of standard immune globulin as compared with
corelipopolysaccharide immune globulin in patients at high risk of postsurgical infection. N. Engl. J. Med. 327:234–240.
63. Welton, L.A., Thal, L.A., Perri, M.B., et al. (1998) Antimicrobial resistance in enterococci isolated from turkey flocks fed virginiamycin. Antimicrob. Agents
Chemother. 42: 705–708.
64. Gold, H.S. and Moellering, R.C. (1996) Antimicrobial drug resistance. N. Engl. J. Med. 335:1445–1453.
65. Thomson, K.S., Prevan, A.M. and Sanders, C.C. (1996) Novel plasmidmediated βlactamases in enterobacteriaceae: Emerging problems for new βlactam
antibiotics. Current Clin. Topics Infect Dis. 16:151–163.
66. Doern, G.H., Pfaller, M.A., Kugler, K., et al. (1998) Prevalence of antimicrobial resistance among respiratory tract isolates of S. pneumoniae in North America.
Clin. Infect Dis. 27:764–780.
67. Kollef, M.H., Vlasnik, J., Sharpless, L., et al. (1997) Scheduled change of antibiotic classes. A strategy to decrease the incidence of ventilatorassociated
pneumonia. Am. J. Respir. Crit. Care Med. 156:1040–1048.
Page 270

Page 271
Chapter 17
Obstructive lung disease: asthma and COPD
Contents
17.1 Introduction
17.2 AutoPEEP
17.3 Nonventilatory management
17.4 Intubation
17.5 Mechanical ventilation
17.6 Weaning from ventricular support
Page 272
17.1 Introduction
Obstructive lung disease, classified largely as chronic obstructive lung disease (COPD) and asthma, is common in the population at large and thus commonly affects
patients in the ICU. Although similar in their primary physiologic abnormality ‘airway obstruction’, the type of obstruction does differ between asthma and COPD. With
COPD the obstruction is dynamic, increasing during expiration but not inspiration and not fully reversible. In contrast, the obstruction during an acute asthmatic attack is
fixed, making both inspiration and expiration difficult. However, the obstruction is fully reversible; between acute events respiratory mechanics are frequently normal.
In both asthma and COPD, patients present with inflammation, bronchospasm and hypersecretion of mucus caused by a variety of infectious and noninfectious
agents. This leads to increased airway narrowing that aggravates ventilationperfusion matching and causes increased air trapping. Treatment is focused on reversing the
cause of the acute problem and reversing the inflammation, bronchospasm, and hypersecretion. However, many of these patients require ventilatory support.
This chapter will discuss the treatment of patients with asthma and COPD in the ICU. Because autoPEEP is an important development in such patients, autoPEEP
is discussed first, followed by a brief review of pharmacologic management. Because many patients with obstructive lung disease in the ICU may be mechanically
ventilated, intensivists must be aware of the unique ventilatory issues associated with airflow obstruction. The remainder of the chapter addresses ventilatory
management.
17.2 AutoPEEP
Air trapping and autoPEEP as a result of severe airflow limitation are the primary pathophysiologic abnormalities leading to mechanical ventilation of patients
presenting in an exacerbation of COPD or severe acute asthma. AutoPEEP is a primary cause of patient ventilator dyssynchrony but also exerts similar
cardiopulmonary effects as applied PEEP1. Functional residual capacity is increased, intrathoracic pressure is increased, hemodynamics are compromised and
ventilationperfusion mismatch is increased1. In addition, work of breathing is markedly increased and the efficiency of diaphragmatic contraction decreased2. Air
trapping and autoPEEP flatten the diaphragm, frequently eliminating its inspiratory muscle capabilities and changing its motion to expiratory in function2. The altered
pulmonary mechanics observed as autoPEEP increases frequently leads to the ventilatory muscle dysfunction and blood gas abnormalities that require initiation of
mechanical ventilation in COPD and asthma.
The obstruction observed during asthma and COPD does differ. In COPD, the obstruction is primarily expiratory and varies greatly with ventilatory pattern. In
asthma, the obstruction is fixed (both inspiratory and expiratory) and frequently so severe that the extent of autoPEEP is underestimated3. As demonstrated by
Leatherman and Ravenscraft in severe asthma, at endexpiration some airways are completely closed.3 As a result, the autoPEEP behind these obstructions is never
measured. As shown in Figure 17.1, the measured autoPEEP is 5 cmH2O but the actual air trapping generates autoPEEP levels up to 20 cmH2O. This problem is
much less likely to occur in COPD; because of the dynamic nature of the autoPEEP, a relaxed endexpiratory measurement generally results in the true average level of
autoPEEP across lung units.
17.3 Nonventilatory management
Tables 17.1 and 17.2 summarize the nonventilatory management of acute exacerbations of COPD and severe asthma4,5. Although similar in many respects, there are
distinct differences between asthma and COPD that affect management. Frequently, severe exacerbations of COPD are a result of both viral and bacterial infection.
Careful identification of the
Page 273
Figure 17.1 AutoPEEP is the presence of airway closure during exhalation, Using a static endexpiratory occlusion method, the
measured level of autoPEEP is 5 cmH2O, The actual level of autoPEEP not communicating with the central airway during
measurement in some severe asthmatic patients may greatly exceed the measured level. This is a result of complete airway
obstruction in these units at end exhalation. From; Leatherman, J.W, and Ravenscraft, S.A. (1996) Low measured autopositive
endexpiratory pressure during mechanical ventilation of patients with severe asthma: Hidden autopositive endexpiratory
pressure. Reproduced with permission from Crit. Care Med. 24:541–549.
causative organism and appropriate antibiotic therapy are indicated especially in acute purulent COPD exacerbation. Table 17.3 provides recommendations for
antibiotic therapy in an acute exacerbation of COPD6. Empirical treatment with antibiotics has become a basic aspect of managing acute exacerbations of COPD
characterized by increased volume and purulence of secretions, even if specific organisms cannot be identified. In asthma, antibiotic therapy is normally reserved for
those situations where fever is present, the sputum contains neutrophils, or clinical evidence of bacterial pneumonia or sinusitis is present5. Most pneumonias in status
asthmaticus are viral in origin.
17.3.1 Bronchodilators
In both asthma and COPD, bronchodilators are the mainstay of therapy. In severe asthma, rapid administration of repeated doses of albuterol is recommended5. In
many situations, continuous or nearcontinuous initial administration is necessary. In COPD, the norm is standard doses of aerosolized bronchodilators administered
every 4 h. However, many also recommend the use of anticholinergics instead of, or in addition to adrenergic agents in managing patients with COPD4,6. In addition to
bronchodilation, anticholinergics reduce mucous hypersecretion and dyspnea4. Many, however, will prefer an adrenergic agent as initial therapy over anticholinergic
agents because of more rapid response (5–10 min)4.
Table 17.1 Medical treatment of an acute exacerbation of COPD
Firstline therapies Example Dosages

Beta2 agonists Albuterol (salbutamol) 2.5 mg in NSS via nebulizer or 4 puffs (400 mg) by metered dose inhaler (MDI) with spacer every 4 h
Anticholinergics Ipratropium bromide 0.5 mg via nebulizer or 2 puffs (18 µg) by MDI with a spacer every 6 h
Corticosteroids Methylprednisolone 60–125 mg IV every 6 h for 1–3 days, then change to oral prednisone and taper off over 2 weeks
Antibiotics See Table 17.3 See Table 17.3
Oxygen Cannula, ventimask or mask Titrate to keep SaO2 >90%
Page 274
Table 17.2 Medical treatment of status asthmaticus
Examples Dosages
Firstline therapies
β2agonists Albuterol (salbutamol) 2.5 mg in normal saline via nebulizer or 4 puffs (400 µg) by MDI with a spacer every 20 min×3
Metaproterenol (orciprenaline) 15 mg via nebulizer every 20 min×3 or 3 puffs (1.95 mg) by MDI with a spacer q 20 min×3
Epinephrine (adrenaline) 0.3 ml 1:1000 solution SC every 20 min×3
Corticosteroids Prednisone 15–240 mg daily in divided doses
Methylprednisolone 60–125 mg IV every 6–8 h
Oxygen Titrate to keep SaO2 >90%
Secondline therapies
Methylxanthines Theophylline Load: 5–6 mg kg–1 IV over 20–30 min (reduced loading dose in patients already taking theophylline preparations)
Maintenance: 0.6 mg kg–1 h–1 IV; titrate to serum theophylline concentration 8–15 µg ml–1
Anticholinergics Ipratropium bromide 0.5 mg via nebulizer every h×3
Adapted from: Levy, B.D., Kitch, B. and Fanta, C.H. (1998) Medical and ventilatory management of status asthmaticus. Reproduced with permission from Blackwell Science publishers.
Intens. Care Med. 24:105–117.
17.3.2 Corticosteroids
Systemically administered corticosteroids are a mainstay in the management of severe asthma. Therapy should be initiated upon admission and continued for 7–14
days, then tapered and eventually changed to inhaled preparations for those appropriately responding. In COPD, the use of inhaled steroids is more controversial. The
American Thoracic Society recommends steroid use in patients with an asthmatic component to their COPD but also recommends rapid discontinuation in those not
responding7. The British Thoracic Society also recommends steroid use in COPD patients failing to respond to bronchodilators8.
17.3.3 Methylxanthines
Although little evidence exists for the beneficial effects of theophylline over beta2 adrenergic agents, these agents are still considered secondline therapy for both acute
exacerbations of COPD and status asthmaticus4,5. In addition to having a bronchodilator effect, theophylline is an antiinflammatory agent, improves respiratory muscle
endurance and accelerates recovery from fatigue of respiratory muscles9,10.
17.3.4 Magnesium sulfate

The use of magnesium sulfate in acute asthma action is unknown. Except in patients with is controversial11,12. The actual mechanism of renal insufficiency, magnesium
sulfate is safe when a dose of less than 2 g is infused intravenously over at least 2 min. This therapy must be considered experimental in asthma and not useful in COPD.
17.3.5 Heliumoxygen mixtures (heliox)

In some severe asthmatics, the use of inhaled heliumoxygen mixtures during spontaneous breathing has lessened the severity of the asthmatic attack13. Because of its
lower density, helium decreases the pressure gradient required to move inhaled gas past an obstruction. In those patients where heliox is effective, respiratory rate
decreases, accessory muscle use diminishes, and intrathoracic pressure change is reduced, resulting in a decrease in pulsus paradoxus and work of breathing14.
Page 275
Table 17.3 Recommendations for classification and antibiotic treatment of an acute exacerbation of COPD
Group Clinical state Risk factors Probable pathogens First choice Alternatives
I Acute tracheobronchitis None Viral, rarely M. pneumoniae or C. No antibiotics Macrolide or tetracycline (for persistent
pneumoniae symptoms)
II Acute exacerbation of None H. influenzae, Haemophilus spp., M. Amoxicillin, tetracycline, Secondgeneration cephalosporin, second
chronic bronchitis catarrhalis, S. pneumoniae TMP–SMX generation macrolide, amoxicillinclavulanate,
fluoroquinolone
III Acute exacerbation of Multiple* Same as above, Also consider Gram Fluoroquinolone Amoxicillinclavulanate, oral secondor third
chronic bronchitis with risk negatives especially in patients with generation cephalosporin, or second
factors severely impaired lung function generation macrolide
IV Chronic suppurative airway Most have Same as group III plus multiresistant Antipseudomonal Consider parenteral therapy with
disease bronchiectasis Gramnegatives, particularly P. fluoroquinolone antipseudomonal agents
aeruginosa (ciprofloxacin)
*FEV1 <50% predicted, frequent exacerbations, significant comorbid conditions, malnutrition, chronic steroid use, mucus hypersecretion, duration of COPD >10 years, previous
pneumonia. TMP–SMX, Trimethoprimsulphamethoxazole. From: Sherk, P.A. and Grossman, R.E. (2000) The chronic obstructive pulmonary exacerbation. Clin. Chest Med. 21:705–721
Page 276
As a result, cardiovascular stress is also reduced.

Not all asthmatics respond to heliox. The reason for the varying response is not clear. However, based on pathophysiology, one can expect patients with primarily
small airways involvement and large amounts of retained secretions causing complete airway obstruction to be least likely to respond to heliox5. In contrast, patients
with primarily large airway obstruction, where gas density determines flow, are most likely to respond. As heliox has no effect on the specific pathology of asthma, its
effects are observed only during heliox use. Most responding patients demonstrate improvement within minutes when heliox is applied and demonstrate loss of effect
within minutes when heliox is removed.
Heliox is normally administered by partial rebreathing mask (Figure 17.2)15. The setup used allows the bronchodilator administered to also be powered by heliox if
desired. FIO2 is generally adjusted by a nasal cannula under the mask or by oxygen titrated into the heliox delivery system. As with all face masks, determination of the
precise FIO2 is impossible. Delivered oxygen is generally adjusted based on pulse oximetry. The higher the heliox concentration, the greater the likelihood for positive
response. Generally, helium concentrations between 60 and 80% are most effective. However, we have noted improvement with even lower helium concentrations.
The use of heliox should be continued until the patient's status has improved to the level that removal of the heliox does not cause an increase in cardiopulmonary stress.
No guidelines are established regarding the timing of heliox administration. However, we find it useful to consider heliox use in asthmatic patients who do not
demonstrate a positive response after three doses of aerosolized bronchodilator. Few data evaluating the use of heliox in COPD are available and at this time we do
not recommend its use in this setting16. Limited information on the use of heliox during mechanical ventilation of severe asthmatics is available17. Technically, few
ventilators are capable of properly operating with this lowdensity gas. Seimens Servo 900C and 300 ventilators are the two most common ventilators recommended
for use with heliox. As with spontaneous breathing, heliox delivers gas past an obstruction requiring a pressure gradient lower than with oxygennitrogen mixtures. In
volume ventilation, peak pressure tends to decrease whereas with pressure ventilation tidal volumes increase. This frequently allows the FIO2 to be decreased because
of better matching. As with spontaneous breathing, the higher the heliox concentration, the greater the likelihood for success and frequently once heliox is started,
FIO2 can be lowered. Even with FIO2 values of 0.6 or 0.7, we would consider the use of heliox in any mechanically ventilated asthmatic patient. One word of caution:
the ventilator display of delivered or exhaled tidal volume is inaccurately low because of the density of the heliox, although airway pressures are accurate. Multiplying
the displayed tidal volume by a factor of 1.4–1.8 (depending on helium concentration) provides an estimate of actual delivered VT18.
Other than the inability to correctly measure actual tidal volume, the cost of the gas mixture, and therapist’s time, there are no side effects or disadvantages of using
heliox. Generally, ‘H’ size tanks need to be changed every 4–6 h.
17.3.6 NIPPV
Published evidence19 and our clinical experience suggest that noninvasive positive pressure ventilation (NIPPV) is not useful in severe acute asthma. Because of the
level of cardiopulmonary distress in severe asthma, patients have a difficult time tolerating NIPPV. The opposite, however, is true in COPD patients. As discussed in
Chapter 19, NIPPV must be considered the standard of care for the management of COPD patients in an acute exacerbation20–22. This statement is based on data
from dozens of case series and at least eight randomized prospective clinical trials demonstrating in COPD patients that NIPPV
Page 277
Figure 17.2 Illustration of a heliox delivery system to spontaneously breathing patients, From: Hess, D. and Chatmongkolchart, S. (2000)
Techniques to avoid intubation: Noninvasive positive pressure ventilation and heliox therapy. Int Anesth. Clin. 38: 161–187.
results in a decrease in endotracheal intubation rate, length of ventilator support and ICU stay, nosocomial pneumonia rate, and mortality (see Chapter 19).
NIPPV is indicated in patients with acute (or acute on chronic) hypercapnic respiratory failure (PaCO2 >45 mmHg, pH <7.35)23. In the absence of severe gas
exchange abnormalities, moderate to severe respiratory distress characterized by respiratory rates >24 min–1, accessory muscle use and paradoxical breathing are also
considered indications for NIPPV23.
Successful application of NIPPV can usually be established within a few hours of initiation. NIPPV should eliminate cardiopulmonary distress and begin to restore
blood gas values to baseline. The success of NIPPV depends on a number of factors: the ventilator, the mask, and the personnel at the bedside. In acute respiratory
failure either an ICU ventilator or a bilevel pressure ventilator designed for ICU use should be employed. Careful monitoring of patientventilator synchrony with
appropriate alarms and ability to precisely titrate the FIO2 are essential.
The mask is also a crucial aspect of successfully applying NIPPV. In an acute exacerbation of COPD, we favor the use of a fullface mask rather than a nasal mask.
Ideally, a mask with exhalation ports as part of the mask is ideal. As recently shown, CO2 rebreathing can be minimized with masks where the exhalation port is
Page 278
located at the top (nose area) of the mask24. This allows for complete flushing of the mask at end exhalation.
Finally, the therapists and nurses caring for the patient must fully understand the principles of NIPPV and the appropriate approach to interacting with the patient.
With NIPPV, because sedation is minimized, time must be taken to carefully explain the procedure, gain the trust of the patient and slowly apply the technique. Failure
can occur simply as a result of the clinician’s approach. (See Chapter 19 for details.)
17.4 Intubation
Many COPD and asthmatic patients require intubation and mechanical ventilation. Although the decision to intubate is often lifesaving, intubation may be associated
with significant morbidity. Data from NIPPV trials clearly show that intubated COPD patients have a much higher risk of nosocomial pneumonia25 and a higher
mortality than nonintubated patients26. The decision to intubate either COPD or asthmatic patients should be made only after exhaustive efforts to manage these
patients’ respiratory failure more conservatively have failed.
17.5 Mechanical ventilation

Although both COPD and asthma are characterized as bronchospastic or obstructive disease, issues during mechanical ventilation are both similar and very different
(Table 17.4). With severe asthma, patients can be expected to return to normal baseline function during symptomfree periods. In contrast, COPD patients may be
chronically limited and symptomatic. Obstruction is fully reversible with asthma but only partially reversible in COPD. Obstruction in COPD is dynamic, mostly affecting
exhalation, whereas in asthma, obstruction is fixed, equally affecting inspiration and expiration. Ventilation is frequently long term in COPD, requiring lengthy weaning,
whereas in asthma, ventilation is commonly very short term, with rapid extubation. Similarities are primarily regarding air trapping and autoPEEP.
17.5.1 Mode of ventilation

As discussed in Chapter 13, two decisions must be made regarding which mode of ventilation should be used. First, should patients be managed with pressure or
volumetargeted ventilation (see Table 17.5)? Second, what specific mode of ventilation should be used? In
Table 17.4 Mechanical ventilation of asthma and COPD
COPD Asthma
Target Pressure Initially volume, pressure when able to assist
Mode PS, PA/C Volume A/C then PS or PA/C
PEEP Improves triggering (5–10 cm H2O) Not indicated
VT (6–10 ml kg–1) Initially may be ≤4 ml kg–1 up to 10 ml kg–1

P PLAT ≤30 cmH2O ≤30 cmH2O
I:E ≤1:3 ≤1:3
Inspiratory time 0.6–0.9 s 1.0–1.5 s
Rate Minimize autoPEEP Minimize autoPEEP
Flow wave If volume decelerating If volume decelerating
Peak flow ≥80 l min
–1
≥80 l min
–1
Page 279
COPD patients, patientventilator synchrony is a major issue, as patients are commonly allowed to continue to interface with the ventilator. As a result, we prefer
pressuretargeted ventilation, as gas delivery varies with patient demand and increases the ability of the ventilator to respond in synchrony with the patient27. Pressure
assist/control or pressure support are the modes of choice. If the patient has control over their ventilatory drive, pressure support is ideal. However, if the patient has
periodic breathing because of sedation or exhaustion, pressure assist/control is ideal. Volume assist/control can be used in these patients but it does not allow variability
in gas delivery (fixed flow pattern) and therefore increases the likelihood of dyssynchrony and enhanced patient work and effort to breathe.
On the other hand, with asthma volume ventilation, at least at the onset of ventilatory support is essential in most patients. With asthma, high driving pressures are
required to move gas into a lung with very high fixed airways resistance. Manually ventilating these patients is a major effort, occasionally requiring ‘peak’ pressures of
100 cmH2O to deliver tidal volumes of 200 ml. Because of the algorithm governing gas delivery, pressure ventilation cannot sustain the high driving pressures needed to
ventilate many asthmatic patients. It should be remembered that lung injury is primarily a result of high distending volumes and high peak alveolar pressures, not high
peak airway pressures28. In asthma, the driving pressures used to deliver small volumes never affect the distal lung. With volume ventilation as long as the end
inspiratory plateau pressure is less than 35 cmH2O (ideally <30 cmH2O), the likelihood of ventilatorinduced lung injury is limited28.
During pressure ventilation, the ventilator drives gas flow to rapidly achieve the pressure target. To avoid exceeding the target, flow must be rapidly decreased,
frequently in this setting to zero, thereby delivering small or no tidal volume. That is, with a pressure control setting of 30 cmH2O (zero PEEP), driving pressure is 30
cmH2O and peak alveolar pressure is limited to 30 cmH2O. However, gas volume may not be delivered. In volume ventilation at a specific VT, resulting in an end
inspiratory plateau pressure less than 30 cmH2O, the initial peak pressure may need to increase to 80 cmH2O, providing a driving pressure of 80 cmH2O (zero PEEP)
to deliver the desired VT, but maintaining a safe peak alveolar pressure.
To ventilate asthmatic patients, control ventilation is required. This may be achieved by sedation in some patients; however, many patients require neuromuscular
paralysis to achieve successful ventilation. If asthmatics are allowed to continue spontaneous ventilatory efforts, it is frequently impossible to ventilate them. In many
asthmatic patients, volume assist/control can be converted to pressure assist/control within a few hours of ventilatory support. However, this is usually unnecessary until
spontaneous ventilation resumes. Once the asthmatic patient starts to breathe spontaneously, every effort to insure patientventilator synchrony should be made,
Table 17.5 Volume versus pressure ventilation
Pressure Volume
Tidal volume Variable Constant
Peak inspiratory pressure Constant Variable
Peak alveolar pressure Constant Variable
Peak flow Variable Preset
Flow pattern Decelerating Constant
Inspiratory Preset or variable Preset
Page 280
and pressure assist/control or pressure support is indicated.
17.5.2 Tidal volume

In COPD patients, a moderate VT is frequently desired by the patient (generally between 6 and 10 ml kg−1). In patients with strong ventilatory demand but end
inspiratory plateau pressures less than 30 cmH2O, VT values up to 10 ml kg–l are not associated with ventilatorinduced lung injury6. However, avoidance of
overdistention and high peak alveolar pressure is always indicated. In asthma, VT may need to be very small initially (≤4 ml kg–1) because of high airway resistance
and air trapping. As airway resistance improves, VT values can be increased to meet the patient’s ventilatory demand. Once pressure support or pressure assist/control
is started (provided peak alveolar pressures are <30 cmH2O) VT values up to 10 ml kg–1 maybe necessary to insure patientventilatory synchrony5.
17.5.3 Peak flow, inspiratory time, I:E ratio

Peak flow settings are unnecessary in pressuretargeted ventilation modes but are critical in volumetargeted modes. In general, we recommend the use of a decelerating
flow pattern over a squarewave flow pattern in both COPD and asthma when volume ventilation is used. A decelerating flow pattern insures better distribution of
inspired gas and distributes the primary effect of airways resistance and compliance to opposite ends of the inspiratory phase, generally decreasing peak airway
pressure. At the onset of inspiration, with the flow rate very high but no volume delivered, impedance to gas delivery is primarily a result of airways resistance. In
contrast, at the end of the inspiratory phase when the flow rate is very low and tidal volume almost completely delivered, impedance to gas delivery is primarily a result
of compliance. A decelerating flow pattern produces a gas delivery pattern similar to pressure ventilation.
Especially in patients breathing spontaneously, initial peak flow should be high in volume ventilation29. In the setting of both asthma and COPD, peak flows should be
set at 80 l min–1 or higher. This insures that flow will better match the patient’s inspiratory demand. In pressure ventilation, peak flow automatically varies with demand
(generally exceeding 80 l min–1) if the pressure level is set greater than 10 cmH2O30.
Inspiratory time in spontaneously breathing patients should be set equal to the patient’s neuroinspiratory time. In most COPD patients this is about 0.6–0.9 s.
Inspiratory times longer than those desired by the patient tend to increase dyssynchrony. During controlled ventilation in COPD, inspiratory time can be set longer, but
considering the problems with air trapping they are generally set at 1.0 s or less.
In asthma, as the impact of airways resistance is fixed and equally affects inspiration and expiration, inspiratory time should not be less than 1.0 s; short inspiratory
times prevent gas delivery through the highly obstructed airway. The tradeoff is that because of air trapping, expiratory times should also be long. In general, each
asthmatic patient defines an experimental setting with inspiratory times adjusted between 1.0 and 1.5 s to insure adequate inspiratory volume and expiratory time
lengthened to the level that air trapping is minimized but ventilation (respiratory rate) is adequate5.
I:E ratio is generally considered a result of the above settings and not a target that is set. However, in both COPD and asthma I:E ratio tends to be small (1:3 to 1:6)
with short inspiratory and long expiratory times.
17.5.4 Respiratory rate and autoPEEP

Respiratory rate is generally set to insure an adequate PaCO2 level. In asthma it may not be possible to achieve normocapnia; permissive hypercapnia is generally the
rule early in the ventilatory course, with PaCO2 in the 70s and 80s common. The rate resulting in the best CO2 elimination with minimal air trapping and autoPEEP is
generally patient dependent. In many asthma patients, there is a specific expiratory time that minimizes autoPEEP;
Page 281
lengthening expiratory time beyond this level has minimal if any effect on autoPEEP until the expiratory time becomes unacceptably long. Frequently, patients can be
maintained with rates in the midtoupper teens if tidal volumes are small. It is overall minute ventilation that has the greatest impact on autoPEEP1. With asthma, care
should always be taken when ventilator settings are adjusted to insure that autoPEEP levels are minimized. Severe asthma is one of the clinical settings in which
increasing the level of ventilation can result in increasing CO2 levels because of decreased cardiac output and increased dead space ventilation as autoPEEP increases.
If an alteration in pressure target, VT, inspiratory time, or rate does not produce a decrease in CO2 it was the wrong adjustment and should be reversed. High CO2
levels may need to be tolerated for days in severe asthma.
With COPD patients, rates should be set to insure minimal autoPEEP and adequate CO2 removal during controlled ventilation. This goal is readily achieved in
managing patients with COPD. During spontaneous breathing, patients frequently select a rate in the mid20s. Insurance that peak flow and inspiratory time as well as
VT or pressure target are set appropriately will assist in avoiding dyssynchrony.
17.5.5 PEEP/FIO2
In COPD patients, adequate oxygenation can usually easily be managed by administering moderate oxygen levels. With asthma, some patients do require high FIO2
values (0.6–0.8) because of both increased shunting (mucous plugging) and mismatch.
The use of PEEP in these two clinical settings is controversial. In asthma we have rarely found PEEP to be beneficial. In some patients, it appears that lowlevel
PEEP improves matching but in most cases because of the fixed nature of the obstruction applied PEEP is simply additive to autoPEEP and detrimental.
COPD patients breathing spontaneously (assisted ventilation) frequently benefit greatly from the use of applied PEEP. Appropriately adjusting PEEP can markedly
improve patientventilator synchrony. AutoPEEP in COPD is the result of a variable dynamic flow resistor. The pressure difference at end exhalation across the
resistor is great; autoPEEP on one side and endexpiratory ventilator circuit pressure on the other (Figure 17.3)31. In this setting, applied PEEP less than about 80%
of the autoPEEP level does not increase overall total PEEP, but does reduce the gradient to trigger the ventilator, thereby avoiding missed triggered breaths32. It is
difficult to measure autoPEEP during spontaneous breathing, but its presence can easily be identified by observing the expiratory flow waveform (Figure 17.4)33. In
addition, autoPEEP is normally the cause of the patient’s inspiratory efforts failing to trigger an assisted breath. Whenever the patient’s respiratory rate exceeds the
ventilator response rate, the patient is having difficulty overcoming the autoPEEP. In this setting slowly increasing PEEP in 1–2 cmH2O increments and observing the
patients ventilatory pattern can dramatically improve synchrony34. When every patient effort triggers an assisted breath, the appropriate PEEP level is set. In all COPD
patients, 5 cmH2O applied PEEP is helpful. However, in many patients, levels as high as 12–15 cmH2O may be necessary because of high autoPEEP levels. During
controlled ventilation there are no data to suggest that the application of PEEP has a beneficial effect. However, PEEP does not appear to be harmful provided its
application does not increase total PEEP.
17.6 Weaning from ventilatory support

Details on approaches to weaning are presented in Chapter 18, but special considerations in patients with asthma and COPD deserve mention here. With asthma,
patients rarely require prolonged weaning. Because they normally have good ventilatory mechanics once the acute asthma is reversed, they tolerate rapid extubation
well. Asthmatic patients who are not heavily sedated rarely tolerate an endotracheal tube without developing bronchospasm. As a result, as soon as airways
Page 282
Figure 17.3 Measurement of autoPEEP by expiratory port occlusion. Normally (see top panel), alveolar pressure is atmospheric at the
end of passive exhalation. With severe airflow obstruction (see middle panel) alveolar pressure remains elevated (in this
example at 15 cmH2O) and slow flow continues even at the end of the set exhalation period. The ventilator manometer senses
nagligible pressure because it is open to atmosphere through largebore tubing and downstream from the site of flow limitation.
With gas flow stopped by occlusion of the expiratory port at the end of the set exhalation period (see lower panel), pressure
equilibrates throughout the lungventilator system and is displayed on the ventilatory manomete From: Pepe, P.E. and Marini, J.
J. (1982) Occult positive endexpiratory pressure in mechanically ventilated patients with airflow obstruction; the autoPEEP
effect. Am. Rev. Respir. Dis. 126: 166–170.
resistance returns to nearnormal levels, and ventilation and oxygenation can be maintained at low ventilator settings (FIO2 ≤0.4, pressure support or control levels of
10–15 cmH2O and VT values in the 6–10 ml kg–1 range), extubation should be considered5.
In contrast, COPD patients frequently require a very long weaning period because of chronic airway obstruction, other organ system dysfunction, and a general state
of debilitation35. These patients require a consistent, standardized weaning program that incorporates general physical rehabilitation. Some do wean rapidly but most
require days to weeks to be successfully liberated from ventilatory support. The use of either a pressure support or spontaneous breathing trial approach to weaning is
appropriate. Care should be taken to insure that patients are well rested (particularly at night), well nourished, and that their ventilatory capabilities are challenged daily
but are not pushed to the level of fatigue.
Page 283
Figure 17.4 Gas flow, airway pressure, and volume waveforms during volumetargeted control ventilation in the presence of auto
PEEP. Expiratory gas flow Shows an initial spike that rapidly decreases to a low level throughout the prolonged expiratory
time. It should be noted that expiratory gas flow does not return to zero (arrow) indicating the presence of autoPEEP. from:
Kacmarek, R.M. and Hess, D.R. (1993) Airway pressure, flow and volume waveforms, and lung mechanics during mechanical
ventilation, In: Monitoring in Respiratory Care (eds R.M.Kacmarek, D.R.Hess and J.K.Stoller). Mosby Yearbook, St Louis,
MO, pp. 497–543.
References
1. Pepe, P.E. and Marini, J.J. (1982) Occult positive endexpiratory pressure in mechanically ventilated patients with airflow obstruction: The autoPEEP effect. Am.
Rev. Respir. Dis. 126: 166–172.
2. Petrof, B.J., Legare, M., Goldberg, P., et al. (1990) Continuous positive airway pressure reduces work of breathing and dyspnea during weaning from mechanical
ventilation in severe chronic obstructive pulmonary disease. Am. Rev. Respir. Dis. 141:281–289.
3. Leatherman, J.W. and Ravenscraft, S.A. (1996) Low measured autopositive endexpiratory pressure during mechanical ventilation of patients with severe asthma:
Hidden autopositive endexpiratory pressure. Crit. Care Med. 24:541–549.
4. Ferguson, G.T. (2000) Update on pharmacologic therapy for chronic obstructive pulmonary disease. Clin. Chest Med. 21:723–738.
5. Levy, B.D., Kitch, B. and Fanta, C.H. (1998) Medical and ventilatory management of status asthmaticus. Intens. Care Med. 24:105–117.
6. Sherk, P.A. and Grossman, R.E. (2000) The chronic obstructive pulmonary exacerbation. Clin. Chest Med. 21:705–721.
7. American Thoracic Society (1995) Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease. Am. J. Respir. Crit. Care Med.
152:584–596.
Page 284
8. British Thoracic Society (1997) Guidelines for the management of chronic obstructive pulmonary disease. Thorax 52:S1S28.
9. Weinberger, M. and Hendeles, L. (1996) Theophylline in asthma. N. Engl J. Med. 334: 1380–1388.
10. Ferguson, G.T. and Cherniack, R.M. (1993) Management of chronic obstructive pulmonary disease. N. Engl J. Med. 328:1017–1022.
11. Skobeloff, E.M., Spivery, W.H., McNamara, R.M. and Greenspan, L. (1989) Intravenous magnesium sulfate for the treatment of acute asthma in the emergency
department. JAMA 262:1210–1213.
12. Green, S.M. and Rothrock, S.G. (1992) Intravenous magnesium for acute asthma: Failure to decrease emergency treatment duration or need for hospitalization.
Ann. Emerg. Med. 21: 260–265.
13. Kass, J.E. and Castriotta, R.J. (1995) Heliox therapy in acute severe asthma. Chest 107: 757–760.
14. Manthous, C.A., Hall, J.B., Meled, A., Caputo, M.A., Walter, J., Klocksieben, J.M., Schmidt, G.A. and Wood, L.D.H. (1995) Heliox improves pulsus
paradoxus and peak expiratory flow in non intubated patients with severe asthma. Am. J. Respir. Crit. Care Med. 152:310–317.
15. Hess, D. and Chatmongkolchart, S. (2000) Techniques to avoid intubation: Noninvasive positive pressure ventilation and heliox therapy. Int Anesth. Clin. 38:161–
187.
16. Jaber, S., Fodil, R., Carlucci, A., et al. (2000) Noninvasive ventilation with heliumoxygen in acute exacerbations of chronic obstructive pulmonary disease. Am. J.
17. Gluck, E.H., Onorato, D.J. and Castriotta, R. (1990) Heliumoxygen mixtures in intubated patients with status asthmaticus and respiratory acidosis. Chest 98:693–
698.
18. Tassaux, D., Jolliet, P., Thouret, J.M., Roeseler, J., Dorne, R. and Chevrolet, J.C. (1999) Calibration of seven ICU ventilators for mechanical ventilation with
heliumoxygen mixtures. Am. j. Respir. Crit. Care Med. 160:22–32.
19. Keenen, S.P., Kernerman, P.D., Cook, D.J., Martin, C.M., McCormack, D. and Sibbald, W.J. (1997) Effect of noninvasive positive pressure ventilation on
mortality in patients admitted with acute respiratory failure: A metaanalysis. Crit. Care Med. 25:1685–1691.
20. Kramer, N., Meyer, T.J., Meharg, J., Cece, R.D. and Hill, N.S. (1998) Randomized, prospective trial of noninvasive positive pressure ventilation in acute
respiratory failure. Am. J. Respir. Crit. Care Med. 151:1799–1805.
21. Brochard, L., Mancebo, J., Wysocki, M., et al. (1995) Noninvasive ventilation for acute exacerbations of chronic obstructive pulmonary disease. N. Engl. J. Med.
338:817–822.
22. Plant, P.K., Owen, J.L. and Elliott, M.W. (2000) Early use of noninvasive ventilation for acute exacerbation of chronic obstructive pulmonary disease on general
respiratory wards: A multicenter randomized controlled trial. Lancet 355:1931–1936.
23. Wysocki, M., Tric, L., Wolff, M.A., Millet, H. and Herman, B. (1995) Noninvasive pressure support ventilation in patients with acute respiratory failure: A
randomized comparison with conventional therapy. Chest 107:761–766.
24. Schettino, G.P.P., Chatmongkolchart, S., Hess, D. and Kacmarek, R.M. (2001) CO2 elimination during BIPAP ventilation: Effects of mask design and PEEP
(abstract). Am. Rev. Respir. Crit. Care Med. 163:A679.
25. Girou, E., Schortgen, F., Delclaux, C., BrunBuisson, C., Blot, F., Lefort, Y., Lemaire, F. and Brochard, L. (2000) Association of noninvasive ventilation with
nosocomial infections and survival in critically ill patients. JAMA 284: 2361–2366.
26. Guerin, C., Girard, R., Chemorin, C., DeVarax, R. and Fournier, G. (1997) Facial mask noninvasive mechanical ventilation reduces the incidence of nosocomial
pneumonia: A prospective epidemiological survey from a single ICU. Intens. Care Med. 23: 1024–1028.
27. Cinnella, G., Conti, G., Lofaso, F., Lorino, H., Harf, A., Lemaire, F. and Brochard, L. (1996) Effects of assisted ventilation on the work of breathing: Volume
controlled versus pressurecontrolled ventilation. Am. J. Respir. Crit. Care Med. 153:1025–1033.
28. Dreyfuss, D. and Saumon, G. (1993) Role of tidal volume, FRC, and endinspiratory volume in the development of pulmonary edema following mechanical ventilation.
Am. Rev. Respir. Dis. 148:1194–1203.
29. Marini, J.J., Rodriquez, M. and Lamb, V. (1986) The inspiratory workload of patientinitiated mechanical ventilation. Am. Rev. Resp. Dis. 134:902–906.
30. Williams, P., Muelver, M., Kratohvil, J., Ritz, R., Hess, D.R. and Kacmarek, R.M. (2000) Pressure support and pressure assist/control: Are there differences?
An evaluation of the newest intensive care unit ventilators. Respir. Care 45: 1169–1181.
31. Pepe, P.E. and Marini, J.J. (1982) Occult positive endexpiratory pressure in mechanically ventilated patients with airflow obstruction: the
Page 285
autoPEEP effect. Am. Rev. Respir. Dis. 126: 166–170.

32. Smith, T.C. and Marini, J.J. (1988) Impact of PEEP on lung mechanics and work of breathing in severe airflow obstruction. J. Appl. Phys. 64: 1488.
33. Kacmarek, R.M. and Hess, D.R. (1993) Airway pressure, flow and volume waveforms, and lung mechanics during mechanical ventilation. In: Monitoring in
Respiratory Care (eds R.M.Kacmarek, D.R.Hess and J.K.Stoller). Mosby Yearbook, St Louis, MO, pp. 497–543.
34. Petrof, B.J., Legare, M., Goldberg, P., et al. (1990) Continuous positive airway pressure reduces work of breathing end dyspnea during weaning from mechanical
ventilation in severe COPD. Am. Rev. Respir. Dis. 141: 281–289.
35. Vitacca, M., Vianello, A., Colombo, D., et al. (2001) Comparison of two methods for weaning patients with chronic obstructive pulmonary disease requiring
mechanical ventilation for more than 15 days. Am. J. Respir. Crit. Care Med. 164:225–230.
Page 286

Page 287
Chapter 18
Weaning from mechanical ventilation
James K.Stoller, MD and
Contents
18.1 Introduction
18.2 Who requires gradual withdrawal of mechanical ventilation?
18.3 Sequelae of prolonged intubation and ventilation
18.4 Overview of the logic of weaning
18.5 Evaluation of breathing pattern
18.6 Work of breathing (WOB)
18.7 Tracheal occlusion pressure
18.8 Endurance measurements
18.9 Techniques of weaning
18.10 Studies comparing weaning modes
18.11 Nonrespiratory impediments to successful weaning
18.12 Respiratory impediments to weaning
18.13 Pharmacologic aids to weaning
18.14 Weaning failure
18.15 Longterm prognosis
Page 288
18.1 Introduction
Weaning, the process of freeing the patient from ventilatory dependence, is a common challenge in the ICU and subacute respiratory care units. Weaning encompasses
both art and science because measured respiratory parameters alone do not always accurately predict a patient’s readiness to breathe without ventilatory assistance.
The term ‘weaning’ must also be distinguished from ‘extubation’ or removal of the endotracheal tube. For example, patients who cannot protect their upper airway
(e.g. in coma) or patients with upper airway obstruction (e.g. with edema or glottic dysfunction) may require placement of an endotracheal tube even though they can
breathe spontaneously. Incomplete reversal of neuromuscular blockade can also produce a situation where ventilation is adequate, but airway protection is impaired1.
This chapter will consider the indications for weaning and extubation, the rationale and clinical utility of weaning predictors, weaning techniques in general and in specific
patient subsets, and specific impediments to weaning.
18.2 Who requires gradual withdrawal of mechanical ventilation?

Not all patients require gradual withdrawal of mechanical respiratory support. The majority of patients undergoing surgical procedures can be extubated immediately
following surgery, and need to be ventilated only until they have sufficient neurologic function and muscular strength to breathe on their own. Thus, the first important
concept in weaning is that the underlying process causing respiratory failure must be improving before weaning can begin. Patients who have undergone uncomplicated
operations generally require ventilation only until the effects of anesthetic and neuromuscular blocking drugs have cleared. Even patients undergoing thoracic or upper
abdominal operations (those are associated with the greatest impairment of ventilatory function) are usually liberated from mechanical ventilation once acute drug effects
resolve, as long as there is no associated cardiovascular instability. Finally, patients for whom intubation is performed for airway protection rather than for ventilatory
support can be promptly extubated once the reason for airway compromise resolves.
A subset of complex postoperative patients require ventilatory support for several hours to several days following their surgical procedure. Common reasons for
prolonged postoperative ventilatory support include low cardiac output states (where the work of breathing from spontaneous ventilation will impair already tenuous
systemic perfusion), unstable hemodynamics, ongoing bleeding and/or the possibility of reexploration, intravascular volume overload, or a requirement for continued
sedation and neuromuscular blockade (as in the neurosurgical patient with elevated intracranial pressure). Although past practice dictated that cardiac surgical patients
were ventilated overnight, newer cardiac anesthetic techniques have made it possible to ‘fasttrack’ most patients2. Criteria for removing the endotracheal tube in
postoperative cardiac patients include hemodynamic stability, normalization of acidbase and electrolyte status, restoration of body temperature to normal, and
adequate muscle strength and mental status3.
Individuals whose respiratory function has been compromised by a serious, protracted insult often require prolonged ventilation and more gradual withdrawal of
ventilatory support. Common examples of such conditions include sepsis, the acute respiratory distress syndrome (ARDS), COPD complicated by an acute insult such
as pneumonia, and neuromuscular disorders compromising ventilatory function. Such patients require prolonged ICU stays, account for a large proportion of ICU days
and attention, and will be the subject of most of the ensuing discussion. Importantly, approaches that have been shown to improve outcomes in such patients have
included use of standardized weaning approaches4,5, protocols that incorporate input from allied health care providers, and teams dedicated to weaning6.
Page 289
18.3 Sequelae of prolonged intubation and ventilation

Intubation and prolonged mechanical ventilation are associated with morbidity, including vocal granulomas and ulceration of the true cords, in more than half of patients
intubated for more than 24 h7. Early tracheal lesions may progress to a circumferential fibrous stenosis that is difficult to later repair. Roughly 20% of patients who
survive their ICU stay, and 95% of those dying while intubated, will have significant tracheal lesions8. Functional concerns also favor removing the endotracheal tube as
early as possible to avert epithelial damage, loss of cilia, and impairment of tracheal mucous clearance9. Endotracheal intubation increases the risk of nosocomial
pneumonia and often precludes oral feeding, mandating the use of enteral or parenteral nutrition. Iatrogenic malnutrition is thus common in the ventilated patient10.
Finally, mechanical ventilation elicits an inflammatory response that can be attenuated by minimizing overdistention of the lung11.
18.4 Overview of the logic of weaning

In approaching the patient on mechanical ventilation for whom weaning is being considered, the clinician must ask:
• Is the patient a candidate to undergo a weaning trial? And, if so,
• Is weaning likely to be successful?
Each question invites several clinical considerations and/or measurements.
18.4.1 Is the patient a candidate to undergo a weaning trial?

For a patient to be considered candidate for a weaning trial, criteria listed in Table 18.1 must be met. Weaning will not succeed unless the process initially responsible
for ventilator dependence has shown improvement. Remediable aspects that can contribute to ongoing ventilator dependence (e.g. bronchospasm, hypermetabolic
state, poor nutrition, cardiac failure) should be optimized.
18.4.2 Is weaning likely to be successful?

Once the patient is deemed a candidate to wean and a weaning trial is planned, predicting the likelihood of weaning success is useful. Weaning prediction is the process
of estimating the likelihood that weaning and/or extubation efforts will succeed or fail in a specific patient at a specific time12. Although no single respiratory parameter
or scoring system performs flawlessly, measuring parameters is helpful as a training device, reduces the dependence on the skills of the individual clinician and helps to
guide management.
Many parameters to enhance weaning prediction have been proposed and include both univariate (single variable) measures and multivariate indices whose
complexity varies. Multivariate predictors generally utilize measurements of:
• static lung function (e.g. forced vital capacity (FVC));
• ventilatory muscle strength and/or endurance (e.g. maximal inspiratory force,
• gas exchange (e.g. alveolararterial gradient maximal voluntary ventilation); or PaO2/FIO2 ratio);
• systemic perfusion (e.g. gastric intramural pH).
The predictive accuracy of most univariate weaning parameters (Table 18.2) is low,
Table 18.1 Questions to ask in assessing the patient’s candidacy to undertake weaning
• Is the process responsible for ventilator dependence improving?

• Is oxygenation adequate?
• Is the patient hemodynamically stable, e.g. free from a requirement for pressors and inotropes?
• Have remediable factors been optimized, e.g. maximum treatment of bronchospasm of metabolic abnormalities that can impair weaning (e.g. hypophosphatemia,
hypokalemia), adequate resolution of effects of all sedating and paralytic agents, adequate nutrition?
Page 290
especially regarding ability to predict weaning failure (i.e. negative predictive value). As a result of this low negative predictive value, slavish adherence to univariate
weaning parameters in initiating weaning trials can delay weaning that is actually possible13. As shown in Table 18.3, some multivariate weaning indices are better
predictors of weaning success or failure.
In the sections that follow, we consider several of the multivariate weaning parameters, particularly those regarding the pattern of breathing, work of breathing,
measurements of tracheal occlusion pressure, and endurance.
18.5 Evaluation of breathing pattern

Assessing the patient’s pattern of breathing provides important insight into the likelihood of weaning outcome, because rapid shallow breathing and paradoxic inward
inspiratory motion of the abdominal muscles (socalled ‘abdominal paradox’) accompany ventilatory failure. Unlike measurements of the maximal inspiratory pressure
(MIP), vital capacity (VC), and maximal voluntary ventilation (MVV), the breathing pattern is not highly dependent on patient cooperation.
Table 18.2 Diagnostic weaning parameters in predicting weaning
Minute ventilation (VE) <10 L min–1

Study n Sensitivity (%) Specificity (%) PPV NPV
Sahn & Lakshminarayan (1973) 86 100 92 100 100 71
Tahvanainen et al. (1983)51 47 45 78 89 25
Krieger et al. (1989) 13† 269 NS NS 93 15
Yang & Tobin (1989) 87 41 24 69 55 37

–1
*VE <10 l min , MIP <−30 cmH2 O, and MVV >2 VE.
–1
†VE <10 l min , MIP <−30 cmH2O, and MVV >2 VE.
PPV, positive predictive value; NPV, negative predictive value; NS, not stated.
Vital capacity (VC)
Study Criterion n Sensitivity (%) Specificity (%) PPV NPV
Milbern et al. (1978) 88
VC >15 ml kg –1 33 25 0 58 0
Tahvanainen et al. (1983)51 VC>10 ml kg–1 47 97 13 83 50
Pardee et al. (1984) 89

VC >17 ml kg –1 133 90 60 88 NS
*VC >15 ml kg–1 and MIP <–25 cmH2 0. NS, not stated.
Maximal inspiratory pressure (MIP) ≤ 30 cmH2 O

Study (date) n Patient type Sensitivity (%) Specificity (%) PPV NPV
Sahn & Lakshminarayan (1973) 86 100 Mean MV duration 37 h 92 100 100 71
Milbern et al. (1978)88† 33 Mean MV 3.1 h 25 0 58 0
Tahvanainen et al. (1983)51 47 Mean MV 5 days 68 0 74 0
DeHaven et al. (1986) 90 48 Mean MV 55 h 49 100 100 12
Krieger et al. (1989) 13 269 Mean age >70 years MV 71 h NS NS 92 21
Yang & Tobin (1989)87 41 NS 76 25 61 40

–1
*MIP <–30 cmH2 O, VE <10 l min , and MVV ≥2 VE.
1.
†MIP <–30 cmH2 O, and VC ≥15 ml kg–
PPV, positive predictive value; NPV, negative predictive value; MV, mechanical ventilator; NS, not stated. From: Stoller, J.K. (1991) Establishing clinical unweanability. Reproduced
with permission from Respir. Care 36:186–198.
Page 291
Table 18.3 Diagnostic performance of selected multivariate indices for weaning prediction
Study n Index Patient type Positive predictive (%) Negative predictive (%)
Hilberman et al. (1976) 76 124 Nurses’ assessments OHS 82 67
Krieger et al. (1984)12 269 NIF <–30 cmH2O, VE >10 l >70 yr, on MV mean 71 h 93 15
–1
min
Yang & Tobin (1991) 13 100 CROP>13 Freq/VT ≥105 On MV 8.2±1.1 days 71 70
Sheinhorn et al. (1995)14 565 P(Aa)O2 BUN, gender On MV >6 weeks 71 67
Epstein (1995)15 184 Freq/VT >100 MICU on MV, f/VT measured within 8 h of 83 40
wean onset
Gluck et al. (1995)22 55 Score (5 variables) Points >3 On MV ≥3weeks 58 100
NIF, negative inspiratory force; MV, mechanical ventilator; VE, minute ventilation; CROP, compliance rate oxygenation pressure; OHS, openheart surgery; NS, not stated; BUN, blood
urea nitrogen; MICU, medical intensive care unit; VT , tidal volume. Hypothesisgenerating study, not confirmed in a separate dataset; hypothesistesting dataset included. From: Stoller, J.
91
(2000). Reproduced with permission from SCCM/ACCP Combined Critical Care Course .
18.5.1 The rapid shallow breathing index

This index, commonly known as the frequency to tidal volume ratio (f/VT), calculates the patient’s spontaneous breathing frequency (breaths per minute) divided by the
patient’s spontaneous tidal volume (in liters) while off mechanical ventilatory assistance. Yang and Tobin proposed that an f/VT value less than 105 would discriminate
between patients who were successfully weaned (defined as maintaining spontaneous breathing for greater than 24 h after extubation) versus those who required
ongoing ventilatory support14. This index was developed on 36 patients and subsequently validated in 64 different patients. Overall, an f/VT greater than 105 carried a
negative predictive value of 0.95, meaning 95% of patients with a high f/VT could not be extubated. In contrast, values of f/VT less than 105 predicted a 78% chance
of weaning success. In patients who had received mechanical ventilation for more than 8 days, the negative and positive predictive values were slightly lower at 0.89
and 0.64 respectively, emphasizing that predicting unweanability in longterm mechanically ventilated patients is more difficult15. Subsequent studies16,17 have confirmed
the usefulness of the rapid shallow breathing index, but have also pointed out its shortcomings when nonrespiratory factors such as congestive heart failure or upper
airway obstruction are present. The diagnostic accuracy of different weaning parameters will vary depending on the underlying cause of respiratory failure. For
example, the rapid shallow breathing index has its highest diagnostic accuracy in predicting weaning success in patients with COPD, but does not perform as well in
patients with neurologic disease or other causes of acute respiratory failure18.
18.5.2 The tensiontime index of the diaphragm (Ttdi)

This is derived from the percentage of maximal transdiaphragmatic pressure (Pdi) expended with each breath (Pdi/Pdimax) and the fraction of the breathing cycle time
spent in inspiration
Page 292
(Ti/TTOT)19. Values of Ttdi below 0.15 predict the ability to sustain unassisted respiration for longer than 45 min. Values of Ttdi above 0.15 predict eventual ventilatory
failure, possibly because of limitation of diaphragmatic blood flow (Figure 18.1). Although transdiaphragmatic pressure and inspiratory time are not routinely measured,
devices such as the Bicore can provide surrogate measurements that have been advocated by some in enhancing weaning20.
18.6 Work of breathing (WOB)

Work of breathing is the work expended by the ventilatory muscles in supporting breathing
Figure 18.1 Tensiontime index for t the human diaphragm (TTdi). The inspiratory time total ventilatory cycle time (TI/TTOT) ratio is
plotted against transdiaphragmatic pressure during normal breathing/maximum possible (Pdi/Pdimax) ratio. The fatigue threshold
represents a TTdi that can be sustained for 1 hour r or longer and has a value of 0;1 5 to 0.18 in the normal human diaphragm.
The TTdi threshold can be achieved with a large variety of Pdi and Ti/TTOT values. Patterns in the fatigue zone result in
fatigue in less than 1 hour, whereas patterns in the nonfatigue zone can be sustained indef finitely. The circle r represents the
TTdi during resting breathing in nomal subjects. From: Grassino A, Bellemare F, and Laporta D. (1984) Diaphragm fatigue
and the strategy of breathing in COPD. Reproduced with pemission from Chest. 85: 515–535.
and is measured by integrating the area under a pressurevolume inspiratory curve21. Elevated WOB predicts inspiratory muscle fatigue with subsequent weaning
failure. Work per minute, defined as the amount of work done per breath multiplied by the respiratory rate, reflects ventilatory requirements. Work per liter is the work
per minute divided by VE and measures efficiency of breathing. In patients requiring prolonged ventilation, work of breathing is a better indicator of successful weaning
than standard bedside weaning criteria (i.e. vital capacity, maximum inspiratory pressure, tidal volume, and minute ventilation)22. Specifically, as WOB values fall below
1.60 kgm min–1 (16 joules min–1), WOB better discriminates between patients with weaning success versus failure than more conventional measures.
In a study of 23 patients comparing a weaning protocol incorporating WOB and f/VT versus a clinical approach using conventional weaning criteria (e.g. minute
ventilation, negative inspiratory force, tidal volume, and static compliance), the protocol incorporating WOB measurements hastened weaning in at least 41% of
instances and shortened the projected mean duration of weaning by 1.68 days23.
18.6.1 The oxygen cost of breathing

This is a related measure assessed by comparing oxygen (O2) consumption during complete inspiratory muscle rest (i.e. during paralysis with controlled mechanical
ventilation) with O2 consumption during spontaneous ventilation24. In normal individuals, oxygen consumption by the respiratory muscles accounts for approximately 2–
5% of total oxygen consumption. In patients who became ventilator dependent, the oxygen cost of breathing was frequently 15% or more of total O2 consumption.
Using a threshold value of 15% as a predictor, the sensitivity of O2 cost of breathing measurements was 100% and specificity 80%25. Despite the physiologic appeal of
O2 cost of breathing measurements in guiding weaning, impediments to clinical use include the needs for metabolic cart measurements and to conduct measurements
during complete paralysis.
Page 293
18.7 Tracheal occlusion pressure

Respiratory center drive is measured by the tracheal occlusion pressure (P0.1), the airway pressure generated during the first tenth of a second during an occluded
inspiratory effort. Using a threshold value of 6 cmH2O, the P0.1 provides a simple, valid way to assess the likelihood of respiratory muscle fatigue. In one series, P0.1
was followed in patients beginning with the first day of acute respiratory failure and was found to decrease before successful extubation26. In contrast, in patients in
whom P0.1 did not change significantly from the initial value, diaphragmatic fatigue eventually developed, and reintubation was necessary within 2–6 days.
18.8 Endurance measurements

Fatigue (lack of ventilatory muscle endurance) is the usual reason for weaning failure. Short of allowing fatigue to develop, it is difficult to predict an individual patient’s
endurance. Experimentally, the power spectrum of the diaphragmatic electromyogram (EMG) has been used to evaluate endurance27. In 12 patients, seven developed
EMG evidence of fatigue (i.e. appearance of EMG power spectral shift); in six of these seven, this shift was associated with an abnormal breathing pattern
characterized by paradoxical abdominal motion with inspiration. Although EMG is not practical for routine clinical evaluation, it has nicely characterized the clinical
sequence of respiratory muscle fatigue (Table 18.4) and suggests that the onset of chest wall and abdominal paradox is an early indicator, followed by hypercapnea as
a late finding. Notably, later studies suggest that inspiratory paradox indicates respiratory muscle loading and does not distinguish between successful and unsuccessful
weaning28. However, measurement of the maximal amplitude of diaphragm excursion divided by VT, and the relationship between rib cage excursion and VT have
been found to be nonspecific but sensitive indicators of respiratory fatigue.
Table 18.4 Clinical sequence of inspiratory muscle fatigue
• Shift of EMG power spectrum

• Tachypnea
• Respiratory alternans
• Paradoxical abdominal motion
• Fall in respiratory rate and minute ventilation
• Elevation of PaCO2
• Arterial acidemia
EMG, diaphragmatic and intercostal electromyogram. Adapted from Cohen et al. 27 (1982) Clinical manifestations of inspiratory muscle fatigue. Am. J. Med. 73:308–
316.
18.9 Techniques of weaning

A variety of techniques for weaning patients from mechanical ventilation have been described, including Tpiece trials of increasing duration alternating with periods of
completely supported breaths, progressive decrease in the intermittent mandatory ventilation (IMV) rate, use of declining levels of pressure support, and combinations
of the above (e.g. pressure support weaning with an IMV backup rate). Strategies in use have changed as new ventilatory modes have became available on newer
ventilators. In 1987, before the widespread availability of pressure support ventilation (PSV), IMV was the primary mode of weaning employed by 72% of
respondents to a large survey29. More recently, a survey of practices in 47 Spanish ICUs30 showed that in 195 patients undergoing weaning, Tpiece trials of
increasing duration were used most commonly (24%), followed by synchronized IMV weaning (18%), pressure support weaning (15%), and combined pressure
support with an IMV backup rate (9%). Other combinations applied concurrently or in succession were used in.33%. Each of the weaning techniques is described
below, followed by a summary of the evidence supporting specific techniques of weaning. Figure 18.2 offers an approach to weaning.
18.9.1 Tpiece trials

Tpiece trials (also known as Briggs trials) allow the patient to breathe spontaneously through the endotracheal tube or tracheostomy connected to a Tpiece setup.
CPAP can be applied
Page 294
Figure 18.2 Approach to weaning.
to the Tpiece circuit, although CPAP pressure levels much above 5 cmH2O would be unlikely during Tpiece trials, as adequate oxygenation at acceptably low levels
of endexpiratory pressure is a criterion for beginning weaning. Tpiece trials are laborintensive, because the patient must be monitored continuously for distress or
fatigue. Beginning intervals may be as brief as 2 min each hour and progressively lengthened by 2 min h–1 or more as endurance develops. A reasonable starting point is
5 min of Tpiece followed by 60 min of rest. If this first trial is successful, the trial can be increased in 5 min intervals, allowing at least 1 h of rest between attempts.
Signs of Tpiece weaning failure include a respiratory rate above 35–40, heart rate above 110, and increased blood pressure, diaphoresis, air hunger, and signs of
Page 295
respiratory muscle fatigue. Elevated endtidal CO2 is a late sign of decompensation6. Optimally, weaning should be stopped just before signs of distress to avoid
inducing 'weaning anxiety'. Between weaning periods, full support should be provided with use of assistcontrol at a rate sufficient to blunt the patient's own respiratory
drive.
18.9.2 Intermittent mandatory ventilation weaning

This was first proposed in 1973 as a new and preferred weaning strategy31. With IMV weaning, a tidal volume and breathing frequency is set so the ventilator does
most, but not all of the work initially, and the intermittent mandatory frequency is then gradually decreased until the patient has assumed most of the minute ventilation.
Early studies suggested that IMV was preferable to controlled mechanical ventilation, resulting in less alkalosis, less time on the ventilator, and less increase in O2
consumption. However, claims that IMV was a preferred weaning mode because it could shorten weaning time have not been supported by later studies. More recent
controlled trials comparing time to weaning using this mode versus progressive Tpiece or pressure support weaning have shown that IMV is associated with the
longest duration of ventilation and is therefore the least attractive initial weaning mode32. In our view, IMV weaning remains an option for the versatile clinician, but
usually after other modes have proven unsuccessful (Figure 18.2).
18.9.3 Pressure support ventilation

Unlike IMV, which is a volumecycled mode of mechanical ventilation, pressure support ventilation33 is flowcycled and delivers gas at a set pressure level for a
duration determined by the patient's inspiratory flow demands. Pressure support weaning involves the gradual diminution of the inspiratory pressure level, allowing the
patient gradually to assume more of the work of breathing (WOB). In contrast to IMV, in which supported breaths are interspersed with spontaneous ventilation, each
PSV breath is partly supported. Advantages of PSV include increased patient comfort, ability to offset increased work of breathing imposed by the ventilator or the
endotracheal tube, and increased efficiency for the clinician overseeing its use34,35. The initial pressure level can be set at approximately half the peak fully supported
inspiratory pressure, and PEEP is usually set at 5–10 cmH2O. The level of pressure support may then be adjusted to maintain adequate tidal volume with a respiratory
rate between 12 and 24. This allows the patient to perform some respiratory muscle training while minimizing tachypnea and its associated stress. Weaning using
inspiratory pressure support can usually be continued for 1 h or more; the patient is placed back on full support if tachypnea or clinical signs of distress develop.
Patients will often tolerate PSV for much of the day and require a return to full support only overnight. The pressuretime product, work of breathing, and f/Vt all
decrease in a stepwise fashion as PSV is increased, so f/Vt can perhaps be used as an indicator of patient WOB during PSV36.
18.10 Studies comparing weaning modes
Three recent controlled trials have compared available weaning strategies. Brochard et al.37 randomized 109 patients who failed initial weaning attempts to one of
three weaning strategies: (i) Tpiece trials, with progressively increasing intervals of spontaneous breathing until the patient tolerated up to three Tpiece trials lasting 120
min (n=35 patients); (ii) synchronized IMV weaning, in which the IMV rate was decreased by 2–4 breaths min–1 twice daily until the patient tolerated 24 h at an IMV
rate of four or less (n=43 patients); (iii) pressure support weaning, in which the level of pressure was decreased by 2 or 4 cmH2O twice daily until the patient could
tolerate breathing at pressure support of 8 cmH2O or less for 24 h (n=31 patients). The study concluded that pressure support was the preferred weaning mode based
on demonstrating that: (i) fewer patients failed to wean with pressure support
Page 296
than with the other modes (23% vs 43% (Tpiece) and 42% (IMV), P <0.05); (ii) the probability of requiring continued ventilatory support was lower with pressure
support than with the other modes (P=0.03); (iii) weaning duration was shorter with pressure support (mean 5.7±3.7 days) than for the other modes pooled (mean
9.3±8.2 days, P <0.05).
In a second trial of weaning modes, Esteban et al.38 randomized 130 patients who had failed initial weaning attempts to one of four weaning modes: (i) IMV
weaning, in which the rate was decreased by 2–4 breaths min–1 at least twice daily until the patient tolerated an IMV rate of five or less for 2 h (n=29 patients); (ii)
pressure support weaning, in which the pressure was decreased by 2–4 cmH2O at least twice daily until a pressure support level of 5 cmH2O was tolerated for 2 h
(n=37 patients); (iii) intermittent trials of spontaneous breathing, in which Tpiece trials of increasing length were undertaken at least twice daily until the patient could
tolerate 2 h of spontaneous breathing (n=33 patients); (iv) oncedaily Tpiece trials, in which a single Tpiece trial was undertaken daily until 2 h of spontaneous
breathing were tolerated without distress (n=31 patients). In contrast to Brochard et al., Estaban et al. concluded that a oncedaily Tpiece trial was the preferred
strategy, based on a higher rate of success and more rapid weaning with oncedaily Tpiece trials than with pressure support or IMV modes.
The most recent controlled trial of different weaning strategies, by Ely et al.39, randomized patients to a oncedaily respiratory assessment and trial of spontaneous
breathing for up to 2 h with physician notification of a successful trial (n=149 patients) versus usual care by the managing physicians (pulmonologist, cardiologist, or
intensivist, n=151 patients). A strategy of conducting daily trials of spontaneous breathing was associated with shorter weaning time (median 1 versus 3 days, P
<0.001), shorter duration of mechanical ventilation (median 4.5 versus 6 days, P <0.003), fewer total complications (20% versus 41%, P < 0.001), a lower rate of
reintubation (4% versus 10%, P=0.04), and lower ICU costs (median $15 740 versus $20 890, P <0.03). Taken together, these three studies show that contrary to
early views and practices, IMV weaning is the least likely to effect successful extubation and requires longer weaning than other available strategies, and that daily
assessment of respiratory status and spontaneous breathing trials was associated with shorter weaning duration than usual practice. Results of a recent metaanalysis of
weaning modes confirm these findings32.
Failure to free the patient from mechanical ventilation using one or more of the above techniques should prompt the clinician to consider special clinical
circumstances. These include unsuspectedly high imposed WOB from a small endotracheal tube or autoPEEP, occult cardiac ischemia, oversedation, and lack of
psychological readiness or motivation to wean.
18.11 Nonrespiratory impediments to successful weaning

Despite extensive investigation of pulmonary function parameters (e.g. gas exchange and mechanical function) as indicators of weaning, little attention has been given to
other, nonrespiratory features that may prove equally important. Indeed, failure to consider such factors as nutritional status and nonpulmonary organ dysfunction may
contribute to conflicting results obtained in studies of traditional (e.g. pulmonary function) weaning parameters40,41. Factors that determine weanability, in addition to
preexisting respiratory status and the reason for mechanical ventilation, include baseline function of other organ systems, complications developing while on the
ventilator, duration of critical illness, and the duration of mechanical ventilation. Nonrespiratory factors of particular importance include nutritional status, fluid balance,
metabolic and acidbase derangements, cardiac and renal function, effects of pharmacologic therapy, and neuropsychiatric factors.
Page 297
18.11.1 Nutritional status

Patients with acute respiratory failure are often malnourished either on admission or during hospitalization as a result of increased metabolic demands, inadequate
nutritional support or inability to utilize provided nutritional support. Malnutrition has adverse effects on the respiratory system, including decreased respiratory muscle
strength and function, decreased diaphragmatic mass (both area and thickness) and contractility, and decreased endurance42.
Malnutrition, particularly amino acid deficiency, decreases the ventilatory response to hypoxemia43. The immune system responds to malnutrition with decreased
cellmediated immunity, altered IgG turnover, and impaired macrophage function44. In 47 patients who received mechanical ventilation for 3 days or more, 93% of
those with adequate nutritional support could be weaned; however, only half of those with inadequate nutritional support could be weaned45. Among 14 'unweanable'
medical patients, all those who responded with increases in albumin and transferrin during parenteral nutrition were eventually weaned, but those who showed no
biochemical response to nutrition remained unweanable46. Amino acids are an essential part of parenteral nutrition, and have been shown to restore ventilatory drive
and improve both minute ventilation and respiratory rate43.
Although undernutrition is obviously an impediment to weaning, overnutrition or the wrong kind of nutrition can also frustrate weaning efforts. Excessive carbohydrate
intake by increasing the respiratory quotient (ratio of CO2 produced to O2 consumed)47, CO2 production and work of breathing can increase other causes of
increased CO2 production including fever, sepsis, shivering, seizures, and agitation. Increased CO2 production may also be caused by inefficient ventilation because of
increased dead space; this occurs in the setting of pulmonary embolism without compensatory hyperventilation, obstructive lung disease, late in the course of the acute
respiratory distress syndrome (ARDS), and with acute hyperinflation as a result of a rapid cycling rate and extrinsic PEEP or autoPEEP.
18.11.2 Metabolic abnormalities

Metabolic causes of weaning failure include hypophosphatemia, hypocalcemia, and hypothyroidism. Hypophosphatemic patients demonstrate increased
transdiaphragmatic pressure after repletion to normal levels of phosphate48. Hypocalcemia causes decreased diaphragmatic strength49. Hypothyroidism can affect
weaning, as the hypoxic ventilatory response is markedly blunted with thyroid deficiency50.
18.11.3 Acidbase disturbances

These are a common problem in patients who develop acute ventilatory failure in the face of chronic CO2 retention. As the baseline PaCO2 is frequently unknown,
these patients may be ventilatorsupported during their acute illness to a presumed normal baseline PaCO2 of 40 mmHg, causing acute alkalemia. Three or 4 days later,
a new compensatory state develops as the kidneys dump the excess bicarbonate (HCO3), leaving the patient unable to be weaned in this new steady state. Allowing
the PaCO2 to gradually normalize to the patient's premorbid baseline over several days may allow weaning to resume. Other acidbase problems that can develop
during weaning include respiratory alkalosis, metabolic alkalosis (common as a result of nasogastric drainage and diuretic therapy), and metabolic acidosis (commonly
as a result of disease or diabetic ketoacidosis). Complex disorders involving partial compensation and failure of two or more systems frequently confound the issue.
18.11.4 Infection
Infection, particularly sepsis, is a contraindication to weaning because of increased metabolic requirements and the propensity for hemodynamic compromise. The
presence of positive blood cultures predicts the need for reintubation in medical patients51. Similarly, an active infectious pulmonary process or tracheobronchitis may
be an indication for
Page 298
continued mechanical ventilation if secretions are thick or copious, especially in the presence of diminished pulmonary reserve.
18.11.5 Acute left ventricular dysfunction and anemia

In postoperative patients, the transition from controlled ventilation at a rate of 10 or 12 breaths min–1 to continuous positive airway pressure (CPAP) is associated with
a 10% increase in both O2 consumption and CO2 production52. Rapid weaning in patients with obstructive lung disease causes marked increases in transmural
pulmonary artery pressures53, which suggests that acute left ventricular dysfunction may occur during the transition from supported to spontaneous ventilation. Anemia,
by virtue of decreasing O2carrying capacity, interferes with weaning by forcing increases in minute ventilation or cardiac output. The role of erythropoietin in reducing
anemia and facilitating ventilation is currently under study.
18.11.6 Renal impairment and volume overload

These can be impediments to successful weaning. Indeed, application of PEEP is used to treat pulmonary edema because increased intrathoracic pressure discourages
venous return. Prolonged (10 day) diuretic therapy with reduction of body weight by an average of 5 kg is sometimes necessary to achieve weaning in patients with left
ventricular dysfunction53. Use of continuous dialysisultrafiltration techniques has been shown to reduce fluid overload in patients after coronary artery bypass surgery
and to facilitate their weaning from mechanical ventilation54.
18.11.7 Sleep deprivation

The ICU is frequently noisy, lights are always on, and the constant level of activity makes it difficult for patients to establish a circadian routine. The result may be a state
of sleep deprivation and twilight awareness where patients are neither fully awake nor asleep, nap throughout the day, or invert their daynight cycle. One common
approach is to give patients sedativehypnotic drugs at bedtime to promote a good night's sleep and to help restore a normal daily cycle. Because depression is
commonly a problem in the longterm ICU patient, antidepressant therapy at bedtime may be useful both for sedative effect and to forestall depression.
18.11.8 Motivation
The attitude of the patient is important, but often difficult to assess or change. Patients who are well informed about what is expected and who strive to meet daily goals
seem to progress more rapidly than those who are uninvolved in the process or who are depressed. Weaning goals set by the clinical team should be attainable, should
avoid precipitating fatigue or undue discomfort, and should cause each session of weaning to end in success rather than fatigue. From the practical viewpoint, a daily
weaning strategy should be developed early in the day and can be canceled if the day is otherwise occupied with laboratory tests or if the patient has a medical setback
that makes it impossible to successfully wean that day. Some patients with psychological dependence will hyperventilate at the onset of weaning trials without relation to
physiologic need and may have to be weaned surreptitiously. Formal biofeedback can also facilitate weaning by allowing the patient to visualize different breathing
patterns55. Potential solutions to the problem of dysfunctional breathing patterns include use of opioids to blunt respiratory drive and very slow changes in the level of
pressure support or IMV rate to prevent anxiety induced by sudden changes in the response of the lung stretch receptors during the active weaning process.
18.11.9 Other impediments

Upper airway obstruction, as caused by vocal cord paralysis, tracheal mucosal edema, or upper airway compression, can pose an unexpected impediment following
extubation, even though the patient has sustained spontaneous breathing well through the endotracheal tube.
Page 299
Use of the ‘leak test’, in which the volume of exhaled air is assessed as the endotracheal tube cuff is deflated, has been advocated as a way of predicting the
postextubation potency of the upper airway. Exhalation of less than 110 ml with the cuff deflated suggests the possibility of upper airway obstruction, although other
studies have challenged the predictive value of the ‘leak test’56. Interference with the weaning process may also be caused by multiple medications, including narcotics,
tranquilizers, sedative hypnotics, paralytic agents, and aminoglycoside antibiotics, which can cause myasthenialike muscle weakness. Indeed, daily interruption of
intravenous sedation, which allows serial reassessment of the patient’s immunologic status and need for sedation, has been shown to accelerate liberation from
mechanical ventilation40.
18.12 Respiratory impediments to weaning
18.12.1 Diaphragmatic dysfunction

This can occur secondary to phrenic nerve injury; the incidence is about 1% following coronary artery bypass grafting. Injury to the left phrenic nerve is more common
and was once thought to occur as a result of cold injury to the nerve during surgery57, although more recent experience suggests that phrenic nerve dysfunction
following internal mammary dissection may be caused by ischemia or traction on nerves. Bilateral diaphragmatic dysfunction should be considered in the differential
diagnosis of orthopnea after extubation and is confirmed if the decrement in vital capacity exceeds 20% on moving from an upright to a supine posture. Atelectasis
following coronary artery bypass surgery may also signify phrenic nerve dysfunction58. Phrenic nerve recovery may require up to 27 months59. When bilateral phrenic
nerve dysfunction precludes weaning, longterm management with a rocking bed or nighttime ventilation is required until respiratory muscle strength and endurance
improve.
Longterm, lowdose corticosteroids have been shown experimentally to impair dia phragmatic function60. Finally, critically ill patients often develop muscle
weakness or polyneuropathy with sepsis and multiple organ failure61,62. Muscle disuse itself may contribute to weakness as a result of progressive fiber atrophy63.
These factors, along with residual neuromuscular blockade, can contribute to weaning difficulty.
18.12.2 Patients with chronic obstructive pulmonary disease (COPD)

These pose a special weaning challenge. Rates of failure to wean in such patients may be as high as 24%, and most ventilated COPD patients (58%) require up to 2
weeks of mechanical ventilation64.
In ventilating and weaning COPD patients, special consideration must be given to the phenomenon of autoPEEP, also known as intrinsic PEEP. AutoPEEP is
defined as endexpiratory pressure present in the alveoli but not measured at the airway manometer without special maneuvers such as an endexpiratory hold65. Auto
PEEP must be distinguished from PEEP purposefully applied and recorded in the proximal airway. AutoPEEP occurs commonly among mechanically ventilated
patients (39–62% of patients assessed) and may produce high alveolar pressures at end expiration (i.e. up to 15 cmH2O)66. AutoPEEP is increased when airflow
resistance is increased and when exhalation time is decreased, both of which impair alveolar emptying before the next ventilator breath. AutoPEEP may frustrate
weaning by increasing the work of breathing associated with having to overcome residual alveolar pressure before inhalation can begin67. Adding conventional PEEP in
COPD patients can offset this inspiratory muscle loading effect of autoPEEP68 that is associated with dynamic hyperinflation and airflow limitation. Maneuvers to
decrease autoPEEP also include treating bronchospasm and secretions, slowing the respiratory rate, maximizing the time for expiration, and using lowcompliance
ventilator tubing to permit rapid gas delivery and to maximize exhalation time.
Page 300
18.12.3 Perception of breathing

In patients with obstructive lung disease, pulmonary stretch receptors are activated with mechanically delivered volumes much larger than those attained during
spontaneous breathing. Failure to recreate these volumes during weaning can lead to agitation69. One approach to this problem is to decrease the tidal volume (VT)
during rest periods of mechanical ventilation and increase the supported respiratory rate as necessary to maintain alveolar ventilation. During weaning periods, opioids
can be added judiciously to manage anxiety and feelings of respiratory insufficiency.
18.12.4 Endotracheal tube caliber and tracheostomy

Airway resistance increases as endotracheal tube size decreases. To the extent that smallercaliber tubes (e.g. for nasal intubation) can greatly increase the imposed
work of breathing, they can impair weaning70. Thus, use of larger caliber (i.e. >7.5 mm inner diameter) endotracheal tubes and avoidance of nasotracheal intubation in
patients experiencing weaning difficulty is advised. Tracheotomy is indicated for relief of upper airway obstruction, for access for pulmonary hygiene, and for control of
the airway for longterm positive pressure ventilation71. Tracheostomy appears to help patients who have trouble clearing secretions, may improve patient comfort, and
allows use of a fenestrated tracheostomy tube, which may be occluded to allow the patient to speak during periods of spontaneous breathing. Complications of
tracheostomy include pneumothorax, pneumomediastinum, subcutaneous emphysema, incisional hemorrhage, aerophagia, aspiration, and tube displacement72. Late
complications also include tracheal stenosis, tracheoinnominate artery and tracheoesophageal fistulae, tube obstruction, aspiration, swallowing dysfunction, and stomal
infections. In the absence of definitive studies regarding the timing of tracheostomy, we suggest the following indications for tracheotomy during the weaning process:
• anticipation of prolonged (e.g. >2 weeks) intubation and mechanical ventilation;
• absence of a contraindication to tracheostomy (e.g. upper airway burn, mediastinitis);
• copious or tenacious secretions requiring direct access to the airway for suctioning;
• chronic aspiration requiring longstanding airway protection in the absence of other strategies to minimize aspiration (e.g. gastrostomy tubes, jejunostomy tubes,
nasogastric feeding).
Clinical predictors for prolonged ventilatory support and eventual tracheostomy include development of nosocomial pneumonia, administration of aerosol treatments,
witnessed aspiration event, and requirement for reintubation73.
18.13 Pharmacologic aids to weaning

Betaadrenergic agents, aminophylline, and anticholinergics are frequently helpful in the patient with reactive airway disease (see Chapter 17). Aminophylline has also
been shown to improve diaphragmatic contractility74 and to partially mitigate the reduced diaphragmatic activity seen after abdominal surgery75. Dopamine improves
diaphragmatic strength and blood flow in patients with COPD and acute respiratory failure. An infusion of 10 µg kg–1 min–1 typically increases heart rate by 17%,
cardiac output by 40%, and diaphragmatic blood flow and transdiaphragmatic strength by 30%76. Opioids, particularly dihydrocodeine, have been shown to decrease
breathlessness and improve exercise tolerance in patients with COPD and may have a role in improving patient comfort during weaning77. Respiratory stimulants such
as doxapram have been successfully employed in patients with primary alveolar hypoventilation78 but are not generally recommended in weaning, because respiratory
drive should be adequate. Androgens such as oxandrolone decrease net nitrogen loss and improve wound healing79 and
Page 301
have anecdotally been used to promote weaning, although studies are lacking. Serum testosterone levels decline in elderly men, and replacement therapy may restore
body weight and even body mass80.
18.14 Weaning failure

Despite the best efforts to prepare the patient for weaning and ultimately extubation, some patients will inevitably fail to sustain respiration on their own and will require
resumption of ventilatory support. The rate of extubation failure in general surgical patients averaging 65 years of age is approximately 5%81. In patients with head
injury, approximately 5% require reintubation, primarily for airway protection. Patients undergoing open heart surgery also have an overall reintubation rate of 5–7%,
with older patients and those undergoing reoperation, emergency surgery, and valve surgery having a much higher rate of reintubation than those undergoing firsttime
isolated coronary artery bypass surgery82. Predictors of extubation failure include age 65 years or over, preoperative hospitalization, arterial vascular disease, COPD,
pulmonary hypertension, severe left ventricular dysfunction, cardiogenic shock, hematocrit of 34% or less, renal insufficiency, malnutrition, reoperation, massive
transfusion, prolonged cardiopulmonary bypass time and complex procedures83. Reintubation and failure to wean after all issues have been addressed suggests the
need for patient placement in a longterm ventilator facility, assuming the patient desires aggressive management.
18.15 Longterm prognosis

The prognosis after successful weaning from mechanical ventilation varies widely, depending on the age of the patient, the underlying disease process, and the presence
of other complicating factors. Diseases such as ARDS are associated with high hospital mortality (up to 65%), but similar survival rates at hospital discharge, 1 year,
and 3 years84. In a study of 100 patients receiving mechanically assisted ventilation in a community hospital, all patients under the age of 50 years who survived the ICU
stay survived to leave the hospital and were alive at 1 year85. However, in the subset of older patients (i.e. age ≥70 years), hospital survival was only 49% and 1 year
survival was only 27%. Patients with COPD who experience acute ventilatory failure have been estimated to have a 57% 1 year and a 51% 2 year survival rate85.
Although longterm survival prediction helps inform the clinician’s understanding of average outcomes, these data must be applied with circumspection to individual
patients and with attention to the patient’s overall progress to date.
References
1. Pavlin, E.G., Holle, R.H. and Schoene, R.B. (1989) Recovery of airway protection compared with ventilation in humans after paralysis with curare. Anesthesiology
70:381–385.
2. Engelman, R.M., Rousou, J.A., Flack, J.E., III, et al. (1994) Fasttrack recovery of the coronary bypass patient. Am. Thorac. Surg. 58:1742–1746.
3. Higgins, T.L. (1995) Safety issues regarding early extubation after coronary artery bypass surgery. J. Cardiothorac. Vasc. Anesth. 9 (5S1): 24–29.
4. Horst, H.M., Mouro, D., HallJenssens, R.A. and Pamukov, N. (1998) Decrease in ventilation time with a standardized weaning process. Arch. Surg. 133:483–
489.
5. Koleff, M.H., Shapiro, S.D., Silver, P., et al. (1997) A randomized, controlled trial of protocoldirected versus physiciandirected weaning from mechanical
ventilation. Crit. Care Med. 25: 567–574.
6. Cohen, I.L., Bari, N., Strosberg, M.A., et al. (1991) Reduction of duration and cost of mechanical ventilation in an intensive care unit by use of a ventilatory
management team. Crit. Care Med. 19:1278–1284.
7. Kastanos, N., Miro, R.E., Perez, A.M., et al. (1983) Laryngotracheal injury due to endotracheal intubation: Incidence, evolution, and predisposing factors. A
prospective longterm study. Crit. Care Med. 11:362–367.
8. Stauffer, J.L., Olson, D.E. and Petty, T.L. (1981) Complications and consequences of endotracheal intubation and tracheostomy. A prospective study of 150
critically ill adult patients. Am. J. Med. 70:65–76.
Page 302
9. Alexopoulos, C., Jansson, B. and Lindholm, C.E. (1984) Mucus transport and surface damage after endotracheal intubation and tracheostomy. An experimental
study in pigs. Acta Anaesthesiol. Scand. 28:68–76.
10. Driver, A.G. and LeBrun, M. (1980) Iatrogenic malnutrition in patients receiving ventilatory support. JAMA 244:2195–2196.
11. Ranieri, V.M., Suter, P.M., Tortorella, C., et al. (1999) Effect of mechanical ventilation on inflammatory mediators in patients with acute respiratory distress
syndrome. JAMA 282:54–61.
12. Stoller, J.K. (1991) Establishing clinical unweanability. Respir. Care 36:186–198.
13. Krieger, B.P., Ershowsky, P.F., Becker, D.A., et al. (1989) Evaluation of conventional criteria for predicting successful weaning from mechanical ventilatory
support in elderly patients. Crit. Care Med. 17:858–861.
14. Yang, K.L. and Tobin, M.J. (1991) A prospective study of indexes predicting the outcome of trials of weaning from mechanical ventilation. N. Engl. J. Med.
324:1445–1450.
15. Scheinhorn, D., Hassenpflug, M., Artinian, B.M., et al. (1995) Predictors of weaning after six weeks of mechanical ventilation. Chest 107: 500–505.
16. Epstein, S.K. (1995) Etiology of extubation failure and the predictive value of the rapid shallow breathing index. Am. J. Respir. Crit. Care Med. 152:545–549.
17. Chatila, W., Jacob, B., Guaglionine, D. and Manthous, C.A. (1996) The unassisted respiratory ratetidal volume ratio accurately predicts weaning outcome. Am.
J. Med. 101:61–67.
18. Valverdu, I., Calaf, N., Subirana, M., et al. (1998) Clinical characteristics, respiratory functional parameters, and outcomes of a twohour Tpiece trial in patients
weaning from mechanical ventilation. Am. J. Respir. Crit. Care Med. 158: 1855–1862.
19. Bellemare, F. and Grassino, A. (1982) Effect of pressure and timing of contraction on human diaphragm fatigue. J. Appl. Physiol. 53: 1190–1195.
20. Petros, A.J., Lamond, C.T. and Bennett, D. (1993) The Bicore pulmonary monitor. A device to assess the work of breathing while weaning from mechanical
ventilation. Anaesthesia 48: 985–988.
21. Banner, M.J., Kirby, R.R., Gabrielli, A., Blanch, P.B. and Lyon, A.J. (1994) Partially and totally unleading respiratory muscles based on realtime measurements
of work of breathing. Chest 106:1835–1842.
22. Fiastro, J.P., Habib, M.P., Shon, B.Y. and Campbell, S.C. (1988) Comparison of standard weaning parameters and the mechanical work of breathing in
mechanically ventilated patients. Chest 94:223–238.
23. Gluck, E.H., Barkoviak, M.J., Balk, R.A., et al. (1995) Medical effectiveness of esophageal balloon pressure manometry in weaning patients from mechanical
ventilation. Crit. Care Med. 23: 504–509.
24. Harpin, R.P., Baker, J.P., Downer, J.P., et al. (1987) Correlation of the oxygen cost of breathing and length of weaning from mechanical ventilation. Crit. Care
Med. 15:807–812.
25. Lewis, W.D., Chwals, W. Bennotti, P.N., et al. (1998) Bedside assessment of the work of breathing. Crit. Care Med. 6:117–122 Q6.
26. Murciano, D., Boczkowski, J., Lecocguir, Y., et al. (1988) Tracheal occlusion pressure: A simple index to monitor respiratory muscle fatigue during acute
respiratory failure in patients with chronic obstructive pulmonary disease. Ann. Intern. Med. 108:800–803.
27. Cohen, C.A., Zagelbaum, G., Cross, D., et al. (1982) Clinical manifestations of inspiratory muscle fatigue. Am. J. Med. 73:308–316.
28. Tobin, M.J., Guenther, S.M., Peres, W., et al. (1987) KonnoMead analysis of rib cageabdominal motion during successful and unsuccessful trials of weaning from
mechanical ventilation. Am. Rev. Respir. Dis. 135: 1320–1328.
29. Venus, B., Smith, R.A. and Mathru, M. (1987) National survey of methods and criteria used for weaning from mechanical ventilation. Crit. Care Med. 15:530–
533.
30. Esteban, A., Alia, L, Ibanez, J., et al (1994) Modes of mechanical ventilation and weaning: A national survey of Spanish hospitals. Chest 106:1188–1193.
31. Downs, J.B., Klein, E.F., Desautels, D., et al. (1973) Intermittent mandatory ventilation: A new approach to weaning patients from mechanical ventilation. Chest
64:331–335.
32. Butler, R., Keenan, S.P., Inman, K.J., Sibbald, W.J. and Block, G. (1999) Is there a preferred technique for weaning the difficulttowean patients? A
systematic review of the literature. Crit. Care Med. 27:2331–2336.
33. MacIntyre, N.R. (1986) Respiratory function during pressure support ventilation. Chest 89: 677–682.
34. Kacmarek, R.M. (1988) The role of pressure support ventilation in reducing work of breathing. Respir. Care 334:99–120.
35. MacIntyle, N.R. (1988) Weaning from mechanical ventilatory support: Volumeassisting intermittent breaths versus pressureassisting every breath. Respir. Care
33:121–125.
Page 303
36. Johannigman, J.A., David, K., Campbell, R.S., et al. (1997) Use of the rapid/shallow breathing index as an indicator of patient work of breathing during pressure
support ventilation. Surgery 122:737–741.
37. Brochard, L., Rauss, A., Benito, S., et al. (1994) Comparison of three methods of gradual withdrawal from ventilatory support during weaning from mechanical
ventilation. Am. J. Respir. Crit. Care Med. 150: 896–903.
38. Esteban, A., Frutoes, F., Tobin, M.J., et al. (1995) A comparison of four methods of weaning patients from mechanical ventilation. N. Engl J. Med. 332:345–350.
39. Ely, E.W., Baker, A.M., Dunagan, D.P., et al. (1996) Effect on the duration of mechanical ventilation of identifying patients capable of breathing spontaneously. N.
Engl. J. Med. 335: 1864–1869.
40. Kress, J.P., Pohlman, A.S., O’Connor, M.F. and Hall, J.B. (2000) Daily interruption of sedative infusions in critically ill patients undergoing mechanical
ventilation. N. Engl. J. Med. 342: 1471–1477.
41. Morganroth, M.L., Morganroth, J.L., Nett, L.M. and Petty, T.L. (1984) Criteria for weaning from prolonged mechanical ventilation. Arch. Intern. Med.
144:1012–1016.
42. Wilson, D.O. and Rogers, R.M. (1989) The role of nutrition in weaning from mechanical ventilation. J. Intensive Care 4:124–133.
43. Weissman, C., Askanazi, J., Rosenbaum, S., et al. (1983) Amino acids and respiration. Ann. Intern. Med. 98:41–44.
44. Pingleton, S.K. and Harmon, G.S. (1987) Nutritional management in acute respiratory failure. JAMA 257:3094–3099.
45. Bassili, H.R. and Deitel, M. (1981) Effect of nutritional support on weaning patients off mechanical ventilators. J. Parenter. Enteral Nutr. 5:161–163.
46. Larca, L. and Greenbaum, D.M. (1982) Effectiveness of intensive nutritional regimes in patients who fail to wean from mechanical ventilation. Crit. Care Med.
10:297–300.
47. Covelli, H.D., Black, J.W., Olsen, M.S. and Beekman, J.E (1981) Respiratory failure precipitated by high carbohydrate loads. Ann. Intern. Med. 95:579–581.
48. Aubier, M., Murciano, D., Lecocguic, Y., et al. (1985) Effect of hypophosphatemia on diaphragmatic contractility in patients with acute respiratory failure. N.
Engl. J. Med. 313: 420–424.
49. Aubier, M., Viires, N., Piquet, J., et al. (1985) Effects of hypocalcemia on diaphragmatic strength generation. J. Appl. Physiol. 58: 2054–2061.
50. Zwillich, C.W., Pierson, D.J., Hofeldt, F.D., et al. (1975) Ventilatory control in myxedema and hypothyroidism. N. Engl. J. Med. 292:662–665.
51. Tahvanainen, J., Salmenpera, M. and Nikki, P. (1983) Extubation criteria after weaning from intermittent mandatory ventilation and continuous positive airway
pressure. Crit. Care Med. 11:702–707.
52. Kemper, M., Weissman, C., Askanazi, J., et al. (1987) Metabolic and respiratory changes during weaning from mechanical ventilation. Chest 92:979–983.
53. Lemaire, F., Teboul, J.L., Cinotti, L., et al. (1988) Acute left ventricular dysfunction during unsuccessful weaning from mechanical ventilation. Anesthesiology
69:171–179.
54. Coraim, F.J., Coraim, H.P., Ebermann, R. and Stellwag, F.M. (1986) Acute respiratory failure after cardiac surgery: Clinical experience with the application of
continuous arteriovenous hemofiltration. Crit. Care Med. 14:714–718.
55. Holliday, J.F. and Hyers, T.M. (1990) The reduction of weaning time from mechanical ventilation using tidal volume and relaxation biofeedback. Am. Rev. Respir.
Dis. 141: 1214–1220.
56. Engoren, M. (1999) Evaluation of the cuffleak test in a cardiac surgery population. Chest 116: 1029–1031.
57. Wheeler, W.E., Rubis, L.J., Jones, C.W., et al. (1985) Etiology and prevention of topical cardiac hypothermiainduced phrenic nerve injury and left lower lobe
atelectasis during cardiac surgery. Chest 88:680–683.
58. Wilcox, P., Baile, E.M., Hards, J., et al. (1988) Phrenic nerve function and its relationship to atelectasis after coronary artery bypass surgery. Chest 93:693–698.
59. Abd, A.G., Braun, N.M.T., Baskin, M.I., et al. (1989) Diaphragmatic dysfunction after open heart surgery: Treatment with a rocking bed. Ann. Intern. Med.
111:881–886.
60. VanBalkom, R.H.H., Dekhuijzen, P.N.R., Folgering, H.T.M., et al. (1997) Effects of longterm lowdose methylprednisone on rat diaphragm function and
structure. Muscle Nerve 20:983–990.
61. Latronico, N., Fenzi, F., Recupero, D., et al. (1996) Critical illness myopathy and neuropathy. Lancet 347:1579–1582.
62. Lorin, S., Sivak, M. and Nierman, D.M. (1998) Critical illness polyneuropathy: What to look for in atrisk patients. Diagnosis requires a high index of suspicion. J.
Crit. Illness 13: 608–612.
Page 304
63. Ibebunjo, C. and Martyn, J.A.J. (1999) Fiber atrophy, but not changes in acetylcholine receptor expression, contributes to the muscle dysfunction after
immobilization. Crit. Care Med. 27:275–285.
64. Menzies, R., Gibbons, W. and Goldberg, P. (1989) Determinants of weaning and survival among patients with COPD who require mechanical ventilation for acute
respiratory failure. Chest 95:398–405.
65. Ranieri, V.M., Grasso, S., Fiore, T., et al. (1996) Autopositive endexpiratory pressure and dynamic hyperinflation. Clin. Chest Med. 17: 379–394.
66. Brown, D.G. and Pierson, D.J. (1986) AutoPEEP is common in mechanically ventilated patients: A study of incidence, severity, and detection. Respir. Care
31:1069–1074.
67. Kirton, O.C., DeHaven, C.B., Morgan, J.P., et al. (1995) Elevated imposed work of breathing masquerading as ventilator weaning intolerance. Chest 108:1021–
1025.
68. Petrof, B.J., Legare, M., Goldberg, P., et al. (1990) Continuous positive airway pressure reduces work of breathing and dyspnea during weaning from mechanical
ventilation in severe chronic obstructive pulmonary disease. Am. Rev. Respir. Dis. 141:281–289.
69. Wolkove, N., Altose, M.D., Kelsen, S.G., et al. (1981) Perception of changes in breathing in normal human subjects. J. Appl. Physiol. 50: 78–83.
70. Habib, M.P. (1989) Physiologic implication of artificial airways. Chest 96:180–184.
71. Heffner, J.E., Scott Miller, K. and Sahn, S.A. (1986) Tracheostomy in the intensive care unit. Part I: Indications, technique, management. Chest 90:269–274.
72. Heffner, J.E., Scott Miller, K. and Sahn, S.A. (1986) Tracheostomy in the intensive care unit. Part 2: Complications. Chest 90:430–436.
73. Kollef, M.H., Ahrens, T.S. and Shannon, W. (1999) Clinical predictors and outcomes for patients requiring tracheostomy in the intensive care unit. Crit. Care Med.
27(9): 1714–1720.
74. Aubier, M., De Troyer, A., Sampson, M., et al. (1981) Aminophylline improves diaphragmatic contractility. N. Engl. J. Med. 305:249–252.
75. Dureuil, B., Desmonts, J.M., Mankikian, B. and Prokocimer, P. (1985) Effects of aminophylline on diaphragmatic dysfunction after upper abdominal surgery.
76. Aubier, M., Murciano, D., Menu, Y., et al. (1989) Dopamine effects on diaphragmatic strength during acute respiratory failure in chronic obstructive pulmonary
disease. Ann. Intern. Med. 110:17–23.
77. Woodcock, A.A., Gross, E.R., Gellert, A., et al. (1981) Effects of dihydrocodeine, alcohol, and caffeine on breathlessness and exercise tolerance in patients with
chronic obstructive lung disease and normal blood gases. N. Engl. J. Med. 305:1611–1616.
78. Lugliani, R., Whipp, B.J. and Wasserman, K. (1979) Doxapram hydrochloride: A respiratory stimulant for patients with primary alveolar hypoventilation. Chest
76:414–419.
79. Demling, R.H. and Orgill, D.P. (2000) The anticatabolic and wound healing effects of the testosterone analog oxandrolone after severe burn injury. j. Crit. Care
15:12–17.
80. Bagatell, CJ. and Bremner, W.J. (1996) Androgens in men—uses and abuses. N. Engl. J. Med. 334:707–714.
81. Demling, R.H., Read, T., Lind, L.J. and Flanagan, H.L. (1988) Incidence and morbidity of extubation failure in surgical intensive care patients. Crit. Care Med.
16:573–577.
82. Higgins, T.L. (1989) Postoperative care of the cardiac surgery patient. Probl. Anesth. 3:211–227.
83. Rady, M.Y. and Ryan, T. (1999) Perioperative predictors of extubation failure and the effect on clinical outcome after cardiac surgery. Crit. Care Med. 27:340–347.
84. Schmidt, C.D., Elliott, C.G., Carmelli, D., et al. (1983) Prolonged mechanical ventilation for respiratory failure: A costbenefit analysis. Crit Care Med. 11:407–
411.
85. Connors, A.F., Jr, Dawson, N.V., Thomas, C., et al. (1996) Outcomes following acute exacerbation of severe chronic obstructive lung disease. The SUPPORT
investigators (Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment). Am. J. Respir. Crit. Care Med. 154:959–967.
86. Sahn, S.A., Lakshminarayan, S. (1973) Bedside criteria for discontinuation of mechanical ventilation. Chest 63:1002–1005.
87. Yang, K.L. and Tobin, M.J. (1989) Decision analysis of parameters used to predict outcome of a trial of weaning from mechanical ventilation (abstract). Am. Rev.
Respir. Dis. 139:(4, part 2) A98.
88. Milbern, S.M., Downs, J.B., Jumper L.C. and Modell, J.H. (1978) Evaluation of criteria for discontinuing mechanical ventilatory support. Arch. Surg.
113:1441–1443.
89. Pardee, N.E., Winterbauer, R.H. and Allen, J.D. (1984) Bedside evaluation of respiratory distress. Chest 85:203–206.
Page 305
90. DeHaven, C.B., Hurst, J.M. and Branson, R.D. (1986) Evaluation of two different extubation criteria; Attributes centributing to success. Crit. Care Med. 14:92–
94.
91. Stoller, J.K. (2000) Weaning from mechanical ventilation. SCCM/ACCP. Combined Critical Care Course Syllabus: 491–506.
Page 306

Page 307
Chapter 19
Noninvasive positive pressure ventilation (NIPPV)
Contents
19.1 Introduction
19.2 NIPPV for acute respiratory failure
19.3 Indication for noninvasive positive pressure
19.4 When should NIPPV be initiated for acute respiratory failure?
19.5 Complications or problems with NIPPV
19.6 The application of NIPPV
19.7 Summary
Page 308
19.1 Introduction
The use of noninvasive applications of ventilatory assistance can be traced to the 1800s where prototypes of the ‘Iron Lung’ were first used1. However, none of these
devices were in general use when Drinker in 1928 developed the first electrically powered Iron Lung2. Jack Emerson in 1931 greatly improved upon Drinker’s original
design and marketed a less expensive, lighter, quieter, and easier to operate unit3. However, because of the size of negative pressure ventilators and the inability to well
coordinate gas delivery with patient demand, the use, in all but the rare neuromuscularly or neurologically diseased patient, has declined since the late 1970s4.
Noninvasive positive pressure is also not new. Barach et al. in the 1930s originally proposed the use of noninvasive continuous positive air pressure (CPAP) for the
management of acute pulmonary edema5. Intermittent positive pressure breathing (IPPB) by mouthpiece or mask was widely employed in the management of critically
ill patients from the 1960s up to the mid1980s4. The use of NIPPV for nocturnal ventilator support dates back to the early 1960s at the Goldwater Rehabilitation
Center in New York for the treatment of neuromuscularly diseased patients6. However, the use of a face mask to apply NIPPV to patients with Duchenne muscular
dystrophy can be credited to Rideau et al. in France in 19847.
The use of NIPPV to manage acute respiratory failure (ARF) is based on data demonstrating that NIPPV reduced esophageal pressure swings and diaphragmatic
electromyogram activity in patients with acute respiratory distress (Figure 19.1)8,9 and three case series that demonstrated that NIPPV could prevent intubation in
patients with acute respiratory failure9–11. In this chapter the use of NIPPV and noninvasive CPAP for the management of ARF will be discussed.
19.2 NIPPV for acute respiratory failure

Acute respiratory failure is characterized by the inability to eliminate CO2 and to oxygenate the blood. As illustrated the Figures 19.2 and 19.3 from a recent
consensus conference, there are many points in the mechanisms that induce hypercarbia and hypoxemia where NIPPV or CPAP can reverse the acute failure12. In the
setting of acute failure noninvasive positive pressure is used to reverse the hypercarbia and hypoxemia by unloading respiratory muscles, augmenting alveolar ventilation
and oxygenation, and improving cardiovascular status.
19.3 Indication for noninvasive positive pressure

Table 19.1 lists the clinical settings where noninvasive positive pressure has been used. Although these settings are extensive, outcome data supporting the use of
NIPPV are limited to a few of these indications. Overall, as noted in the following sections, more randomized controlled trials demonstrating improved outcome
associated with the use of NIPPV exist than in any other aspect of critical respiratory care.
Table 19.1 Settings in which noninvasive positive pressure has been used during acute respiratory failure
• Acute exacerbation of COPD

• NonCOPD hypercarbic acute respiratory failure
• Cardiogenic pulmonary edema
• Hypoxemic respiratory failure
• Immunocompromised patients
• Weaning from ventilatory support
• Communityacquired pneumonia
• Asthma
• Do not resuscitate/do not intubate patients
• Postoperative patients
Page 309
Figure 19.1 Recording from a representative patient during spontaneous unassisted breathing and after 45 min of treatment with 10
cmH2O inspiratory positive airway pressure, zero PEEP. from: Brochard, Isabey, D., Piquet, J., Amaro, P., Mancebo, J. and
Messadi, A.A. (1990) Reversal of acute exacerbations of chronic obstructive lung disease by inspiratory assistance with a face
mask. N. Engl. J. Med, 95:865–870.
19.3.1 COPD
The use of NIPPV in COPD patients with an uncomplicated acute exacerbation is now considered a standard of care. At least seven randomized, controlled
prospective trials primarily enrolling COPD patients have demonstrated that NIPPV in this population avoids intubation, decreases mortality, decreases the length of
mechanical ventilation, and decreases ICU and hospital stay13–19. In addition, a recent metaanalysis by Keenan et al.20 concluded that NIPPV in COPD decreases
frequency of intubation and mortality.
Figure 19.4 from Kramer et al. illustrates the percentage of patients randomized to NIPPV and those randomized to standard medical management without NIPPV
who required intubation14. Overall a greater percentage in the nonNIPPV group required intubation; however, in those patients with COPD approximately 70%
without NIPPV required intubation, whereas only about 10% in the NIPPV group required intubation. Similar data were provided by Brochard et al. in a large
European randomized trial of NIPPV15. In the NIPPV group 11 of 43 (26%) patients required endotracheal intubation compared with 31 of 42 (74%) patients in the
standard therapy group. In addition, significant differences in mortality were observed between these groups. In the NIPPV group four (9%) patients died versus 12
Page 310
Figure 19.2 When PaCO2 is increased, and minute ventilation is normal or increased, the respiratory muscles are failing to generate
sufficient alveolar ventilation to eliminate the CO2 being produced, Means of correcting this pathophysiology include increasing
alveolar ventilation by increasing tidal volume and/or respiratory rate, and reducing CO2 production ( CO2) by decreasing the
work of breathing. Respiratory muscle failure can occur when the work of breathing is normal (e.g. numerous acute or chronic
neuromuscular problems), or increased (e.g. patients with COPD, asthma, or the obesity hypoventilation syndrome), and
presumably because of inadequate delivery of oxygen to the respiratory muscles (e.g. approximately onethird of patients
presenting with cardiogenic pulmonary edema), When PaCO2 is increased and minute ventilation is low the level of
consciousness is generally impaired. Such patients usually require intubation for airways protection in addition to ventilatory
assistance, unless the hypercapnia can be rapidly reversed. From: Consensus statement (2001) International Consensus
Conference in Intensive Care Medicine: Noninvasive Positive Pressure Ventilation in Acute Respiratory Failure. Reproduced
with permission from Am. J, Respir, Crit. Care Med. 163:283–291.
(29%) in the standard therapy group. The authors went on to comment that in COPD patients mortality was increased by intubation because of the increased likelihood
of nosocomial pneumonia, the high probability of a prolonged ventilator course and the general complications associated with endotracheal intubation in the group of
patients.
Two recent studies, one observational21 and the other using historical controls22, concluded that endotracheal intubation in COPD increased the likelihood of all
types of nosocomial infections. Guerin et al.21 observed in 230 consecutive patients staying 2 days or longer in the ICU and requiring 1 day or more of mechanical
ventilation that none of the 60 patients managed only with NIPPV developed nosocomial pneumonia, whereas 18% of the patients first treated with NIPPV then
intubated, 22% of the patients first intubated than
Page 311
Figure 19.3 Hypoxemia develops as a result of alveolar hypoventilation (which is accompanied by increases in PaCO2 and is addressed
in Figure 19.1) and from perfusion going to areas where the ratio of alveolar ventilation ( the less the effect), and by
recruiting airspaces. Airspace derecruitment occurs when the transpulmonary pressure falls below the airspace collapsing or
the closing pressure (as occurs in numerous conditions that alter surfactant or that decrease the lung or the chest wall
compliance), and when the transpulmonary pressure applied during inhalation fails to exceed airspace opening pressure.
Accordingly, airspace opening can be facilitated by increasing the transpulmonary pressure applied at end exhalation (CPAP)
and at end inhalation (i.e. Inspiratory positive airway pressure (IPAP)). An additional beneficial effect of CPAP and IPAP
may be seen in patients with cardiogenic pulmonary edema, as they all reduce venous return and functionally reduce left
ventricular afterload, abns, abnormalities; Crs, respiratory system compliance. From: Consensus statement (1982) international
Consensus Conference in Intensive Care Medicine: Noninvasive Positive Pressure Ventilation in Acute Respiratory Failure.
Reproduced with permission from Am. J, Respir. Crit. Care Med 163:283–291.
Page 312
changed to NIPPV, and 28% of the patients intubated only developed nosocomial pneumonia. Girou et al.22 compared 50 patients with COPD or severe
cardiogenetic pulmonary edema who were managed with NIPPV with 50 matched controls endotracheally intubated
Figure 19.4 Need for intubation (percentage of patients) among all patients (a) and among COPD patients (b) plotted against time in the
study. Control patients; □, patients receiving noninvasive positive pressure ventilation (NIPPV). n=16 16 NPPV patients (12
with COPD) and 15 control patients (11 with COPD). *P<0.05 compared with control patients, From; Kramer, N., Meyer,
T.J., Meharg, J., Cece, R.D. and Hill, N.S. (1995) Randomized, prospective trial of noninvasive positive pressure ventilation in
acute respiratory failure. Reproduced with permission from Am. J. Rspir. Crit. Care Med. 151:1799–1806.
and ventilated. They also found that the nosocomial infection rates were significantly lower in the NIPPV group (Figure 19.5).
The largest randomized, controlled trial of NIPPV in COPD was conducted by Plant et al. in patients not admitted to the ICU but maintained on medical wards18.
The results were similar to those of Brochard et al.15 and Kramer et al.14; that is, only 18 of 118 (15%) patients in the NIPPV group required intubation versus 32 of
118 (27%) in the standard therapy group. The single negative study of NIPPV in COPD was reported by Barbe et al.19, who found no difference in outcome between
the NIPPV and control group. However, no patient in either the control or treatment group died or required intubation. On close examination it would appear that none
of the patients recruited into this study was sick enough to need any type of ventilatory assistance. In no
Figure 19.5 Frequency of nosocomial infections in 50 cases treated with noninvasive ventilation (NIV) and 50 controls treated with
intubation and conventional mechanical ventilation (MV). P values between the two groups are 0.04 for nosocomial pneumonia,
0.03 for urinary tract infections and 0.002 for catheterrelated infection. From: Girou, E., Schortgen, F., Delclaux, C., Brun
Buisson, C., Blot, F., Lefort, Y., Lemaire, F. and Brochard, L. (2000) Association of noninvasive ventilation with nosocomial
infection and survival in critically ill patients. JAMA 284:2361–2367,
Page 313
study have patients randomized to the NIPPV group done more poorly on any outcome variable than those randomized to the control group.
Clearly, the data on COPD patients would seem to indicate that NIPPV should be available to all patients with an acute exacerbation of COPD in any institution
treating these patients. As noted in Table 19.2, NIPPV is indicated in patients with acute hypercarbic respiratory failure that do not have cardiovascular instability to
the level requiring immediate intubation. Additional exclusions from NIPPV are facial or airway trauma, inability to protect the airway, and inability to clear secretions in
spite of the use of NIPPV. Of course, NIPPV cannot be used in patients who will not tolerate NIPPV. However, patient tolerance may be more a problem with the
caregiver and available equipment (see later sections) than with
Table 19.2 Selection guidelines: noninvasive ventilation for patients with COPD and acute respiratory failure
Step 1. Identify patients in need of ventilatory assistance:

A. Symptoms and signs of acute respiratory distress:
(i) moderate to severe dyspnea, increased over usual and
(ii) RR > 24, accessory muscle use, paradoxical breathing
B. Gas exchange abnormalities:
(i) PaCO2>45 mmHg, pH<7.35 or
(ii) PaCO2/FIO2<200
Step 2. Exclude those at increased risk with noninvasive ventilation:
A. Respiratory arrest
B. Medically unstable (hypotensive shock, uncontrolled cardiac ischemia or arrhythmias)
C. Unable to protect airway (impaired cough or swallowing mechanism)
D. Excessive secretions
E. Agitated or uncooperative
F. Facial trauma, burns, or surgery, or anatomic abnormalities interfering with mask fit
From: Mehta, S. and Hill, N.S. (2001) Noninvasive ventilation. Reproduced with permission f from Am. J. Respir. Crit. Care Med. 163:540–577.
the patient. Patients most likely to be successfully managed with NIPPV are young (<70 years), with lower acuity of illness scores (APACHE) and better able to
cooperate. They also have minimal air leak, have moderate to severe hypercarbia (PaCO2 >45 but <90 mmHg) and moderate to severe acidemia (pH <7.35 but
>7.10)4,23.
19.3.2 NonCOPD hypercapnic respiratory failure

Few systematically acquired data and no randomized controlled trials are available on the use of NIPPV in nonCOPD hypercapnic respiratory failure. However, the
use of NIPPV in patients with neurologic and neuromuscular disease has been available for acute and chronic management of these patients for over 20 years24–27. We
would encourage the early use of NIPPV in this group of hypercapnic respiratory failure patients.
There has been interest by a few groups in the use of NIPPV in the management of acute asthma24,28,29. Although there have been some isolated successes, we find
the use of heliox (see Chapter 17) more suited than NIPPV for the management of these patients. Many asthmatics, in our experience, in severe acute attacks simply
will not tolerate NIPPV.
Others have successfully used NIPPV as a bridge to transplantation in cystic fibrosis patients4,30. We have also been able to successfully manage some
transplantation candidates without endotracheal intubation for days to weeks with NIPPV. NIPPV is also useful in the management of postoperative, hypercapnic
respiratory failure31–33.
19.3.3 Cardiogenic pulmonary edema

Positive pressure has been shown to be beneficial in the management of acute cardiogenic pulmonary edema because the elevated mean intrathoracic pressures
decrease venous return, reducing left ventricular preload and afterload, improving cardiac output34. In addition, oxygenation is generally improved because of better
ventilationperfusion matching35. In fact, four prospective, randomized controlled trials
Page 314
have demonstrated a decrease in endotracheal intubation rate in patients in acute cardiogenic pulmonary edema receiving mask CPAP versus controls without
CPAP34,36–38. As a result, CPAP has been considered a standard for the management of hypoxemia in acute cardiogenic pulmonary edema4,12.
As NIPPV became more popular, a number of groups have used NIPPV for the management of cardiogenic pulmonary edema. In the largest case series,
Rusterholtz et al.40 successfully applied (avoided intubation) NIPPV to 21 out of 26 patients. All patients successfully treated were hypercarbic (PaCO2 54.2±15
mmHg versus 32±2.1 mmHg failures) and had a lower Creatine phosphokinase (CPK) (176±149 versus 1282±2080 IU/I in the failures) than those failing NIPPV. In
fact, four of the five failures versus two of the 21 successes had acute myocardial infarctions. Three randomized controlled trials have evaluated NIPPV in acute
cardiogenic pulmonary edema39–42. In two of these trials the number of patients developing acute myocardial infarctions was higher than the group receiving NIPPV
versus CPAP40 or versus oxygen therapy with highdose nitrates41. In addition, a recent metaanalysis concluded that there was too little evidence to make conclusions
regarding the use of NIPPV in acute cardiogenic pulmonary edema43. As listed in Table 19.3, CPAP should be used during acute cardiogenic pulmonary edema in all
patients presenting with hypoxemia who are eucapnic. NIPPV should be reserved for patients with hypercarbia. Patients presenting with an acute myocardial infarction
should be considered for endotracheal intubation.
Table 19.3 Use of positive pressure in acute cardiogenic pulmonary edema
• Hypoxemia but eucapnia, CPAP 8–12 cmH2O
• Hypoxemia and hypercarbia, BIPAP

• Acute cardiogenic pulmonary edema associated with myocardial infarction; consider intubation
19.3.4 Hypoxemic respiratory failure

NIPPV has been most successful during hypercarbic respiratory failure. In fact, Wysocki et al.44 showed that the lack of hypercarbia was associated with failure of
NIPPV (Figure 19.6). This makes intuitive sense; ventilation can be expected to have the most impact on patients who require ventilatory assistance not oxygenation
assistance. However, patients with severe hypoxemic experience increased cardiopulmonary work and hypoxemic respiratory failure can be expected over time to
show hypercarbic respiratory failure if the hypoxemia is not corrected.
A single prospective randomized trial has evaluated the use of NIPPV in nonhypercarbic, hypoxemic respiratory failure45. Antonelli et al.45 randomized patients to
NIPPV or endotracheal intubation. Seventeen patients in the endotracheally intubated group (n=32) and 23 in the NIPPV group (n=32) survived. More patients in the
endotracheally intubated group experienced serious complication and had a longer hospital stay. However, this was a singlecenter trial and it is difficult to conclude
from the data presented that the patients randomized to conventional mechanical ventilation required endotracheal intubation at the time of randomization.
More recently, Delclaux et al.46 compared the use of CPAP with oxygen therapy in patients presenting with hypoxemic respiratory failure. They found no
differences between groups in the rate of endotracheal intubation, hospital mortality or intensive care length of stay. They did, however, find differences in the number of
adverse events. More adverse events occurred in the CPAP group (Table 19.4). Of most concern was the fact that four patients in the CPAP group had cardiac
arrests and four other patients developed stress ulcers. No patients in the oxygenonly group developed either complication. They concluded that CPAP, although
improving oxygenation, delayed necessary endotracheal intubation and placed patients at greater risk for complications than simply oxygen therapy.
Page 315
Figure 19.6 Percentage of patients requiring endotracheal intubation, length of ICU stay and inICU mortality between patients treated
with NIPPV (NIPSV) and those treated with conventional therapy in subgroups of patients with (a) initial PaCO2 greater than
45 mmHg (n=17) or (b) initial PaCO2 of 45 mmHg or less (n=24). ***P<0.02; **P<0.04; *P<0.06 between NIPSV and
conventional therapy. From: Wysocki, M., Tric, L, Wolf, M.A., Millet, H. and Herman, B. (1995) Noninvasive pressure support
ventilation in patients with acute respiratory failure: A randomized comparison with conventional therapy. Reproduced with
permission from Chest 107:761–768.
Table 19.4 Adverse events occurring in patients during their intensive care unit stay*
Adverse event Oxygen alone (n=60) Oxygen plus CPAP (n=62) P value between treatment groups
During spontaneous ventilation
Facial skin necrosis 0 2 0.50
Gastric distention 0 1 0.54
Nosocomial pneumonia 1 0 0.97
Cardiac arrest 0 4* 0.14
During mechanical ventilation
Nosocomial pneumonia 4 6 0.74
Sinusitis 1 0 0.99
Pneumothorax 0 1 0.54
Stress ulcer 0 4 0.14
Any adverse event, n (%) 6 (10) 18(29) 0.01
Patients with adverse events, n (%) 5 (8) 14 (23) 0.03
CPAP, continuous positive airway pressure.
*Three patients experienced cardiac arrest at the time of scheduled intubation, in patients receiving CPAP to maintain oxygenation. In another patient, cardiac arrest occurred when CPAP
delivered by a face mask was disconnected to allow nursing care, resulting in a rapid worsening of hypoxemia. From: Delclaux, C, L’Her, E., Alberti, C, et al. (2000) Treatment of acute
hypoxemic nonhypercapnic respiratory insufficiency with continuous positive airway pressure delivered by a face mask. Reproduced with permission from the American Medical
Association. JAMA 289:2352–2360.
Page 316
As a result, it is difficult to recommend the use of NIPPV for hypoxemic respiratory failure, and the use of CPAP may need to be restricted to use in patients with acute
cardiogenic edema who can maintain eucapnia. When either NIPPV or CPAP is used in these patients careful ongoing assessment is required to avoid delays in
endotracheal intubation in those patients where it is required.
19.3.5 Immunocompromised patients

The single group of patients where NIPPV for the management of hypoxemic respiratory failure seems appropriate is immunocompromised patients. The avoidance of
endotracheal intubation in these patients reduces the risk of infection and hemorrhagic complications4. In a number of case series, endotracheal intubation has been
successfully avoided with the use of NIPPV47,48. Patients with acute respiratory failure of various etiologies after solid organ transplantation were randomly assigned by
Antonelli et al.49 to receive NIPPV or standard therapy. Patients treated with NIPPV had a reduced need for endotracheal intubation and a lower mortality. There
was also a trend toward fewer ventilatorassociated pneumonias and a significant reduction in the rate of septic shock. Considering the problems inherent with
endotracheal intubation with this group of patients. NIPPV should be considered as firstline therapy.
19.3.6 Communityacquired infection

Patients with communityacquired infections have been included in some of the case series on the use of NIPPV24,50. However, only one randomized trial recruiting
only patients with communityacquired infections has been performed47. Fiftysix patients with severe communityacquired pneumonias were randomized to receive
NIPPV plus conventional therapy or conventional therapy alone. Patients receiving NIPPV avoided intubation (21 versus 50%, P<0.03) and had a shorter duration of
ICU stay (1.8 versus 6 days, P< 0.04). However, closer examination of the data indicates that the beneficial results of NIPPV occurred only in patients with COPD.
As a result, the use of NIPPV for communityacquired pneumonia appears beneficial only in patients with COPD.
19.3.7 Do not resuscitate/do not intubate

This is a controversial indication for NIPPV. However, a number of case series have demonstrated successful use of NIPPV in this setting51–53. The central issue is an
ethical one, for in patients who choose not to be intubated or resuscitated is NIPPV truly following the wishes of the patient? Patients should fully understand that
NIPPV is being used as a form of life support, although noninvasive. On the other hand, for a terminally ill patient NIPPV may relieve dyspnea, preserve the patient's
autonomy and permit verbal communication between the patient and family4. Indeed, we have seen patients in this category discharged home after successful use of
NIPPV. We believe NIPPV is appropriate in many do not intubate/do not resuscitate patients but appropriate communication with the patient and family regarding
NIPPV should always occur.
19.3.8 Weaning
A number of case series have indicated that NIPPV is a useful bridge to weaning, especially in patients with COPD54–57. Two randomized trials have specifically
addressed this issue58,59. In both of these trials COPD patients receiving conventional mechanical ventilation were screened 48 h after the initiation of mechanical
ventilation. Only those patients failing this initial weaning trial were randomized to either extubation and NIPPV or continued intubation and pressure support weaning.
As noted in Figure 19.7 from the Nava et al. trial58, patients extubated to NIPPV were weaned more quickly than patients remaining intubated. Mean duration of
mechanical ventilation, ICU stay, survival and nosocomial pneumonia rates were all lower in the NIPPV group. In the Girault et al.59 trial no difference in the number of
patients successfully weaned was observed. As expected by protocol, the
Page 317
number of days patients remained intubated was shorter with NIPPV. However, the total days of ventilatory support (7.69±3.79 days versus 16.10±5.24 days
NIPPV) was longer in the NIPPV group.
It is difficult to recommend against the use of NIPPV in the patient who is inadvertently extubated or who develops respiratory failure after extubation. But it is also
difficult to recommend planned extubation of COPD patients who fail weaning trials to initiate NIPPV, as these patients probably failed NIPPV before intubation. We
have attempted to electively extubate COPD patients failing weaning trials but have not been successful maintaining extubation with NIPPV. In our opinion, this is
Figure 19.7 KaplanMeier curves for patients who could not be weaned from mechanical ventilation (defined as weaning failure or
death linked to mechanical ventilation) in two groups. The probability of weaning failure was significantly lower for the
noninvasive ventilation group (P<0.01). The vertical line represents day 21, The continuous line represents the NIPPV group;
the dashed line the invasive pressure support group. From; Nava, S., Ambrosino, N., Clini, E., Prato, M, Orlanodo, G., Vitacca,
M., B Brigada, P., Fraccia, C and Rubinni, F. (1998) Noninvasive mechanical ventilation in t the weaning of patients with
respiratory failure due to chronic obstructive pulmonary disease: A randomized study. Reproduced with permission from the
American College of Physicians. Ann. Intern. Med. 128:721–728.
an indication for NIPPV requiring greater research.
19.4 When should NIPPV be initiated for acute respiratory failure?

Table 19.5 summarizes the clinical settings where noninvasive positive pressure is indicated based on data in the literature and our experience. For hypoxemic
respiratory failure, asthma, and elective extubation for weaning we believe insufficient data are available to support the unquestioned use of NIPPV.
19.5 Complications or problems with NIPPV

NIPPV is not without complications; however, provided airway pressure is not excessive, the complications are not serious. Table 19.6 lists the complications
observed during NIPPV. Of those listed, gastric distention and subsequent regurgitation are the most potentially dangerous, but, as addressed later in the chapter, if
airway pressure is maintained less than 20 cmH2O gastric distention is unlikely. Most of the other problems listed can be corrected with proper equipment (see later
discussion). In addition, the use of artificial skin on the bridge of the nose can reduce the likelihood of skin
Table 19.5 Clinical setting where NIPPV is indicated
• COPD acute exacerbation

• Communityacquired pneumonia in COPD patients
• Acute cardiogenic pulmonary edema presenting with hypercarbia but no evidence of mycocardial infarction
• Acute hypercapnic respiratory failure in:
• Neuromuscular or neurologic disease
• Postoperative patients
• Immunocompromised patients
• Do not resuscitate/do not intubate patients
• Inadvertently extubated patients
• Patients failing extubation
Page 318
Table 19.6 Complication or problems with NIPPV
Problems Solutions
• Air leak Different size or type of mask
• Mask discomfort or facial soreness, skin breakdown Different size or type of mask; artificial skin; lower mask securing pressure
• Eye irritation Reposition mask; different size or type of mask
• Sinus congestion or oronasal drying Humidifier; decongestant
• Epistaxis Humidifier; petroleum jelly
• Gastric insufflation Lower peak pressure; nasogastric tube
breakdown. This may especially be useful in the elderly with poor peripheral perfusion.
19.6 The application of NIPPV

Establishing an NIPPV program is not without problems. It is common that initial results are not as successful as those published in randomized, controlled trials. The
reason for this difference is multifactorial (Table 19.7). Most importantly, all involved in the care of the patient receiving NIPPV must fully understand the benefits and
problems associated with NIPPV. If the therapists and nurses caring for these patients do not understand the benefits associated with NIPPV or how to properly
manage patients and select equipment during NIPPV, their actions (whether conscious or subconscious) can markedly affect the success rate.
Four major factors affect the efficacy of NIPPV and can influence its success: (i) the ventilator; (ii) the facial interphase; (iii) the interaction between the patient and
the clinician applying NIPPV; (iv) the interaction between the patient and the mechanical ventilator.
19.6.1 The mechanical ventilator

Any ventilator ever manufactured can be used for NIPPV. However, for the successful application of NIPPV to patients in acute respiratory failure a number of
ventilator features are desirable (Table 19.8). First, pressuretargeted ventilation should be available. In patients experiencing marked cardiopulmonary stress
Table 19.7 Successful application of NIPPV requires
• Clinician knowledge of the process

• Proper selection of the ventilator
• Proper selection of the mask
• Interaction between patient and clinician
• Interaction between patient and ventilator
Table 19.8 The ideal ventilator for NIPPV use in acute respiratory failure
• Pressuretargeted ventilation available

• Leak compensation
• CO2 rebreathing minimal
• Precise, accurate, and high FIO2 available
• Pressure, volume, and flow waveforms
• Appropriate alarms and monitors
with varying ventilatory demand it has been clearly demonstrated that pressure ventilation more appropriately responds to patient’s flow demand than volume
ventilation60. This does not mean that volume ventilation is always inappropriate, but it does mean that the likelihood for success is greater with pressure than volume
ventilation. Second, in acute respiratory failure visual feedback of patientventilatory synchrony is essential. As with invasive ventilation, knowledge of the patient's
respiratory rate, tidal volume, and coordination with triggering and cycling helps to insure ventilator settings are appropriate. The ventilator must also be able to provide
high and precise FIO2. In addition, methods of monitoring the patient’s status
Page 319
and notifying staff of problems are essential. The ventilator should be able to compensate for leaks and CO2 rebreathing should be minimal. Can ventilators without
these features provide successful NIPPV? Yes, but the likelihood of success is greatly diminished.
19.6.2 ICU ventilators

There is no question that ICU ventilators meet most of the above requirements. Most provide waveforms and monitors, pressure ventilation, precise FIO2 delivery, and
alarms. In addition, the newest generation of ICU ventilators are very responsive to patient demand61. Some of these ventilators (Puritan Bennett 840 and Hamilton
Galileo) allow for adjustment of inspiratory termination criteria during pressure support and most of these new generation ventilators allow for rise time adjustment
during both pressure support and pressure assist/control ventilation62,63. In addition, CO2 rebreathing is not a problem with ICU ventilators. However, the major
problem with the use of ICU ventilators for NIPPV is their inability to compensate for leaks. None of the current generation of ICU ventilators compensates for leaks;
none of them incorporates a NIPPV mode. However, as is obvious from the literature, many trials of NIPPV have been conducted with these ventilators8,21,22,24,45. It
is also important to note that these ventilators impose very little effort when used in the CPAP mode64.
19.6.3 Bilevel pressure ventilators

This group of ventilators was developed out of the need to provide pressure ventilation to select patients requiring nasal CPAP for sleep apnea. In this application, a
very simple approach to ventilation was all that was required. This approach has been shown to work well during the chronic application of NIPPV in the home. In fact
it has been shown that the responsiveness of many of these ventilators is equivalent to or better than that of current ICU ventilators61,62,65. However, the vast majority
of these ventilators are not designed for use in the ICU (Figure 19.8). They do not include alarms or monitors. They are not capable of delivering high or precise FIO2
and it is virtually impossible with most to determine if the patient in acute respiratory failure is in synchrony with the ventilator. Most do not include any indication of
phase of ventilation. The only ventilator in this group specifically designed for ICU use is the Respironics Inc. Vision Ventilator (Figure 19.8). However, it can be
expected that similar type ventilators from other manufacturers will soon be available. The Vision incorporates all of the features for successful application of NIPPV to
patients in acute respiratory failure and would have to be considered the model ventilator for application in this setting.
19.6.4 CO2 rebreathing

A problem with all bilevel pressure ventilators that can affect success in acute respiratory failure is CO2 rebreathing. All of these units have low internal resistance
turbinedriven compressors. In addition, none contains a true exhalation valve. Expiration is via a small orifice near the patient’s face. As a result, total patient exhalation
into the ventilator circuit has the potential of retrograde movement of exhaled gas all the way up to the ventilator66. This can be avoided by setting PEEP equal to or
greater than about 5 cmH2O (Figure 19.9). PEEP maintains a high flow from the ventilator during exhalation, minimizing the likelihood of CO2 rebreathing65.
Furthermore, the use of an isolation (plateau) valve that directs gas out of the circuit prevents rebreathing. In addition, it has been recently demonstrated that full face
masks with the exhalation port located in the mask prevent or markedly reduce CO2 rebreathing67. Consideration of CO2 rebreathing is critical during acute application
with a oronasal or full face mask. Modifications to prevent CO2 rebreathing may be the difference between success and failure in a given patient.
19.6.5 The mask

The most critical factor in the successful application of NIPPV is the mask. If the mask does not fit properly, if the mask is too large, if there
Page 320
Figure 19.8 Bilevel pressure ventilators.
is a large leak near the patient’s eyes, or if the patient finds the mask uncomfortable the likelihood of successful application of NIPPV is markedly diminished. As a
general rule, the smaller the mask the better the fit. If gas leaks from the mask into the patient's eyes, the mask is too big. The top of a properly sized and fitted mask
should rest about a quarter to a third down from the top of the bridge of the nose and fit snugly along the lateral border of the nose. For nasal masks, the bottom of the
mask should sit above the upper lip snug to the nose. With fullface masks the lower border should fit just below the lower lip or down to the chin depending upon
design. In sizing masks, it is useful to use the sizing templates available with many masks (Figure 19.10).
Especially with full face masks, the internal volume of the mask should be considered. A large internal volume can result in CO2 rebreathing. The ResMed oronasal
mask avoids this concern by placing the orifice for exhalation at the top of the mask itself67. As a result, flow to maintain PEEP flushes the mask during exhalation.
No single type of mask will work on all patients, especially during the acute application of NIPPV. Multiple types of both oronasal
Page 321
Figure 19.9 Volume of CO2 inhaled from the ventilator tubing at various IPAP and EPAP settings during BIPAP ventilatory assistance.
WhisperSwivel—standard exhalation valve. Plateau and nonrebreathing valves—prevent retrograde movement of exhaled
gas into the ventilator circuit. *P<0.05 compared with other devices at similar BIPAP settings. From: Ferguson, G.T. and
Gilmartin, M. (1995) CO2 rebreathing during ring BIPAP ventilatory assistance, Reproduced with permission from Am. J.
Respir. Crit. Care Med 151:1126–1132,
and nasal masks should be available and taken to the bedside so that the mask that best conforms to the patient’s anatomy can be selected (Figure 19.11).
During chronic NIPPV either a nasal or oronasal mask can be used. However, when patients are in acute respiratory distress they breathe through their mouths; as a
result, oronasal masks are the masks of choice. Although oronasal masks are less comfortable and more claustrophobic, because they prevent mouth leaks they insure
better ventilation (Table 19.9). A number of companies have chinstraps to prevent mouth leak with nasal masks; however, during acute respiratory failure they do not
work. After the first 12–24 h of NIPPV, when acutely distressed patients are stabilized, many can be switched from oronasal to nasal masks for continued NIPPV.
A fullface shield has recently become avail
Figure 19.10 Template for sizing masks (Respironics Inc.).
Table 19.9 Nasal versus oronasal masks
Nasal Oronasal
Adequacy of ventilation Low High
Claustrophobia Low High
Ability to speak Easy Difficult
Ability to expectorate Easy Difficult
Air leak High Low
Dead space Low High
able from Respironics, Inc. This is a onesizefitsall mask that covers the whole face. No data are currently available on the use of this mask; however, it appears to be
well tolerated by patients in acute failure.
19.6.6 Patientclinician interaction

Successful application of NIPPV especially in the acute setting requires a clinician who understands what the patient is experiencing, is empathetic to the patient’s
needs, and is patient and willing to educate and work with the patient during the initial application. Initiation of NIPPV is more difficult than initiation of invasive
ventilatory support. The clinician needs to first educate the patient, to explain carefully what is about to happen and why. After the mask has been selected it should be
fitted to the patient's face briefly without ventilation to confirm fit. When ventilation is started, PEEP should be set at zero or minimum and inspiratory pressure set at 2–
5 cmH2O. The mask should be held to the patient's face by the clinician or patient before it is strapped in place. The mask should never be strapped until
Page 322
Figure 19.11 Masks used for NIPPV.
the patient is totally accepting and comfortable with the process. The wrong approach is to set the ventilator at 15 cmH2O peak pressure and 5 cmH2O PEEP and on
the first breath strap the mask to the patient’s face. It is very difficult to rapidly acclimate to these pressures. A gradual buildup of pressure over time insures patient
acclimation, elimination of patient’s fears and increases the patient’s confidence in the process.
Throughout the initial application phase (first 1–2 h) an experienced clinician capable of answering all of the patient’s questions and adjusting the ventilator should be
immediately available. During this period, the clinician may have to stay at the bedside talking to the patient, eliminating their fears, and insuring the ventilator is set to
meet the patient’s ventilatory and comfort needs.
19.6.7 Patientventilator synchrony

In the acute application of NIPPV it is impossible to insure patientventilatory synchrony without visual assessment of the inspiratory and expiratory phase of ventilation.
Some of the bilevel pressure ventilators are so quiet and the sound they do make is so similar during inspiration and expiration that it is difficult to tell if the patient is in
phase with the ventilator. The most common causes of dyssynchrony are a lack of coordination between the patient and the ventilator during triggering and cycling of
inspiration, and the delivery of inadequate or excessive flow at the onset of inspiration67. Dyssynchrony during acute respiratory failure is more common with nasal than
oronasal masks, as stressed patients breathe through their mouths (Table 19.10).
Page 323
Table 19.10 Patientventilator synchrony
Inability to trigger to inspiration

• Improper sensitivity setting
• AutoPEEP
• PEEP
• Reduce peak pressure or VT
System leaks
• Too large a mask
• Use oronasal or full face mask
• Change to pressure assist/control if ICU ventilator
Cycling to exhalation
• Increase cycling criteria if available
• Use pressure assist/control
• Set inspiratory time 0.6–0.9 s
Gastric distention
• Peak pressure too high
• Cough or dryness of airway or nose
• Add humidifier
Dyssynchrony with triggering is most commonly a result of autoPEEP68. All of the ICU and bilevel pressure ventilators are very responsive to patient demand and
trigger with small pressure changes and short time delays. However, autoPEEP as with intubated patients can result in an inability of patients to trigger the ventilator
with each inspiratory effort. Whenever the patient’s respiratory rate is greater then the ventilator response rate autoPEEP should be considered69. Applied PEEP can
be adjusted to minimize the difference between central airway and ventilator circuit pressure, and alveolar pressure improving triggering70. This works to decrease the
gradient to trigger only if the autoPEEP is a result of dynamic airway obstruction as seen in COPD patients. The autoPEEP level in COPD patients is increased by
excessive airway pressure and large VT values, particularly if the respiratory rate is high. Dyssynchrony during the transition to exhalation can also increase the level of
autoPEEP.
19.6.8 Transition to exhalation

Termination of the inspiratory phase during pressure support occurs when the patient’s inspiratory flow is equal to the inspiratory flow termination criteria of the
ventilator62. With most bilevel pressure ventilators, the size of the leak is learned and the patient’s desire to exhale is well coordinated with the inspiratory flow
terminating criteria69. This is not true with any other group of currently available ventilators.
During NIPPV lack of coordination between the patient and ventilator during transition from inspiration to expiration with ICU ventilators is common70. This problem
can be approached in a number of ways. First, only oronasal or full face masks should be used with ICU ventilators. Second, if available the inspiratory termination
criteria should be set to a high percentage of the patient’s peak flow. Third, the patient can be switched from pressure support to pressure assist/control (Figure 19.12)
70. The use of pressure assist/control is frequently required even with an oronasal mask in ventilators with a low flow inspiratory termination criteria (PuritanBennett
7200 5 L min–1; Seimens Servo 300 5% of peak flow). When pressure assist/control is used, a careful assessment of the patient’s inspiratory time should be made. In
most adults in acute respiratory failure inspiratory time should be set at about 0.6–0.9 s. Spontaneously breathing patients rarely exceed 1.0 s inspiratory time.
19.6.9 Ventilatory pressure

It is rare that high airway pressure is needed during NIPPV. As NIPPV is primarily used in patients with hypercarbic not hypoxemic respiratory failure, peak pressures
can easily be kept at 20 cmH2O or less. If airway pressures are allowed to exceed 20 cmH2O gastric distention becomes a concern and may result in dyssynchrony.
However, gastric distention of clinical significance is quite rare if peak pressures are kept below 20 cmH2O. Gastric opening pressure is about 20–25 cmH2O. If a
nasogastric tube is necessary it is unlikely that NIPPV will be successful. PEEP levels are adjusted to insure inspiratory trigger synchrony. Rarely is more than 7 cmH2O
PEEP required. Ventilating pressure (pressure support
Page 324
Figure 19.12 Representative flow, esophageal pressure (Pes) and airway pressure (Paw) during NIPPV with pressure assist/control
(left side) and during pressure support (right side) using an ICU ventilator. In both, a large mask leak is present. From;
Calderini, E., Confalonieri, M., Puccio, P.G., Francavilla, N., Stella, L, and Gregoretti, C. (1999) Patientventilator assynchrony
during noninvasive ventilation: the role of expiratory trigger. Intensive Care Med. 25:662–667. Reproduced with permission
from SpringerVerlag.
or pressure control level) exceeding 12–15 cmH2O is rarely needed; in fact, many COPD patients can be adequately ventilated with a pressure support level of 8–10
cmH2O.
19.6.10 Humidification
Many patients require humidification during the acute application of NIPPV. This is particularly true if a high concentration of oxygen is required. In most patients an
unheated, passover humidifier is sufficient. If a patient persistently coughs after the start of NIPPV, humidify the inspired gas.
19.7 Summary
In acute exacerbations of COPD NIPPV should be considered the treatment of choice. NIPPV has also been shown to be effective in other forms of acute
hypercarbic respiratory failure and postextubation failure (Table 19.5). However, NIPPV should only be used cautiously in hypoxemic respiratory failure. In fact, we
have found limited success with this application. Numerous factors contribute to a successful NIPPV program. Education of staff and appropriate equipment selection
are most
Page 325
important. The learning curve for NIPPV is steep once all the pieces are in place but it does require a defined program be established.
References
1. Woollam, C.H.M. (1976) The development of apparatus for intermittent negative pressure respiration. Anaesthesia 31:537–547.
2. Drinker, P. and Shaw, L.A. (1929) An apparatus for the prolonged administration of artificial respiration. I. Design for adults and children. J. Clin. Invest 7:229–247.
3. Hodes, H.L. (1955) Treatment of respiratory difficulty in poliomyelitis. In: Poliomyelitis Papers and Discussions Presented at the Third International
Poliomyeliytis Conference. Lippincott, Philadelphia, PA, pp. 91–113.
4. Mehta, S. and Hill, N.S. (2001) Noninvasive ventilation. Am. J. Respir. Crit. Care Med. 163: 540–577.
5. Barach, A.L., Martin, J. and Eckman, M. (1938) Positive pressure respiration and its application to the treatment of acute pulmonary edema. Ann. Intern. Med.
12:754–795.
6. Alba, A., Khan, A. and Lee, M. (1984) Mouth IPPV for sleep. Rehabil. Gazette 24:47–49.
7. Rideau, Y., Gatin, G., Bach, J. and Gines, G. (1983) Prolongation of life in Ducheme’s muscular dystrophy. Acta Neurol. Belg. 5:118–124.
8. Carrey, Z., Gotfried, S.B. and Levy, R.D. (1990) Ventilatory muscle support in respiratory failure with nasal positive pressure ventilation. Chest 97:150–158.
9. Brochard, L., Isabey, D., Piquet, J., Amaro, P., Mancebo, J. and Messadi, A.A. (1990) Reversal of acute exacerbations of chronic obstructive lung disease by
inspiratory assistance with a face mask. N. Engl. J. Med. 95:865–870.
10. Meduri, G.U., Conoscenti, C.C., Menashe, P. and Nair, S. (1989) Noninvasive face mask ventilation in patients with acute respiratory failure. Chest 95:865–870.
11. Elliott, M.W., Steven, M.H., Phillips, G.D. and Branthwaite, M.A. (1990) Noninvasive mechanical ventilation for acute respiratory failure. Br. Med. J.
300:358–360.
12. International Censensus Conference (2001) International Consensus Conference in Intensive Care Medicine: Noninvasive Positive Pressure Ventilation in Acute
Respiratory Failure. Am. J. Respir. Crit. Care Med. 163:283–291.
13. Bott, J., Carroll, M.P., Conway, J.H., et al. (1993) Randomized controlled trial of nasal ventilation in acute ventilatory failure due to chronic obstructive airways
disease. Lancet 341: 1555–1557.
14. Kramer, N., Meyer, T.J., Meharg, J., Cece, R.D. and Hill, N.S. (1995) Randomized, prospective trial of noninvasive positive pressure ventilation in acute
respiratory failure. Am. J. Respir. Crit. Care Med. 151:1799–1806.
15. Brochard, L., Mancebo, J., Wysocki, M., et al. (1995) Noninvasive ventilation for acute exacerbations of chronic obstructive pulmonary disease. N. Engl J. Med.
333:817–819.
16. Angus, R.M., Ahmed, A.A., Fenwick, L.J. and Peacock, A.J. (1996) Comparison of the acute effects on gas exchange of nasal ventilation and doxapram in
exacerbations of chronic obstructive pulmonary disease. Thorax 51: 1048–1050.
17. Celikel, T., Sungur, M., Ceyhan, B. and Karakurt, S. (1998) Comparison of noninvasive positive pressure ventilation with standard medical therapy in hypercapnic
acute respiratory failure. Chest 114:1636–1642.
18. Plant, P.K., Owen, J.L. and Elliott, M.W. (2000) Early use of noninvasive ventilation for acute exacerbations of chronic obstructive pulmonary disease on general
respiratory wards: A multicenter randomized controlled trial. Lancet 355:1931–1935.
19. Barbe, F., Togores, B., Rubi, M., Pons, S., Maimo, A. and Agusti, A.G.N. (1996) Noninvasive ventilatory support does not facilitate recovery from acute
respiratory failure in chronic obstructive pulmonary disease. Eur. Respir. J. 9:1240–1245.
20. Keenan, S.P., Kernerman, P.D., Cook, D.J., Martin, C.M., McCormack, D. and Sibbald, W.J. (1997) The effect of noninvasive positive pressure ventilation
on mortality in patients admitted with acute respiratory failure: A metaanalysis. Crit. Care Med. 25:1685–1692.
21. Guerin, C., Giard, R., Chemorin, C., De Varox, R. and Fournier, G. (1997) Facial mask, noninvasive mechanical ventilation reduces the incidence of nosocomial
pneumonia; a prospective epidemiological survey from a single ICU. Intensive Care Med. 23:1024–1032.
22. Girou, E., Schortgen, F., Delclaux, C., BrunBuisson, C., Blot, F., Lefort, Y., Lemaire, F. and Brochard, L. (2000) Association of noninvasive ventilation with
nosocomial infection and survival in critically ill patients. JAMA 284: 2361–2367.
23. Anton, A., Guell, R., Gomez, J., Serrano, J., Castellano, A., Carrasco, J.L. and Sanchis, J.
Page 326
(2000) Predicting the result of non invasive ventilation in severe acute exacerbations of patients with chronic airflow obstruction. Chest 117:828–836.
24. Meduri, G.U., Turner, R.E., AbouShala, N., Wunderink, R. and Tolley, E. (1996) Noninvasive positive pressure ventilation via face mask. Chest 109:179–193.
25. Bach, J.R. (1996) Conventional approaches to managing neuromuscular ventilatory failure. In: Pulmonary Rehabilitation: the Obstructive and Paralytic
Conditions (ed. J.R.Bach). Henley & Belfus, Philadelphia, PA, pp. 285–301.
26. Finlay, G., Conconnon, D. and McDonell, T.J. (1995) Treatment of respiratory failure due to kyphoscoliosis with nasal intermittent positive pressure ventilation
(NIPPV). J. Med. Sci. 164: 28–30.
27. Bach, J.R., Ishikawa, Y. and Kim, H. (1997) Prevention of pulmonary morbidity for patients with Duchenne muscular dystrophy. Chest 112: 1024–1028.
28. Meduri, G.U., Cook, T.R., Turner, R.E., Cohen, M. and Leeper, K.V. (1996) Noninvasive positive pressure ventilation in status asthmaticus. Chest 110:767–
774.
29. Levy, B.D., Kitch, B. and Fanta, C.H. (1998) Medical and ventilatory management of status asthmaticus. Intensive Care Med. 24:105–117.
30. Hodson, M.E., Madden, B.P., Steven, M.H., Tsang, V.T. and Yacoub, M.H. (1991) Noninvasive mechanical ventilation for cystic fibrosis patients—a potential
bridge to transplantation. Eur. Respir. J. 4:524–527.
31. Pennock, B.E., Kaplan, P.D., Carlin, B.W., Sabangan, J.S. and Magovern, J.A. (1991) Pressure support ventilation with a simplified ventilatory support system
administered with a nasal mask in patients with respiratory failure. Chest 100:1371–1376.
32. Pennock, B.E., Crawshaw, L. and Kaplan, P.D. (1994) Noninvasive nasal mask ventilation for acute respiratory failure. Chest 105:441–444.
33. Gust, R., Gottschalk, A., Schmide, H., Bottiger, B., Bohrer, H. and Martin, E. (1996) Effects of continuous (CPAP) and bilevel positive airway pressure
(BIPAP) on extravascular lung water after extubation of the trachea in patients following coronary artery bypass grafting. Intensive Care Med. 22:1345–1350.
34. Rasanen, J., Heikkila, J., Downs, J., Nikki, P., Vaisanen, I. and Viitanen, A. (1985) Continuous positive airway pressure by face mask in acute cardiogenic
pulmonary edema. Am. J. Cardiol. 55:296–300.
35. Vaisanen, I.T. and Rasanen, J. (1987) Continuous positive airway pressure and supplemental oxygen in the treatment of cardiogenic pulmonary edema. Chest
92:481–485.
36. Lin, M. and Chiang, H. (1991) The efficacy of early continuous positive airway pressure therapy in patients with acute cardiogenic pulmonary edema. J. Formos.
Med. Assoc. 90: 736–743.
37. Bersten, A.D., Holt, A.W., Vedig, A.E., Skowronski, G.A. and Baggolery, CJ. (1991) Treatment of severe cardiogenic pulmonary edema with continuous
positive airway pressure delivered by face mask. N. Engl J. Med. 325: 1825–1830.
38. Lin, M., Yang, Y., Chiany, H., Chang, M., Chainy, B.N. and Chitlin, M.D. (1995) Reappraisal of continuous positive airway pressure therapy in acute
cardiogenic pulmonary edema: Shortterm results and longterm followup. Chest 107:1379–1386.
39. Mehta, S., Jay, G.D., Woolard, R.H., Hipona, R.A., Connelly, E.M., Cimini, D.M., Drinkwine, J.H. and Hill, N.S. (1997) Randomized prospective trial of
bilevel versus continuous positive airway pressure in acute pulmonary edema. Crit. Care Med. 25:620–628.
40. Rusterholtz, T., Kempf, J., Berton, C., Gayol, S., Tournoud, C., Zaehringer, M., Jaeger, A. and Sauder, P. (1999) Noninvasive pressure support ventilation
(NIPSV) with face mask in patients with acute cardiogenic pulmonary edema (ACPE). Intensive Care Med. 25:21–28.
41. Masip, J., Betlese, A.J., Vecilla, F., Canizares, R., Padro, J., Paz, M.A., deOtero, J. and Ballus, J. (2000) Noninvasive pressuresupport ventilation versus
conventional oxygen therapy in acute pulmonary edema: A randomized trial. Lancet 356:2126–2132.
42. Sharon, A., Shpiner, I., Kaluski, E., et al. (2000) Highdose intravenous isosorbidedinitrate is safer and better than BIPAP ventilation combined with conventional
treatment for severe pulmonary edema. J. Am. Coll. Cardiol. 36:832–837.
43. Pang, D., Keenan, S.P., Cook, D.I. and Sibbald, W.J. (1998) The effect of positive pressure airway support on mortality and the need for intubation in cardiogenic
pulmonary edema. Chest 114:1185–1192.
44. Wysocki, M., Tric, L., Wolf, M.A., Millet, H. and Herman, B. (1995) Noninvasive pressure support ventilation in patients with acute respiratory failure: A
randomized comparison with conventional therapy. Chest 107:761–768.
45. Antonelli, M., Conti, G., Rocco, M., Bufi, M., DeBlasi, R.A., Vivino, G., Gasparetto, A. and Meduri, G.U. (1998) A comparison of noninvasive positive
pressure ventilation and conven
Page 327
tional mechanical ventilation in patients with acute respiratory failure. N. Engl J. Med. 339: 429–435.
46. Delclaux, C., L’Her, E., Alberti, C., et al. (2000) Treatment of acute hypoxemic nonhyperapnic respiratory insufficiency with continuous positive airway pressure
delivered by a face mask. JAMA 289:2352–2360.
47. Ambrosino, N., Rubini, F., Callegari, G., Nava, S. and Fracchia, C. (1994) Noninvasive mechanical ventilation in the treatment of acute respiratory failure due to
infectious complications of lung transplantation. Monaldi Arch. Chest Dis. 49:311–314.
48. Conti, G., Marino, P., Cogliati, A., Dell’Utri, D., Lappa, A., Rosa, G. and Gasparetto, A. (1998) Noninvasive ventilation for the treatment of acute respiratory
failure in patients with hematologic malignancies: A pilot study. Intensive Care Med. 24:1283–1288.
49. Antonelli, M., Conti, C., Bufi, M., Costa, M.G., Lappa, A., Rocco, M., Gasparetto, M. and Meduri, G.U. (2000) Noninvasive ventilation for treatment of
acute respiratory failure in patients undergoing solid organ transplantation. JAMA 283:235–241.
50. Confalonieri, M., Potena, A., Carbone, G., Della Porta, R., Tolley, E.A. and Meduri, G.U. (1999) Acute respiratory failure in patients with severe community
acquired pneumonia. Am. J. Respir. Crit. Care Med. 160:1585–1591.
51. Benhamon, D., Girault, C., Faure, C., Portier, F. and Muir, J.F. (1992) Nasal mask ventilation in acute respiratory failure: Experience in elderly patients. Chest
102:912–917.
52. Meduri, G.U., Fox, R.C., AbouShala, N., Leeper, K.V. and Wunderink, R.G. (1994) Noninvasive mechanical ventilation via face mask in patient with acute
respiratory failure who refused endotracheal intubation. Crit. Care Med. 22:1584–1590.
53. Benhamon, D., Muir, J.F. and Melen, B. (1998) Mechanical ventilation in elderly patients. Monaldi Arch. Chest Dis. 53:547–551.
54. Udwadia, Z.F., Santis, G.K., Stevan, M.H. and Simonds, A.K. (1992) Nasal ventilation to facilitate weaning in patients with chronic respiratory insufficiency.
Thorax 47:715–718.
55. Goodenberger, D.M., Couser, J. and May, J.J. (1992) Successful discontinuation of ventilation via tracheostomy by substitution of nasal positive pressure
ventilation. Chest 102:1277–1279.
56. Restrick, L.J., Scott, A.D., Ward, E.M., French, R.O., Cornwell, W.E. and Wedjicha, J.A. (1993) Nasal intermittent positive pressure ventilation in weaning
intubated patients with chronic respiratory disease from assisted intermittent positive pressure ventilation. Respir. Med. 87:199–204.
57. Gregoretti, C., Beltrame, F., Lucangelo, U., Burbi, L., Conti, G., Turello, M. and Gregori, D. (1998) Physiologic evaluation of noninvasive pressure support
ventilation in trauma patients with acute respiratory failure. Intensive Care Med. 24:795–790.
58. Nava, S., Ambrosino, N., Clini, E., Prato, M., Orlando, G., Vitacca, M., Brigada, P., Fraccia, C. and Rubinni, F. (1998) Noninvasive mechanical ventilation
in the weaning of patients with respiratory failure due to chronic obstructive pulmonary disease: A randomized study. Ann. Intern. Med. 128:721–728.
59. Girault, C., Daudenthun, I., Chevron, V., Tamion, F., Leroy, J. and Bonmarchand, G. (1999) Noninvasive ventilation as a systematic extubation and weaning
technique in acuteonchronic respiratory failure. Am. J. Respir. Crit. Care Med. 160:86–92.
60. MacIntyre, N.R., McConnell, R., Cheng, K.C. and Sane, A. (1997) Patientventilator flow dyssynchrony: flowlimited versus pressurelimited breaths. Crit. Care
Med. 25:1671–1677.
61. Williams, P., Muelver, M., Kratohvil, J., Ritz, R., Hess, D.R. and Kacmarek, R.M. (2000) Pressure support and pressure assist/control: are there differences?
An evaluation of the newest intensive care ventilators. Respir. Care 45: 1169–1181.
62. Chatmongkolchart, S., Williams, P., Hess, D. and Kacmarek, R.M. (2001) Evaluation of inspiratory rise time and inspiratory termination criteria in new
generation mechanical ventilators: A lung model study. Respir. Care. 46: 666–678.
63. Chalburn, R.L. (1999) Mechanical ventilation. In: Respiratory Care Equipment (eds R.D.Branson, D.R.Hess and R.L.Chatburn). Lippincott, Williams and Wilkins,
New York, pp. 395–527.
64. Takeuchi, M., Williams, P., Hess, D. and Kacmarek, R.M. (2002) Continuous positive airway pressure in new generation mechanical ventilation: A lung model
study. Anesthesiology 96: 162–172.
65. Bunburaphong, T., Imanaka, H., Nishimura, M., Hess, D. and Kacmarek, R.M. (1997) Performance characteristics of bilevel pressure ventilators. Chest
111:1050–1058.
66. Ferguson, G.T. and Gilmartin, M. (1995) CO2 rebreathing during BIPAP ventilatory assistance. Am. J. Respir. Crit. Care Med. 151:1126–1132.
67. Schettino, G., Chatmongkolchart, S., Hess, D. and Kacmarek, R.M. (2001) CO2 elimination during BIPAP ventilation: Effects of mask design
Page 328
and PEEP (abstract). Am. J. Respir. Crit. Care Med. 163:A679.

68. Kacmarek, R.M. (1999) NIPPV: Patientventilator synchrony, the difference between success and failure? Intensive Care Med. 25:645–647.
69. Hess, D.R. and Kacmarek, R.M. (1999) Noninvasive ventilation. In: Respiratory Care Equipment, (eds R.D.Branson, D.R.Hess and R.L.Chatburn). Lippincott,
Williams and Wilkins, New York, pp. 593–613.
70. Calderini, E., Confalonieri, M., Puccio, P.G., Francavilla, N., Stella, L. and Gregoretti, C. (1999) Patientventilator assynchrony during noninvasive ventilation:
the role of expiratory trigger. Intensive Care Med. 25:662–667.
Page 329
Chapter 20
Postoperative care of the cardiac surgical patient
Contents
20.1 Introduction
20.2 Delivery of the patient to the ICU
20.3 Respiratory management
20.4 Cardiovascular support
20.5 Postoperative complications
20.6 Discharge criteria
20.8 Summary
Page 330
20.1 Introduction
As cardiac surgery is a highvolume procedure, postoperative patients account for a substantial percentage of intensive care admissions at many hospitals. The
postoperative care of uncomplicated cardiac surgical patients primarily involves:
• monitoring and control of cardiac output;
• control of high or low blood pressure;
• recovery from anesthesia and hypothermia while controlling the stress response;
• respiratory management and weaning from mechanical ventilation;
• adequate sedation and pain relief to facilitate patient comfort and the above goals;
• monitoring for adverse events that can be prevented or ameliorated.
20.2 Delivery of the patient to the ICU
Transport of a critically ill patient is associated with acute physiologic changes and the potential for mishap1. The intensity of monitoring may be inadvertently relaxed
while attention is focused on physically moving the patient and associated infusion pumps and monitors. Problems commonly encountered during transport range from
sudden hypotension as a result of vasodilation to abrupt awakening with extreme hypertension and the risk for serious bleeding. Marked changes in blood pressure can
be minimized with volume loading of the patient, maintaining anesthesia during transport, and assuring all pumps supplying vasoactive medication remain connected and
functioning.
The anesthesiologist and surgeon have learned valuable information about a particular patient’s response to fluid and vasoactive agents from direct observation and
possibly echocardiography in the operating room. This information plus data on the dose and timing of anesthetic agents and antibiotics, blood product and fluid
administration, and expected postoperative course should be communicated to the intensive care team. Table 20.1 gives a list of important concerns at the time of ICU
arrival.
20.3 Respiratory management
Immediate respiratory system changes include up to a 40% decrease in vital capacity, total lung capacity, inspiratory capacity, and functional residual capacity2.
Cardiopulmonary bypass (CPB) alters pulmonary function, and the severity of pulmonary edema correlates with the length of bypass3. Atelectasis, which occurs in up
to 90% of patients, causes increased work of breathing. Maintaining the lungs at 5 cmH2O of CPAP with room air during CBP improves oxygenation and lowers shunt
fraction4. Direct trauma to the lungs or pleura may occur during surgical retraction or insertion of chest tubes. Pain and splinting cause the patient to limit lung expansion,
and reduce the ability to cough and clear secretions. Goals for respiratory management are outlined in Table 20.2.
20.3.1 Mechanical ventilation

Conventional practice has been to institute intermittent mandatory ventilation (IMV) at relatively high tidal volumes (8–10 cm3 kg–1) and a low respiratory rate (8–10)
with 60% or higher FiO2 and low levels (5 cm) of PEEP. If the patient is spontaneously breathing at ICU arrival, it may be possible to proceed directly to pressure
support ventilation with levels of 5–15 cmH2O. With ‘fasttrack’ protocols, many patients are extubated in the operating room or shortly after ICU arrival.
Assist/control or IMV ventilation are more appropriate for heavily narcotized patients, until spontaneous breathing occurs. In the event of difficulty with oxygenation,
lungprotective ventilation strategies with lower tidal volumes (6–8 cm3 kg–1), and higher levels of PEEP may be needed. Higher airway pressures may necessitate fluid
loading to maintain cardiac output. In patients with airflow obstruction, occult positive endexpiratory pressure (‘AutoPEEP’) is a cause of hypotension and low
cardiac output (see Chapters 13 and 17).
Page 331
Table 20.1 ICU admission concerns
Preparation
Ventilator set to Vt 8–10 cm3 kg–1 with healthy lungs; rate 8–10; FiO2 0.60; PEEP 2–5 cm
Bedside monitor ready for ECG, CVP, Aline, ± PA or LA pressures
Infusion pumps for inotropes, vasopressors, and vasodilators
Defibrillator, resuscitation cart, surgical ‘bleeder’ tray available in ICU
On arrival in ICU
Connect monitors and ventilator
Auscultate breath sounds and observe chest excursion
Check pulse oximetry and endtidal CO2 if available
Zero pressure transducers, and correlate with cuff pressure
Determine cardiac output if applicable
Check function of IABP and/or assist devices if applicable
Check function of chest tubes and amount of drainage
Order any necessary laboratory studies and CXR
Information to be obtained from operating room team
Patient’s name, age, gender, and brief history
Operation performed and any major problems encountered
CPB and circulatory arrest times
Optimum filling pressures as determined in OR
Current drug infusions and titration plans
Pacemaker and antiarrhythmic information
Anesthetics given and plans for awakening (‘fasttrack’)
Fluids and blood products given; urine output
Assessment of hemostasis and heparin reversal
Last hematocrit, potassium, and antibiotic dose in OR
Availability of blood products and surgical salvage
ECG, electrocardiography; CVP, central venous pressure; ICU, intensive care unit; PA, pulmonary artery; LA, left atrial; IABP, intraaortic balloon pump; OR, operating room; CXR, chest
radiograph.
Table 20.2 Respiratory management
Early Later
Goals Complete O2 saturation of hemoglobin Extubation with adequate reserve
Normocarbia Restoration of FRC

Minimization of ventilation’s hemodynamic effects
Assessment ABGs Spontaneous tidal volume and f/Vt rate
Cardiac output and filling pressures
CXR Vital capacity and MIP
ABGs, CXR, fluid balance
Chest tube output
Interventions Changes in rate and tidal volume Control of secretions
Addition or subtraction of PEEP Upright position
Sedation and neuromuscular blockade Incentive spirometry
Lungprotective ventilator strategy Ambulation
ABGs, arterial blood gases; CXR, chest radiograph; FRC, functional residual capacity; MIP, maximal inspiratory pressure. f/Vt, frequency to total volume ratio; PEEP, positive end
expiratory pressure.
Page 332
20.3.2 Timing of extubation

Factors that enter into the timing of extubation include the total dose and patient’s ability to metabolize or eliminate anesthetic agents; whether hypothermia has occurred
and has been corrected, hemodynamic stability, and the likelihood of return to the operating room for tamponade or control of bleeding (Table 20.3). Until the early
1990s, overnight mechanical ventilation was accepted practice for the cardiac surgical patient, largely because of habit, staffing considerations, and a belief that it was
beneficial to the patient’s respiratory and cardiovascular function. Shorteracting anesthetic agents now make it possible to provide a nearly stressfree anesthetic
without the need for prolonged postoperative ventilation5. Documented advantages of early endotracheal extubation after cardiac surgery include improved left
ventricular performance, decreased length of intensive care stay, lower costs, diminished requirement for postoperative medication, lessened cardiopulmonary
morbidity, and improved patient comfort6–8. In the otherwise stable patient, the ability to maintain arterial pH greater than 7.35 while the IMV rate is decreased to
CPAP is a convenient and easily applied test for ability to sustain spontaneous respiration9.
20.3.3 Longterm ventilation

About 5% of patients will require mechanical ventilation for longer than 72 h following coronary artery bypass grafting (CABG); the corresponding figure for valve
surgery patients is only slightly higher. The odds of requiring prolonged ventilation increase with preoperative presence of congestive heart failure, azotemia, valvular
heart disease, advanced age, diffuse vascular disease, and with reoperation and emergency procedures10. Early enteral feeding or parenteral nutrition should be
considered so that malnutrition does not occur as a result of ventilator dependence. Chapter 18 has general advice for the patient with difficulty in weaning.
20.3.4 Common factors limiting removal of mechanical ventilatory support

These factors in the cardiac surgical patient are sepsis, overfeeding11 renal failure with fluid overload, marked neurologic dysfunction, and unstable hemodynamic
status. Acute left ventricular dysfunction occurs in patients with COPD during the shift from mechanical to spontaneous ventilation12. Careful attention to fluid balance
and aggressive diuresis or use of ultrafiltration13 ameliorates the abrupt rise in left ventricular filling pressure and allows
Table 20.3 Extubation criteria
Neurologic Awake, following simple commands, able to protect airway

Neuromuscular blockade and opiates fully dissipated
Cardiac Stable; not on IABP, index >2.2 min–1 m–2
MAP >70; no serious arrhythmias
Respiratory Acceptable CXR, ABGs (pH ≥7.35) on minimal vent support
Secretions minimal; cough and gag reflexes present
Comfortable on CPAP or Tpiece with spontaneous rate ≤24 and/or f/Vt ratio <100
Optional: MIP <–25; Vt >5 cm3 kg–1. Vc >10 cm3 kg–1
Renal Diuresing well; not markedly edematous
Urine output >0.5 cm3 kg–1 min–1
Hematologic Chest tube drainage <50 cm3 h–1
Temperature Temperature >36.0°C
IABR intraaortic balloon pump; MAP, mean arterial pressure; CXR, chest radiograph; ABGs, arterial blood gases; MIP, maximum inspiratory pressure; F, respiratory rate; Vt, tidal volume;
Vc, vital capacity; CPAP, continuous positive airway pressure.
Page 333
weaning from ventilation to proceed. It is not usual to find patients ‘unweanable’ until fluid removal reduces body weight to several kilograms below the preoperative
value.
20.3.5 Diaphragmatic paralysis

This may complicate any thoracic procedure, but is more common in reoperative patients because of difficulty in identifying the phrenic nerve in fibrotic pericardial
tissue. The incidence of diaphragmatic paralysis has not been well characterized; temporary diaphragmatic weakness occurs in about 4% of patients. Diaphragmatic
paralysis should be suspected whenever a patient fails to wean from mechanical ventilation, and can be clinically identified by paradoxical movement of the diaphragm
during inspiration, and by comparing vital capacity and tidal volumes in the supine and seated positions. Differences in supine and seated vital capacity of more than 10–
15% should prompt fluoroscopic examination of the diaphragm (‘sniff’ test). Measuring gastric pressure during spontaneous breathing also is useful. A negative
inspiratory gastric pressure indicates diaphragmatic paralysis. Transient diaphragmatic paralysis may occur secondary to cold injury to the phrenic nerve14. Less often,
the phrenic nerve is injured or transected during dissection of the internal mammary arteries or in mobilizing the heart in reoperative patients. Other causes of apparent
diaphragmatic weakness include noncompliant lung, muscle weakness and wasting, hypophosphatemia, and hypomagnesemia. Longterm management of
diaphragmatic paralysis is facilitated by performing a tracheostomy, which allows better secretion control and use of intermittent ventilation.
20.4 Cardiovascular support

There are four immediate cardiovascular goals:
• maintain appropriate perfusion pressure (see Sections 20.4.1–20.4.3);
• deliver adequate oxygen to the brain and other tissues (Sections 20.4.5–20.4.7);
• minimize oxygen demand to the available supply (Sections 20.4.8 and 20.5.3);
• maintain a stable cardiac rhythm (Sections 20.4.9).
20.4.1 Perfusion pressure

This must be adequate to supply the brain, kidney, visceral organs and ultimately all tissues, and also must be high enough that newly placed bypass grafts remain
patent, without becoming so high that leaks occur at freshly placed anastamoses. Empirically, a mean arterial pressure of more than 70 mmHg is considered adequate,
but elderly patients, and those with renal or cerebrovascular stenoses may require higher pressures.
20.4.2 Postoperative hypotension

This is commonly precipitated by hypovolemia, decreased systemic vascular resistance, or insufficient cardiac output. Hypovolemia can result from diuresis (mannitol in
pump prime), release of vasoactive mediators during cardiopulmonary bypass, or blood loss. Volume status is not always easy to assess, even when a pulmonary artery
(SwanGanz) catheter is available. Ventricular compliance may change during the first several hours after bypass15, altering the expected relationship between filling
pressures and ventricular filling. Observation of the response to fluid challenge is the best measurement of optimal filling pressure for an individual patient.
Echocardiography can be used to determine optimal ventricular loading and to help identify new onset cardiac tamponade or outflow obstruction.
Routine use of calcium to treat hypotension and low output following cardiac surgery is controversial. Calcium inhibits the cardiotonic action of betaadrenergic
agonists such as dobutamine, although not the effects of phosphodiesterase inhibitors16. Hyperamylasemia, a marker of pancreatic cellular injury, is associated with
calcium chloride therapy17. Bolus administration of calcium reduces internal mammary artery (IMA) flow and can potentially trigger vasospasm18. On the other hand,
low ionized calcium affects both cardiac
Page 334
output and systemic resistance, and repletion of hypocalcemia is associated with clinical improvement in these variables.
20.4.3 Postoperative hypertension

In patients with normal or only mild left ventricular impairment and normal volume status, blood pressure changes are influenced more by changes in systemic vascular
resistance than by cardiac output19. Systemic vascular resistance may initially be low following CPB, but typically rises during the first few postoperative hours, and can
later require vasodilator therapy. Nitroglycerine and nitroprusside are the most commonly used vasodilators; however, other agents such as nicardipine, labetalol,
hydralazine and prazosin are alternatives (see Chapter 9). The vasodilating properties of propofol may also be useful in this setting20. Nitroprusside is widely used
because it can be titrated in response to sudden changes in preload and afterload. Labetalol is associated with significant reductions in heart rate and cardiac index,
whereas sodium nitroprusside causes significant increases in both21. When cardiac output is borderline (<2.3 1 min–1 m–2) sodium nitroprusside is the agent of choice.
Labetalol is preferred in situations where the circulation is hyperdynamic. Esmolol is also effective at reducing blood pressure by depressing a hyperdynamic circulation
after cardiac surgery, and its halflife is sufficiently short that stopping the infusion of the drug will quickly reverse any adverse effects.
Other causes of postoperative hypertension include hypoxia, hypercarbia, and hypothermia with shivering, visceral distention, and inadequate sedation, particularly in
the presence of residual neuromuscular blockade. In contrast to traditional highdose opioid anesthesia, ‘fasttrack’ regimens do not automatically provide the patient
with residual analgesia, requiring specific attention to pain and discomfort. Cardiopulmonary bypass results in variable increases in circulating epinephrine and
norepinephrine levels, which correlate with blood pressure changes and may be the result of sympathetic overdrive.
20.4.4 Assessing cardiac output

Acute myocardial dysfunction occurs in nearly all patients, with the nadir of cardiac function occurring 2–6 h after cardiopulmonary bypass15. Clinical indicators such as
capillary refill and pedal pulses are unreliable markers for peripheral perfusion after cardiac surgery. Better evidence of adequate perfusion can be obtained by
monitoring mixed venous oxygenation, and the generation of lactic acid. Even these markers may be misleading, particularly in the ‘washout’ period following
cardiopulmonary bypass, where elevated lactic acid may represent improved perfusion and clearance of lactate from the periphery. A measured cardiac index of more
than 2.2 1 min–1 m–2 is generally acceptable; higher cardiac outputs are preferred.
20.4.5 Causes of low cardiac output

In the early postoperative period these include hypovolemia, myocardial dysfunction, and cardiac tamponade. Unless specific blood component therapy is indicated,
volume expansion can be achieved with electrolyte solutions. Albumin and hetastarch are more expensive than electrolyte solutions, but will remain in the intravascular
compartment longer than crystalloids. If ventricular filling pressures appear adequate (CVP or PCWP between 10 and 18 cmH2O), further increases in cardiac output
will require inotropic agents, reductions in afterload, or a combination of the two (see Chapter 8). Right heart failure may occur as the result of prolonged ischemic
arrest, and is more common following mitral valve replacement and cardiac transplantation than postCABG. Pulmonary hypertension and right ventricular failure will
limit left ventricular filling and cardiac output. Unfortunately, there are no intravenous agents that specifically dilate the pulmonary vasculature without causing systemic
vasodilation. Agents that are frequently helpful with right ventricular failure include catecholam ines with dilating action (dobutamine, isoproterenol) and
phosphodiesterase inhibitors such as milrirone. Diastolic dysfunction (failure of the heart to
Page 335
relax and fill properly) frequently responds to milrinone.
20.4.6 Mechanical cardiac support

This is needed when low cardiac output persists despite adequate volume loading and administration of inotropes and dilators. Intraaortic balloon counterpulsation is
indicated for low cardiac output states, cardiogenic shock, failure to wean from cardiopulmonary bypass, and prophylactic perioperative support in the highrisk
patient. The intraaortic balloon pump (IABP) is capable of increasing cardiac output by about 0.7 l min–1, and works best when the cardiac rhythm is stable, preferably
sinus. Severe cardiac dysfunction occasionally requires a true pump capable of replacing the function of the heart until recovery occurs. A number of left and right
ventricular assist devices are available, and fall into two major categories: pulsatile (e.g. Abiomed, Heartmate) and nonpulsatile (e.g. Biomedicus centrifugal pump).
Generally, ventricular assist is used to achieve a cardiac index greater than 2.2 l min–1 for a minimum of 24 h. Weaning involves decreasing the pump's contribution
while watching for sustained cardiac output without a significant increase in atrial pressures. If adequate systemic flows can be sustained with the device providing
minimal support (<500 ml min–1 augmentation), the device can usually be removed while maintaining inotropic and/or IABP support.
20.4.7
Myocardial oxygen supply and demand
Myocardial oxygen supply and demand must be balanced. Loss of blood supply can be acute (early graft occlusion or coronary artery spasm) and subacute (IMA
hypoperfusion syndrome, systemic hypotension). Myocardial oxygen demand increases with tachycardia, shivering, and agitation. Inadequate myocardial oxygen supply
may present as electrocardiographic STsegment changes, ventricular dyssynergy on echocardiographic examination, and increased left or rightsided filling pressures,
but can be difficult to recognize clinically.
20.4.8 Maintaining a stable cardiac rhythm

Maintenance of a stable cardiac rhythm, optimally normal sinus rhythm at a rate of 70–100 beats min–1, is the final hemodynamic goal. Perioperative heart rates above
100 beats min–1 are associated with ischemia and postoperative myocardial infarction; bradycardia may depress cardiac output. If pacing wires have been placed,
bradycardia can be treated electrically, and the pacing wires may also be used for arrhythmia detection. Arrhythmias may result from electrolyte imbalance, myocardial
ischemia, acidbase abnormalities, and excess catecholamines. Supraventricular tachycardias associated with hemodynamic instability should be treated with
synchronous cardioversion at 50–150 joules. Adenosine (initial dose 6 mg IV) is the drug of choice for reentry supraventricular tachycardias22.
Atrial arrhythmias are common; and occur in about one out of five CABG patients and one out of two valve patients. Magnesium prophylaxis can lessen the
incidence of atrial fibrillation following CABG. Rate control can be accomplished with betaadrenergic blockade, calciumchannel blockers (typically diltiazem),
amiodarone, or electrical cardioversion. Aggressive treatment and control of atrial arrhythmias shorten total hospital length of stay.
Ventricular arrhythmias are also common, and frequently can be caused by pulmonary artery (PA) catheter placement and electrolytes. Initial treatment of
hemodynamically significant ventricular dysrhythmias is immediate defibrillation and administration of intravenous lidocaine (1 mg kg–1 bolus) followed by a continuous
infusion of 2–4 mg min–1. Amiodarone, despite higher cost and the potential for pulmonary toxicity, is also considered firstline therapy for ventricular dysrhythmias.
Other agents such as ibutalide, bretylium, procainamide, quinidine, and diphenylhydantoin may occasionally be
Page 336
useful. (See Chapter 11 for a complete discussion of arrhythmia management.) Torsades de pointes, a special form of ventricular tachycardia associated with a long QT
interval, are commonly seen in patients with ongoing ischemia. Torsades may be precipitated or worsened by type 1A antiarrhythmic drugs and therapy of torsades in
the postCPB setting should first address repletion of Mg2+ and control of heart rate and ischemia.
20.5 Postoperative complications
20.5.1 Cardiac tamponade

This presents differently in the postoperative cardiac patient as compared with the trauma patient or in those with constrictive diseases of the pericardium. Because the
pericardial sac is generally left open, the compliance of the cavity is altered, and classic equalization of pressures may not be seen. Tamponade may be caused by acute
cardiac dilation without hemopericardium or by a small clot strategically placed so that it compresses a vital structure and decreases venous return. The diagnosis of
tamponade should always be suspected when hemodynamic status deteriorates, and the etiology of mediastinal tamponade following heart surgery is almost always
excessive postoperative bleeding with accumulated clot producing compression of one or more chambers of the heart. Although pericardiocentesis is lifesaving in non
surgical tamponade, it is rarely useful in the postbypass patient as clot cannot be aspirated, and the procedure risks disrupting newly placed grafts. Transesophageal
echocardiogram (TEE) can be helpful if the time course of deterioration is not rapid. With a high index of suspicion, the usual course is to transport the patient back to
the operating room, open the chest and evacuate the clot, at which point hemodynamic improvement is usually dramatic. If immediate return to the operating room is
not possible, the lower end of the sternotomy incision can be opened in the intensive care unit, and clot evacuated manually or with a suction catheter.
20.5.2 Perioperative myocardial infarction

This can be difficult to diagnose, as conventional criteria (new electrocardiographic changes; enzyme elevation) can result from surgery itself. Using the criteria of
creatinine kinase MB fraction (CPKMB) greater than 8% and aspartate aminotransferase greater than 80 units dl–1, the rate of infarction is in the range of 3–6%.
Reinfusion of shed mediastinal and chest tube drainage also complicates enzyme diagnosis of infarction, as autotransfusion raises the serum levels of creatinine kinase,
aspartate aminotransferase, and lactic dehydrogenases23. The criteria proposed by the Society of Thoracic Surgeons (STS) (Table 20.4) provide a reasonable
benchmark for quality assurance24.
20.5.3 Postoperative shivering

This occurs commonly following CPB, and the ability to abate shivering with steroid pretreatment suggests cytokine or mediator release rather than rewarming as the
major instigator25. Shivering increases oxygen consumption
Table 20.4 STS criteria for myocardial infarction (Adapted from reference 24)
A perioperative myocardial infarction (MI) is diagnosed by finding at least two of the following four criteria:
• Prolonged (<20 min) typical chest pain not relieved by rest and/or nitrates
• Enzyme level elevation; one of: (i) CKMB >5% of total CPK; (ii) CK greater than 2× normal; (iii) LDH subtype 1> LDH subtype 2; or (iv) troponin >0.2 µg ml–1
• New wall motion abnormalities
• Serial ECG (at least two) showing changes from baseline or serially in STT and/or Q waves that are 0.03 s in width and/or > or + onethird of the total QRS complex in
two or more contiguous leads
Page 337
beyond the level a stressed heart can deliver, and results in lactic acidosis. Patients may also develop acute respiratory acidosis as a result of the inability to increase
spontaneous minute ventilation appropriately because of residual opioids and neuromuscular blocking drugs. Shivering responds to radiant heat and/or small doses
(12.5–25 mg IV) of meperidine. Full ventilation and neuromuscular blockade may be needed if shivering is severe and cannot be controlled by other methods.
20.5.4 Oliguric or anuric renal failure

This occurs in 1–4% post open heart surgery patients; a rise in serum creatinine can be documented in 2–30% of patients26,27. Univariate predictors of renal
dysfunction include low cardiac output at the end of CPB, advanced age, preoperative cardiac failure, need for postoperative circulatory support, blood transfusions,
and prolonged time on CPB28. Despite aggressive ICU care and the availability of dialysis and hemofiltration, mortality from postoperative renal failure may exceed
80%, particularly if other organ system failure occurs simultaneously29.
When renal failure occurs, it generally follows one of three welldefined patterns (Figure 20.1)30. Abbreviated acute renal failure occurs after an isolated insult,
results in a peak in serum creatinine around the fourth postoperative day, and generally has a favorable prognosis if no other events occur. The second pattern is similar
to the first, except that the acute insult is accompanied by prolonged circulatory failure. This pattern runs a longer course, with recovery typically occurring in the second
or third week after injury, in tandem with improvements in cardiac output. The final pattern begins like the second, but recovery is complicated by a second insult such
as sepsis, massive gastrointestinal bleeding or myocardial infarction. As fluid overload with renal failure may precipitate respiratory and cardiac failure, there has been a
trend towards early application of continuous arteriovenous hemofiltration (CAVH) and related techniques to remove excess fluid13. There is speculation that
aggressive fluid removal may reduce tissue edema and intraabdominal pressure, thus improving renal perfusion pressure and consequently renal function. Other
speculation involves the role of renal replacement techniques in removing circulating cytokines, but there are few data proving that cytokine removal results in better
outcome.
20.5.5 Neurologic complications

Neurologic complications following openheart surgery are varied, and findings depend partly on how aggressively they are sought. Minor neuropsychiatric changes can
be demonstrated in over half of CPB patients, but gross sensorymotor abnormalities such as hemiparesis, aphasias, and global encephalopathy are relatively
uncommon. Despite anticoagulation, subarachnoid hemorrhage is also an uncommon event. Peripheral nerve injury, phrenic nerve injury, and delirium requiring
treatment are more common. Major neurologic disability occurs in just over 1% of patients, and can almost always be identified very early in the postoperative period.
ICU care of these patients is generally concerned with maintaining reasonable brain perfusion pressure, and ensuring continued protection of the airway, as aspiration is
a major risk. Management of the patient who fails to awaken includes withholding all sedatives, narcotics and muscle relaxants, and assessing peripheral neuromuscular
function by nerve stimulations. CT or MRI scanning is indicated to rule out structural lesions such as multiple emboli or cerebral edema. EEG and evoked potential
monitoring can assess brain electrical activity and rule out seizure activity, but only when other tests have failed to deliver a diagnosis.
Transient postoperative delirium occurs in about 7% of patients, resolving spontaneously or with reassurance and pharmacologic intervention in almost all patients by
the sixth postoperative day31. Initial management of agitation consists of reassurance and orientation of the patient, and control of pain and anxiety with opioids and
benzodiazepines respectively. Additional agents may worsen agitation accompanied by disorientation, and careful
Page 338
Figure 20.1 Postoperative renal dysfynction typically follows one of three patterns. In Pattern A, an acute insult at the time of surgery
results in an abrupt drop in creatinine clearance, which resolves over several days. The patient’s serum creatinine peaks around
day 4, and returns to normal. If the acute insult is accompanied by impaired cardiac function, the rise in creatinine will be higher
and the duration longer. Dialysis may be needed temporarily. If septic shock or other instability occurs during the period of poor
cardiac function, Pattern C will occur, and is likely to be associated with a need for chronic dialysis if the patient survives,
Modified from Myers, B.D. and Moran, S.M. (1987) Hemodynamically mediated acute renal failure. N. Engl. J. Med. 314:97.
Copyright (1987) Massachusetts Medical Society, All rights reserved.
observation off medication and ample reassurance and reorientation may suffice. If the patient remains agitated and discontented, haloperidol is often useful.
20.5.6 Postoperative bleeding

This can result from numerous causes, including inadequate heparin neutralization, fibrinolysis, complement activation, quantitative or qualitative platelet defects, and less
often, consumption and decreased levels of factors V, VIII, XIII, fibrinogen and plasminogen. No single test of hemostasis is sufficient. The activated clotting time
(ACT) is a good screen for adequate heparin reversal. Ddimer or fibrin split products may be ordered if there is suspicion of disseminated intravascular coagulation as
the result of low cardiac
Page 339
output and hypoperfusion. Prothrombin time (PT) and international normalized ratio (INR) are almost always abnormally elevated in any patient undergoing
cardiopulmonary bypass in the immediate postoperative period, even without excessive bleeding32. Platelet function can be depressed despite sufficient platelet
quantity. One approach to postoperative bleeding is to obtain an ACT, and, if elevated, administer protamine 50 mg intravenously, and then recheck the ACT. If
bleeding persists and the ACT is normal, then a clotting profile (PT, INR, platelet count) is drawn. At this point, platelets may be indicated even in the presence of a
normal count, as function may be impaired. Desmopressin33, Amicar34 and aprotinin35 have all been recommended either for prophylaxis or treatment. Substantial
bleeding (more than 300 ml in the first hour, 250 ml in the second hour, or 150 ml h–1 thereafter) will probably require surgical exploration36, although the decision must
be individualized based on the surgeon’s knowledge of operating room events and the stability of the patient. Surgical exploration is also indicated if a large hemothorax
or pericardial tamponade occurs.
20.5.7 Infectious complications

These range from simple wound infections through mediastinitis to sternal dehiscence, the latter occurring in about 1%, but associated with a 13% mortality rate37.
Predisposing factors for wound complications after open heart surgery include diabetes, low cardiac output, use of bilateral internal mammary grafts, and reoperation
for control of bleeding. Mediastinal infection can present as failure to wean, unexplained fever, and an unstable sternum. Management of mediastinitis requires selective
antibiotic therapy, debridement, irrigation, drainage, and reclosure. In cases of severe sternal necrosis, early debridement and plastic surgical reconstruction using
omental and/or myocutaneous flaps may be required.
Pneumonia, tracheobronchitis, line sepsis, and urinary tract infections are frequently seen in the ventilatordependent longterm ICU patient (see Chapter 16). As
routine antibiotic prophylaxis for cardiac surgery typically ‘covers’ for staphylococci and enteric Gramnegative rods, postoperative infections commonly involve
relatively resistant organisms such as Candida albicans, and Gramnegative rods such as Pseudomonas aeruginosa, Klebsiella and Providencia species, and
Serratia marcesans. Methicillinresistant Staphylococcus aureus (MRSA) and vancomycinresistant enterococci (VRE) are of increasing concern. Reintubation,
nasogastric tubing, prior broadspectrum antibiotic therapy and blood transfusion are risk factors for acquisition of nosocomial pneumonia38.
20.5.8 Gastrointestinal complications

Gastrointestinal complications requiring intervention occur in only 1–3% of patients, but are associated with morbidity and mortality39. Postoperative ileus and upper
gastrointestinal bleeding are the most common complications. Pancreatitis, intestinal ischemia, perforation, or bleeding elsewhere in the gastrointestinal tract tend to
occur in complicated patients who have an initial low cardiac output syndrome resulting in ventilator dependence, renal dysfunction, or multiorgan failure. Mesenteric
ischemia can occur as a result of low perfusion or embolization of atheroma from largevessel manipulation. The incidence of gastrointestinal bleeding can be minimized
with antacid therapy, histamine blockers, or barrier protection agents such as sucralfate. Enteral nutritional support also appears to protect the gastric mucosa40.
Jaundice occurs in approximately 20% of patients post cardiopulmonary bypass, and increases risk of subsequent mortality41. Acute pancreatitis is less common, but
carries a poor prognosis, especially in the setting of multisystem organ failure. Transient elevation of amylase may occur following CPB and be unrelated to pancreatic
injury, making hyperamylasemia less valuable as a diagnostic test, unless it is combined with serum lipase assay.
Page 340
20.6 Discharge criteria

Patients recovering from cardiac surgery are typically monitored and treated for several hours to overnight in a traditional intensive care unit. Increasingly, some centers
are using specialized recovery areas, postanesthesia care units (PACU), or highdependence nursing units as substitutes for the ICU. In many hospitals, what the ward
is called is less important than the nursetopatient ratio. The nursetopatient ratio for a fresh postoperative patient may well be greater than 1:1 if the patient is
hemodynamically unstable, requiring an IABP or other assist device, or actively bleeding. A 1:1 nursetopatient ratio is reasonable for the first few hours of instability,
and a 1:2 ratio is typically required for the first 6–24 h, depending on the type of anesthesia utilized and rapidity of extubation. Once the patient meets the criteria in
Table 20.5, less nursing attention will be needed, and the patient can typically be moved to a lowerdependence area. The risks of premature ICU discharge include
respiratory compromise, particularly aspiration as a result of residual sedation and/or neurologic changes, and unrecognized renal or metabolic problems, plus continued
bleeding. Cardiac arrhythmias are more difficult to predict, and may not occur until the third or fourth postoperative day, so continued electrocardiographic monitoring
is appropriate after ICU discharge. ICU readmission rates are typically in the 5–9% range, and are most often precipitated by pulmonary and cardiac complica
Table 20.5 Criteria for ICU discharge following open heart syrgery
Neurologic and psychological

Able to protect airway (adequate gag reflex; or tracheostomized)
Able to signal distress (verbal or using call button)
Cardiac
Stable MAP≥70 without pressor support
Adequate perfusion (Cl ≥2.2; warm extremities; satisfactory urine output)
No inotropic support other than lowdose dopamine (≤3 µg kg–1 min–1)
Heart rhythm: sinus with few or no PVCs; or atrial fibrillation with controlled rate
Not pacer dependent (unless suitable junctionalventricular escape rate)
Respiratory
Maintaining pH >7.35 and PCO2 within 8 mmHg of baseline value
Oxygen saturation ≥92% on nasal cannula or simple mask (<50% FiO2)

Coughing well and clearing secretions; suctioned twice per shift or less
No pneumothorax, large effusions, or pulmonary edema on CXR
Renal or metabolic
Urine output at least 0.5 cm3 kg–1 h–1
Weight within 5 kg of preoperative value
Renal function (creatinine) near baseline (or known problem followed by consultant)
Serum potassium ≥3.5 mequiv
Blood glucose adequately controlled
Hematologic
Hematocrit ≥6%
Chest tube output <50 cm3 h–1
Clotting parameters within normal range unless deliberately anticoagulated
Infectious
SIRS resolving (temperature ≤38.0°C, WBC <18 000)
The above criteria may be liberalized if nursing and monitoring facilities on a ‘stepdown’ ward are available to meet patient needs. MAP, mean arterial pressure; CI, cardiac index; PVC,
premature ventricular contraction; CXR, chest radiograph.
Page 341
tions; less often by infection, gastrointestinal bleeding, or neurologic events.
Several different systems are available for retrospective analysis of outcome in a cardiac surgical program42. In general, the risk for morbidity and mortality in an
individual patient is increased with emergency procedures, poor left ventricular function (ejection fraction <35%), complex surgery, reoperation, advanced age, low
body weight, diabetes, peripheral or carotid vascular disease, and chronic obstructive pulmonary disease43. Severityadjustment tools such as the Acute Physiology
and Chronic Health Evaluation (APACHE) and the Mortality Probability Model (MPM) are generally not appropriate for evaluating the cardiac surgical patient whose
physiology is deliberately altered in the operating room.
Patients presenting for CABG surgery are older and more seriously ill now than a decade ago. Riskadjusted mortality has not changed substantially, although
complications have diminished, possibly as the result of better intraoperative and postoperative care44. The risk of subsequent morbidity and mortality can be predicted
based on a patient's condition on arrival in the ICU45 (Table 20.6, Figure 20.2) but all risk predictions are based on population averages and do not necessarily apply
to the individual. Patients with singleorgan failure have a reasonable chance of functional recovery, although they may require chronic hemodialysis, or tracheostomy
and longterm weaning from ventilatory support. Multiple organ failure is a poor prognostic sign29,37, and mortality correlates with the number of failed organs45,46.
Physiologic insults (upper gastrointestinal bleeding, aspiration, sepsis) that occur during recovery from the initial bout of organ failure are frequently fatal. Deciding when
further care approaches futility is difficult and open to challenge, but should be addressed with the patient’s family and the health care
Figure 20.2 The risk of mortality, serious morbidity and prolonged ICU length of stay is a function of condition at ICU admission,
calculated by adding the point values in Table 20.6. Patients with five points or less have excellent outcomes, whereas those
with high scores are likely to have complicated, lengthy postoperative courses. Mortality is included within the definition of
morbidity, as patients who die early in the postoperative course may not have time to manifest ventilator dependence, sepsis,
renal or neurologic failure. It should be noted, however, that a small percentage of patients with the highest score may still have
a good outcome, From: Higgins, T.L, Estafanous, F.G., Loop, F.D., et al. (1997) ICU admission score for predicting morbidity
and mortality risk after coronary artery bypass grafting. Reproduced with permission from Elsevier Science. Ann. Thorac
Surg. 64:1050–1058,
Page 342
Table 20.6 ICU riskassessment score for cardiac surgical patients
Preoperative factors ICU admission factors

Preop albumin <3.5 mg % 5 IABP post CPB 7
Preop creatinine ≥1.9 mg % 4 CPB time≥160min 3
Age ≥70 years 3 Arterial bicarb <21 4
History PVD or vasc surg 3 CVP≥17mmHg 4
≥2 prior heart operations 2 Cardiac index <2.1 3
One prior heart operation 1 Heart rate≥100 3
Small size (BSA <1.72 m2) 1 (Aa)PO2 gradient ≥250 torr 2
From: Higgins, T.L, Estafanous, F.G., Loop, F.D., et al. (1997) ICU admission score for predicting morbidity and mortality risk after coronary artery bypass grafting. Reproduced with
permission from Elsevier Science. Ann. Thorac Surg. 64:1050–1058.
team if the patient is not improving by the second or third postoperative week. Characteristics of patients who survive multiorgan failure and prolonged ICU stay are
younger age (<70 years), awake neurologic status with a will to live, and unambiguous signs of recovery such as less ventilator or pressor support over time. However,
there are no good studies to quantify these impressions.
20.8 Summary
Management of the cardiac surgery patient is becoming ever more complex with changes in demographics and extent of surgery. Successful ICU outcome following
cardiac surgery requires constant vigilance and attention to details. Most postCABG patients are relatively healthy, and will recover nicely on their own as long as
important complications are avoided. Statistically, certain patients are more likely to suffer postoperative complications, and it is in this highrisk group that greater
attention and interventions can be directed. By understanding the risk factors for catastrophic deterioration, and undertaking preventive measures, outcome can be
improved.
References
1. Smith, I., Fleming, S. and Cernaianu, A. (1990) Mishaps during transport from the intensive care unit. Crit. Care Med. 18:278–281.
2. Braun, S.R., Brinbaum, M.L. and Chopra, P.S. (1978) Pre and postoperative pulmonary function abnormalities in coronary artery revascularization surgery. Chest
73:316–320.
3. Ratliff, N.B., Young, G., Jr, Hackel, D.B., Mikat, E. and Wilson, J.W. (1973) Pulmonary injury secondary to extracorporeal circulation. J. Thorac. Cardivasc.
Surg. 65:425–432.
4. Michalopoulos, A., Anthi, A., Rellos, K., et al. (1998) Effects of positive endexpiratory pressure (PEEP) in cardiac surgery patients. Respir. Med. 92:858–862.
5. Higgins, T.L. (1992) Early endotracheal extubation is preferable to late extubation in patients following coronary artery surgery. J. Cardiothorac. Vasc. Anesth.
6:488–493.
6. Cheng, C.H., Karski, J., Peniston, C., Raveendran, G., Asokumar, B., Carroll, J., David, T. and Sandler, A. (1996) Early tracheal extubareduces costs and
improves resource use. A tion after coronary artery bypass graft surgery prospective, randomized, controlled trial. Anesthesiology 85:1300–1310.
7. Higgins, T.L. (1995) Safety issues regarding early extubation after coronary artery bypass surgery. J. Cardiothorac. Vasc. Anesth. 9:24–29.
8. Engelman, R.M., Rousou, J.A., Flack, J.E., (1994) Fasttrack recovery of the coronary bypass patient. Ann. Thorac. Surg. 58:1742–1746.
9. Millbern, S.M., Downs, J.B., Jumper, L.C. and Modell, J.H. (1978) Evaluation of criteria for discontinuing mechanical ventilatory support.Arch. Surg. 113:1441–
1443.
10. Higgins, T.L., Yared, J.P., Paranandi, L., Baldyga, A. and Starr, N.J. (1991) Risk factors for respiratory complications after cardiac surgery. Anesthesiology
75:3A.
11. Covelli, H.D., Black, J.W., Olsen, M.S. and Beekman, J.F. (1981) Respiratory failure precipitated by high carbohydrate loads. Ann. Intern. Med. 95:579–581.
12. Lemaire, E, Teboul, J.L., Cinotti, L., Giotto, G., Abrouk, F., Steg, G., MacquinMavier, I. and Zapol, W.M. (1988) Acute left ventricular dysfunction during
unsuccessful weaning from mechanical ventilation. Anesthesiology 69: 171–179.
Page 343
13. Coraim, F.J., Coraim, H.P., Ebermann, R. and Stellwag, F.M. (1986) Acute respiratory failure after cardiac surgery: clinical experience with the application of
continuous arteriovenous hemofiltration. Crit. Care Med. 14:714–718.
14. Wilcox, P., Baile, E.M., Hards, J., Muller, N.L., Dunn, L., Pardy, R.L. and Pare, P.D. (1988) Phrenic nerve function and its relationship to atelectasis after
coronary artery bypass surgery. Chest 93:693–698.
15. Breisblatt, W.M., Stein, K.L., Wolfe, C.J., et al. (1990) Acute myocardial dysfunction and recovery: a common occurrence after coronary artery bypass. J. Am.
Coll. Cardiol. 15: 1261–1269.
16. Butterworth, J.F., Zaloga, G.P., Prielipp, R.C., Tucker, W.Y., Jr and Royster, R.L. (1992) Calcium inhibits the cardiac stimulating properties of dobutamine but
not of amrinone. Chest 101: 174–180.
17. FernandezDel Castillo, C., Harringer, W., Warshaw, A.L., Vlahakes, G.J., Koski, G., Zaslavsky, A.M. and Rattner, D.W. (1991) Risk factors for
pancreatic cellular injury after cardiopulmonary bypass. N. Engl J. Med. 325: 382–387.
18. Janelle, G.M., Urdaneta, F., Martin, T.D. and Lobato, E.B. (2000) Effects of calcium chloride on grafting internal mammary artery flow after cardiopulmonary
bypass. J. Cardiothorac. Vasc. Anesth. 14:4–8.
19. Estafanous, F.G., Urzua, J., Yared, J.P., Zurick, A.M., Loop, F.D. and Tarazi, R.C. (1984) Pattern of hemodynamic alterations during coronary artery
operations. J. Thorac. Cardiovasc. Surg. 87:175–182.
20. Higgins, T.L., Yared, J.P., Estafanous, F.G., et al. (1994) Propofol versus midazolam for intensive care unit sedation after coronary artery bypass grafting. Crit.
Care Med. 22: 1415–1423.
21. Chauvin, M., Deriaz, H. and Viars, P. (1987) Continuous IV infusion of labetalol for postoperative hypertension. Br. J. Anesth. 59: 1250–1256.
22. Pinski, S.L. and Maloney, J.D. (1990) Adenosine: a new drug for acute termination of supraventricular tachycardia. Clev. Clin. J. Med. 57:383–388.
23. Wahl, G.W., Feins, R.H., Alfieres, G. and Bixby, K. (1992) Reinfusion of shed blood after coronary operation causes elevation of cardiac enzyme levels. Ann.
Thorac. Surg. 53: 625–627.
24. Society of Thoracic Surgeons (1999) Definition of Terms. STS, National Cardiac Database, Chicago, IL, p. 25.
25. Yared, J.P., Starr, N.J., HoffmanHogg, L., Bashour, C.A., Insler, S.R., O’Connor, M., Piedmonte, M. and Cosgrove, D.M. (1998) Dexamethasone
decreases the incidence of shivering after cardiac surgery: A randomized, doubleblind, placebocontrolled study. Anesth. Analg. 87:795–799.
26. Bhat, J.G., Gluck, M.C., Lowenstein, J. and Baldwin, D.S. (1976) Renal failure after open heart surgery. Ann. Intern. Med. 84:677–682.
27. Hilberman, M., Myers, B.D., Carrie, B.J., Derby, G., Jamison, R.L. and Stinson, E.B. (1979) Acute renal failure following cardiac surgery. J. Thorac.
Cardiovasc. Surg. 77:880.
28. Koning, H.M., Koning, A.J. and Leusink, J.A. (1985) Serious acute renal failure following open heart surgery. Thorac. Cardiovasc. Surg. 33: 283–287.
29. Higgins, T.L., Blum, J.M. and Lytle, B.W. (1992) Postoperative complications following coronary bypass grafting: implications of elevated bilirubin and elevated
creatinine. Abstracts of Society of Cardiovascular Anesthesiologists Meeting. Society of Cardiovascular Anesthesiologists, Richmond, Virginia, p. 27.
30. Myers, B.D. and Moran, S.M. (1987) Hemodynamically mediated acute renal failure. N. Engl. J. Med. 314:97.
31. Calabrese, J.R., Skwerer, R.G., Gulledge, A.D., et al. (1987) Incidence of postoperative delirium following myocardial revascularization. Clev. Clin. J. Med.
54:29–32.
32. Milam, J.D., Austin, S.F., Martin, R.F., Keats, A.S. and Cooley, D.A. (1981) Alteration of coagulation and selected clinical chemistry parameters in patients
undergoing open heart surgery without transfusions. Am. Soc. Clin. Pathol. 76:155–162.
33. Salzman, E.W., Weinstein, M.J., Weintraub, R.M., et al. (1986) Treatment with desmopressin acetate to reduce blood loss after cardiac surgery: a doubleblind
randomized trial. N. Engl J. Med. 314:1402–1406.
34. DelRossi, A.J., Cernaiann, A.C., Botros, S., Lemole, G.M. and Moore, R. (1989) Prophylactic treatment of postperfusion bleeding using EACA. Chest 96:27–
30.
35. Royston, D., Taylor, K.M., Bidstrup, B.P. and Sapsford, R.N. (1987) Effect of aprotinin on need for blood transfusion after repeat openheart surgery. Lancet
2:1289–1291.
36. Michelson, E.L., Torosian, M., Morganroth, J. and MacVaugh, H., III (1980) Early recognition of surgically correctable causes of excessive mediastinal
bleeding after coronary artery bypass graft surgery. Am. J. Surg. 139:313–317.
37. Mahfood, S.S., Higgins, T.L. and Loop, F.D. (1991) Management of complications related to
Page 344
perioperative care in cardiothoracic surgery . In:Management of Complications Related to Coronary Artery Bypass Surgery (eds J.A Waldhausen and
M.B.Orringer). MosbyYear Book, St. Louis, MO, pp. 265–280.
38. LealNoval, S.R., MarquezVarcaro, J.A., GarciaCuriel, A., CamachoLarana, P., RinconFerrari, M.D., OrdonezFernandez, A., FloresCordero, J.M.
and LoscertalesAbril, J. (2000) Nosocomial pneumonia in patients undergoing heart surgery. Crit. Care Med. 28:935–940.
39. Krasna, M.J., Flanchbaum, L., Trooskin, Z.S., et al. (1988) Gastrointestinal complications after cardiac surgery. Surgery 104:733.
40. Ephgrave, K.S., KleinmanWexler, R.L. and Adar, C.G. (1990) Enteral nutrients prevent stress ulceration and increase intragastric volume. Crit. Care Med.
18:621.
41. Sanderson, R.G., Ellison, J.H., Benson, J.A. and Starr, A. (1967) Jaundice following openheart surgery. Ann. Surg. 165:217.
42. Peterson, E.D., DeLong, E.R., Muhlbaier, L.H., Rosen, A.B., Buell, H.E., Kiefe, C.I. and Kresowik, T.F. (2000) Challenges in comparing riskadjusted
bypass surgery mortality results. J. Am. Coll Cardiol. 36:2174–2184.
43. Higgins, T.L. (1998) Quantifying risk and assessing outcome in cardiac surgery. J. Cardiothorac. Vasc. Anesth. 12:330–340.
44. Estafanous, F.G., Loop, F.D., Higgins, T.L., TekyiMensah, S., Lytle, B.W., Cosgrove, D.M., RobertsBrown, M. and Starr, N.J. (1998) Increased risk
and decreased morbidity of coronary artery bypass grafting between 1986 and 1994. Ann. Thorac. Surg. 65:383–389.
45. Higgins, T.L., Estafanous, F.G., Loop, F.D., et al. (1997) ICU admission score for predicting morbidity and mortality risk after coronary artery bypass grafting.
Ann. Thorac. Surg. 64: 1050–1058.
46. Vincent, J.L., de Mendoca, A., Cantraine, F., et al. (1998) Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units:
results of a multicenter, prospective study. Crit. Care Med. 26: 1793–1800.
Page 345
Chapter 21
Postoperative care of the thoracic surgical patient
Contents
21.1 Introduction
21.2 Identifying the highrisk patient
21.3 Operating room events affecting ICU care
21.4 Immediate postoperative issues
21.5 Extubation and airway concerns
21.6 Postoperative fluid management
21.7 Prolonged ventilator support
21.8 Specific postoperative complications
21.9 Pulmonary parenchymal complications
21.10 Pleural complications
21.11 Other considerations
Page 346
21.1 Introduction
The thoracic surgery patient is typically older, has concurrent medical problems, may be debilitated as a result of cancer and/or malnutrition, and often has significant
abnormalities of lung function. Pulmonary abnormalities may have arisen from tobacco use, occupational exposure, or a primary disease process. A history of asthma,
wheezing, or allergic airway responses is important, both as a risk factor, and to identify patients in whom bronchodilator management may be needed in the post
operative period.
21.2 Identifying the highrisk patient

Forced vital capacity (FVC) alone is insufficiently sensitive and specific as the sole physiologic predictor of postoperative pulmonary function. Forced expiratory
volume in 1 s (FEV1), is determined by inspiratory muscle strength, elastic recoil and degree of obstructive air trapping, and provides a good indicator of the patient’s
postoperative ability to cough effectively and clear secretions. The decrease in FEV1 after lung resection for cancer is not necessarily a simple proportional relationship,
as lesser decreases in FEV1 than calculated will be noted if an obstructed lobar or mainstem bronchus is present. A cutoff value for postpneumonectomy FEV1 of 800
ml is commonly used as a criterion of resectability.
Lung surgery affects postoperative lung function, and entering the chest cavity, even without resection of tissue, produces substantial changes. Posteriorlateral
thoracotomy is associated with greater postoperative impairment than median sternotomy. FVC and functional residual capacity (FRC) can fall to less than 60% of their
preoperative values on the first postoperative day. Subsequent return to baseline can take up to 14 days. The decline in FRC is especially important, because the
resulting atelectasis causes physiologic shunting and hypoxemia. In patients with severe chronic obstruction, the best predictors of postoperative ventilation requirements
were arterial pO2 less than 70% of predicted for age and the presence of dyspnea at rest1. For elective thoracic or abdominal surgery, the risk of postoperative
pneumonia increases with:
• low serum albumin concentration on admission;
• high American Society of Anesthesiologists (ASA) physical status classification;
• history of smoking;
• longer preoperative stay;
• longer operative procedure;
• thoracic or upper abdominal surgical site2.
Numerous studies implicate advanced age, particularly age over 80 years3 as a risk factor. Elderly patients have a number of agerelated changes in pulmonary
function, including decreased elastic recoil and progressive stiffening of the chest wall, increase in the ratio of FRC to total lung capacity, and diminished vital capacity
and FEV14. The activity of upper airway reflexes is blunted, which may result in impaired clearance of secretions and ability to protect the airway. In patients over the
age of 70 undergoing thoracotomy, hospital mortality can be 7.5%, with a complication rate of 17%5.
Obesity results in decreases in FRC and expiratory reserve volume (ERV), to the point that ERV may drop below closing volume, resulting in perfused, unventilated
segments of lung and a widened (Aa)pO2 gradient. Obese patients are more likely to cough poorly, retain secretions and develop basilar atelectasis.
Cigarette smoking is well recognized for its contribution to perioperative morbidity via effects on the cardiovascular system, mucus secretion and clearance, and small
airway narrowing. The benefit of smoking cessation before elective surgery is well recognized, although data in coronary artery bypass patients suggests that cessation
of smoking within 8 weeks of surgery may actually increase the risk of postoperative pulmonary complications6, possibly as a result of a transient increase in sputum
volume.
A system for classification of the risk of pulmonary complications following thoracic and
Page 347
abdominal procedures was developed by Shapiro et al.7 (Table 21.1). By applying 0–3 points for spirometric valves, cardiovascular history, arterial blood gases,
central nervous system, and expected postoperative ambulation, patients can be classified as low risk, moderate risk, or high risk. Lowrisk patients do not generally
require oxygen therapy after recovery room discharge. Moderaterisk patients require several days’ observation. The highrisk patient often requires postoperative
intensive care and significant prophylactic interventions.
21.3 Operating room events affecting ICU care
21.3.1 Anesthetic agents

The speed of postoperative recovery will depend in part on the amount and types of anesthetic agents given as premedication and during the operative procedure.
Considerations for the anesthesiologist involve the need to use high inspired oxygen concentrations, particularly during onelung anesthesia (limiting the use of nitrous
oxide) and the goal of early extubation (limiting the use of opioids). Regional techniques alone (spinal, epidural) are generally not applicable to operative anesthesia in
thoracic procedures because of the difficulty in providing a high enough spinal level to allow surgery without affecting the brainstem or muscles of respiration. Controlled
ventilation is necessary to sustain respiration during openthorax procedures. The usual plan is to have the patient awake, comfortable and extubated at the end of the
procedure, particularly in volume reduction surgery for emphysema where stress on fresh suture lines is undesirable.
Table 21.1 Calculating the risk of pulmonary complication of thoracic and abdominal procedures
Category Points*
Expiratory spirogram
Normal (%FVC+%FEV1/FVC>150) 0
%FVC+%FEV1/FVC=100–150 1
%FVC+%FEV1/FVC<100 2
Preoperative FVC <20 ml kg–1 3
Postbronchodilator FEV1/FVC <50% 3
Cardiovascular system
Normal 0
Controlled hypertension, myocardial infarction without sequelae for more than 2 years 0
Dyspnea on exertion, orthopnea, paroxysmal nocturnal dyspnea, dependent edema, congestive heart failure, angina 1
Arterial blood gases
Acceptable 0
PaCO2>50 mmHg or PaO2<60 mmHg on room air 1
Metabolic pH abnormality >7.50 or <7.30 1
Nervous system
Normal 0
Confusion, obtundation, agitation, spasticity, discoordination, bulbar malfunction 1
Significant muscular weakness 1
Postoperative ambulation
Expected ambulation (minimum, sitting at bedside) within 36 h 0
Expected complete bed confinement for at least 36 h 1
*Zero points indicate low risk; 1–2 points, moderate risk, 3 points, high risk. From: Shapiro, B.A., Harrison, R.A., Kacmarek, R.M. and Can, R.D. (1985). Reproduced with permission from
Clinical Application of Respiratory Care, 3rd Edn. Year Book, Chicago, IL.
Page 348
21.3.2 Selective endobronchial intubation

This allows isolation of the right and left lung, permitting the surgeon to operate on a quiet, collapsed lung, while ventilating on the opposite side. Disposable polyvinyl
doublelumen tubes are available in odd sizes between 35 and 41 French in both rightsided and leftsided configurations. Single lumen tubes with a retractable blocker
(such as the UniVent tube) can also be used to achieve lung isolation. The nonoperative bronchus is usually chosen for selective intubation in lobectomies and
pneumonectomies, so that surgical manipulation does not displace the tube and disrupt lung isolation, and to allow resection of the mainstem bronchus if necessary.
When selective endobronchial intubation is impossible (pediatric patients, very small adults, laryngectomy patients), a bronchial blocker can be placed under fiberoptic
guidance to selectively occlude a bronchus (Figure 21.1).
Onelung ventilation alters the ventilationperfusion relationship, and blood passing through the unventilated lung effectively causes a righttoleft shunt and reduces
arterial oxygenation. Perfusion of the unventilated lung will be reduced by physical collapse of the lung and hypoxic pulmonary vasoconstriction. As the doublelumen
endotracheal tube is large with the potential to cause airway trauma and edema, prone to shifts in position, and more difficult to suction through, it is generally removed
at the end of the operation and replaced by a singlelumen tube if continued mechanical ventilation is required. Specific indications for continued selective endobronchial
intubation would include the need to protect against soilage (pus or blood) and provision of different levels of positive endexpiratory pressure to lung of different
compliance (Table 21.2). The need for differential levels of PEEP frequently occurs in emphysematous patients undergoing singlelung transplantation. A fiberglass resin
tubechanger or a pediatric fiberoptic bronchoscope are essential tools, which should be available for placement and adjustment of doublelumen tubes.
Figure 21.1 Fogarty embolectomy catheter used as a bronchial blocker. From; Rice, T.W., Higgins, T.L, Kirby T.J. (1995):
Management the General Thoracic Surgical Patient. In: Sivak, G.D., Higgins, T L, Seiver, A. (eds) The high risk patient:
Management of the critically ill. Reproduced with permission from Lippincott, Williams and Wilkins, Baltimore.
21.3.3 ICU versus PACU versus stepdown for recovery

The choice of recovery area following operation will depend primarily on the degree of patient illness, and the ability of a particular unit to deal with postoperative
ventilation
Table 21.2 Indication for use of endobronchial intubation
Absolute
To prevent soilage of contralateral lung with pus, blood or secretions
To control distribution of ventilation
Bronchopleural fistula
Cystic lesions
Differential lung compliance
For unilateral bronchopulmonary lavage
For unilateral lung transplantation
Relative
Surgical exposure—pneumonectomy and thoracic aneurysm
Surgical convenience—esophageal resection and lobectomy
rd
From: Shapiro, B.A., Harrison, R.A,, Kacmarek, R.M. and Can, R.D. (1985). Reproduced with permission from Clinical Application of Respiratory Care, 3 Edn. Year Book, Chicago, IL.
Page 349
and/or hemodynamic monitoring. At many hospitals, patients undergoing bronchoscopy, mediastinoscopy, esophageal dilation, esophagoscopy, gastrostomy,
jejunostomy, laryngoscopy, video thoracoscopy, or scalene node biopsy can spend a short time in the postanesthesia recovery unit (PACU) and then be transferred to
a regular nursing floor, or even sent home. Patients undergoing lobectomy, segmental or wedge pulmonary resections, or hiatal hernia repair (transabdominal Nissen
repairs) can generally be recovered in PACU and then sent to a stepdown unit if there are no complications. Patients undergoing esophagectomy,
esophagogastrectomy, and pneumonectomy, may have ongoing monitoring needs or require postoperative ventilation, and thus require management in an intensive care
setting.
21.4 Immediate postoperative issues
21.4.1 Usual ICU monitoring

This includes intermittent blood pressure determinations, continuous electrocardiography and pulse oximetry to detect hypotension, dysrrhythmia and arterial
desaturation, respectively. In selected patients, assessing intravascular volume status and cardiopulmonary function may require use of central venous pressure or
pulmonary artery catheters.
21.4.2 Chest tubes

These are usually inserted to drain the surgical site at the end of the procedure, except with pneumonectomy patients where standard practice is to avoid a chest tube
unless there is an infected pleural space. Chest tubes should never be clamped, even during patient transport, because of the dangers of unrecognized bleeding and
tension pneumothorax. Immediately upon arrival in the PACU or ICU, patients should have a quick but comprehensive examination including vital signs, auscultation of
breath sounds, and visual inspection of monitoring lines and chest tube connections. Chest tubes, except for tubes in pneumonectomy spaces which require a ‘balanced’
drainage system, are usually connected to a vacuum regulator to provide −20 cmH2O of suction (Figure 21.2). A chest
Figure 21.2 The workings of a chest tube drainage system can best understood when simplified to a threebottle system. Bottle (a) is
the collection chamber; bottle (b) provides a water seal and escape mechanism for extrapleural or intrathoracic air; bottle (c) is
a safety device to limit amount of negative pressure that can be applied. From Rice, T.W., Higgins, T.L., Kirby,T.J. (1995):
Management of the General Thoracic Surgical Patient. In Sivak, G.D., Higgins., T.L., Seiver, A. (eds) The high risk patient:
Management. of the ciritically ill. Reproduced with perimssion from Lippincott, Williams and Wilkins, Baltimore.
Page 350
radiograph should be ordered immediately for confirmation of endotracheal, nasogastric and chest tube placement, and identification of pneumothorax, mediastinal shift
or significant atelectasis. Routine chest Xrays are not necessary after uncomplicated removal of chest tubes and the decision to reinsert a chest tube is usually based on
clinical appearance rather than radiologic findings8.
21.4.3 Commercially available chest tube systems

These vary in their appearance, but all provide calibrated drainage chambers, a method to release excess positive pressure, and regulated amounts of negative pressure.
The mechanics of these systems can best be understood as a threebottle system (Figure 21.2). Air bubbles are normally expected in the chamber that limits the
amount of applied suction; air bubbles in the water seal chamber represent an active leak.
Hourly output from chest tubes should be recorded, and the surgeon notified if drainage is more than 100 cm3 h–1 for more than 4 h, or if more than 200 cm3 of
drainage is recorded in any 1 h observation period. Expected chest tube drainage in the first 24 h is 300–600 cm3, tapering to less than 200 cm3 by the second day.
The level of fluid in the water seal chamber should fluctuate with each respiration (assuming no air leak) and should be checked frequently to assure chest tube patency.
Most pulmonary resection patients will return with mild to moderate air leaks, which become problematic only if the underlying lung parenchyma does not completely
expand to fill the pleural space, or when the patient loses a significant percentage of tidal volume out of the chest tubes. In such circumstances, additional pleural
drainage may be required, or changes in ventilation made to minimize the air leak and optimize ventilation. Frequently, leaks occur only above a given inflation pressure,
and ventilation techniques such as smaller volumes at higher rates or pressurecontrolled inverse ratio ventilation can minimize leaking and allow a seal to develop. Once
all air leaks have resolved and drainage is minimal (less than 100 cm3 per 24 h), chest tubes may be removed during the expiratory phase of ventilation or while the
patient performs a Valsalva maneuver.
21.5 Endotracheal extubation and airway concerns

Extubation can often be accomplished in the operating room, but specific considerations (concurrent cardiac illness, inability to protect airway, malnutrition, coexisting
lung disease) may mandate continued intubation. Silent aspiration of gastric contents is an important complication following pulmonary resections, and maintenance of
endotracheal intubation for 24 h postoperatively has been shown to decrease the occurrence of pneumonia and the operative mortality rate9 in highrisk patients.
Measurement of maximal inspiratory pressure (MIP, often called negative inspiratory force, NIF is helpful in determining respiratory muscle strength, particularly in
patients recovering from thymectomy for myasthenia gravis, and in those who have received longacting neuromuscular blocking agents in the operating room. Residual
neuromuscular blockade can be assessed using a trainoffour monitor and reversed, if necessary, with small doses of neostigmine plus vagolytic agents such as
atropine or glycopyrrolate. Criteria for extubation include:
• awake and following instructions;
• presence of gag reflex (airway protection) and cough (secretion clearance);
• respiratory rate to tidal volume (f/Vt) ratio of less than 100 (no rapid/shallow breathing);
• MIP of greater than 25 cmH2O;
• adequate oxygen saturation (92% or more) on FiO2 of 50% or less at PEEP of five or less.
Although many patients will not strictly meet these criteria for extubation, it is usually best to attempt weaning and extubation rather than risk the complications of
continued
Page 351
ventilation in this group of patients. Specific indications for delaying extubation include:
• airway compromise as a result of edema or bleeding (see below);
• inadequate pulmonary reserve post surgery;
• compromised myocardial function especially with perioperative infarction;
• expected large fluid shifts with thoracoabdominal procedures;
• severe neurologic impairment;
• continued bleeding with likelihood of return to operating room;
• esophageal surgery patients (at risk for reflux and aspiration—delay extubation until airway reflexes have fully recovered as for ‘fullstomach’ intubation).
Considerations for longterm ventilator support, weaning and tracheostomy are covered in Chapter 18.
21.5.1 Laryngeal and glottic edema

This may occur after airway manipulation or intubation with a large doublelumen endotracheal tube. The presence of serious laryngeal edema can be detected by
deflating the endotracheal tube cuff (after first suctioning the posterior pharynx), and occluding the endotracheal tube and watching for evidence of airway obstruction.
Endotracheal intubation may need to be maintained while edema resolves. Racemic epinephrine and corticosteroids are useful adjuncts. If there is any doubt about
airway patency, the endotracheal tube should be removed only under direct laryngoscopic or fiberoptic observation, with a thyrocricoidotomy set immediately at hand
to facilitate airway access should reintubation be impossible because of airway swelling.
21.6 Postoperative fluid management

Fluid administration must be individualized based on patient condition, and a ‘cookbook’ approach is impossible. In patients where pulmonary capillary integrity is likely
to be a problem (significant lung retraction, pneumonectomy, lung transplant recipients, postaspiration) fluids should be minimized. On the other hand, patients who have
had considerable dissection and mobilization of tissue planes where ‘third space’ losses are likely (esophagectomy, esophagogastrectomy, excision of large mediastinal
tumors) will require large volumes (typically 2 l or more) of fluid replacement. This fluid is generally spontaneously mobilized beginning on the second to third
postoperative day, which may pose a problem if the patient has cardiac or renal compromise. In some cases, continued ICU care with invasive monitoring and
aggressive intervention with inotropes, pressors, fluid boluses or diuretics will be necessary.
Urinary output more than 0.5 cm3 kg–1 h–1 is considered acceptable. If intravascular fluid volume is required, the usual considerations apply relative to hemodilution
and translocation of fluid (see Chapter 7). Because colloid (albumin, hetastarch, plasma) tends to remain in the intravascular compartment longer than crystalloid,
administration of colloid will reduce the total amount of volume required by up to 66%. This can be an important consideration in the patient with preexisting cardiac
compromise, or when pulmonary edema is likely (posttransplant, postpneumonectomy). To maintain oxygencarrying capacity, the patient’s hematocrit should be
followed, and transfusions given when necessary (generally if the hematocrit drops below 25% in a healthy patient or below 30% in a patient with coronary disease or
severe organ system compromise).
Most patients who have undergone routine pulmonary resections can begin oral intake within the first 24 h. Patients who have undergone a pneumonectomy can
develop mediastinal shift, diaphragmatic elevation or alteration of the esophageal hiatus, and are at higher risk for aspiration. As aspiration is a lifethreatening
complication (in any patient, but particularly in the pneumonectomy patient), it is helpful to assess swallowing and the gag reflex before the resumption of oral intake.
Consideration should be given to early institution of enteral or
Page 352
parenteral nutrition in patients with complex clinical problems, particularly those expected to be nil per os (NPO) for more than 48 h, those expected to require long
term ventilation and those with preoperative cachexia. Patients who suffered from dysphagia preoperatively are often severely malnourished and parenteral or enteral
support via jejunostomy tube should be employed until recovery is complete10.
21.7 Prolonged ventilator support

Only a small minority of thoracic surgery patients requires postoperative ventilation. The goals in these patients are to provide enough ventilation (air movement) to
prevent CO2 retention and consequent respiratory acidosis, and distend the parenchyma sufficiently to prevent hypoxemia. In patients with otherwise healthy lungs, a
tidal volume of 8–10 ml kg–1 delivered at a rate of 8–10 breaths min–1 with 3–5 cmH2O of PEEP will accomplish these goals. In thoracic surgery patients, reduction of
barotrauma becomes an additional consideration, which limits both total inflation volume and PEEP. Low tidal volumes (6 ml kg–1) are recommended in the population
at risk for ARDS, but this approach has not been well studied in routine thoracotomy patients. The normal inspiratory to expiratory ratio is about 1:2, and inspiratory
times longer than 1 s are poorly tolerated in awake patients. Longer inspiratory times will reduce peak airway pressure, but require addition of sedative agents, and in
patients with significant airway obstruction may not allow sufficient time for exhalation, resulting in autoPEEP with consequent hemodynamic compromise.
Intermittent mandatory ventilation or CPAP with pressure support are commonly employed for early postoperative management, as patients can breathe
spontaneously with some support as they awaken from anesthesia and metabolize residual neuromuscular blockers. The FiO2 in the early postoperative period is
generally set at 50–60%, and then adjusted as clinically appropriate. The combination of pulse oximetry and endtidal carbon dioxide monitoring may obviate the need
for frequent arterial blood gas sampling. Controversy still exists as to the optimal level of PEEP in the thoracic surgery patient. Low levels of PEEP (3–5 cm) may be
helpful in restoring functional residual capacity and substituting for the ‘physiologic PEEP’ of the glottis.
21.7.1 Highfrequency jet ventilation (HFJV)

This has a role in the operating room during laryngoscopy, bronchoscopy, microlaryngeal procedures, and airway surgery. The role of HFJV in the intensive care unit,
particularly for management of hypoxemic respiratory failure, is poorly defined. With bronchopleural fistulae, required ventilation can sometimes be attained with HFJV
at lower airway pressures than when using conventional ventilation. The reduction in ventilation pressure will ideally minimize the amount of air passing through the fistula
and may promote healing by allowing adjacent tissues to approximate and possibly seal the fistula. In the face of decreased pulmonary compliance, the beneficial effect
of HFJV in lowering airway pressure may be lost11.
21.7.2 Independent lung ventilation

With adequate separation of the right and left lungs, it is possible to ventilate each lung independently. Some commercially available ventilators (such as the Siemens
Servo 900C) can be linked together in a masterslave combination allowing a synchronized respiratory cycle, with independent control of CPAP, tidal volume, and
pressure limits for each lung. A single ventilator may also be used to ventilate one lung while a small amount of CPAP is applied to the other lung. These systems allow
the establishment of differential airway pressures. This is beneficial in the setting of bronchopleural fistula, or other unilateral lung pathology, where application of
sufficient pressure to ventilate the diseased lung could produce overdistention and mismatching in the healthy lung. Disadvantages of this technique include the
difficulty of placing and maintaining the double
Page 353
lumen endotracheal tube, and the limitations of suctioning and bronchoscopy through the narrower individual lumens of the doublelumen endotracheal tube. This
technique should be utilized for the shortest time possible, usually only when the differential level of pressure between the lungs is greater than 10 cmH2O. Ventilation
techniques such as pressurecontrolled inverse ratio ventilation (PCIRV) can sometimes achieve reexpansion of atelectatic segments even when there is substantial
inhomogeneity of lung tissue.
21.7.3 Respiratory therapy

Patients presenting for general thoracic procedures often have significant underlying COPD, impairment of mucociliary clearance, excessive secretions, and increased
closing volumes, all of which predispose to atelectasis. The respiratory therapist plays an important role in providing intensive care with secretion management and chest
physiotherapy (percussion and vibration). Other modalities supporting recovery include humidified oxygen, adequate hydration, aerosolized bronchodilators, and early
identification and treatment of infection of the tracheobronchial tree. Instruction for chest physiotherapy should begin preoperatively and treatment started as soon as
possible postoperatively. Many patients will be unable to cough effectively, unless provided with adequate analgesia. Mucolytic agents (such as Nacetylcysteine) are
helpful in solubilizing thick secretions, but may cause bronchospasm. Oral or nasotracheal suctioning is reserved for selected patients because of discomfort and the
possibility of complications (hypoxemia, vagalmediated bradycardia or cardiac arrest). The minitracheostomy (bedside percutaneous cricothyroidotomy for suctioning)
can occasionally be utilized for access to the lower airway in patients with thick secretions. Inadequate clearance of secretions occasionally requires flexible
bronchoscopy, which is of greatest benefit in the extubated patient who cannot adequately be suctioned. When pulmonary parenchymal involvement is confined to one
lung, altering body position can improve gas exchange, by changing the relationships between ventilation and perfusion. The lateral decubitus position, with the
uninvolved lung down, allows maximal blood flow to ventilated areas during spontaneous ventilation. This relationship may be altered with mechanical breaths and
application of PEEP. In postoperative pneumonectomy patients, placing the operative side down may improve ventilation of the remaining lung. Specialized beds can be
set to supine, lateral or rotating modes to optimize oxygenation12.
When possible, patients should be positioned in a semiFowler’s position to maximize efficiency of cough and clearance of secretions13.
21.7.4 Postoperative analgesia

The majority of patients undergoing a lateral thoracotomy will require medication for adequate pain relief, and to minimize the splinting and atelectasis that will otherwise
result. Patients undergoing median sternotomy appear to require substantially lower amounts of postoperative narcotics than those undergoing lateral thoracotomy,
partly as a result of the nature of the incision and more secure bony closure. Systemic pain relief may be accomplished with intravenous opiates such as morphine and
meperidine; however, the potential for respiratory depression exists. Nonopiate agents, such as the parenteral prostaglandin inhibitor ketorolac14, or selective cyclo
oxygenase (COX) inhibitors such as roficoxib or celecoxib offer a viable alternative to opioids.
Epidural analgesia is highly effective compared with systemic morphine in postthoracotomy patients and improves respiratory function15. Epidural catheters may be
placed at either the lumbar or thoracic level, although locations above the cauda equina require an added degree of operator experience. Although a single bolus
injection of preservativefree morphine provides 6 h or more of effective analgesia, continuous epidural infusions of opiates with or without local anesthetics can
Page 354
provide superior control for up to several days. Disadvantages to epidural analgesia are the potential for inadvertent entry of the needle or catheter into the
subarachnoid space (‘wet tap’) with subsequent spinal headache, urinary retention, pruritus (common), and delayed respiratory depression as the result of circulation of
opiates from the epidural space to the brainstem respiratory center.
Intrathecal administration of morphine has also been used extensively in the thoracotomy population, and offers the advantage of ease of administration. Like epidural
administration, intrathecal administration carries the risk of delayed respiratory depression. Typical doses in the thoracotomy patient are 0.25–0.5 mg of preservative
free morphine in normal saline given at the 13 or 14 interspace16.
Intercostal nerve blocks effectively relieve postoperative pain and muscle spasm. These may be performed intraoperatively (often upon opening as ‘preemptive
analgesia’) or in the ICU. Percutaneous intercostal block (Figure 21.3) is generally performed in the posterior lateral area at the level of the surgical incision, and at
interspace; above and below the site. If pain occurs at the chest tube insertion site(s), these may also be treated. Intercostal nerve blocks are relatively contraindicated
in the postpneumonectomy patient, because of the risks of entering or contaminating the empty chest cavity, and also because splinting on the side of the
pneumonectomy may actually be beneficial in reducing atelectasis in the remaining lung. With a properly performed intercostal block, analgesia will generally last 6–12
h, and in some patients a single procedure is all that is required. Contraindications to intercostal block include chest wall resection with loss of appropriate landmarks,
infection or tumor at the injection site and pneumonectomy.
Intrapleural catheters (Figure 21.4) have been used to administer local anesthetics following thoracic, breast, renal and upper abdominal surgery and in the treatment
of multiple rib fractures. A catheter is inserted in the posterior
Figure 21.3 (a) The intercostal bundle is located by first Identifying the inferior border of the adjacent rib. (b) The needle is then
advanced below and deep to the inferior rib border. From: Contemp. Surg. (1992) front cover, June.
pleural cavity and threaded towards the apex of the lung, and local anesthetics (commonly bupivacaine or lidocaine) are supplied by intermittent bolus or continuous
infusion. A single 20 ml injection of 0.5% bupivacaine with epinephrine can be expected to supply between 3 and 10 h of pain relief17. Complications include difficulty
in placement, pneumothorax, toxicity to the local anesthetic (often in patients with abnormal pleura), and tachyphyl
Page 355
Figure 21.4 Placement of intrapleural catheter. Blind percutaneous placement is associated with a high incidence of lung puncture, so
preferred placement is under direct vision in :the operating room prior to full lung reexpansion. The Tuoheytype needle,
available in epidural anesthesia kits, must be placed over the top of the rib to avoid damage to the neurovascular bundle. From:
Rice, T.W., Higgins, T.L., Kirby, J.J. (1995) Management of the general thoracic surgical patient. In: Sivak, E.D., Higgings, ins,
T.L., Seiver, A. (eds). The high risk patient: Management of the critically ill. Reproduced with permission from Lippincott,
Williams and Wilkins, Baltimore.
axis to the local anesthetic with time. Contraindications to interpleural catheters include pleural effusion (which will dilute the local anesthetic), and pleural fibrosis.
21.8 Specific postoperative complications

General and specific complications are listed in Tables 21.3 and 21.4.
21.8.1 Airway complications

These occur with prolonged intubation using large or doublelumen endotracheal tubes, the passage of bronchial blockers, rigid bronchoscopes, or frequent
reintubation. All can
Table 21.3 Postoperative complications following
Airway edema or stridor

Arrhythmias (especially atrial fibrillation,
multifocal atrial tachycardia)
Arytenoid dislocation
Aspiration of gastric contents
Atelectasis
Bronchospasm
Bronchopleural fistula
Chylothorax
Congestive heart failure
Deep venous thrombosis
Empyema
Hemorrhage
Hemothorax
Infection (superficial, deep)
Lobar collapse
Lobar torsion
Myocardial infarction
Pain and splinting
Pleural effusion
Pneumothorax
Pulmonary embolus
Reexpansion pulmonary edema
Respiratory failure
Retaining secretions
Subcutaneous emphysema
Tension pneumothorax
From: Higgins, T.L and Chernow, B. (1995) Receptor physiology and pharmacology in circulatory shock. In: Sivak, E.D., Higgins, T.L. and Seiver, A. The High Risk Patient: Management
of the Critically ill. Lippincott, Williams and Wilkins, Baltimore.
result in trauma or edema of the larynx or trachea. Routinely assessing the patient for an air leak around an occluded endotracheal tube just before extubation can help
identify laryngeal and supraglottic edema. Upright or sitting position, intravenous corticosteroids and racemic epinephrine respiratory treatments are the mainstay of
edema reduction. A critical airway may be converted to an adequate airway by the administration of heliox, a helium and oxygen mixture18. Helium, being denser and
less viscous than nitrogen, allows maintenance of laminar flow through the critically swollen upper airway. Prolonged endotracheal intubation or temporary
tracheostomy may be required to allow resolution of airway edema.
Page 356
Table 21.4 Complications of specific thoracic procedures
Procedure Complications
Anterior mediastinotomy (Chamberlain) Damage to recurrent laryngeal nerve (particularly left)
Bronchoscopy or mediastinoscopy Bleeding from major vessels if torn, air leak (rare)
Bronchopleural fistula repair Persistent leak, dehiscence
Bronchopulmonary lavage Respiratory distress or contralateral spillage
Bullectomy Tension pneumothorax; air leak
Chest wall reconstruction Blood loss, altered chest wall compliance, unstable chest, infection of prosthetic material
CollisBelsey hiatal hernia repair Gastric leak
Decortication Blood loss, air leak(s)
Esophageal dilation Esophageal perforation, pleural effusion, elevated CPK
Esophagoscopy Esophageal perforation
Esophagogastrectomy 'Thirdspacing' of fluids, anastomotic leak, gastric devascularization, splenic injury, gastric torsion, elevated CPK
Heller myotomy Esophageal tear
Lobectomy Bronchial leak, lobar collapse, lobar torsion
Mediastinal tumor excision Airway obstruction with sedation or anesthesia; damage to recurrent laryngeal nerve
Nissen fundoplication Esophageal obstruction (with tight wrap), splenic injury
Pectus repair Costrochondritis, unstable sternum
Pneumonectomy Atrial arrhythmias (atrial fibrillation, MAT), mediastinal shift, cardiac torsion, air embolism, disrupted bronchus)
Thoracic aortic aneurysm Paraplegia, bleeding, aortobronchial fistula, esophageal injury
Thymectomy In myasthenics, possible weakness and respiratory failure
Lung transplant Rejection (day 5), reperfusion injury, infection, overdistention of native lung, dehiscence
Tracheal resection Fixed neck flexion postoperatively; dehiscence, air leak
From: Higgins, T.L. and Chernow, B. (1995) Receptor physioiogy and pharmacology in circulatory shock. In: Sivak, E.D., Higgins, T.L and Seiver, A. The High Risk Patient: Management
of the Critically III. Lippincott, Williams and Wilkins.
21.8.2 Damage to recurrent laryngeal nerves

The recurrent laryngeal nerves originate from the vagus nerves as they enter the chest. The right recurrent laryngeal nerve arises high in the apex of the right chest and
loops around the right subclavian artery to travel back to the larynx in the tracheoesophageal groove. The left recurrent laryngeal nerve, which is more susceptible to
injury, wraps around the aortic arch in the left chest before it enters the tracheoesophageal groove. Excessive traction, aggressive dissection about these nerves, or the
possible sacrifice of these nerves can result in postoperative palsy. Mediastinoscopy, anterior mediastinotomy, left pulmonary resection
Page 357
with subaortic exenteration and resections of mediastinal tumors are common operations in which the recurrent laryngeal nerve may be damaged. Associated airway
and laryngeal edema may allow for adequate coaptation of the vocal cords for the first days postextubation preventing identification of cord injury until after discharge
from the ICU, resulting in ineffective cough or aspiration of secretions. If there is permanent damage or division of the recurrent laryngeal nerve then augmentation of the
vocal cord with a longlasting substance such as Teflon may be considered. In many instances, aggressive chest physiotherapy, careful airway management, and
temporary avoidance of oral feeding may eliminate the need for any intervention until recovery of the nerve function has occurred. Intermittent noisy inspiration and
painful swallowing characterize arytenoid dislocation, an uncommon cause of postextubation respiratory failure19. Treatment consists of surgical reduction, using gentle
pressure with a laryngeal spatula, and must be accomplished before the cricoarytenoid joint becomes fibrosed in poor position.
21.8.3 Retained secretions and blood in the airway

These are especially common if the airway was opened, as during a bronchoplastic procedure or closure of a bronchial stump. If there have been excessive or bloody
secretions in the airway, flexible fiberoptic bronchoscopy with clearance of secretions is wise before extubation. The mechanical airway obstruction secondary to
secretions may be aggravated by bronchospasm. Many patients have reactive airways and their preoperative bronchodilators should be continued in the perioperative
and postoperative course, as secretions can precipitate coughing and bronchospams.
21.8.4 Postoperative air leaks

Most postoperative air leaks are the result of very distal fistulae between tiny bronchioles or respiratory units and the pleural cavity. One of the main functions of the
chest tube is to evacuate air from these small air leaks and assure complete expansion of the lung and coaption of the cut surface of the lung to the parietal pleura. With
complete expansion of the lung, these small air leaks will seal. Careful examination of the chest Xray, insertion of additional chest tubes into undrained spaces,
adequate suction applied to the pleural cavity, and full expansion of the lung with vigorous chest physiotherapy help to close these small distal fistulae. If there is a
substantial persistent air leak from the chest tube, or there is incomplete or no expansion of the lung in the pleural cavity, a significant bronchopleural fistula should be
suspected. Major proximal airway problems such as failure of the anastomosis, disruption of a bronchial closure, or retained secretions or foreign bodies should be
excluded early in the postoperative course by bronchoscopy. Within the first seven postoperative days, the fistula is usually secondary to a technical problem. In the
subacute phase, usually more than 1 week after the operation but usually within the first 6 weeks, the fistula is often associated with an empyema or local peribronchial
abscess.
Occurrence of a bronchopleural fistula in the early postoperative course of a pneumonectomy patient is a surgical emergency. The typical presentation is the sudden
expectoration of copious amounts of pink, frothy sputum. Rather than assume that this is pulmonary edema, a bronchopleural fistula should be suspected. The patient
should be positioned with the operated or pneumonectomy side down, to trap remaining fluid in the pneumonectomy space and protect the patient from drowning. A
chest Xray will show loss of fluid from the pneumonectomy space. A chest tube should be placed into the pneumonectomy cavity, remembering that there may be
significant shift in the mediastinum and diaphragm; the tube should be placed high in the midaxillary line. Further management, by the thoracic surgeon, will probably
include bronchoscopy to assess the stump closure, and probably reoperation.
Page 358
21.8.5 Empyema
This is initially treated with closed tube drainage and antibiotic therapy. Once the patient has been stabilized and any bronchopleural fistula identified and treated,
drainage of the empyema cavity is converted from closed tube drainage to open tube drainage. A chest Xray is then taken to determine if the mediastinum is fixed or
has shifted and compressed the contralateral remaining lung. If the mediastinum is stable, then the drainage of the cavity may be permanently converted to open
drainage. This may take the form of rib resection and marsupialization of the pneumonectomy cavity with later closure as the pneumonectomy cavity shrinks in size,
using either the Clagett procedure or obliteration of the space with a muscle flap.
21.8.6 Pulmonary or systemic tumor emboli

In the postoperative period, pulmonary emboli are sometimes difficult to diagnose as hypoxemia may be due to sepsis, ARDS or pneumonia. If pulmonary emboli are
suspected, ventilationperfusion scanning, spiral computed tomography or pulmonary angiography should be performed. Treatment is anticoagulation or lytic therapy
(Chapter 12). If these are contraindicated, then an inferior vena caval filter should be placed. Systemic tumor emboli, although uncommon, may be seen after pulmonary
resections for primary bronchogenic carcinomas or metastatic sarcomas.
21.8.7 Massive postoperative hemorrhage

This can be a lethal complication, and can manifest as significant shock, and generally requires emergent reoperation. Slower postoperative hemorrhage is more
commonly the result of small bleeding arteries or veins in the mediastinum or chest wall. Reoperation may be required for control of bleeding. Failure of the hemothorax
to resolve can necessitate later decortication to prevent fibrothorax and restrictive lung disease in the future.
21.9 Pulmonary parenchymal complications
21.9.1 Atelectasis
This is the most common complication following thoracic surgery and can occur as a result of splinting, hypoventilation, bronchospasm, inadequate cough, retained
secretions, and intraoperative trauma to the lung. The increased work of breathing and impaired gas exchange, as a result of the inherent ventilationperfusion mismatch,
may precipitate respiratory failure in a patient with marginal pulmonary reserve. Retained bronchial secretions and the frequent colonization of the airway in patients with
COPD facilitates the progression of atelectasis to postoperative pneumonia. With clinically significant atelectasis, the patient may demonstrate tachypnea, tachycardia,
arrhythmias, hypoxia, or respiratory failure. Atelectasis alone, however, is unlikely to be the source of postoperative fever20. Percussion and auscultation may reveal the
signs of a consolidated portion of lung. The chest Xray may show patterns of platelike atelectatic collapse, to segmental, lobar, or total atelectasis of the lung. Although
the operative side is more frequently involved, contralateral lung and bilateral involvement may be seen. Treatment (incentive respirometry, chest physiotherapy
inhalation therapy) is directed at expanding the collapsed lung and clearing the airway of retained secretions. Routine bronchoscopy has been shown not to be
advantageous; however, it has a role in established atelectasis21.
21.9.2 Lobar collapse

This is most commonly seen following right upper lobectomy or right upper plus right middle bilobectomy, because of altered bronchial geometry. Following right upper
lobectomy, the horizontal fissure rises and rotates to lie perpendicularly along the mediastinal pleura, and the resultant twisting and kinking of the long and narrow
middle lobe bronchus renders it susceptible to occlusion. This complication should prompt early bronchoscopy to rule out
Page 359
torsion and repeated bronchoscopy to remove retained secretions.
21.9.3 Lobar torsion

This is an uncommon complication, which results from the twisting of a mobile piece of pulmonary parenchyma about its hilar structures with resultant bronchial, arterial,
and venous occlusion. Torsion also occurs most commonly following right upper lobectomy when the middle lobe has no anchoring attachments save the middle lobe
bronchus, artery and vein. This onepoint fixation and the absence of the upper lobe allow the middle lobe to freely rotate. Lobar torsion may also be seen after
bilobectomy on the right with torsion of the remaining upper or lower lobe, or after lobectomy on the left with torsion of the remaining lobe. The patient may be
asymptomatic, or may present with atelectasis often radiographically more than expected in the early postoperative period. Later in the course, the Xray may show
complete ‘white out’ of the involved segment or lobe. If ischemia progresses to infarction, biochemical signs of tissue necrosis may be seen, and the patient may cough
up bloody, purulent, malodorous secretions. Similar chest tube drainage may be found22. Bronchoscopy will demonstrate abrupt occlusion of the involved airway;
however, it may be possible to pass the bronchoscope distally into the twisted airway. Pulmonary angiography and perfusion scans will demonstrate lack of flow to the
involved portion of lung; however, these tests are not diagnostic. Once diagnosed, urgent thoracotomy with detorsion and likely resection of necrotic tissue is required.
21.9.4 Reexpansion pulmonary edema

This may occur following rapid removal of a large pleural effusion or after reexpansion of a pneumothorax. It is characterized by unilateral alveolar pulmonary edema,
which resolves rapidly23.
21.9.5 Respiratory insufficiency

This may be irreversible and due to inadequate pulmonary reserve, as the result of too aggressive a resection, either from preoperative assessment or technical
problems that required a larger than predicted resection. If there is failure to wean from the ventilator and to extubate, arrangements must be made for tracheostomy,
feeding tube, and longterm ventilator support (see Chapter 18).
Respiratory failure may be reversible, the result of superimposed treatable complication, such as infection, fluid overload, aspiration, pulmonary edema, pulmonary
embolus or pneumothorax.
21.10 Pleural complications

Fluid and air may collect in the pleural space and require correct positioning of a sufficient number of appropriately sized chest tubes. It is often difficult to adequately
place the tube with only twodimensional imaging as with a portable chest Xray. CT scanning with placement after or during the scan is beneficial, as is drainage with
surface ultrasound guidance.
Pneumothorax post thoracic surgery can occur with or without an air leak. The drainage system should be checked to guarantee it is closed with an adequate
underwater seal. Chest tube patency should next be ascertained. If the chest tube and drainage system are functioning properly then a loculated pneumothorax with no
communication with the drainage system or large bullae are likely explanations. A persistent pneumothorax not associated with an air leak or signs of sepsis is generally
not a problem, but may respond to increased suction on the present system. A persistent pneumothorax associated with an air leak indicates an established
bronchopleural fistula.
Significant persistent pleural fluid drainage should prompt analysis of the fluid. A transudative pleural effusion can be drained by thoracentesis, with attention to
treatment of the extrathoracic cause. An exudative pleural effusion leads to consideration of an empyema even if cultures of the fluid are negative.
Chylothorax can complicate operations in the chest about the aorta, esophagus or lung.
Page 360
Any dissection in the posterior mediastinum may result in an injury to the thoracic duct. If there is a suspected injury to the thoracic duct, and this injury cannot be
found, then supradiaphragmatic ligation of the thoracic duct is an acceptable approach. Damage to the thoracic duct noted can present as highoutput fistula with
volumes of 1–3 l day–1, but diagnosis can be delayed as volume will be minimal while the patient remains NPO. The diagnosis should be confirmed by analysis of the
fluid, which will have a high protein content, a large proportion of lymphocytes, and elevated concentrations of fat and chylomicrons. A contrast lymphangiogram or
nuclear lymphangiogram using technetium99m antimony colloid is usually not necessary, but may be used to locate the site of injury24. Initial management consists of
placing the thoracic duct at rest (patient NPO) and complete drainage of the pleural space to promote full expansion of the lung. Nutrition is provided with either
intravenous hyperalimentation or feedings through the GI tract using mediumchain triglyceride (MCT) formulas. If the chylothorax does not resolve, then thoracotomy
with ligation of the thoracic duct is indicated.
21.11 Other considerations
21.11.1 Infection of thoracic incisions

This is uncommon. The incidence increases following operations where there is a septic complication, operations for empyema, lung abscess and mediastinitis, or if the
GI tract is opened as for a perforated esophagus.
21.11.2 Subcutaneous emphysema

This is the result of tracking of air into the subcutaneous space and dissecting along the path of least resistance. Positive pressure ventilation and the positive pressure
generated on expiration or coughing generate the driving pressure. Although alarming, subcutaneous emphysema does not generally adversely affect the patient’s
outcome. Treatment consists of the placement of a properly functioning chest tube into the pleural space, and suction drainage. Rarely, massive emphysema in a non
intubated patient will compromise the airway and require urgent intubation or possibly tracheostomy for control of the upper airway. Release of paratracheal air will
often permit emergent endotracheal intubation. The patient with subcutaneous emphysema should be assured this temporary problem would resolve.
21.11.3 Patients with myasthenia gravis

These patients may be safely extubated in the operating room or shortly after arrival in the ICU if neuromuscular blockade is monitored and adequately reversed. Less
than 9% require prolonged mechanical ventilation, which may range from 15 h to 3 days25; with routine preoperative plasmapheresis, the incidence of postoperative
problems may be even smaller. The need for postoperative mechanical ventilation can be predicted by:
• duration of myasthenia gravis greater than 6 years;
• history of chronic respiratory disease not related to myasthenia;
• pyridostigmine dose greater than 750 mg day–1;
• preoperative vital capacity less than 2.9 l.
21.11.4 Esophageal surgery

The major complication following esophageal surgery includes leakage of esophageal contents from the surgical site. In 138 patients undergoing gastric interposition
following transhiatal esophagectomy, the leak rate was 11%, with aspiration occurring in 4%, and gastric perforation in 1.5%26. More recent literature quotes even
lower rates of leakage27.
Early esophageal perforation is usually managed with thoracotomy, reinforced closure with a pedicle wrap and drainage. Manipulation of the esophagus generates
release of all three isoenzymes of creatine kinase (CPK), limiting the use of CPK for diagnosis of perioperative myocardial infarction.
21.11.5 Reasons for ICU readmission

The major reason for ICU readmission is respiratory decompensation, as a result of fluid
Page 361
overload, aspiration, or failure to cough and breathe well enough to prevent atelectasis and retained secretions. Atelectasis is a continuing threat, and can be reduced
with the use of incentive spirometry. Bronchodilators and mucolytics are added as needed. More intensive intervention, in the form of percussion, postural drainage,
and nasotracheal suctioning, may be needed when secretions are thick or atelectasis persistent. Followup care should include periodic chest radiographs, and at least
daily physical examination.
References
1. Nunn, J.F., Miledge, S., Chen, D. and Dore, C. (1988) Respiratory criteria for fitness for surgery and anesthesia. Anaesthesia 43:543–611.
2. Garibaldi, R.A., Britt, M.R., Coleman, M.L., Reading, J.C. and Pace, N.L. (1981) Risk factors for postoperative pneumonia. JAMA 70:677–680.
3. Stein, M. and Cassara, E.L. (1970) Preoperative pulmonary evaluation and therapy for surgery patients. JAMA 211:787–790.
4. Wahba, W.M. (1983) Influence of aging on lung function—clinical significance of changes from age twenty. Anesth. Analg. 62:764–776.
5. Edner, H., Sudkamp, N., Wex, P. and Dragojevic, D. (1985) Selection and preoperative treatment of overseventyyearold patients undergoing thoracotomy.
Thorac. Cardiovasc. Surg. 33:268–271.
6. Warner, M.A., Offord, K.P., Warner, M.E., Lennon, R.L., Conover, M.A. and JanssonSchumacher, U. (1989) Role of preoperative cessation of smoking and
other factors in postoperative pulmonary complications: A blinded, prospective study of coronary artery bypass patients. Mayo Clin. Proc. 64:609–616.
7. Shapiro, B.A., Harrison, R.A., Kacmarek, R.M. and Can, R.D. (1985) Clinical Application of Respiratory Care, 3rd Edn. Year Book, Chicago, IL.
8. Palesty, J.A., McKelvey, A.A. and Dudrick, S.J. (2000) The efficacy of xrays after chest tube removal. Am. J. Surg. 179:13–16.
9. DeHaven, C.B., Hurst, J.M. and Branson, R.D. (1986) Evaluation of two different extubation criteria: Attributes contributing to success. Crit. Care Med. 14:92.
10. Stizman, J.V. (1990) Nutritional support of the dysphagic patient: Methods, risks, and complications of therapy. Parent Enternal Nutr. 14:60–63.
11. Baumann, M.H. and Sahn, S.A. (1990) Medical management and therapy of bronchopleural fistulas in the mechanically ventilated patient. Chest 97:721.
12. Nelson, L.D. and Anderson, H.B. (1989) Physiologic effects of steep positioning in the surgical intensive care unit. Arch. Surg. 124: 352–355.
13. Brandi, L.S., Bertolini, R., Janni, A. et al. (1996) Energy metabolism of thoracic surgical patients in the early postoperative period: Effects of posture. Chest
109:630–637.
14. Yee, J.P., Koshiver, J.E., Allbon, C. and Brown, C.R. (1986) Comparison of intramuscular ketorolac tromethamine and morphine sulfate for analgesia of pain
after major surgery. Pharmacotherapy 6 (5):253–261.
15. Shulman, M., Sandler, A.N., Bradley, J.W., Young, P.S. and Brebner, J. (1984) Postthoracotomy pain and pulmonary function following epidural and systemic
morphine. Anesthesiology 61:569–575.
16. Nordberg, G., Hedner, T., Mellstrand, T. and Dahlstrom, B. (1985) Pharmacokinetic aspects of intrathecal morphine analgesia. Anesthesiology 60:448–454.
17. Reiestad, F. and Stromskag, K.E. (1986) Interpleural catheter in the management of postoperative pain—A preliminary report. Reg. Anesth. 11:89–91.
18. Skrinskas, G.J., Hyland, R.H. and Hutcheon, M.A. (1983) Using heliumoxygen mixtures in the management of acute upper airway obstruction. Can. Med.
Assoc. J. 128: 555–558.
19. Castella, X., Gilabert, J. and Perez, C. (1991) Arytenoid dislocation after tracheal intubation: An unusual cause of acute respiratory failure? Anesthesiology
74:613–615.
20. Marik, P.E. (2000) Fever in the ICU. Chest 117: 855–869.
21. Marini, J.J., Pierson, D.J. and Hudson, L.D. (1979) Acute lobar atelectasis: A prospective comparison of fiberoptic bronchoscopy and respiratory therapy. Am.
Rev. Respir. Dis. 119: 971–978.
22. Mulin, M.J., Zumbro, G.L., Fishback, M.E., et al. (1972) Pulmonary lobar gangrene complicating lobectomy. Ann. Surg. 175:62–66.
23. Hymphreys, R.L. and Berne, A.S. (1970) Rapid reexpansion of pneumothorax: a cause of unilateral pulmonary edema. Radiology 96: 509–512.
24. Rice, T.W., Kirsh, J.C., Schacter, I.B. and Goldberg, M. (1987) Simultaneous occurrence of chylothorax and subarachnoid pleural fistula after thoracotomy. Can.
J. Surg. 30:256–258.
25. Eisenkraft, J.B., Papatestas, A.E., Kahn, C.H., et al. (1986) Predicting the need for postoperative mechanical ventilation in myasthenia gravis. Anesthesiology
65:79–82.
Page 362
26. Agha, E.P., Orringer, M.B. and Amendola, M.A. (1985) Gastric interposition following transhiatal esophagectomy: radiographic evaluation. Gastrointest Radiol.
10:17–21.
27. Adoumie, R., Shennib, H., Brown, R., Slinger, P. and Chiu, R.C. (1993) Differential lung ventilation. Application beyond the operating room. J. Thorac.
Cardiovasc. Surg. 105:229233.
Page 363
Chapter 22
Thoracic trauma
Imtiaz A.Munshi, MD
Contents
22.1 Introduction
22.2 Anatomy
22.3 Pathophysiology
22.4 Initial evaluation—ABCs
22.5 Injuries to the chest wall
22.6 Injuries to the bony thorax
22.7 Injuries to the lung
22.8 Tracheobronchial injury
22.9 Blunt aortic injury
22.10 Blunt cardiac injury
Page 364
22.1 Introduction
Thoracic injuries are directly responsible for over 25% of traumarelated deaths in the United States each year. Complications from thoracic trauma can contribute
another 20% to overall traumarelated deaths1. The mechanism of injury and type of death are related (Table 22.1). Prompt diagnosis and treatment can prevent the
early deaths and reduce the rates of complications and late deaths.
The most common type of thoracic trauma is injury to the ribs. Over 85% of patients with thoracic injuries can be successfully managed ‘conservatively’ with
adequate analgesia, tube thoracostomy, judicious fluid management and ventilatory support. The remaining 15% of patients with blunt and penetrating trauma require
thoracotomy to treat injuries2,3.
22.2 Anatomy4
With suspected thoracic injury, it is helpful to think of the thorax as being composed of three units: the chest wall, the lungs and the mediastinum. The bony chest wall is
bounded anteriorly by the sternum, laterally by the ribs and the costal cartilages, and posteriorly by the vertebral column (Figure 22.1). The wall of the chest is made
up of layers of muscle and fascia supported by the thoracic skeleton. The
Table 22.1 Time sequence of mortality from chest trauma
Immediate death
• Disruption of the heart or aorta
Early death (minuteshours)
• Cardiac tamponade
• Tension pneumothorax
• Airway obstruction
• Uncontrolled hemorrhage
Late death (daysweeks)
• Infections (empyema, pneumonia)
• Missed tracheobronchial or esophageal injuries
• Multisystem organ failure (MSOF)
• Acute respiratory distress syndrome (ARDS)
nerves, blood vessels and lymphatics of the chest wall pass within these layers. The chest or thoracic cavity is separated into two hemithoraces by a broad median
septum called the mediastinum (Figure 22.2). The mediastinum is divided into superior, anterior, middle and posterior regions and contains numerous vital structures
including the heart, great vessels, esophagus and the tracheobronchial tree. The inferior thoracic aperture is bounded by the diaphragm, a muscle that separates the
thoracic and abdominal cavities. The position of the diaphragm is dynamic; upon exhalation the right leaflet rises to the fourth intercostal space and the left to the fifth
intercostal space; during inhalation both expand to reach the eight intercostal space.
22.3 Pathophysiology
22.3.1 Blunt injury

One factor common to blunt injuries to the bony thorax is pain, aggravated by movement. Pain by itself can lead to a decrease in vital capacity and the inability to clear
secretions, which ultimately manifest as hypoxia and hypercarbia. From the mechanical standpoint bony thoracic injuries such as rib fractures and flail chest result in
ventilationperfusion abnormalities, increase respiratory work and decrease functional residual capacity. Untreated pain thus contributes to worsening hypoxia,
hypercarbia and respiratory insufficiency. Lung contusion, hemothorax or pneumothorax also cause ventilationperfusion mismatching and decrease in lung compliance.
Associated hypovolemia secondary to hemothorax can compromise oxygen delivery to tissues, resulting in metabolic acidosis. Methods of pain relief including
intercostal nerve block and epidural medications for thoracic surgical patients are discussed in Chapter 21.
22.3.2 Penetrating injury

Penetrating injuries to the chest are divided into stab wounds and gunshot wounds. The gunshot wounds are further classified into low
Page 365
Figure 22.1 Anatomy of the chest wall.
Figure 22.2 Anatomy of the thoracic cavity.
and highenergy injuries based on the velocities of the missiles. Highvelocity missiles travel at more than 1500 ft s–1 and cause significant tissue destruction secondary
to blast effect. Patients with injuries from these weapons often require early operative intervention with debridement and resection of devitalized tissues. Patients with
lowvelocity wounds and stab wounds have little blast effect and primarily local tissue damage, which may be managed conservatively. Like blunt injury, penetrating
injuries can cause lung contusion, pneumothorax and hemothorax, resulting in hypoxia, ventilationperfusion mismatching and decrease in lung compliance.
22.4 Initial evaluation—ABCs

Initial management of all trauma patients follows the guidelines outlined by the American
Page 366
College of SurgeonsCommittee on Trauma (ACS–COT) in the Advanced Trauma Life Sup– port (ATLS) Manual5. Assessment of a trauma patient begins with the
ABCs (airway, breathing and circulation), securing a patent airway while simultaneously ensuring adequate ventilation; insertion of largebore intravenous catheters;
administration of crystalloid solutions; control of obvious sources of hemorrhage; maintenance of cardiac output and evaluation of the patient for lifethreatening injuries
such as tension pneumothorax, cardiac tamponade, hemothorax, and pneumothorax. Subsequently, a complete, rapid physical examination is performed. Physical
examination findings of chest trauma include crepitus, jugular venous distention, abnormal chest wall movement, diminished or absent breath sounds, diminished heart
sounds or new heart murmurs, dullness to percussion of the chest wall or percussion tympany, and identification and localization of wounds or bruises on the chest wall.
22.5 Injuries to the chest wall
22.5.1 Soft tissue injuries

The diagnosis of chest wall wounds is often obvious. Patients present with symptoms of pain from the wound and may exhibit respiratory insufficiency secondary to an
associated pneumothorax or hemothorax. Chest wall wounds must never be probed because of the risk of causing further injury to underlying structures, resulting in
bleeding or pneumothorax. Small open chest wounds can act as oneway valves with air entering during inspiration and not exiting during exhalation, thus creating a
pneumothorax (‘open pneumothorax’). A resultant decrease in tidal volume and venous return to the heart occurs, leading to further compromise in cardiopulmonary
status. Chest wounds exceeding twothirds the diameter of the trachea result in ineffective ventilation as air preferentially moves across the chest wall defect rather than
through the trachea. Open sucking or hissing wounds should be covered with a sterile, airtight dressing while a chest tube is inserted at a separate site on the side of the
injury to avoid development of a tension pneumothorax. Chest tubes must never be inserted through existing wounds for fear of following the wound tract and injuring
underlying organs. Chest wall wounds will often require surgical debridement and irrigation, allowing the wound to heal secondarily or by delayed primary closure on
the fifth postinjury day. Large or complex wounds, particularly from shotgun blasts, require operative debridement with removal of all foreign bodies, including the
wadding from the shell. If the wound is extensive enough that closure is not possible a plastic surgical consultation for coverage with a myocutaneous flap is indicated.
Myocutaneous flaps can be fashioned from the pectoralis, latissimus dorsi, and rectus abdominus muscles6.
22.5.2 Subcutaneous emphysema

This is a manifestation of thoracic injury. It results when air from the lung or the tracheobronchial tree communicates with the soft tissues of the chest wall through an
opening in the parietal pleura. Injuries to the larynx, pharynx or esophagus may also cause subcutaneous emphysema. The air often tracks through the planes of the soft
tissues into the neck, face, back and abdomen. Swelling can reach massive proportions but is rarely life threatening and resolves without intervention. Patients with
subcutaneous emphysema should be assumed to have an underlying pneumothorax. Consequently, such patients should have a chest tube(s) placed before operation or
placement on positive pressure ventilation.
22.5.3 Traumatic asphyxia

This is caused by an acute increase in the intravascular pressure within the superior vena cava as a result of sudden compression of the chest, usually by a heavy weight.
Venous hypertension produces capillary engorgement, resulting in edema and cyanosis of the head, neck and upper torso. Diffuse reddening above the site of
compression is characteristic and is secondary to extravasation of blood from
Page 367
ruptured vessels. Subconjunctival hemorrhage and increased intracranial pressures can also be present. Treatment is administration of supplemental oxygen with
elevation of the head of the bed to 30°. Prognosis is good, and the skin, ocular and neurologic manifestations resolve over a period of several weeks7,8.
22.6 Injuries to the bony thorax
22.6.1 Simple rib fractures

These are the most common injuries seen in blunt chest trauma and are present in 25% of all injured patients. Although rarely life threatening in themselves, rib fractures
are an indicator of possible underlying intrathoracic or intraabdominal injury. In addition, patients with rib fractures have significant pain and muscle splinting, which
interferes with respiratory mechanics, leading to pulmonary complications. These complications further compromise oxygenation and ventilation. The bony thorax in
children is relatively more elastic or compliant than that of the adult. The presence of rib fractures in children should raise suspicion that the mechanism involved a
significant transfer of kinetic energy and potential child abuse. In the elderly population, rib fractures significantly increase overall morbidity and mortality, often
warranting hospital admission in an otherwise stable and uninjured patient. Each additional rib fractured in an elderly person increases mortality by 19% and the
pneumonia risk by 27%9. The presence of hypotension in a patient with low rib fractures (ribs 9–11), without hemothorax or tension pneumothorax, should raise
suspicion for intraabdominal sources of hemorrhage. Low rib fractures on the right side may injure the liver and those on the left may injure the spleen with resultant
hemorrhage.
Diagnosis of rib fractures is largely on clinical grounds and should be suspected in any patient with localized chest wall tenderness following blunt thoracic trauma. A
negative chest radiograph does not rule out rib fracture, as up to 50% of rib fractures are not visualized on anteriorposterior (AP) chest radiographs10. Chest
radiographs in patients with suspected rib fractures are obtained to search for hemoor pneumothorax, lung contusion or other injuries. Treatment of rib fractures is
largely supportive, emphasizing adequate analgesia and chest physiotherapy. Intercostal nerve blocks or epidural analgesia may be appropriate in patients with severe
pain from multiple fractures or in the elderly. Epidural catheters provide the most effective pain control with minimal adverse effects. This method is the treatment of
choice in patients with significant pain from multiple fractures and who are at high risk for respiratory complications. Strapping of the chest is not advocated because it
can promote atelectasis and reduce vital capacity. Operative repair of rib fractures with plates and wires is indicated only if the bony chest wall injury is the primary
reason for prolonged mechanical ventilation. Repair would be of no benefit if the mechanical ventilation were used to support the underlying pulmonary parenchymal
contusion.
22.6.2 Fractures of the first and/or second ribs

These usually indicate that a significant transfer of kinetic energy to the thoracic cage has occurred, as considerable force is needed to fracture these bones. Fractures of
these ribs are associated with mortality rates between 15 and 36%, higher than with any other rib fracture pattern11. Also, 40% of patients with first and/or second rib
fractures have other significant injuries such as aortic rupture, severe head injuries and cervical spine fractures12,13. Local trauma to the brachial plexus and subclavian
vessels can occur14. Operative repair of first and second rib fractures is not required.
22.6.3 Flail chest

This is defined as fracture of three or more consecutive ribs, which are each fractured in two or more areas (Figure 22.3). Mechanical malfunction of the injured chest
wall is manifested by paradoxical movement (inward collapse) of the affected segment during inspiration. Although this paradoxical motion of
Page 368
Figure 22.3 Left flail chest.
the chest wall contributes to an increase in work of breathing, the primary cause of hypoxia in these patients is the underlying pulmonary contusion15. ICU admission is
warranted for initial monitoring with flail chest. As with other bony injuries of the thorax, adequate analgesia, judicious fluid administration and chest physiotherapy are
of paramount importance.
22.6.4 Sternal fractures

These are most often the result of a direct blow, classically by impact with the steering column of an automobile or secondary to seat belt use16,17. Sternal fractures
occur with an incidence of 1.5–4%18, most often at the junction between the body and manubrium. The manubrium is the densest part of the sternum and is rarely
fractured. A fracture of the manubrium requires tremendous force and is associated with more severe underlying injuries and with higher mortality than are fractures of
the body18. Indirect forces can cause sternal fractures as a consequence of hyperflexion of the spine. Wedge compression fractures of the cervical or thoracic spine are
present in 9.6% of patients with sternal fractures19. Associated rib fractures occur in 21% of cases20.
Clinically, these patients may exhibit anterior chest wall pain, point tenderness, swelling and crepitus. Diagnosis is best confirmed with lateral chest radiograph,
although deformity may be present on AP films as well. Suspicion or presence of a sternal fracture should direct focus to the diagnosis of underlying cardiac or great
vessel injuries. Associated blunt cardiac injury can occur in up to 70% of these patients17. Cardiac workup should include serial electrocardiograms and cardiac
enzyme studies whereas echocardiography and angiography are indicated in a select few patients. Treatment of sternal fractures consists of effective analgesia. Surgical
repair is indicated in patients with severe displacement, associated costochondral disruptions, or associated flail segments in which operative fixation may permit earlier
withdrawal of mechanical ventilation21.
22.6.5 Clavicular fractures

These are a relatively common injury and are suspected in those with obvious deformity, crepitus or point tenderness (Figure 22.4). These fractures rarely result in
significant pathophysiologic respiratory changes. However, injury to the brachial plexus or subclavian vessels, although rare, can occur. Diagnosis is usually made on
AP chest radiographs. Treatment largely consists of immobilization with an arm sling and adequate analgesia. Surgical repair is unnecessary but
Page 369
Figure 22.4 Left clavicle fracture.
sometimes indicated in patients with significantly displaced fractures.
22.6.6 Scapular fracture

This, like sternal and first and second rib fractures, is often an indicator that a significant force has been inflicted upon the bony thoracic cavity. The scapulae are well
protected by surrounding musculature and are rarely injured. Clinically, pain and swelling are often the only indicators of the injury. Diagnosis is usually made on chest
radiography. However, 28% of these injuries will show little or no radiographic evidence of fracture and as many as 43% are missed on initial chest radiographs22.
Specific oblique radiographic views are necessary to confirm the diagnosis (Figure 22.5). Treatment consists of immobilization with arm sling but with early range of
motion activity.
22.7 Injuries to the lung
22.7.1 Pulmonary contusions

These are a frequent sequela of blunt and penetrating injury to the thorax. It is believed that the blunt force results in compression of the chest wall with underlying
compression of the lung parenchyma; subsequently, a rapid overexpansion with recoil of the lung tissue against the chest wall occurs, damaging the alveoli. Interestingly,
the impact can be transmitted across the mediastinum to the contralateral lung resulting in a countercoup type injury (analogous to cerebral contusions). The
pathophysiologic mechanism for the development of contusions is not well characterized. However, it is known that capillary disruption with resulting interstitial and
intraalveolar edema decrease lung compliance and increase physiologic shunt, leading to hypoxemia. The hypoxemia usually lasts between 24 and 48 h23,24. During this
time period it is imperative that chest physiotherapy and adequate analgesia be provided to prevent atelectasis.
Diagnosis of pulmonary contusion can be made by chest radiography (Figure 22.6). Approximately twothirds of patients exhibit patchy infiltration or consolidation
on radiographs as early as 6 h after injury, which then develop over the next 24 h16. The remaining onethird of patients often have evidence of contusion on Xray by
24 h after injury25. As a general rule, chest Xray findings often lag by hours behind the typical blood gas changes of respiratory alkalosis and hypoxia. Pulmonary
contusions are generally much larger than evidenced by chest radiographic findings. Poor
Page 370
Figure 22.5 Left scapular fracture.
Figure 22.6 Left pulmonary contusion.
outcomes have been noted in patients with: (i) contusions apparent on admission chest Xray; (ii) presence of three or more rib fractures; (iii) a PaO2/FiO2 ratio on
admission of less than 250 mmHg25. The initial treatment of pulmonary contusions is chest physiotherapy, analgesia, and judicious fluid replacement. Ventilatory support
is rarely required if the initial therapy is effective. Ventilatory support using intermittent mandatory ventilation (IMV) or pressure support ventilation (PSV) and positive
endexpiratory pressures (PEEP) provide improved ventilationperfusion matching and better venous return. Administration of excessive amounts of crystalloid solution
can aggravate the pathophysiologic changes that occur with pulmonary contusions25,26. In general, enough fluid should be provided to maintain the urine output at 0.5
ml kg–1 h–1 or more. Central venous pressure or pulmonary arterial catheter monitoring systems are indicated in patients with severe injuries or those with
Page 371
significant cardiac comorbidities that may affect the resuscitation. Patients with unilateral lung injury may benefit from synchronous independent lung ventilation (SILV)
provided through a doublelumen endotracheal tube27. The SILV technique helps prevent overinflation of the normal lung and underinflation of the damaged lung.
22.7.2 Hemothorax
Hemothorax is blood accumulating within the pleural space and is most frequently caused by blunt or penetrating injury to the lung parenchyma. Other sources of
bleeding may include the chest wall, intercostal or internal mammary arteries, great vessels, heart, or diaphragm. The combination of high concentrations of
thromboplastin and low arterial pressures within the lung along with the compressive effect of the hematoma help reduce the bleeding from the injured lung parenchyma
in most cases. The presence of clot within the thoracic cavity results in the release of substances causing ongoing fibrinolysis, thus continued bleeding may occur until the
clot is evacuated. Large blood clots may impede ventilation and affect venous return to the heart. Often, once the blood and blood clot are removed from the chest
bleeding from small vessels will stop. Severe hemorrhage is frequently seen with penetrating injuries to the chest. Injuries to the great vessels or even significant amounts
of the lung parenchymal damage can result in massive hemothorax.
Knowledge of the mechanism of injury in combination with the physical examination findings and a chest radiograph allow the diagnosis of hemothorax to be made
readily. Physical examination findings include decreased breath sounds with dullness to percussion on the affected side. Diagnostic workup with an upright or decubitus
chest Xray can usually reveal amounts greater than 200 ml of fluid or blood within the pleural space (Figure 22.7). Interestingly, as much as 1 l of fluid or blood may
be unappreciated on a supine chest radiograph. Ultrasonography can also be useful in diagnosing fluid within the pleural cavity; this can be performed as an extension of
the focused abdominal sonogram for trauma (FAST) exam.
General principles of treatment of a hemothorax are to restore blood volume, to provide drainage of the blood within the pleural space using closed tube
thoracostomy and to control bleeding. Hemothorax following blunt trauma to the thorax does not often result in massive hemorrhage in contrast to penetrating thoracic
injuries. Small collections of blood are resorbed by the body and thus can be
Figure 22.7 Right hemothorax.

Page 372
observed. Thoracentesis is not indicated in the treatment of every hemothorax. Large collections of blood, however, must be drained as quickly and efficiently as
possible using closed tube thoracostomy. A 32–40 French chest tube is ideal for the evacuation of blood from the pleural space with the tube positioned posteriorly.
Antibiotic prophylaxis is beneficial in reducing the incidence of empyema in patients having chest tubes placed after penetrating trauma28. Most clinicians provide at
least 24 h of antibiotic coverage against Staphylococcus aureus. Patients with associated intraabdominal injuries or esophageal injuries, patients with AIDS, or with
highvelocity gunshot wounds should also receive prophylactic antibiotics with coverage for anerobic and Gramnegative contamination as needed. If blood
accumulates after initial chest tube placement and cannot be removed, another chest tube should be inserted and positioned appropriately.
Significant hemorrhage is defined as more than 1500 ml of blood output after initial placement of a chest tube or continuous hemorrhage at 250 ml h–1 for 4 h.
Significant hemorrhage and/or an associated adverse change in the patients’ vital signs are indications for thoracotomy, preferrably under controlled conditions in the
operating room. Delay in performing a thoracotomy in such patients can result in sudden cardiac arrest. Blood drained from the pleural space is defibrinated and thus
can be collected and rapidly reinfused for resuscitation. Significant hemorrhage from chest tubes can be autotransfused to reduce the need for donor blood. Caution
should be used in regulating the amount of autotransfused blood replaced, as reinfusion of greater than 2500 ml of blood can lead to coagulopathy29.
Thoracoscopy is an important tool that can be utilized for diagnosis and treatment in the patient with hemothorax. Diagnostic thoracoscopy allows direct visualization
of the lung, chest wall, heart and great vessels. Most hemothoraces resorb spontaneously within 3–4 weeks, but retained or clotted hemothorax, most often after a
penetrating injury, can occur as a result of failure of adequate tube thoracostomy drainage. CT scan can differentiate between hemothorax, atelectasis and contusion.
Enzymatic liquefaction of the clot by instilling lytic enzymes has not proven satisfactory and is not advocated. Operative intervention with either thoracoscopy or limited
thoracotomy is necessary if the hemothorax (i) causes greater than onethird of lung volume loss in one hemithorax, (ii) causes significant atelectasis, or (iii) is associated
with infection. Delay of operative intervention beyond 4 weeks usually results in the development of fibrous adhesions along with increased capillary ingrowth from the
pleura into the clot. Fibrothorax may result in significant permanent loss of functional residual capacity.
Posttraumatic empyema is a significant problem that can occur after blunt or penetrating thoracic trauma. Causes of empyema include iatrogenic introduction of
contaminants, direct infection by penetrating missiles, secondary infection from an abdominal source (via hematogenous or transdiaphragmatic lymphatic routes), or
infection secondary to postpneumonic pneumonia. The organisms responsible vary greatly, but are commonly Staphylococcus aureus or Streptococcus species.
Secondary contamination often involves Gramnegative or mixed organisms. Appropriate cultures of the fluid drained from the chest cavity (by thoracentesis, chest tube
or CTultrasound guided aspiration) should be obtained and broadspectrum intravenous antibiotic therapy begun until therapy can be narrowed by culture results.
Wellestablished empyemas often require formal thoracotomy and decortication, in which all infected debris is removed, the lung is freed from adhesions and the
inflammatory peel is removed. Thoracoscopy may also provide definitive management of empyemas30.
22.7.3 Pneumothorax
Pneumothorax is the most common intrathoracic injury following blunt and penetrating trauma. A pneumothorax occurs when air orig
Page 373
inating from the lung parenchyma is trapped between the parietal and visceral pleura. In blunt trauma a pneumothorax may occur secondary to rib fractures, crush injury
with damage to alveoli, or from sudden increases in intrathoracic pressure. In penetrating wounds to the chest, particularly with gunshot wounds, a hemothorax often
accompanies the pneumothorax, whereas with stab wounds one often sees only a pneumothorax. Large pneumothoraces can easily be detected on physical
examination of the chest. Small pneumothoraces are more difficult to diagnose; diagnostic evaluation includes chest radiography with upright films at end expiration. In
the supine position, pleural air does not collect at the lung apex, but rather in the basilar and subpulmonic areas. Pneumothoraces are often expressed as a percentage
relating the distance across which the lung is collapsed from the chest wall compared with the total lung size, determined on the AP film. The treatment for
pneumothorax is tube thoracostomy. For patients with blunt thoracic injury, who often have a simple pneumothorax, a 22–28 French chest tube can be inserted in the
third or fourth intercostal space in the midclavicular line. If an associated hemothorax is present, as with penetrating trauma, a larger 32–40 French tube is
recommended positioned optimally for both fluid and air evacuation in the fifth or sixth intercostal space in the anterior axillary line. The tube is then connected to an
underwater seal with negative suction (usually –20 cmH2O) to facilitate drainage. A chest Xray is obtained to check the position of the tube and to determine the
amount of air or fluid remaining within the chest cavity.
A subset of pneumothorax patients may not require chest tube insertion. These patients, who are hemodynamically stable and have isolated, small pneumothoraces
(<20%) and no other injuries, can be followed with a chest radiograph at 6 and 24 h. Resorption of the air within the pleural space often occurs within days and can be
facilitated by the administration of supplemental oxygen. However, all patients with a pneumothorax, regardless of size, should have a chest tube inserted before being
placed on positive pressure ventilation.
After resolution of the pneumothorax, chest tubes are ready to be removed when no air leak has been present for 24 h and/or when serous drainage from the tube is
less than 250 ml day–1. Chest tubes may be removed while on suction. A chest Xray is obtained within 6 h after removal of the tube. Complications of removal include
a recurrent pneumothorax, for which a chest tube is reinserted, if necessary.
A continued pneumothorax is a complication that may be due to improper connections in the external tubing or to the water seal device, improper position of the
tube, obstruction of or tear in the bronchi, or a large tear in the lung parenchyma. When the lung is not expanded and a large air leak is present bronchoscopy should be
performed to clear the bronchi and to attempt to identify a tracheobronchial injury. Continued large air leak for 24–48 h along with a pneumothorax, with failure of the
described measures above, is an indication for thoracotomy. In these patients the incidence of empyema and bronchopleural fistula is very high. A posterior lateral
thoracotomy is performed and the patient will require stapling of the lung parenchyma, a limited pulmonary resection or repair of a tracheobronchial injury. Some
success in sealing air leaks is possible when autologous blood (‘blood patch’) or fibrin gel is administered in patients with persistent air leaks via a closed chest tube
system31.
22.7.4 Tension pneumothorax

This is an immediately lifethreatening condition in which air at supraatmospheric pressure accumulates in a hemithorax. The air results in an elevation of the intrapleural
pressure that leads to impairment in venous return and a decrease in cardiac output. The underlying pathophysiology is very similar to that seen with cardiac tamponade
except that in tamponade the increased pressure is within the pericardium not the pleural space. Elevated intrapleural pressure also causes a shift in the
Page 374
mediastinum to the opposite side manifested by tracheal deviation. Respiration is adversely affected along with circulatory compromise. Physical examination findings
include absent breath sounds and hyperresonance to percussion on the affected side. Immediate diagnosis and treatment must be instituted for this lifethreatening
emergency. Once suspected, time must not be wasted in obtaining a chest Xray. Rather a large bore (14–16 French) intravenous catheter should be inserted into the
second intercostal space in the midclavicular line on the affected side to relieve the tension component. Subsequently, a tube thoracostomy is performed to treat the
pneumothorax.
22.7.5 Inhalation injuries

These occur when hot gases, usually secondary to fires in closed spaces, are inhaled. A high index of suspicion based on mechanism necessitates investigation. The
pathophysiology of inhalation injury involves: (i) direct thermal injury from heat exposure, usually limited to the supraglottic area; (ii) carbon monoxide poisoning—a
byproduct of organic combustion, carbon monoxide binds avidly to hemoglobin (300 times greater affinity than oxygen) decreasing the oxygencarrying capacity of
hemoglobin, decreasing oxygen delivery and leading to hypoxia; (iii) chemical injury to the mucosa (tracheobronchitis) and alveoli (pneumonitis) from corrosive gases
and irritants. The last of these results in damage to type 2 pneumocytes, leading to decreased surfactant production and decreased alveolar volume.
Evaluation and control of the airway is essential in management of these patients and should be performed as early as possible. Diagnosis of inhalation injury requires
direct laryngoscopy or flexible bronchoscopy to evaluate the glottic region for supraglottic edema, erythema, ulceration, hemorrhage or soot particles. Presence of these
findings implies upper airway injury and usually necessitates prophylactic endotracheal intubation to prevent eventual loss of the airway as a result of edema. Other
physical findings include tachypnea, hoarseness, labored breathing, altered phonation, stridor, and cyanosis. Flexible bronchoscopy allows the evaluation of the distal
tracheobronchial tree and the ability to irrigate and suction the distal airways. Injuries to the distal bronchioles and alveoli are best evaluated by xenon133 ventilation
perfusion scanning32. Arterial blood should be sent for determination of carbon monoxide levels along with the oxygen and acidbase status of the patient. Management
involves airway control, postural drainage and treatment of bronchospasm. Prophylactic intravenous antibiotics or steroids are not indicated in the management of
patients with inhalation injury. Intravenous antibiotics are initiated cultures. only when an organism is identified from
Physical manifestations of carbon monoxide poisoning are related to the carboxyhemoglo bin level in the blood. Normal carboxyhemoglobin is less than 2%, but may
approach 8% in smokers. Carboxyhemoglobin levels between 20 and 40% result in headaches and nausea, levels of 40–60% cause neurologic dysfunction, and levels
greater than 60% are often fatal. The treatment of carbon monoxide poisoning requires the administration of 100% oxygen via a nonrebreather mask or an
endotracheal tube, which decreases the halflife of carbon monoxide from 270 min to 50 min. Treatment is continued until the carboxyhemoglobin level is less than 5%
(normal values 0–5%). Hyperbaric oxygen, if available, can reduce the carboxyhemoglobin levels faster and can reverse the neurologic symptoms of carbon monoxide
poisoning. Indications for hyperbaric treatment are given in Table 22.2.
22.8 Tracheobronchial injury
Traumatic injuries to the trachea or major bronchi are rare. Bronchial injuries occur predominantly on the right side33,34. When present they are often life threatening,
with the majority of patients presenting with severe respiratory distress or in extremis. Autopsy studies have revealed that the majority of patients with major airway
injuries die at the scene
Page 375
Table 22.2 Indications for hyperbaric oxygen treatment
Air or gas embolism

Complications of uncontrolled decompression during diving
Carbon monoxide intoxication
Necrotizing soft tissue infections
Chronic nonhealing wounds
Clostridial myonecrosis and myositis
Cyanide intoxication
Compartment syndrome or crush injuries
Multiple sclerosis
Intracranial abscess
Significant anemia
Thermal burns
Osteomyelitis
Mucormycosis
from asphyxia as a result of obstruction of the airway compounded by aspiration of blood. Patients may present with severe dyspnea, hemoptysis, aphonia, cyanosis
and hypotension. A pneumothorax is often present and a large, persistent air leak notable once a chest tube is inserted. Pneumomediastinum with subcutaneous
emphysema will be present if the air leak from the injury does not communicate with the pleural space.
The chest radiograph will be abnormal in 90% of patients with major airway injury, with pneumomediastinum, pneumothorax, or pleural effusion often noted. Also,
although uncommon, the presence of a ‘dropped lung’ sign on Xray (the apex of the lung is sitting at the level of the hilum, below the bronchial transection) is
characteristically seen with mainstem bronchial injury35. Atelectasis and pulmonary infiltrates may also appear on Xray. Flexible bronchoscopy should be performed in
all patients with suspected tracheobronchial injury33,36 to establish the diagnosis, localize the injury and allow for safe endotracheal intubation. Bronchoscopy should be
performed in the operating room while maintaining cervical spine precautions. Once the airway is secure and the diagnosis made, consultation with a thoracic surgeon is
indicated for repair of the injury. Both blunt and penetrating tracheobronchial trauma is often accompanied by major vascular and esophageal injuries.
Many small tracheobronchial injuries are not diagnosed until after they cause atelectasis and severe pulmonary infections. Delay occurs because the peribronchial
tissue may remain intact, allowing for ventilation to continue. Within 2–6 weeks stenosis secondary to the development of granulation tissue occurs, resulting in distal
airway collapse with subsequent atelectasis and infection34,37.
22.9 Blunt aortic injury
Blunt aortic injury (BAI) is the second most common cause of death in blunt trauma patients38,39. Only 13–15% of patients with this type of injury reach the hospital
with signs of life and those who initially survive have a 30% risk of subsequent rupture and death38. Diagnosis of BAI can be made using chest Xray, aortography
(sensitivity 99%, specificity 96%), CT, or transesophageal echocardiography (sensitivity 60–250%, specificity 90%). The ‘gold standard’ for the diagnosis of an aortic
injury is aortography, although CT scanning of the chest is reported to be 100% sensitive, 81–99% specific with a 100% negative predictive value40,41 (Figure 22.8). A
widened superior mediastinum is defined as a measured width greater than 8 cm, a mediastinal/chest width ratio of greater than 0.38, or simply the physician’s
impression or suspicion that the mediastinum is widened, on an erect posterioranterior view of the chest (Figure 22.9). Other chest radiographic findings suggestive
for aortic rupture are listed in Table 22.3. Once the diagnosis of BAI is made, immediate surgical repair by a cardiovascular surgeon is indicated. Unfortunately, this
may not always be possible in patients with multiple trauma who are unstable from intraabdominal injuries or severe head injuries. Additionally, elderly trauma patients
with complicated comorbidities that preclude immediate repair are best managed with pharmacological manipulation of their heart rate and blood pressure (see Chapter
9).
Page 376
Table 22.3 Chest Xray findings suggestive of aortic injury
Widened mediastinum Indistinct aortic knob

Depression of left mainstem bronchus Opacification of aortopulmonary window
Widened paratracheal stripe Deviation of nasogastric tube to right
Apical capping First or second rib fracture
Figure 22.8 Chest CT with intravenous contrast demonstrating active extravasation of contrast f from the thoracic aorta with resultant
left hemothorax.
Figure 22.9 Chest Xray showing widened mediastinum.
In BAI the aortic tear typically occurs distal to the ligamentum arteriosum, and disruption may not always occur through all three layers of the aortic wall. Patients with
BAI who survive to reach the hospital often have a subadventitial hematoma or dissection limited
Page 377
to a short segment near the rupture. Traumatic aortic dissection occurs with a frequency between 2 and 12%42,43 although the true incidence of traumarelated
dissection is not known, as aortography can result in false negative results if the false lumen is not opacified. Contrast enhanced chest CT scan and TEE have been used
to investigate aortic diseases and both have been shown to be accurate for the diagnosis of dissection44. The underlying pathophysiology of aortic dissections involves
an intimal tear with formation of a false lumen within the medial layer. The false lumen can extend to the abdominal aorta to occlude visceral or renal branches, resulting
in organ infarction43,45. Medial aortic dissection seems to occur commonly in patients predisposed to aortic disease (atherosclerosis or hypertension)43,46,47 although
traumarelated dissection may develop without prior aortic disease.
Complications after aortic surgery include paraplegia (5%), renal failure, respiratory complications (pneumonia, empyema), and aortic aneurysm formation from
suture line failure. Paraplegia occurs when there is hypoperfusion of the spinal cord resulting from crossclamping above the artery of Adamkowitz, which supplies the
spinal cord but takes off from the aorta at variable levels. Risk factors for the development of paraplegia include intraoperative hypotension and prolonged duration of
aortic crossclamping (>30 min) if using the clamp and sew technique. Cardiopulmonary bypass should be considered if the repair is anticipated to take longer than 30
min.
22.10 Blunt cardiac injury

Blunt cardiac injury is usually the result of a direct precordial blow, and can result in myocardial contusion, cardiac rupture, valvular tears, or coronary thrombosis.
22.10.1 Myocardial contusion

Most patients with myocardial contusions are asymptomatic. Diagnosis is made based on the mechanism of injury (steering wheel damage), bruising and tenderness
over the sternum, associated sternal fractures, multiple rib fractures and the presence of cardiogenic shock. There is no highly sensitive or specific test available to make
the diagnosis. Cardiac enzyme fractions specific for myocardium (creatinine phosphokinase, lactate dehydrogenase and troponin) are not predictive48. Tachyarrythmias
are the most common manifestation of cardiac contusions and the admission 12lead ECG is the most important diagnostic screening tool49. Other dysrhythmias include
atrial fibrillation or flutter, premature ventricular contractions, ventricular tachycardia or fibrillation. Twave and STsegment abnormalities may also be seen and are
indicative of a bruised ventricular myocardium. Echocardiogram and radionuclide scans are not effective screening tools. A normal ECG in a hemodynamically stable
blunt trauma patient rules out blunt cardiac injury of any clinical significance. Current practice guidelines recommend observation in a telemetry unit to monitor for the
development of dysrhythmias or hemodynamic instability, which will require treatment in about 2–5% of patients (see Chapter 11). Delayed complications from
myocardial contusions are rare.
22.10.2 Cardiac rupture

This usually occurs at the atriocaval junction, left atrium, or right ventricular outflow tract50, often in the setting of acute deceleration, crush injury, or lacerations from
chest wall fractures. Patients with blunt cardiac rupture are often dead at the scene or die en route to the hospital. Valvular disruption occurs in 9% of patients with
blunt cardiac injury and most often involves the aortic valve51. The new onset of a harsh holosystolic murmur implies rupture of a papillary muscle. Subsequent severe
left ventricular failure is an indication for early operative repair. Ventricular septal rupture is rare, occurring most often at the muscular portion of the septum near the
apex. Coronary vascular thrombosis is rare following blunt thoracic trauma. Its presentation is
Page 378
indistinguishable from atherosclerotic coronary occlusion, but the prognosis is more favorable52. Hemodynamic monitoring, fluid managment and inotropic support for
blunt cardiac injury follow usual practice as discussed in Chapters 5–8.
References
1. LoCicero, J. and Mattox, K.L. (1989) Epidemiology of chest trauma. Surg. Clin. N. Am. 59: 15–19.
2. Wilson, R.F. and Steiger, Z. (1999) Thoracic trauma: Chest wall and lung. In: Trauma: Pitfalls and Practice, 2nd Edn (ed. R.F.Wilson). Lippincott, Williams &
Wilkins, Philadelphia, PA, pp. 314–342.
3. Richardson, J.D., Miller, F.B., Carrillo, E.H. and Spain, D.A. (1996) Complex thoracic injuries. Surg. Clin. N. Am. 76(4):725–748.
4. Hollinshead, W.H. and Rosse, C. (eds) (1985) The thorax in general. In: Textbook of Anatomy, 4th Edn. Harper & Row, Philadelphia, PA, pp. 465–488.
5. The American College of Surgeons (1997) Advanced Trauma Life Support (ATLS) Manual, 6th Edn. First Impression, Chicago, IL pp. 21–46.
6. Chaikhouni, A., Dayas, C.L. Jr, Robinson, J.H., et al. (1981) Latissimus dorsi free myocutaneous free flap. J. Trauma 21:398–402.
7. Landercasper, J. and Cogbill, T.H. (1985) Longterm followup after traumatic asphyxia. J. Trauma 25(9):838–841.
8. Jongewaard, W.R., Cogbill, T.H. and Landercasper, J. (1992) Neurologic consequences of traumatic asphyxia. /. Trauma 32(1):28–31.
9. Bulger, E.M., Arneson, M.A., Mock, C.N. and Jurkovich, G.J. (2000) Rib fractures in the elderly. J. Trauma 48(6):1040–1046.
10. Hunt, D.M. and Schwab, F.J. (1992) Chest trauma. In: Diagnostic Radiology in Emergency Medicine (ed. P. Rosen). Mosby Yearbooks, St. Louis, MO,
p. 77.
11. Bassett, J.S., Gibson, R.D. and Wilson, R.F. (1968) Blunt injuries to the chest. J. Trauma 8 (3): 418–429.
12. Logan, P.M. (1999) Is there an association between fractures of the cervical spine and firstand secondrib fractures? Can. Assoc. Radio. J. 50(1):41–43.
13. Wilson, J.M., Thomas, A.N., Goodman, P.C., et al. (1978) Severe chest trauma: morbidity implications of first and second rib fractures in 120 patients. Arch.
Surg. 113(7):846–849.
14. Collins, J. (2000) Chest wall trauma. J. Thorac. Imag. 15(2):112119.
15. Maloney, J.V., Jr, Schnutzer, K.J. and Raschke, F. (1961) Paradoxical respiration and 'pendelluft'. J. Thorac. Cardiovasc. Surg. 41:291.
16. Restifo, K.M. and Kelen, G.D. (1994) Case Report: sternal fracture from a seatbelt. J. Emerg. Med. 12:321–323.
17. Hamilton, J.R., Dearden, C. and Rutherford, W.H. (1984) Myocardial contusion associated with fracture of the sternum: important features of the seat belt
syndrome. Injury 16(3):155–156.
18. Purkiss, S.F. and Graham, T.R. (1993) Sternal fractures. Br. J. Hosp. Med. 50(2–3): 107–112.
19. Brookes, J.G., Dunn, R.J. and Rogers, I.R. (1993) Sternal fractures: A retrospective analysis of 272 cases. J. Trauma 35(1):46–54.
20. RoyShapiro, A., Levi, I. and Khoda, J. (1994) Sternal fractures: A red flag or a red herring. J. Trauma 37(1):59–61.
21. Henley, M.B., Peter, R.E., Benirschke, S.K., et al. (1991) External fixation of sternum for thoracic trauma. J. Ortho. Trauma 5 (4):493–497.
22. Harris, R.D. and Harris, J.H., Jr (1988) The prevalence and significance of missed scapular fractures in blunt trauma. Am. J. Radiol. 151(4): 747–750.
23. Fulton, R.L. and Peter, E.T. (1970) The progressive nature of pulmonary contusion. Surgery 67(3):499–506.
24. Trinkle, J.K., Furman, R.W., Hinshaw, M.A., et al. (1973) Pulmonary contusion. Ann. Thorac. Surg. 16(6):568–573.
25. Hoff, S.T., Shotts, S.D., Edd, V.A., et al. (1994) Outcome of isolated pulmonary contusion in blunt trauma patients. Am. Surg. 60(2):138–142.
26. Adoumie, R., Shennib, H., Brown, R., et al. (1993) Differential lung ventilation. Applications beyond the operating room. J. Thorac. Cardiovasc. Surg. 105
(2):229–233.
27. Richardson, J.D., Franz, J.L., Grover, F.L., et al. (1974) Pulmonary contusion and hemorrhage: crystalloid versus colloid replacement. J. Surg. Res. 16(4):330–
336.
28. Luchette, F.A., Barie, P.S., Oswanski, M.F., et al. (2000) Practice management guidelines for prophylactic antibiotic use in tube thoracostomy for traumatic
hemothorax: The EAST practice management guidelines work group. J. Trauma 48(4):753–757.
29. Thomas, A.N. (1982) Discussion of Graham et al Innominate vascular injury. J. Trauma 22: 655.
30. O’Brien, J., Cohen, M., Solit, R., et al. (1994) Thoracoscopic drainage and decortication as definitive treatment for empyema thoracis following penetrating chest
injury. J. Trauma 36(4):536–539.
Page 379
31. Nicholas, J.M. and Dulchavsky, S.A. (1992) Successful use of autologous fibrin gel in traumatic bronchopleural fistula: case report. J. Trauma 32(1):87–88.
32. Moylan, J.A., Jr, Witmore, D.W., Mouton, D.E., et al. (1972) Early diagnosis of inhalation injury using 133xenon lung scan. Ann. Surg. 176(4):477–484.
33. Baumgartner, F., Shepard, B., de Virgilo, C., et al. (1990) Tracheal and main bronchial disruptions after blunt chest trauma: presentation and management. Ann.
Thorac. Surg. 50(1):569–574.
34. Deslauriers, J., Beaulieu, M., Archambault, G. et al. (1982) Diagnosis and long term followup of major bronchial disruptions due to nonpenetrating trauma. Ann.
Thorac. Surg. 33(1):32–39.
35. Taskinen, S.O., Salo, J.A., Halttunen, P.E.A. and Sovijarvi, A.R.A. (1989) Tracheobronchial rupture due to blunt chest trauma: a followup study. Ann. Thorac.
Surg. 48(6):846–849.
36. Jones, W.S., Mavroudis, C., Richardson, J.D., et al. (1984) Management of tracheobronchial disruption resulting from blunt trauma. Surgery 95(3):319–323.
37. Roxburgh, J.C. (1987) Rupture of the tracheobronchial tree. Thorax 42(9):681–688.
38. Fabian, T.C., Richardson, J.D., Croce, M.A., et al. (1997) Prospective study of blunt aortic injury: Multicenter trial of the American Association for the Surgery of
Trauma. J. Trauma 42(3):374–380.
39. Nagy, K., Fabian, T., Rodman, G., Fulda, G., et al. (2000) Guidelines for the diagnosis and management of blunt aortic injury: An EAST practice management
guidelines work group. J. Trauma 48(6):1128–1143.
40. Fabian, T.C., Davis, K.A., Gavant, M.L., et al (1998) Prospective study of blunt aortic injury: helical CT is diagnostic and antihypertensive therapy reduces rupture.
Ann. Surg. 227(5): 666–676.
41. Mirvis, S.E., Shanmuganathan, K., Buell, J. and Rodriguez, A. (1998) Use of spiral computed tomography for the assessment of blunt trauma patients with
potential aortic injury J. Trauma 45(5):922–930.
42. Fischer, R.G. and Hadlock, F. (1981) Laceration of the thoracic aorta and brachiocephalic arteries by blunt trauma: report of 54 cases and review of the literature.
Radiol. Clin. N. Am. 19(1):91–110.
43. Gates, J.D., Clair, D.G. and Hechtman, D.H. (1994) Thoracic aortic dissection with renal artery involvement following blunt thoracic trauma: Case report. J.
Trauma 36(3):430–432.
44. Nienaber, C.A., von Kodolitsch, Y., Nicholas, V., et al. (1993) The diagnosis of thoracic aortic dissection by noninvasive imaging procedures. N. Engl J. Med.
328(1):1–9.
45. Gammie, J.S., Katz, W.E., Swanson, E.R. and Peitzman, A.B. (1996) Acute aortic dissection after blunt trauma. J. Trauma 40(1):126–127.
46. Perchinsky, M., Gin, K. and Mayo, J.R. (1998) Traumaassociated dissection of the aorta. J. Trauma 45(3):626–629.
47. Faraci, R.M. and Westcott, J.L. (1976) Dissecting hematoma of the aorta secondary to blunt chest trauma. Radiology 123(3):569–574.
48. Fabian, T.C., Cicala, M.R., Croce, M.A., et al. (1991) A prospective evaluation of myocardial contusion: correlation of significant arrythmia and cardiac output with
CPKMB measurements. J. Trauma 31(5):653–659.
49. Feliciano, D.V. and Rozycki, G.S. (1999) Advances in the diagnosis and treatment of thoracic trauma. Surg. Clin. N. Am. 79(6):1417–1429.
50. Williams, J.B., Silver, D.G. and Laws, H.L. (1981) Successful management of heart rupture from blunt trauma. J. Trauma 21:534.
51. Parmley, L.F., Manion, W.C. and Mattingly, T.W. (1958) Nonpenetrating traumatic injury of the heart. Circulation 18:371.
52. Ivatury, R.R. (1996) Injury to the heart. In: Trauma, 3rd Edn (eds E.E.Moore, K.L.Mattox and D.V.Feliciano DV). Appleton & Lange, Stamford, CT, pp. 409–421.
Page 380

Page 381
Chapter 23
Techniques of vascular access for invasive hemodynamic monitoring
William T.McGee, MD, MHA,
Jay S.Steingrub, MD, and
Contents
23.1 Introduction
23.2 Intraarterial blood pressure monitoring
23.3 Central venous access
23.4 Femoral venous cannulation
23.5 Internal jugular vein
23.6 Subclavian vein cannulation
23.7 Introducer placement
23.8 Complications and their prevention
Page 382
23.1 Introduction
Accessing the circulation is often an initial lifesaving procedure in critically ill patients. Central venous catheterization is the most common invasive procedure performed
in hospitalized patients. Although routinely performed safely, significant morbidity and even mortality are reported to complicate central venous cannulation. Physicians
performing vascular access procedures require an excellent understanding of relevant anatomy and awareness of potential complications.
23.2 Intraarterial blood pressure monitoring

Systemic arterial blood pressure and continuous electrocardiographic (EKG) recording is an essential component of hemodynamic monitoring routinely provided to
critically ill patients. Although indirect blood pressure measurements via sphygmomanometry are generally accurate for normotensive or hypertensive individuals,
measurements are intermittent, subject to technical error, and tend to underestimate systolic blood pressure in the critically ill, hemodynamically unstable patient.
Advantages of indwelling catheters include accuracy, access for blood sampling, and continuous acquisition and display of data. Ongoing surveillance of pressure both
permits immediate recognition of hypotension or hypertensive episodes and facilitates titration of pharmacotherapy.
23.2.1 Blood pressure physiology

The pressure in the arterial tree is highest at the aortic valve and falls as a result of the resistance of arterioles, but is also affected by the dynamics of blood flow during
each cardiorespiratory cycle. Thus, the arterial waveform and the values of systolic and diastolic pressure may be dissimilar at various locations in the arterial tree or
during phases of the cardiorespiratory cycle. Invasive direct measurement of blood pressure best corresponds to actual clinical events and thus best reflects the true
‘blood pressure’.
Systolic and diastolic pressures represent estimates of the highs and lows of systemic arterial pressure generated during a cardiac cycle, mean systemic pressure
being the average pressure. Pressure transducers convert the pressureinduced movement of a diaphragm to an electrical signal proportional to the inducing pressure.
The accuracy of physiological data accumulation is dependent on the system’s ability to translate these data without distorting the arterial signal as reflected by a visible
waveform.
23.2.2 Site selection for invasive blood pressure measurement

Intraarterial catheters may be inserted in the femoral, radial, axillary or dorsalis pedis artery with no data supporting an optimal site. However, large proximal vessels
and larger catheters provide more accurate pressure measurements. Smaller catheter diameter in relation to the vessel diameter minimizes the risk of vessel thrombosis.
Ideal arteries are those with rich collateral circulation that will assure distal perfusion in the event of thrombosis. The site should be easily accessible for nursing care and
not prone to contamination. The radial arteries are the most frequently cannulated site; success rates are approximately 92%1,2. In the presence of circulatory shock,
the femoral artery may be most easily cannulated and less prone to thrombosis. Employment of the Seldinger technique, which employs a small needle for vessel
cannulation followed by a guidewire over which the catheter is placed, has minimized hemorrhagic complications. Axillary artery cannulation is more difficult and has a
greater risk of producing particulate or air embolization when retrograde flow occurs during catheter flushing. Brachial artery catheterization is generally avoided
because of the lack of collateral circulation around the vessel and the possibility of median nerve injury. As a result of the distal location of the dorsalis pedis artery,
invasive pressure monitoring from this site is more prone to waveform artifact.
Page 383
23.2.3 Complications
Complications with arterial cannulation are usually related to minor bleeding, resulting in localized hematomas. Although commonly asymptomatic and not clinically
significant, arterial thromboses occur in up to 20–48% of the catheterizations of greater than 48 h1,2. Most data on arterial thromboses involve radial artery cannulation,
with the incidence of thrombosis increasing as the ratio of the catheter diameter to vessel diameter increases. Severe ischemia leading to permanent damage occurs in
less than 0.5% of arteries cannulated despite the high prevalence of partial or complete arterial thrombosis, if good collateral flow is available. The Allen test is used
clinically to test for collateral flow. The ulnar and radial arteries are simultaneously occluded while the patient’s hand is held upright. The patient is then asked to open
and close his or her fist until the fingers appear blanched. To check for collateral flow, the ulnar artery is released; evidence of return of circulation should appear
promptly. If prolonged beyond 4 s, collateral flow is thought to be insufficient and this site should probably be abandoned in favor of another site. Using a Doppler
probe beyond the occlusion of the radial artery to assess the return of a signal when the ulnar artery is released is an alternative method to the visual test. Local
neurologic complications of radial and axillary catheterization are rare, but have been attributed to median nerve or brachial plexus compression by hematomas. The
incidence of catheterrelated sepsis is rare.
23.2.4 Cannulation procedure

The procedure should be explained to the patient, informed consent obtained for nonemergency procedures, and the patient made comfortable with appropriate
sedatives and analgesia. An operative field should be large enough so that the sterility of the required catheters and wires can be maintained. Local anesthesia should be
obtained through a 25gauge needle utilizing a 1% lidocaine solution. Infiltration of deeper tissue with local anesthetic is often required for femoral and brachial sites.
23.2.5 Radial artery

The Allen test is performed. The patient’s wrist is hyperextended and held in place with an arm board and gauze dressing so that the wrist is exposed. Once the wrist is
extended and fixed to the arm board, a wash is performed and a sterile field obtained. The artery is then palpated between the first and second fingers. The entry site is
typically 1–1 1/2 cm cephalad from the junction of the arm and the hand. Typically, the best approach is to find the area of maximal arterial pulsation. A 20gauge or
smaller catheter over the needle cannulation is typically used with or without a guidewire. The needle is held at an approximately 30–45° angle above the skin toward
the direction of blood flow (i.e. cephalad) aiming directly at the pulsation. Once pulsatile blood flow is noted, the catheter should be gently slipped into the artery with a
slight rotatory motion while holding the needle guide perfectly still. If a guidewire is to be used at this point, it should now be inserted through the needle into the vessel
and the catheter advanced over it. Occasionally the back wall of the radial artery is penetrated. In this circumstance the pulsatile blood flow will be transient. Simply
backing the needle catheter assembly out slowly and assessing for the return of pulsatile blood flow will often locate the vessel in this situation.
The catheter should then be hooked up to the pressure tubing, quickly flushed and then assessed for the adequacy of the arterial tracing. Once confirmed within the
artery by visual waveform analysis, the catheter should be secured using tape, a stitch or one of the securing devices without needles such as Stat Lock3. If cannulation
fails, there may be anatomic or technical problems why this vessel cannot be entered. Radial artery cannulation is associated with a very high success rate, thus we
recommend no greater than four separate arterial sticks before abandoning one site and moving to another. Patency of radial
Page 384
artery catheters is enhanced by placing the catheter close to the bend of the wrist4.
23.2.6 Femoral artery

The femoral artery is palpated just below the inguinal ligament in the groin. It has the advantage of being a fairly large vessel and may be easier to palpate in shock
patients. The artery is palpated between the first and second digits attempting to define the medial and lateral extent of the vessel. A fairly steep angle of entry, often
between 60 and 90°, must be employed because of the depth of the vessel and the lengths of commonly employed needles. A guidewire is always used as part of this
technique and should be inserted through the needle once pulsatile blood return is identified. As with all arterial cannulations, the needle and catheter are advanced
toward the blood flow. It is often necessary to drop the insertion angle down more in line with the expected course of the artery to facilitate guidewire entry. Once the
guidewire is in place, the needle is removed and then the femoral arterial catheter is placed. Dilation of the skin and muscle is recommended to provide a tract through
which the catheter can be easily slid. If the skin entry site is tight, a small stab incision can be made with the Number 11 blade facing away from the wire to allow easier
passage of the dilator and catheter.
23.2.7 Dorsalis pedis artery

Assessment of the accuracy of collateral flow should be obtained. Similar to the Allen test, the dorsalis pedis and posterior tibial artery should be simultaneously
occluded while assessing blood flow to the great toe. This can be performed by either placing a Doppler along the course of the dorsalis pedis artery distal to the
occlusion or using a visual assessment of the great toe while the posterior tibial artery is released. Again, either return of color or prompt return of pulsation would
indicate that adequate collateral flow is present. The procedure proceeds similarly to radial artery cannulation. Hyperextension of the foot is sometimes thought to be
helpful and typically a catheter of no greater than 20 gauge should be used.
23.2.8 Axillary artery

Axillary cannulation is rarely used and associated with a higher failure rate. This approach has a more significant potential for cerebral embolization during catheter
flushing especially if done on the right side because of the proximity to the central circulation and the carotid artery. For this reason, axillary artery cannulation is almost
always done on the left side.
The patient is positioned with their hand under their head to allow maximal exposure of the axilla. The vessel is then palpated and can typically be entered with a
catheter over the needle type of device, as the vessel is fairly superficial at this point.
23.2.9 Brachial artery

The brachial artery is uncommonly used, as the collateral circulation around this vessel is often inadequate. The Allen test should be performed before brachial artery
cannulation. Because of the close proximity of the radial nerve in this area, arterial bleeding has been associated with a compressive neuropathy of the median nerve in
those patients in whom bleeding is a complication of the technique.
23.3 Central venous access

Access to the circulation is required for almost all therapies in hospitalized patients. Besides inability to cannulate the peripheral circulation, there are many other reasons
why central venous access may be required. Indications for central venous cannulation are numerous and are listed in Table 23.1.
23.3.1 Site selection

Site selection should consider the operator’s expertise with any particular technique, and anatomic considerations for a particular patient. Often the subclavian or
femoral route is more accessible in an acutely injured patient who may require intubation and also have a
Page 385
Table 23.1 Indications for central venous cannulation
Diagnostic uses
• Pulmonary artery catheterization (right internal jugular preferred)
• Central venous pressure monitoring
• Endomyocardial biopsies (right internal jugular preferred)
• Pulmonary angiography
• Differentiation of supraventricular from ventricular arrhythmias
• Transvenous hepatic and cardiac biopsy and other rare indications
Therapeutic uses
• Administration of large amounts of fluid, hemodialysis, or total parenteral nutrition
• Access for hemodialysis, hemofiltration or plasmapheresis
• Transvenous pacemaker insertion (right internal jugular preferred)
• Removal of embolized thrombi or catheters
• Aspiration of air from the right atrium during neurosurgical procedures
• Valvuloplasty, angioscopy, and other rare indications
cervical collar in place when first seen. Other anatomic considerations peculiar to individual patients such as a short, thick neck, prior arterial surgery along one of the
adjacent arteries, presence of localized skin infection, burns or other unique anatomic variations guide site selection. In patients with chest tubes, it is preferable to use
the side where the tube is already in place for subclavian vein cannulations. In patients with a single lung, however, a pneumothorax will not be tolerated, and the side of
the solitary lung should be avoided. Femoral venous catheters may pose an independent risk for venous thrombosis, and in those patients already at high risk for this
complication this route should be avoided if possible4,5. For patients with traumatic brain injury or other causes of increased intracranial pressure, occlusion of the
internal jugular vein secondary to thrombosis or hematoma may increase the risk of that approach. In patients with coagulopathy, a compressible vessel is preferred, as
bleeding is a common complication of venipuncture and especially inadvertent arterial puncture. The subclavian vein is not a compressible vein. The relative advantages
and disadvantages of internal jugular vein (IJV) versus subclavian vein (SCV) are listed in Table 23.2. Although femoral catheters are not truly central venous catheters
because they typically do not enter the chest, other than for pressure monitoring, they are used identically to other central venous catheters.
23.3.2 Preparation and equipment

All equipment necessary for this technique should be available at the bedside before a cannulation attempt. This would also include any drugs necessary for sedation,
pain control, and local anesthesia, along with cleansing materials for the skin and those materials necessary for the preparation of a sterile field. Patient hair likely to
interfere with sterility should be clipped, not shaved, as shaving may increase infection risk. In our practice, cap, mask, gloves and gown are utilized following a
thorough washing of the hands.
The surgical site is prepared with antimicrobial solution and gently scrubbed with a circular motion progressing from the center of the site outward. Skin preparation
should be performed first so that there is an adequate amount of time (approximately 5 min) for antimicrobial action.
After sterile skin preparation, new gloves should be utilized, and the insertion site draped with sterile towels to prevent any equipment, specifically the catheter or the
guidewire, from touching a nonsterile area. A sterile field large enough to accommodate at least twice the length of the guidewire is recommended. The opening in the
barrier protection to the skin should be large enough to facilitate cannulation and palpation of landmarks. For cooperative patients, local infiltration of 1% lidocaine
often provides adequate analgesia. For patients who are particularly anxious or uncomfortable, systemic therapy with anxiolytics and narcotics may be necessary.
Attempting vascular access procedures especially in the chest and neck in an uncooperative or combative patient can result in
Page 386
Table 23.2 Advantages and disadvantages of neck versus clavicular approaches
Advantages Disadvantages
Internal Good landmarks, vein may be visible and palpable Small pneumothorax risk Obscured landmarks in obese patients Tracheostomy appliances problematic
jugular vein (<0.1%) Compressible bleeding site with good hemostasis after inadvertent Dressing hard to secure, possible secretion contamination hazard Relative
arterial puncture Easiest SwanGanz and pacemaker passage from right discomfort in some patients Contraindication with high intracranial pressure
side
Subclavian May be the largest target in vasoconstricted states Fair landmarks, if patient 1–5% pneumothorax risk Avoid with chronic obstructive pulmonary disease,
vein* obese, edematous Dressing easily secured Generally comfortable high positive and expiratory pressure, coagulopathies Noncompressible
artery if punctured Slightly less ease of passage for SwanGanz, pacemaker
6
*The supraclavicular approach to the subclavian and brachiocephalic veins, as taught and performed by Parsa , appears to have comparable success, and lower immediate complication
rates, than reported for subclavian and IJV approaches. Reproduced with permission from W.B. Saunders.
inappropriate motion by the patient at critical junctures during the procedure and lead to potentially lifethreatening injuries. In select circumstances, intubation and
paralysis are required. While the prep solution dries and the local anesthetic takes effect, the operator should remove the catheter from its packaging, flush the port(s)
with saline or heparin flush solution, place any necessary endcaps, and lay the prepared catheter on the sterile field where it will be quickly accessible. Using a 25
gauge needle, a skin weal is raised over the anticipated entry site. Following this, a 22gauge needle is passed through the anesthetized dermis to infiltrate subcutaneous
tissue and occasionally bone if contact with the clavicle is expected to be made.
23.3.3 Positioning
Positioning of the patient typically is the final preparation step, as internal jugular and subclavian vein approaches typically involve the head down (Trendelenburg)
position. Many patients find this position uncomfortable and it is relatively contraindicated with elevated intracranial pressure as well as associated with increased
nosocomial pneumonia risk. For internal jugular and subclavian vein cannulations, however, the head down (Trendelenburg) position has been shown to significantly
increase the diameter of these vessels. A secondary benefit of this position is that not only is the size of the target vessel increased, but also the pressure within the
vessel is greater, making the venipuncture simpler. Finally, in spontaneously breathing patients, because of the increased venous pressure, this position significantly
reduces the risk of venous air emboli. In general, the more Trendelenburg that the patient can tolerate, the better, as these advantages are enhanced as the venous
pressure increases.
For femoral venous approaches the patient should be flat without any bend at the hip with the toes positioned toward the ceiling or allowed to fall slightly to the side.
Patient comfort is essential as in complicated cases these positions may have to be maintained for longer than anticipated. Operator comfort when performing this
procedure is also important as an awkward or uncomfortable position during the cannulation attempt will result in fatigue and a lower chance of success. Beds that can
be placed in full Trendelenburg
Page 387
position and then raised to the appropriate height for the operator are ideal. In the ICU setting, significant attention to positioning of ventilator tubing, IV tubing, and
monitors is often necessary.
Patients with difficulty breathing, especially patients with congestive heart failure, may not be able to tolerate even lying flat, let alone head down position. For these
patients, attempting to cannulate the vein quickly and then allowing the patient to resume a semirecumbent position once the risk for air embolization is minimized (i.e.
the guidewire is in place), is a possible alternative. Once the guidewire is in place, the rest of the procedure can be performed with the patient in a semirecumbent
position with the head even as high as 45°. Some patients will require intubation and mechanical ventilation before this line placement. We do not recommend
attempting this procedure with the head up in a spontaneously breathing patient. In patients with increased intracranial pressure, an attempt can be made with the patient
lying flat, but Trendelenburg position is also acceptable as long as the intracranial pressure (ICP) is being monitored. Pretreatment with pentobarbital, hyperventilation
and mannitol can also be employed. The goal should be to have the head down for as short a period of time as possible.
23.3.4 Cannulation procedure

The vessel is entered with either an angiocath or a hollow steel needle. Once the vessel is entered and free flow of venous blood is confirmed, the syringe is gently
twisted off the needle or the angiocath advanced into the vessel and the guidewire inserted to approximately half its length. Most guidewires are approximately twice the
length of the catheter. For those insertions where there is a potential for the guidewire to enter the heart (internal jugular and subclavian vein approaches) the EKG
should be monitored for atrial and ventricular dysrrhythmias. If arrhythmias are encountered, simply withdrawing the wire from the heart, but not out of the vessel, is
usually all that is necessary to terminate the dysrhythmias. Persistent arrhythmias generally respond to a bolus injection (100 mg) of lidocaine (see Chapter 11).
Once the wire is in place, the needle or angiocath is withdrawn. The wire should be held at all times to prevent inadvertent loss within the venous system. Wires do
not become pulled into the venous circulation by the flow of blood, but can be inadvertently pushed inward with catheter insertion. Simply clamping the guidewire
above the skin frees up the operator’s other hand as needed. A Number 11 scalpel blade is be used to cut down along the wire, with the cutting edge of the blade
facing away from the wire. The scalpel is then removed, turned 180° and a similar cut is made on the other side of the wire. The depth and width of the cut should be
individualized. In thin patients, especially where the vein is visible just below the skin (i.e. some internal jugular vein sites), a very superficial nick of the skin will suffice.
In others with lots of muscle mass, fatty tissue or old scarring, deeper cuts will be necessary. In all circumstances, the goal is to open the skin and possibly some of the
adjacent subcutaneous tissue and muscle, not the vein itself.
The dilator is now threaded over the wire through the skin and muscle and into the vein. Rarely does more than half the length of the dilator need to go through the
skin. Dilators tend to be very stiff and should be advanced only if they move fairly smoothly. Occasionally, the dilator will become stuck on a kink in the wire and will
be very difficult to advance. This is remedied by pulling the wire back 12 cm, moving the kink back within the dilator, and then proceeding. At no time should the
dilator be further advanced if significant resistance is encountered. If the tip of the dilator splays out, it becomes useless and a new dilator will be needed.
Once the dilator is removed, expect fairly significant bleeding through the venipuncture. This is minimized by direct pressure, but more completely by rapid insertion
of the central
Page 388
venous catheter. The catheter should have already been prepared by its removal from the package, placement of appropriate caps and flushing of the catheter, typically
with a heparinized solution. For a multilumen catheter, it should also be noted, ahead of time, which port the wire will exit from so that this cap can be intentionally left
off.
The guidewire is now threaded back through the catheter until it can be grabbed at the distal end. Once the guidewire is securely grasped, the catheter is advanced
over the wire to the proper position within the vein. When using the femoral approach, the entire catheter should be inserted into the vein. For insertions where the tip of
the catheter may potentially enter the heart, i.e. subclavian vein or internal jugular vein sites of entry, the length of catheter to be inserted will vary based on the patient’s
size. Correct insertion should leave the catheter in the distal superior vena cava, but outside the right atrium. Optimal insertion length for those catheters placed via
subclavian or internal jugular veins is 15 cm7–9. In shorter patients, especially for rightsided approaches, shorter insertion distances (typically 13 cm) are needed to
keep the catheter tip outside of the heart.
The wire is now withdrawn, the catheter is recapped, and all ports are checked for blood return and then flushed. Occasionally, one or more ports will flush easily,
but not produce a blood return on aspiration. This can be caused by suction between the catheter port and the vessel wall, and manipulating the catheter within the vein
or twisting it often solves the problem.
23.3.5 Postinsertion tasks

Once the catheter has been flushed, it should be secured in place, either with a suture through the appropriate attachment points on the catheter, or using one of the
newly available securing devices such as the Stat Lock3. Sterile blood cultures may be drawn at this point if desired. It is the operator’s responsibility to dispose of all
sharp items (needles, scalpels, sutures) and bloodcontaminated supplies (dilator, angiocath) in the proper receptacles. A dressing is now applied. Radiographic
confirmation of placement and potential complications (i.e. pneumothorax) is not always needed clinically but may be mandated by individual hospital policy.
The catheter tip should be located outside the heart, ideally parallel to the superior vena cava (Figure 23.1). A procedure note including the date, time, indication,
consent, anesthetic, and particular anatomic or procedural issues should be recorded.
Figure 23.1 Catheter positions relative to the superior vena cava (SVC) based on insertion site and proximity to the right atrium.
Catheter angulation relative to the SVC is minimized when the catheter tip is in the distal SVC near the right atrium for all
insertion locations other than the right internal jugular vein (RIJV). RIJV insertions tend to be parallel to the vessel wall
regardless of location within the SVC.
Page 389
23.3.6 Precautions
Venous catheterization from any selected site will be unsuccessful up to 35% of the time. We do not recommend more than four needle passes at any selected site.
Success is most likely with the first needle pass and diminishes on subsequent attempts, but each pass of the needle carries a risk for puncture of adjacent structures.
Thus, the riskbenefit ratio increases with subsequent passes of the needle.
If inadvertent arterial puncture occurs with a finder needle or angiocath, the needle or catheter should be withdrawn as soon as this is recognized and pressure
applied to compressible vessel. Direct pressure will be required for longer periods in those patients who do have a coagulopathy. In those patients with coagulopathy in
whom this complication is recognized with the catheter already in place, it is probably prudent to correct the coagulopathy before withdrawal of the catheter. Every
effort should be made to confirm intravenous and exclude intraarterial placement before using a vessel dilator. If there is any doubt, the pressure in the vessel can be
transduced or estimated by the height of blood in attached tubing, or a sample sent for blood gas analysis. Visual inspection of blood pulsatility and color can be
unreliable in patients who are hypoxic, in shock, or compromised by tricuspid regurgitation and volume overload.
If the arterial picture is not recognized with the finder needle and the vessel has been dilated (as for triplelumen or cordis sheath placement) it is best to leave the
catheter or dilator in place and obtain an immediate vascular surgery consultation for repair of the vessel.
Besides the hemodynamic clues or obvious external or subcutaneous bleeding, a new pleural effusion in a patient in whom a subclavian vein or low internal jugular
venopuncture has been attempted or completed should warrant investigation for hemothorax and evaluation for possible vessel injury.
Never withdraw guidewires against resistance through a needle as the spiral wire may stretch apart or break completely. In those circumstances where the needle is
used to guide the wire into the vein, special precautions must be taken to insure the integrity of the wire. If the wire must be withdrawn through the needle and resistance
is met, the needle and catheter should be withdrawn simultaneously to prevent shearing of the wire against the needle. Shearing off the wire in an intravascular location is
a serious morbid complication. If the wire is withdrawn and intravenous location is reconfirmed, often flipping the wire over from the J end to the straight end will
facilitate easier passage.
23.4 Femoral venous cannulation

The femoral venous site typically does not result in central venous cannulation, thus pressure measurements, specifically central venous pressure, cannot be taken.
However, for all other reasons for central venous cannulation, the femoral vein is equivalent to other sites. Although some data suggest increased infectious disease risks
for this approach, other data do not10,11. Thrombotic complications of the deep venous system are probably related to use of femoral venous catheters, especially in a
pediatric population11–13.
Benefits of the femoral approach include accessibility during acute resuscitation, reduced risk of significant mechanical complications, an easily compressible vessel,
and, for patients with respiratory failure or increased intracranial pressure, no need to place the patient in Trendelenburg position. Femoral venous cannulation may be
the simplest and safest approach for inexperienced operators13,14.
23.4.1 Anatomy
The femoral pulse is the primary guide for femoral venous cannulation. The vein is located just medial to where the pulse is palpated. The entry site of choice is below
the inguinal ligament wherever the pulse is most prominent. Above the inguinal ligament, the vein runs posterior by and becomes the iliac vein. Below the inguinal
ligament the vein
Page 390
lies within the femoral sheath just medial to the femoral artery. In pulseless patients, 2 cm below the inguinal ligament and approximately onethird of the way from the
pubic tubercle to the anteriorsuperior iliac spine will serve as a good starting point (Figure 23.2).
Once the artery is located by palpation, inserting the needle approximately 1 cm medial to the pulse often results in success. The needle is either inserted in a
perpendicular plane or 15–30° off the perpendicular plane with the needle oriented in the direction of blood flow. Negative pressure should be applied to the syringe
throughout the entire procedure once the needle has passed through the skin. Evacuating air from the syringe before the procedure will minimize clotting. Although
anatomically a superficial structure at this location, 4–6 cm depth of penetration or more is typically required to access the vein. Once a free flow of blood is
established, the procedure proceeds as described in Section 23.3.4.
23.4.2 Caveats
Hip flexion has the potential to cause femoral catheter kinking. For sick ICU patients who must be maintained in bed, this is often not an issue. Femoral catheters tend
to function even with the head of the bed up to 30 or 45°. If catheter location is inhibiting patients’ weaning or rehabilitation progress, a new site should be chosen.
Although a finder needle may be helpful for femoral venous cannulation, those included in most kits may not be long enough to locate the vessel. For the same
reason, using a catheter over the needle technique for initial femoral venous access does not provide any advantage and may actually make the procedure more
cumbersome.
23.5 Internal jugular vein

The approach from the right internal jugular vein provides the most direct access to the heart, facilitating placement of SwanGanz catheters, transvenous pacemakers
and
Figure 23.2 Anatomic location of femoral vein below the inguinal ligament
Page 391
endomyocardial biopsy. Internal jugular venous cannulation is usually the preferred route for anesthesiologists who are already working at the head of the bed and may
have limited access to other anatomic locations. In patients with coagulopathies, the clavicular routes are relatively contraindicated because of inability to compress a
bleeding vessel. Finally, because of the extremely low incidence of pleural complications associated with internal jugular venous cannulation, this technique is also
preferred in those patients in whom a pneumothorax would not be well tolerated; specifically patients with severe acute or chronic lung disease, or these about to
undergo anesthesia and positive pressure ventilation without a chest radiograph to exclude pneumothorax.
23.5.1 Anatomy
The internal jugular vein emerges from the base of the skull through the jugular foramen and enters the carotid sheath anterior and lateral to the internal carotid artery.
The cephalad internal jugular vein usually lies medial to the anterior portion of the sternocleidomastoid muscle, runs beneath the triangle formed by the two heads of the
sternocleidomastoid in its middle segment, and medial to the anterior portion of the clavicular head of the muscle in the lower segment (Figure 23.3). Beneath the
clavicle, the right internal jugular vein joins the subclavian vein to form the innominate vein, which continues in a straight path to the superior vena cava. The left internal
jugular vein joins the left subclavian vein at almost a right angle. Consequently, any catheter inserted through leftsided approaches must negotiate this turn. The local
anatomy is equally important, as complications occur from injury to these structures. The carotid artery usually lies medial to the internal jugular vein throughout its
course within the carotid sheath. The artery is usually posteriormedial to and sometimes partially enveloped by the internal jugular vein. Medial and posterior to the
internal jugular vein, and the carotid artery, lie the stellate ganglion and the cervical sympathetic trunk. Deep to this are the roots of the brachial plexus. The phrenic
nerve is usually lateral to the vascular bundle, whereas the vagus nerve lies between the internal jugular vein and the carotid artery. Branches of the laryngeal nerves are
medial to the vascular bundle. Near the junction of the internal jugular vein and the subclavian vein is the pleural dome. This is usually higher on the left. The thoracic
duct lies behind the left internal jugular vein and usually enters the superior margin of the subclavian near the jugular subclavian junction.
23.5.2 Ultrasound and internal jugular venous cannulation

The use of ultrasound has significantly enhanced safety while better defining anatomy of internal jugular venous cannulation15,16. Information derived from ultrasound
examination gives better appreciation of the following:
• The caliber of the vein usually increases as it progresses toward the clavicle.
• The distance from the skin surface to the vein increases as it approaches the clavicle.
Figure 23.3 Typical anatomy of IJV. The vein lies just lateral to the carotid artery (CA) below the sternocleidomastoid muscle (SCM).
Multiple nerves are located in this area.
Page 392
• Palpation or retraction of the carotid artery during cannulation attempts markedly decreases and occasionally obliterates the ultrasound image of the lowpressure
internal jugular vein. Often during venipuncture, the vein will collapse simply as a result of the pressure exerted by the needle. For this reason, it is important to
maintain negative pressure on the syringe as the needle is backed out through the skin, as this is often the time when successful cannulation is noted. Intravascular
volume loading either during or before the procedure may be helpful in increasing the size of the target vessel.
• Occasionally the carotid artery is found posterior to the internal jugular vein, making it difficult to cannulate with usual familiar landmarks (Figure 23.4).
• Placing the patient in the head down Trendelenburg position or using the Valsalva maneuver increases the size of the vessel.
• The internal jugular vein is a superficial structure. The average perpendicular distance from the skin to the vein at the level of the cricoid cartilage is approximately 1.5
cm. For this reason, the needle should rarely be inserted to beyond 4 cm in most patients (Figure 23.4).
• Multiple anatomic variations exist, and for those patients in whom cannulation proceeds with difficulty we recommend early application of ultrasound. Use of
ultrasound for cannulation is now recognized as an important safety initiative by the Agency for Healthcare Research and Quality (AHRQ).
23.5.3 Cannulation procedure: median approach

Although multiple approaches to the internal jugular vein have been described, our experience has been that knowledge of two approaches (median and anterior) is
sufficient.
For the median approach, the two heads of the sternocleidomastoid should be located and followed to their insertions on the clavicle and sternum, respectively. The
clavicle forms the base of the sternocleidomastoid triangle. The sternal notch should also be identified as this aids identification of the sternal head of the
sternocleidomastoid. The carotid artery should be gently palpated and localized. The angle of the mandible along with the junction between the clavicle and the sternum
should also be identified. Identification of the thyroid
Figure 23.4 Ultrasound image of IJV overlying carotid artery.

Page 393
cartilage and the cricothyroid membrane may also be useful in difficult cannulations. The more landmarks that can be identified, the easier it will be locate the vessel.
However, in many patients the only identifiable landmark will be the carotid pulsation. Even in massively obese patients, the angle of the mandible and medial
articulation of the clavicle can be identified. Bisecting a line between these two points will approximate the apex of the sternocleidomastoid triangle. This is often found
at the level of the cricothyroid membrane. Identifying these additional structures will provide additional assurance that initial site selection is correct.
At a minimum, the carotid artery needs to be identified. Except for urgent or emergent venous access, we use a 22gauge 2inch finder needle. The incidence of
carotid artery puncture during internal jugular vein cannulation may be as high as 10%17,18. If this occurs with a small finder needle, there is typically no adverse
consequence and the procedure may still proceed rapidly. If the carotid artery is punctured with either the 18gauge hollow wall steel needle or the 16gauge angiocath
over 20gauge needle, a minimum of 5 min of direct pressure over the vessel must be applied.
The needle is placed at the apex of the triangle just lateral to where the carotid pulsation is felt and is advanced in a perpendicular direction with the plane of the
needle being midway between the midsternum and ipsilateral nipple (Figure 23.5). Once the vein is found, this needle can be left in place while the catheterizing needle
is placed just caudad to it using the same angle of entry and approximate depth. If this proves too cumbersome, the finder needle can be withdrawn, mentally recording
the angle and the depth. The cannulating needle is then passed through the identical skin puncture site.
We recommend vessel cannulation using an angiocath rather than a needle, although both are typically provided in most kits. Most kits include a 20gauge needle
with a 16gauge angiocath over it. Puncturing the carotid artery with a 20gauge needle is less problematic then when larger cannulating needles are used. The
angiocath provides very secure venous access once the vessel is entered, as the entire length of the angiocath can be inserted down the vessel. A problem with this
technique is that if the vein is not entered on the first pass of the needle, the angiocath must be held together as it is withdrawn, otherwise the angiocath will separate and
potentially make that system nonfunctional.
A perpendicular approach to the vessel or approximately 30° off the perpendicular with the tip of the needle pointed caudad is the preferred angle to the skin. Once
the vessel
Figure 23.5 Position of cannulating needle for median approach to the IJV. The needle is placed at the apex of the sternocleidomastoid
triangle using a steep angle of incidence to the skin (60–90°). It is advanced in a plane midway between the center of the
sternum and the ipsilateral nipple. The carotid pulsation should be just medial to the venipuncture.
Page 394
is entered, the angiocath is further advanced slightly (1–2 mm) while simultaneously placing the syringe more in line with the anatomic direction of the vessel. The syringe
is now reaspirated to confirm intravenous location and the angiocath is threaded into the vessel with a twisting motion. Once the angiocath is in place, direct pressure
over the internal jugular vein can be applied using three or four fingers to minimize the risk of venous air embolization and back leak. If a hollow steel needle is used,
once the vein is accessed the needle should be advanced slightly to assure intravenous location of the entire needle tip. The needle is not advanced further at this point,
to prevent laceration of the back wall of the vessel. The guidewire is inserted through the needle hub. Using the plain steel needle requires more attention to detail as
slight movements of the needle may inadvertently dislodge the needle tip. The hub of the needle needs to be occluded at all times to prevent air entry. Once the wire is
inserted through the needle securely into the vein, further catheterization is similar to the other technique. With the angiocath in place, the wire is threaded through the
angiocath, and the angiocath is withdrawn.
If the angiocath becomes unusable and a replacement is to be used, two rules should be followed: assure that the wire will fit through the newly selected angiocath;
most guidewires will fit through either a 16 or 18gauge angiocath. A 1 3/4 inch angiocath would be the minimum length with 2 1/4 inch being preferred.
23.5.4 Cannulation procedure: anterior approach

One advantage of the anterior approach is that it can be performed with the patient’s head in the midline position. Insertion will be at the level of the cricoid, which is
palpated as the first landmark. An imaginary horizontal line is drawn to the anterior border of the sternocleidomastoid. The carotid should be palpable here, and the
initial jugular will be just lateral to the carotid. With the needle parallel to the body’s long axis and entering at a 30–45° angle to the perpendicular plane, the finder
needle is inserted 1–1.5 cm lateral to the carotid. The internal jugular is usually very superficial, no more than 2–4 cm under the skin except in obese or edematous
patients. Once the vessel is identified with the finder needle, the angiocath or largerbore needle is inserted. As blood is aspirated, the needle or angiocath is aligned
more closely to the vessel’s direction and free flow assured. The Selinger technique is then used to insert first a guidewire and then the final catheter or introducer. As
with the median approach, the carotid artery must be identified and care taken not to exert pressure sufficient to activate carotid reflexes. The risk of carotid puncture
using the anterior approach argues for the finder needle technique previously described.
23.6 Subclavian vein cannulation

All approaches to the subclavian vein involve the chest and for this reason the risk of pneumothorax with this approach is significant. Most reports indicate an
approximately 5% risk of this complication; however, in experienced hands it is significantly less19,20. Although the risk of pneumothorax is greater with this approach,
the subclavian approach might be preferable to internal jugular venous cannulation during acute resuscitation when the airway is being secured, with difficult landmarks
in the neck, and for patient comfort with longterm cannulations.
23.6.1 Anatomy
The anatomic relationship of the subclavian vein to the subclavian artery and the clavicle along with some of the muscular insertions are shown in Figure 23.6. The
subclavian vein continues from the axillary vein as it crosses over the first rib then proceeds medially, arching cephalad towards its junction with the internal jugular vein
close to the manubrialclavicular junction. Once joined with the internal jugular vein it forms the brachicephalic vein where it then enters the thorax. As the vein
proceeds medially over the first rib and under the clavicle, it is anterior and caudad to the subclavian
Page 395
Figure 23.6 Sagittal view of subclavian vein and pertinent superf icial and deep perivascular anatomy. The vein is an anterior structure
found just underneath the clavicle. Deeper venipuncture attempts are likely to injure either the subclavian artery or pleura.
artery. The vein and artery at this level are separated by the anterior scalenus muscle. As illustrated in Figure 23.6, the vein resides just below the clavicle at this point.
This is particularly pertinent regarding cannulation of the vein. Approaches just posterior to the clavicle and parallel to it are most likely to result in successful
venipuncture. If the position of the needle is moved to form a greater angle relative to the clavicle, the likelihood of subclavian artery or pleural puncture is greater. The
vein is covered anteriorly by the pectoralis major muscle, which will always be penetrated first. The medial portions of the vein are completely covered by the clavicle,
which will also serve as a guide during cannulation. The pleural dome, which is higher on the left, lies posterior and superior to the subclavian artery. The thoracic duct,
which may be a large structure on the left side, crosses the vein just medial to its junction with the internal jugular vein.
Subclavian veins are usually large and often patent as a result of their fibrous attachments. This may be the preferred approach for patients who are hypovolemic or
in shock. Multiple approaches to the subclavian vein either above or below the clavicle have been described, but only the two most useful will be presented in detail. All
approaches to the subclavian vein are made simpler by placing a small rolled towel between the scapulae to allow the shoulders to fall back. This positioning increases
the distance between the first rib and the clavicle.
23.6.2 Supraclavicular approach to the subclavian vein

Supraclavicular approaches to the subclavian vein have primarily been popularized in the emergency room setting especially for hypovolemic patients. This approach to
the central circulation provides probably the largest target
Page 396
vessel in the body; the junction of the internal jugular vein (IJV) and subclavian vein. The clavicular notch approach provides easily identifiable landmarks with a very
high success rate. As this approach is not terribly different from the medial approach to the internal jugular vein, we find it is easiest with the operator standing at the
head of the bed. The base of the triangle formed by the clavicle and the two heads of the sternocleidomastoid muscle is located first. The insertion point for the needle
will be just lateral to the midpoint of the base of the triangle, which will also be lateral to where the carotid pulsation is felt. The needle is advanced parallel to the
sagittal plane at approximately a 30° angle to place the tip of the needle just under the clavicle (Figure 23.7). The vein is typically entered within 4 cm.
Figure 23.7 Superclavicular approach to subclavian vein (see text).
Picking the needle up towards a more perpendicular location relative to the chest wall is more likely to injure posterior structures. If venipuncture is not accomplished, a
slightly more lateral approach may be useful. Essentially, access is at the junction between the internal jugular and subclavian veins. Once blood is aspirated from within
the vein, cannulation proceeds as previously described. Because of its similarities to median approaches to the IJV, the use of a finder needle and the angiocath
techniques described for the IJV are appropriate in this location. Several variations on this technique exist. Simply moving higher up within the triangle, but using the
same needle angles, approximates the low median approach to the internal jugular vein. This is particularly useful in patients who have thick clavicles, which may cause
the insertion angle to be greater than desired.
23.6.3 Alternate supraclavicular approach

A second variation on the supraclavicular technique is to place the tip of the cannulating needle just lateral to the clavicular head of the sternocleidomastoid. The tip
of the needle is placed at the junction between the clavicle and the clavicular head of the sternocleidomastoid muscle. The anterior scalene muscle behind which lies the
subclavian artery may be appreciated in some patients behind the sternocleidomastoid. The needle is directed somewhat medially to a location below the clavicle above
the first rib using a fairly steep approach, approximately 45–60° until blood flow is returned. At that point, the needle is placed more in line with the expected route of
the subclavian vein at this location. Again, because the depth of insertion is not expected to be beyond 4 cm in this location, a finder needle and angiocath techniques
are appropriate. Although not reported in the literature, because of the initial steeper angle of incidence relative to chest wall, there is some concern for a higher
incidence of pneumothorax using these approaches.
Page 397
23.6.4 Infraclavicular approaches to the subclavian vein

Infraclavicular approaches to the subclavian vein are probably the most common access to this vessel. Positioning a patient for this approach will help facilitate venous
access. In addition to placing a roll between the scapulae, another useful technique is to have the arm pulled down under slight traction on the side to be cannulated.
For all infraclavicular approaches, we recommend standing at the side of the bed. The important anatomic landmarks are the clavicle and the suprasternal notch.
Multiple initial skin punctures along the clavicle beginning within the distal third and progressing out toward the lateral third have been described. We have had the most
success beginning with a point at the midclavicle, particularly where the clavicle begins to bend cephalad and the bulk of the insertion of the pectoralis muscle can be
felt. These two structures, the clavicle on top and the pectoralis on bottom, form the head of an arrow that aims at the suprasternal notch (Figure 23.8). The
cannulating needle is grasped such that the tip of the needle is controlled by the index finger. The needle is then inserted directly in line with the arrowhead formed by
the clavicle and the major insertion of the pectoralis muscle approximately 1 cm lateral to the clavicle so that it can easily be slipped underneath. The purpose of
controlling the tip with the index finger is so that the needlesyringe apparatus can be pressed down toward the patient’s bed and the needle advanced under the
clavicle with minimal lifting of the syringe. The plane in which the needle travels is as close to parallel to the plane of the floor as possible. If the clavicle is encountered
and the hub of the syringe needs to be picked up so that the needle can slide underneath the clavicle, we do not recommend picking the syringe up by any more than
30° off the horizontal plane. Angles above this are much more likely to proceed into deeper structures, such as the subclavian artery and the pleura. It is significantly
easier and less risky in
Figure 23.8 Infraclavicular subclavian vein approach. The clavicle on top and the insertion of the pectoralis major on the bottom form
the head of an arrow (where the intensivists’ index finger is In this figure) that aims at the suprasternal notch. The needle tip is
pushed toward the bed and controlled by the index finger to allow the most tangential (close to 0°) insertion possible relative to
the chest wall. The needle is advanced toward the suprasternal notch.
these circumstances to choose an insertion point slightly more lateral so that the tip of the needle can be slipped under the clavicle without having to angle the syringe up
significantly. The direction of needle travel is now toward the suprasternal notch; constant aspiration on the syringe is maintained as the needle is advanced. In obese
patients or those with significant muscle mass, it is not uncommon to have to place the entire length of the needle through the skin, and even encounter dimpling of the
skin before the vein is entered.
Page 398
This provides a particularly challenging situation as the syringe is twisted off the needle, because the pressure from the skin alone may push it out of the vein. It is very
important in this circumstance, therefore, to hold the needle in the position in which it is known to be within the vein. The Raulerson syringe, which allows for placement
of the guidewire through the hub of the syringe, is particularly useful for infraclavicular subclavian vein approaches when encountering this problem. With this device,
once the vein is entered the guidewire can simply be inserted along the hub of the syringe without having to twist the needle off. If in the first pass toward the
suprasternal notch the vein is not entered, withdrawing the needle and adjusting the tip slightly more caudad and then reinserting often locates the vessel. If unsuccessful
after several passes of the needle (typically no more than four), it is likely that a location more proximal along the clavicle toward the suprasternal notch will be
successful compared with a more lateral initial skin puncture.
The clavicle will often be encountered during this approach. As the vein sits just below the clavicle, it is important that the most posterior portion of the clavicle is
appreciated and that the needle does not track significantly below this. For this reason, we recommended pushing the needlesyringe assembly down in a plane parallel
to the floor utilizing the index finger. Other physicians like to ‘march’ the needle down the clavicle in small increments until the needle finally slips past. Either approach is
fine, as the emphasis should be that the entire assembly of syringe and needle should not be angled to greater than 30° off the plane parallel to the floor.
Orienting the bevel of the needle caudad will safely position the guidewire within the superior vena cava. Because of the typically deeper insertion depth encountered
when accessing the subclavian vein, we do not recommend using either a finder needle or an angiocath in this location. Once the vein is entered, central venous
cannulation proceeds as previously described.
23.7 Introducer placement

Techniques for placing introducers suitable for the insertion of SwanGanz catheters or transvenous pacemakers are different only because of their size. Otherwise,
placement of these devices is similar to central venous catheter placement, with a few specific pointers.
23.7.1 Techniques for placing Cordistype introducers

The introducer should be fully assembled with the dilator in place before venous access. The dilator typically goes through the hemostatic valve and then exits the distal
tip of the introducer. It is sometimes easier to first use the dilator by itself to dilate the skin and muscle and then place it through the introducer rather than assembling the
whole device initially. A larger skin incision will be necessary than with a central venous catheter. Although more pressure is required to place the introducer through the
skin, excessive pressure is not warranted and may damage the introducer. If the introducer is damaged in any way, it should be replaced. Once the introducer assembly
is securely placed within the vein, the guidewire and dilator are withdrawn simultaneously. At this point, the stopcock, if utilized, should be open to allow the tubing to
fill with blood, and then connected to the desired IV set. These connections should be screwed securely together to prevent inadvertent loss of continuity of the system
and the attendant risk of either venous air embolism or bleeding. Intravenous fluids should be continuously run through the introducer to assure patency and the
introducer should be securely affixed to the skin. Multiple types of introducers are available depending on the specific requirement. The introducer is typically 1 French
greater in diameter than the catheter that will be inserted through it; that is, the common 8 1/2 French introducer will accept a 7 or 7 1/2 French pulmonary artery
catheter. The technique for floating a pulmonary artery (SwanGanz) catheter is covered in Sections 5.3.3 and 5.3.4.
Page 399
23.8 Complications and their prevention

Complications related to central venous catheterization generally fall into two categories:
• complications occurring during central venous catheter insertion: mechanical complications (Table 23.3);
• complications occurring during the indwelling phase of central venous catheterization: septic complications.
23.8.1 Septic complications, catheters and antibiotic coating

Septic complications related to central venous catheterization typically occur after the catheter has been in place for greater than 72 h and are rare before 48 h.
Catheterrelated blood stream infection occurs in up to 6% of catheterized patients21–23. This complication is more likely in patients who are catheterized for longer
lengths of time and are more severely ill, with multiple comorbidities and/or therapy with total parenteral nutrition through the central venous catheter. Frequent routine
changes of the central venous catheters have not been shown to decrease the incidence of this complication. Antibioticimpregnated or coated catheters, on the other
hand, have been shown to result in statistically significant decreases in catheterrelated blood steam infections, and may be a valuable technique for those patients with
expected prolonged duration of catheterization24–26. In general, central venous catheters
Table 23.3 Major complications of central venous catheterization
IJV SCV Femoral

Arterial puncture 5–10% 1–4% 7%
Pneumothorax Rare 2–5% 0%
DVT – – Up to 20% esp. pediatrics
IJV or SCV
Tip malposition within heart ~50% (20 cm CVC)
<4% (CVC inserted to 15 cm)
Miscellaneous complications
IJV SCV
Tracheal puncture Thoracic duct laceration
Perforation of endotracheal tube cuff Phrenic nerve paralysis
Horner’s syndrome Brachial plexus palsy
Vocal cord paralysis
Cerebral vascular accident
Both
Air embolism
Myocardial injury with tamponade
Hydromediastinum
Catheter embolism
Subcutaneous fluid collection
Fatal exsanguination
Dysrhythmias
Compression of adjacent structures by hematoma
Others
CVC, central venous catheter.
Page 400
should be removed once they are no longer needed.
23.8.2 Guidewire exchanges

Guidewire exchanges of catheters are reserved for those patients when the risk of venopuncture is thought to be significant (i.e. in patients with severe coagulopathy or
serious problems with the alveolararterial oxygen gradient who would not tolerate the complications of a pneumothorax). In these circumstances, we recommend that
as part of the guidewire exchange procedure, quantitative culture of the catheter tip is obtained. If the culture ultimately grows out greater than 15 colony forming units,
the tract through which the catheter was exchanged is considered to be contaminated and central access should then be moved to a new site22,23. However, for those
catheters with less than 15 colony forming units of growth using this technique, the newly inserted catheter may remain in place and the patient has avoided
venipuncture.
For pulmonary artery catheters, there is no consensus on the length of the time that these catheters may remain in place. On the basis of a fairly exhaustive review of
current literature, the CDC (Centers for Disease Control) recommends that these catheters be replaced every 5 days22,23. It is our experience that the majority of
relevant information that can be uncovered with pulmonary artery catheterization has been incorporated into the treatment plan before 5 days, and that pulmonary artery
catheterization beyond 5 days is rarely necessary.
23.8.3 Guidewire exchange technique

Sterile guidewire exchanges can be tricky and require a fair amount of expertise. The procedure is as follows:
1. A sterile field is prepared by first scrubbing the skin and then saturating the visible parts of the catheter with an antiseptic solution. The solution should be allowed to
soak on the catheter and skin for at least 5 min.
2. The catheter, now disconnected from any IV tubing, is now slowly withdrawn from the skin approximately 4–6 cm. The catheter is now firmly grasped at the skin
level or clamped, and cut with sterile scissors approximately 2–3 cm above the skin. This will assure that the part of the catheter being cut was either intravenous or
below the dermis. The catheter is cut at a sharp angle to allow easy identification of its internal lumens. The proximal portion of the catheter is discarded and the
guidewire is now inserted through the cut portion of the catheter back into the vein. Close attention to detail is important so that the remaining catheter fragment will
not be pushed intravenously.
3. Once the guidewire is in place, the catheter can then be withdrawn, leaving the guidewire in an intravenous location. Further catheterization can proceed as
previously described.
23.8.4 Catheter malposition

Malposition of central venous catheter tips is the most common complication following central venous catheterization. A review of the practice at five medical centers
involving 350 catheterizations showed the incidence of malpositioned tips within the right atrium to be greater than 50%7. Using an ECGguided technique, we were
able to eliminate this complication. Simply choosing an insertion length of 15 cm regardless of the site of either internal jugular or subclavian vein catheterization will
further reduce the incidence of this complication to less than 5%8. Catheter tip position within the right atrium may result in fatal cardiac tamponade typically from the
intravenous solution. The Food and Drug Administration (FDA) and manufacturer’s guidelines recommend placing the catheter in the superior vena cava above the
heart27, 28.
Superior vena caval injuries may also occur especially for those catheters that contact the superior vena cava at high angles of incidence29,30. These injuries often
require thoracotomy drainage of the pleural space and
Page 401
possibly repair of the blood vessel itself. Placing catheters from the right internal jugular vein virtually eliminates the possibility of this complication. Multiple other
malpositioned catheters have been described, although if intravenous these rarely cause significant problems. The primary purpose of the postprocedure chest film is to
check for catheter course and tip location.
23.8.5 Dysrhythmias
Dysrhythmias are fairly common during central venous catheterization, especially during pulmonary artery catheter insertion31–33. This is thought to be due to
mechanical irritation of the right ventricle and rarely causes significant hemodynamic morbidity as it is almost terminated by either advancing the catheter into the
pulmonary artery or withdrawing it back to the right atrium. This complication may also occur during guidewire insertion or during placement of the central venous
catheter inadvertently within the heart.
23.8.6 Pneumothorax
Pneumothorax is the most significant common complication following clavicular approaches to the central circulation. It is reported to occur up to 5% of the time
depending on which series is reviewed34,35. Pneumothorax is a rare complication of internal jugular vein cannulation. Of 1338 cases of IV venipuncture reported in
large series, the overall pleural complication rate was less than 0.1%7. Numerous unique and rare complications have been reported to occur with both internal jugular
and subclavian vein catheterizations, usually resulting from laceration of adjacent anatomic structures, vessel dissection or injury caused by vessel dilators or errantly
placed catheters.
23.8.7 Ultrasound guidance

Numerous data reported over the past 10 years have shown that ultrasound localization of the internal jugular vein results in faster and safer central venous
catheterization15,36,37. The primary advantage of the ultrasound technique is realtime visualization of the vessel to be cannulated. Although we do not use this technique
in all patients, we strongly recommend it for those patients in whom the catheterization is anticipated to be difficult because of anatomic variations (i.e. obesity, short
neck) or the risk of a complication related to inadvertent puncture of adjacent structures are high (i.e. arterial puncture in a patient with serious coagulopathy or pleural
injury for patients with large alveolararterial oxygen gradients). Early recognition of complications in therapy is mandatory for anyone putting in a central venous
catheter. Although not considered the standard of care at this time, ultrasound guidance is now recognized to improve patient safety38,39. We have found ultrasound
techniques extremely useful, especially as part of our training program in vascular access for housestaff.
References
1. Bedford, R.F. (1977) Radial artery function following percutaneous cannulation with 18 and 20 gauge catheters. Anesthesiology 47:37.
2. Weiss, B.M. and Gattiker, R.I. (1986) Complications during and following radial artery cannulation: A prospective study. Intensive Care Med. 12:424.
3. Ventec (2000) Stat lock. Catheter Securement Device. Ventec International, Inc., San Diego, CA.
4. Kaye, J., Heald, G.R., Morton, J., et al. (2001) Patency of radial arterial catheters. Am. J. Crit. Care 10:104–111.
5. Joynt, G.M., Kew, J., Gomersall, C.D., et al. (2000) Deep venous thrombosis caused by femoral venous catheter in critically ill adult patients. Chest 117:178–183.
6. Parsa, M.H., Taboraf. (1986) Central Venous Access in critically ill patients in the Emergency Department. Emergency Department management of critical illness.
Emer. Med. Clin. North. Am. 4:709–717.
7. Durbec, O., Viviand, X., Potie, F., et al. (1997) Lower extremity deep vein thrombosis: A prospective, randomized, controlled trial in comatose or sedated patients
undergoing femoral vein catheterization. Crit. Care Med. 25: 1982–1985.
8. McGee, W.T. and Moriarty, K.P. (1996) Accurate placement of central venous catheters using a 16cm catheter.
Page 402
J. Intensive Care Med. 11: 19–22.

9. McGee, W.T., Ackerman, B.L., Rouben, L.R., et al. (1993) Accurate placement of central venous catheters: A prospective, randomized, multicenter trial. Crit.
Care Med. 21:1118–1123.
10. McGee, W.T. and Martin, R.T. (1999) Safe placement of central venous catheters. Crit. Care Med. 27:A47.
11. Kanter, R.K., Zimmerman, J.J., Strauss, R.H., et al. (1986) Central venous catheter insertion by femoral vein: Safety and effectiveness for the pediatric patient.
Pediatrics 77:842–847.
12. Daily, R.H. (1985) Femoral vein cannulation: A review. J. Emerg. Med. 2:367–372.
13. Newman, B.M., Jewett, T.C., Karp, M.O., et al. (1986) Percutaneous central venous catheterization in children. First line choice for venous access. J. Ped. Surg.
21:685–688.
14. Williams, J.F., Seneff, M., Friedman, B.C., et al. (1991) Use of femoral venous catheters in critically ill adults. Prospective study. Crit. Care Med. 19:550–553.
15. Durbec, O. and Viviand, X. (1997) A prospective evaluation of the use of femoral venous catheters in critically ill adults. Crit. Care Med. 25:1986–1989.
16. Mallory, D.L., McGee, W.T., Shawker, T.H., et al. (1990) Ultrasound guidance improves the success rate of internal jugular vein cannulation: A prospective,
randomized trial. Chest 98: 157–160.
17. Mallory, D.L., Shawker, T.H., Evans, G., et al. (1990) Effects of clinical maneuvers on sonographically determined internal jugular vein size during venous
cannulation. Crit. Care Med. 18:1269–1273.
18. Brinkman, A.J. and Costley, D.O. (1973) Internal jugular puncture. JAMA 223:182.
19. Johnson, F.E. (1978) Internal jugular vein catheterization prospective study. N.Y. State J. Med. 78:2168.
20. Eisenhauer, E.D., Derveloy, R.J. and Hastings, P.R. (1982) Prospective evaluation of central venous pressure (CVP) catheters in a large citycountry hospital.
Ann. Surg. 196:560.
21. Sznajder, J.I., Zvebil, F.R., Bitterman, H., et al. (1986) Central vein catheterization failure and complication rates by three percutaneous approaches. Arch. Intern.
Med. 146:259.
22. Connors, A.F., Speroff, T., Dawson, N.V., et al. (1996) The effectiveness of right heart catheterization in the initial care of critically ill patients. JAMA 276:889–
918.
23. The Hospital Infection Control Practices Advisory Committee (1996) Guideline for prevention of intravascular devicerelated infections. Part I. Intravascular
devicerelated infections: An overview (Review). Am. J. Infection Control 24:262–277.
24. The Hospital Infection Control Practices Advisory Committee (1996) Guideline for prevention of intravascular devicerelated infections. Part II.
Recommendations for the prevention of nosocomial intravascular devicerelated infections (Review). Am. J. Infection Control 24:277–293.
25. Maki, D.G., Stolz, S.M., Wheeler, S., et al. (1997) Prevention of central venous catheterrelated blood stream infection by use of an antisepticimpregnated catheter:
A randomized, controlled trial. Ann. Intern. Med. 127:257–266.
26. Marin, M.G., Lee, J.C. and Skurnick, J.H. (2000) Prevention of nosocomial blood stream infections: Effectiveness of antimicrobialimpregnated and heparinbonded
central venous catheters. Crit. Care Med. 28: 3332–3338.
27. Darouiche, R.P., Raad, I.I., Heard, S.O., et al. (1999) A comparison of two antimicrobialimpregnated central venous catheters. N. Engl. J. Med. 340:1–8.
28. Food and Drug Administration (1989) Precautions necessary with central venous catheters. FDA Task Force. FDA Drug Bull. July: 15–16.
29. Arrow (1989) Arrow educational advisory and cautions included as a package insert with all CVC kits. Arrow, Reading, PA.
30. Tocino, I.M. and Watanabe, A. (1986) Impending catheter perforation of superior vena cava: Radiographic recognition. Am. J. Roentgenol. 146:487–490.
31. Ellis, L.M., Vogel, S.B. and Copeland, E.M. (1989) Central venous catheter vascular erosions. Ann. Surg. 209:475–478.
32. Sprung, C., Pozen, R., Rozanski, J., et al. (1982) Advanced ventricular arrhythmias during bedside pulmonary artery catheterization. Am. J. Med. 72:203.
33. Iberti, T., Ernest, B., Gruppi, L., et al. (1985) Ventricular arrhythmias during pulmonary artery catheterization in the intensive care unit. Am. J. Med. 78:451.
34. Patel, C., Laboy, V., Venus, B., et al. (1986) Acute complications of pulmonary artery catheter insertion in critically ill patients. Crit. Care Med. 14:195.
35. Feliciano, D.V., Mattox, K.L., Graham, J.M., et al. (1979) Major complications of percutaneous subclavian vein catheters. Am. J. Surg. 173:184.
36. McGee, W.T. and Mallory, D.L. (1988) Cannulation of the internal and external jugular veins.
Page 403
In: Problems in Critical Care: Vascular Cannulation, Vol. 2 (eds R.R.Kirby and R.W.Taylor) J.B.Lippincott, Philadelphia, PA, pp. 217–241.
37. Randolph, A.G., Cook, D.J., Gonzalez, C.A. and Pribble, C.G. (1996) Ultrasound guidance for placement of central venous catheters: A metaanalysis of the
literature. Crit. Care Med. 24:2053–2058.
38. Gilbert, T.B., Seneff, M.G. and Becker, R.B. (1995) Facilitation of internal jugular venous cannulation using an audioguided Doppler ultrasound vascular access
device: Results from a prospective, dualcenter, randomized, crossover clinical study. Crit. Care Med. 23:60–65.
39. Agency for Healthcare Research and Quality (2001) Making Health Care Safer—A Critical Analysis of Patient Safety Practices; Summary. Evidence
Report/Technology Assessment: Number 43. AHRQ Publications, Rockville, MD.
Page 404

Page 405
Index
2, 3diphosphoglycerate (2, 3DPG), 39
A
A wave, 64
Abdominal paradox, 290
Accelerated hypertension, 128
ACEinhibitors, 28
Acetazolamide, 59
Acetylcholine, 112
Acidbase abnormalities, 50
Acidbase nomogram, 50
ACTH (cortrosyn) stimulation test, 89
Activated protein C, 91
Acute aortic dissection, 136, 141
Acute exacerbation of COPD, treatment, 273
Acute lung injury (ALI), 222
Acute myocardial infarction and cardiogenic shock, 94
Acute pulmonary hypertension and shock, 96
Acute respiratory distress syndrome, 222–232
Acute spinal cord injury and DVT, 198
Adaptive pressure support (APS) ventilation, 215
Adenosine (Adenocard), 169
Adrenal insufficiency, 89
Adrenergic receptors, 113
Advanced Cardiac Life Support (ACLS), 174, 180
Afterload, 18, 20
Airway edema, 92
Airway pressure release ventilation, 211, 213
Albumin, 100, 107
Albuterol (salbutamol), 273
Alcoholic ketoacidosis, 53
Aldosterone deficiency, 58
Alkaline diuresis, 55
Allen test, 383
Alphaadrengergic agents, 139
Alphaadrengergic receptors, 113
Alteplase (tPA), 156
Alveolar gas equation, 11
Alveolar rupture during mechanical ventilation, 236
Amicar (epsilon aminocaproic acid), 339
Aminoglycoside antibiotics and weaning, 299
Aminophylline and weaning, 300
Amiodarone, 123, 176
Amyl nitrate, 133
Anerobic metabolism, 38, 42
Anaphylaxis, 86, 91, 93
Anatomy of the chest wall, 365
Angioedema, 92
Angiotensin converting enzyme (ACE) inhibitors, 140, 153
Angiotensin II, 82, 132
Anion gap, 51
Antibiotic impregnated catheters, 399
Antibiotic resistance, 265
Antibiotic treatment of COPD, 275
Antibiotics by class, 263
Anticoagulation before cardioversion, 172
Antiendotoxin therapies, 90
Antihistamines for anaphylaxis, 94
Antihypertensive agents, 132
Antiphospholipid syndrome, 185
Antithrombin III, 91
Aortic dissection, 32, 33, 96, 375
Aortic injury Xray findings, 376
Aortic regurgitation, 32
Aortic stenosis, 31
Approaches to the internal jugular vein, 392
Aprotinin, 339
ARDS incidence, 222
ARDS: ventilatory settings, 246
Arterial O2 content (CaO2 ), 39
Aspiration of subglottic secretions, 264
Assist/control mode ventilation, 207
Atenolol (Temormin), 153, 169
Atrial “kick”, 27
Atrial fibrillation, 170, 335
Atrioventricular nodal reentrant tachycardia, 167, 168
Atropine, 179
Automatic tube compensation (ATC), 215
Autonomic hyperreflexia, 142
AutoPEEP, 7, 13, 206, 272, 299
Page 406
B
Baroreceptor reflexes, 82
Barotrauma, 228, 236
Betaadrenegic receptors, 113
Betahydroxybutaric acid, 52
Betalactamases, 265
Bifascicular block, 161
Bilevel pressure ventilators, 319
Biomedicus centrifugal pump, 335
Bivalirudin, 153
Blood pressure adaptation/control mechanisms, 131
Blood transfusion risks, 106
Blunt cardiac injury, 377
Bohr equation, 15
Bradyarrhythmias, 179
Bromide intoxication, 51
Bronchial blocker, 348
Bronchodilators, 273
Bronchopleural fistula, 357
C
C wave, 64
Calcium channel blockers, 137, 152
Calcium chloride therapy, 333
Cannon A waves, 173
Carbicarb, 54
Carbon monoxide poisoning, 374
Carbonic anhydrase inhibitors, 57
Carboxyhemoglobin, 374
Cardiac index, 18, 77
Cardiac rupture, 377
Cardiac tamponade, 78, 96, 157, 336
Cardiogenic shock, 19, 94
Cardiopulmonary bypass (CPB), 330
Carotid sinus massage, 167
CatecholamineOmethyl transferase (COMT), 120
Catecholaminesecreting tumors, 141
Cephalosporin antibiotics (table), 263
Cellular oxygen utilization, 40
Central venous pressure monitoring, 64
Cephalosporinases, 265
Cerebral autoregulation, 130
Chemoreceptors, 130
Chest physiotherapy, 40
Chesttube drainage system, 349
Chloride responsive alkalosis, 59
Chylothorax, 359
Ciprofloxacin, 263
Citrate intoxication, 105
CKMB, 150
Clagett procedure, 358
Clavicular fractures, 368
Clonidine (Catapres), 135, 139, 140, 144
Clopidogrel (Plavix), 152
Clostridium difficile cytotoxic assay, 87
Coagulation cascade in shock, 91
CollisBelsey hiatal hernia repair, 356
Colloid osmotic pressure (COP), 100
Communityacquired pneumonia (CAP), 255
Complete heart block, 23
Compliance, 3, 6
Complications of central venous catheterization, 399
Complications of thoracic procedures, 356
Congestive heart failure, diagnosis, 26
Constrictive pericarditis, 78, 96
Continuous renal replacement therapy in sepsis, 89
Contractility, 21
Contraction alkalosis, 59
Controlmode ventilation, 207
Cordistype introducers, 398
Coronary revascularization, 158, 161
Corticosteroids, 94, 274, 299
Corticosteroids in ARDS, 229
Cortrosyn stimulation test, 89
Criteria for extubation, 350
Criteria for ICU discharge after open heart surgery, 340
Crop rotation, 265
Cushing reflex, 142
Cyanide poisoning, 40, 42
Cyanmethemoglobin, 133
Cytokines and mechanical ventilation, 240
D
Dalteparin (Fragmin), 155
DDAVP, 124
Ddimer assay, 187
Dead space, 13
Derived hemodynamic parameters, 77
Desmopressin, 339
Dextrans, 108
Diabetic ketoacidosis, 52
Diaphragmatic dysfunction, 293, 299, 333
Diastolic function, 22
Diazoxide, 134, 137
Dichloroacetate, 54
Digoxin (Lanoxin), 28, 31, 169
Diltiazem (Cardizem), 138, 169
Disseminated intravascular coagulation (DIC), 89
Distributive shock, 86
Diuretics, 19, 27, 141
Dlactic acidosis, 54
Dobutamine (Dobutrex), 118, 120, 160
Dopamine, 117, 119
Dopamine receptors, 113
Dopexamine, 121
Dorsalis pedis artery, 384
Down regulation, 121
Doxapram, 300
DVT prophylaxis, 197
Dynamic compliance, 227
E
Echocardiography, 27, 160
Eclampsia, 128, 136, 143
Page 407
Ejection fraction, 21
Elastin fibers, 256
Electrocardiogram in acute MI, 149
Electrolyte disorders and adrenergic response, 122
Empyema, 358, 372
Enalaprilat, 135, 140
Endobronchial intubation, 348
Endocarditis, 30, 32
Endotracheal aspirates (ETAs), 256
Enoximone, 123
Epidural analgesia, 353
Epinephrine, 93, 118, 120, 274
Eptifibatide (integrelin), 154, 155
Esmolol (Breviblock), 134, 138, 153, 169
Esophageal perforation, 360
Esophageal pressure, 4
Esophagogastrectomy, 356
Ethanol infusion, 55
Ethylene glycol intoxication, 51, 56
Extended spectrum betalactamases (ESBLs), 265
Extracorporeal membrane oxygenation (ECMO), 229
Extubation criteria following cardiac surgery, 332
F
Factor V Leiden mutation, 184
Fasttrack protocols, 330
Femoral vein anatomy, 390
Fenoldopam, 118, 119, 135, 137
Fiberoptic bronchoscopy, 257
Fibroproliferation in ARDS, 224
Fibrothorax, 372
Fick equation, 37
Firstdegree AV block, 161
Flail chest, 364, 367
Flail leaflets, 30
Fluid compartments, 101
Folic acid, 55
Fomepizole (4methylpyrizole), 55
Fosinopril, 141
FrankStarling curve, 18
Frequency to tidal volume ratio (f/Vt), 291
Fresh frozen plasma (FFP), 85
Functional residual capacity (FRC), 2
Fungal pneumonia, 262
Furosemide, 27
G
Ganglionic blockers, 141
Gastric interposition, 360
Gastric opening pressure, 323
Gastric tonometry, 45
Gentamicin, 263
Glycoprotein IIb/IIIa antagonists, 154
Goaldirected therapy with DO2 and VO2 , 44
Greenfield filters, 195
Guidewire exchanges of catheters, 400
Gunshot wounds, 364
GUSTO III trial, 156
H
Hampton’s hump, 188
Heartmate, 335
Helical (spiral) CT, 190
Heliox delivery system, 277
Heliumoxygen mixtures (heliox), 6, 274, 355
Heller myotomy, 356
Hemodynamic profile, 26
Hemopericardium, 157
Hemothorax, 371
HendersonHaselbalch equation, 51
Heparin, 191
Hetastarch 100, 107
Highfrequency jet ventilation (HFJV), 352
Hirudin, 153
Humidification of inhaled gas, 324
Hyaline membrane formation, 223
Hydralazine (Apresoline), 134, 137
Hydrocortisone, 89
Hydrostatic pressure, 100
Hydroxycobalamin, 133
Hyperbaric oxygen treatment indications, 375
Hyperkalemia, 52
Hypertensive crisis, 128
Hypertonic saline solution (HSS), 106
Hypertropic cardiomyopathy, 115
Hyponatremia, 52
Hypophosphatemia, 122, 297
Hypothyroidism and weaning, 297
Hypotonic solutions, 100, 106
Hypovolemia, 19
Hypovolemic shock, 83
Hypoxemia, 10, 12, 311
Hypoxemic respiratory failure and NIPPV, 314
Hypoxic vasoconstriction, 13
I
Ibuprofen in shock, 91
ImipenemCilastin (Primixin), 263
Immunocompromised patients and NIPPV, 316
Impedance plethysymography, 191
IMV weaning, 295
Increased intracranial pressure, 142
Independent lung ventilation, 352, 371
Indicator dilution curve, 76
Infraclavicular approaches to the subclavian vein, 397
Inhalation injuries, 374
Inhaled nitric oxide, 13
Inositol 1, 4, 5triphosphate (IP3), 114
Insect stings and anaphyaxis, 92
Inspiratory positive pressure waveform, 7
Inspiratory time fraction, 206, 280
Inspiratory work, 211
Page 408
Intercostal nerve blocks, 354

Intermittent positive pressure breathing (IPPB), 308
Internal cardiac defibrillators, 177
Internal jugular vein anatomy, 391
Interstitial edema, 41
Interstitial osmotic pressure, 103
Interstitial pulmonary edema, 223
Intraaortic balloon pump (IABP), 30, 95, 160, 335
Intraembolic thrombolysis, 196
Intrapleural catheters, 354, 355
Intrapulmonary shunt, 13
Intrathecal morphine, 354
Intraventricular septal rupture, 158
Invasive hemodynamic monitoring, 382–401
Ipratroplum bromide (Atrovent), 273
Iron lung, 308
Isoproterenol, 118, 121
Isotonic solutions, 100, 106
Isradipine, 135, 138
IVC filters, 195
J
Jejunoileal bypass, 54
K
Kerley Blines, 27
Kussmaul’s sign, 29, 157
L
Labetalol (Normodyne), 134, 139, 153
Lactated Ringer’s solution (LR), 106
Lactic acidosis, 37, 42, 53
Lactate dehydrogenase (LDH), 150
Laryngeal edema, 92, 351
Leftatrial pressure (LAP), 18
Leftventricular strokework index, 77
Leftventricular compliance, 74
Leftventricular enddiastolic blood volume (LVEDV), 18
Leionella infection, 258, 260
Levofloxacin (Levagvin), 263
Lidocaine, 176
Linezolid (Zyvox), 263
Lisinopril, 141
Lobar torsion, 359
Lobectomy, 356
Lovenox (Enoxaparin), 155
Low molecular weight heparin, 153, 193
Lower inflection point, 226, 242
Lowstretch ventilation, 228
Lung Injury Score, 222
Lung recruitment, 225, 246
Lungprotective ventilation strategies, 236–250
M
Magnesium sulfate, 177, 274
Magnetic resonance angiography, 191
Malignant hypertension, 128
Mandatory minute ventilation (MMV), 212
Maximal inspiratory pressure (MIP), 290
Mediastinitis, 339
Meropenem, 263
Metabolic acidosis, 50, 52, 59
Methanol, 51, 55
Methicillinresistant Staphylococcus aureus (MRSA), 259
Methoxamine, 139
Methyldopa, 135
Methylprednisone, 273, 274
Methylxanthines, 274
Metoprolol (Lopressor), 152, 153, 169
Milrinone, 118, 123
Minimal inhibitory concentration (MIC), 261
Minute ventilation, 290
Mitral regurgitation, 29, 76
Mitral stenosis, 30
Mixed venous 02 saturation, 39, 41, 43, 77
Monoamine oxidase (MAO), 120, 128
Monoclonal antiTNF (tumor necrosis factor), 90
Morphine sulfate for acute MI, 151
Multifocal atrial tachycardia, 168
Multiple myeloma, 51
Multiple organ dysfunction syndrome (MODS), 87
Multivariate indices for weaning prediction, 291
Murray Lung Injury Score, 222
Myasthenia gravis, 360
Myocardial contusion, 377
Myocardial free wall rupture, 157
Myocardial oxygen consumption, 22
N
Nadalol (Corgard), 153
Narcotic withdrawal, 144
Necrotizing pneumonia, 256
Negative pressure ventilators, 308
Neuromuscular paralysis in asthma, 279
Nicardipine (Cardene), 135, 137
Nifedipine (Adalat), 135, 138
NIH ARDS Network trial, 227
Nimodipine, 135
Nissen fundoplication, 356
Nitric oxide (NO), 230
Nitrogen washout atelectasis, 13
Nitroglycerin, 28, 134, 151
Nitroprusside (Nipride), 13, 133
Nontransmural MI, 148
Noninvasive lower extremity testing, 190
Noninvasive positive pressure ventilation, 276, 308–325
Norepinephrine (Levophed), 117, 118
Normal saline solution, 106
Noscomial pneumonia—common pathogens, 255
Nosocomial pneumonia, 254–266, 312
Nutritional status and weaning, 297
Page 409
O
Obstructive shock, 96
Onelung ventilation, 348
Openlung biopsy, 258
Osmolal gap, 51
Osmolarity, 101
Oxandrolone, 300
Oxidative phosphorylation, 40
Oxygen consumption (VO2 ), 37, 40, 46, 77
Oxygen content (CaO2 ), 36
Oxygen cost of breathing, 292
Oxygen delivery, 36, 45
Oxygen extraction ratio (O2 ER), 38, 77
Oxygen free radical production in ARDS, 224
Oxygen saturation, 39
Oxyhemaglobin dissociation curve, 40
P
P mitrale, 31
P waves, 166
Pacemakers, 22
Palla sign, 188
PaO2 to FiO2 ratio (P/F ratio), 222
Papillary muscle rupture, 94, 158, 377
Paradoxical motion of the chest wall, 367
Paraplegia, 377
Parapneumonic effusion, 258
Partial liquid ventilation (PLV), 230
Pectus repair, 356
Penetrating injuries to the chest, 364
Penicillin and anaphylaxis, 91
Penicillin resistance, 265
Pentoxyfylline, 229
Perfluorocarbons, 230
Perioperative myocardial infarction, 336
Permanent pacing, 161
Permissive hypercapnia, 228, 246
Phentolamine, 116, 134, 140
Phenylephrine (neosynephrine), 116, 118, 169
Pheochromocytoma, 128, 136
Phosphodiesterase inhibitors, 123
Phrenic nerve injury, 299
Pindolol (Visken), 153
PIOPED trial, 189
Piperacillintazobactam (Zosyn), 263
Plasma osmolality, 51
Plasminogen activator inhibitor I (PAII), 91
Pleural effusion, 359
Pneumomediastinum, 375
Pneumothorax, 96, 222, 359, 372, 401
Positive endexpiratory pressure (PEEP), 226, 238, 248, 281
Postantibiotic effect, 261
Postoperative renal dysfunction patterns, 338
Prednisone, 274
Preeclampsia, 143
Preload, 18, 19
Premature ventricular contractions (PVCs), 179
Pressure control inverse ratio ventilation (PCIRV), 209
Pressure support (PS), 207 295
Pressure time product, 211
Pressureregulated volume control (PRVC), 214
Pressurevolume curves, 19, 239, 242
Proarrhythmia, 177
Procainamide, 176
Proinflammatory cytokines, 91
Prone positioning, 10, 228, 248–9
Propranalol (Inderal), 153
Prostacyclin and prostagladin E1, 229
Protamine, 339
Protected brush specimens (PBSs), 257
Protein C, 91, 188
Protein S, 185
Pseudomonas aeruginosa, 260
Pulmonary angiography, 190
Pulmonary artery catheter, 66, 85 160
Pulmonary artery daistolic pressure, 69
Pulmonary artery occlusion pressure (PAOP), 18, 68
Pulmonary compliance, 227
Pulmonary contusions, 369
Pulmonary edema, 19, 24
Pulmonary emboli, 13, 78, 186, 195
Pulmonary embolism—diagnostic approach, 188
Pulmonary vascular resistance, 77
Pulseless electrical activity (PEA), 174
Pulsus alternans, 25
Pwave, 27
Pyridoxine, 56
Pyrivate utilization, 42
Q
Quinupristin+dalfopristin (Synercid), 263
R
Racemic epinephrine, 351
Radiocontrast agents, 92–3
Radiofrequency catheter ablation, 170
Rapid shallow breathing index, 291
Rate pressure product, 77
Recombinant Activated Protein C (rhAPC), 91
Recombinant tissue type plasminogen activator (rtPA), 195
Recurrent laryngeal nerves, 356
Reexpansion pulmonary edema, 359
Renal tubular acidosis (RTA), 56–57
Reninangiotensin system, 132
Respiratory acidosis, 50, 58
Respiratory alkalosis, 60
Respiratory cycle and PAWP, 71
Respiratory quotient, 11, 297
Resting oxygen consumption, 40
Retaplas (rtPA), 156
Page 410
Rib fractures, 367

Rightatrial pressure, 64
Rightventricular failure, 25, 29, 76, 157, 196
Rightventricular infarction, 29, 76, 157
Rightventricular pressure overload, 20
Risetime adjustment, 209
Risk assessment for cardiac surgical patients, 342
Risk factors for nosocomial pneumonia, 254
Risk factors for PE, 184
Routine changes of the central venous catheters, 399
S
S3 gallop, 25, 94
Salicylate intoxication, 55
Scapular fracture, 369
Secondary messenger, 114
Seldinger technique, 382
Selective digestive decontamination, 265
Semiquantitative cultures, 257
Septal shift, 20
Serum glutamic oxatoacetic acid transferase (SGOT), 150
Severe sepsis/SIRS, 87
Shivering, 40, 336
Shunt fraction, 13
Signal transduction, 114
Sinus bradycardia, 166, 168
Sinus tachycardia, 166, 168
Sinusitis, 264
Sleep deprivation, 298
Sodium nitroprusside, 134
Spinal cord injuries, 86, 142
Spironolactone, 28
ST segment elevation, 149
Staphylococcus aureus, 259
Starling curve, 105
Starling’s law, 18, 102
Starvation ketoacidosis, 53
Static compliance (Cs), 227
Status asthmaticus, treatment, 274
Sternal dehiscence, 339
Sternal fractures, 368
Sternocleidomastoid triangle, 392
Streptokinase, 156, 195
Stress testing, 159
Stroke index, 77
Stroke volume, 27, 77
Stroke work, 21
Subclavian vein anatomy, 395
Subconjunctival hemorrahge, 367
Subcutaneous emphysema, 228, 360, 366
Supraglottic edema, 355
Supranormal DO2 , 45
Supraventricular tachycardiac (SVT), 166
Surfactant, 4, 230
Sympathetic nervous system, 82, 112
Synchronized intermittent mandatory ventilation (SIMV), 206
Systemic inflammatory response syndrome (SIRS), 86, 87
Systemic vascular resistance, 77, 130
Systolic ejection murmur, 32
T
Tachycardia, 22
Tachyphylaxis, 121
Temporary pacemaker, 22, 160, 179
Tenectaplase (TNKtPA), 156
Tension pneumothorax, 373
Tensiontime index of the human diaphragm (Ttdi), 291, 292
THAM, 54
Thebesian veins, 13
Theophylline, 274
Thiamine, 42, 56
Thiocyanate, 133
Thirddegree AV block, 161, 166
Thirdspace, 83
Thoracentesis, 258, 372
Thoracic aortic aneursym, 356
Thoracoscopy, 372
Thrombolytic therapy, 155, 194
Thymectomy, 356
Thyroid storm, 86, 136, 143
Ticarcillin Clavulanate (Timentin), 263
Ticlopidine (Ticlid), 152
Tidal volume in COPD, 280
Timedependent antibiotics, 262
Timolol (Blocadren), 153
Tirofiban (aggrastat), 154, 155
Tissue factor protein inhibitor (TFPI), 91
Tobramycin, 263
Torsade de Pointes (TdP), 177, 178
Total body water (TBW), 101
Total lung capacity (TLC), 2
Tracheal occlusion pressure, 293
Tracheostomy, 300
Transmural pressure, 73
Transpulmonary pressure gradient, 237
Transthoracic needle aspiration, 258
Tricuspid regurgitation, 66, 78
Trimethaphan, 134, 141
Tromethamine, 54
Troponin, 150
Tube thoracostomy, 373
Tumor emboli, 358
U
Ultrasound image of IJV, 392
Upper inflection point, 242
Upregulation, 121
Urinary anion gap, 51
Urokinase, 195
Page 411
V
V wave, 29, 64, 76
V/Q lung scanning, 189
V/Q relationships, 8, 11
Vancomycin, 263
Vasopressin, 123
VaughnWilliams classifcation, 169
Ventilator tubing changes, 264
Ventilatorinduced lung injury (VILI), 224, 236
Ventilatory management of ARDS, 225
Ventricular assist device, 335
Ventricular compliance, 73
Ventricular dysrhythmias, 176
Ventricular interdependence, 19
Ventricular septal rupture, 76
Ventricular tachycardia, 174
Verapamil (Calan, Isoptin), 138, 169
Viscosity, 5, 40
Vital capacity (VC), 2, 290
VO2 , 37
Volume overload and weaning, 298
Volume support (VS) ventilation, 213
Volumeassured pressure support (VAPS), 216, 217
Volutrauma, 227, 236
W
Waveform during pressure ventilation, 205
Waveforms during PCIRV, 212
Waveforms during volume A/C ventilation, 203
Weaning from mechanical ventilation, 288–301
Weaning parameters, 290
Weaning pharmacologic support, 124
Wenckebach block, 166
Westermark sign, 188
Wide complex tachycardias, 172
Wolfe ParkinsonWhite Syndrome, 169
Work of breathing (WOB), 40, 292

Cardiopulmonary Critical Care Compress

Uploaded by

Copyright:

Available Formats

Cardiopulmonary Critical Care Compress

Uploaded by

Document Information

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

Cardiopulmonary Critical Care Compress

Uploaded by

Copyright:

Available Formats

Cover

This page intentionally left blank.

CARDIOPULMONARY CRITICAL CARE

This page intentionally left blank.

CARDIOPULMONARY CRITICAL CARE

© BIOS Scientific Publishers Limited, 2002

ISBN 0­203­42767­X Master e­book ISBN

ISBN 0­203­44667­4 (OEB Format)

BIOS Scientific Publishers Ltd

Production Editor: Aimie Haylings

1. Respiratory physiology/pulmonary gas exchange 1

Supraventricular tachycardias 166

16. Nosocomial pneumonia 253

Complications or problems with NIPPV 317

Central venous access 384

ECG electrocardiogram/ electrocardigraphy

C.Allen Bashour MD FACS FCCP

Howard L.Corwin MD FCCM FACP

Thomas L.Higgins MD MBA FACP FCCM FACC

Robert M.Kacmarek RRT PhD

William T.McGee MD MHA

Lawrence S.Rosenthal MD PhD FACC

Marc Schweiger MD FACC

Jay S.Steingrub MD FACP FCCM

James K.Stoller MD FACP FCCM FCCP

1.2 Lung mechanics

Table 1.1 Volume subdivisions and capacities of the lung

Residual volume RV Gas remaining in the lung after maximum exhalation

1.2.1 Pulmonary mechanics

1.2.2 Elastic properties of the respiratory system

1.2.4 Surfactant and effect of surface tension

1.2.5 Flow resistance

where B is a constant, and by the Reynold’s Number (RN)5:

1.2.6 Normal and abnormal values

1.2.7 Measurements during mechanical ventilation

where PEEPA is applied PEEP and PEEPI is auto­PEEP.

1.3 The law of motion

Total respiratory systema (RS) 0.08–0.1 l cm–1 H2O

Lunga 0.16–0.2 l cm–1 H2O

Thoraxa 0.16–0.2 l cm–1 H20

1.5.3 Shunting and dead space

Table 1.3 Cause of altered relationships

1.6 Evaluation of hypoxemia

The last part of Equation (1.13) is frequently simplified to

1.6.2 True shunting

1.6.3. Dead space

1.6.4 Estimation of shunt and dead space

Table 1.4 Estimation of intrapulmonary shunting and dead space

Index Normal value Abnormal value Limitation

PaO2/PAO2 >0.75 <0.75 –

PaO2/FIO2 450–500 <300–350 Varies with PaCO2 and FIO2

VD/VT 25–40% >40% –

determinant of pulmonary blood flow distribution. J. Appl. Phys. 71:620–629.

2.1 Cardiac function

2.1 Cardiac function

2.2.1 Frank­Starling curve

2.2.2 Pressure­volume relationships

2.2.3 Alterations of preload

2.2.4 Ventricular interdependence and LV preload

Table 2.1 Ventricular interdependence

ISBN 020342767X Master ebook ISBN

ISBN 0203446674 (OEB Format)

where PEEPA is applied PEEP and PEEPI is autoPEEP.

2.2.1 FrankStarling curve

2.2.2 Pressurevolume relationships

(av)O2 content (C(av)O2) (ml dl–1) 5 10 5 5

3.10 Goaldirected therapy with DO2 and VO2