Modern Monitoring in Anesthesiology and Perioperative Care 2020

Freemedicalbooks4download
Modern Monitoring
in Anesthesiology
and Perioperative Care
Downloaded from https://www.cambridge.org/core. University of New England, on 05 May 2020 at 15:57:09, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://www.cambridge.org/core/product/0DB17B714238E5060D91E51A9E694F10
Modern Monitoring
in Anesthesiology
and Perioperative Care
Edited by
Andrew B. Leibowitz, MD
Icahn School of Medicine at Mount Sinai
Suzan Uysal, PhD

University Printing House, Cambridge CB2 8BS, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India
79 Anson Road, #06–04/06, Singapore 079906
Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781108444910
DOI: 10.1017/9781108610650
© Cambridge University Press 2020
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2020
Printed in the United Kingdom by TJ International Ltd, Padstow Cornwall
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloging-in-Publication Data
Names: Leibowitz, Andrew B., editor.
Title: Modern monitoring in anesthesiology and perioperative care / edited by Andrew B. Leibowitz.
Description: Cambridge ; New York, NY : Cambridge University Press, 2020. | Includes bibliographical
references and index.
Identifiers: LCCN 2019042016 | ISBN 9781108444910 (paperback)
Subjects: LCSH: Anesthesia. | Intraoperative monitoring. | Point-of-care testing. | Anesthesiology – Apparatus
and instruments. | Surgical technology.
Classification: LCC RD82 .M62 2020 | DDC 617.9/6–dc23
LC record available at https://lccn.loc.gov/2019042016
ISBN 978-1-108-44491-0 Paperback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
................................................................................................................................
Every effort has been made in preparing this book to provide accurate and up-to-date
information that is in accord with accepted standards and practice at the time of
publication. Although case histories are drawn from actual cases, every effort has been
made to disguise the identities of the individuals involved. Nevertheless, the authors,
editors, and publishers can make no warranties that the information contained herein is
totally free from error, not least because clinical standards are constantly changing
through research and regulation. The authors, editors, and publishers therefore
disclaim all liability for direct or consequential damages resulting from the use of
material contained in this book. Readers are strongly advised to pay careful attention to
information provided by the manufacturer of any drugs or equipment that they plan to use.
Contents
List of Contributors vi
Preface ix
1 Statistics Used to Assess Monitors and 11 Assessing Intravascular Volume Status and
Monitoring Applications 1 Fluid Responsiveness: A Non-Ultrasound
Lester A. H. Critchley Approach 109
David S. Beebe
2 Multimodal Neurological Monitoring 10
Samson Zarbiv, Erica Fagleman, and Neha 12 Assessment of Extravascular Lung Water 117
S. Dangayach Torsten Loop
3 Cerebral Oximetry 20 13 Point-of-Care Hematology 123
Benjamin Salter and Elvera Baron Jacob Raphael, Liza Enriquez, Lindsay
Regali, and Linda Shore-Lesserson
4 The Oxygen Reserve Index 30
Andrew B. Leibowitz 14 Assessment of Intraoperative Blood Loss 139
Kyle James Riley and Daniel Katz
5 Point-of-Care Transesophageal
Echocardiography 35 15 Respiratory Monitoring in Low-Intensity
Ronald A. Kahn Settings 148
Andrew B. Leibowitz and Adel Bassily-
6 Point-of-Care Transthoracic
Marcus
Echocardiography 55
Julia Sobol and Oliver Panzer 16 The Electronic Health Record as a Monitor for
Performance Improvement 154
7 Point-of-Care Lung Ultrasound 67
David B. Wax
Zachary Kuschner and John M. Oropello
17 Future Monitoring Technologies: Wireless,
8 Point-of-Care Ultrasound: Determination of Wearable, and Nano 164
Fluid Responsiveness 80
Ira S. Hofer and Myro Figura
Subhash Krishnamoorthy and Oliver
Panzer 18 Downside and Risks of Digital
Distractions 173
9 Point-of-Care Abdominal Ultrasound 92
Peter J. Papadakos and Albert Yu
Shaun L. Thompson and Daniel
W. Johnson
10 Noninvasive Measurement of Cardiac
Output 100 Index 180
Samuel Gilliland, Robert H. Thiele, and
Karsten Bartels The colour plate section can be found between
pp. 110 and 111.
v
https://www.cambridge.org/core/terms. https://www.cambridge.org/core/product/954F8F3883F17962369350760AE1E18C
Contributors
David Geffen School of Medicine
Elvera Baron, MD, PhD
University of California
Assistant Professor
Los Angeles, CA
Department of Anesthesiology, Perioperative
and Pain Medicine Samuel Gilliland, MD
Department of Medical Education Assistant Professor
The Icahn School of Medicine at Mount Sinai Department of Anesthesiology
New York, NY University of Colorado
Aurora, CO
Karsten Bartels, MD, PhD
Associate Professor Ira S. Hofer, MD
Department of Anesthesiology Assistant Professor
University of Colorado Director, Division of Bioinformatics and Analytics
Aurora, CO Department of Anesthesiology and Perioperative
Medicine
Adel Bassily-Marcus, MD
David Geffen School of Medicine
Associate Professor of Surgery
University of California, Los Angeles
New York, NY Daniel W. Johnson, MD, FCCM
Division Chief & Fellowship Director,
David S. Beebe, MD
Critical Care
Professor
Medical Director, Cardiovascular ICU
Department of Anesthesiology
Associate Medical Director, Nebraska
University of Minnesota Medical School
Biocontainment Unit
Minneapolis, MN
Associate Professor
Lester A. H. Critchley, MD Department of Anesthesiology
Retired Professor and Consultant Anesthetist University of Nebraska Medical Center
The Chinese University of Hong Kong Omaha, NE
Neha S. Dangayach, MD, MSCR Ronald A. Kahn, MD

Assistant Professor of Neurology and Neurosurgery Professor of Anesthesiology, Perioperative & Pain
The Icahn School of Medicine at Mount Sinai Medicine, and Surgery
New York, NY Icahn School of Medicine at Mount Sinai
New York, NY
Liza Enriquez, MD Professor of Anesthesiology, Pain,
Montefiore Medical Center and Critical Care
Bronx, NY The Tel Aviv Medical Center
Tel Aviv
Erica Fagleman, MD
Department of Anesthesiology, Perioperative and Daniel Katz, MD
Pain Medicine Vice Chair of Education
Icahn School of Medicine at Mount Sinai Associate Professor of Anesthesiology, Perioperative
New York, NY and Pain Medicine
Myro Figura, MD New York, NY
Clinical Instructor
Department of Anesthesiology and Perioperative Subhash Krishnamoorthy, MD
Medicine Department of Anesthesiology
Downloaded from https://www.cambridge.org/core. University of Exeter, on 03 May 2020 at 05:53:06, subject to the Cambridge Core terms of use, available at
vi
https://www.cambridge.org/core/terms. https://www.cambridge.org/core/product/C32ABF3E1EA128479DEC54BDBA47754A
List of Contributors
College of Physicians & Surgeons of Columbia Department of Anesthesiology and Perioperative

University Medicine
New York, NY University of Virginia Health System
Charlottesville, VA
Zachary Kuschner, MD
Department of Emergency Medicine Lindsay Regali, MD
Long Island Jewish Medical Center, Northwell Health Anesthesiology Consultants of Virginia, Inc.
New Hyde Park, NY Roanoke, VA
Department of Internal Medicine, Division of Critical
Care Kyle James Riley, MD
Mather Hospital, Northwell Health Department of Anesthesia, Critical Care and Pain
Port Jefferson, NY Medicine
Massachusetts General Hospital
Andrew B. Leibowitz, MD Boston, MA
Chair of Anesthesiology, Perioperative & Pain
Medicine Benjamin Salter, DO
Mount Sinai Health System Assistant Professor of Anesthesiology
Professor of Anesthesiology, Perioperative & Pain Department of Anesthesiology, Perioperative and
Medicine, and Surgery Pain Medicine
Icahn School of Medicine at Mount Sinai The Icahn School of Medicine at Mount Sinai
New York, NY New York, NY
Torsten Loop, MD Linda Shore-Lesserson, MD, FAHA, FASE

Professor and Vice Chair Past-President, Society of Cardiovascular
Department of Anesthesiology and Critical Care Anesthesiologists
University of Freiburg Medical Center Professor of Anesthesiology
Freiburg, Germany Zucker School of Medicine at Hofstra Northwell
Vice Chair for Academic Affairs
John M. Oropello, MD, FACP, FCCP, FCCM Director, Cardiovascular Anesthesiology
Professor of Surgery and Medicine Long Island, NY
Program Director, Critical Care Medicine
Icahn School of Medicine at Mount Sinai Julia Sobol, MD, MPH
Director, Transplant Intensive Care Unit Assistant Professor of Anesthesiology
Mount Sinai Health System College of Physicians & Surgeons of Columbia
New York, NY University
New York, NY
Oliver Panzer, MD
Assistant Professor of Anesthesiology Robert H. Thiele, MD
College of Physicians & Surgeons of Columbia Associate Professor
University Department of Anesthesiology
New York, NY University of Virginia
Charlottesville, VA
Peter J. Papadakos, MD, FCCM, FAARC
Director of Critical Care Medicine Shaun L. Thompson, MD
Tenured Professor of Anesthesiology and Assistant Professor
Perioperative Medicine, Surgery, Neurology and Associate Program Director of the Critical Care
Neurosurgery Medicine Fellowship
University of Rochester Medical Director of ECLS Services at UNMC
Rochester, NY Department of Anesthesiology, Division of Critical
Care
Jacob Raphael, MD University of Nebraska Medical Center
Carl Lynch III Professor of Anesthesiology Omaha, NE
Downloaded from https://www.cambridge.org/core. University of Exeter, on 03 May 2020 at 05:53:06, subject to the Cambridge Core terms of use, available at vii
List of Contributors
David B. Wax, MD University of Rochester

Professor of Anesthesiology, Perioperative & Pain Rochester, NY
Medicine
Icahn School of Medicine at Mount Sinai Samson Zarbiv, MD, MPH
New York, NY Department of Critical Care Medicine
Cooper University Health Care
Albert Yu MD Cape Regional Health System
Assistant Professor of Anesthesiology and Cape May Court House, NJ
Perioperative Medicine
viii
Preface
Our current practice environment is daunting. provide a framework for evaluation of all monitors,
Providing safe quality care using technology and and reveal the risk that too much information poses to
receiving feedback that didn’t even exist 15 years ago the care that we provide. Other chapters focus on the
should be easy, but it isn’t. The plethora of information “why” monitoring a certain variable may be advanta-
and the number and variety of monitors available for geous, the basics of “how” the monitor works, and
use has increased in an almost exponential fashion. On “what” is the evidence of the impact on patient outcome.
October 27, 2019, a PubMed search of “noninvasive A basic theme of the chapters is that just because
monitoring of cardiac output” yielded 809 references, a we can do something does not mean we should, and
Google search for “anesthesiology guidelines” 25,700 sometimes less may be more, but a working knowl-
references, and there are at least 9 different technolo- edge of the whys, hows, and whats of modern mon-
gies for monitoring of cardiac output incorporated into itoring in anesthesiology and perioperative care will
commercially available products. allow every provider to optimize their patient’s safety
This book aims to be different than other monitor- and the quality of care they provide.
ing books and focuses on the “practical.” Chapters on
statistics and electronic distraction are unique and Andrew B. Leibowitz, MD
Downloaded from https://www.cambridge.org/core. University of Exeter, on 03 May 2020 at 05:57:17, subject to the Cambridge Core terms of use, available at ix
https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108610650.021
Chapter
Statistics Used to Assess Monitors
1 and Monitoring Applications

Lester A. H. Critchley
Introduction Instead, at-a-distance (e.g., transthoracic Doppler)

or surrogate (e.g., bioimpedance) methods are uti-
An evidence-based approach now prevails when
lized, which result in lack of precision.4,5 Compared
recommending medical treatments. This applies as
to measuring other more accessible hemodynamic
much to the latest therapies as to appropriate methods
variables such as blood pressure or heart rate, lack of
to monitor patients and their response to treatment.
accuracy and precision has hampered the develop-
For an evidence-based approach to be successful,
ment of routine CO monitoring in the clinical
however, it must be based on good-quality clinical
setting.5
data from well-conducted research. The quality of
Cardiac output can be measured accurately using
clinical studies and their data is now graded according
techniques such as the Fick method and radionuclide
to the level of evidence they provide,1 and guidelines
imaging studies. These methods, however, have sev-
exist on how to properly conduct clinical research.
eral limitations. They are only applicable in settings
Cochrane reviews have set standards for best evi-
such as the physiology laboratory or radiology depart-
dence. Working groups such as the National
ment; they are inapplicable at the point of care and
Institute for Clinical Excellence (NICE) and
therefore cannot be used in operating room, emer-
Resuscitation Council (UK) demonstrate how such
gency medicine, or critical care settings. Furthermore,
an approach can be transformed into up-to-date
Fick and radionuclide studies only provide single
guidelines and courses. When assessing the value of
readings, and there is need for technologies that mea-
emerging clinical monitoring technologies for peri-
sure CO on a frequent or continuous basis. The clin-
operative, emergency room, and critical care use,
ical significance of being able to assess changes or
researchers should be aware that clinical validation
trends in CO is only now being recognized, and this
studies must be of a sufficient standard to be of use in
is highlighted by the designs of recently marketed CO
evidence-based reviews. This perspective drives the
devices and the statistical approaches to their
approach of this chapter, with a focus on cardiac out-
validation.
put (CO) monitoring, since most of the literature on
All validation studies require a reliable reference
these statistical methods has arisen from analysis of
method against which comparisons are made. For CO
this variable.
monitoring, the accepted reference method has been
and remains single bolus thermodilution using a pul-
Cardiac Output Measurement monary artery Swan–Ganz catheter. The pulmonary
Cardiac output is the sum of stroke volumes expelled artery catheter, however, is now seldom used in clin-
from the heart over one minute; it can be measured ical practice, and its use is associated with significant
from either the pulmonary or the systemic circula- risk to patients.6,7 Clinical validation studies incor-
tions. As the arterial system leaving the heart porating pulmonary artery catheter measurements
branches, it is not possible to measure total CO at a are mostly restricted to cardiac surgery and liver
distal point such as the arm or descending aorta, and transplant. Some recent research studies have used
corrections are needed (e.g., arterial pulse contour the less invasive transpulmonary thermodilution
analysis and esophageal Doppler).2,3 Measurement method, which is employed in the PiCCO (Pulsion,
of CO at its source, the heart, is also difficult to achieve Munich, Germany) and VolumeView (Edwards
in the clinical setting because of restricted access, Lifesciences, Irvine, CA, USA) systems. Errors arise
unless one is performing open-heart surgery. in thermodilution measurement because of injectate
Downloaded from https://www.cambridge.org/core. University of Exeter, on 03 May 2020 at 06:01:18, subject to the Cambridge Core terms of use, available at 1
Statistics to Assess Monitoring Applications
and dead space issues,8 and the degree of inaccuracy The whole approach to validation statistics chan-
varies between clinical settings and different manu- ged in the 1980s when J. M. Bland and D. G. Altman
factured devices.9 The precision of the thermodilution introduced a new method of comparing measure-
method is generally accepted to be ± 20%,10,11 and this ments based on bias, the difference between pairs of
margin of error has played a significant role in the comparative readings.14 Bias was plotted against the
ongoing development of validation statistics. average of each pair, and the standard deviation of the
Cardiac output is not a static variable; its value bias provided a statistic called limits of agreement (i.e.,
constantly changes. Achieving a steady state in which 95% confidence intervals for the bias). Bland and
simultaneous comparative readings can be taken Altman, however, never provided guidance as to
often proves difficult, and this hampers the collection how the limits of agreement should be used to confirm
of good-quality validation data. clinical utility, leaving this to the discretion of the
user. This was particularly unsatisfactory when
Protocol Design and Data Collection Bland–Altman analysis was applied to CO studies
where the reference method, usually thermodilution,
The need for ethical approval and patient consent
was imprecise. Limits of agreement of less than
is an obvious prerequisite for publication. Poorly
1 liter/min were considered to be acceptable,15,16 but
planned data collection and inadequate sample size
no provision for (i) variations in baseline CO or
will limit the usefulness of collected data and thus
(ii) imprecision of the reference method was made.
the ability to publish the study findings. Common
To enable outcomes from Bland–Altman style CO
mistakes are (i) failure to blind investigators to
studies to be compared in 1999, Critchley proposed
comparative readings, (ii) failure to achieve simul-
the use of percentage error (PE), a statistic calculated
taneous readings during steady-state hemody-
from the limits of agreement (i.e., 95% confidence
namics, (iii) failure to have sufficient range of
interval of the bias) divided by the baseline CO for
readings, (iv) failure to collect sufficient data
the study.10 A benchmark for acceptance of a new
resulting in inconclusive results, (v) inconsistent
technique of less than 28.4% was set, which was
number and timing of repeated measurements
rounded up to less than 30%. This benchmark was
from individuals (i.e., irregular data collection),
based on a reference method’s precision of 20% and
and (vi) failure to collect serial data pairs that
acceptance of the test method also being set at 20%.
show adequate changes and hence fail to facilitate
Although PE has been criticized over the years for
trend analysis. A well-designed study has clearly
being too strict,11,17,18 its simplicity and robustness
defined times of data collection, which are of suffi-
as an analytical tool have withstood the test of time.
cient number to allow comprehensive analysis.12
In more recent decades, following advances in
Sample size is difficult to calculate in this type of
clinical medicine and monitoring technology, it has
research, even if a pilot study is performed, because of
become increasingly important to have bedside moni-
the range of different variables and outcomes
tors that accurately follow the vital signs of hospita-
involved. A more pragmatic approach may be based
lized patients. Unfortunately, Bland–Altman analysis
on reviewing the sample sizes used in previous studies
does not assess the ability of devices to detect changes;
that were successful in detecting effects. Comparative
it is limited to assessing accuracy of readings and
studies with cohorts of over 30 patients and 6 or more
agreement between methods.19,20 Thus, new statistical
serial data pairs are recommended.12
approaches were developed, referred to as trend ana-
lysis.21 Many researchers new to clinical monitoring,
Background to Validation however, fail to recognize the need to show trending
Thirty years ago scatter plots and regression and cor- and restrict data collection to that suitable for Bland–
relation analyses were the principal analytical meth- Altman analysis.
ods used to show how reliably a new measurement How to effectively address the issue of trending
method compared to a reference standard.13 capability has not been fully resolved in the literature.
Regression and correlation, however, only evaluate In a recent review of CO studies, Critchley and col-
the degree of association between two measurement leagues reported that only 20% of the studies per-
methods; they do not quantify accuracy. Quoting formed some form of trend analysis; the analytical
correlation coefficients and p values confirms little. methods employed were (i) Bland–Altman analysis
2
of tables and histograms, (ii) regression analysis of Each plot should display horizontal lines indicating
scatter plots, and (iii) analysis of direction of mean bias and the 95% confidence intervals or limits
change.21 of agreement. Inspection of the plot allows one to
When analyzing CO data from hospital patients, assess (i) the distribution or spread of data, (ii) the
commonly used trend analysis methods are (i) con- degree of agreement between methods (i.e., size of the
cordance on a four-quadrant plot and (ii) polar plot limits of agreement), and (iii) any systematic changes
analysis.22,23 Both these analyses rely on comparing in bias as CO increases (i.e., offsets in calibration).
serial data from reference and test methods, calculat- One common problem with presenting Bland–
ing the serial change in consecutive readings (ΔCO), Altman plots is using inappropriate scales, especially
and excluding data where the change is small (i.e., when more than one plot is shown. Rather than
< 10–15% change). The polar method involves trans- choosing scales that fill the page with data points,
forming the data from a simple (x, y) Cartesian format the axis of each plot should have similar scales and
to a radial format (radius, angle). Polar plots provide ranges. Otherwise visual comparisons between plots
greater information about the agreement between two are difficult to perform. Very often the Bland–Altman
methods that is lost when just direction of change is plot is accompanied by an (x, y) scatter plot that shows
used. Criteria for acceptable trending have been pro- the raw data (Figure 1.1), but regression lines and
posed for CO monitoring.21,23 A more detailed correlation coefficients are often omitted.
description of these methods follows. Bland–Altman analysis requires each data pair to
be independent of all other pairs and ideally from
Bland–Altman Analysis separate subjects.14 If data pairs are related (i.e., they
come from the same subject), the size of 95% confi-
Practically all CO validation studies published today
dence intervals and limits of agreement for the analy-
use Bland–Altman analysis and provide a Bland–
sis will be reduced. Use of repeated measures (i.e., data
Altman plot (Figure 1.1). The plot shows bias col-
pairs from the same subject) is common in CO stu-
lected from the whole or subgroups of the study.
dies; thus, the data analysis should correct for
Scatter plot Bland–Altman plot

12 5
4
Bias [COcardioq-COuscom] (L/min)
10
Cardiac output (uscom) (L/min)
2
8
1
6 0
–1
4
–2
–3
2
Regression line: Bland–Altman analysis:
COuscom = Cocardioq x 0.6 + 2.7 –4 meanCO = 6.20 L/min, Bias = –0.23 L/min
n = 285 data pairs, R 2 = 0.53, r = 0.73 LOA – 2.54 to 2.08 L/min, PE = 37%
0 –5
0 2 4 6 8 10 12 0 2 4 6 8 10 12
Cardiac output (cardioq) (L/min) Cardiac output (L/min)
Figure 1.1 Scatter plot with regression line and accompanying Bland–Altman plot. Statistical analysis data are added to each plot. The
Bland–Altman plot also displays the mean bias and limits of agreement of the analysis (dashed horizontal lines). Data are from a study that
compared two Doppler CO measurement methods, transthoracic (USCOM) and esophageal (CardioQ).
Source: Huang L, Critchley LA. An assessment of two Doppler-based monitors to track cardiac output changes in anaesthetized patients
undergoing major surgery. Anaesth Intens Care 2014;42:631–9. LOA: limits of agreement, PE: percentage error.
repeated measures by either (i) Bland and Altman or consecutive readings. Direction of change in CO can
(ii) Myles and Cui methods, which differ slightly in either be increased (i.e., positive direction change) or
complexity.24,25 Statistical software programs that decreased (i.e., negative direction change); the magni-
perform Bland–Altman analysis should also adjust tude of change is not included in the analysis. In the
for repeated measures; journal editors and reviewers trial a test method is compared to a reference method,
expect that authors will employ such corrections and which provides pairs of directions of change of read-
describe them in their manuscripts. ings that can either agree (i.e., concord) or disagree.
Percentage error is a key outcome statistic arising Concordance is measured as the proportion of read-
from CO studies that perform a Bland–Altman analy- ings that agree.
sis.10 It is used to compare findings of CO studies with To make concordance analysis easier to visualize,
findings of other published studies. It also allows criteria a four-quadrant plot is drawn of ΔCO reference
to be set for acceptance of a new CO monitor prior to against ΔCO test (Figure 1.2). Data where directions
starting a study. Most authors will use the less than 30% of change agree fall into the right upper and left lower
benchmark, from Critchley’s 1999 paper that based the quadrants. The ratio of the number of data pairs
criteria on a 20% precision for thermodilution where directions of change agree over the total num-
CO measurement and the need for less than 20% ber of data pairs for the study provides the concor-
measurement error (i.e., 95% confidence intervals or dance presented as a percentage.
precision).11 A 20% error represented up to a 1 liter/ Data pairs where the serial change in CO is small,
min variation in CO if the mean CO was 5 liter/min. however, can often have directions of change that
Cecconi and colleagues have questioned the logic disagree due to random errors in measurement; this
of assuming a 20% error in the reference method.26 is referred to as statistical noise. To eliminate statisti-
They recommended measuring its precision and cal noise from the concordance analysis an exclusion
using the error to set new acceptance criteria a priori. zone is used that removes data where the change in
Their rational was that (i) the error in thermodilution CO is less than 10–15% of the mean CO for the study.
or other reference method is very variable and 20% is The setting of limits for the exclusion zones is based
just an approximation, and (ii) any significant varia- on a receiver operator characteristic (ROC) curve
tion from 20% would result in lesser or greater errors analysis.22
in the test method to be accepted, if the acceptance Current advice for acceptable trending ability in
criteria are set at the standard 30%. Their approach to CO studies is greater than 92%.21 Ideally, confidence
measuring the reference method’s precision was to limits should be calculated for the concordance,
perform serial steady-state measurements from which is based on sample size. The ΔCO data is
which the coefficient of variation was calculated and treated as a binomial (i.e., direction of change either
precision derived.26 agrees or disagrees), and the standard deviation of the
concordance ratio (p) is √[np(1−p)], where n is the
Trend Analysis number of data points. A good example of how this
Trending capability, the ability to follow changes in statistic is generated and used is found in Axiak-
CO, can be assessed either by (i) multiple paired Flammer and colleagues.27
comparisons in a small number of subjects (i.e.,
n = 6–10 laboratory animals) or (ii) as part of a larger Polar Plots
scale clinical trial with up to 8–10 comparative mea- The introduction of polar plots (Figure 1.2) was to
surements in 20 or more patients. Statistical address the problems that (i) the four-quadrant plot
approaches are different for the two settings. Small method did not include magnitude of change and (ii)
cohort studies are dealt with later in the section Time all data pairs were treated equally despite size.14,21,23
Plots and Regression Analysis. By converting the data to (i) a radial distance that
represented the size of the combined changes in CO
Concordance Analysis from the two paired readings (i.e., average absolute
For larger cohort clinical trials the current approach is change in ΔCO) and (ii) an angle that represented the
concordance analysis using direction of change.21,22 degree of agreement (i.e., the greater the degree of
This analysis is based on serial data, and ΔCO is the disagreement the larger the angle), more information
study variable calculated from the difference between about the comparison between the two measurement
4
Four-quadrant plot 90 Polar plot – half moon
ΔCl (litre.min–1 min–2) [i.e., mean of ΔCluscom & ΔCOcardioq]

2.0 75
60 Polar plot set up:
2.0 Excl. zone 0.3 litre.min–1.m–2
1.5 45 n = 159 of 256 data pairs
1.5
1.0 30
ΔCluscom (litre.min–1 min–2)
1.0
0.5 15
0.5
0.0 0.0 0
–0.5 0.5 –15
–1.0
Concordance: 1.0 –30
Excl. zone 0.5 litre.min–1.m–2
–1.5
n = 129 of 256 data pairs 1.5 –45
rate = 97%
–2.0 Polar analysis:
–2.0 –1.5 –1.0 –0.5 0.0 0.5 1.0 1.5 2.0 –60 Angular bias = –1.7 degrees
2.0
–75 Radial limits (95% c.i.) = –25.8 to +22.4
ΔClcardioq (litre.min–1 min–2) –90
Polar concordance = 98%
[Boundary limits – 30 to +30 degrees]
Figure 1.2 Four-quadrant and polar plots showing changes (ΔCO). Four-quadrant plot has zero axes crossing at its center, creating four
zones. A central exclusion zone (square) is shown: Data lying within this zone are excluded because they contain a high level of random
variation compared to changes in CO (i.e., statistical noise). The line of identity y = x (dashed line) also is shown. Ideally, all data points should lie
along this line. Data that lie within the upper right and lower left quadrants agree (i.e., direction of change agrees). Results of concordance
analysis are printed in the plot. Polar plot is of a semicircle, or half-moon, design in which both positive and negative changes in CO are shown
together. A central exclusion zone also is shown (half circle). Zero- and 30-degree axes are highlighted (solid lines). Mean polar angle and 95%
radial limits of agreement for the polar analysis also are shown (dotted lines). Polar concordance rate is based on the proportion of data points
that lie within 30 degrees of the polar axis (zero degrees). Results of the polar analysis are shown.
Source: Huang L, Critchley LA. An assessment of two Doppler-based monitors to track cardiac output changes in anaesthetized patients
undergoing major surgery. Anaesth Intens Care 2014;42:631–9.
methods was retained. The concept of excluding data negative change data through 180 degrees to become
pairs where changes in CO were small and statistical a positive change, thus producing a half-moon rather
noise may corrupt the analysis was also applied. than full-moon plot. Generating polar data from
However, the exclusion zone was reduced from 15% Cartesian (x, y) ΔCO data and drawing polar plots
to 10% of the mean CO for the study because the can be technically challenging. Some of the newer
combined change in ΔCO on the polar diagram (i.e., statistical programs now provide polar plot drawing
radial length) was derived from the average of the two and analysis software. Guidance can also be found in
ΔCO values, whereas in the four-quadrant plot the the original paper describing polar plots.23 The polar
combined change was derived from the hypotenuse of method is probably best reserved for research groups
a triangle produced by test and reference values and performing high-quality validation studies. Mastering
was √2 (or 1.42) times larger in size. The mean angle the technique of polar plots provides a greater appre-
for all the data pairs provided a measure of misalign- ciation of the data and trending ability.
ment in calibration or offset between methods.
Empirically, a limit of ± 5% was set as the criterion Time Plots and Regression Analysis
for an acceptable offset. The radial limits of agreement Understanding the structure of one’s data is the key to
were set at ± 30% and were based on a 2:1 ratio in size knowing which statistical methods are most appro-
between ΔCO readings. These limits, however, were priate. Data arising from validation studies can be
not based on sound statistical theory. To make the considered as a two-dimensional matrix of paired
polar plot more visually friendly, one can rotate readings representing subjects in one plane and serial
measurements from individual subjects in the other when trending capability of the test bioimpedance
plane. Bland–Altman analysis is most appropriate method was considered acceptable. However, when
when there are many subjects and few, if any, serial regression analysis was applied to group data, the
measurements, because the primary attribute being systematic differences in calibration between subjects
tested is the accuracy of a measurement technique as introduced a second source of variation, and trending
it is applied to a study sample. In studies where trend- capability could no longer be easily evaluated using
ing capability is being analyzed, multiple serial mea- the correlation coefficients.
surement pairs (n = 10 or more) are needed. For this
type of study design, data can be analyzed on an Reporting Validation Study Data
individual subject basis (i.e., within subject) using
Since 1999 there have been concerns in the literature
regression analysis. Huang and colleagues performed
regarding how validation study data have been
a number of clinical studies comparing Doppler CO
reported, especially for studies using Bland–Altman
with bioimpedance CO methods during anesthesia for
analysis.29–32 As recently as 2016, Abu-Arafeh and
major surgery.3,28 Their surgical model provided a
colleagues published a review of 111 papers from a
range of ever-changing CO values. They plotted
two-year period, which concluded that Bland–Altman
within subject serial changes in CO over time, for
study data were poorly reported and of limited useful-
each monitoring modality and for each patient
ness to evidence-based reviews.33 Additionally, they
(n = 7 to 27 data points). They were able to visually
proposed a list of 13 key issues to be included in
identify divergences in the trend lines for CO between
reports and called for journals to provide more gui-
the different monitoring modalities and relate them to
dance on how Bland–Altman studies should be con-
interventions during the surgery (Figure 1.3). They
ducted and reported. In 2010, Critchley and
also used regression analysis as a method of quantify-
colleagues reported similar findings in relation to
ing the degree trending between the monitoring mod-
reporting trend analysis data.21 Based on the present
alities for each subject. For CO studies using Doppler
author’s experience as a researcher and journal
methods as the reference, they were able to set criteria
8
Inflation
7
Cardiac output (litre/min)
IV Fluids
6 1000ml
4 Deflation
3
Head
down tilt
2
0
0 20 40 60 80 100 120 140 160 180 200
Time from start of case (min)
Figure 1.3 Time plot of comparative CO data collected during a laparoscopic surgical procedure. The two uppermost trend lines represent
reference Doppler readings (diamond-USCOM and square-CardioQ). The lower dashed line represents a new bioimpedance device (triangles).
Note how after 100 minutes there was a definite downward divergence of bioimpedance readings relative to the Doppler readings. Major
interventions such as inflation and deflation of the pneumoperitoneum are shown. The precise cause of the downward divergence was
unclear yet present in a number of other cases. Detection of this type of change in monitor readings is very important to developers yet does
not show up in more classical group statistics using Bland–Altman and concordance analyses.
6
reviewer in the field of validation studies, the follow- used.24,25 Some authorities are now asking for
ing recommendations are provided: confidence intervals of the limits of agreement to
also be included.34 It is best to stick to simple
1. Provide a clear and thorough description of the
numerical measurement units (i.e., CO in liter/
study design, including (i) recruitment, (ii)
min) rather than percentage changes; however,
number of subjects, (iii) how readings were taken,
indexing variables to body surface area (BSA) (i.e.,
including blinding of investigators and steady-
cardiac index = CO/BSA) is acceptable. Diagrams
state synchronous readings, and (iv) timing of data
and legends should display essential numerical
collection points. Remember to mention ethical
information about the plot(s).
approval and consent.
6. Sufficient data to calculate the PE should be
2. Provide a well-described plan for analyzing the
provided, including (i) the standard deviation of
data in the methods section. Ideally, one should
the bias or 95% confidence interval and (ii) mean
measure the precision of the reference method and
CO for all the study data. Ideally, the PE should
use it when setting a priori criteria for acceptance
also be presented. The PE facilitates comparison of
of the test method.26 A typical sequence for a
data with previous studies; one may wish to make
simple test versus reference comparison study
such comparisons in the discussion section. In the
would be (i) results of any pilot studies such as
methods section the criterion threshold should be
reference method precision and power calculation
set a priori that defines a PE that supports
(i.e., study size), (ii) inspection of study data using
acceptance of the new technique. This requires
scatter plots, (iii) Bland–Altman analysis with
some consideration regarding the precision of the
details, and (iv) trend analysis using concordance
reference method. Cecconi and colleagues
and possibly polar plots. Acceptance criteria with
recommend estimating the precision from
references should be added to the relevant
coefficient of variation measurements for the
subsections.
reference method.26 The current benchmark for
3. The results section should start with the general
PE is less than 30%, but this criterion should be set
demographics of the study population, including
in the context of the precision of the reference
number of subjects and how many subjects were
method as a 20% error is presumed (see Axiak-
excluded and why. The power calculations
Flammer et al. for guidance if an alternative
justifying the size of the study, if performed, could
reference method has been used27).
be included at this point (see previous comments
on study size). 7. Depending on the study design and data structure,
if a trend analysis is performed, then a four-
4. Draw a scatter plot (optional) that shows the
quadrant plot should be drawn (Figure 1.2).
distribution of raw data (Figure 1.1). Multiple
Concordance analysis should be performed for
plots may be needed if subgroups of subjects have
studies with grouped data of sufficient numbers
been included in the study design. Addition of a
(e.g., n > 20 subjects) and serial data pairs (e.g.,
regression line and correlation coefficients is
n > 3). An exclusion zone should be employed (i.e.,
optional, as Bland–Altman recommended their
15% of mean CO for the study) to remove data
exclusion.14 Plots should contain, within the
where changes are small and data points lie close
diagram or legend, essential information such as
to zero. For CO studies the zone is set at 15% of the
number of data points and relevant statistical
mean CO value. Remember that concordance is
outcomes, for example regression line equation
based on the variable ΔCO, not CO. Criteria for
and correlation coefficient (i.e., r or R2).
accepting a CO monitor as having good trending
5. Draw the Bland–Altman plot(s) (Figure 1.1).
ability have been set at greater than 92%, where the
Make sure axes are appropriately scaled with
reference method was single bolus
sensible data ranges. If more than one plot is
thermodilution.21 For studies with a small number
presented, the scales and ranges should be similar
of data pairs, the confidence intervals for the
to facilitate visual comparison. Add horizontal
concordance also need to be calculated.27
lines for the mean bias and limits of agreement
(i.e., 95% confidence intervals of the bias). Make 8. A polar plot analysis may also be employed
sure the limits have been corrected for repeated (Figure 1.2), following advice on generating the
measures, citing which methodology was data from paired readings, creating the plots, and
interpreting the results.11,21,23 Exclude central 6. Koo KK, Sun JC, Zhou Q, et al. Pulmonary artery
zone data that are less than 10% of the mean CO catheters: evolving rates and reasons for use. Crit Care
for the study. Key outcome data are the mean Med 2011;39:1613–8.
angle and 95% radial limits of agreement. They 7. Harvey S, Harrison DA, Singer M, et al. Assessment of
should be added to the polar plot as radial lines. the clinical effectiveness of pulmonary artery catheters
The 30-degree radial axes should also be in management of patients in intensive care (PAC-
Man): a randomized controlled trial. Lancet
highlighted. Negative direction data points can be 2005;366:472–7.
rotated through half a turn (i.e., 180 degrees), but
8. Reuter DA, Huang C, Edrich T, Shernan SK, Eltzschig
not reflected, to provide a half-moon plot. Ideally,
HK. Cardiac output monitoring using indicator
the main data outcomes, including number of data dilution techniques: basics, limits, and perspectives.
points, exclusion zone size, mean angle, and radial Anesth Analg 2010;110:799–811.
limits of agreement, should be added to the
9. Yang XX, Critchley LA, Rowlands DK, Fang Z, Huang
diagram or legend. Polar plots demonstrate (i) L. Systematic error of cardiac output measured by
offsets in calibration between methods (i.e., mean bolus thermodilution with a pulmonary artery catheter
angle of greater than 5 degrees) and (ii) the level of compared with that measured by an aortic flow probe
agreement between the methods (i.e., tightness of in a pig model. J Cardiothorac Vasc Anesth
alignment of radial data points to the zero-degree 2013;27:1133–9.
axis or mean angle line). The 30-degree lines act as 10. Critchley LA, Critchley JA. A meta-analysis of studies
guides to good trending when 95% of data points using bias and precision statistics to compare cardiac
fall within their boundaries. output measurement techniques. J Clin Monit Comput
1999;15:85–91.
9. For less commonly used methods of assessing
trending, one should refer to the papers that 11. Critchley LA. Bias and precision statistics: should we
describe them. still adhere to the 30% benchmark for cardiac output
monitor validation studies? Anesthesiology
2011;114:1245.
Noncardiac Output Studies 12. Biancofiore G, Critchley LA, Lee A, et al.
The application of validation statistics is not limited Evaluation of an uncalibrated arterial pulse contour
to just CO monitoring data. They can also be applied cardiac output monitoring system in cirrhotic
to blood pressure, oxygen saturation, and hemoglobin patients undergoing liver surgery. Brit J Anaesth
level monitoring. The main difference is the criteria 2009;102:47–54.
used to determine acceptance thresholds and exclu- 13. Fuller HD. The validity of cardiac output measurement
sion zones, because of their reliance on the precision by thoracic impedance: a meta-analysis. Clin Invest
Med 1992;15:103–12.
of the reference method.
14. Bland JM, Altman DG. Statistical methods for
assessing agreement between two methods of clinical
References measurement. Lancet 1986;1:307–10.
1. Oxford Centre for Evidence-based Medicine – Levels of 15. LaMantia KR, O’Connor T, Barash PG. Comparing
Evidence (March 2009). www.cebm.net/oxford-centre- methods of measurement: an alternative approach.
evidence-based-medicine-levels-evidence-march-2009/ Anesthesiology 1990;72:781–3.
(Accessed November 2017).
16. Wong DH, Tremper KK, Stemmer EA, et al.
2. Sun JX, Reisner AT, Saeed M, Heldt T, Mark RG. The Noninvasive cardiac output: simultaneous comparison
cardiac output from blood pressure algorithms trial. of two different methods with thermodilution.
Crit Care Med 2009;37:72–80. Anesthesiology 1990;72:784–92.
3. Huang L, Critchley LA. An assessment of two Doppler- 17. Michard F. Thinking outside the (cardiac output) box.
based monitors to track cardiac output changes in Crit Care Med 2012;40:1361–2.
anaesthetised patients undergoing major surgery.
Anaesth Intens Care 2014;42:631–9. 18. Peyton PJ, Chong SW. Minimally invasive
measurement of cardiac output during surgery and
4. Chong SW, Peyton PJ. A meta-analysis of the accuracy critical care: a metaanalysis of accuracy and precision.
and precision of the ultrasonic cardiac output monitor Anesthesiology 2010;113:1220–35.
(USCOM). Anaesthesia 2012;67:1266–71.
19. Critchley LA. Validation of the MostCare pulse
5. Wang DJ, Gottlieb SS. Impedance cardiography: more contour cardiac output monitor: beyond the Bland and
questions than answers. Curr Cardiol Rep 2006;8:180–6. Altman plot. Anesth Analg 2011;113:1292–4.
8
20. Critchley LA. Meta-analyses of Bland-Altman-style 28. Huang L, Critchley LA, Zhang J. Major upper
cardiac output validation studies: good, but do they abdominal surgery alters the calibration of
provide answers to all our questions? Brit J Anaesth BioReactance cardiac output readings, the NICOM,
2017;118:296–7. when comparisons are made against suprasternal and
21. Critchley LA, Lee A, Ho AM. A critical review of the esophageal Doppler intraoperatively. Anesth Analg
ability of continuous cardiac output monitors to 2015;121:936–45.
measure trends in cardiac output. Anesth Analg 29. Mantha S, Roizen MF, Fleisher LA, Thisted R, Foss J.
2010;111:1180–92. Comparing methods of clinical measurement:
22. Perrino AC, O’Connor T, Luther M. Transtracheal reporting standards for Bland and Altman analysis.
Doppler cardiac output monitoring: comparison to Anesth Analg 2000;90:593–602.
thermodilution during noncardiac surgery. Anesth 30. Dewitte K, Fierens C, Stöckl D, Thienpont LM.
Analg 1994;78:1060–6. Application of the Bland-Altman plot for
23. Critchley LA, Yang XX, Lee A. Assessment of trending interpretation of method comparison studies: a
ability of cardiac output monitors by polar plot critical investigation of its practice. Clin Chem
methodology. J Cardiothorac Vasc Anesth 2002;48:799–801.
2011;25:536–46. 31. Berthelsen PG, Nilsson LB. Researcher bias and
24. Bland JM, Altman DG. Agreement between methods generalization of results in bias and limits of agreement
of measurement with multiple observations per analyses: a commentary based on the review of 50 Acta
individual. J Biopharm Stat 2007;17:571–82. Anaesthesiologica Scandinavica papers using the
Altman Bland approach. Acta Anaesthesiol Scand
25. Myles PS, Cui J. Using the Bland-Altman method to 2006;50:1111–3.
measure agreement with repeated measures. Br J
Anaesth 2007;99:309–11. 32. Bein B, Renner J, Scholz J, Tonner PH. Comparing
different methods of cardiac output determination: a
26. Cecconi M, Rhodes A, Poloniecki J, Della Rocca G, call for consensus. Eur J Anaesthesiol 2006;23:710.
Grounds RM. Bench-to-bedside review: the
importance of the precision of the reference technique 33. Abu-Arafeh A, Jordan H, Drummond G. Reporting of
in method comparison studies – with specific reference method comparison studies: a review of advice, an
to the measurement of cardiac output. Crit Care assessment of current practice, and specific suggestions
2009;13:201. for future reports. Br J Anaesth 2016;117:569–75.
27. Axiak-Flammer SM, Critchley LA, Weber A, et al. 34. Drummond GB. Limits of agreement with confidence
Reliability of lithium dilution cardiac output in intervals are necessary to assess comparability of
anaesthetized sheep. Brit J Anaesth 2013;111:833–9. measurement devices. Anesth Analg 2017;125:1075.
Chapter
Multimodal Neurological Monitoring
2 Samson Zarbiv, Erica Fagleman, and Neha S. Dangayach
Introduction The Monroe–Kellie Doctrine

Patients with acute neurological injuries are suscepti- Intracranial pressure is determined by the volume of
ble to injury progression and clinical deterioration. intracranial contents (i.e., brain parenchyma, cere-
The goal of neurological monitoring in these patients brospinal fluid [CSF], and blood) in patients with an
is to rapidly treat the primary neurological injury, intact cranial vault. Under normal circumstances ICP
prevent progression to secondary neurological injury, ranges from 5 to 15 mmHg, with brain parenchyma
and ultimately improve outcome. accounting for 80% of the intracranial volume and
The current monitoring paradigm for patients CSF and blood each accounting for 10%. The
with acute neurological injuries includes frequent Monroe–Kellie doctrine stipulates that the total
neurological examination, neuroimaging, and volume of these three intracranial components
multimodal monitoring, depending upon the remains constant. An increase in the volume of any
severity of the primary neurological injury and one of these three components must be accompanied
likelihood of deterioration. Clinical examination, by either a reduction in the volume of the other
including assessment of mental status and level of components or a shift of volume to outside of the
consciousness, may be confounded by comorbid- cranial vault.
ities (e.g., facial trauma confounding eye-opening In a compensated state, ICP will remain normal
response, tracheostomy precluding verbal when there is an increase in volume of any of the three
response) and use of sedatives or analgesics. intracranial contents, primarily through reducing
Furthermore, clinical examination and neuroima- venous blood volume by displacement out of the
ging assessment are limited because they are per- intracranial space, reducing intracranial CSF volume
formed intermittently and may miss crucial by diversion into the spinal subarachnoid space, or
changes that occur between examinations, and a combination of the two.1 When these compensatory
because they are to some extent subjective. By mechanisms fail, ICP will increase rapidly and brain
contrast, continuous multimodal monitoring pro- parenchyma will herniate from areas of high pressure
vides objective real-time measurement and trend- to areas of low pressure.1,2 Intracranial pressure mon-
ing of critical parameters such as intracranial itoring can guide therapy to reduce the intracranial
pressure (ICP), cerebral blood flow (CBF), cere- volume (e.g., hyperventilation, hyperosmolar therapy,
bral oxygenation, cerebral metabolism, and sys- CSF diversion, targeted temperature management)
temic hemodynamics. This chapter presents an and prevent or treat increases in ICP.3,4
overview of the principles of multimodal neuro-
monitoring that may be used in perioperative and Cerebral Autoregulation
intensive care unit (ICU) settings. The second physiological principle in the manage-
ment of ICP is cerebral autoregulation. Cerebral auto-
Brain Physiological Principles regulation maintains CBF despite variation in mean
There are two fundamental aspects of brain physiology arterial pressure (MAP). The cerebral perfusion pres-
that provide the framework for understanding neuro- sure (CPP), which is the pressure gradient driving
monitoring and management of brain injury: (1) the CBF, is simply the difference between the MAP and
Monroe–Kellie doctrine and (2) cerebral the ICP. Under normal circumstances (i.e., with
autoregulation. MAPs in the range of 50–150 mmHg) autoregulation
10
is accomplished by vasoconstriction or vasodilation of as intracranial hemorrhage, seizures, and catheter-

the cerebral vasculature to maintain a constant CBF. associated infections.
With MAPs outside the 50–150 mmHg range, The Brain Trauma Foundation guidelines recom-
autoregulation is impaired. Cerebral blood flow will mend invasive ICP monitoring with an EVD to guide
vary with MAP, and may be inadequate and cause management of patients with severe TBI, defined as
secondary ischemic insult.5 Glasgow Coma Scale (GCS) score of less than 8 after
initial resuscitation, or an abnormal head computed
Intracranial Pressure Monitoring tomography (CT) scan (class II evidence).6,7 The
guidelines offer a weaker recommendation for ICP
Intracranial pressure monitoring is the oldest and
monitoring in patients with a normal head CT scan
most commonly used physiologic neurologic moni-
and two of the three following criteria: systolic blood
tor. There are several methods of ICP monitoring that
pressure less than 90 mmHg, age greater than 40
are invasive or noninvasive, and continuous or inter-
years, or early unilateral or bilateral motor posturing
mittent. Intracranial pressure monitoring is always
(class III evidence).7 A study conducted in severe
used in conjunction with other forms of systemic
TBI patients found that therapy guided by ICP mon-
hemodynamic and neuromonitoring. Regardless of
itoring was independently associated with
the method of ICP monitoring, the objective always
a reduction in two-week mortality by 64%, compared
is to detect changes in ICP before irreversible injury
to patients in whom neurological examination and
occurs. A transient and/or minor increase in ICP may
routine imaging alone were used to guide clinical
not be harmful, but a sustained and/or significant
care.3
increase in ICP may cause a critical decrease in CPP
In patients with poor-grade aneurysmal SAH,
and CBF, herniation, and death. Intracranial pressure
EVD placement is recommended for ICP monitor-
monitoring provides an early “danger” warning that
ing and CSF diversion.8 In spontaneous ICH, EVD
cannot be detected by neurological exam in patients
placement should be considered in patients with
who already have a depressed level of consciousness
a GCS less than 9, clinical signs of transtentorial
or are pharmacologically sedated.
herniation, significant intraventricular hemorrhage
(IVH), or hydrocephalus.9
Invasive Intracranial Pressure Monitoring The ICP waveform may be used to evaluate
Invasive ICP monitoring should be considered in all intracranial compliance. Under normal physiologic
patients with clinical and/or radiographic findings conditions the morphology of the ICP waveform has
indicative of increased ICP, as often accompanies three components: P1 (percussion wave) represents
severe traumatic brain injury (TBI), subarachnoid arterial pulsation, P2 (tidal wave) represents the
hemorrhage (SAH), intracerebral hemorrhage overall intracranial compliance, and P3 (dicrotic
(ICH), bacterial meningitis, and fulminant hepatic wave) represents the venous pulsations. When intra-
failure. Invasive ICP monitoring devices directly mea- cranial compliance diminishes, the P2 wave will
sure ICP via probes that are placed in the intraven- increase in amplitude and may exceed the P1
tricular, intraparenchymal, subarachnoid, subdural, wave’s amplitude. Pathological waveforms, known
or epidural compartments. Two methods of invasive as Lundberg A and B waves, occur during sustained
ICP monitoring that are commonly employed in clin- intracranial hypertension and represent a state of
ical practice are external ventricular drains (EVDs) reduced compliance. Lundberg A waves last several
and implantable microtransducer systems such as minutes and represent sustained ICP elevation to
strain gauge devices and fiber optic sensors. greater than 50 mmHg and concomitant reduction
External ventricular drains are considered the gold in CPP; they indicate evolving brain injury.
standard of ICP monitoring. In addition to measuring Lundberg B waves last only 1–3 minutes and repre-
ICP, they can be used therapeutically to drain CSF sent an elevation of ICP that is less than the
(i.e., CSF diversion) to treat increased ICP. 50 mmHg range. They may indicate a compromise
Microtransducer devices measure localized ICP, and in the intracranial compliance and may precede the
their measurements correlate well with those of development of A waves. Lundberg A waveforms are
EVDs.2 They lack therapeutic potential but compared always pathological, but B waveforms vary in their
to EVDs they have lower rates of complications such clinical significance.
Downloaded from https://www.cambridge.org/core. University of New England, on 03 May 2020 at 06:06:28, subject to the Cambridge Core terms of use, available at 11
Noninvasive Intracranial Pressure Invasive Cerebral Blood Flow Monitoring

Monitoring Thermal diffusion flowmetry (TDF) is an invasive
There are two noninvasive indirect measures of ICP technique that provides quantitative assessments of
regional CBF (rCBF). The TDF catheter utilizes
that can be performed at the bedside, transcranial
a distal thermistor heated to a few degrees above the
Doppler ultrasound and transocular ultrasound.
regional tissue temperature, and a second, more prox-
Transcranial Doppler (TCD) is the primary
imal temperature probe.15 The difference in tempera-
modality for noninvasive measurement of ICP. It
ture between the two probes reflects the absolute
emits ultrasound waves that travel through various
blood flow between them and is expressed in mL/
layers of tissue until they encounter red blood cells
100g/min. Thermal diffusion flowmetry can provide
(RBCs) in blood vessels. After contacting the RBCs,
the ultrasound waves are reflected back and the continuous CBF data and has been shown to correlate
with rCBF data obtained from simultaneous xenon-
frequency of the returning ultrasound waves indi-
enhanced CT imaging. Normal rCBF values range
cates the direction and velocity of blood flow in the
between 40 and 70 mL/100g/min; values less than
large arteries of the brain, most commonly the
20 mL/100g/min will result in cerebral ischemia with-
middle cerebral artery (MCA). Transcranial
out a concomitant reduction in metabolic demand
Doppler can be used to calculate the pulsatile
(e.g., hypothermia, drug-induced coma). Thermal dif-
index (PI), which is the difference between systolic
fusion flowmetry catheters are usually placed in white
and diastolic blood flow velocities divided by the
mean blood flow velocity. The PI correlates with matter regions that are deemed to be at greatest risk of
ischemia. In conjunction with other monitoring
invasively measured ICP.10 Pulsatile index values
devices, they may also assist detection of cerebral
greater than 1.4 are indicative of elevated ICP and
vasospasm in SAH patients.16
are associated with worse clinical outcomes (sensi-
Cerebral blood flow monitoring with TDF can also
tivity 88%, specificity 97%).11 Transcranial Doppler
be used to estimate brain water content (BWC). In
does not have the risks of invasive methods, but its
a study of 36 comatose brain-injured patients, brain
main disadvantage is that it also does not provide
regions that appeared edematous on CT imaging
continuous data.
Transocular ultrasound of the optic nerve sheath showed higher TDF-calculated BWC than normal-
appearing brain regions.17 These same patients had
diameter is another noninvasive ICP monitoring
a calculated decrease in BWC after administration of
modality. Increased optic nerve sheath diameter cor-
hyperosmolar therapy, as would be expected. These
relates with increased ICP, because in between the
findings suggest that BWC monitoring utilizing TDF
optic nerve sheath and the nerve itself, there is sub-
is feasible.
arachnoid space that is in communication with the
Unfortunately, TDF may only be possible in
subarachnoid space of the brain. Optic nerve sheath
30–40% of cases due to high rates of monitor failure
diameter greater than 5 mm is pathological.12–14 This
noninvasive method of ICP monitoring is easily per- and loss of data during recalibration.5 Given the high
rates of failure, TDF is usually used in conjunction
formed at the bedside, but its accuracy is inferior to
with other brain monitoring.
invasive ICP measurement and it does not provide
continuous data.
Noninvasive Cerebral Blood Flow
Cerebral Blood Flow Monitoring Monitoring
As reviewed above, CBF is normally maintained In addition to noninvasive measurement of ICP, TCD
within a wide range of MAPs (50–150 mmHg) by can also be used to noninvasively measure CBF (ml/
adjusting blood vessel diameter, so that CPP is main- gm/min). A blood flow velocity (BFV) in the
tained between 50 and 70 mmHg. Impaired autore- MCA greater than 120 cm/sec is considered patholo-
gulation, increased ICP, or decreased MAP may gical and correlates with cerebral vasospasm (sensi-
reduce CBF and result in brain injury. Cerebral tivity 73%, specificity 80%).8,15
blood flow monitoring may detect changes before Patients with SAH should have periodic TCD
they lead to injury and be used to guide preventive assessments for early diagnosis of cerebral vasospasm.
measures. Prompt management of vasospasm may prevent
12
delayed cerebral ischemia (DCI), a frequent and critical this monitoring technique are vascular injury of the
complication of SAH that occurs in up to 70% of cases.8 internal jugular vein, inadvertent cannulation of the
The severity of vasospasm can be graded using the carotid artery, and thrombosis of these vessels.
Lindegaard ratio (LR), which is the ratio of the MCA
BFV to the extracranial internal carotid artery (ICA) Intraparenchymal Catheters
BFV. An LR less than 3 is considered normal, whereas The partial pressure of oxygen in brain parenchyma
LR 3–6 is associated with mild to moderate vasospasm (PbtO2) can be directly measured using catheters that
and LR greater than 6 with severe vasospasm.18 are usually inserted into the subcortical white matter.
There are two main limitations to this applica- These catheters employ the Clark cell method, utiliz-
tion. Although cerebral vasospasm is the most coming a semipermeable membrane that allows dissolved
mon cause of elevated CBF, it is not the only oxygen to pass through it and generate an electrical
pathologic cause. Common conditions such as fever current proportional to the PbtO2.21 Most clinical
and increased cardiac output due to any cause (e.g., data regarding PbtO2 monitoring are from studies in
peripheral shunting, hyperthyroidism, catechola- severe TBI and, to a lesser extent, SAH patients.22
mine administration) will increase BFV.15 When placed in normal parenchyma, PbtO2 values
Therefore, BFV should be interpreted in the context represent global cerebral oxygenation; placement in
of the patient’s condition. Another limitation is that injured areas will result in regional oxygenation
TCD is not a continuous monitoring technique, and assessment.
there is risk of missing significant CBF changes In general, PbtO2 monitors are placed when the
between exams. GCS is ≤ 8 (i.e., similar to indications for ICP mon-
itoring). There is no consensus where in the brain
Brain Oxygen Monitoring these devices should be placed, but in TBI the right
frontal lobe is usually chosen, and in a focal injury the
Critically ill patients recovering from neurologic
side of maximal pathology is chosen. In SAH, the
injury are vulnerable to decreased brain oxygenation.
PbtO2 monitor should be placed in the brain region
As with increased ICP, transient decreased brain oxy-
most likely impacted by the expected vasospastic
genation may not have any consequence, but sus-
response.23 Once placed and confirmed with CT ima-
tained cerebral hypoxia can be devastating. There
ging, an “oxygen challenge” should be conducted by
are several invasive and noninvasive methods of mon-
increasing FiO2 from baseline to 1.0 for approxi-
itoring brain oxygenation used to guide therapy.
mately five minutes; a functional probe will reveal an
increase in the PbtO2. The increase is less robust when
Invasive Brain Oxygen Monitoring the probe is in a hypoperfused region (e.g., CBF is less
than 20 ml/100 g/min).24
Jugular Venous Oxygen Saturation The PbtO2 is a more complex variable to interpret
Jugular venous oxygen saturation (SvjO2) can be mea- than SvjO2 because it may vary significantly by site of
sured intermittently by sampling blood from the placement and might be more sensitive to small
jugular venous bulb from a catheter placed in changes. It is influenced by global determinants of
a retrograde direction, or continuously using a fiber oxygen delivery (i.e., cardiac output and blood oxygen
optic catheter similarly placed. It is used to assess content), ICP, CPP, autoregulation, and specifically
oxygen demand–supply mismatch. An increase in cerebral oxygen tissue gradients (i.e., nonhomogen-
the demand-to-supply ratio will result in a decrease ous cerebral oxygenation), which are more prominent
in the oxygen saturation of blood draining from the in the injured brain.
brain. The normal range of SvjO2 is 55–75%, and Normal PbtO2 values range from 25 to 35 mmHg.
lower values correlate with poorer outcomes.19 Values less than 20 mmHg are suggestive of cerebral
Current guidelines recommend maintaining SvjO2 ischemia, and those less than 15 mmHg predict
greater than 50%.7 Any condition resulting in arter- poor outcome in severe TBI and SAH patients.25
iovenous shunting or decreased brain metabolism Monitoring of PbtO2 should be used when secondary
(e.g., hypothermia, drug-induced coma, and brain brain injury is likely, and it is best used in an integrated
death) will result in a higher SvjO2; fever, shivering, fashion with other monitors, clinical evaluation, and
and seizures will reduce the SvjO2.20 The main risks of imaging studies.26 In TBI and SAH, PbtO2 monitoring
should always be used in conjunction with ICP mon- closer to +1.0 indicate a positive linear correlation
itoring because cerebral hypoxia may occur even when and that autoregulation is impaired. In patients with
ICP and CPP are normal and would otherwise go TBI such values predict greater morbidity and mor-
undetected and be potentially injurious.25,27 tality. A plot of CPP on the x-axis and PRx on the
y-axis results in a U-shaped curve, and the nadir point
Noninvasive Brain Oxygen Monitoring represents the optimal CPP. This point is patient-
specific and differs based on underlying injury and
Near-infrared spectroscopy (NIRS) is the primary
degree of autoregulation preservation.32
noninvasive method of measuring cerebral oxyge-
The oxygen reactivity index (ORx) is the Pearson
nation. This technique determines the concentra-
correlation coefficient of the CPP and PbtO2.30
tion of oxygenated hemoglobin in the brain (rSO2)
Understanding the relationship between CBF and
using sensors placed on the forehead that deter-
oxygen/metabolic demand and delivery can help
mine the absorption of light in the near-infrared
guide management in real time.33
wavelength by brain tissue and hemoglobin.28
Monitoring of PRx and ORx requires continuous
Changes in cerebral oxyhemoglobin concentration
data processing, which may be expensive and time-
help detect impaired regional brain oxygenation
consuming, limiting their use as a bedside modality.
that may be due to reduced blood oxygen content
An observational study of low-frequency, minute-by-
or blood flow. Values for cerebral oxyhemoglobin
minute assessment of MAP and ICP data found that
less than 60% are considered pathologic and are
real-time measurement of optimal CPP was possible
associated with cerebral hypoxia. Unfortunately,
and that continuous processing may not be
incorporating NIRS monitoring with other inva-
required.26 Despite encouraging results from small
sive strategies has not been shown to improve
clinical trials supporting the hypothesis that indivi-
outcome (see Chapter 3).29
dualized CPP monitoring yields improved outcomes
in TBI patients, there is insufficient evidence to sup-
Cerebral Autoregulation Monitoring port widespread adoption.
Current Brain Trauma Foundation guidelines recom-
mend maintaining CPP between 60 and 70 mmHg in Brain Metabolism Monitoring:
patients with TBI; this goal is often extrapolated to
other conditions.7 Cerebral Microdialysis
Preservation of normal cerebral autoregulation is Cerebral microdialysis (MD) is a technique that
the most important factor for determining the lowest allows for direct assessment of the brain’s extracellu-
target CPP.30 Aggressive attempts to increase CPP lar biochemical milieu by a fine-tipped double-lumen
above 70 mmHg by administering fluids and/or probe with a semipermeable membrane inserted into
vasoactive agents are not recommended due to the the subcortical white matter. The fundamental prin-
associated risk of acute lung injury.7 ciple of MD is the same as that of hemodialysis; sub-
Cerebral perfusion pressure is a crude target used stances will cross a semipermeable membrane along
in the absence of more sophisticated monitoring. their concentration gradient. The final concentration
Intracranial pressure, cerebral oximetry, and TCD of substances in the dialysate depends on the differ-
measurements, however, may be used to assess cere- ence between their baseline concentration in the
bral autoregulation and have been studied primarily probe and their uptake from the brain’s extracellular
in the setting of TBI.31 fluid (ECF). Perfusate, a fluid isotonic to the ECF, is
Intracranial pressure data can be used to compute introduced into the catheter, and molecules at high
the pressure reactivity index (PRx), which reflects the concentration in the ECF equilibrate through the
degree of cerebral autoregulation. The PRx is the semipermeable membrane. The resulting microdialy-
Pearson correlation coefficient of the MAP and ICP sate can then be analyzed. Several substances, includ-
derived from 30 consecutive, 10-second averaged ing glucose, lactate, pyruvate, glutamate (an excitatory
values over a 5-minute period. Correlation coeffi- neurotransmitter associated with the inflammatory
cients range from +1.0 to −1.0; PRx values near 0 cascade), and glycerol (a marker of neuronal cell
indicate that there is no correlation between the ICP breakdown associated with irreversible cell injury),
and MAP and that autoregulation is intact. Values can be measured at the bedside.33
14
Glucose serves as the brain’s primary energy status does not recover following a clinical
source, and its metabolism to adenosine triphosphate seizure.43,44 In comatose patients, it is estimated
(ATP) provides the fuel for brain function. During that 30% have a cEEG-detected seizure after 24
glycolysis the production of pyruvate enters the citric hours of monitoring that would have gone undiag-
acid cycle to produce energy at the cellular level. nosed in the absence of monitoring.43 As with
During ischemia, seizures, and other situations in other neuromonitoring modalities, the goal of
which oxygen demand exceeds supply, pyruvate is cEEG in the ICU setting is early detection so that
shunted to the anaerobic metabolic pathway and is potential secondary neurologic injury may be
converted to lactate.33 The lactate-to-pyruvate ratio avoided by early administration of antiepileptics.
(LPR) is considered to be the most reliable marker of Prompt recognition and treatment of seizures pre-
cerebral metabolic derangement due to impaired oxi- vents resistance to antiepileptic therapy. 45,46
dative metabolism. Elevated LPR may be the result of Continuous EEG also has been used to detect
ischemic injury and is associated with neuronal cell vasospasm and increased ICP as a component of
death.34 multimodal prognostication in cardiac arrest.43
Microdialysis can be used to detect early signs
of secondary neurological insult and may be clini- Evoked Potentials
cally useful as part of multimodality monitoring Somatosensory evoked potentials (SSEPs) are elec-
paradigm in patients with acute brain injury. trical signals emitted from the cerebral cortex in
Microdialysis studies, conducted primarily in response to a sensory stimulus delivered at the
patients with severe TBI and poor-grade SAH,22 periphery. These are often used in combination
reveal significant and predictable patterns of meta- with motor evoked potentials (MEPs), electrical
bolic derangement, including elevated levels of signals emitted from muscles in response to stimu-
lactate, pyruvate, LPR, glutamate, and glycerol, as lation of the motor cortex via scalp electrodes.
well as low levels of glucose after periods of sus- Evoked potential monitoring helps to evaluate the
tained cerebral ischemia.35 integrity of the ascending somatosensory and des-
Values of LPR greater than 40 represent cerebral cending motor pathways, and is an important com-
“metabolic distress,” whereas the combination of ponent of intra-operative monitoring for surgical
LPR greater than 40 and brain glucose levels less than complications during spine and brain surgery.47
0.7 mmol/L represents brain “metabolic crisis.”35–37 Somatosensory evoked potential signals are much
These abnormalities have been associated with poor smaller in amplitude than typical EEG signals, and
outcomes in TBI patients.33,38 Detected metabolic therefore the SSEP waveform represents an average
changes can predict vasospasm in patients with high- of many signals in response to multiple stimuli
grade SAH and may be used to guide timely treat- recorded over a brief period of time.48
ment to help prevent DCI and neurologic Somatosensory evoked potentials serve as an
deterioration.39–41 important noninvasive method of assessing subcor-
tical structures, and unlike many other neuromo-
Electrophysiology nitoring modalities, they are not subject to
interference by intravenous hypnotics or
Electroencephalography sedatives.43,48 Evoked potential monitoring, how-
Continuous electroencephalography (cEEG) moni- ever, requires specialized equipment and onsite
toring with video patient recording and remote interpretation by a technician who is dedicated to
monitoring capability is now fairly routine in neu- a single patient and in communication with
rological ICUs. In addition to its value in diagnos- a neurologist. The equipment and expertise
ing non-convulsive seizures, cEEG is integral to required make the timely application of this mon-
guide careful titration of sedatives and antiepilep- itoring challenging.
tics in status epilepticus, minimizing deleterious
side effects while suppressing seizure activity.42 It “Big Data” in Critical Care
is recommended in patients with an unexplained The goal of multimodal neuromonitoring is contin-
persistently altered mental status or whose mental uous assessment of ICP, CBF, PbtO2, brain
metabolism, and neurophysiological function in Some programs developed for research pur-
order to establish baseline status, identify early poses have also shown promising results to under-
signs of deterioration that are treatable, and pre- stand time-synchronized relationships between
vent secondary neurologic injury. Multimodal neu- different neuromonitoring parameters. The ICM+,
romonitoring data may also be used to refine developed at Cambridge University, integrates ICP
prediction of morbidity and mortality, and help waveform data with MAP data to monitor
to allocate expensive and scarce resources such as autoregulation. Another recent addition is the
ICU beds. Moberg Research© multimodal neuromonitoring
Intensive care units are data-rich environ- platform that includes measures of cerebrovascular
ments. The complexity of critical care and the autoregulation, brain tissue oxygenation,
enormous amount of data generated present an microdialysis, and cEEG to enhance decision
exciting opportunity. Display and trending of support.58
information and data from clinical examination, In sum, neurocritical care depends on com-
devices, laboratories, and imaging should be able plex bedside monitoring, even though the speci-
to drive evidence-based decision support and fics of what should be monitored and how remain
improve outcomes.49 Despite the promise, most unanswered. The exact impact and costs of the
research has focused on using this information various strategies discussed herein are largely
to validate patient outcome scoring systems, and unknown. Successfully integrating monitoring
has not utilized it in decision support to with all the other data available into the clinical
a significant degree.50 Commercially available pro- workflow while avoiding injury and data misin-
grams that integrate data from various monitors terpretation is difficult. Achieving improved
and electronic health records (EHRs) for display patient outcomes as a result of these efforts, how-
and decision support have not been widely ever, is clearly within reach.
adapted. Electronic health records alone could
potentially evolve into artificial intelligence sys- References
tems and move beyond their current feature
1. Jones RF, Dorsch NW, Silverberg GD, Torda TA.
“best practice alerts” into continuous evidence- Pathophysiology and management of raised
based practice drivers.51,52 intracranial pressure. Anaesth Intensive Care
There are several promising databases that are 1981;9:336–51.
designed for integrating data from general critical 2. Kirkman MA, Smith M. Intracranial pressure
care patients.53,54 The Multiparameter Intelligent monitoring, cerebral perfusion pressure estimation,
Monitoring in Intensive Care (MIMIC) database and ICP/CPP-guided therapy: a standard of care or
is a large open-access database that contains optional extra after brain injury? Br J Anaesth
demographic, physiological, medication, and out- 2014;112:35–46.
come data recorded from thousands of ICU 3. Treggiari MM, Schutz N, Yanez ND, Romand, JA.
patients admitted to Beth Israel Deaconess Role of intracranial pressure values and patterns
Medical Center (Boston, Massachusetts, USA).55 in predicting outcome in traumatic brain
These types of databases can be important injury: a systematic review. Neurocrit Care
2007;6:104–12.
resources for patient monitoring research and
will support efforts in medical data mining and 4. Muizelaar JP, Wei EP, Kontos HA, Becker DP. Mannitol
causes compensatory cerebral vasoconstriction and
knowledge discovery.56
vasodilation in response to blood viscosity changes.
The International Initiative for Traumatic Brain J Neurosurg 1983;59:822–8.
Injury Research has developed the Federal
5. Akbik OS, Carlson AP, Krasberg M, Yonas H. The
Interagency TBI Research Informatics System plat-
utility of cerebral blood flow assessment in TBI. Curr
form to collect patients’ phenotypic, genomic, and Neurol Neurosci Rep 2016;16:72.
imaging information. As part of this initiative, the
6. Brain Trauma Foundation. Guidelines for the
Center-TBI database collects detailed information management of severe traumatic brain injury.
from thousands of ICU patients in Europe.57 These VI. Indications for intracranial pressure
databases have been shown to help better understand monitoring. J Neurotrauma 2007;24 Suppl 1:
patient outcomes.57 S37–44.
16
7. Carney N, et al. Guidelines for the management of 19. Robertson CS, Gopinath SP, Goodman JC, et al. SjvO2
severe traumatic brain injury, Fourth Edition. monitoring in head-injured patients. J Neurotrauma
Neurosurgery 2017;80:6–15. 1995;12:891–6.
8. Bederson JB, et al. Guidelines for the management of 20. Schell RM, Cole, DJ. Cerebral monitoring:
aneurysmal subarachnoid hemorrhage: a statement for jugular venous oximetry. Anesth Analg
healthcare professionals from a special writing group 2000;90:559–66.
of the stroke council, American Heart Association. 21. Siegemund M, van Bommel J, Ince C. Assessment of
Stroke 2009;40:994–1025. regional tissue oxygenation. Intensive Care Med
9. Steiner T, Al-Shahi Salman R, Christensen H, 1999;25:1044–60.
et al. European Stroke Organisation (ESO) 22. Francoeur CL, Pain M, Mayer SA. Multimodality
guidelines for the management of spontaneous monitoring: illuminating the Comatose Human Brain.
intracerebral hemorrhage. Int J Stroke Semin Neurol 2016;36:560–9.
2014;9:840–55.
23. Bohman LE, Pisapia JM, Sanborn MR, et al. Response
10. Bathala L, Mehndiratta MM, Sharma VK. of brain oxygen to therapy correlates with long-term
Transcranial Doppler: technique and common outcome after subarachnoid hemorrhage. Neurocrit
findings (part 1). Ann Indian Acad Neurol Care 2013;19:320–8.
2013;16:174–9.
24. Hlatky R, Valadka AB, Gopinath SP, Robertson CS.
11. Cardim D, Robba C, Bohdanowicz M, et al. Non- Brain tissue oxygen tension response to induced
invasive monitoring of intracranial pressure using hyperoxia reduced in hypoperfused brain. J Neurosurg
transcranial Doppler ultrasonography: is it possible? 2008;108:53–8.
Neurocrit Care 2016;25:473–91.
25. Nangunoori R, Maloney-Wilensky E, Stiefel M,
12. Rajajee V, Vanaman M, Fletcher JJ, Jacobs TL. et al. Brain tissue oxygen-based therapy and
Optic nerve ultrasound for the detection of raised outcome after severe traumatic brain injury:
intracranial pressure. Neurocrit Care a systematic literature review. Neurocrit Care
2011;15:506–15. 2012;17:131–8.
13. Robba C, Santori G, Czosnyka M, et al. Optic nerve 26. Johnston AJ, Steiner LA, Coles JP, et al. Effect of
sheath diameter measured sonographically as cerebral perfusion pressure augmentation on
non-invasive estimator of intracranial pressure: regional oxygenation and metabolism after head
a systematic review and meta-analysis. Intensive Care injury. Crit Care Med 2005;33:189–95; discussion
Med 2018;44:1284–94. 255–187.
14. Suarez JI, Qureshi AI, Yahia AB, et al. Symptomatic 27. Oddo, M, Levine JM, Mackenzie L, et al. Brain
vasospasm diagnosis after subarachnoid hemorrhage: hypoxia is associated with short-term outcome after
evaluation of transcranial Doppler ultrasound and severe traumatic brain injury independently of
cerebral angiography as related to compromised intracranial hypertension and low cerebral perfusion
vascular distribution. Crit Care Med pressure. Neurosurgery 2011;69:1037–45; discussion
2002;30:1348–55. 1045.
15. Vajkoczy P, Roth H, Horn P, et al. Continuous 28. Naidech AM, Bendok BR, Ault ML, Bleck TP.
monitoring of regional cerebral blood flow: Monitoring with the somanetics INVOS 5100C after
experimental and clinical validation of a novel aneurysmal subarachnoid hemorrhage. Neurocrit Care
thermal diffusion microprobe. J Neurosurg 2000;93: 2008;9:326–31.
265–74.
29. Rosenthal G, Furmanov A, Itshayek E, Shoshan Y,
16. Al-Mufti F, Amuluru K, Damodara N, et al. Singh V. Assessment of a noninvasive cerebral
Novel management strategies for oxygenation monitor in patients with severe
medically-refractory vasospasm following traumatic brain injury. J Neurosurg
aneurysmal subarachnoid hemorrhage. 2014;120:901–7.
J Neurol Sci 2018;390:44–51.
30. Jaeger M, Dengl M, Meixensberger J,
17. Ko SB, Choi HA, Parikh G, et al. Real time Schuhmann MU. Effects of cerebrovascular
estimation of brain water content in comatose pressure reactivity-guided optimization of cerebral
patients. Ann Neurol 2012;72:344–50. perfusion pressure on brain tissue oxygenation
18. Lindegaard KF, Nornes H, Bakke SJ, Sorteberg W, after traumatic brain injury. Crit Care Med
Nakstad P. Cerebral vasospasm diagnosis by means of 2010;38:1343–7.
angiography and blood velocity measurements. Acta 31. Jaeger M, Schuhmann, MU, Soehle M,
Neurochir (Wien) 1989;100:12–24. Meixensberger J. Continuous assessment of
cerebrovascular autoregulation after traumatic brain 43. Caricato A, Melchionda I, Antonelli M. Continuous
injury using brain tissue oxygen pressure reactivity. electroencephalography monitoring in adults in the
Crit Care Med 2006;34:1783–8. intensive care unit. Crit Care 2018;22:75.
32. Czosnyka M, Smielewski P, Piechnik S, Steiner LA, 44. Herman ST, et al. Consensus statement on
Pickard JD. Cerebral autoregulation following head continuous EEG in critically ill adults and children,
injury. J Neurosurg 2001;95:756–63. part I: indications. J Clin Neurophysiol
33. Zeiler FA, Thelin EP, Helmy A, et al. A systematic 2015;32:87–95.
review of cerebral microdialysis and outcomes in 45. Claassen J, Mayer SA, Kowalski RG, Emerson RG,
TBI: relationships to patient functional outcome, Hirsch LJ. Detection of electrographic seizures with
neurophysiologic measures, and tissue outcome. continuous EEG monitoring in critically ill patients.
Acta Neurochir (Wien) 2017;159:2245–73. Neurology 2004;62:1743–8.
34. Sahuquillo J, Merino MA, Sánchez-Guerrero A, et al. 46. Farrokh S, Tahsili-Fahadan P, Ritzl EK, Lewin JJ 3rd,
Lactate and the lactate-to-pyruvate molar ratio cannot Mirski MA. Antiepileptic drugs in critically ill
be used as independent biomarkers for monitoring patients. Crit Care 2018;22:153.
brain energetic metabolism: a microdialysis study in 47. Chang SH, Park YG, Kim DH, Yoon SY.
patients with traumatic brain injuries. PLoS One Monitoring of motor and somatosensory evoked
2014;9:e102540. potentials during spine surgery: intraoperative
35. Hlatky R, Valadka AB, Goodman JC, Contant CF, changes and postoperative outcomes. Ann Rehabil
Robertson CS. Patterns of energy substrates during Med 2016;40:470–80.
ischemia measured in the brain by microdialysis. 48. Rosenthal ES. The utility of EEG, SSEP, and other
J Neurotrauma 2004;21:894–906. neurophysiologic tools to guide neurocritical care.
36. Helbok R, Schmidt JM, Kurtz P, et al. Systemic Neurotherapeutics 2012;9:24–36.
glucose and brain energy metabolism after 49. Sanchez-Pinto LN, Luo Y, Churpek, MM. Big data
subarachnoid hemorrhage. Neurocrit Care and data science in critical care. Chest
2010;12:317–23. 2018;154:1239–48.
37. Hillered L, Vespa PM, Hovda DA. Translational 50. Yang S, Stansbury LG, Rock, P, Scalea T, Hu PF.
neurochemical research in acute human brain Linking big data and prediction strategies: tools,
injury: the current status and potential future for pitfalls, and lessons learned. Crit Care Med 2019;47
cerebral microdialysis. J Neurotrauma 2005;22:3–41. (6):840–8.
38. Vespa, P, Prins M, Ronne-Engstrom E, et al. Increase 51. Learning Health Systems | Agency for Healthcare
in extracellular glutamate caused by reduced cerebral Research & Quality www.ahrq.gov/professionals/sys
perfusion pressure and seizures after human traumatic tems/learning-health-systems/index.html (2019).
brain injury: a microdialysis study. J Neurosurg
1998;89:971–82. 52 Li P, Xie C, Pollard T, et al. Promoting secondary
analysis of electronic medical records in China:
39. Sarrafzadeh AS, Sakowitz OW, Kiening KL, et al. summary of the PLAGH-MIT critical data
Bedside microdialysis: a tool to monitor cerebral conference and health Datathon. JMIR Med Inform
metabolism in subarachnoid hemorrhage patients? 2017;5:e43.
Crit Care Med 2002;30:1062–70.
53. Flechet M, Grandas FG, Meyfroidt G. Informatics in
40. Sarrafzadeh A, Haux D, Kuchler I, Lanksch WR, neurocritical care: new ideas for Big Data. Curr Opin
Unterberg AW. Poor-grade aneurysmal subarachnoid Crit Care 2016;22:87–93.
hemorrhage: relationship of cerebral metabolism to
outcome. J Neurosurg 2004;100: 400–6. 54. ICM+ Features | Cambridge Enterprise ICM+., https://
icmplus.neurosurg.cam.ac.uk/home/icm-features/
41. Skjoth-Rasmussen J, Schulz M, Kristensen SR, (2018).
Bjerre P. Delayed neurological deficits detected by an
ischemic pattern in the extracellular cerebral 55. Johnson AE, Pollard TJ, Shen L, et al. MIMIC-III,
metabolites in patients with aneurysmal subarachnoid a freely accessible critical care database. Sci Data
hemorrhage. J Neurosurg 2004;100:8–15. 2016;3:160035.
42. Ferguson M, Bianchi MT, Sutter R, et al. Calculating 56. Saeed M, Villarroel M, Reisner AT, et al.
the risk benefit equation for aggressive treatment of Multiparameter intelligent monitoring in intensive
non-convulsive status epilepticus. Neurocrit Care care II: a public access intensive care unit database. Crit
2013;18:216–27. Care Med 2011;39:952–60.
18
57. Maas AI, Menon DK, Steyerberg EW, et al. 58. Multimodal Data Integration – Moberg ICU
Collaborative European NeuroTrauma Effectiveness Solutions: Transforming Neurocritical Care.,
Research in Traumatic Brain Injury (CENTER-TBI): www.moberg.com/solutions/multimodal-data-
a prospective longitudinal observational study. integration (2018).
Neurosurgery 2015;76:67–80.
Chapter
Cerebral Oximetry
3 Benjamin Salter and Elvera Baron
Introduction Cerebral Oximetry

The risk of perioperative neurologic injury has
decreased over the years with advances in surgical Principles of Near-Infrared Spectroscopy
and anesthetic techniques; however, it remains a sig- In 1977, Franz Jabsis presented the concept of a
nificant clinical problem that may result in severe noninvasive real-time measurement of cerebral
debilitation and death. Of non-neurological surgeries, tissue oxygenation using wavelengths of light in
cardiac surgery is associated with the highest risk of the near-infrared spectrum.6 His theory was based
perioperative neurologic injury, including cognitive on the observation that near-infrared light (wave-
dysfunction, postoperative visual loss, cerebrovascu- length range 660–940 nm) passes easily through
lar accidents (CVA), nonfatal diffuse encephalopathy, skull and brain tissue. In this range, light is
and peripheral nerve injury. The incidence of new absorbed by certain biological molecules in the
ischemic lesions following cardiac surgery is 27.6% brain, including hemoglobin, and to a much lesser
in coronary artery bypass graft (CABG) surgery, degree by water, lipids, skin, and bone, allowing
63% in aortic surgery, and 84% in valve replacement.1 for little interference.7–10 Hemoglobin molecules,
In particular, the incidence of a stroke in the cardiac present in both oxygenated and deoxygenated
surgical population is estimated to range from 1 to 7% forms, have different absorption spectra.
and approaches 12% in patients with severe carotid Deoxygenated hemoglobin absorbs light waves in
stenosis.2,3 Most of these events occur during the the 650–1000 nm range, and oxygenated hemoglo-
operation, or at least their genesis begins intraopera- bin absorbs light waves in the 700–1150 nm
tively. Thus, there is a need to better assess for poten- range.10 Using the principles of the Beer–
tial neurologic injury, guide hemodynamic therapy, Lambert law and proprietary algorithms, a ratio
and alert physicians that a neurologic rescue or pro- or percentage of oxygenated hemoglobin to deox-
tective therapy is indicated. ygenated hemoglobin is generated and an rScO2
Noninvasive intraoperative monitoring of regio- value is displayed. Cerebral oximeters do not rely
nal cerebral oxygen saturation (rScO2) by near-infra- on pulsatile flow and assume a fixed proportion
red spectroscopy (NIRS) is increasingly used in this (70:30 or 75:25) of venous-to-arterial blood
vein to ensure an adequate balance between cerebral volume. Bilateral cerebral oximetry probes provide
oxygen supply and demand. Since perioperative neu- information on the regional supply and demand
rologic injury is not unique to cardiac surgical balance of the frontal cortical tissue supplied by
patients, NIRS technology is increasingly employed the anterior and middle cerebral arteries. This
in orthopedic, vascular, and neurosurgical procedures watershed area is the most vulnerable to hypoxe-
too. It is also being utilized in other clinical settings, mia.8,11,12 Very basically, the lower the cerebral
including trauma, critical care, and cardiopulmonary oximetry value (i.e., rScO2), the greater the
resuscitation.4,5 This chapter reviews the basic princi- likelihood that there is an oxygen delivery–
ples and interpretation of cerebral oximetry, research consumption imbalance. Similar to the systemic
studies that examine its efficacy, and novel uses of this circulation, oxygen delivery is dependent on both
technology. For the purpose of simplification, the flow and oxygen content of the blood, while con-
terms “NIRS,” “cerebral oximetry,” and “rScO2” will sumption is determined by metabolic demand and
be used interchangeably throughout the following is most affected by temperature and depth of
discussion. anesthesia.
20
Cerebral Oximetry
Figure 3.2 CASMED FORESIGHT™ monitor, cables, and oximetry

probes. (A black and white version of this figure will appear in some
Figure 3.1 CASMED FORESIGHT™ cerebral oximetry adult probe
formats. For the color version, please refer to the plate section.)
placement. *Printed with permission from the manufacturer.
*Printed with permission from the manufacturer.
Deep detector
Shallow
detector
emitter
Figure 3.3 Somanetics INVOS™ cerebral oximetry monitor. (A black ~2.5 cm Depth of penetration
and white version of this figure will appear in some formats. For the
color version, please refer to the plate section.) *Printed with
permission from the manufacturer.
Figure 3.4 CASMED FORESIGHT™ oximetry probe and depth of
penetration. (A black and white version of this figure will appear in
some formats. For the color version, please refer to the plate section.)
Normally, the cerebral oximetry system consists of *Printed with permission from the manufacturer.
an oximeter probe (or two) attached to a cable and
device monitor (Figures 3.1–3.3). In the United States,
a variety of NIRS monitoring devices are commer- detector receiving the reflection from deeper tissues
cially available, including the INVOS™ (Somanetics/ (approximately 2–2.5 cm) (Figure 3.4). The
Covidien, Inc., Boulder, CO), FORESIGHT™ (CAS EQUANOX device uses two emitters and two detec-
Medical Systems, Branford, CT), and EQUANOX™ tors (Figure 3.5). Although all three of the proprietary
(Nonin Medical Inc., Plymouth, MN).11 The INVOS sensors have some amount of extracranial signal con-
and FORESIGHT probes utilize a single light-emit- tamination, EQUANOX has the least compared to the
ting source and two detectors spaced several centi- other two.13 Unfortunately, the different proprietary
meters apart, with the proximal detector receiving the algorithms used by each system make comparisons of
reflection from superficial tissues and the more distal the different NIRS technologies difficult.12
Cerebral Oximetry
BOX 3.1 Common Causes of Increased rScO2
• Hypoventilation
• Hypertension
• Hypothermia
BOX 3.2 Common Causes of Decreased rScO2

↓Supply ↑Demand
Figure 3.5 Nonin EQUANOX™ cerebral oximetry sensors. (A black • Perfusion pressure: • Inadequate
and white version of this figure will appear in some formats. For the cardiac output, mean anesthesia
color version, please refer to the plate section.) *Printed with arterial pressure • Seizure with/without
permission from the manufacturer.
• Hypoxemia convulsions
• Anemia • Hyperthermia
Interpretation (hemorrhage,
hemodilution)
It is important for clinicians utilizing cerebral oximetry
• Hyperventilation
to understand factors that influence the values dis-
• Acid–base
played on the monitors. Alterations in the emitter-to-
disequilibrium
sensor distance by tissue edema and changes in the
arterial and venous hemoglobin distribution (e.g., by
arterio-venous shunts, hematomas, and hemodilution)
do not necessarily affect tissue oxygenation but can content. Oxygen content is determined by the hemo-
impact oximeter readings. Motion artifact, jaundice, globin concentration, oxygen saturation, and, to a
and hair pigmentation can all cause decreased oximeter lesser degree, oxygen tension (which depends on the
accuracy; however, skin color and melanin content do inspired oxygen concentration and the alveolar-arterial
not appear to have an effect, unlike in pulse oximetry. gradient). Common perioperative factors associated
In order to establish a reference point for measure- with decreased rScO2 values that often are of great
ments throughout the procedure, it is recommended concern are low cardiac output, hypoxemia, anemia
that baseline oximetry values be determined in the (e.g., secondary to hemorrhage or hemodilution),
awake, spontaneously breathing patient. Low baseline hyperventilation that causes cerebral vasoconstriction,
oximetry values are commonly observed in patients extreme perturbations in acid–base balance, increased
with brain atrophy, diastolic dysfunction, poor left ven- oxygen demand (e.g., under-anesthetized patients,
tricular function, anemia, and hemodialysis,14–18 and hyperthermia, convulsions), and circulatory collapse
they are associated with higher incidences of morbidity, (Box 3.1).4,7,24 In cardiovascular surgery specifically,
including postoperative delirium and mortality.4,19–21 unilateral and bilateral cerebral desaturations occur
Baseline rScO2 values, however, have significant inter- with catastrophic events such as acute dissection, vas-
individual variation and can range from 50% to 80% in cular compression, inadequate anterograde cerebral
healthy patients.8,10,22 Thus, trending values may be perfusion, and cannula misplacement. By contrast,
more indicative of the ensuing cerebral supply–demand common perioperative factors associated with
balance than absolute values. Additionally, cerebral increased rScO2 values are usually of lesser concern
desaturation load (CDL), defined as the product of the and include hypoventilation that causes cerebral vaso-
duration and severity of desaturations, has been inves- dilation, hypertension, and hypothermia-associated
tigated as a trigger for intervention.4,23 decreased oxygen extraction (Box 3.2).
Cerebral oximetry data reflect the balance between As rScO2 perturbations often reflect a mismatch
regional tissue oxygen supply and demand. Factors between cerebral oxygen supply and demand, they
that affect this balance, therefore, will contribute to warrant a thorough and systematic evaluation. Once
alterations in the displayed rScO2 value. Supply is a possible etiology has been identified, interventions
directly influenced by cerebral blood flow (CBF) that should be discussed and considered with all involved
is autoregulated, cardiac output, and arterial oxygen team members. Deschamps et al.4 proposed a
22
Cerebral Oximetry
Cerebral desaturation
Head position Unilateral Reposition or replace

Verify probe placement
*Other probe
Bilateral
Cerebral desaturation
SV
Optimize cardiac
output
HR
Mean arterial pressure Low MAP
Assess systemic
vascular resistance
Normal MAP
Low SaO2/
PaO2
SaO2/PaO2 Treat and find etiology
SaO2/PaO2 Normal
Low PaCO2
PaCO2 Correct hyperventilation
Normal PaCO2
Convulsions
Cerebral O2 Sepsis Treat
consumption Hyperthermia accordingly
Inadequate
anesthesia
Normal O2
consumption
Low SvO2
Venous O2 saturation Optimize oxygen
consumption and delivery
Normal SvO2
Low
hematocrit
Hemoglobin Consider transfusion
concentration
Proposed algorithm for the treatment fo cerebral desaturations

* Correlate Clinically: etiology can include cannula malposition, venous obstruction, dissection, and one-
lung ventilation.
Figure 3.6 Proposed algorithm for the treatment of cerebral desaturations.

Cerebral Oximetry
systematic approach to the correction of cerebral cardiac surgery, as it showed no evidence of reduction
desaturations in patients undergoing high-risk car- in neurocognitive dysfunction or biomarkers of renal,
diac surgery. A modified version of this intervention myocardial, and neurologic injury.33
protocol is presented here (Figure 3.6).
Carotid Endarterectomy
Near-Infrared Spectroscopy in Clinical The rate of perioperative stroke associated with carotid
Practice endarterectomy (CEA) ranges from 1% to 6%.34
The relationship between cerebral tissue oxygena- Intraoperative hypoperfusion is one of the major causes
tion and clinical outcomes has been most studied of this devastating complication. Transcranial Doppler
(TCD), electroencephalography (EEG), and NIRS have
in patients undergoing cardiac surgery due to con-
all been utilized to minimize perioperative hypoperfu-
cerns regarding end-organ perfusion while on car-
sion, as well as assist in the decision whether to place a
diopulmonary bypass (CPB).25 As NIRS technology
carotid shunt. While EEG and TCD can be cumbersome
has become more sophisticated, its clinical use has
and intrusive, NIRS offers clinicians a simple, afford-
expanded into other surgical settings such as vas-
able, and noninvasive method of monitoring bilateral
cular, orthopedic, and general surgery, as well as
nonsurgical settings such as management of cere- cerebral supply during CEA. When compared to EEG
bral autoregulation, traumatic brain injury (TBI), and TCD, NIRS has been shown to reliably predict
changes in intraoperative cerebral supply during carotid
and cardiopulmonary resuscitation.
cross-clamping.35,36 Several studies, however, failed to
identify reliable rScO2 cutoff values that provide reason-
Cardiopulmonary Bypass able sensitivity and specificity for the onset of ischemia;
Cardiac surgery can result in significant complications, some authors proposed that a decrease of 20% from
including death, stroke, delirium, and respiratory fail- baseline indicates the need for intervention.35,37 It there-
ure. Several studies have shown associations between fore remains unclear whether cerebral oximetry alone
cerebral oxygen saturation monitoring and these can serve as a reliable clinical monitor during CEA. One
events.17,20,24,26 As previously mentioned, in patients must also consider that a false-positive identification of
undergoing cardiac surgery requiring CPB, low base- ischemia with rScO2 monitoring may lead to unneces-
line and intraoperative rScO2 values are associated with sary treatment with associated risk.11,38 No well-
higher rates of mortality and other adverse outcomes, powered studies have examined the association between
including cardiopulmonary dysfunction.17,20 NIRS data and stroke in patients undergoing CEA.25 If
Studies performed in patients undergoing CABG cerebral oximetry were shown to have similar predictive
surgery suggest that an algorithmic approach to inter- values as other modalities, its use would be invaluable.
ventions to increase cerebral oxygenation can result in
improved outcomes, including decreased incidence of Orthopedic Surgery
stroke resulting in permanent loss of function,27 bet- Patients undergoing orthopedic surgery in the beach
ter cognitive outcomes,28,29 and less injury to other chair position are more likely to have blunted cerebral
organs.30 A retrospective, non-blinded study con- autoregulation that can lead to rare but devastating
ducted in patients undergoing aortic arch surgery neurologic complications.39 The use of cerebral oxi-
with anterograde cerebral perfusion demonstrated metry as a monitor of adequate cerebral perfusion has
that cerebral oximetry desaturations predicted perio- been examined in beach chair positioning.40,41 Fischer
perative neurological complications.31 et al. reported that decreases in blood pressure during
By contrast, several reviews have found no benefit beach chair positioning were accompanied by epi-
of NIRS monitoring in the perioperative setting. sodes of cerebral oxygen desaturation. These desa-
Serraino et al. conducted a systematic review and turations resolved with the administration of
meta-analysis involving almost 1,500 adult patients phenylephrine and increases in the mean arterial
and found that the use of cerebral oximetry did not pressure (MAP). The authors proposed that cerebral
reduce brain, heart, or kidney injury during CPB.32 A oximetry monitoring may be of value in the manage-
multicenter randomized controlled trial (PASPORT) ment of patients undergoing surgery in the beach
did not support the use of an oximetry-based algo- chair position, in order to determine the adequacy of
rithm for optimization of tissue perfusion in adult blood pressure at the level of the brain.40 Murphy et al.
24
Cerebral Oximetry
also reported significant reductions of cerebral oxyge- data have suggested that the lower limit of CBF auto-
nation in patients undergoing surgery in the beach regulation is in the range of 45–80 mmHg.45
chair position compared to the lateral decubitus posi- Autoregulation is also affected by factors other than
tion.41 In a literature review, Salazar et al. concluded the MAP, including intracranial pressure, partial
that accurate intraoperative cerebral perfusion moni- pressure of carbon dioxide, as well as a variety of
toring by cerebral oximetry or other methods (invasive brain pathologies, including traumatic injury,
blood pressure monitoring at the brain level, EEG), as tumor, stroke, ruptured cerebral aneurysm, and
well as alternatives to general anesthesia and judicious ischemic cerebrovascular disease.46 Even in patients
use of intraoperative blood pressure control, may with neurologic disease, there is still an optimal blood
improve patient safety in this patient population.42 pressure range in which autoregulatory function is
intact.47 By identifying an individual’s autoregulatory
General Surgery range, we can provide targeted and patient-specific
The number of elderly patients undergoing surgical management in a variety of perioperative settings.
procedures requiring general anesthesia has dramati- It has been suggested that clinicians can generate
cally increased over time. Because this patient popula- an index of autoregulation, known as the cerebral
tion has lower physiologic reserve and more oximetry index (COx),45,47–49 by plotting cerebral
comorbidities, their risk for postoperative complica- oxygen saturation over blood pressure. More specifi-
tions is significant. Although the brain is particularly cally, the COx is a continuous, moving Pearson’s
at risk in geriatric surgical patients, it is rarely mon- correlation coefficient to measure the strength of
itored. Investigations have sought to decrease the risk association between MAP and cerebral oximetry
of post-procedural neurocognitive decline by monitor- values.45 A value that approaches zero indicates CBF
ing cerebral supply and demand.43 Thus, the associa- autoregulation, while values closer to 1 indicate dys-
tion between cerebral oximetry monitoring and other regulation. Several authors have concluded that loss of
outcomes has also been investigated. In a study by autoregulation due to hypotension can be accurately
Casati et al., 122 elderly patients undergoing major assessed by the COx for patients with acute brain
abdominal surgery were randomized into a control injury and those undergoing CPB.45,48,50 The COx
group in which the anesthesiologist was blinded to potentially has value in traditional intraoperative set-
the rScO2 and an intervention group with their rScO2 tings and postoperative critical care settings, but
values visible and maintained above 75%. They con- further investigation is needed.
cluded that low cerebral saturations were associated
with longer post-anesthesia care unit stays and an Postoperative Cognitive Decline
increased rate of hospital readmissions.43
Many non-neurosurgical procedures are associated
Furthermore, the maintenance of cerebral saturations
with a risk for POCD, particularly cardiac surgery.
equal to or above 75% was associated with a significant
Cerebral hypoperfusion is a major risk factor for
reduction in hospital length of stay. In a prospective
POCD;51,52 therefore, as a noninvasive monitor of
observational study, Casati et al. again demonstrated
cerebral supply, NIRS has the potential to be valuable
similar results, confirming an association between cer-
in guiding therapy and possibly preventing POCD. A
ebral desaturation in the elderly undergoing abdominal
number of studies have found an association between
surgery, postoperative cognitive decline (POCD), and a
cerebral desaturation events and POCD in cardiac
longer hospital stay.44 While these findings are
surgery patients. In an observational study by Yao et
encouraging, larger prospective randomized investiga-
al., multivariate data analysis suggested that time
tions are needed. Given that these mostly observational
spent at cerebral saturations less than 40% predicted
studies were performed more than a decade ago, how-
cognitive deficits.53 Several smaller studies have
ever, there seems to be waning enthusiasm.
demonstrated that low rScO2 during CPB was asso-
ciated with POCD54 and that maintenance with a
Cerebral Autoregulation higher rScO2 during CPB was associated with a
Cerebral autoregulation is typically maintained lower incidence of POCD.55 In contrast, Hong et al.
between MAPs of 50–150 mmHg. The specific MAP found that low intraoperative cerebral saturations
threshold needed to ensure constant CBF for any were unrelated to POCD in cardiac surgery patients.56
individual patient, however, is unknown, and some Furthermore, a review of six studies of 962 patients in
Cerebral Oximetry
total presented moderate-quality evidence that NIRS practice, and cerebral oximetry is no exception. While
monitoring is not associated with decreased incidence many technologies measure variables heretofore
of POCD at one week post-surgery.57 The use of impossible to obtain, or may noninvasively measure
perioperative cerebral NIRS monitoring to decrease variables that previously required interventional or
the occurrence of POCD, therefore, is not supported impractical complex methods, the standard for accep-
by the existing evidence. tance into practice is increasingly an improved out-
come driven by use of the technology. Routine NIRS
Cardiopulmonary Resuscitation use remains controversial despite a significant market
penetrance in cardiac surgery and increasing use in
Advanced cardiac life support (ACLS) is often deliv-
other procedures (e.g., carotid endarterectomy).
ered without 100% compliance to guidelines.
While some studies suggest that NIRS monitoring
Inadequate rate and depth of chest compressions, com-
and NIRS-guided interventions may decrease cerebral
bined with errors in medication administration, con-
injury and improve neurological outcomes, others do
tinue to be problems, despite extensively developed
not support its ubiquitous use and officially the “jury
evidence-based guidelines and considerable provider
is still out.”
education and training. Methods to improve the qual-
ity of ACLS are continually being investigated in an
effort to improve patient outcomes.58 It has been pro- References
posed that cerebral oximetry may be used to monitor 1. Torres J, Ishida K. Neuroprotection after major
ACLS adequacy.59,60 This field is in its infancy, but cardiovascular surgery. Curr Treat Options Neurol
investigators have demonstrated that higher cerebral 2015;17(7):28.
oximetry saturations are associated with better neuro- 2. Costa MA, Gauer MF, Gomes RZ, Schafranski MD.
logic outcomes, and that mean rScO2 values over the Risk factors for perioperative ischemic stroke in
course of the ACLS are more strongly correlated with cardiac surgery. Rev Bras Cir Cardiovasc 2015;30
the return of spontaneous circulation than the initial (3):365–72.
rScO2 values.61,62 It has been suggested that a sustained 3. Andersen ND, Hart SA, Devendra GP, et al.
inability to achieve cerebral saturations greater than Atheromatous disease of the aorta and perioperative
stroke. J Thorac Cardiovasc Surg 2018;155(2):508–16.
30% could be included in the criteria for ceasing
ACLS efforts.62 There is no evidence, however, that 4. Deschamps A, Hall R, Grocott H, et al. Cerebral
the cerebral saturation data can be used to successfully oximetry monitoring to maintain normal cerebral
oxygen saturation during high-risk cardiac surgery.
guide ACLS, as the studies are primarily observational.
Anesthesiology 2016;124(4):826–36.
5. Nielsen HB. Systematic review of near-infrared
Traumatic Brain Injury spectroscopy determined cerebral oxygenation during
Near-infrared spectroscopy has been used to predict non-cardiac surgery. Front Physiol 2014;5:93.
and assess the impact of goal-directed treatment in 6. Jöbsis FF. Noninvasive, infrared monitoring of
patients with TBI.63 Rosenthal et al. compared jugular cerebral and myocardial oxygen sufficiency and
bulb venous measurements with NIRS in an attempt circulatory parameters. Science 1977;198(4323):1264–
7.
to provide an accurate estimation of cerebral oxyge-
nation. They concluded that measuring regional cere- 7. Fischer GW, Silvay G. Cerebral oximetry in cardiac
bral tissue oxygenation by NIRS was feasible in and major vascular surgery. HSR Proc Intensive Care
Cardiovasc Anesth 2010;2(4):249–56.
patients with severe TBI and could provide an estima-
tion of cerebral oxygenation status in a noninvasive 8. Green DW, Kunst G. Cerebral oximetry and its role in
manner.63 There is a need for larger-scale studies adult cardiac, non-cardiac surgery and resuscitation
from cardiac arrest. Anaesthesia 2017;72:48–57.
examining the utility of NIRS monitoring to predict
clinical outcomes in the complex physiologic state of 9. Smythe PR, Samra SK. Monitors of cerebral
oxygenation. Anesthesiol Clin North Am 2002;20
TBI.25,64
(2):293–313.
10. Tosh W, Patteril M. Cerebral oximetry. BJA Educ
Conclusion 2016;16(12):417–21.
Novel medical technologies need to be investigated 11. Steppan J, Hogue CW Jr. Cerebral and tissue oximetry.
and scrutinized prior to their acceptance into clinical Best Pract Res Clin Anaesthesiol 2014;28(4):429–39.
26
Cerebral Oximetry
12. Ghosh A, Elwell C, Smith M. Cerebral near-infrared 26. Fischer GW, Lin HM, Krol M, et al. Noninvasive
spectroscopy in adults. Anesth Analg 2012;115 cerebral oxygenation may predict outcome in patients
(6):1373–83. undergoing aortic arch surgery. J Thorac Cardiovasc
13. Davie SN, Grocott HP. Impact of extracranial Surg 2011;141(3):815–21.
contamination on regional cerebral oxygen saturation: 27. Goldman S, Sutter F, Ferdinand F, Trace C.
a comparison of three cerebral oximetry technologies. Optimizing intraoperative cerebral oxygen delivery
Anesthesiology 2012;116(4):834–40. using noninvasive cerebral oximetry decreases the
14. Kobayashi K, Kitamura T, Kohira S, et al. Factors incidence of stroke for cardiac surgical patients. Heart
associated with a low initial cerebral oxygen saturation Surg Forum 2004;7(5):E376–81.
value in patients undergoing cardiac surgery. J Artif 28. Slater JP, Guarino T, Stack J, et al. Cerebral oxygen
Organs 2017;20(2):110–6. desaturation predicts cognitive decline and longer
15. Madsen PL, Nielsen HB, Christiansen P. Well-being hospital stay after cardiac surgery. Ann Thorac Surg
and cerebral oxygen saturation during acute heart 2009;87(1):36–45.
failure in humans. Clin Physiol 2000;20(2):158–64. 29. Colak Z, Borojevic M, Bogovic A, Ivancan V, Biocina
16. Paquet C, Deschamps A, Denault AY, et al. Baseline B, Majeric-Kogler V. Influence of intraoperative
regional cerebral oxygen saturation correlates with left cerebral oximetry monitoring on neurocognitive
ventricular systolic and diastolic function. J function after coronary artery bypass surgery: a
Cardiothorac Vasc Anesth 2008;22(6):840–6. randomized, prospective study. Eur J Cardiothoracic
Surg 2015;47(3):447–54.
17. Sun X, Ellis J, Corso PJ, et al. Mortality predicted by
preinduction cerebral oxygen saturation after cardiac 30. Murkin JM, Adams SJ, Novick RJ, et al. Monitoring
operation. Ann Thorac Surg 2014;98:91–6. brain oxygen saturation during coronary bypass
surgery: a randomized, prospective study. Anesth
18. Wen SYB, Peng AZY, Boyle S, et al. A pilot study using Analg 2007;104(1):51–8.
preoperative cerebral tissue oxygen saturation to
stratify cardiovascular risk in major non-cardiac 31. Olsson C, Thelin S. Regional cerebral saturation
surgery. Anaesth Intensive Care 2017;45(2):202–9. monitoring with near-infrared spectroscopy
during selective antegrade cerebral perfusion:
19. Lei L, Katznelson R, Fedorko L, et al. Cerebral oximetry diagnostic performance and relationship to
and postoperative delirium after cardiac surgery: a postoperative stroke. J Thorac Cardiovasc Surg
randomised, controlled trial. Anaesthesia 2017;72 2006;131(2):371–9.
(12):1456–66.
32. Serraino GF, Murphy GJ. Effects of cerebral near-
20. Heringlake M, Garbers C, Käbler JH, et al. infrared spectroscopy on the outcome of patients
Preoperative cerebral oxygen saturation and clinical undergoing cardiac surgery: a systematic review of
outcomes in cardiac surgery. Anesthesiology 2011;114 randomised trials. BMJ Open 2017;7(9):e016613.
(1):58–69.
33. Rogers CA, Stoica S, Ellis L, et al. Randomized trial of
21. Ghosal S, Trivedi J, Chen J, et al. Regional cerebral near-infrared spectroscopy for personalized
oxygen saturation level predicts 30-day mortality rate optimization of cerebral tissue oxygenation during
after left ventricular assist device surgery. J cardiac surgery. Br J Anaesth 2017;119(3):384–93.
Cardiothorac Vasc Anesth 2018;32(3):1185–90.
34. Wu TY, Anderson NE, Barber PA. Neurological
22. Thavasothy M, Broadhead M, Elwell C, Peters M, complications of carotid revascularisation. J Neurol
Smith M. A comparison of cerebral oxygenation as Neurosurg Psychiatry 2012;83(5):543–50.
measured by the NIRO 300 and the INVOS 5100 near-
infrared spectrophotometers. Anaesthesia 2002;57 35. Cho JW, Jang JS. Near-infrared spectroscopy versus
(10):999–1006. transcranial Doppler-based monitoring in carotid
endarterectomy. Korean J Thorac Cardiovasc Surg
23. Deschamps A, Lambert J, Couture P, et al. Reversal of 2017;50(6):448–52.
decreases in cerebral saturation in high-risk cardiac
surgery. J Cardiothorac Vasc Anesth 2013;27(6):1260– 36. Perez W, Dukatz C, El-Dalati S, et al. Cerebral
6. oxygenation and processed EEG response to clamping
and shunting during carotid endarterectomy under
24. Vretzakis G, Georgopoulou S, Stamoulis K, et al. general anesthesia. J Clin Monit Comput 2015;29
Cerebral oximetry in cardiac anesthesia. J Thorac Dis (6):713–20.
2014;6 (Suppl 1):S60–9.
37. Mauermann WJ, Crepeau AZ, Pulido JN, et al.
25. Bickler P, Feiner J, Rollins M, Meng L. Tissue oximetry Comparison of electroencephalography and cerebral
and clinical outcomes. Anesth Analg 2017;124(1):72– oximetry to determine the need for in-line arterial
82. shunting in patients undergoing carotid
Cerebral Oximetry
endarterectomy. J Cardiothorac Vasc Anesth 2013;27 49. Steiner LA, Pfister D, Strebel SP, Radolovich D,
(6):1253–9. Smielewski P, Czosnyka M. Near-infrared
38. Pennekamp CWA, Bots ML, Kappelle LJ, Moll FL, de spectroscopy can monitor dynamic cerebral
Borst GJ. The value of near-infrared spectroscopy autoregulation in adults. Neurocrit Care 2009;10
measured cerebral oximetry during carotid (1):122–8.
endarterectomy in perioperative stroke prevention. A 50. Brady KM, Lee JK, Kibler KK, et al. Continuous time-
review. Eur J Vasc Endovasc Surg 2009;38(5):539–45. domain analysis of cerebrovascular autoregulation
39. Laflam A, Joshi B, Brady K, et al. Shoulder surgery in using near-infrared spectroscopy. Stroke 2007;38
the beach chair position is associated with diminished (10):2818–25.
cerebral autoregulation but no differences in 51. Pappa M, Theodosiadis N, Tsounis A, Sarafis P.
postoperative cognition or brain injury biomarker Pathogenesis and treatment of post-operative
levels compared with supine positioning: the cognitive dysfunction. Electron Physician 2017;9
anesthesia patient safety foundation beach chair study. (2):3768–75.
Anesth Analg 2015;120(1):176–85. 52. Messerotti Benvenuti S, Zanatta P, Longo C,
40. Fischer GW, Torrillo TM, Weiner MM, Rosenblatt Mazzarolo AP, Palomba D. Preoperative cerebral
MA. The use of cerebral oximetry as a monitor of the hypoperfusion in the left, not in the right, hemisphere
adequacy of cerebral perfusion in a patient undergoing is associated with cognitive decline after cardiac
shoulder surgery in the beach chair position. Pain Pract surgery. Psychosom Med 2012;74(1):73–80.
2009;9(4):304–7. 53. Yao FS, Tseng CC, Ho CY, Levin SK, Illner P. Cerebral
41. Murphy GS, Szokol JW, Marymont JH, et al. Cerebral oxygen desaturation is associated with early
oxygen desaturation events assessed by near-infrared postoperative neuropsychological dysfunction in
spectroscopy during shoulder arthroscopy in the beach patients undergoing cardiac surgery. J Cardiothorac
chair and lateral decubitus positions. Anesth Analg Vasc Anesth 2004;18(5):552–8.
2010;111(2):496–505. 54. Fudickar A, Peters S, Stapelfeldt C, et al. Postoperative
42. Salazar D, Hazel A, Tauchen AJ, Sears BW, Marra G. cognitive deficit after cardiopulmonary bypass with
Neurocognitive deficits and cerebral desaturation preserved cerebral oxygenation: a prospective
during shoulder arthroscopy with patient in beach- observational pilot study. BMC Anesthesiol 2011;11
chair position: a review of the current literature. Am J (1):7.
Orthop (Belle Mead NJ) 2016;45(3):E63–8. 55. Colak Z, Borojevic M, Bogovic A, Ivancan V, Biocina
43. Casati A, Fanelli G, Pietropaoli P, et al. Continuous B, Majeric-Kogler V. Influence of intraoperative
monitoring of cerebral oxygen saturation in elderly cerebral oximetry monitoring on neurocognitive
patients undergoing major abdominal surgery function after coronary artery bypass surgery: a
minimizes brain exposure to potential hypoxia. randomized, prospective study. Eur J Cardiothoracic
Anesth Analg 2005;101(3):740–7. Surg 2015;47(3):447–54.
44. Casati A, Fanelli G, Pietropaoli P, et al. Monitoring 56. Hong SW, Shim JK, Choi YS, Kim DH, Chang BC,
cerebral oxygen saturation in elderly patients Kwak YL. Prediction of cognitive dysfunction and
undergoing general abdominal surgery: a patients’ outcome following valvular heart surgery and
prospective cohort study. Eur J Anaesthesiol 2007;24 the role of cerebral oximetry. Eur J Cardiothoracic Surg
(1):59. 2008;33(4):560–5.
45. Brady K, Joshi B, Zweifel C, et al. Real-time continuous 57. Yu Y, Zhang K, Zhang L, Zong H, Meng L, Han R.
monitoring of cerebral blood flow autoregulation Cerebral near-infrared spectroscopy (NIRS) for
using near-infrared spectroscopy in patients perioperative monitoring of brain oxygenation in
undergoing cardiopulmonary bypass. Stroke 2010;41 children and adults. Cochrane Database Syst Rev
(9):1951–6. 2018;1:CD010947.
46. Tameem A, Krovvidi H. Cerebral physiology. Contin 58. Sutton RM, Nadkarni V, Abella BS. Putting it all
Educ Anaesth Crit Care Pain 2013;13(4):113–8. together to improve resuscitation quality. Emerg Med
47. Moerman A, De Hert S. Recent advances in cerebral Clin North Am 2012;30(1):105–22.
oximetry. Assessment of cerebral autoregulation with 59. Genbrugge C, Dens J, Meex I, et al. Regional
near-infrared spectroscopy: myth or reality? cerebral oximetry during cardiopulmonary
F1000Research 2017;6:1615. resuscitation: useful or useless? J Emerg Med
48. Joshi B, Ono M, Brown C, et al. Predicting the limits of 2016;50(1):198–207.
cerebral autoregulation during cardiopulmonary 60. Ibrahim AW, Trammell AR, Austin H, et al. Cerebral
bypass. Anesth Analg 2012;114(3):503–10. oximetry as a real-time monitoring tool to assess
28
Cerebral Oximetry
quality of in-hospital cardiopulmonary resuscitation and meta-analysis. Acad Emerg Med 2016;23(8):851–
and post cardiac arrest care. J Am Heart Assoc 2015;4 62.
(8):e001859. 63. Rosenthal G, Furmanov A, Itshayek E, Shoshan Y,
61. Sinha N, Parnia S. Monitoring the brain after cardiac Singh V. Assessment of a noninvasive cerebral
arrest: a new era. Curr Neurol Neurosci Rep 2017;17 oxygenation monitor in patients with severe traumatic
(8):62. brain injury. J Neurosurg 2014;120(4):901–7.
62. Cournoyer A, Iseppon M, Chauny JM, Denault A, 64. Weigl W, Milej D, Janusek D, et al. Application of optical
Cossette S, Notebaert É. Near-infrared spectroscopy methods in the monitoring of traumatic brain injury: A
monitoring during cardiac arrest: a systematic review review. J Cereb Blood Flow Metab 2016;36(11):1825–43.
Chapter
The Oxygen Reserve Index
4 Andrew B. Leibowitz
The introduction of pulse oximetry into operating methemoglobin (SpMet®), and carboxyhemoglobin
room and intensive care units in the early 1980s was (SpCO®).
revolutionary. In 1986 pulse oximetry became the Pulse oximetry, however, has limitations: (1) its
American Society of Anesthesiologists formal stan- accuracy is ± 1.5–4% compared to arterial oxygen
dard for intra-operative monitoring,1 and in 1988 it saturation (SaO2) measured by a CO oximeter utiliz-
was referred to as the “fifth vital sign.”2 Real-time ing multiwavelength spectrophotometry of a blood
continuous noninvasive measurement of the pulse sample;4,5 (2) an SpO2 ≥ 98% may represent arterial
oxygen saturation (SpO2) is now so ingrained in our oxygen tension (PaO2) between 75 mmHg and
practice that it is hard to imagine an era when physi- 650 mmHg (e.g., on a FiO2 of 1.0), so decreases in
cians relied on patient skin color to assess the ade- PaO2 from 650 mmHg to 75 mmHg may go unde-
quacy of oxygenation. Intuitively, SpO2 was so useful tected while the SpO2 remains in a reassuring ≥ 94%–
that even when a large study published in the journal 98% range; (3) the relationship between PaO2 and
Anesthesiology reported no difference in outcomes in SaO2 is not linear but sigmoidal, thus when the
over 20,000 surgical patients randomly assigned to PaO2 is 60 mmHg and the SpO2 is 90% at the inflec-
monitoring with versus without oximetry, it did not tion point of the O2 hemoglobin dissociation curve,
deter its widespread acceptance.3 the PaO2 and SpO2 may decline rapidly without a
In brief review, pulse oximetry relies on light- clinically predictable nadir (see Figure 4.1); and (4)
emitting diodes (LEDs) that emit at least two wave- it is difficult to predict the time to significant desatura-
lengths of light, typically a red wavelength of 660 nm tion to SpO2 less than 90% during apnea or rapid
and an infrared wavelength of 905 or 940 nm. These changes in ventilation and perfusion, particularly in
wavelengths of light are absorbed to different degrees certain high-risk patients.
by oxygenated (oxyhemoglobin) and deoxygenated Use of current monitoring technology to help
(reduced hemoglobin) hemoglobin. A photodiode avoid desaturation has cardiopulmonary physiologic
used to measure the amount of absorbed light is limitations too. In the operating room, measurement
positioned opposite the LEDs, usually on the other of end tidal expired O2 is used during pre-oxygena-
side of a finger or earlobe, and resultant light that tion to assess denitrogenation of the functional resi-
reaches it allows for the calculation of the ratio of dual volume and maximize O2 reserve in order to
the two absorbed wavelengths. This ratio is then con- increase the time to desaturation during apnea. The
verted to SpO2 using a lookup table created by mea- success of a strategy aimed toward optimizing end
suring the ratios in volunteers whose saturations were tidal expired O2, however, may be altered by reduced
altered from 100% to 70% by breathing increasingly functional residual capacity (FRC), increased closing
hypoxic gas mixtures. Pulse oximeters are designed to capacity-FRC ratio, increased O2 consumption, and
measure the absorbance of these wavelengths in the decreased O2-carrying capacity, each of which inde-
pulsatile arterial blood by filtering out the signal con- pendently changes the rate of decline in the PaO2 and
tributed by venous blood and other tissues, and in SpO2 during apnea and thereby complicates the accu-
many models motion artifact. Additional wavelengths rate estimation of time to desaturation. These pertur-
of light may also be incorporated, in order to non- bations occur quite frequently in obese patients, as
invasively and continuously determine the percent well as in emergency intubations and critically ill
concentration of total hemoglobin (SpHb®), patients where continuous changes in ventilation
30
curve in the PaO2 of 100–200 mmHg range provides a

PVO2 PaO2
detectable change in absorbance signals utilizing
100 other wavelengths of light in addition to the standard
660 and 905 or 940 nm. Currently, above that range,
Hemoglobin % saturation
80 no usable absorbance signals have been detected.

Masimo pulse CO oximeters use additional wave-
60 lengths of light and advanced signal processing to
collect optical absorbance information that enables
40 them to resolve extremely small differences in absor-
bance in the PaO2 100–200 mmHg range. The signal is
20 then processed and expressed as a unitless ORi value
from 0.0 (no reserve) to 1.0 (good or “much” reserve).
In the original publication in 2014 the ORi scale was
0 20 40 60 80 100 120 140 100–200, reflecting the original focus on the utility of
Partial pressure O2 (mmHg) this new index associated with changes in this PaO2
range.9
Figure 4.1 Oxygen hemoglobin saturation curve highlighting the
typical mixed venous partial pressure of oxygen (PvO2) and arterial Devices with this functionality do not have FDA
oxygen tension (PaO2) ranges. 510(k) clearance and are not available for purchase in
the United States, but they are commercially available
and perfusion due to extreme and acute respiratory in Japan and countries that require CE marking. In a
and circulatory compromise are more common.6,7 very recent study, 20 healthy volunteers breathing gas
Not only do they experience greater risk of rapid mixtures ranging from hypoxemic to hyperoxic had
decline in PaO2 and SpO2, but they also have an their ORi and PaO2 simultaneously measured, yield-
associated increased risk of morbidity and mortality.8 ing 1,090 paired data points.10 The ORi trended the
Children are also at greater risk of rapid decline in PaO2 within the 100–200 mmHg range with a con-
PaO2 and SpO2 during apnea because their relative cordance of 94%, and the ORi predicted a PaO2 less
consumption-reserve ratio is greater than that of than 100 mmHg with an area under the curve of 0.91.
adults. While avoidance of hypoxemia has been a domi-
A warning that hypoxemia is imminent would nant clinical focus since the advent of arterial blood
potentially be a useful addition to standard pulse gas analysis, the ORi potentially may also be used to
oximetry and capnometry. The association between prevent hyperoxia without the need for arterial
PaO2 and SpO2 decreases and increased morbidity blood gas analysis and reduce the exposure to harm-
and mortality during emergency intubations suggests ful oxygen supplementation, a long-recognized risk
that the ability to predict the time to desaturation to a in newborns and more recently appreciated risk in
hypoxemic range may result in better real-time man- post-resuscitative states.
agement and improved outcomes. This capability Much of the attention to this technology’s poten-
would complement the myriad functionality that has tial has focused on the time taken for patients to
been engineered into pulse oximeters since their desaturate to an SpO2 94%–98%, versus the time to
inception, such as reduction of motion artifact, assess- the start of an ORi alarm triggered by either a decrease
ment of pulse pressure variation, detection of dyshe- to an absolute value or a rate of change, or both. The
moglobins, and miniaturization. difference in time between the actual desaturation
Newer technology aims to overcome the limita- versus the alarm preceding the desaturation is
tions of pulse oximetry outlined above by measure- referred to by Masimo as the “increase in the warning
ment of an oxygen reserve index (ORi). The ORi is time,” but alternatively it might be considered as a
determined from the very small changes to the clinical “window of opportunity” to intervene.
absorption pattern of blood in states where the SpO2 Szmuk et al. tested the hypothesis that the ORi
is ≥ 98%. While the ratio of the difference in absor- provides warning of impending desaturation during
bance to the traditional red and infrared wavelengths induction of anesthesia.11 Twenty-five pediatric sur-
decreases as the oxygen saturation reaches 100%, the gical patients aged 7.6 ± 4.6 years with the American
moderate hyperoxic region of the oxygen dissociation Society of Anesthesiologists physical status I and II
Figure 4.2 The ORi and SpO2 values at
ea
9 0 Ox
t
m
r
ta
to
%
%
%
%
different times of study. I-Start = beginning
ar
d
e-
I-S
I-S
Ap
En
98
90
98
Oxygen Reserve Index
Al
R
1.0 of intubation; I-Stop = end of intubation;
0.8
Alarm = start of the oxygen reserve alarm;
Re-Ox = reoxygenation; End = end of
0.6 recording. The points and error bars
0.4 represent mean (SD) values. Reproduced
0.2 with permission from Szmuk P, Steiner JW,
Olomu PN, Ploski RP, Sessler DI, Ezri T.
0.0 100
Oxygen reserve index: A novel noninvasive
SpO2 [%]
95 measure of oxygen reserve – A pilot study.
Anesthesiology 2016;124(4):779–84.
90
85
2 4 6 8 10 12 14
Elapsed time since start of induction [min]
1.0
0.8
Oxygen Reserve Index
0.6
0.4
0.2
0.0
50 100 150 200 240

PaO2 mmHg
Figure 4.3 Plot of ORi compared with arterial PaO2 obtained from 106 patients undergoing surgery in whom measured PaO2 from 485
arterial blood gas analyses was between 62 and 534 mmHg. Patients had more than one sensor applied, with analysis done using 1,594 ORi
values. Locally weighted regression analysis showed a nonlinear relationship overall, with a more positive relationship for PaO2 up to
240 mmHg compared to > 240 mmHg.
Reproduced with permission from Applegate RL II, Dorotta IL, Wells B, Juma D, Applegate PM. The relationship between oxygen reserve index
and arterial partial pressure of oxygen during surgery. Anesth Analg 2016;123(3):626–33.
were pre-oxygenated, general anesthesia was induced, in a median of 31.5 seconds (interquartile range 19–
the trachea intubated, the anesthesia circuit was dis- 34.3) before the SpO2 decreased to 98%.
connected, and the oxygen saturation was allowed to Applegate et al. investigated the relationship
decrease to 90% before resuming ventilation. The ORi between the ORi and PaO2 during surgery in 106
slowly decreased over a mean apneic period of 5.9 ± patients who had simultaneous arterial blood gas
3.1 minutes, from 0.73 ± 0.16 to 0.37 ± 0.11, while the analysis and ORi monitoring, and found a correlation
SpO2 remained at 100% (see Figure 4.2). The ORi between the ORi and PaO2 of r2 = 0.546 (see Figure
alarm was triggered by its fractional rate of change 4.3).12 An ORi > 0.55 indicated a PaO2 ≥ 150 mmHg,
32
and an ORi > 0.24 indicated a PaO2 of about 100 studies performed thus far have occurred in these
mmHg. The findings also suggested that an ORi environments or evaluated patient outcomes.
decrease to 0.24 may provide an increase in the warn- Both the fundamental principle of detecting very
ing time when the SpO2 is greater than 98%. small changes in light absorption using multiple
In an abstract presented at the International wavelengths and the new advanced signal processing
Anesthesia Research Society (IARS) in 2017, Lee et underlying this technology are enticing. The clinical
al. studied 40 critically ill adult patients undergoing value of detecting such minimal changes, given the
elective surgery requiring tracheal intubation; 33 known limitations of SpO2 measurement (i.e., accu-
completed the study and had their data compositely rate to ± 1.5–4% in arterial blood), and using that
analyzed, excluding 4 outliers considered separately.13 change in a proprietary algorithm to estimate the
The time from the ORi alarm triggered by a “decrease PaO2 between 100 mmHg and 200 mmHg, requires
in the absolute value and rate of change” to the time more evidence, especially under real-world condi-
the SpO2 was 94% was compared to the time taken for tions. Common potential confounding factors
the SpO2 to decrease from 98% to 94%. The increase include:
in warning time was considered to be the time differ- (1) Anemia and polycythemia
ence between the ORi alarm start to the time when the
(2) Shifts in the oxygen–hemoglobin dissociation
SpO2 was 98%. The average time from the ORi alarm
curve due to acid base balance, temperature, and
until an SpO2 of 94% was 80 ± 38 seconds (range 29–
2,3-diphosphoglycerate (2,3-DPG) levels
227), while the average time for SpO2 to decrease from
(3) Increased cardiac output states
98% to 94% was 46 ± 23 seconds (range 12–108). The
(4) Increased oxygen consumption
increase in warning time was 34 ± 23 seconds (range
4–119), or on a percentage basis 96 ± 92% (range 5%– (5) Poor peripheral circulation impacting both SpO2
479%) and ORi measurement
These investigations demonstrate that in specific (6) Skin pigmentation and tissue absorption in the
circumstances the ORi may be used to increase the ORi range may be greater sources of error than in
warning time to impending desaturation or, at least, the typical SaO2 range
desaturation to SpO2 ≤ 98%. These findings, however, (7) Pre-oxygenation efforts that do not reliably
require further consideration of several points before increase the PaO2 to ≥ 200 mmHg (a common
adopting this technology into general practice. finding in intensive care patients) would
Assuming the investigations published thus far are predictably result in a shorter indeterminate
accurate in their findings, it is unclear whether a 30– warning time and void the potential benefit of this
40 second advanced warning that the SpO2 will early alarm
decrease to ≤ 98% is clinically advantageous. It is (8) The impact of dyshemoglobins on the absorption
even conceivable that such an advanced warning of these new wavelengths of light has not been
may be harmful. Interruption of tracheal intubation reported
(which has a success rate greater than 99%) in order to Routine use of this technology, however, may lead to a
address the alarm by resumption of mask ventilation change in customary practice once the warning time is
or other maneuvers may delay time to intubation, understood in the context of observed results. For
increase the risk of airway management failure, example, while the premise of pre-oxygenation was a
increase time spent in the hypoxemic zone, and known physiologic concept, after routine measure-
increase risk for trauma and aspiration. ment of end tidal oxygen became available, correct
Further investigation is necessary by a rando- application of a face mask and the time needed to
mized controlled trial comparing this technology ver- achieve optimal denitrogenation were widely appre-
sus standard pulse oximetry in patient groups with ciated, and even in the absence of the monitor these
high rates of hypoxemia complicating intubation, as is improved practices would contribute a safety margin.
the case for intubations done under emergent condi- Although studies have not demonstrated a relation-
tions and in intensive care units. Outcomes of interest ship between SpO2 monitoring and outcomes, simply
would be incidence and time of SpO2 less than 90%, using SpO2 monitors may have led to better manage-
hemodynamic instability, morbidity including airway ment in hundreds of millions of anesthetics once
injury and aspiration, and mortality. None of the previously unobserved phenomena were appreciated
and customary practice changed as a result. In 5. Chan ED, Chan MM, Chan MM. Pulse oximetry:
essence, using the pulse oximeter was like simulator understanding its basic principles facilitates
training that allowed persons to learn from near appreciation of its limitations. Respir Medicine
2013;107:789–99.
misses and incorporate changes in management that
prevented harm even when the monitor was not used. 6. Peppard PE, Ward NR, Morrell MJ. The impact of
In conclusion, the ORi is an interesting innovation obesity on oxygen desaturation during sleep-
disordered breathing. Am J Respir Crit Care Med
with the potential to improve patient care. The litera- 2009;180;788–93.
ture thus far is preliminary, with only a few small
7. Leibowitz AB. Persistent preoxygenation efforts before
studies that have varied in terms of patient popula-
tracheal intubation in the intensive care unit are of no
tions and alarm triggers (i.e., an absolute value versus use: who would have guessed? Crit Care Med 2009;37
a rate of change, or both) and no study of clinical (1):335–6.
outcomes. It is necessary to determine if this alarm
8. Jaber S, Amraoui J, Lefrant JY, et al. Clinical practice
function results in a reduced incidence or severity of and risk factors for immediate complication of
desaturation, particularly in high-risk patients. The endotracheal intubation in the intensive care unit: a
underlying physiologic assumptions may be limited prospective multiple-center study. Crit Care Med
by common pathologies, and this also requires inves- 2006;34;2355–61.
tigation. Further, the algorithm and technology 9. Szmuk P, Steiner JW, Olomu PN, Curuz JD, Sessler D.
require more detailed understanding in order for Oxygen reserve index – a new, noninvasive method of
clinicians to have confidence in a displayed ORi’s oxygen reserve measurement. October 14, 2014
utility. Nonetheless, development of the ORi is an American Society of Anesthesiologists Annual
advance and reveals that the limits of pulse oximeter Meeting, New Orleans, LA Abstract BOC12.
technology have not yet been reached, despite the fact 10. Vos JJ, Willems CH, Van Amsterdam K, et al. Oxygen
that they have been a bedside staple for more than 30 reserve index: validation of a new variable. Anesth
Analg 2019:129(s):409–15.
years.
11. Szmuk P, Steiner JW, Olomu PN, Ploski RP, Sessler DI,
Ezri T. Oxygen reserve index: a novel noninvasive
References measure of oxygen reserve – A pilot study.
1. American Society of Anesthesiologists: standards for Anesthesiology 2016;124(4):779–84.
basic intraoperative monitoring. Anesth Patient Safety 12. Applegate RL II, Dorotta IL, Wells B, Juma D,
Newslett 1987;2(1):1–8. Applegate PM. The relationship between oxygen
2. Neff TA. Routine oximetry: a fifth vital sign. Chest reserve index and arterial partial pressure of oxygen
1988;9:277. during surgery. Anesth Analg 2016;123(3):626–33.
3. Moller JT, Johannessen NW, Espersen H, et al. 13. Lee L, Singh A, Applegate R, Fleming N. Oxygen
Randomized evaluation of pulse oximetry in 20,802 Reserve Index: An Early Warning for Desaturation in
patients. Anesthesiology 1993;78:436–53. Critically Ill Patients. Proceedings from the 2017
4. Milner QJ, Mathews RG. An assessment of the accuracy IARS Annual Meeting, Washington, DC. Abstract
of pulse oximeters. Anaesthesia 2012;67:396–401. #A1406.
34
Chapter
Point-of-Care Transesophageal
5 Echocardiography
Ronald A. Kahn
Introduction Ultrasound waves are reflections when the

width of the reflecting object is larger than one-
Transesophageal echocardiography (TEE) was intro-
fourth of the ultrasound wavelength. To visualize
duced into anesthesiology practice in the late 1980s
smaller objects, ultrasound waves of shorter wave-
before becoming a standard in cardiac anesthesiol-
lengths must be used. Because the velocity of
ogy within the following decade. Slowly it has spread
sound in soft tissue is approximately constant,
outside of the cardiac operating rooms and is fre-
shorter wavelengths are obtained by increasing
quently used in other high-risk operations, in
the frequency of the ultrasound beam. The ultra-
patients with significant cardiac disease undergoing
sonic impedance of a structure (density times the
only moderate risk surgery, and in intensive care
ultrasound velocity through the structure) must be
units. All perioperative physicians, even those not
significantly different from the ultrasonic impe-
adept at personally performing TEE, should have a
dance of surrounding structures, otherwise ultra-
working knowledge of the TEE principles and
sound cannot distinguish the different structures.
practice.
The ultrasound impedances of air and bone are
significantly different from that of blood.
Technical Concepts Ultrasound is strongly reflected from air and
The complex physics of ultrasound may be bone, thereby limiting the ability of ultrasound to
reviewed in detail elsewhere, but the salient con- visualize structures deep to them.
cepts are as follows.1 An ultrasound beam is a Most modern echo scanners combine Doppler
continuous or intermittent train of sound waves capabilities with their two–dimensional (2D) ima-
emitted by a transducer or wave generator. ging capabilities. In pulsed-wave (PW) Doppler,
Ultrasound waves are characterized by their wave- blood flow parameters can be determined at precise
length, frequency, and velocity. Wavelength is the locations within the heart by emitting repetitive
distance between the two nearest points of equal short bursts of ultrasound at a specific frequency
pressure or density in an ultrasound beam, and (pulse repetition frequency, or PRF) and analyzing
velocity is the speed at which the waves propagate the frequency shift of the reflected echoes at an
through a medium. The number of cycles per identical sampling frequency (fs). The trade-off for
second (hertz) is called the frequency of the wave. the ability to measure flow at precise locations is that
The relationship among the frequency (f), wave- ambiguous information is obtained when flow velo-
length (λ), and velocity (v) of a sound wave is city is very high. By contrast, continuous-wave (CW)
defined by the formula: Doppler technique uses continuous, rather than dis-
crete, pulses of ultrasound waves. As a result, the
v=f×λ region in which flow dynamics are measured cannot
be precisely localized. Blood flow velocity is mea-
Because the frequency of an ultrasound beam is deter-
sured with great accuracy, however, even at high
mined by the properties of the emitting transducer,
flows.
and the velocity is a function of the tissue density
During color flow Doppler (CFD) mapping, real-
through which the sound travels, wavelengths vary
time blood flow within the heart is displayed in colors,
according to the relationship expressed in the above
along with 2D images in black and white. In addition
equation.
to showing the location, direction, and velocity of
Point-of-Care Transesophageal Echocardiography
cardiac blood flow, the images produced by these Fortunately, in the anesthetized patient there are
devices allow estimation of flow acceleration and dif- rarely hemodynamic consequences to esophageal pla-
ferentiation of laminar and turbulent blood flow. cement of the probe, and thus there are no studies that
Color flow Doppler echocardiography is based on specifically address this question. More important for
the principle of multi-gated, PW Doppler where the anesthesiologist is the problem of distraction from
blood flow velocities are sampled at many locations patient care.
along many lines covering the entire imaging sector. Absolute contraindications to TEE in intu-
At the same time, the sector is also scanned to gen- bated patients include esophageal stricture, diver-
erate a 2D image. ticula, tumor, recent suture lines, and known
esophageal interruption. Relative contraindica-
Complications and Contraindications tions include symptomatic hiatal hernia, esopha-
gitis, coagulopathy, esophageal varices, and
Complications resulting from intraoperative TEE
unexplained upper gastrointestinal bleeding. It
can be separated into two groups: injury from
should be noted that, despite these relative
direct trauma to the airway and esophagus, and
contraindications, TEE has been used in patients
indirect effects of TEE. In the first group, potential
undergoing hepatic transplantation without
complications include esophageal bleeding, burn-
reported sequelae.6,7
ing, tearing, dysphagia, and laryngeal discomfort.
Many of these complications could result from
pressure exerted by the tip of the probe on the Basic American Society
esophagus and the airway. Although in most of Echocardiography TEE Views
patients even maximal flexion of the probe will
There is general agreement on the echocardiography
not result in pressure above 17 mmHg,
views that should be obtained during all examinations
occasionally, even in the absence of esophageal
(Figure 5.1).8
disease, pressures greater than 60 mmHg will
result.2
Further confirmation of the low incidence of Transgastric Mid-Papillary Short-Axis View
esophageal injury from TEE is apparent in the few While the American Society of Echocardiography
case reports of complications. In a study of 10,000 recommends multiple views for the definition of
TEE examinations, there was one case of hypo- cardiac pathology and function, the assessment of
pharyngeal perforation (0.01%), two cases of cer- preload, function, and pericardial effusions are the
vical esophageal perforation (0.02%), and no cases most important aspects of the evaluation. All of these
of gastric perforation.3 Kallmeyer et al. reported parameters may be evaluated in the transgastric (TG)
overall incidences of TEE-associated morbidity and mid-papillary short-axis view. The probe is advanced
mortality of 0.2% and 0%, respectively.4 The most into the stomach and slightly anteroflexed until the
common TEE-associated complication was severe posterolateral and anterolateral papillary muscles
odynophagia (0.1%); other complications were are visualized at their attachment to the ventricular
dental injury (0.03%), endotracheal tube malposi- wall. It is important to image the insertion of the
tioning (0.03%), upper gastrointestinal hemorrhage papillary muscles to insure accurate intra- and inter-
(0.03%), and esophageal perforation (0.01%). observer examinations. The left ventricle (LV) is
Piercy et al. reported a gastrointestinal complica- centered on the screen. If the mitral valve apparatus
tion rate of approximately 0.1%, with a great fre- is visualized, the probe should be further advanced
quency of injuries among patients older than 70 or posteroflexed. Similarly, if the ventricular apex is
years and women.5 If resistance is met while visualized, the probe should be either withdrawn or
advancing the probe, the procedure should be anteroflexed. All six middle segments of the ventricle
aborted to avoid these potentially lethal are visualized, which represents perfusion from each
complications. of the three coronary arteries. The size and function
The second group of complications that result of both the right and left ventricles may be evaluated.
from TEE includes hemodynamic and pulmonary Although qualitative evaluation of right ventricular
effects of airway manipulation and, particularly for function may be performed, quantitative measure-
new TEE operators, distraction from patient care. ment of left ventricular fractional area of shortening
36
(a) (b) (c)
(d) (e) (f)
(g) (h) (i)
Figure 5.1 American Society of Echocardiography basic TEE views. (a) Transgastric mid-papillary short axis. (b) Mid-esophageal four
chamber. (c) Mid-esophageal two chamber. (d) Mid-esophageal long axis. (e) Mid-esophageal aortic valve short axis. (f) Mid-esophageal
ascending aortic short axis. (g) Mid-esophageal ascending aortic long axis. (h) Mid-esophageal right ventricular inflow-outflow. (i) Mid-
esophageal bi-caval.
may be calculated. Regional wall motion abnormal- Both anterior and posterior leaflets of the mitral
ities may be visualized, which may represent myo- valve may be imaged and evaluated for motion and
cardial ischemia. pathology. With the echo probe optimally retroflexed
with full visualization of the left ventricular apex, the
Mid-Esophageal Four-Chamber View most posteromedial aspects of the anterior and pos-
terior mitral valve leaflets are usually imaged. As the
Maintaining the probe at approximately 0°, the
probe is either slowly anteroflexed or withdrawn, the
probe is withdrawn into the mid-esophagus. Slight
middle and final anterolateral aspects of these leaflets
retroflexion is usually necessary to avoid foreshor-
may be imaged. The septal and posterior tricuspid
tening of the left ventricle. The array may need to be
valve leaflets are usually visualized; however, if the
rotated approximately 10° to avoid imaging the left
probe is withdrawn too far or is too far anteroflexed,
ventricular outflow tract and optimize visualization
the anterior tricuspid valve leaflet may be visualized
of the basal aspect of the inferoseptal left ventricular
instead of the posterior leaflet. Pulsed wave Doppler at
wall. In this view, the basal, middle, and apical
the level of the mitral valve leaflet tips may be used to
aspects of the anterolateral and inferoseptal left ven-
evaluate diastolic function and the severity of mitral
tricle walls may be evaluated. The presence of left
stenosis; with moderate to severe mitral stenosis, CW
atrial spontaneous echo contrast (“smoke”) may be
Doppler will probably be required. Color flow
observed in patients with left atrial stasis; the left
Doppler may be used to evaluate the severity of mitral
atrial appendage may be examined for evidence of
and tricuspid regurgitation (TR) as well as the pre-
thrombus. Right ventricular size and function may
sence of an atrial septal defect. Because of the small
be evaluated.
expected gradients between the atria, a lower CFD difficult. The left main coronary artery may be
setting is required in order to appreciate interatrial imaged as it emerges adjacent to the left coronary
flow if it is present. cusp. Its bifurcation into the left anterior descending
and circumflex artery is occasionally appreciated.
Mid-Esophageal Two-Chamber View
Rotation of the transducer array to 90° will develop Mid-Esophageal Ascending Aortic Short-
the mid-esophageal (ME) two-chamber view. All seg- Axis View
ments of the anterior and inferior left ventricular The array probe is returned to 0° and the aortic
walls may be evaluated for thickness and contracti- valve is once again centered on the screen. The
lity. The anterior and posterior leaflets of the mitral
probe is withdrawn through the base of the heart
valve are imaged, and evidence of stenosis and regur-
until the pulmonary artery and its bifurcation are
gitation may be obtained. The left atrial size may be
imaged. The severity of proximal and middle
measured, and “smoke” may be imaged if present.
ascending aortic atherosclerosis and pathology
The left atrial appendage can be examined for
may be evaluated.
thrombus, which is not uncommon with left atrial
stasis.
Mid-Esophageal Ascending Aortic Long-
Mid-Esophageal Long-Axis View Axis View
Further rotation to 120° develops the long-axis view. The transducer is rotated to 90–120° to visualize
A single axis from the left ventricular apex to the left the ascending aorta in its long axis. The proximal
ventricular outflow tract to the aortic valve to the and middle section of the aorta can usually be seen
ascending aorta should be developed; further rotation and examined for atherosclerotic disease up to the
to 130° or 140° may be necessary to obtain this axis crossing of the right pulmonary artery. The distal
alignment with cardiac rotation. In this view, both ascending aorta, however, usually cannot be
segments (basal and middle) of the anteroseptal and visualized because it is interposed between the
inferolateral walls may be evaluated. Posterior leaflet esophagus and either the trachea or the bronchus.
restriction may be seen with ischemic cardiac disease,
and systolic anterior motion of the anterior mitral Mid-Esophageal Right Ventricular Inflow-
valve leaflet may be easily appreciated with hyper-
trophic obstructive cardiomyopathy. The severity of
Outflow View
mitral regurgitation may be estimated by CFD. The transducer array is returned to 0° and the
four-chamber view is redeveloped. The tricuspid
valve is centered on the screen and the array is
Mid-Esophageal Aortic Valve Short-Axis rotated to approximately 60°. In addition to the
View left atrium, the right atrium, tricuspid valve, right
The aortic valve is centered on the screen and ventricle, right ventricular outflow tract, pulmon-
the array is rotated to approximately 30°. The ary valve, and proximal pulmonary artery may be
three cusps of the aortic valve (right, left, and visualized. The septal and anterior leaflet of the
non-coronary) may be visualized along with its tricuspid valve may be seen. Color flow Doppler
triangular opening. Evidence of pathology on the may be superimposed, evaluating the severity of
aortic valve cusps may be appreciated. Aortic valve tricuspid regurgitation. If there is tricuspid regur-
stenosis may be qualitatively estimated by obser- gitation, CW Doppler may be used to estimate
ving the aortic valve leaflet opening (see section on systolic pulmonary artery pressure. Moving the
aortic stenosis). Quantization of aortic stenosis by color Doppler field to the pulmonary valve allows
planimetry may be attempted in this view; how- estimation of pulmonary regurgitation.
ever, its use may be limited by shadowing caused
by valvular calcification. Color flow Doppler may Mid-Esophageal Bi-Caval View
be used to identify the site of aortic valve regur- If the array is rotated to approximately 90° and clock-
gitation, but quantification using this view may be wise, the bi-caval view may be obtained. The superior
38
and inferior vena cava may be visualized as they the areas in end-diastole and end-systole divided by
enter the right atrium. The Eustachian valve may the area in end-diastole, represents the FAC.
be seen as a fibrinous structure at the junction of
the inferior vena cava with the right atrium. FAC = (EDA – ESA)/EDA
Septum secundum may be visualized as it extends
EDA = end-diastolic area
superiorly into the right atrium, while septum
primum may be visualized as it extends from the ESA = end-systolic area
inferior aspect towards the left atrium. Low-velocity Since the biplane disk summation method
CFD may be used to image either of these two atrial (modified Simpson’s) corrects for shape distor-
septal defects, which usually are associated with left tions, it is currently the recommended method of
to right flows. 2D volume measurements (Figure 5.2). With
biplane disk summation, the total LV volume is
Descending Aortic Short-Axis View calculated from the summation of a stack of ellip-
The probe is turned leftward toward the descending tical disks. The height of each disk is calculated as
thoracic aorta. Because the probe is severely rotated, a fraction of the LV long axis, based on the longer
the orientation of the images will have changed. of the two lengths from the two- and four-chamber
From this angle, the top of the image is the antero- views. The cross-sectional area of the disk is based on
medial aortic wall while the bottom of the image is the diameters obtained, and the volume of the disk is
the posterolateral wall. The probe is advanced estimated. The volume is obtained by summing these
toward the stomach until the image is lost. It is values.
then withdrawn slowly, imaging the entire descend- While in conventional hemodynamics preload
ing aorta and observing for atherosclerotic disease. is often estimated by measuring left-heart filling
When the probe reaches the level of the upper eso- pressures (i.e., pulmonary capillary wedge pres-
phagus, a long-axis view of the aortic arch may be sure, left atrial pressure, or LV end-diastolic pres-
seen. sure), in echocardiography it can be determined by
measuring LV end-diastolic dimensions. There are
two main echocardiographic signs of decreased
Descending Aortic Long-Axis View preload. (1) Decreased EDA (< 5.5 cm2/m2) invari-
The probe is returned to the stomach and the ably reflects hypovolemia. It is, however, difficult
array is rotated to 90° to develop a long-axis to set an upper limit of EDA below which hypo-
view of the descending aorta. Once again, the volemia can be confirmed. This is particularly true
probe is slowly withdrawn. It may be necessary in patients with impaired contractility where a
to rotate the probe right and left to optimize this compensatory baseline increase in preload makes
long-axis visualization. The aortic arch is visua- the echocardiographic diagnosis of hypovolemia
lized in the upper esophageal area when a cross- difficult. (2) Obliteration of the end-systolic area
sectional image is developed. (ESA), also known as the “kissing ventricle” sign,
often accompanies a decreased EDA in severe
Evaluation of Ventricular Size hypovolemia.
and Function
Right Ventricular Size and Function
Left Ventricular Size and Function The right ventricle (RV) is a complex structure that
The fractional area change (FAC) is the parameter pumps venous blood to the normally low pressure/
most often requested from the echocardiographer. low resistance pulmonary arterial circuit. Owing to
While this measurement is dependent upon loading the historical focus on the left side of the circulation,
conditions, it is frequently used to guide clinical man- lack of geometrical assumptions of RV shape, and the
agement of patients with cardiovascular disease. This difficulty in imaging the right heart, information
measurement is simple and reproducible. The LV regarding the RV has been limited until relatively
cavity is imaged in a transgastric midpapillary short- recently. When right ventricular function and loading
axis (SAX) view and the area measured in end-systole conditions are normal, the RV is typically triangular
and end-diastole (Figure 5.2). The difference between when viewed in the mid-esophageal four-chamber
39
Figure 5.2 Calculation of fractional area

of change (FAC) and ejection fraction (EF).
The left column is diastole and the right
column is systole. The first row illustrates
the calculation of left ventricular (LV) FAC
using the transgastric mid-papillary view.
The second row illustrates the calculation
of LV EF using a modified Simpson’s
method with the mid-esophageal four-
chamber view. The third row illustrates the
calculation of right ventricular FAC using
the mid-esophageal four-chamber view.
view and crescent-shaped when viewed in the trans- RV-focused apical four-chamber view.9 There are no
gastric mid-papillary short-axis view. guidelines specific to TEE. Right ventricle pressure or
Right ventricle dilation is readily identified volume overload may cause distortion or flattening of
by echocardiography and may be assessed qualita- the interventricular septum that is most easily identi-
tively or quantitatively. Qualitatively, the RV size is fied in the TG mid-papillary short-axis view (Figure
compared to the LV size in the mid-esophageal four- 5.3). The overload of the RV, as well as the under-
chamber view; its cross-sectional area normally occu- filling of the LV from reduced RV output, leads to a
pies two-thirds of the normal LV cross-sectional area leftward deviation of the septum and a “D-shaped” LV
(Figure 5.3). Mild enlargement is defined as an RV chamber appearance.
size that is more than two-thirds the LV size, moder- Right ventricular FAC is another 2D-based
ate enlargement is defined as the chambers being method of systolic function evaluation. This measure-
equal in size, and severe enlargement is defined as ment may be obtained in an ME four-chamber view,
the RV area greater than the LV area. Quantitatively, tracing the RV from the lateral tricuspid annulus,
the RV is difficult to assess due to its complex shape down the RV free wall to the apex, and returning
and the poor interobserver reproducibility of RV along the septum to the tricuspid annulus (Figure
chamber size measurements. Current RV chamber 5.2). The change in this area measurement between
quantification guidelines suggest upper reference diastole and systole is calculated as a percentage. An
values of a 4.1 cm diameter at the base and a 3.5 cm RV FAC less than 35% is indicative of RV
diameter in the midlevel of the RV on a transthoracic dysfunction.8
40
Figure 5.3 Large right ventricle. Mid-esophageal four-chamber view on the left demonstrates a right ventricle that is significantly larger than the
left ventricle. A transgastric mid-papillary view demonstrates a “D-shaped” left ventricle because of the increase in right ventricular pressure.
Assessment of Coronary Ischemia ventricle and classifying it as normal, hypokinetic,

akinetic, dyskinetic, or aneurysmal. The degree of
The ability to reliably detect regional wall motion
thickening (i.e., the percentage change in wall thick-
abnormalities (RWMA) is clinically relevant because
ness during systole) can also be used to assess overall
of its diagnostic and therapeutic implications. Not
function of the observed segment. A thickening
every RWMA detected by TEE is diagnostic for myo-
greater than 30% is normal, 10–30% indicates mild
cardial ischemia. Myocarditis, ventricular pacing, and
hypokinesia, 0–10% indicates severe hypokinesia, no
bundle branch blocks can easily lead to wall motion
thickening indicates akinesia, and bulging of the seg-
abnormalities that can be misinterpreted and poten-
ment during systole indicates dyskinesia.
tially lead to clinical mismanagement.
Regional wall motion abnormalities anatomical
localization and degree of dysfunction should be
Assessment of Pressures and Flow
described according to the ASE classification, which Doppler echocardiographic measurements may be
is based on a 16-segment model of the LV (Figure used to calculate pressure gradients. The Bernoulli
5.4).8 This model subdivides the LV into three zones principle states that an increase in the speed of the
(basal, mid, and apical). The basal (segments 1–6) and fluid occurs simultaneously with a decrease in its
mid-ventricular (segments 7–12) zones are each sub- pressure or a decrease in the fluid’s potential
divided into six segments, while the apical zone is energy. When flow acceleration and viscous fric-
subdivided into four (segments 13–16). tion variables of blood are ignored, and flow velo-
The coronary distribution related to the indivi- city proximal to a fixed obstruction is significantly
dual segments can be assessed, enabling the echocar- less than flow velocity after the obstruction, a sim-
diographer to make assumptions regarding the plified modified Bernoulli equation may be defined
localization of a potential coronary lesion. as:
Segments 1, 2, 7, 13, 14, and 17 are in the distribution
ΔP = 4(V)2
territory of the left anterior descending artery.
Segments 5, 6, 11, 12, and 16 are associated with where ΔP is the pressure difference between two
the circumflex artery, and segments 3, 4, 9, 10, and structures and V is the velocity across the structures.
15 are supplied by the right coronary artery (Figure With this formula, the pressure gradient across a
5.4). This segmental distribution, however, can be fixed orifice can be estimated.
variable among patients due to the variability of the
coronary arteries.
In addition to defining anatomical segments of the
Aortic Valve Evaluation
LV, it is important to grade segment thickening and Two-dimensional transesophageal echocardiographic
excursion. The simplest assessment is achieved by the interrogation provides information on valve area,
echocardiographer “eyeballing” the motion of the leaflet structure, and mobility. The valve is composed
Segmental model of the left ventricle
Four-chamber view Two-chamber view Long-axis view Mid-short axis view Basal short-axis view
4
10
11 3è 5
9
12
6 1 8 2 6
3 4
7
2 1
10 7 5
9 12
RCA RCA or CX
13 11 8
14 16 15 LAD LAD or CX
LCX RCA or LAD
17 17
15 or 16 17 13 or 14
1. Basal anterior 7. Mid-anterior 13. Apical anterior

2. Basal anteroseptal 8. Mid-anteroseptal 14. Apical septal
3. Basal inferoseptal 9. Mid-inferoseptal 15. Apical inferior
4. Basal inferior 10. Mid-inferior 16. Apical lateral
5. Basal inferolateral 11. Mid-inferolaterale 17. Apex
6. Basal anterolateral 12. Mid-anterolateral
Figure 5.4 was taken from Badano LP, Picano E. (2015) Standardized Myocardial Segmentation of the Left Ventricle. In: Stress
Echocardiography. Springer, Cham. (A black and white version of this figure will appear in some formats. For the color version, please refer to the
plate section).
of three fibrous cusps (right, left, and non-coronary) immobile leaflets and is usually associated with
attached to the root of the aorta. The spaces between concentric left ventricular hypertrophy and a
the attachments of the cusps are called the “commis- dilated aortic root. The valve leaflets may be
sures” and the circumferential connection of these domed during systole; this finding is sufficient for
commissures is the sinotubular junction. The aortic a diagnosis of aortic stenosis.25
wall bulge behind each cusp is known as the sinus of The quantification of aortic stenosis is sum-
Valsalva. The sinotubular junction, the sinuses of marized in Table 5.1. The aortic valve area may
Valsalva, the valve cusps, and the junction of the be measured by planimetry (Figure 5.5).11 A cross-
aortic valve with the ventricular septum and anterior sectional view of the aortic valve orifice may be
mitral valve leaflet comprise the aortic valve complex. obtained by TEE, and measurement of the aortic
The three leaflets of the aortic valve are easily visua- valve area corresponds well to those obtained by
lized, and vegetations or calcifications can be identi- transthoracic echocardiography and cardiac cathe-
fied on basal transverse imaging or longitudinal terization, assuming the degree of calcification is
imaging. not severe. The severity of aortic stenosis may be
quantified using Doppler echocardiography (Figure
Aortic Stenosis 5.6).12 The evaluation of severity, however, may be
Aortic stenosis may be caused by congenital uni- limited by difficulty aligning the ultrasonic beam
cuspid, bicuspid, tricuspid, or quadricuspid valves, with the direction of blood flow through the left
rheumatic fever, or degenerative calcification of the ventricular outflow tract. Normal Doppler signals
valve in the elderly.10 Valvular aortic stenosis is across the aortic valve have a velocity of less than
characterized by thickened, echogenic, calcified, 1.5 m/sec and have peak signals during early
42
Table 5.1 Summary of aortic stenosis.
Aortic sclerosis Mild Moderate Severe

Aortic jet velocity (m/s) ≤ 2.5 2.6–2.9 3.0–4.0 > 4.0
Mean gradient (mmHg) < 20 20–40 > 40
2
Aortic valve area (cm ) > 1.5 1.0–1.5 < 1.0
Source: C.M. Otto, Valvular aortic stenosis: disease severity and timing of intervention. J Am Coll Cardiol 2006;47:2141–51.
Figure 5.5 Aortic valve stenosis by planimetry. The left panel indicates a normal aortic valve while the right panel indicates an aortic valve
with stenosis. Since there is no significant calcification of the valve, planimetry may be used.
Figure 5.6 Doppler spectrum through a

stenotic aortic valve.
systole. With worsening aortic stenosis, the flow of either CW or high PRF Doppler. Aortic velocity
velocities increase and the signal’s peak is later allows classification of stenosis as mild (2.6 to
during systole. These high velocities will limit the 2.9 m/sec), moderate (3.0 to 4.0 m/sec), or severe
use of pulsed wave Doppler and necessitate the use (> 4 m/sec).13
Aortic Regurgitation The vena contracta is the narrowest portion of a

Aortic regurgitation or aortic insufficiency (AI) may regurgitant jet that usually occurs at or immediately
result from either diseases of the aortic leaflets or the upstream from the valve (Figure 5.7). This jet width is
aortic root.14 Valvular lesions that may result in AI directly proportional to the severity of the AI. It is
include leaflet vegetations and calcifications, perfora- usually characterized by high velocity and laminar
tion, or prolapse. Aortic insufficiency may be caused flow and is slightly smaller than the regurgitant ori-
by annular dilation secondary to a variety of etiolo- fice.16 A vena contra diameter less than 0.3 cm is
gies, including annulo-aortic ectasia, Marfan’s syn- consistent with mild AI, and a diameter greater than
drome, aortic dissection, collagen vascular disease, 0.6 cm is consistent with severe AI.
and syphilis.
Cusp pathology (e.g., redundancy, restriction, Mitral Valve Evaluation
mobility, thickness, integrity), commissural varia- The mitral value consists of two leaflets, chordae
tions (fusion, splaying, alignment, and attachment tendineae, two papillary muscles, and a valve annulus.
site), and root morphology (septal hypertrophy and The anterior leaflet is larger than the posterior leaflet
root dimensions) should be ascertained.13 Leaflet and is semicircular; the posterior mitral valve leaflet
movement (excessive, restricted, or normal), origin has a longer circumferential attachment to the mitral
of jet (central or peripheral), and direction of regur- valve annulus.18 The leaflets are connected to each
gitant jet (eccentric or central) should be determined other at junctures of continuous leaflet tissue called
to provide insight into the underlying pathology. “commissures.” Primary, secondary, and tertiary
Other signs that may be associated with AI include chordal structures arise from the papillary muscle,
high-frequency diastolic fluttering of the mitral valve, subdivide as they extend, and attach to the free edge
premature closing of the mitral valve, and reverse and several millimeters from the margin on the ven-
doming of the mitral valve.15,16 tricular surface of both the anterior and posterior
The criteria for qualitative grading of AI are sum- valve leaflets.19 The annulus of the mitral valve pri-
marized in Table 5.2. Color flow Doppler has tradi- marily supports the posterior mitral valve leaflet,
tionally been the major method of assessing the while the anterior mitral valve leaflet is continuous
severity of valvular regurgitation. Nyquist limits with the membranous ventricular septum, aortic
should provide an aliasing velocity of approximately valve, and aorta.
50–60 cm/sec and a color gain that just eliminates the
random color speckle from nonmoving regions.17 Mitral Stenosis
Aortic regurgitant flow through the outflow tract is The most common etiology of mitral stenosis is rheu-
characteristically a high-velocity turbulent jet extend- matic disease; other causes include congenital valvular
ing through the left ventricular outflow tract and left stenosis, vegetations and calcifications of the leaflets,
ventricle during diastole. parachute mitral valve, and annular calcification. In
Table 5.2 Quantification of aortic regurgitation.
Mild Moderate Severe

Left atrial size Normal Normal or dilated Usually dilated
Aortic cusps Normal or abnormal Normal or abnormal Abnormal/flail or wide
coaption defect
Jet width in LVOT* Small in central jets Intermediate Large in central jets; variable
in eccentric jets
Vena contracta width (cm)* < 0.3 0.3–0.6 ≥ 0.6
* At Nyquist limits of 50–60 cm/sec

CSA = cross-sectional area
LVOT = left ventricular outflow tract
Source: Zoghbi WA, Adams D, Bonow RO, et al. Recommendations for noninvasive evaluation of native valvular regurgitation: a report from
the American society of echocardiography developed in collaboration with the society for cardiovascular magnetic resonance. J Am Soc
Echocardiogr 2017; 30:303–71.
44
Figure 5.7 Vena contra aortic valve. Mid-esophageal aortic valve long axis. The vena contracta may be used to differentiate among degrees
of aortic regurgitation. The image on the left has a narrow vena contracta, which is associated with a mild degree of aortic regurgitation. By
contrast, the large vena contracta on the right is associated with a severe degree of aortic regurgitation. (A black and white version of this figure
will appear in some formats. For the color version, please refer to the plate section).
addition to structural valvular abnormalities, mitral underestimation of the degree of hemodynamic com-
stenosis may be caused by non-valvular etiologies promise may occur when determining mitral valve
such as intra-atrial masses (myxomas or thrombus) area by planimetry.26
or extrinsic constrictive lesions.20,21 Generally, mitral A transmitral Doppler spectrum is measured along
stenosis is characterized by restricted leaflet move- the axis of transmitral blood flow, which may usually
ment, a reduced orifice, and diastolic doming be obtained in a mid-esophageal four-chamber view or
(Figure 5.8).22 The diastolic doming occurs when the two-chamber view (Figure 5.8). Transmitral valve flow
mitral valve is unable to accommodate all the blood is characterized by two peaked waves of flow away from
flowing from the left atrium into the ventricle, so the the transducer. The first wave (E) represents early
body of the leaflets separates more than the edges. In diastolic filling, while the second wave (A) represents
rheumatic disease, calcification of the valvular and atrial systole. The transvalvular gradient may be
subvalvular apparatus, as well as thickening, deforma- estimated using the modified Bernoulli equation:27
tion, and fusion of the valvular leaflets at the antero- pressure gradient = 4 × (maximal velocity)2. Since
lateral and posteromedial commissures, produces a peak gradient is heavily influenced by left atrial com-
characteristic fish-mouth-shaped orifice.23 Other pliance and ventricular diastolic function, the mean
characteristics that may be associated with chronic gradient is the relevant clinical measurement.24 The
obstruction to left atrial outflow include an enlarged high velocities that may occur with mitral stenosis
left atrium, spontaneous echo contrast or “smoke” limit the use of pulsed wave Doppler echocardiogra-
(which is related to low-velocity blood flow with sub- phy, thus CW Doppler echocardiography should be
sequent rouleaux formation by red blood cells24), utilized.
thrombus formation, and RV dilation.
The assessment of the severity of mitral stenosis is Mitral Regurgitation
summarized in Table 5.3. Because planimetry of the Mitral regurgitation (MR) may be classified as primary
mitral valve orifice is not influenced by assumptions or secondary. Primary causes of regurgitation are
of flow conditions, ventricular compliance, or asso- structural or organic, while secondary causes are func-
ciated valvular lesions, it is the reference standard for tional (i.e., without evidence of structural abnormal-
the evaluation of the mitral valve area in mitral ste- ities of the mitral valve). The most common causes of
nosis.25 While at times technically difficult, care primary mitral regurgitation are degenerative (e.g.,
should be taken to image the orifice at the leaflet Barlow’s disease, fibroelastic degeneration, Marfan’s
tips. Severe calcification of the mitral valve may inter- syndrome, Ehler–Danos syndrome, annular calcifica-
fere with mitral valve area determination and, in tion), rheumatic disease, toxic valvulopathies, and
patients with significant subvalvular stenosis, endocarditis.13 Mitral regurgitation may be caused by
Table 5.3 Quantification of mitral stenosis.

2
Valve area (cm ) > 1.5 1.0–1.5 > 1.0
Mean gradient (mmHg) <5 5–10 > 10
Source: Baumgartner H, Hung J, Bermejo J, et al. Echocardiographic assessment of valve stenosis: EAE/ASE recommendations
for clinical practice. J Am Soc Echocardiogr 2009;22:1–23.
Figure 5.8 Top row: Mid-esophageal four-chamber view. The mitral valve is severely stenotic with severe calcification of the annulus and
leaflets. Bottom row: Transmitral Doppler spectrum. The panel on the left is the normal transmitral Doppler flow and the panel on the right is
the transmitral flow in the presence of mitral stenosis. (A black and white version of this figure will appear in some formats. For the color
version, please refer to the plate section).
disorders of any component of the mitral valve appa- annulus and atrium dilate and the annulus loses its
ratus (i.e., the annulus, the leaflets and chordae, or normal elliptical shape, becoming more circular.28
papillary muscles). With chronic regurgitation the Annular dilation, in turn, leads to poor leaflet
46
Figure 5.9 Mitral regurgitation. The top row illustrates a Barlow’s mitral valve with multiple billowing and prolapsed segments. The second
row illustrates a prolapsed posterior leaflet. The posterior leaflet is above the level of the mitral annulus. By contrast, a flail segment is illustrated
in the bottom row, with the mitral valve segment pointing toward the atrium. (A black and white version of this figure will appear in some
formats. For the color version, please refer to the plate section).
coaptation and worsening mitral regurgitation. while older populations present with fibroelastic defi-
Although increased left atrial and ventricular dimen- ciency disease. A Barlow’s valve is usually character-
sions may suggest severe mitral regurgitation, smaller ized by gross redundancy of multiple segments of the
dimensions do not exclude the diagnosis.29 anterior or posterior leaflets and chordal apparatus
The most common cause of chronic primary MR (Figure 5.9). The leaflets are bulky and billowing with
in developed countries is mitral valve prolapse.30 multiple areas of prolapse.31 The chordae are more
Younger individuals present with Barlow’s syndrome, often elongated rather than ruptured. The leaflets are
thickened with severe myxomatous degeneration. The moderate and moderate-severe grades.13,16 The most
annulus is usually severely dilated and may be common method of grading the severity of mitral
calcified. regurgitation is CFD mapping of the left atrium.
By contrast, fibroelastic deficiency usually only With the Nyquist limit set at 50–60 cm/sec, jet areas
affects a single segment. The nonaffected leaflets less than 4 cm2 or 20% of the left atrial size are usually
tend to be thin, with a thickening of the affected classified as mild, while jets greater than 10 cm2 or
segment. Elongated chords may produce prolapse of 40% of the atrial volume are classified as severe.16 An
one or both attached leaflets. Excessively mobile alternative method of grading mitral regurgitation is
structures near the leaflet tips during diastole may based upon the vena contracta width.34 While the
represent elongated chords or ruptured minor chords. vena contracta is commonly circular, it may be ellip-
These structures do not prolapse into the atrium dur- tical in shape with secondary etiologies or functional
ing systole (Figure 5.9). By contrast, ruptured major regurgitation.13 In these cases, multiple views of the
chords are identified as thin structures with a flutter- vena contra along different axes should be obtained
ing appearance in the atrium during systole and are and averaged. A vena contracta width of less than
associated with marked prolapse of the affected leaflet. 0.3 cm is associated with mild MR, while a width
In this instance, the valve segment is termed as “flail.” greater than 0.7 cm is associated with severe MR.16
A flail leaflet segment generally points in the direction
of the left atrium, and this directionality of leaflet Tricuspid Valve
pointing is the principal criterion for distinguishing
The tricuspid valve consists of three leaflets, an annu-
a flailed leaflet from severe valvular prolapse.32,33 Flail
lar ring, chordae tendineae, and multiple papillary
leaflets are most commonly caused by ruptured
muscles.35 The anterior leaflet is usually the largest,
chordae and less commonly caused by papillary
followed by the posterior and septal leaflets. The sep-
muscle rupture.
tal leaflet of the tricuspid valve is usually further apical
With secondary or functional mitral regurgitation,
than the septal attachments of the mitral valve.
the mitral valve is structurally normal.13,29 Left ven-
Chordae arise from a large single papillary muscle,
tricular dilation secondary to another process such as
double or multiple septal papillary muscles, and sev-
myocardial infarction or idiopathic dilated cardio-
eral small posterior papillary muscles attached to the
myopathies result in papillary muscle displacement
corresponding walls of the right ventricle.
and annular dilation with resultant tethering of the
Intrinsic structural abnormalities of the tricuspid
mitral valve leaflets with incomplete leaflet coaption.
valve that can be well characterized by TEE include
Since the valvular regurgitation is only one compo-
rheumatic tricuspid stenosis, carcinoid involvement
nent of the disease process, its progress is worse than
of the tricuspid valve, tricuspid valve prolapse, flail
primary mitral regurgitation and its treatment is less
tricuspid valve, Ebstein’s anomaly, and tricuspid
clear.
endocarditis. Rheumatic involvement of the tricuspid
Mitral regurgitation is graded semi-quantitatively
valve, which is typically seen with concomitant mitral
as mild, moderate, or severe, as summarized in Table
valve involvement, is characterized by thickening of
5.4. Regurgitation less than mild may be classified as
the leaflets (particularly at their coaptation surfaces),
either trivial or trace. Some authors have suggested
fusion of the commissures, and shortening of the
the subdividing of moderate regurgitation into mild-
chordal structures resulting in restricted leaflet
Table 5.4 Summary of mitral regurgitation (data from reference 17).

Left atrial size Normal Normal or dilated Usually dilated
Color flow jet area* Small central jet (< 4 cm2 Large central jet (> 10 cm2 or > 40% LA)
or < 20% LA area) or variable-sized wall impinging jet
Vena contracta width (cm) < 0.3 0.3 – 0.69 ≥ 0.7
* At Nyquist limits of 50–60 cm/sec.
48
Table 5.5 Quantification of tricuspid regurgitation.

Right atrial size Normal Normal or dilated Usually dilated
Tricuspid valve leaflets Usually normal Normal or abnormal Abnormal/flail or wide
coaption defect
Jet area – central jets (cm2)* <5 5–10 > 10
Vena contracta width (cm)* Not defined Not defined, but < 0.7 > 0.7
* At Nyquist limits of 50–60 cm/sec.

Source: Zoghbi WA, Adams D, Bonow RO, et al. Recommendations for noninvasive evaluation of native valvular regurgitation: a report from
the American society of echocardiography developed in collaboration with the society for cardiovascular magnetic resonance. J Am Soc
Echocardiogr 2017;30:303–71.
Figure 5.10 Functional tricuspid

regurgitation. A mid-esophageal four-
chamber view demonstrates a large
functional tricuspid regurgitant jet. The
vena contracta is measured as the jet
emanates from the right ventricle. (A black
and white version of this figure will appear
in some formats. For the color version,
please refer to the plate section).
motion.36 Carcinoid syndrome results in a diffuse While tricuspid regurgitation (TR) may have pri-
thickening of the tricuspid valve (and pulmonic mary etiologies, most causes are secondary or functional
valve), and endocardial thickening of right-heart as a result of either tricuspid annular dilation (greater
structures that may result in restricted tricuspid than 40 mm) or RV dilation (Figure 5.10).13 Right ven-
valve motion (mixed stenosis and regurgitation).37 tricle enlargement results in annular dilation and papil-
Supravalvular, valvular, or subvalvular restriction lary muscle displacement with tethering of the tricuspid
may cause tricuspid stenosis. The most common etiol- valve leaflets. This tethering may result in poor leaflet
ogy of tricuspid stenosis is rheumatic heart disease, coaption. This TR results in additional RV enlargement
while less common causes include carcinoid syn- and further leaflet tethering with worsening TR.
drome and endomyocardial fibrosis. Tricuspid steno- The quantification of TR is summarized in Table
sis is characterized by a domed, thickened valve with 5.5. The apparent severity of tricuspid regurgitation
restricted movement. Tricuspid regurgitation may be is exquisitely sensitive to right-heart loading condi-
secondary to annular or right ventricular dilation, or tions. A central jet area of less than 5 cm2 is consis-
pathology of the leaflets or subvalvular apparatus. tent with mild regurgitation, while a jet area greater
Continuous wave Doppler measurements of the than 10 cm2 is consistent with severe regurgita-
inflow velocities across the tricuspid valve can be tion.16 Recent guidelines, however, suggest that
employed to estimate the mean diastolic tricuspid color flow area of the regurgitant jet should not be
valve gradient with the modified Bernoulli equation.38 used to quantitate the severity of TR.13 A vena
contra width less than 0.3 mm is consistent with pericardial layers. Under pathological conditions,
mild regurgitation, while a VC greater than 0.7 cm fluid accumulation can occur, resulting in the devel-
is consistent with severe regurgitation.39 opment of a pericardial effusion (Figure 5.11).
According to the 2013 guidelines, small effusions are
Pericardial Disease usually defined as 50 to 100 mL, moderate as 100 to
500 mL, and large as greater than 500 mL.40 The size
The pericardium is a two-layered structure reflecting
of the effusion correlates poorly with its hemody-
from a visceral layer to a parietal layer approximately
namic effect; the rapidity of effusion accumulation
1–2 cm distal to the origin of the great vessels and
has a greater effect on hemodynamics. Different etiol-
around the pulmonary veins. Under normal circum-
ogies of pericardial effusions have characteristic sizes
stances 5–10 ml of fluid is contained within the peri-
and progressions. Idiopathic or viral infections may
cardial sack, allowing for practically frictionless
result in small pericardial effusions, while large effu-
motion of the heart during the cardiac cycle. The
sions may be associated with hypothyroidism, tuber-
parietal layer of the pericardium is rich in collagen
culosis, or neoplasms. Rapid pericardial blood
fibers, making it a low-compliance structure confin-
accumulation may be from blunt trauma, ascending
ing the volume of the four cardiac chambers. In other
aortic dissection, or cardiac rupture (either secondary
words, a volume increase of one chamber requires a
to myocardial infarction or iatrogenic, such as during
reduction of volume within another. Likewise, if an
invasive cardiac procedures).41 Free effusions are typi-
increase in volume is seen within the pericardial sack,
cally seen in medical conditions leading to pericardial
a reduction of chamber volumes must occur.
effusions, whereas loculated effusions are seen after
surgery or inflammatory processes. In many patients,
Pericardial Effusion however, the etiology of pericardial effusions must be
Under normal circumstances the echocardiographer classified as idiopathic.
is unable to visualize the fluid film between the two Most echocardiographers use a qualitative grading
system to characterize the quantity of pericardial effu-
Table 5.6 Severity of pericardial effusions.
sion. This qualitative grading uses the diameter of the
Diameter of effusion Severity effusion in two dimensions. The 2013 guidelines are
0–1.0 cm Mild
summarized in Table 5.6. Trivial effusions are only
seen during systole, mild effusions are less than 1 cm,
1.1–2.0 cm Moderate
and large effusions are greater than 2 cm in diameter.
> 2.0 cm Severe
Figure 5.11 Transgastric short-axis view

of left ventricle. A large pericardial effusion
seen surrounding the LV.
50
Additionally, the effusion can either encompass the venous fullness with systolic blunting of hepatic
entire heart (free) or be loculated. While transthoracic blood flow velocities, excessive ventricular septal
echocardiography may not be sensitive for the detec- movement with respiration, and a small left ventricle.
tion of loculated effusions, TEE is an excellent mod- An IVC greater than 2.1 cm with less than a 50%
ality for the detection of these effusions. TEE has been reduction in diameter during inspiration is indicative
demonstrated to be a more sensitive modality for the of an elevation in systemic venous pressure that
detection of loculated effusions or intra-pericardial accompanies an increased pericardial pressure.45
clots, compared with the transthoracic route.42
Aortic Disease
Cardiac Tamponade The diagnosis of aortic dissection is based on the
Cardiac tamponade and pericardial effusion are not presence of an intimal flap (Figure 5.12). In the diag-
synonymous. A pericardial effusion is an anatomical nosis of dissection, TEE has overcome some of the
diagnosis that may or may not lead to hemodynamic major disadvantages of the alternative diagnostic
alterations. Cardiac tamponade is a pathophysiologi- modalities (CT, MRI). In comparison to these other
cal diagnosis. Echocardiographically, cardiac tampo- modalities, TEE has been shown to have high sensi-
nade may be identified as right atrial and ventricular tivity and specificity.46,47 According to the 2010
collapse during their relaxation phase. This collapse guidelines for the diagnosis and management of aortic
occurs when the intra-pericardial pressure exceeds the disease, TEE, CT, and MRI are all recommended as
intrachamber pressures. The severity and duration of definitive methods for the identification or exclusion
this collapse increase with further increases in the of thoracic aortic dissection.48 The sensitivity of TEE
pericardial pressure. Duration of right atrial collapse for the detection of proximal aortic dissection is 88–
exceeding one-third of the cardiac cycle is nearly 98%, with a specificity of 90–95%.49 With further
100% sensitive and specific for clinical cardiac tampo- improvement in technology, the sensitivity of TEE in
nade.43 Since the right ventricle is thicker than the detecting aortic dissections approaches 100%.50
right atrium, a higher pericardial pressure is required Ideally the specific locations of the entry and exit
for right ventricular diastolic collapse; right ventricu- sites are also identifiable. The primary tear usually
lar collapse is a more specific but less sensitive sign of has a diameter of over 5 mm and is frequently located
pericardial tamponade.44 In addition to the presence in the proximal part of the ascending aorta in type A
of a pericardial effusion, other signs of tamponade dissections and immediately below the origin of the
include a dilated inferior vena cava (IVC), hepatic left subclavian artery in type B dissections. Flow in
Figure 5.12 Aortic dissection. An aortic

long-axis view is presented with an intimal
flap from an aortic dissection visualized in
the proximal ascending aorta.
both the true and false lumens can be analyzed with case series of 7200 cardiac surgical patients. Anesth
Doppler color flow imaging. TEE performed in real Analg 2001;92:1126–30.
time utilizes its unique ability to give functional and 5. Piercy M, McNicol L, Dinh DT, Story DA, Smith JA.
hemodynamic information. This enables the detec- Major complications related to the use of
tion of the common complications of aortic dissec- transesophageal echocardiography in cardiac surgery.
tion: aortic valve regurgitation, pericardial J Cardiothorac Vasc Anesth 2009;23:62–5.
tamponade, and left ventricular dysfunction second- 6 Ellis JE, Lichtor JL, Feinstein SB, et al. Right heart
ary to coronary artery involvement in the dissection dysfunction, pulmonary embolism, and paradoxical
embolization during liver transplantation. Anesth
process.
Analg 1989;68:777.
Pericardial effusions are commonly seen with
acute aortic dissection and may occur as a result of 7 Suriani RJ, Cutrone A, Feierman D, Konstadt S.
Intraoperative transesophageal echocardiography
two etiologies.47 The most common etiology is the
during liver transplantation. J Cardiothorac Vasc
transudation of fluid across the thin-walled false Anesth 1996;10:699–707.
lumen into the pericardium. This is usually mild and
8. Reeves ST, Finley AC, Skubas NJ, et al. Council on
does not lead to a hemodynamically significant peri- Perioperative Echocardiography of the American
cardial effusion. Alternatively, the aorta may rupture Society of Echocardiography; Society of Cardiovascular
directly into the pericardium with resultant pericar- Anesthesiologists. Basic perioperative transesophageal
dial tamponade and hemodynamic collapse. echocardiography examination: a consensus statement
of the American Society of Echocardiography and the
Society of Cardiovascular Anesthesiologists. J Am Soc
Conclusion Echocardiogr 2013;26:443–56.
Transesophageal echocardiography is a powerful
9. Lang RM, Badano LP, Mor-Avi V, et al.
diagnostic tool that revolutionized the practice of Recommendations for cardiac chamber quantification
cardiac anesthesia. While its full functionality is by echocardiography in adults: an update from the
most often utilized by cardiac anesthesiologists, its American society of echocardiography and the
application to other high-risk surgeries (e.g., liver European association of cardiovascular imaging. J Am
transplantation), high-risk patients with diseases like Soc Echocardiogr 2015;28:1–39.
pulmonary hypertension and aortic stenosis under- 10. Rapaport E, Rackley CE, Cohn LH. Aortic valve
going routine surgery, and as a “rescue” diagnostic disease. In: Schlant RC, Alexander RW, O’Rourke RA,
method to access hemodynamic collapse is unpa- et al. eds. Hurst’s The Heart: Arteries and Veins. New
trolled. Perioperative physicians should have a basic York: McGraw Hill, 1994.
knowledge of the principles of TEE, and may enhance 11. Stoddard MF, Arce J, Liddell NE, et al. Two
their everyday practice by incorporating some of the dimensional transesophageal echocardiographic
determination of aortic valve area in adults with aortic
basic aspects of TEE examination.
stenosis. Am Heart J 1991;122:1415.
12. C.M. Otto, Valvular aortic stenosis: disease severity
References and timing of intervention. J Am Coll Cardiol
2006;47:2141–51.
1. Kahn RA, Fischer GW. Transesophageal
echocardiography. In: Reich DL, Kahn RA, Mittnacht 13. Baumgartner H, Hung J, Bermejo J, et al.
AN, et al. eds. Monitoring in Anesthesia and Recommendations on the echocardiographic
Perioperative Care. New York, NY: Cambridge assessment of aortic valve stenosis: a focused update
University Press, 2011;105–35. from the European Association of Cardiovascular
Imaging and the American Society of
2. Urbanowicz JH, Kernoff RS, Oppenheim G, et al.
Echocardiography. J Am Soc Echocardiogr 2017;30
Transesophageal echocardiography and its potential for
(4):372–92.
esophageal damage. Anesthesiology 1990;72:40.
14. Lancellotti P, Tribouilloy C, Hagendorff A, et al.
3. Min JK, Spencer KT, Furlong KT, et al. Clinical features
Scientific Document Committee of the European
of complications from transesophageal
Association of Cardiovascular Imaging.
echocardiography: a single-center case series of 10,000
Recommendations for the echocardiographic
consecutive examinations. J Am Soc Echocardiogr
assessment of native valvular regurgitation: an
2005;18:925–9.
executive summary from the European Association of
4. Kallmeyer IJ, Collard CD, Fox JA, et al. The safety of Cardiovascular Imaging. Eur Heart J Cardiovasc
intraoperative transesophageal echocardiography: a Imaging 2013;14(7):611–44.
52
15. Roberson WS, Stewart J, Armstrong WF, Dillon JC, 28. Ormiston JA, Shah PM, Tei C, Wong M. Size and
Feigenbaum H. Reverse doming of the anterior mitral motion of the mitral valve annulus in man. Circulation
leaflet with severe aortic regurgitation. J Am Coll 1981;64:113.
Cardiol 1984;3:431. 29. Burwash IG, Blackmore GL, Koilpillai CJ. Usefulness
16. Ambrose JA, Meller J, Teichholz LE, Herman MV. of left atrial and left ventricular chamber sizes as
Premature closure of the mitral valve: predictors of the severity of mitral regurgitation. Am J
echocardiographic clue for the diagnosis of aortic Cardiol 1992;15:774.
dissection. Chest 1978;73:121. 30. Nishimura RA, Otto CM, Bonow RO, et al. 2017 AHA/
17. Zoghbi WA, Adams D, Bonow RO, et al. ACC focused update of the 2014 AHA/ACC guideline
Recommendations for noninvasive evaluation of for the management of patients with valvular heart
native valvular regurgitation: a report from the disease: a report of the American College of
American Society of Echocardiography developed in Cardiology/American Heart Association Task Force
collaboration with the Society for Cardiovascular on Clinical Practice Guidelines. J Am Coll Cardiol
Magnetic Resonance. J Am Soc Echocardiogr 2017;70(2):252–89.
2017;30:303–71. 31. Anyanwu AC, Adams DH. Etiologic classification of
18. Ranganathan N, Lam JHC, Wigle ED, Silver MD. degenerative mitral valve disease: Barlow’s disease and
Morphology of the human mitral valve: II. The valve fibroelastic deficiency. Semin Thorac Cardiovasc Surg
leaflets. Circulation 1970;41:459–67. 2007;19:90–6.
19. Perloff JK, Roberts WC. The mitral apparatus: 32. Mintz GS, Kotler MN, Segal BL, Parry WR. Two
functional anatomy of mitral regurgitation. Circulation dimensional echocardiographic recognition of
1972;46:227–39. ruptured chordae tendineae. Circulation 1978;57:244.
20. Hammer WJ, Roberts WC, deLeon AC Jr. Mitral 33. Ogawa S, Mardelli TJ, Hubbard FE. The role of cross-
stenosis secondary to combined massive mitral sectional echocardiography in the diagnosis of flail
annular calcific deposits and small, hypertrophied left mitral leaflet. Clin Cardiol 1978;1:85.
ventricles: hemodynamic documentation in four 34. Tribouilloy CB, Shen WF, Quere JP, et al. Assessment
patients. Am J Med 1978;64:371–6. of severity of mitral regurgitation by measuring
21. Pai RG, Tarazi R, Wong S. Constrictive pericarditis regurgitant jet width at its origin with transesophageal
causing extrinsic mitral stenosis and a left heart mass. Doppler color flow imaging. Circulation 1992;85:1248–
Clin Cardiol 1996;19:517. 53.
22. Felner JM, Martin RP. The echocardiogram. In: 35. Silver MD, Lam JHC, Ranganathan N, Wigle ED.
Schlant RC, Alexander RW, O’Rourke RA, et al. eds. Morphology of the human tricuspid valve. Circulation
Hurst’s The Heart: Arteries and Veins. New York: 1971;43:333–48.
McGraw Hill, 1994, 375–422. 36. Guyer DE, et al. Comparison of the echocardiographic
23. Roberts WE. Morphological features of the normal and hemodynamic diagnosis of rheumatic tricuspid
and abnormal mitral valve. Am J Cardiol 1983;51: stenosis. J Am Coll Cardiol 1984;3:1135.
1005–28. 37. Lundin L, Landelius J, Andrea B, Oberg K.
24. Chen YT, Kan MN, Chen JS, et al. Contributing Transesophageal echocardiography improves the
factors to formation of left atrial spontaneous echo diagnostic value of cardiac ultrasound in patients with
contrast in mitral valvular disease. J Ultrasound Med carcinoid heart disease. Br Heart J 1990;64:190–4.
1990;9:151–5. 38. Perez JE, Ludbrook PA, Ahumada GG. Usefulness of
25. Baumgartner H, Hung J, Bermejo J, et al. Doppler echocardiography in detecting tricuspid valve
Echocardiographic assessment of valve stenosis: EAE/ stenosis. Am J Cardiol 1985;55:601.
ASE recommendations for clinical practice. J Am Soc 39. Tribouilloy CM, Enriquez-Sarano M, Bailey KR, Tajik
Echocardiogr 2009;22:1–23. AJ, Seward JB. Quantification of tricuspid
26. Feigenbaum H. Acquired valvular heart disease in regurgitation by measuring the width of the vena
Echocardiography, Philadelphia: Lea & Febiger, 1994, contracta with Doppler color flow imaging: a clinical
pp. 239–349. study. J Am Coll Cardiol 2000;36:472–8.
27. Currie PJ, Seward JB, Reeder GS, et al. Continuous- 40. Klein AL, Abbara S, Agler DA, et al. American society
wave Doppler echocardiographic assessment of of echocardiography clinical recommendations for
severity of calcific aortic stenosis: a simultaneous multimodality cardiovascular imaging of patients with
Doppler-catheter correlative study in 100 adult pericardial disease: endorsed by the Society for
patients. Circulation 1985;71:1162. Cardiovascular Magnetic Resonance and Society of
Cardiovascular Computed Tomography. J Am Soc Society of Cardiovascular Anesthesiologists; Society

Echocardiogr 2013;26(9):965–1012. for Cardiovascular Angiography and Interventions;
41. Sagrista-Sauleda J, Merce J, Permanyer-Miralda G, Society of Interventional Radiology; Society of
Soler-Soler J. Clinical clues to the causes of large Thoracic Surgeons; Society for Vascular Medicine.
pericardial effusions. Am J Med 2000;109:95–101. 2010 ACCF/AHA/AATS/ACR/ASA/SCA/SCAI/SIR/
STS/SVM guidelines for the diagnosis and
42. Berge KH, Lanier WL, Reeder GS. Occult cardiac management of patients with thoracic aortic
tamponade detected by transesophageal disease: executive summary. A report of the
echocardiography. Mayo Clin Proc 1992;67:667–70. American College of Cardiology Foundation/
43. Gillam LD, Guyer DE, Gibson TC, et al. American Heart Association Task Force on Practice
Hydrodynamic compression of the right atrium: a new Guidelines, American Association for Thoracic
echocardiographic sign of cardiac tamponade. Surgery, American College of Radiology, American
Circulation 1983;68:294–301. Stroke Association, Society of Cardiovascular
Anesthesiologists, Society for Cardiovascular
44. Wann S, Passen E. Echocardiography in pericardial
Angiography and Interventions, Society of
disease. J Am Soc Echocardiogr 2008;21: 7–13.
Interventional Radiology, Society of Thoracic
45. Himelman RB, Kircher B, Rockey DC, Schiller NB. Surgeons, and Society for Vascular Medicine. J Am
Inferior vena cava plethora with blunted respiratory Coll Cardiol 2010–4;55:e27–e129.
response: a sensitive echocardiographic sign of cardiac
49. Shiga T, Wajima Z, Apfel CC, et al. Diagnostic
tamponade. J Am Coll Cardiol 1988;12:1470–7.
accuracy of transesophageal echocardiography, helical
46. Ballal RS, Nanda NC, Gatewood R, et al. Usefulness of computed tomography, and magnetic resonance
transesophageal echocardiography in assessment of imaging for suspected thoracic aortic dissection:
aortic dissection. Circulation 1991;84:1903–14. systematic review and meta-analysis. Arch Intern Med
47. Simon P, Owen AN, Havel M, et al. Transesophageal 2006;166:1350–6.
echocardiography in the emergency surgical 50. Goldstein SA, Evangelista A, Abbara S, et al.
management of patients with aortic dissection. J Thor Multimodality imaging of diseases of the thoracic aorta
Card Surg 1992;103:1113–8. in adults: from the American Society of
48. Hiratzka LF, Bakris GL, Beckman JA, et al.; American Echocardiography and the European Association of
College of Cardiology Foundation/American Heart Cardiovascular Imaging: endorsed by the Society of
Association Task Force on Practice Guidelines; Cardiovascular Computed Tomography and Society
American Association for Thoracic Surgery; American for Cardiovascular Magnetic Resonance. J Am Soc
College of Radiology; American Stroke Association; Echocardiogr 2015;28:119–82.
54
Chapter
Point-of-Care Transthoracic
6 Echocardiography
Julia Sobol and Oliver Panzer
Introduction blood flow at a specific site, but it is limited by its

inability to measure high velocities (Nyquist limit)
Point-of-care transthoracic echocardiography (TTE)
due to signal aliasing. Continuous-wave Doppler
has become integral to the practice of acute care
(CWD) transmits ultrasound waves continuously,
medicine. It can help assess patients quickly, accu-
allowing measurement of very high velocities without
rately, and noninvasively,1 and allows repeated
the ability to localize the origin.5 Color flow Doppler
examinations to evaluate the effects of
depicts flow velocity and direction, with red moving
interventions.2 The key aspect of this examination
toward and blue away from the probe; optimal ima-
is to rapidly provide diagnostic information to the
ging occurs when blood flow is parallel to the ultra-
acute care practitioner. Therefore, this examination
sound beam.4 Pulsed-wave tissue Doppler imaging
is focused primarily on differentiating normal from
(TDI) evaluates contraction velocity of myocardial
severe cardiac dysfunction as the cause of hemody-
tissue.6
namic instability and does not replace a detailed
Standard views used for the point-of-care TTE
cardiology TTE. Some educators have argued that
examination consist of parasternal long-axis
formal, basic perioperative ultrasound training
(PLAX), parasternal short-axis (PSAX), apical four-
should be included in anesthesiology residency,
(A4 C) and five-chamber (A5 C), and subcostal long-
with more advanced ultrasound teaching during spe-
(SLAX) and short-axis (SSAX). Information that may
cific fellowship training.3 In this chapter, we will first
be acquired from each view is summarized in Table
discuss how point-of-care TTE works in terms of
6.1.
logistics and diagnostic capabilities and then exam-
ine its effect on patient care.
Parasternal Long-Axis View
TTE Views and Possible Assessments The PLAX view allows visualization of the left
atrium (LA), left ventricle (LV), mitral valve
in Each View (MV), aortic valve (AV), aortic root and LV out-
Point-of-care TTE utilizes a phased-array, low- flow tract (LVOT), part of the right ventricle (RV),
frequency (2.5–3 MHz) probe with a high frame rate and the descending aorta (descAO). This view is
(at least 24 frames/second). Imaging relies on B-mode obtained on the left side of the sternum at or near
ultrasonography in which high-density, high- the third intercostal space along the long axis of
impedance structures appear hyperechoic, and low- the heart.4
impedance, fluid-filled structures appear anechoic. Left ventricle systolic function may be evaluated
Other imaging modalities to assess cardiac in the PLAX view by visual estimation, fractional
physiology include M-mode and Doppler shortening, and E-point septal separation (EPSS).
ultrasonography. M-mode focuses on one anatomical Visual estimation of the change in volume in the
area over time, whereas Doppler examines the motion LV between systole and diastole can differentiate
of fluid (pulsed-wave, continuous-wave, and color - normal contractility from mildly, moderately, or
flow Doppler) or structures (tissue Doppler) in rela- severely decreased LV function as well as from
tion to the ultrasound probe, with optimal imaging hyperdynamic motion.4 Both endocardial border
when movement is parallel to the ultrasound beam.4 movement toward the center of the LV cavity and
Pulsed-wave Doppler (PWD) sends out ultrasound increased wall thickness during systole help assess
waves in pulses and is able to measure the velocity of LV systolic function visually.2
Point-of-Care Transthoracic Echocardiography
Table 6.1 Summary of the information and data acquired from each point-of-care transthoracic echocardiographic view.
PLAX PSAX A4 C A5 C Subcostal IVC

LV Systolic Function
Visual estimation ++ ++ ++ ++
Fractional shortening ++ ++
LV FAC ++
Stroke Volume
LVOT ++
VTI ++
EPSS ++
RWMA + ++ + + +
LV Diastolic Function
LVH ++ + + + +
Mitral E/A ++
Mitral E/e ++
RV Function
Visual estimation + + ++ + ++
TAPSE ++
RVSP/sPAP ++ ++
RVH + ++
RAP of CVP ++
Valves
Aortic 2D ++ ++ ++
Doppler ++
Mitral 2D + ++
Doppler ++
Tricuspid 2D ++ ++
Doppler ++
Pulmonary 2D ++
Doppler ++
AV, aortic valve; CVP, central venous pressure; EPSS, E-point septal separation; FAC, fractional area change; IVC, inferior vena cava; LV, left
ventricle; LVH, left ventricular hypertrophy; LVOT, left ventricular outflow tract; MV, mitral valve; PV, pulmonic valve; RAP, right atrial pressure;
RV, right ventricle; RVH, right ventricular hypertrophy; RVSP, right ventricular systolic pressure; RWMA, regional wall motion abnormality;
sPAP, systolic pulmonary artery pressure; SV, stroke volume; TAPSE, tricuspid annular plane systolic excursion; TV, tricuspid valve; VTI,
velocity-time integral.
Left ventricle function may be evaluated semi- measurement of LV contractility in the PLAX view is
quantitatively with the M-mode-based method of EPSS, which correlates well with LV ejection fraction8
fractional shortening (FS), which compares the end- (Figure 6.1). Normally, the anterior leaflet of the
diastolic and end-systolic diameters of the LV cavity.4 mitral valve (ALMV) approaches the LV septal wall
The accuracy of FS may be diminished by misalign- when the valve opens during diastole, but with LV
ment of the M-mode cursor and by any abnormality dilation and decreased LV contractility, the distance
in the examined LV segment shape or movement,7 between the open ALMV and the interventricular
limiting its utility in predicting LV function. Another septum increases. E-point septal separation is
56
RV
IVS
E A Anterior leaflet
Poterior leaflet
LVPW
Figure 6.1 E-point septal separation (EPSS) in the parasternal long-axis view of a patient with severe left ventricular systolic dysfunction. EPSS
> 8 mm is associated with systolic dysfunction. IVS, interventricular septum; LVPW, left ventricular posterior wall; RV, right ventricle. (A black and
white version of this figure will appear in some formats. For the color version, please refer to the plate section.) Reprinted with permission from
McKaigney CJ, Krantz MJ, La Rocque CL, et al. E-point septal separation: a bedside tool for emergency physician assessment of left ventricular
ejection fraction. Am J Emerg Med 2014;32:493–7.
measured with M-mode to evaluate the distance anterior to the descAO, differentiating the fluid from
between the E wave (the first wave that shows passive a pleural effusion, which lies posterior to the descAO.4
ventricular filling) and the LV septum; EPSS greater
than 1 cm is associated with decreased LV contracti- Parasternal Short-Axis View
lity. It is inaccurate in cases of mitral stenosis or
By rotating the ultrasound probe 90° clockwise from
regurgitation, aortic regurgitation, or severe LV
the PLAX view, the PSAX view will appear with the
hypertrophy (LVH).4
LV in cross-section. By fanning the probe from the
The PLAX view allows assessment of ventricular
base to the apex of the heart, the ultrasound beam can
wall thickness, AV, MV, and LVOT. M-mode or cal-
visualize several levels, including the AV and mid-
iper measurements can evaluate LV wall thickness. Left
papillary levels. At the AV level, the AV, both atria,
ventricular hypertrophy is associated with LV wall
tricuspid valve (TV), RV, pulmonic valve (PV), and
thickness greater than 1 cm,4 with severe LVH char-
pulmonary artery (PA) may be seen.4 The AV should
acterized by LV wall thickness greater than 1.5 cm.7 In
be assessed visually in this view for leaflet calcification
the PLAX view, RV-free wall thickness greater than
and mobility to evaluate for AS.11 In addition, as flow
5 mm implies that RV afterload is likely chronically
through the TV and PV is aligned parallel to the
high.9,10 The valves should also be examined for leaflet
ultrasound beam, tricuspid and pulmonic regurgitant
thickness and excursion, as well as calcifications that
jets may be assessed by Doppler.9
might suggest aortic stenosis (AS)11 or mitral valve
At the mid-papillary muscle level, the LV short-
pathophysiology such as systolic anterior motion of
axis view can be utilized to estimate LV systolic func-
the MV.12 Mitral and aortic regurgitation can also be
tion visually or by FS. Alternatively, the percentage
detected in this view by color Doppler but should be
change between end-diastolic and end-systolic LV
evaluated in other views as well.1 The LVOT area may
area can be measured as the fractional area change
be estimated in this view as part of stroke volume (SV)
(FAC).13 The formula for FAC is:
and cardiac output (CO) calculations. The LVOT dia-
meter is measured between the right and non-coronary FAC(%) = 100 × (LV end-diastolic area – LV end-
cusps of the AV during systole, and the LVOT area is systolic area)/(LV end-diastolic area)
calculated from the diameter (Figure 6.2a). The PLAX
view may also show a pericardial effusion appearing as Normal values are 36–64%.13 Area measurements,
an anechoic region at the pericardial border tracking however, assume normal ventricular geometry.7 In
Figure 6.2(a) Close-up of the left

ventricular outflow tract (LVOT) in the
parasternal long-axis view with the cursor
measuring the LVOT diameter during
systole. LVOT area is calculated from the
diameter by the formula:
LVOTarea = π (diameter/2)2.
Figure 6.2(b) Apical five-chamber view

showing pulsed-wave Doppler (PWD) at the
left ventricular outflow tract (LVOT). The
velocity-time integral (VTI) function on the
ultrasound machine traces the area under
the PWD curve, which, when multiplied by
the area of the LVOT obtained in the
parasternal long-axis view, allows
calculation of left ventricular stroke volume.
Stroke volume multiplied by heart rate
determines cardiac output. (A black and
white version of this figure will appear in
some formats. For the color version, please
refer to the plate section).
case of myocardial ischemia or infarction, regional chambers and their relative size and function, and this is
wall motion abnormalities can be detected as the the best view for Doppler evaluation of the MV and TV
entire coronary distribution may be seen in this as the direction of blood flow is aligned parallel to the
view. The mid-papillary muscle PSAX view is also ultrasound probe. By angling the probe more superiorly
helpful in assessing RV pressure or volume overload. from the A4 C view, the A5 C view is obtained, with the
As RV pressure increases, the interventricular septum LVOT and AV as the fifth chamber.4 The apical views
(IVS) bows into the LV during systole, and as RV are important in assessing LV and RV function as well
volume increases, the IVS moves toward the LV dur- as in evaluating valvulopathies.
ing diastole. Both of these scenarios lead to One purpose of the A5 C view is to estimate SV
a D-shaped LV in the PSAX view.10 and CO.4 The SV is calculated by the formula:
SV = LVOTarea × VTI
Apical Four- and Five-Chamber Views
The apical views are obtained by placing the probe at the The velocity-time integral (VTI) is the area under the
point of maximal impulse. The A4 C view shows all four PWD curve at the LVOT (Figure 6.2b), representing
58
Figure 6.3(a) Apical four-chamber view

with pulsed-wave Doppler (PWD) across
the mitral valve showing biphasic
diastolic flow. The E wave represents
early passive filling, whereas the A wave
shows atrial contraction during late
diastole. Peak E- and A-wave velocities
are used to calculate the E/A ratio. (A
black and white version of this figure will
appear in some formats. For the color
version, please refer to the plate section).
blood flow moving through the LVOT in each cardiac and preload affect the E wave, while LV compliance
cycle.4 Velocity-time integral measurements should and LA contraction determine the A wave.14 By cal-
be repeated at least three times to improve culating the mitral E/A ratio, different filling patterns
accuracy.13 The LVOT area is obtained from measur- may be described. An E/A ratio ≥ 2 in patients older
ing the AV diameter in the PLAX view as described than 40 years usually signifies severe diastolic dys-
above.4 Alternatively, Simpson’s method is a different function. However, normal and pseudo-normal filling
way to calculate LV ejection fraction using multiple patterns are difficult to distinguish from each other,
apical views, but it is time-consuming and requires especially in patients with normal ejection fraction.6
clear endocardial borders with perfect orientation,10 Furthermore, filling patterns are affected by patient
limiting its utility in point-of-care TTE. age, hypertrophic or restrictive cardiomyopathy,
The A5 C view also allows for the assessment of mitral valve disease, and mitral valve repair/replace-
AV leaflet calcifications and excursion. The degree of ment, and they cannot be utilized in patients with
AS may be estimated quantitatively by placing the atrial arrhythmias.6,10
CWD cursor across the AV, which may show elevated If the E/A ratio is inconclusive or if the patient
peak velocities that indicate moderate (3–4 m/sec) to has normal LV function, it may be helpful to
severe (> 4 m/sec) AS. However, as the peak AV obtain the mitral annular peak velocity via TDI of
velocity depends on LV contractility, the degree of the mitral annulus10 (Figure 6.3b). TDI evaluates
AS may be underestimated in patients with severe early (e’) and late (a’) mitral annular diastolic
LV dysfunction.11 excursion. Peak e’ velocity is due primarily to LV
The A4 C view is helpful for assessing LV diastolic relaxation, independent of preload.14 The E/e’
function and RV function. Impaired LV relaxation ratio – a combination of PWD and TDI measure-
and/or elevated LV stiffness contribute to diastolic ments in early diastole – is less dependent on LV
dysfunction.14 The degree of LV diastolic dysfunction function and age than other indices of diastolic
can be evaluated using PWD across the MV and TDI function.6,14 If E/e’ is less than 8, the LV filling
at the MV annulus.10 On PWD, diastolic flow across pressure is normal, but if the ratio is greater than
the MV is biphasic, with the E wave occurring during 14, the LV filling pressure is high.6 High LV filling
early passive filling and the A wave during late dia- pressure identifies patients at risk of cardiogenic
stole due to atrial contraction.6 The PWD is placed pulmonary edema.10 Other markers of LV diastolic
between the MV leaflet tips to measure peak E-wave dysfunction may include LA enlargement (a sign of
and A-wave velocities10 (Figure 6.3a). LV relaxation elevated LA pressure in patients without atrial
Figure 6.3(b) Apical four-chamber view

with tissue Doppler imaging (TDI) of the
mitral annulus showing early (e’) and late (a’)
excursion of the mitral annulus during
diastole. The E/e’ ratio utilizing peak e’
velocity is another measure of left
ventricular diastolic function. (A black and
white version of this figure will appear in
some formats. For the color version, please
refer to the plate section).
arrhythmias or MV disease) and/or LVH, which is systolic pressure (RVSP) using the simplified
often associated with diastolic dysfunction.6 Bernoulli equation and RA pressure (RAP):
The A4 C view also allows the assessment of the
RV. RV pressure or volume overload may lead to RV RVSP = 4(Vmax)2 + RAP
dilation and dysfunction.9 RV dilation may be evalu-
where Vmax is the peak TR regurgitant jet velocity
ated by comparing LV and RV size.4 Normally, the
and RAP is obtained from inferior vena cava (IVC)
end-diastolic RV cavity is 60% of the end-diastolic LV
measurements in the subxiphoid view9 or from
size. RV cavity size greater than 100% the size of the
a central venous catheter.10 RVSP and sPAP are
LV denotes severe RV enlargement.10 While the LV
equivalent if there is no gradient across the PV or
normally occupies the apex of the heart, RV dilation
RVOT.9
may displace the LV and take over the apex instead.9
The apical window is also especially helpful for
As RV free-wall contraction relies on longitudinal
diagnosing tamponade, as pericardial effusion with
movement, tricuspid annular plane systolic excursion
diastolic collapse of the RA or RV may be visualized
(TAPSE) correlates well with RV systolic function.10
easily in this view.4
TAPSE is measured in M-mode, with values less than
16 mm associated with decreased RV systolic
function9 (Figure 6.4). Subcostal Long- and Short-Axis Views
Tricuspid regurgitation (TR), which is often asso- The subcostal views are particularly useful to evaluate
ciated with RV dysfunction, can be assessed by color- cardiac morphology, the pericardium, and the IVC.16
flow Doppler and allows estimation of PA systolic They are obtained by pointing the ultrasound beam
pressure (sPAP).10 The maximal TR jet velocity of toward the heart under the sternum by using the liver
blood is converted to a pressure gradient across the as an acoustic window, thereby avoiding lung inter-
TV by the modified Bernoulli equation.15 The CWD is ference. The SLAX view allows visual assessment of
positioned along the TR jet main axis to calculate both ventricles and atria as well as the pericardial
blood flow velocity and then pressure across the TV space, similar to the A4 C view. Given its proximity
is calculated10 (Figure 6.5). For greater accuracy of to the ultrasound probe, the RV free wall is easily
sPAP calculation, the spectral profile should be dense assessed in this view,4 with RV free wall thickness
with well-defined borders and the TR jet should be greater than 5 mm suggesting RV hypertrophy.9 By
examined from multiple views as angle affects velocity rotating the probe counterclockwise into a sagittal
measurements. Peak TR jet velocity estimates RV plane, the SSAX view is obtained, which resembles
60
Figure 6.4 Tricuspid annular plane systolic

excursion (TAPSE) measured in the apical
four-chamber view with M-mode. Values <
16 mm are associated with decreased right
ventricular systolic function.
Figure 6.5 Continuous-wave Doppler

(CWD) of tricuspid regurgitation (TR) in the
apical four-chamber view. Maximal TR jet
velocity can be used to calculate right
ventricular systolic pressure using the
simplified Bernoulli equation and right atrial
pressure. (A black and white version of this
figure will appear in some formats. For the
color version, please refer to the plate
section).
the PSAX view.5 While the subcostal views provide immediately proximal to the hepatic vein entry,9
images similar to the A4 C and PSAX views, the although some recommend measuring it 2 cm caudal
ultrasound beam does not align parallel to the direc- to the entry of the hepatic vein.17 Elevated RAP leads to
tion of blood flow in the subcostal position, limiting IVC dilation and reduced IVC collapse with inspira-
its utility in calculating hemodynamic parameters.16 tion during spontaneous ventilation. If the IVC dia-
In the subcostal position, the IVC may be visua- meter is ≤ 2.1 cm with inspiratory collapse greater than
lized in short and long axes to measure diameter and 50% with sniffing, RAP is likely 0–5 mmHg, while
respiratory variability, which help estimate central a diameter greater than 2.1 cm with collapse less than
venous pressure (CVP).4 The IVC should be assessed 50% indicates RAP 10–20 mmHg. IVC diameter

showing a pericardial effusion during
systole.

illustrating a pericardial effusion
accompanied by right ventricular diastolic
collapse, suggesting tamponade.
should be measured at end-expiration. In mechanically act as an acoustic window for an alternative view of the
ventilated patients, RAP should be measured by pres- IVC from a lateral, mid-axillary probe position.4
sure transduction as the IVC may be dilated with no
respiratory variation. However, if the IVC diameter is Impact of Point-of-Care TTE on Patient
≤ 1.2 cm in these patients, RAP is probably
< 10 mmHg.9 Inferior vena cava size and collapsibility Outcomes
measures should not be used in patients with elevated Increased utilization of point-of-care TTE could be
intra-abdominal pressure.2 In addition, the liver may of great benefit, particularly in patients at risk of
62
life-threatening shock or cardiac arrest. In non-trauma than 5 mm, the segment is contracting abnormally.10
emergency room patients with undifferentiated Evaluation for RWMA should focus on thickening
hypotension, goal-directed TTE has been shown to rather than motion, because tethering to nearby wall
help guide treatment interventions, narrow the dif- segments can affect adjacent wall motion.7
ferential diagnosis,18 and improve diagnostic Point-of-care TTE is also helpful in distinguishing
accuracy.19 In the perioperative setting, anesthesiol- different forms of obstructive shock. Tamponade pre-
ogists performing TTE prior to urgent surgical pro- sents with pericardial effusion plus RA or RV diastolic
cedures discovered unexpected pathology that collapse, often best viewed in the A4 C or subcostal
affected anesthetic management in high-risk windows (Figures 6.6a and 6.6b). RV diastolic collapse
patients.20,21 In addition, even brief training of prac- is more specific for tamponade than RA collapse, and
titioners allows accurate identification of myocardial progresses with the degree of tamponade from
dysfunction when compared to formal TTE.22 While a subtle inward movement of the RV free wall to
data have not shown that point-of-care TTE complete collapse.4 An increase in intra-pericardial
improves patient outcomes or mortality,20,21 point- pressure with pericardial fluid accumulation deter-
of-care ultrasound can certainly help determine the mines whether tamponade physiology occurs, rather
type and possibly the cause of shock,23 potentially than the size of the pericardial effusion itself.2 The
allowing earlier, more appropriate intervention and IVC is usually enlarged with no respiratory
treatment. variation.24 In the A4 C view, PWD through the MV
In patients with hypovolemic or distributive and TV may identify changes in valvular flow with the
shock, the ventricles are usually both small and some- respiratory cycle; tamponade physiology is suggested
times hyperkinetic.23 Obliteration of the LV cavity at by an increase in flow of greater than 25% across the
end-systole with vigorous systolic wall motion indi- TV with inspiration, or a decrease in flow of greater
cates hyperdynamic function, hypovolemia, and low than 15% across the MV with inspiration.4
afterload.2 As both hypovolemic and distributive Other forms of obstructive shock are primarily
shock are associated with an empty LV in systole, due to RV dysfunction. Acute inferior wall myocar-
the LV size in diastole may help distinguish the two dial ischemia or infarction could cause RV failure,
entities: small LV diastolic size suggests hypovolemia, which might present with RV dilation and dysfunc-
while fuller LV cavity diastolic size may indicate dis- tion without elevation of sPAP.10 The LV posterior or
tributive shock due to low afterload.12 In either type of inferior free wall may also show wall motion abnorm-
shock, the diameter of the IVC will likely be small with alities. Acute cor pulmonale with an abrupt increase
collapsibility through the respiratory cycle.24 The sub- in RV afterload is most often due to pulmonary embo-
set of patients with hypovolemic and/or distributive lism (PE) or acute respiratory distress syndrome
shock due to sepsis may exhibit unique findings on (ARDS),25 with an incidence of approximately
point-of-care TTE, such as LV systolic and diastolic 30–55% of patients with PE and 25% of patients with
dysfunction, as well as RV dysfunction with dilated ARDS.15 Acute cor pulmonale due to PE would likely
IVC.2,23 exhibit RV failure, elevated sPAP, and flattening of
Cardiogenic shock presents with poor LV contrac- the interventricular septum, best viewed in the PSAX
tility and function as assessed by visual estimation, FS, view10,24 (Figure 6.7). The IVC is often dilated with-
FAC, EPSS, and calculation of CO.4 Diastolic heart out respiratory variability, and clot may be visible in
failure would be associated with changes in flow and the right side of the heart or the PA.4,24 With a large,
motion at the MV, as discussed above.6 Assessing for acute PE, McConnell’s sign may be present, in which
regional wall motion abnormality (RWMA) in the LV the RV base and free wall become hypokinetic, but the
may identify myocardial ischemia or infarction. The apex contracts normally.24
PSAX view is the most suitable for RWMA assessment To distinguish acute from chronic RV dysfunc-
as it allows visualization of all LV wall segments, tion, RV wall thickness may be measured; RV wall
which can each be inspected individually. Wall diameter greater than 6 mm indicates chronic RV
motion may be deemed normal/hyperkinetic, hypo- strain.24 Increased RV diastolic wall thickness may
kinetic (decreased wall thickening), akinetic (minimal occur in response to a sudden rise in RV afterload in
thickening), or dyskinetic (systolic thinning).7 If the as little as 48 hours. In addition, marked trabecula-
wall thickens less than 50%, or if wall excursion is less tions within the RV cavity often accompany chronic
return of spontaneous circulation (ROSC) than no

cardiac activity.28 A recent large, multicenter, pro-
spective observational study showed that non-
traumatic cardiac arrest patients who presented to
the emergency department with cardiac activity on
point-of-care TTE at the beginning of resuscitative
efforts had significantly improved odds of survival
than those with no cardiac activity.29 A secondary
analysis of this study showed that the type of cardiac
activity seen may also be important: patients with
disorganized cardiac activity (agonal or twitching
myocardial activity on TTE) had significantly lower
rates of ROSC compared to those with organized
activity (synchronized ventricular wall motion with
change in the size of the ventricular cavity).30
However, utilization of point-of-care TTE in cardiac
Figure 6.7 Parasternal short-axis view of right ventricular pressure/ arrest has not been shown to affect patient out-
volume overload with flattening of the interventricular septum in comes. In a small trial of cardiac arrest patients
a patient with acute cor pulmonale. The left ventricle appears randomized to TTE or no TTE during resuscitation
D-shaped.
efforts, those who underwent TTE did not have
improved rates of ROSC or survival compared to
cor pulmonale.25 Chronic conditions that may be those without TTE, even though potentially reversi-
associated with RV dysfunction include obstructive ble causes of cardiac arrest were found in those who
sleep apnea, chronic obstructive pulmonary disease, were evaluated by TTE.31
pulmonary hypertension, and chronic pulmonary Although other disorders may precipitate circula-
emboli.4,24 Pulmonary hypertension would be sug- tory failure, the pathophysiologic states described here
gested by sPAP ≥ 35 mmHg, peak TR jet velocity > represent some potentially reversible causes of shock
2.8 m/sec, TAPSE < 16 mm, RV dilation, and esti- that have widely different therapeutic interventions
mated RAP > 15 mmHg.26 Finally, it is important to and management. Point-of-care TTE can help distin-
remember that the RV may become dilated and dys- guish these causes of shock and expeditiously narrow
functional due to primary LV dysfunction.9 the differential diagnosis in shock and arrest situations.2
Point-of-care TTE may also be useful in cardiac A focused TTE examination could take as little as
arrest. An international expert panel in emergency 1 minute and may be performed by trained novices.23
medicine recommends that point-of-care TTE be per- While point-of-care TTE is fast becoming essential to
formed during cardiac arrest while pausing for the practice of acute care medicine and anesthesiology,3
rhythm checks, thereby minimizing any interruptions future research in point-of-care TTE should focus on
of chest compressions. The subcostal view is pre- whether patient outcomes are actually improved by its
ferred, with the PLAX view as an alternative window. use.
The initial core views should focus on potentially
reversible causes of cardiac arrest such as tamponade, References
hypovolemia, RV failure, and myocardial dysfunc-
1. Shillcutt SK, Bick JS. A comparison of basic
tion, as well as evaluation of organized cardiac activity transthoracic and transesophageal echocardiography
or asystole. Supplementary views including the apical views in the perioperative setting. Anesth Analg
window, the IVC, and the lungs should be performed 2013;116:1231–6.
if time allows and if further information is required.27 2. Griffee MJ, Merkel MJ, Wei KS. The role of
Performing TTE during cardiac arrest may also echocardiography in hemodynamic assessment of septic
help with prognostication. One systematic review shock. Crit Care Clin 2010;26:365–82.
showed that cardiac activity detected on point-of- 3. Mahmood F, Matyal R, Skubas N, et al. Perioperative
care TTE during cardiac arrest and resuscitation ultrasound training in anesthesiology: a call to action.
efforts was associated with a much higher rate of Anesth Analg 2016;122:1794–804.
64
4. Perera P, Lobo V, Williams SR, et al. Cardiac 18. Shokoohi H, Boniface KS, Pourmand A, et al.
echocardiography. Crit Care Clin 2014;30:47–92. Bedside ultrasonography reduces diagnostic
5. Sturgess DJ. Transthoracic echocardiography: an uncertainty and guides resuscitation in patients
overview. In: Lumb P, Karakitsos D, eds. Critical Care with undifferentiated hypotension. Crit Care Med
Ultrasound. Philadelphia: Elsevier Saunders, 2015;43:2562–9.
2015;139–45. 19. Jones AE, Tayal VS, Sullivan M, et al. Randomized,
6. Nagueh SF, Smiseth OA, Appleton CP, et al. controlled trial of immediate versus delayed
Recommendations for the evaluation of left ventricular goal-directed ultrasound to identify the cause of
diastolic function by echocardiography. Eur Heart nontraumatic hypotension in emergency department
J Cardiovasc Imaging 2016;17:1321–60. patients. Crit Care Med 2004;32:1703–8.
7. Lang R, Badano L, Mor-Avi V, et al. Recommendations 20. Bøtker MT, Vang ML, Grøfte T, et al. Routine
for cardiac chamber quantification by pre-operative focused ultrasonography by
echocardiography in adults: an update from the anesthesiologists in patients undergoing urgent
American Society of Echocardiography and the surgical procedures. Acta Anaesthesiol Scand
European Association of Cardiovascular Imaging. Eur 2014;58:807–14.
Heart J Cardiovasc Imaging 2015;16:233–70. 21. Canty DJ, Royse CF, Kilpatrick D, et al. The impact
8. McKaigney CJ, Krantz MJ, La Rocque CL, et al. of pre-operative focused transthoracic
E-point septal separation: a bedside tool for emergency echocardiography in emergency non-cardiac
physician assessment of left ventricular ejection surgery patients with known or risk of cardiac
fraction. Am J Emerg Med 2014;32:493–7. disease. Anaesthesia 2012;67:714–20.
9. Rudski LG, Lai WW, Afilalo J, et al. Guidelines for the 22. Melamed R, Sprenkle MD, Ulstad VK, et al.
echocardiographic assessment of the right heart in Assessment of left ventricular function by intensivists
adults: a report from the American Society of using hand-held echocardiography. Chest
Echocardiography. J Am Soc Echocardiogr 2009;135:1416–20.
2010;23:685–713. 23. De Backer D. Ultrasonic evaluation of the heart. Curr
10. Narasimhan M, Koenig SJ, Mayo PH. Advanced Opin Crit Care 2014;20:309–14.
echocardiography for the critical care physician, Part 2. 24. Wu TS. The CORE scan. Crit Care Clin
Chest 2014;145:135–42. 2014;30:151–75.
11. Cowie B, Kluger R. Evaluation of systolic murmurs 25. Vieillard-Baron A, Prin S, Chergui K, et al. Echo-
using transthoracic echocardiography by anaesthetic Doppler demonstration of acute cor pulmonale at the
trainees. Anaesthesia 2011;66:785–90. bedside in the medical intensive care unit. Am J Respir
12. Zimmerman JM, Coker BJ. The nuts and bolts of Crit Care Med 2002;166:1310–9.
performing focused cardiovascular ultrasound 26. Hellenkamp K, Unsöld B, Mushemi-Blake S, et al.
(FoCUS). Anesth Analg 2017;124:753–60. Echocardiographic estimation of mean pulmonary
13. Slama M, Maizel J. Echocardiographic measurement of artery pressure: a comparison of different
ventricular function. Curr Opin Crit Care approaches to assign the likelihood of pulmonary
2006;12:241–8. hypertension. J Am Soc Echocardiogr 2017;31
(1):89–98.
14. Maizel J, El-Dash S, Slama M. Evaluation of left
ventricular diastolic function in the intensive care unit. 27. Atkinson P, Bowra J, Milne J, et al. International
In: Lumb P, Karakitsos D, eds. Critical Care Federation for Emergency Medicine Consensus
Ultrasound. Philadelphia: Elsevier Saunders, Statement: Sonography in hypotension and cardiac
2015;175–8. arrest (SHoC): an international consensus on the use of
point of care ultrasound for undifferentiated
15. Krishnan S, Schmidt GA. Acute right ventricular hypotension and during cardiac arrest. CJEM
dysfunction. Chest 2015;147:835–46. 2017;19:459–70.
16. Maizel J, Salhi A, Tribouilloy C, et al. The subxiphoid 28. Blyth L, Atkinson P, Gadd K, et al. Bedside focused
view cannot replace the apical view for transthoracic echocardiography as predictor of survival in cardiac
echocardiographic assessment of hemodynamic status. arrest patients: a systematic review. Acad Emerg Med
Crit Care 2013;17:R186. 2012;19:1119–26.
17. Wallace DJ, Allison M, Stone MB. Inferior vena cava 29. Gaspari R, Weekes A, Adhikari S, et al. Emergency
percentage collapse during respiration is affected by department point-of-care ultrasound in out-of-
the sampling location: an ultrasound study in healthy hospital and in-ED cardiac arrest. Resuscitation
volunteers. Acad Emerg Med 2010;17:96–9. 2016;109:33–9.
30. Gaspari R, Weekes A, Adhikari S, et al. A retrospective 31. Chardoli M, Heidari F, Rabiee H, et al.
study of pulseless electrical activity, bedside ultrasound Echocardiography integrated ACLS protocol versus
identifies interventions during resuscitation associated conventional cardiopulmonary resuscitation in
with improved survival to hospital admission. patients with pulseless electrical activity cardiac arrest.
Resuscitation 2017;120:103–7. Chin J Traumatol 2012;15:284–7.
66
Chapter
Point-of-Care Lung Ultrasound
7 Zachary Kuschner and John M. Oropello
Introduction lie deep within the thorax.14 The three most common
ultrasound transducers are the phased array, linear,
Lung ultrasound was initially thought to be impossi-
and curvilinear probes. They differ with respect to
ble due to the acoustic impedance of air. In the 1960s
their piezoelectric crystal arrangement, aperture
it was demonstrated that ultrasound could be used to
(footprint), and frequency range. Lung ultrasound is
image pleural effusions due to their fluid composition
performed with phased array or linear probes. The
and proximity to the chest wall and diaphragm.1,2
curvilinear (curved linear) probe is generally not used
Subsequent work in the 1980s established that in a
in thoracic ultrasound; because the crystals are orga-
normal thoracic cavity the air–fluid interface and the
nized in an arc, it results in a wide footprint that
presence of two mobile opposing pleural surfaces
makes it difficult to find adequate sonographic win-
generated stereotypical imaging artifacts. Deviations
dows between the ribs.14
from these normal artifacts correspond to specific
There are several modes of processing and display-
pathologic processes,3,4 and a systematic approach to
ing the ultrasonic echo reflections, including A-mode
assessing and monitoring these pathologies in criti-
(amplitude mode), B-mode (brightness mode, also
cally ill patients has been developed.5 Bedsides, lung
known as 2D mode), and M-mode (motion mode).
sonography can be adapted for use by anesthesiolo-
The phased array probe emits low ultrasound fre-
gists in the perioperative care setting,6 where pulmon-
quency waves in the 2–8 MHz range.14 The grouping
ary complications occur in 5–13% of patients and
of crystals in the phased array probe yields a fan-
significantly increase mortality and the cost of
shaped footprint (Figure 7.1a). A variant of the
care.7–9
phased array probe is the rectangular-shaped “car-
The standard method of assessing pulmonary
diac” phased array probe. This probe can be used to
pathology in the perioperative setting remains the
image lung as well as other structures, such as heart,
chest x-ray. Chest x-rays, especially when performed
abdomen, and inferior vena cava; utilizing this single
portably, have significant diagnostic limitations and
probe permits easy transition between examinations
take significantly longer to obtain than bedside sono-
of multiple sites.
graphy.10,11 Lung ultrasound may yield information
The linear probe contains crystals arrayed in a
not obtained by, and in some situations may even
straight line and emits acoustic waves in higher fre-
substitute for, a chest x-ray. Lung ultrasound can
quencies ranging from 5 to 13 MHz.14 The grouping
also be used to determine volume status and guide
of crystals in the linear probe yields a rectangular-
management.12,13 This chapter describes the techni-
shaped footprint (Figure 7.1b). The higher frequen-
ques involved in the application of lung ultrasound to
cies result in better resolution but have less tissue
the monitoring and management of perioperative
penetration. The linear probe can occasionally clarify
patients.
superficial artifacts if the phased array probe images
are indeterminate. It is also the best probe for imaging
Techniques the trachea, as when performing percutaneous
Medical diagnostic ultrasound technologies transmit tracheostomies.
frequencies in the 1–20 megahertz (MHz) range. Low Lung sonographic exams, similar to other noncar-
ultrasound frequencies result in lower resolution diac sonographic exams, are performed with the
images than higher frequencies, but have greater probe marker oriented cephalad when scanning in
penetration and permit evaluation of structures that longitudinal orientation, and toward the patient’s
(a) (b) Figure 7.1 (a) Phased array

with fanned output. (b) Linear
probe with rectangular output.
right side when scanning in transverse orientation. interface in between the two ribs.12 This should be
When using M-mode, the provider should set the repeated in the midaxillary line in both the superior
scan speed to slow, because sonographic studies and inferior hemithorax. The diaphragm should be
using the medium or fast speeds may produce repeat- visualized when examining the inferior hemithorax in
ing horizontal lines that give the erroneous impres- the midaxillary line.
sion of pneumothorax (Figure 7.2).14 The gain should
be adjusted to permit visualization of the artifacts
without obscuring the adjacent pulmonary tissue.
Artifacts
The thorax may be divided into superior and Unlike traditional sonography in which structures
inferior zones, each of which are further divided into and pathology are visualized directly, pulmonary
anterior, lateral, and posterior regions defined by the sonography relies on ultrasonic artifacts created by
anterior and posterior axillary lines (Figure 7.3).5 The the interface between air contained within the lungs
probe should be placed in a longitudinal (sagittal) and fluid within tissue. This interface, in combination
orientation in the midclavicular line, producing an with the phasic sliding between the visceral and par-
image in which both the superior and inferior ribs ietal pleura that accompanies respiration, creates
are in view, as well as the soft tissue and air–pleural stereotypical artifacts when imaging normal lung.
68
Chest wall
Pleural line
Lung
Slow Medium Fast

Figure 7.2 M-mode sweep speed settings. As the sweep speed is increased from slow to medium and then from medium to fast, more
horizontal lines are generated giving the false appearance (at medium and fast sweep speeds) of absent pleural sliding (bar-code sign). On the
proper slow sweep setting, horizontal lines appear only in the chest wall, not below the pleural line where the lung has a sandy appearance
indicating pleural sliding (seashore sign).
Pleural line
Anterior axillary line
Posterior axillary line Rib shadows

A lines
Figure 7.3 Thoracic probe placement for lung ultrasonography.

The thorax may be divided into regions: superior (width of patient’s
hand beginning from right below the clavicle), inferior (another
hand’s breadth below the superior hand), anterior (anterior to the
anterior axillary line), lateral (in between the anterior and posterior
axillary lines), and posterior (posterior to the posterior axillary line).
Probe placement in the superior anterior region below the
midclavicular line is the first place to look for a pneumothorax in a
supine patient. As a pneumothorax enlarges, it moves laterally and
then posteriorly. Pleural effusions usually begin posteriorly, moving
Figure 7.4 Normal ultrasound artifacts present in healthy lung
laterally and anteriorly as they enlarge. In critically ill patients with
tissue. 2D lung still image captured by a linear probe in the
abdominal distension, or on expiration, most of the lung may be
sagittal plane, midclavicular line, 2nd interspace. The pleural line
close to or above the nipple level. Begin to search for the diaphragm
consists of the parietal and visceral pleura. The A-lines are
laterally and at the nipple line.
reverberation artifacts off the pleural line at a distance equal to
the distance between the ultrasound probe and the pleural line.
Identification of these normal artifacts can be used to A-lines indicate “air.” The anechoic borders are produced by rib
rule out pathology. Absence of normal artifacts and shadows. If this were a video, the pleural line would twinkle or
shimmer, indicating pleural sliding.
identification of abnormal artifacts can be used to rule
in pathology.
d5fe4b031a76584e558/577dba7a1b631b1456316e53/
Normal Artifacts 577dba7a37c581b33317e898/1467857682349/normal
The normal artifacts that are indicative of healthy lungslide.gif?format=500w).
lung tissue are lung sliding (also known as pleural The lung sliding artifact is a shimmering appear-
sliding), A-lines, the bat sign, and the seashore sign. ance of the lung-chest wall border caused by move-
Lung sliding (also known as pleural sliding) and ment of the visceral and parietal pleurae against one
A-lines can both be visualized using the longitudinal another.3,15 A-lines are repeating, equidistantly
orientation in the 2D mode (see Figure 7.4 and video spaced horizontal lines that appear deep to the pleural
link 1: https://static1.squarespace.com/static/549b0 line (i.e., the most anterior and brightest line seen on
Pleural line
Pleural line
A B
Figure 7.5 (a) M-mode applied to normal lung parenchyma creates an artifact known as “seashore sign” in which horizontal lines
corresponding to superficial chest wall tissue appear above the pleural line and pulmonary tissue below the pleura produces a granular
pattern. Make sure the sweep speed is set to slow. (b) M-mode in the absence of lung sliding: stratosphere sign or barcode sign in which
horizontal lines corresponding to superficial chest wall tissue appear above the pleural line and the air or nonmoving pulmonary tissue below
the pleura.
ultrasound) and are the acoustic mirror of the air– The main principle that governs interpretation of
fluid interface between apposed pleurae.12,13 these images is that air rises to occupy the most
The bat sign, also seen in the longitudinal 2D ascendant point of a body cavity while fluid falls to
mode, is visualization of the pleura of normal lung dependent locations. As with the normal pulmonary
in which lung sliding and A-lines are flanked by the artifacts, since fluid has greater acoustic impedance
two ribs’ (a superior rib and an inferior rib) acoustic than air, fluid-filled tissue and compressed tissue will
shadows. An image obtained via an abdominal win- better conduct the ultrasound signal.
dow where there are no ribs will not show a bat sign;
thoracentesis or chest tube insertion should not be Pneumothorax
performed in that location. Ultrasound can only rule out a pneumothorax in the
In M-mode, normal lung parenchyma creates the thoracic interspace where lung sliding is seen. Given
seashore sign in which horizontal lines corresponding that free air within the pleural cavity rises to the most
to superficial chest wall tissue appear superficial to the superior regions if there are no adhesions to trap it, a
pleural line and pulmonary tissue deep to the pleura pneumothorax can only definitively be ruled out
produces a granular pattern (Figure 7.5a).16 The sign when the ultrasound is applied to the most superior
is so named because the horizontal lines of the super- region for the patient’s position. Several adjacent
ficial tissue resemble ocean waves and the granular superior thoracic interspaces should be imaged to
pattern of the pleural tissue resembles a sandy beach. assure a sufficiently thorough exam and maximize
The M-mode seashore sign is used to augment evalua- specificity.
tions where lung sliding is subtle and the shimmer of The main abnormal artifacts produced by pneu-
the opposing pleurae is difficult to perceive in the mothorax are abnormal A-lines and the barcode sign.
standard 2D view. Note that M-mode captures lung In addition, there are several complementary artifacts
sliding in a static image that may be printed, whereas that can be used to augment indeterminate exams;
video is required to demonstrate lung sliding in 2D these are lung point, heart point, lung pulse and B-
mode. lines.
Under 2D ultrasound, absence of lung sliding pro-
Abnormal Artifacts duces A-lines throughout the entire sonographic field
Abnormal artifacts that identify lung pathology are and may reflect a pneumothorax, depending on the
due to deviations from the normal air–pleural inter- clinical scenario.12 (video link 2: https://static1.square
face. These deviations may be generated by the patho- space.com/static/549b0d5fe4b031a76584e558/577dba7
logic interposition of air or fluid in the pleural space, a1b631b1456316e53/577dba7c37c581b33317e8a2/146
or by the accumulation of fluid within the lung tissue. 7857674242/noslide.gif?format=1000w).
70
In M-mode, absence of lung sliding produces a bleb generates a sonographic appearance that is simi-
stereotypical finding known as the barcode sign (also lar to pneumothorax.23
known as the stratosphere sign). Horizontal lines cor- The heart point sign is a transient visualization of
responding to the chest wall tissue appear superficial to the heart in diastole that is seen in cases of incomplete
the pleural line, and air in the lung deep to the pleura pneumothorax in the left hemithorax. Heart point,
produces a pattern of thin straight repeating horizontal imaged in 2D posterior to the left parasternum, is
lines and thicker A-lines extending throughout the caused by the intermittent displacement of air ante-
depth of the sonographic field (Figure 7.5b).17 The rior to the pericardium by the expanding diastolic
absence of lung sliding on 2D is 100% sensitive and heart. As the heart contracts in systole, the interposing
78% specific for pneumothorax; specificity increases to air moves away from the chest wall and the heart point
94% when absence of lung sliding on 2D is accompa- sign disappears (video link 4: https://images.readcube-
nied by the M-mode barcode sign.12 Conditions other cdn.com/publishers/wiley/videos/67a52e479d133bf0
than pneumothorax that may eliminate or reduce lung d91b95983681a10db66ea507425852f67226dd89
sliding include apnea, mainstem intubation, phrenic ba54e53d/1.mp4).24
nerve palsy (in spontaneously breathing patients), Left hemithorax M-mode ultrasound may also
pleural effusion, consolidation, atelectasis, acute reveal a lung pulse sign caused by transmitted
respiratory distress syndrome (ARDS), peripheral bul- impulses of the heart beat through consolidated or
lae, asthma or COPD exacerbation with hyperinflation, atelectatic lung.25 These transmitted impulses appear
pulmonary fibrosis, and pleural adhesions.4,15,17–19 as linear, vertical ripples that travel from the lung to
Most of these conditions can be differentiated on the the pleural line simultaneous with systole. If the rip-
basis of lung ultrasound in combination with clinical ples extend above the pleural line into the chest wall,
exam (Table 7.1). then they are artifacts due to probe movement and do
In 2D mode the lung point sign (also referred to not represent a true lung pulse.
simply as lung point) is highly specific for pneu-
mothorax. It represents the junction between normal Increased Lung Water
lung that slides with respiration and abnormal lung Fluid content of the lungs can be assessed by examin-
that does not slide due to the pneumothorax; the point ing for B-lines (also known as comet tails), vertical
at which lung sliding disappears moves with respira- hyperechoic rays that emanate from the pleura and
tion because an incomplete pneumothorax moves radiate into the lung tissue (Figure 7.7). B-lines are
with lung expansion and collapse. This artifact is indicative of interstitial and/or alveolar edema, and
produced by the edge of an incomplete pneumothorax the sonographic effect is generated by fragmentation
where intrapleural air abuts the normally of the air–fluid interface at the pleura by fluid within
apposed visceral and parietal pleura20 (video link 3: inter-lobular septa.26 The difference in acoustic impe-
https://images.readcube-cdn.com/publishers/wiley/vi dance at the air–fluid interface produces internal
deos/67a52e479d133bf0d91b95983681a10d reverberations that result in this streaking artifact
b66ea507425852f67226dd89ba54e53d/2.mp4). Lung that crosses through A-lines and can obliterate
point may also be imaged in M-mode; the seashore them.26 This finding correlates with the chest x-ray
sign will appear where the visceral and parietal finding of Kerley B-lines, after which it is named.27 B-
pleurae are opposed, and the barcode sign will appear lines occur not only in cases of cardiogenic (e.g., left
at the point where they are separated (Figure 7.6). In heart dysfunction) and non-cardiogenic pulmonary
M-mode, as in 2D mode, the lung point moves with edema (e.g., ARDS), but also in cases of interstitial
respiration. When lung point is visualized, it can be lung diseases (e.g., idiopathic pulmonary fibrosis) that
used to define the approximate size of a pneu- cause similar changes in acoustic impedance.
mothorax and serial sonographic exams may be per- B-lines detect alveolar-interstitial syndrome due
formed to monitor its progress.21,22 Lung point, to congestive heart failure (CHF), noncardiogenic
however, will not be seen in cases of a very large edema, or interstitial thickening, (e.g., interstitial
pneumothorax because the air completely surrounds lung diseases, pulmonary fibrosis, inflammatory
the lung and there is no point where the visceral and disease), with a sensitivity of 92.5% and specificity
parietal pleura meet.20 Conversely, lung point can be 86.7%.26 The number of B-lines and the width
mimicked by pulmonary blebs since air within the between each B-line correlate with the degree of
Table 7.1 Sonographic findings that may mimic a pneumothorax.
Condition Etiology of sonographic findings Diagnosis and differentiation from

pneumothorax
Apnea Lack of diaphragmatic movement. Bilateral absence of lung sliding, absence of
diaphragmatic contraction in the lower axillary
views, clinical absence of chest wall movement,
loss of respiratory waveform on end-tidal
capnography.
Esophageal Displacement or misplacement of the Loss of expired tidal volume, bilateral absence of
Intubation/Absence endotracheal tube, disconnection with the lung sliding, presence of two sources of A-line
of Ventilation respiratory circuit. artifact in a trans-tracheal view, loss of respiratory
waveform on end-tidal capnography.
Mainstem Intubation Isolation of the unintubated bronchus from Noted distal displacement of endotracheal tube
ventilation with loss of ventilation of the non- from initial depth, increased peak and plateau
intubated lung. pressures, return of lung sliding with withdrawal of
endotracheal tube.
COPD/Asthma Limited lung excursion due to profound Clinical history, return of lung sliding with “sigh”
exacerbation hyperinflation. induced by allowing prolonged expiration or
temporary removal from vent.
Consolidation Alveolar filling results in hepatization of tissue with Focal areas of involvement that correlate with
spared bronchi producing internal areas of air- lobar distribution. May or may not be stratified by
fluid interface resulting in sonographic air gravity. Presence of sonographic air-
bronchograms. Peripheral consolidations may bronchograms within hepatized pulmonary
focally reduce or eliminate pleural sliding. tissue. Subpleural consolidations seen as
shredded pleural border (shred sign) and focally
absent pleural sliding.
Atelectasis Collapsed pulmonary tissue results in increased Focal areas of involvement correlate with
local concentration of fluid and decreased local pulmonary anatomy. Do not always respect
air. Hepatization without sonographic air gravity, although tend to occur in dependent lung
bronchograms. regions. Hepatization with lack of air
bronchograms. Reduced or absent pleural sliding.
May improve with recruitment maneuvers.
ARDS Patchy and diffuse loss of respirophasic B-lines persist in areas even if lung sliding has
movement with scattered areas of increased been lost. There are no B-lines with a
extravascular lung water. pneumothorax.
Pleural Adhesions Focal restriction of respirophasic movement due History of previous thoracic surgery or
to local “tacking” of the lung to the pleura. inflammatory disease, absence of lung sliding
noted on pre-procedural sonographic exam prior
to interventions.
Pleural Effusion Interposition of fluid between the parietal and Loss of lung sliding is gravity dependent,
visceral pleura prevents lung sliding by separating echolucent fluid can be seen in the pleural space
the pleurae. and may permit the visualization of thoracic
structures, the lung tip may be seen moving
within the pleural fluid.
Pulmonary Bleb Focal restriction of respirophasic movement at the History of COPD, noted on pre-procedural scan,
site of the bleb. remains stable when observed over time.
Phrenic Nerve Palsy Loss of respirophasic movement due to Clinical history, loss of diaphragmatic contraction
diaphragmatic paralysis. in lower axillary view, return of lung sliding with
positive pressure ventilation.
26
extravascular pulmonary fluid. The presence of mul- inferiorly located dependent portions of the lung
tiple confluent B-lines is 97% sensitive for pulmonary due to the effect of gravity, B-lines are first seen in
edema.5,28–30 Conversely, in cases where there are inferior lung fields. With greater degrees of edema, B-
fewer than three B-lines per field, pulmonary edema lines become apparent in superior lung fields.26
is unlikely.26,30–33 In volume overload and cardio- Interstitial lung diseases usually can be differentiated
genic pulmonary edema, as fluid accumulates in from pulmonary edema by clinical history and
72
Lung begins to
Lung falls away B lines
touch the chest
from chest wall
wall at this point
Figure 7.7 B-lines are vertical hyperechoic rays based from the
Figure 7.6 A lung point imaged in M-mode. The seashore sign is pleural line radiating down into the pulmonary tissue to the end of
visualized over a point on the chest where the apposed visceral and the screen (regardless of the depth setting, which in this case is
parietal pleura touch and the stratosphere (or barcode) sign appears 19 cm) that correlate with interstitial thickening (e.g., edema,
over the point where the lung contracts and air interposes between inflammation, fibrosis) and alveolar edema.
the visceral and parietal pleura.
Figure 7.8 Pleural effusions appear

Pleural effusion anechoic and are visualized as an
echolucent space between the diaphragm
and pulmonary tissue, or chest wall and
pulmonary tissue. The compressed lung
edge may be seen swimming in fluid (called
Diaphragm jellyfish sign, elephant trunk sign, or simply
lung flapping), a dynamic finding that
indicates fluid versus an anechoic mimic
such as peripheral consolidation, lipoma, or
mass that is static without lung flapping.
Liver Lung
imaging. In any case, B-lines always rule out greater than 100 ml with 100% sensitivity.37,38
pneumothorax.4,34 However, pleural thickening may mimic small pleural
effusions and subcutaneous emphysema may obscure
Pleural Effusions small pleural effusions.39
Pulmonary sonography can also be used to diagnose In indeterminate cases, two other findings should
pleural effusions and add to the determination of be considered. Normally, vertebral bodies are
volume status. The sonographic evaluation of pleural obscured by the lung in all thoracic views. In an
effusions was first described in 1964 and confirmed by effusion, vertebral bodies appear as saw-toothed
thoracentesis in 1967.2,35 Pleural effusion and hyperechoic lines known as V-lines through the
hemothorax appear as an echolucent (hypoechoic or acoustic window created.40 Pleural effusions can also
anechoic) space between the diaphragm and lung.36 be verified using M-mode that detects the cyclic
With large effusions, the compressed lung edge may movement of collapsed lung parenchyma within the
be seen floating in fluid (Figure 7.8). Modern ultra- effusion during respiration and reveals a sinusoidal
sound machines can reliably visualize effusions pattern.13
(a) (b) (c)
Figure 7.9 (a) Consolidation appears as a granular, tissue-like transformation of the lung termed hepatization for its resemblance to the
sonographic appearance of liver parenchyma. The hyperechoic white dots and linear structures represent air bronchograms cut in transverse
and longitudinal planes respectively. There is also a pleural effusion indicated by the surrounding anechoic area. (b) Subpleural (peripheral)
consolidation is identified by an irregular pleural border (arrow) on ultrasound, termed “shred sign”. (c) One interspace away from this focal
peripheral consolidation the pleural line can be seen and is sharp and smooth (arrow).
consolidation without sonographic air broncho-

grams. Depending on the size of the atelectatic area,
lung sliding may be present, reduced, or absent.
Role of Lung Ultrasound

in Perioperative Care
The Preoperative Exam
Preoperative pulmonary sonography should be con-
sidered if placement of a central venous catheter is
planned, there is suspected cardiopulmonary disease,
volume status is unclear, or there is underlying thor-
acic pathology. ASA IV and V status usually confers
Figure 7.10 Atelectasis. Lung hepatization with rare or absent air some or all of these indications.
bronchograms (indicated by arrow). If there were more hyperechoic Preoperative sonographic evaluation can assess
white dots and linear structures representing air bronchograms volume status and establish a baseline that can be
(shown in Figure 7.9), this would be termed “consolidation.”
used for comparison to later intraoperative and post-
operative studies. When performing a preoperative
Consolidation and Atelectasis exam, the bilateral upper and lower thorax should be
The alveolar filling that occurs with pulmonary con- evaluated in the midclavicular and midaxillary lines at
solidation alters the acoustic impedance of lung tis- the same points the anesthesiologist might perform
sue. These areas with adjacent patent hyperechoic air intraoperative sonography. Marking these points can
bronchi41,42 appear granular and are similar to ultra- ensure fidelity with subsequent studies. The lung apex
sound images of liver parenchyma, an effect known as in the midclavicular line and the diaphragm in the
hepatization (Figure 7.9a).43 Subpleural consolidation midaxillary line should also be visualized bilaterally.
may be identified by an irregular border, termed shred In patients with known lung pathology or prior thor-
sign, with 90% sensitivity and 98% specificity (Figures acic surgery, a more thorough exam visualizing all
7.9b, 7.9c).43 Lung ultrasound also can be used to interspaces may identify pleural adhesions and avoid
detect atelectasis with 88% accuracy.41 Atelectasis later confusion when considering potential
(Figure 7.10) typically appears as juxtapleural pneumothorax.
74
Pleural adhesions will appear as areas of absent confirms that both lungs are effectively ventilated
lung sliding with pleural thickening and can occur in and localizes the tracheal tube tip to between the
otherwise healthy patients.44,45 Locations with absent carina and the vocal cords. Esophageal intubation
lung sliding and without B-lines may indicate pleural will result in absence of lung sliding that is a highly
adhesions or bullae, and should be noted or marked sensitive and specific finding.48,49 Unilateral lung slid-
when detected to avoid confounding later intra- ing may indicate a mainstem intubation, and correc-
operative exams. Other imaging (e.g. chest x-ray, CT tion can be visualized if withdrawing the tube
scan) may be sought in select cases. demonstrates sliding or a comet tail (B-line) arti-
A preoperative sonographic study can also be used fact.49–51
to assess volume status and predict how fluid admin- Esophageal intubation can be detected by trans-
istration might be tolerated by establishing the pre- tracheal ultrasound that is quite sensitive and speci-
sence or absence of B-lines and pleural effusions.28 fic.52–54 In this technique the trachea is examined by
Patients with confluent B-lines may require further placing the ultrasound probe in transverse orientation
workup by electrocardiogram and or echocardio- at the level of the cricothyroid membrane, where the
gram, further preoperative optimization (e.g., diur- trachea appears as a single well-defined column of A-
esis, dialysis), or, if the case is emergent, limited lines produced by the tracheal mucosa-air interface. A
fluid administration and invasive monitoring. tracheal tube will not alter this artifact.55 While the
Atelectasis detected by preoperative sonography esophagus also has the potential for a mucosa-to-air
should prompt incentive spirometry, administration interface, it remains masked behind the trachea and
of bronchodilators, careful preoxygenation, and an esophageal tube will displace the esophagus, result-
aggressive pulmonary toilet before and after intuba- ing in a second, adjacent A-line. Thus, the presence of
tion. While preoperative pneumonia is unlikely in a single anatomically normal A-line confirms tracheal
elective cases, patients requiring emergency opera- tube placement in the trachea, while the presence of a
tions, especially inpatients, often have comorbidities second non-anatomical A-line is evidence that the
such as undiagnosed pneumonia. Establishing this tube is in the esophagus.55 The esophagus is displaced
diagnosis preoperatively will help guide antibiotic leftward in most esophageal intubations, but may
administration, sepsis workup, and postoperative occasionally be displaced to the right or posterior to
management. the trachea.55 When the esophagus is posterior to the
trachea, esophageal intubation may be identified by
Sonography for Intubation tilting the probe at an oblique angle to the left or right
to reveal the presence of two adjacent A-line
Sonography can confirm tracheal tube placement. The
artifacts.53,55
current standard of care is auscultation and end-tidal
capnography.46,47 Auscultation in morbidly obese
patients and those with COPD or acute bronchos- Intraoperative Exam
pasm may result in a false negative exam (suspicion Unlike routine preoperative lung ultrasound, the
that the tracheal tube is NOT in the trachea when it intraoperative sonographic exam is performed in
is), and transmission of sounds from an esophageal response to changes in the patient’s clinical condition
intubation may result in a false positive exam (con- and/or the need to make a management decision. The
firmation that the tracheal tube is in the trachea when issues usually prompting an intraoperative lung ultra-
it is not). Erroneous capnographic findings can occur sound exam are hypoxemia, hypotension, question-
in low cardiac output states, especially cardiac arrest, able tracheal tube position, and volume status. The
or by occlusion of the end-tidal tubing and monitor by study should first rule out tension pneumothorax.
debris. The most superior portion of thorax for the patient’s
Transthoracic and transtracheal views can be used current position should be imaged immediately. This
to confirm tracheal tube placement. Transthoracic often entails imaging the patient’s anterior chest wall,
imaging may be performed with either a linear but in certain seated positions may require imaging
probe or the phased array probe, but a linear probe the lung apex. The presence of bilateral anterior lung
is required to obtain the transtracheal view. After sliding effectively rules out tension pneumothorax.
tracheal intubation, thoracic ultrasound that demon- Smaller or loculated pneumothoraces might be pre-
strates bilateral lung sliding or a seashore sign sent, but they should not cause severe hypoxemia or
hypotension. After tension pneumothorax has been any case, B-lines and pleural effusions indicate that
ruled out, the exam should continue with views fluid administration may be contraindicated and that
obtained in the midaxillary and midclavicular lines in vasopressors and inotropes may optimize hemody-
the superior and inferior thorax and visualize the namic status. The absence of B-lines supports a diag-
diaphragm to the extent that the operative field allows. nosis of hypovolemic shock, distributive shock, or
If a pneumothorax is found and the patient is right heart failure.58,59 As with hypoxemia, in patients
hemodynamically stable, the size of the pneu- with hypotension and an unremarkable lung ultra-
mothorax can be approximated utilizing the lung sound exam, pulmonary embolus should be
point.22 Conservative management of a stable pneu- considered.
mothorax with serial sonography may permit avoid- The presence or absence of B lines helps determine
ance of chest tube placement.21 Ultrasound the safety of a fluid bolus58 when combined with
significantly outperforms supine portable chest x-ray cardiac and IVC examinations. A unique benefit of
for the detection of pneumothorax.11,56 The presence pulmonary ultrasound is that the midclavicular and
of B-lines also rules out a pneumothorax, suggests the midaxillary thorax can be reached during most opera-
alternative diagnosis of pulmonary edema, and tive procedures without disturbing the sterile field,
expands the differential to include volume overload and images can be obtained even if cardiac and IVC
and new left ventricular failure. Pulmonary edema views are not available.
can be further confirmed by the presence of a pleural
effusion. Postoperative Exam
Atelectasis is a significant risk factor for periopera-
Postoperative pulmonary ultrasound should be con-
tive complications and mortality. Ultrasound can
sidered in patients who develop hypoxemia, become
rapidly detect atelectasis as a cause of hypoxemia
hypotensive, or are having difficulty weaning from
and other etiologies of increased peak airway pres-
mechanical ventilation. When a patient with difficulty
sure.9 The presence of juxtapleural infiltrates indi-
weaning from mechanical ventilation has B-lines,
cates atelectasis as the etiology of hypoxemia, and
diuresis may be indicated. Sonographic evidence of
the patient may be responsive to recruitment maneu-
atelectasis suggests that bronchodilators and recruit-
vers and bronchodilators. There are case reports of
ment maneuvers might be effective; incentive spiro-
direct visualization of the resolution of atelectasis by
metry and optimization of pain relief may also be
ultrasound while performing recruitment maneuvers,
helpful. These interventions, provided in a timely
and real-time imaging can guide the recruitment
fashion, should result in more rapid extubation and
maneuver.57 In hypoxemic patients the presence of a
avoidance of prolonged mechanical ventilation – a
completely unremarkable pulmonary sonographic
source of significantly increased morbidity and mor-
exam, including bilateral lung sliding, A-lines, and
tality in postoperative patients.60 If sonography
no B–lines, raises the possibility of pulmonary
demonstrates a shred sign, hepatization, air broncho-
embolus.
grams, reduced pleural sliding, or diffuse pathology,
If there is concern that the endotracheal tube has
then simple atelectasis or overload is unlikely and
been displaced during a procedure, this can be eval-
aspiration, developing pneumonia, or ARDS should
uated by repeating the intubation exam described
be considered.43,61
above and may be accompanied by a complementary
Pulmonary ultrasound can differentiate ARDS
chest x-ray.
from volume overload, which is usually associated
As with hypoxemia, the sonographic evaluation
with B-lines only. The ultrasound hallmarks of
for hypotension begins with ruling out tension pneu-
ARDS include dyshomogenous distribution of the
mothorax. Pulmonary ultrasound can then be paired
increased extravascular water, scattered pleural thick-
with cardiac ultrasound, clinical assessment, and IVC
ening, reduced pleural sliding, and the presence of a
examination to distinguish between hypovolemic and
lung pulse.62 These findings also suggest that in
cardiogenic shock due to left ventricular failure.58 The
mechanically ventilated patients weaning will be dif-
presence of B-lines and pleural effusion should
ficult, and in spontaneously breathing patients close
prompt a search for cardiac failure, reduces the like-
observation and preparation for tracheal intubation is
lihood that hypovolemia is the source of hypotension,
warranted. On mechanical ventilation, any new
and increases suspicion for cardiogenic shock.58,59 In
hypoxemia or hypotension can be evaluated by
76
pulmonary ultrasound to rule out pneumothorax due 9. Restrepo RD, Braverman J. Current challenges in the
to barotrauma. In conjunction with cardiac and infer- recognition, prevention and treatment of perioperative
ior vena cava studies, pulmonary ultrasound can be pulmonary atelectasis. Expert Rev Respir Med 2015;9
(1):97–107.
used to assess volume status and fluid responsiveness.58
10. Zieleskiewicz L, Fresco R, Duclos G, et al. Integrating
extended focused assessment with sonography for
Conclusion trauma (eFAST) in the initial assessment of severe
Lung ultrasound is a point-of-care test particularly trauma: Impact on the management of 756 patients.
well suited to diagnose pneumothorax, pleural effu- Injury 2018;49(10):1774–80.
sion, pulmonary edema, and consolidation. In the 11. Blaivas M, Lyon M, Duggal S. A prospective
operating room setting, imaging can establish a base- comparison of supine chest radiography and bedside
line for comparison to maximize confidence in later ultrasound for the diagnosis of traumatic
examinations displaying new pathology. In conjunc- pneumothorax. Acad Emerg Med 2005;12(9):844–9.
tion with cardiac and IVC ultrasound the etiology of 12. Lichtenstein DA, Meziere G, Lascols N, et al.
hypotension can be ascertained as hypovolemic, dis- Ultrasound diagnosis of occult pneumothorax. Crit
tributive, or cardiac. In addition, even more basic Care Med 2005;33(6):1231–8.
assessment regarding volume status and fluid respon- 13. Lichtenstein D, van Hooland S, Elbers P, Malbrain ML.
siveness is improved. Ultrasound can be used to Ten good reasons to practice ultrasound in critical
care. Anaesthesiol Intensive Ther 2014;46(5):323–5.
guide, confirm, and localize tracheal tube placement.
In the postoperative setting pulmonary ultrasound 14. Koh DM, Burke S, Davies N, Padley SP. Transthoracic
can facilitate clinical decisions regarding weaning US of the chest: clinical uses and applications.
Radiographics 2002;22(1):e1.
patients from mechanical ventilation and the need
for ICU admission. 15. Lichtenstein DA, Menu Y. A bedside ultrasound sign
ruling out pneumothorax in the critically ill. Lung
sliding. Chest 1995;108(5):1345–8.
References 16. Husain LF, Hagopian L, Wayman D, Baker WE,
1. Joyner CR, Jr., Herman RJ, Reid JM. Reflected Carmody KA. Sonographic diagnosis of
ultrasound in the detection and localization of pleural pneumothorax. J Emerg Trauma Shock 2012;5(1):76–
effusion. JAMA 1967;200(5):399–402. 81.
2. Pell R. Ultrasound for routine clinical investigations. 17. Targhetta R, Bourgeois JM, Chavagneux R, et al.
Ultrasonics 1964;2(2):87–9. Ultrasonic signs of pneumothorax: preliminary work. J
3. Rantanen NW. Diseases of the thorax. Vet Clin North Clin Ultrasound 1993;21(4):245–50.
Am Equine Pract 1986;2(1):49–66. 18. Slater A, Goodwin M, Anderson KE, Gleeson FV.
4. Wernecke K, Galanski M, Peters PE, Hansen J. COPD can mimic the appearance of pneumothorax on
Pneumothorax: evaluation by ultrasound–preliminary thoracic ultrasound. Chest 2006;129(3):545–50.
results. J Thorac Imaging 1987;2(2):76–8. 19. De Luca C, Valentino M, Rimondi MR, et al. Use of
5. Lichtenstein DA, Meziere GA. Relevance of lung chest sonography in acute-care radiology. J
ultrasound in the diagnosis of acute respiratory Ultrasound 2008;11(4):125–34.
failure: the BLUE protocol. Chest 2008;134(1): 20. Lichtenstein D, Meziere G, Biderman P, Gepner A. The
117–25. “lung point”: an ultrasound sign specific to
6. Mittal AK, Gupta N. Intraoperative lung ultrasound: a pneumothorax. Intensive Care Med 2000;26(10):1434–40.
clinicodynamic perspective. J Anaesthesiol Clin 21. Sato I, Kanda H, Kanao-Kanda M, Kurosawa A,
Pharmacol 2016;32(3):288–97. Kunisawa T. A case of iatrogenic pneumothorax in
7. Shander A, Fleisher LA, Barie PS, et al. Clinical and which chest tube placement could be avoided by
economic burden of postoperative pulmonary intraoperative evaluation with transthoracic
complications: patient safety summit on definition, risk- ultrasonography. Ther Clin Risk Manag 2017;13:843–5.
reducing interventions, and preventive strategies. Crit 22. Aspler A, Pivetta E, Stone MB. Double-lung point sign
Care Med 2011;39(9):2163–72. in traumatic pneumothorax. Am J Emerg Med 2014;32
8. Abstracts of the 30th International Symposium on (7):819e81–2.
Intensive Care and Emergency Medicine. Brussels, 23. Gelabert C, Nelson M. Bleb point: mimicker of
Belgium. March 9–12, 2010. Crit Care 2010;14 Suppl 1: pneumothorax in bullous lung disease. West J Emerg
P1–602. Med 2015;16(3):447–9.
24. Stone MB, Chilstrom M, Chase K, Lichtenstein D. The 39. Hasan AA, Makhlouf HA, Mohamed ARM.
heart point sign: description of a new ultrasound Discrimination between pleural thickening and
finding suggesting pneumothorax. Acad Emerg Med minimal pleural effusion using color Doppler chest
2010;17(11):e149–150. ultrasonography. Egyptian Chest Diseases Tuberculosis
25. Lichtenstein DA, Lascols N, Prin S, Meziere G. The 2013;62:429–33.
“lung pulse”: an early ultrasound sign of complete 40. Atkinson P, Milne J, Loubani O, Verheul G. The V-
atelectasis. Intensive Care Med 2003;29(12):2187–92. line: a sonographic aid for the confirmation of pleural
26. Lichtenstein D, Meziere G, Biderman P, Gepner A, fluid. Crit Ultrasound J 2012;4(1):19.
Barre O. The comet-tail artifact. An ultrasound sign of 41. Acosta CM, Maidana GA, Jacovitti D, et al. Accuracy of
alveolar-interstitial syndrome. Am J Respir Crit Care transthoracic lung ultrasound for diagnosing
Med 1997;156(5):1640–6. anesthesia-induced atelectasis in children.
27. Kerley P. Radiology in Heart Disease. Br Med J 1933;2 Anesthesiology 2014;120(6):1370–9.
(3795):594–612.3. 42. Weinberg B, Diakoumakis EE, Kass EG, Seife B, Zvi
28. Jambrik Z, Monti S, Coppola V, et al. Usefulness of ZB. The air bronchogram: sonographic demonstration.
ultrasound lung comets as a nonradiologic sign of AJR Am J Roentgenol 1986;147(3):593–5.
extravascular lung water. Am J Cardiol 2004;93 43. Lichtenstein DA, Lascols N, Mezière G, Gepner A.
(10):1265–70. Ultrasound diagnosis of alveolar consolidation in the
29. Baldi G, Gargani L, Abramo A, et al. Lung water critically ill. Intensive Care Med 2004;30(2):276–81.
assessment by lung ultrasonography in intensive 44. Cassanelli N, Caroli G, Dolci G, et al. Accuracy of
care: a pilot study. Intensive Care Med 2013;39 transthoracic ultrasound for the detection of pleural
(1):74–84. adhesions. Eur J Cardiothorac Surg 2012;42(5):813–18;
30. Platz E, Lewis EF, Uno H, et al. Detection and discussion 818.
prognostic value of pulmonary congestion by lung 45. Wei B, Wang T, Jiang F, Wang H. Use of transthoracic
ultrasound in ambulatory heart failure patients. Eur ultrasound to predict pleural adhesions: a prospective
Heart J 2016;37(15):1244–51. blinded study. Thorac Cardiovasc Surg 2012;60(2):101–4.
31. Theerawit P, Touman N, Sutherasan Y, Kiatboonsri S. 46. Grmec S. Comparison of three different methods to
Transthoracic ultrasound assessment of B-lines for confirm tracheal tube placement in emergency
identifying the increment of extravascular lung water intubation. Intensive Care Med 2002;28(6):701–4.
in shock patients requiring fluid resuscitation. Indian J 47. Neumar RW, Otto CW, Link MS, et al. Part 8: adult
Crit Care Med 2014;18(4):195–99. advanced cardiovascular life support: 2010 American
32. Aras MA, Teerlink JR. Lung ultrasound: a “B-line” to Heart Association guidelines for cardiopulmonary
the prediction of decompensated heart failure. Eur resuscitation and emergency cardiovascular care.
Heart J 2016;37(15):1252–4. Circulation 2010;122(18 Suppl 3):S729–767.
33. Dietrich CF, Mathis G, Blaivas M, et al. Lung B-line 48. Weaver B, Lyon M, Blaivas M. Confirmation of
artefacts and their use. J Thorac Dis 2016;8(6):1356–65. endotracheal tube placement after intubation using the
34. Lichtenstein D, Meziere G, Biderman P, Gepner A. The ultrasound sliding lung sign. Acad Emerg Med 2006;13
comet-tail artifact: an ultrasound sign ruling out (3):239–44.
pneumothorax. Intensive Care Med 1999;25(4):383–8. 49. Rajan S, Surendran J, Paul J, Kumar L. Rapidity and
35. Joyner CR Jr, Miller LD, Dudrick SJ, Eskin DJ, Bloom efficacy of ultrasonographic sliding lung sign and
P. Reflected ultrasound in the study of diseases of the auscultation in confirming endotracheal intubation in
chest. Trans Am Clin Climatol Assoc 1967;78:28–37. overweight and obese patients. Indian J Anaesth
2017;61(3):230–4.
36. Kataoka H, Takada S. The role of thoracic
ultrasonography for evaluation of patients with 50. Sim SS, Lien WC, Chou HC, et al. Ultrasonographic
decompensated chronic heart failure. J Am Coll Cardiol lung sliding sign in confirming proper endotracheal
2000;35(6):1638–46. intubation during emergency intubation. Resuscitation
2012;83(3):307–12.
37. Kalokairinou-Motogna M, Maratou K, Paianid I, et al.
Application of color Doppler ultrasound in the study 51. Sustic A, Protic A, Cicvaric T, Zupan Z. The addition
of small pleural effusion. Med Ultrason 2010;12(1):12– of a brief ultrasound examination to clinical
16. assessment increases the ability to confirm placement
of double-lumen endotracheal tubes. J Clin Anesth
38. Rothlin MA, Naf R, Amgwerd M, Candinas D, Frick T, 2010;22(4):246–9.
Trentz O. Ultrasound in blunt abdominal and thoracic
trauma. J Trauma 1993;34(4):488–95. 52. Ma G, Davis DP, Schmitt J, Vilke GM, Chan TC,
Hayden SR. The sensitivity and specificity of
78
transcricothyroid ultrasonography to confirm 58. Lichtenstein D. FALLS-protocol: lung ultrasound in

endotracheal tube placement in a cadaver model. J hemodynamic assessment of shock. Heart Lung Vessel
Emerg Med 2007;32(4):405–7. 2013;5(3):142–7.
53. Chou HC, Chong KM, Sim SS, et al. Real-time tracheal 59. Kajimoto K, Madeen K, Nakayama T, et al. Rapid
ultrasonography for confirmation of endotracheal tube evaluation by lung-cardiac-inferior vena cava (LCI)
placement during cardiopulmonary resuscitation. integrated ultrasound for differentiating heart failure
Resuscitation 2013;84(12):1708–12. from pulmonary disease as the cause of acute dyspnea
54. Ma Gene, Hayden SR, Chan Theodore C, et al., Using in the emergency setting. Cardiovasc Ultrasound
ultrasound to visualize and confirm endotracheal 2012;10(1):49.
intubation. Academic Emergency Medicine 1999;6 60. Smetana GW. Postoperative pulmonary
(5):515. complications: an update on risk assessment and
55. Werner SL, Smith CE, Goldstein JR, Jones RA, Cydulka reduction. Cleve Clin J Med 2009;76 Suppl 4:S60–65.
RK. Pilot study to evaluate the accuracy of 61. Lichtenstein D, Goldstein I, Mourgeon E, et al.
ultrasonography in confirming endotracheal tube Comparative diagnostic performances of auscultation,
placement. Ann Emerg Med 2007;49(1):75–80. chest radiography, and lung ultrasonography in acute
56. Ball CG, Ranson K, Dente CJ, et al. Clinical predictors respiratory distress syndrome. Anesthesiology 2004;100
of occult pneumothoraces in severely injured blunt (1):9–15.
polytrauma patients: a prospective observational 62. Copetti R, Soldati G, Copetti P. Chest sonography:
study. Injury 2009;40(1):44–7. a useful tool to differentiate acute cardiogenic
57. Du J, Tan J, Yu K, Wang R. Lung recruitment pulmonary edema from acute respiratory
maneuvers using direct ultrasound guidance: a case distress syndrome. Cardiovasc Ultrasound
study. Respir Care 2015;60(5):e93–96. 2008;6:16.
Chapter
Point-of-Care Ultrasound: Determination
8 of Fluid Responsiveness
Subhash Krishnamoorthy and Oliver Panzer
Introduction status, baseline laboratories, or imaging, in which

case rapid, efficient diagnosis and initiation of treat-
Hypotension and shock in the perioperative setting
ment can be difficult, if not impossible. The most
may arise from a variety of etiologies that require
common initial intervention is fluid administration
different therapies for successful treatment.
with blood pressure monitoring. Different types of
A common initial intervention is administration of
shock, however, require different therapies, and it is
a fluid bolus to increase the cardiac stroke volume and
therefore helpful to classify shock on the basis of
mean arterial blood pressure; however, fluid respon-
etiology as hypovolemic, distributive, obstructive, or
siveness is absent in up to 50% of all hypotensive
cardiogenic. The underlying etiology may also be
patients, and if they receive excessive fluid without
mixed, increasing the difficulty in deciding on treat-
the desired hemodynamic response, morbidity and
ment that may include a combination or permutation
mortality may result. Therefore, there is a clinical
of administration of fluids, vasopressors, inotropes,
need to determine the likelihood that an individual
and interventions, including thrombolysis and
patient will respond to fluid administration. Advances
pericardiocentesis.
in understanding the physiologic response to fluid
Fluid administration is the most common inter-
administration have led to the development of
vention for shock, gaining favor after the 2001 pub-
dynamic predictors of the hemodynamic response.
lication of the Early Goal Directed Therapy (EGDT)
The respiratory variation of stroke volume associated
trial by Rivers et al., one of the most cited references in
with positive pressure ventilation (heart–lung interac-
critical care medicine.1 The study protocol empha-
tion model) and the passive leg raise (PLR) (endogen-
sized fluid administration to target central venous
ous fluid challenge model) are the two best-validated
pressure (CVP) ≥ 8–12 mmHg, mean arterial pressure
means to quickly augment stroke volume. The
(MAP) ≥ 65 mmHg, urine output ≥ 0.5 ml/kg/hr, and
respiratory variation model is most applicable during
central venous oxygen saturation (ScvO2) ≥ 70%,
mechanical ventilation and is useful in the operating
resulting in patients receiving upwards of 10 liters of
room, postoperative arenas, and intensive care units
fluid over 72 h. More recent trials such as PROCESS2
(ICUs). Passive leg raise is only possible in nonoper-
and ARISE,3 however, did not replicate the EGDT
ating room settings when there is no contraindication
trial findings. In addition, there is strong evidence
to moving the lower extremities or pelvis. This chap-
that CVP is not a good marker of volume status and
ter will review ultrasound-based parameters such as
is a poor predictor of fluid responsiveness.4,5 Central
inferior vena cava diameter change and velocity time
venous pressure may actually decrease with volume
integral variation utilizing these two methods of
resuscitation secondary to the release of vasodilatory
stroke volume augmentation to predict fluid
substances like nitrous oxide (NO) and atrial
responsiveness.
natriuretic peptide (ANP).6 Furthermore, excess
fluid administration may harm patients. Monnet
Shock et al. suggest that excess fluid may decrease oxygen
It is a challenge to manage hypotension in its most delivery to the patient’s tissues.7 In fact, only 50% of
extreme form when it is accompanied by symptoms ICU patients in septic shock will respond to fluid
and signs of hypoperfusion (i.e., shock). Oftentimes administration with an increase in their MAP.8
management must proceed without knowledge of the Excessive fluid has been associated with increased
patient’s pertinent medical history, prior physical mortality in sepsis patients,9 and pulmonary edema
80
Point-of-Care Ultrasound: Determination of Fluid Responsiveness
and the need for mechanical ventilation support in

Normal heart
acute respiratory distress syndrome (ARDS) SV
patients.10 The appropriate amount and timing of Non-responsive
fluid administration, therefore, has become a key
goal of hemodynamic resuscitation and management
of hypotension and shock, and requires the accurate
Responsive
prediction of fluid responsiveness prior to the admin-
Failing heart
istration of large quantities of fluid in order to prevent
iatrogenic harm.
Non-responsive
Fluid Responsiveness Preload

A B C D
Fluid responsiveness is defined as an increase in
stroke volume (SV) and cardiac output (CO) that Figure 8.1 FSC of two different hearts, a normal heart versus a
failing heart with decreased contractility. The diagram illustrates
results from a fluid bolus, leading to an increase in how both hearts start with the same preload (point A). If the preload
the MAP and improved perfusion and oxygen deliv- is increased to point B, only the normal heart has a significant
ery. Most of the literature defines fluid responsiveness increase in SV (preload-dependent state), but the failing heart only
has a small increase in SV and therefore is not volume-responsive
as an increase in SV by 10–15% from baseline.11 As (pre-load independent state). However, if the normal heart is
CO = SV × Heart Rate (HR), for the sake of simplicity challenged at point C, it also only shows a minimal increase in SV (i.e.,
SV will be the variable discussed. Stroke volume is is not volume-responsive as it is working on the flat part of the FSC).
A hypotensive patient in that state would not benefit from fluid
dependent on preload, contractility, and afterload. boluses; instead, vasopressors or other interventions should be
Common surrogates for preload are central venous considered.
pressure (CVP), pulmonary artery occlusion pressure
(PAOP), left or right ventricular end-diastolic pres-
sure (LVEDP, RVEDP), and left or right ventricular of fluid responsiveness (i.e., is the heart in a preload-
end-diastolic volume (LVEDV, RVEDV). The effect dependent or preload-independent state), preferably
of a preload increase on SV depends on the myocar- before fluid is administered so that excessive fluid
dium’s position on the Frank Starling Curve (FSC). administration can be avoided in patients who are
Generally, an increase in preload will have predictably nonresponsive. This chapter focuses on
a significant effect on SV if the myocardium is operat- ultrasound techniques for determining fluid respon-
ing on the steep (ascending) portion of the curve and, siveness, including inferior vena cava diameter
therefore, will be preload-dependent and fluid- change in response to ventilation and the velocity
responsive. The patient is preload-independent and time integral of the left ventricular outflow tract
fluid-nonresponsive when the attempt to increase (VTILVOT); arterial waveform interrogation and
preload has only a minimal effect. This will occur pulse pressure variation are reviewed in Chapter 11.
when the myocardium’s position is on the flat portion Most dynamic measurements are performed using
of the FSC and an increase in SV is not possible, or one of two common models to augment SV: (1) the
with impaired contractility when the curve is flattened heart–lung interaction during positive pressure ven-
(Figure 8.1). Most hemodynamic parameters, how- tilation (PPV) model, and (2) the PLR endogenous
ever, cannot assess where a patient’s cardiac function fluid challenge model. The concepts behind these
lies on the FSC. Furthermore, these static parameters models are key to understanding all fluid responsive-
are measured at a single point in time and cannot ness tests and avoiding errors in interpretation.
predict a significant SV increase in response to fluid
administration, and therefore cannot discern if Heart–Lung Interaction Model
a patient may benefit or be harmed by the adminis- The heart–lung interaction model (HLIM) rests on the
tration of fluid. Surprisingly, even in normal volun- fact that SV is dependent on the interactions between
teers, cardiac filling pressures are not reflective of the heart and thoracic cavity during PPV. In
volume status and cannot determine the likelihood a hypovolemic patient, changes in intrathoracic
of fluid responsiveness.6 pressure cause very noticeable variations in SV. In an
A dynamic assessment of the likely SV response to arterial pressure tracing, the area under the curve is
a preload challenge allows for a precise determination directly proportional to SV, and variations in this area
(a) (b)
Thorax Thorax
Palv Palv
Ppl 3 Ppl
Zone I-II
Zone III 4
2
RA RV LA LV RA RV LA LV
Palv Palv
Ppl Zone III Ppl Zone II
1 Ppl 5 Ppl 5
4 4
Figure 8.2 Illustration of the heart–lung interaction with positive pressure ventilation in (a) a hypovolemic patient, and (b) a normo- or
hypervolemic patient. Ppl = pleural pressure; Palv = alveolar pressure; RA = right atrium; RV = right ventricle; LA = left atrium; LV = left ventricle.
(a) In a hypovolemic patient, (1) Ppl compresses IVC, decreasing RV preload; (2) Ppl increases RA intramural pressure decreasing RV preload; (3)
Palv compresses pulmonary capillaries in West Zones I and II; (4) capillary pressure exceeds Palv in West Zone III; and (5) an increase in Ppl
decreases LV intramural pressure and therefore afterload. (b) In a normo- or hypervolemic patient, (4) capillary pressure exceeds Palv in West
Zones I, II and III, and (5) an increase in Ppl decreases LV intramural pressure, and therefore afterload is no different than in hypovolemia.
Source: Michard F. Changes in arterial pressure during mechanical ventilation. Anesthesiology 2005;103(2):419–28.
are easily detected. Positive pressure breaths increase

intrathoracic pressure, which compresses the right
atrium (Figure 8.2) and impedes flow from the low-
pressure abdominal venous system, and thereby
decreases right ventricular preload and right ventricular
SV.
In the hypovolemic patient’s West lung zones I and
II, alveolar pressure is greater than pulmonary capillary
pressure, the pulmonary capillaries have lower filling
pressure and are easily compressed, and RV afterload
increases and right ventricular SV decreases. In West
zone III, where alveolar pressure is less than pulmonary
capillary pressure, the brief increase in alveolar pressure
accompanying positive pressure pushes the blood con-
tained in the capillaries toward the left heart, causing an
Figure 8.3 Illustration of the heart–lung interaction with positive increase in LV preload and SV for a few heartbeats
pressure ventilation in a hypovolemic patient. (1) As the pressure
increases with the initiation of a mechanical breath, the flow to the (Figure 8.3) before the decreased right ventricular SV
RA through the vena cava decreases significantly. (2) With a minimal decreases the left ventricle’s preload and SV.
delay the flow in the main pulmonary artery decreases as well, Normovolemic and hypervolemic patients have
secondary to both an afterload increase and preload decrease. (3)
The LV SV initially increases as the pulmonary capillaries are drained normal to high filling pressures, the RV is less com-
by the increased intrathoracic pressure and increase the LV preload pressible, the pulmonary capillaries are less compres-
transiently, however, thereafter preload drops significantly and LV SV sible, and there is no significant variation in SV over
drops (red bars). (A black and white version of this figure will appear
in some formats. For the color version, please refer to the plate the course of the respiratory cycle (Figures 8.2 and 8.3).
section). The HLIM has been studied extensively with var-
Source: Michard F. Changes in arterial pressure during mechanical ious LV SV surrogates, such as calculation of the area
ventilation. Anesthesiology 2005;103(2):419–28. under an arterial pressure tracing and the more easily
82
Table 8.1 Summary of established limitations of the HLIM and the associated type of error.
Limitations Mechanisms for failure Type of error

1 Spontaneous breathing Irregular variations in intrathoracic pressure and False positive (may
activity thus variation in stroke volume do not correlate occasionally be false
with preload dependency negative depending on the
type of breathing)
2 Cardiac arrhythmias The variation in stroke volume is related more to False positive
the irregularity in diastolic time than to the heart-
lung interaction
3 Mechanical ventilation using The small variations in intrathoracic pressure False negative
low tidal volume ventilation secondary to the low tidal volume are insufficient to
(< 8 ml/kg) produce significant changes in intrathoracic pressure
4 Low lung compliance The transmission of changes in alveolar pressure False negative
to the intrathoracic structures is attenuated
5 Open thorax No changes in intrathoracic pressure occur False negative
6 Increased intra-abdominal Threshold values for PPV will be elevated False positive
pressure
7 HR/RR ratio < 3.6 (e.g., If the RR is very high, the number of cardiac cycles False negative
severe bradycardia or high per respiratory cycle may be too low to cause
frequency ventilation) variation in stroke volume
Source: Myatra SN, Monnet X, Teboul JL. Use of “tidal volume challenge” to improve the reliability of pulse pressure variation. Crit
Care. 2017; 21(1):6.
measured pulse pressure variation and systolic pressure blood from the lower extremities back to the heart
change.11 In most studies the threshold for fluid respon- by rapidly lifting the legs straight up. The patient is
siveness has been found to be between 10 and 15%; if usually first placed in a semi-recumbent position with
with positive pressure breaths the calculated SV, pulse the head elevated to 45 degrees, and a baseline mea-
pressure variation, or systolic pressure change varies surement of the SV or surrogate parameter is taken.
more than that, there is about a 90% chance that the Subsequently, by manipulating the patient’s bed, their
administration of fluid will raise the SV and MAP. legs are elevated passively to 30–45 degrees as the
This model has several limitations, the two most upper body is lowered into a horizontal plane, and
common being that it requires mechanical ventilation a second measurement of the SV surrogate is per-
with relatively high tidal volumes (e.g., ≥ 10 ml/kg formed immediately after the position is reached.
versus the standard 6 ml/kg) and it is not applicable in This maneuver mobilizes approximately
the presence of cardiac arrhythmias (Table 8.1). 300–500 mL of venous blood from the extremities
Additionally, this method only assesses the LV for and splanchnic circulation toward the heart (Figure
fluid responsiveness, thus patients with isolated RV fail- 8.4), which increases preload to the RV and then the
ure may have a false positive fluid-responsive state LV, and results in a significant increase in SV in
because the LV preload is decreased when the RV is patients who are fluid-responsive. The effect wears
failing. Fluid administration in this case would worsen off in about 60–90 seconds,12 so it is important to
RV function and overall hemodynamics. Ideally, when perform the PLR and SV measurement rapidly. The
using the HLIM, SV surrogates for the RV and LV advantages of this test are that the fluid challenge is
would both be measured.11 completely reversible and is not dependent on the
presence of mechanical ventilation or sinus rhythm.
Reversible Fluid Challenge It is, however, limited in that it should not be per-
The second model utilizing dynamic parameters for formed in patients who may not tolerate a change in
fluid responsiveness is the reversible fluid challenge. position, have intra-abdominal hypertension, or have
The best validated method of assessment resting on increased intracranial pressure.13,14
this model is the PLR “test” in which an endogenous As an alternative to the PLR test, a small fluid bolus
fluid bolus is delivered by gravitationally draining can be administered over a very short period of time.
(a) (b)
Figure 8.4 Two people are required to perform the passive leg raise test, while ultrasound is performed continuously. (a) The first set of SV
surrogates are recorded after the patient is placed into a semi-recumbent position, with the torso at a 30–45 degree angle and the legs in a
horizontal position. (b) The patient’s legs are placed at a 30–45 degree angle and the torso moved in a horizontal position. Approximately
300–500 ml of venous blood will be mobilized toward the heart. Preload increases rapidly, therefore the second set of measurements must be
obtained quickly.
There is little agreement, however, on the minimal (parasternal long axis, parasternal short axis, four-
fluid bolus size that may increase the SV, and repeated chamber apical, subcostal) and was adequate to drive
administration of small boluses of fluid in an effort to therapy that improved outcome.
increase SV may result in fluid overload.15
Inferior Vena Cava Diameter Evaluation
Ultrasound Techniques to Detect Fluid The most basic ultrasound technique to assess fluid
responsiveness is measurement of the IVC diameter
Responsiveness and its variation during respiration. The IVC is very
The concepts behind fluid responsiveness can be compliant and as such its diameter is dependent on
applied to ultrasound-based techniques. The advan- transmural pressure. During positive pressure
tages of ultrasound are that it is noninvasive, it does inspiration intrathoracic pressure increases and
not require arterial line or stand-alone noninvasive venous return to the heart decreases as blood flow
cardiac output monitoring devices, it is easily porta- from the IVC into the right atrium (RA) is reduced.
ble, and it is repeatable. Increased blood volume in the IVC causes distension
A historical controlled trial compared outcomes in and a measurable increase in diameter. During
220 consecutive shock patients treated with standard expiration the opposite occurs; venous blood in the
management versus limited echocardiography (LE)- IVC returns to the RA, decreased blood volume con-
guided therapy after initial resuscitation. Fluid pre- tracts the IVC, and the diameter is measurably
scription during the first 24 hours was significantly decreased.16 During spontaneous breathing the oppo-
lower in LE patients (49 [33–74] versus 66 site reactions occur and variations in the IVC dia-
[42–100] mL/kg, p = 0.01), although more LE patients meter are just as demonstrable and clinically
received dobutamine (22% versus 12%, p = 0.01). The significant.
patients treated using LE had better 28-day survival
(66% versus 56%, p = 0.04), less stage 3 acute kidney Inferior Vena Cava Image Acquisition
injury (20% versus 39%), and more days alive and free Ultrasound assessment of IVC diameter requires
of renal support (28 [9.7–28] versus 25 [5–28], p = a subcostal long axis (i.e., sagittal plane) view. The
0.04). Thus, an exam that took only 10 minutes to probe is placed to the right of the xiphoid process with
perform provided four main echocardiographic views the index marker pointing to the head (Figure 8.5).
84
Figure 8.5 Subcostal view required for measurement of the IVC diameter. IVC= inferior vena cava, RA = right atrium, HV = hepatic vein, L = Liver,
Ao = aorta, ST = soft tissue layer. (a) Subcostal long axis view of the IVC: The IVC is directly attached to the liver parenchyma, receives blood from
the hepatic vein (blue arrow) just distal to the diaphragm, and clearly drains into the right atrium; the measurement of the diameter is performed
within 3 cm distal to the diaphragm and distal to the HV inlet. (b) It is important to differentiate the IVC (blue dot) from the descending aorta (red
dot). This is done by tilting the ultrasound beam to the patient’s right. (c) The resulting view is a subcostal long-axis view of the descending aorta.
The aorta is clearly separated from the liver by a soft tissue layer and is not connected to a cardiac chamber at this level.
(A black and white version of this figure will appear in some formats. For the color version, please refer to the plate section).
The M-mode is activated and the cursor is placed 1 cm preload-independent (i.e., fluid nonresponsive)
caudal to the branch of the hepatic vein, cutting the patients have only minimal variation. As described
IVC perpendicularly. The IVC diameter is then above for the HLIM, the patient must be mechanically
recorded at end-expiration and end-inspiration. ventilated with tidal volumes of at least 8 mL/kg and
There is no universally agreed upon single location in sinus rhythm in order for this approach to have
that is ideal to measure the IVC diameter. A study in reasonable sensitivity and specificity. The IVC disten-
healthy volunteers assessed the percentage collapse at sibility index (IVC DI) can be calculated using the
three locations along the IVC – the diaphragm, the formulas:
hepatic vein inlet, and the left renal vein. Analysis of
the minimum and maximum IVC diameters showed IVC DI = (IVCmax – IVCmin ) /IVCmin
that the percentage of IVC collapse was equivalent at the or
level of the left renal vein and the hepatic inlet, and were
different from the diameter measured at the level of IVC DI = (IVCmax – IVCmin ) /[(IVCmax + IVCmin )/2]
diaphragm.17 We therefore suggest measuring the IVC
just below the hepatic vein inlet, as it is reproducible and Depending on the study, an IVC DI greater than
the view is easy to obtain. 12–18% signals fluid responsiveness.19,20 An example
The subcostal view may not result in an ade- of this is provided in Figure 8.6.
quate image, particularly in patients after abdom-
inal surgery or recent ingestion of a large meal. In Inferior Vena Cava Collapsibility Index
these cases the IVC may be imaged using a lateral The respiratory variation in IVC diameter can also be
view along the right mid-axillary line, with the used in spontaneously breathing patients to gauge
probe marker pointing cephalad and the probe fluid responsiveness; however, the driving force for
angled horizontally (i.e., coronal plane). The the IVC diameter change is very different during
image obtained should be similar to the subcostal spontaneous ventilation and PPV. During a positive
view, although the IVC collapses asymmetrically in pressure breath the IVC is “distended” during inspira-
an elliptical form and therefore assessment of fluid tion secondary to the increasing intrathoracic pres-
responsiveness based upon distensibility requires sure, whereas during a spontaneous breath the IVC
a different guidance that has only recently been actually “collapses” because it contains less blood as
investigated.18 the pressure decreases within the chest. This pressure
decrease can vary significantly between breaths
Inferior Vena Cava Distensibility Index depending on the patient’s effort and can be
The respiratory variation in the IVC diameter (ΔIVC) exaggerated in a patient who is short of breath, espe-
during PPV can be used to evaluate fluid responsive- cially in the case of airway obstruction or bronchos-
ness; preload-dependent (i.e., fluid-responsive) pasm. This is a significant disadvantage compared to
patients have a large variation in IVC diameter while the patient on PPV, where the change in inspiratory
Figure 8.6 M-mode IVC examination.: (1) Patient with minimal change form inspiration (A) to expiration (B), therefore unlikely to be fluid
responsive. (2) Patient with large change from inspiration (A) to expiration (B), therefore likely to be fluid responsive.
intrathoracic pressure is the same for every breath. independent. Ideally, the test is combined with an
The IVC collapsibility index (IVC CI) is calculated LV fluid responsiveness test or at least interpreted
using the following formula: with knowledge of the underlying LV function at
baseline.
IVC CI = (IVCmax – IVCmin )/IVCmax
IVCmax occurs during expiration and IVCmin occurs Left Ventricular Outflow Tract/Aortic Valve
during inspiration. Studies suggest that a subcostal Velocity Time Integral Evaluation
long-axis IVC CI greater than 40% signals fluid Another technique to evaluate preload responsiveness
responsiveness with a sensitivity of 70% and specifi- involves measuring the variation in the velocity time
city of 80%.21 Zhang et al. performed a systematic integral (VTI) of the left ventricular outflow tract
review pooling eight studies involving 235 patients (VTILVOT) or aortic valve (VTIAV), as both closely
determining the ΔIVC prediction of fluid responsive- correlate with stroke volume. Conceptually, the stroke
ness under both spontaneous breathing and mechan- volume in echocardiography is calculated using the
ical ventilation. The respiratory variation in the IVC same formula in both the LVOT and through the
varied from 12 to 40%, with a pooled sensitivity and aortic valve (AV):
specificity of 0.76 and 0.86, respectively. The predic-
tion of fluid responsiveness based upon ΔIVC was SV = AreaLVOT × VTILVOT
better in mechanically ventilated than in sponta-
neously breathing patients.22 Similar findings were where SV is stroke volume, AreaLVOT is the area of the
reported in a larger meta-analysis including both LVOT, and VTILVOT is the volume time integral of the
pediatric and adult patients, although the sensitivity LVOT. As the areas of the LVOT and AV remain
and specificity were lower.23 Overall, the ΔIVC should constant during systole across varying preload condi-
be interpreted cautiously within the clinical context, tions, the SV is directly proportional to the VTI.
especially if the patient is breathing spontaneously. Therefore, this single echocardiographic parameter
Use of IVC images has the same limitation as the may be used to determine fluid responsiveness.
HLIM. In addition, these measurements only evaluate
RV fluid responsiveness. ΔIVC may result in a false Image Acquisition
positive if the RV is preload-dependent and the LV is The LVOT image is captured in the five-chamber
functioning on the plateau of the FSC and preload- apical view. Once the view is achieved, the pulsed
86
Figure 8.7 Acquisition of the LVOT VTI/ Vmax. (a) The apical five-chamber view is obtained with the LVOT (*) positioned in the center of the
image. (b) The pulsed wave Doppler function is activated and the sample volume (SVol) is placed into the LVOT just above the aortic valve. (c)
The LVOT is represented by the yellow quadrangle and the aortic valve is highlighted in blue. Once in pulsed wave Doppler mode, the
envelopes below the time axis (red dotted line) can be highlighted and the area under the curve can be measured at the point of expiration
and inspiration. (A black and white version of this figure will appear in some formats. For the color version, please refer to the plate section).
wave Doppler mode is activated and the cursor is ΔVTILVOT = (VTImax–VTImin )/[(VTImax+VTImin )/2]
placed to sample the volume in the LVOT. The
Doppler signals appear under the time axis and Studies suggest that a ΔVTILVOT greater than 20%
should have an intense (bright white) border with indicates preload responsiveness.24 It has been proposed
a sharp delineation. The VTILVOT is obtained by tra- that the variation of the maximal velocity (Vmax) may
cing the circumference of the signal, as demonstrated also be used as a surrogate for preload responsiveness.
in Figure 8.7. By simply measuring the difference in the inspiratory
and expiratory LVOT Vmax by pulsed wave Doppler, the
VTILVOT Variation: Heart–Lung Interaction Model variation may be calculated using the same formula as
Using the HLIM, the VTILVOT tracing can be used to for the VTI (Figure 8.7). If the change in Vmax is greater
assess fluid responsiveness. The maximal VTI of the than 12%, the patient is likely fluid-responsive.25
Doppler signal traced during early inspiration Just as IVC assessment does not account for the
(VTImax) and the minimal VTI during expiration possible presence of LV dysfunction that may impair
(VTImin) are measured; VTI variation (ΔVTILVOT) is potential fluid responsiveness, VTI measurements do
then calculated by the formula: not account for the possible presence of RV
dysfunction that may impair the potential LV respon- assessment. Most of the studies evaluating carotid
siveness to a fluid bolus. Therefore, when using the artery imaging for this purpose have used the cor-
HLIM it is advisable to perform fluid responsiveness rected flow time (FTc) or the Vmax as an SV
assessments of both the RV and LV simultaneously surrogate.
(i.e., IVC and VTI assessment) to reduce false positive
results. As these calculations are dependent on heart– Image Acquisition
lung interactions, all of the limitations of the HLIM To obtain a Doppler tracing of the CCA a linear
apply. transducer is used. The patient’s anterior neck is
scanned lateral to the trachea and the CCA is visua-
VTILVOT Variation: Passive Leg Raise Test lized medial to the internal jugular vein (IJV), usually
The limitations of the HLIM are avoided by using the on the medial side of the triangle formed by the
VTILVOT while performing a PLR test. To properly medial and lateral heads of the sternocleidomastoid
perform this exam, the LVOT view must be main- muscle. Subsequently, the transducer is rotated 90
tained throughout the PLR. Several heartbeats must degrees to visualize the CCA in its long axis. The
be averaged (minimum of three in sinus rhythm and 5 transducer is then moved cranially until the carotid
in an atrial arrhythmia) on the pre- and post-PLR bifurcation is seen. The pulsed wave Doppler is acti-
Doppler tracing, and then compared using the same vated, and the cursor and sample volume are placed in
formula presented for calculation of the respiratory the middle of the CCA approximately 2 cm distal to
VTI variation. If the difference is greater than 12.5– the bifurcation. The insonation angle between blood
15%, the patient is likely preload-responsive, with flow and the pulsed wave cursor should be less than 60
a sensitivity range of 77–81% and specificity range of degrees. The orientation of the cursor can be either
93–100%.26 cephalad or caudal, but should remain the same
throughout the assessment (Figure 8.8).
VTILVOT Variation: Mini-Fluid Bolus
The focus of this chapter has thus far been on PLR and Carotid Doppler Evaluation: Heart–Lung Interaction
respiration to augment preload, and measurement of Model
a hemodynamic variable in an attempt to predict fluid In mechanically ventilated patients in septic shock,
responsiveness, usually to a fluid bolus of ≥ 500 ml. the respiratory variation of the common carotid max-
A less well-established method for predicting fluid imal velocity (CCVmax) may be used to determine
responsiveness is to administer a smaller bolus of fluid responsiveness. In a study of 59 unstable
fluid (e.g., 100 ml) and measure the VTILVOT varia- patients, fluid responsiveness was determined using
tion, diminishing the risk of fluid overload. In a small a CO from a pulmonary artery catheter and compared
study, 39 mechanically intubated patients received to the change in CCVmax. CCVmax change was super-
a 100 mL hydroxyethyl starch bolus over 1 min to ior to stroke volume variation and pulse pressure
assess pre- and post-bolus VTILVOT, followed by an variation in differentiating fluid responders from
additional 400 mL and repeated measurement. In nonresponders; a change of 14% was the cutoff
response to this mini-fluid challenge, an increase in value.27 In a similar study of 40 coronary artery bypass
VTI variation ≥ 10% predicted fluid responsiveness to grafting patients with a normal LVEF, CCVmax was
a total volume of 500 ml ≥ 15%, with a sensitivity of the most accurate predictor of fluid responsiveness,
95% and specificity of 78%.21 The literature support- with a cutoff value of 11%.28
ing this approach, however, is quite scant.
Carotid Doppler Evaluation: Passive Leg Raise Test
Carotid Velocity Time Integrals Different carotid parameters have been evaluated in
Obtaining the apical five-chamber view is often diffi- conjunction with the PLR test. One study assessed the
cult in acutely ill patients, especially in the operating change of carotid blood flow (carotid area x carotid
room. Therefore, evaluation of more accessible vascu- VTI x HR) and found it to be a very accurate predictor
lature such as the central arteries close to the aortic of fluid responsiveness; with a cutoff value of 20% the
arch is of interest. There is increasing evidence that sensitivity was 94% and the specificity was 86%.29 This
common carotid artery (CCA) Doppler assessment was confirmed by other studies, with similar cutoff
may be a valuable alternative to LVOT flow values ranging from 22.6 to 24.6% the sensitivity was
88
Figure 8.8 Acquisition of the CCA Vmax. As

indicated in the 2D image at the top of the
image, the sample volume is placed in the
middle of the artery and at an angle less
than 60 degrees to the blood flow. Once the
Doppler signal has been recorded the
maximal and minimal flow velocities are
measured, in this case 114.5 cm/s and
132 cm/s.27 (A black and white version of
this figure will appear in some formats. For
the color version, please refer to the plate
section).
60–76% and specificity was 88.7–92%.30,31 While increased to greater than 8 ml/kg for 1 minute,
these studies have all been small (e.g., 20–40 patients) which will improve the accuracy of the HLIM with-
the results are nevertheless promising. out worsening lung function.32 Assessing SV varia-
tion via the CCA is especially appealing, since the
Conclusions anesthesiologist frequently has access to the CCA,
while access to the chest for performing TTE is often
Hypotension is common in hospitalized patients and
limited. In situations in which the HLIM is not
delineating the etiology is vital to successful treat-
applicable, the PLR test can be used as an alternative.
ment. The long-standing practice of administering
This method has fewer limitations with regard to
large volumes of fluid to target the static parameters
determining fluid responsiveness, with LVOT VTI
of heart rate, blood pressure, and CVP have greater
measurements being the most accurate. Limitations
potential risk than previously recognized, as it is now
of the PLR test are that it requires two people to
appreciated that fluid overload may cause morbidity
perform quickly, and can be difficult to perform in
and mortality. Echocardiography can be used to elu-
the OR as the necessary positional changes are pro-
cidate the causes of shock, determine fluid respon-
hibitive during most surgical procedures. User skill
siveness, identify patients who likely will respond to
in acquiring and interpreting the images is variable,
fluid administration, and spare nonresponders the
but with widespread training and availability of
increased risk of morbidity and mortality that is
ultrasound, these tests can provide guidance in
associated with fluid overload.
fluid resuscitation and help prevent unnecessary
It is known that stroke volume variations occur
fluid overload.
secondary to heart–lung interactions, and that
LVOT VTI variation and PPV are the most accurate
methods of assessing preload responsiveness. There References
are several limitations to acquiring reliable data, 1. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed
including the need for mechanical ventilation with therapy in the treatment of severe sepsis and septic
tidal volume greater than 8 ml/kg, which is not the shock. N Engl J Med 2001;345(19):1368–77.
current standard of care. In patients undergoing 2. ProCESS Investigators, Yealy DM, Kellum JA, et al.
general anesthesia, some of the limitations can be A randomized trial of protocol-based care for early
overcome. For example, the tidal volume can be septic shock. N Engl J Med 2014;370(18):1683–93.
89
3. Peake SL, Bailey M, Bellomo R, et al. Australasian mini-fluid challenge accurately predicts fluid
resuscitation of sepsis evaluation (ARISE): a responsiveness. Br J Anaesth 2015;115(3):449–56.
multi-centre, prospective, inception cohort study. 16. Jue J, Chung W, Schiller NB. Does inferior vena cava
Resuscitation 2009;80(7):811–8. size predict right atrial pressures in patients receiving
4. Eskesen TG, Wetterslev M, Perner A. Systematic mechanical ventilation? J Am Soc Echocardiogr 2014;5
review including re-analyses of 1148 individual data (6):613–9.
sets of central venous pressure as a predictor of fluid 17. Wallace DJ, Allison M, Stone MB. Inferior vena cava
responsiveness. Intensive Care Med 2016;42 percentage collapse during respiration is affected by
(3):324–32. the sampling location: an ultrasound study in healthy
5. Osman D, Ridel C, Ray P, et al. Cardiac filling volunteers. Acad Emerg Med 2010;17(1):96–9.
pressures are not appropriate to predict hemodynamic 18. Shah R, Spiegel R, Lu C, Crnosija I, Ahmad S.
response to volume challenge. Crit Care Med 2007;35 Relationship between the subcostal and right lateral
(1):64–8. ultrasound views of inferior vena cava collapse:
6. Kumar A, Anel R, Bunnell E, et al. Pulmonary artery implications for clinical use. Chest 2018;153
occlusion pressure and central venous pressure fail to (4):939–45.
predict ventricular filling volume, cardiac 19. Barbier C, Loubi Res Y, Schmit C, et al. Respiratory
performance, or the response to volume infusion in changes in inferior vena cava diameter are helpful in
normal subjects. Crit Care Med 2004;32(3):691–9. predicting fluid responsiveness in ventilated septic
7. Monnet X, Julien F, Ait-Hamou N, et al. Lactate and patients. Intensive Care Med 2004;30(9):1–7.
venoarterial carbon dioxide difference/arterial-venous 20. Feissel M, Michard F, Faller J-P, Teboul J-L. The
oxygen difference ratio, but not central venous oxygen respiratory variation in inferior vena cava diameter as
saturation, predict increase in oxygen consumption in a guide to fluid therapy. Intensive Care Med 2004;30
fluid responders. Crit Care Med 2013;41(6):1412–20. (9):1834–7.
8. Michard F, Teboul J-L. Predicting fluid responsiveness 21. Muller L, Bobbia X, Toumi M, et al. Respiratory
in ICU patients: a critical analysis of the evidence. variations of inferior vena cava diameter to predict
Chest 2002;121(6):2000–8. fluid responsiveness in spontaneously breathing
9. Boyd JH, Forbes J, Nakada T-A, Walley KR, Russell JA. patients with acute circulatory failure: need for
Fluid resuscitation in septic shock: a positive fluid a cautious use. Crit Care 2012;16:R188.
balance and elevated central venous pressure are 22. Zhang Z, Xu X, Ye S, Xu L. Ultrasonographic
associated with increased mortality. Crit Care Med measurement of the respiratory variation in the
2011;39(2):259–65. inferior vena cava diameter is predictive of fluid
10. National Heart, Lung, and Blood Institute Acute responsiveness in critically ill patients: systematic
Respiratory Distress Syndrome (ARDS) Clinical Trials review and meta-analysis. Ultrasound Med Biol
Network, Wiedemann HP, Wheeler AP, et al. 2014;40(5):845–53.
Comparison of two fluid-management strategies in 23. Long E, Oakley E, Duke T, Babl FE, Paediatric research
acute lung injury. N Engl J Med 2006;354(24):2564–75. in emergency departments international collaborative
11. Marik PE, Cavallazzi R, Vasu T, Hirani A. Dynamic (PREDICT). Does respiratory variation in inferior
changes in arterial waveform derived variables and vena cava diameter predict fluid responsiveness:
fluid responsiveness in mechanically ventilated a systematic review and meta-analysis. Shock 2017;47
patients: a systematic review of the literature. Crit Care (5):550–9.
Med 2009;37(9):2642–7. 24. Slama M, Masson H, Teboul J-L, et al. Monitoring of
12. Monnet X, Teboul J-L. Assessment of volume respiratory variations of aortic blood flow velocity
responsiveness during mechanical ventilation: recent using esophageal Doppler. Intensive Care Med 2004;30
advances. Crit Care 2013;17(2):217. (6):1182–7.
13. Monnet X, Rienzo M, Osman D, et al. Passive leg 25. Feissel M, Michard FDR, Faller J-P, Teboul J-L. The
raising predicts fluid responsiveness in the critically ill. respiratory variation in inferior vena cava diameter as
Crit Care Med 2006;34(5):1402–7. a guide to fluid therapy. Intensive Care Med 2004;30
14. Boulain T, Achard J-M, Teboul J-L, et al. Changes in (9):1–4.
BP induced by passive leg raising predict response to 26. Lamia B, Ochagavia A, Monnet X, et al.
fluid loading in critically ill patients. Chest 2002;121 Echocardiographic prediction of volume
(4):1245–52. responsiveness in critically ill patients with
15. Mallat J, Meddour M, Durville E, et al. Decrease in spontaneously breathing activity. Intensive Care Med
pulse pressure and stroke volume variations after 2007;33(7):1125–32.
90
27. Ibarra-Estrada MÁ, López-Pulgarín JA, Mijangos- induced by passive leg raising. Am J Med Sci 2018;355
Méndez JC, Díaz-Gómez JL, Aguirre-Avalos G. (2):168–73.
Respiratory variation in carotid peak systolic velocity 31. Taggu A, Darang N, Patil S. Carotid artery flow
predicts volume responsiveness in mechanically time corrected (FTc) changes induced by passive leg
ventilated patients with septic shock: a prospective raise (PLR) can predict fluid responsiveness in
cohort study. Crit Ultrasound J 2015;7(1):12. mechanically ventilated (MV) patients. Chest
28. Song Y, Kwak YL, Song JW, Kim YJ, Shim JK. 2016;150(4):294A.
Respirophasic carotid artery peak velocity variation as 32. Myatra SN, Prabu NR, Divatia JV, et al. The changes in
a predictor of fluid responsiveness in mechanically pulse pressure variation or stroke volume variation
ventilated patients with coronary artery disease. Br after a “Tidal Volume Challenge” reliably predict fluid
J Anaesth 2014;113(1):61–6. responsiveness during low tidal volume ventilation.
29. Marik PE, Levitov A, Young A, Andrews L. The use of Crit Care Med 2017;45(3):415–21.
bioreactance and carotid Doppler to determine volume 33. Michard F. Changes in arterial pressure during
responsiveness and blood flow redistribution following mechanical ventilation. Anesthesiology 2005;103
passive leg raising in hemodynamically unstable (2):419–28.
patients. Chest 2013;143(2):364–70.
34. Myatra SN, Monnet X, Teboul JL. Use of “tidal volume
30. Jalil B, Thompson P, Cavallazzi R, et al. Comparing challenge” to improve the reliability of pulse pressure
changes in carotid flow time and stroke volume variation. Crit Care 2017;21(1):6.
Chapter
Point-of-Care Abdominal Ultrasound
9 Shaun L. Thompson and Daniel W. Johnson
Introduction depth limited to 5–10 cm; these higher-frequency

linear probes are best used for vascular ultrasound
Ultrasonography is an indispensable imaging mod-
and peripheral nerve blocks.
ality that is increasingly utilized by perioperative
physicians at the bedside. Abdominal ultrasound
should be performed when patient instability in set-
Focused Assessment with Sonography
tings that suggest abdominal pathology may be the in Trauma
cause. The focused assessment with sonography in The FAST exam was most commonly performed to
trauma (FAST) examination is the most common detect free fluid following abdominal trauma. This
Point-of-Care Ultrasound (PoCUS) approach uti- modality was first described by Asher et al. in 1976,
lized in trauma and is applicable in both ICUs and when it was used specifically to assess possible splenic
PACUs. Abdominal sonography may be used to eval- injury after blunt abdominal trauma (BAT);4,5 it came
uate details of the kidneys, aorta, liver, spleen, into common use in the United States in the 1990s and
gallbladder, bowel, bladder, and intrauterine preg- is now routine in this setting.2,4 Numerous studies
nancy.1 This chapter will describe various ultrasound have shown that the FAST exam has high sensitivity
techniques and protocols for examining the abdo- and specificity for detecting free fluid in the abdo-
men in critical care and perioperative medicine that men.1,4–7
are of interest to anesthesiologists. The original purpose of the FAST exam was to
guide trauma surgeons to opt for immediate surgi-
Types and Performance of Abdominal cal intervention in BAT patients with ultrasound
evidence of free fluid that is likely hemoperitoneum
Ultrasound secondary to splenic injury. While BAT patients go
to the operating room far less frequently now than
Transducer Selection in the past decades (for a variety of reasons), the
Most modern ultrasonography machines can be used imaging techniques of the FAST exam remain use-
to perform abdominal ultrasonography. There are ful. It is a valuable tool for anesthesiologists, as it
several types of ultrasound transducers, but only can be used to detect abdominal bleeding in peri-
two types allow for appropriate imaging of the operative and critical care settings.2 Additionally,
abdominal structures. These are the phased array serial FAST exams allow anesthesiologists and
transducer (the ideal transducer for transthoracic intensivists to evaluate persistent hypotension. For
echocardiography) and the curvilinear transducer, example, in a patient with multiple traumatic inju-
both of which emit ultrasound waves at frequencies ries undergoing emergency craniotomy for
from 1 to 5 MHz.2,3 These low frequencies provide intracranial hemorrhage, the anesthesiologist can
deep tissue penetration up to 30 cm and allow the concurrently perform periodic FAST exams to
sonographer to obtain ultrasound views of the rule out intra-abdominal bleeding.
abdominal viscera and vasculature, although ima- Because the FAST examination enables rapid,
ging at increased depths comes at the expense of noninvasive detection of free fluid in the abdomen,1
image resolution. Abdominal imaging usually is not it reduces the need for computed tomography (CT)
of the same image quality as more superficial ima- scans and diagnostic peritoneal lavages that were the
ging with high-frequency linear transducers that prior standard,8–13 and thereby decreases the time to
have ultrasound frequencies of 5–15 MHz and a initiate potentially life-saving interventions and
92
surgical procedures.9 Reduced need for CT scans

decreases transportation time and its associated risk,
radiation exposure, and cost.8,14–17
The classic FAST examination involves imaging
four regions: perihepatic, perisplenic, pelvic, and peri-
cardial (known as the 4 Ps).9,18,19 The extended FAST
(e-FAST) exam includes lung ultrasound to evaluate
for possible pneumothorax or hemothorax in the
unstable patient.3 The e-FAST exam will not be dis-
cussed in detail in this chapter, as lung ultrasound is
discussed in Chapter 7.
It is important to perform each component of the
FAST exam in the same order every time to ensure
that all areas are examined thoroughly. In the perio-
Figure 9.1 Free fluid in the RUQ: note the black (hypoechoic) area
perative arena the exam order should be right upper in Morrison’s pouch, between the liver and the kidney.
quadrant (RUQ), left upper quadrant (LUQ), pelvis,
and subcostal. In order to rule out the presence of free
fluid, multiple planes must be viewed within each level of the xiphoid process in order to visualize both
window of the examination. This is accomplished by the liver and the right kidney. Image depth should be
“fanning” the ultrasound sector through organs to adjusted to ensure simultaneous visualization of the
ensure the entire space has been examined. liver and the right kidney.2,9 Once the liver and the
Consistency in exam performance between examiners right kidney are visualized, the transducer should be
reduces variation and increases interobserver fanned through this area to ensure that the entire
reliability. space is imaged.2,22 The transducer should also be
moved cephalad to visualize the right diaphragm in
Right Upper Quadrant View (Perihepatic) order to detect fluid in the subphrenic space or in the
Beginning the FAST exam in the RUQ (the peri- right pleural space.3 Placing the patient in
hepatic view) has many advantages. In comparison Trendelenberg position can increase the sensitivity
to the other windows, it is easiest to perform and is of the FAST examination in the RUQ by moving
the most sensitive for detecting free peritoneal fluid free fluid up into this area.2,3,22,24 A common pitfall
following BAT.2,6,20–22 In this view, the clinician in the RUQ view is failure to fan the transducer
seeks to rule out free fluid in three spaces: (1) the caudad in order to rule out free fluid adjacent to the
subphrenic space (i.e., between the diaphragm and inferior pole of the right kidney. In upright patients,
the dome of the liver); (2) the hepato-renal space; this area is a frequent site for accumulation of blood
and (3) the inferior pole of the right kidney. In this (Figure 9.1).
quadrant free fluid most commonly accumulates in
the hepato-renal space (also known as Morrison’s Left Upper Quadrant View (Perisplenic)
pouch).1,2,6,22,23 Recent literature, however, has The left upper quadrant (LUQ) contains the spleno-
shown that free fluid in the abdomen may initially renal space, left subphrenic space, and left paracolic
be seen at the caudal edge of the liver, prior to its gutter. Of the four FAST examination views, this view
appearance in the hepato-renal space.23 If free fluid is the most difficult to acquire.21 The view is obtained
is found in the RUQ then it is treated as a positive by placing the probe along the posterior axillary line
examination and patient status will determine the with the superior edge of the probe at the level of the
course of action from that point onward.2 The xiphoid process. Starting with the “knuckles to the
sensitivity with a single view FAST exam using bed” and aiming the ultrasound beam 45 to 60 degrees
only the RUQ ranges from 51 to 82%.20 anteriorly increases the likelihood of obtaining ade-
To perform the RUQ exam the transducer is quate LUQ views. The probe is fanned in an anterior
placed in a longitudinal orientation with the probe to posterior fashion in order to completely scan the
indicator directed cephalad in the mid-axillary line.2 LUQ to include the spleno-renal and subphrenic
The superior aspect of the transducer should be at the space.2 The subphrenic space can be difficult to
vesicles in males and the small amount of common

benign fluid present in females.2,26,27
The urinary bladder is the main anatomic land-
mark to be visualized.2,3,20 Visualization of the
bladder ensures that the image depth will include the
common sites of free fluid collection. The rectovesical
pouch is the most common area where free fluid is
seen in males in the pelvic view.2,3,20 In females free
fluid can collect in both the uterovesicular pouch and
the rectouterine space (also known as the pouch of
Douglas).2 A full bladder is helpful for visualization,
as it allows not only for proper anatomic guidance but
also for ease of visualizing free fluid in the pelvis.2,3,20
If the urinary bladder is not full, a Foley catheter can
Figure 9.2 Free fluid in the LUQ: note the black (hypoechoic) area
between the diaphragm and the spleen.
be placed and fluid instilled into the bladder to assist
in visualization.3 In comparison with the RUQ view,
smaller volumes of fluid are detectable in the pelvic
visualize due to artifact from aerated lung and rib view.6
shadows.2
The left paracolic gutter located inferior to the Subcostal View
spleen is the area where free fluid is most commonly The final window to be examined in the FAST exam-
found in the LUQ view.21 The paracolic gutter can be ination is the pericardial window or the subcostal
most easily visualized by moving the probe caudad view. Two main questions must be answered by the
until this space is found.3 Cephalad movement of the subcostal view: (1) Does the patient have cardiac
probe will allow for visualization of the left pleural activity? (2) Is there fluid within the pericardial
space.3 Free fluid is typically not isolated to the LUQ, space?
and if it is seen in the LUQ it is usually found in other This view is obtained by placing the transducer
areas.21 If free fluid is isolated to the LUQ, it is most just below the xiphoid process and tilting the ultra-
commonly seen in the left paracolic gutter.21 Because sound beam cephalad almost parallel to the skin sur-
fluid is generally not isolated to this area, if free fluid is face in order to visualize the heart. The marker of the
found within the LUQ, full visualization of the other transducer should be oriented toward the patient’s left
areas of the FAST examination are warranted and an side. The liver acts as an acoustic window and is found
exam limited to the LUQ would be inadequate (Figure in the near field in this view.2 The right ventricle
9.2).21,25 occupies the majority of the border with the liver
and is well visualized in this window.3 In this view
Pelvic View the left ventricle lies beneath the right ventricle, and
Following ultrasound of the right and left upper quad- the atria are well visualized. Global ventricular func-
rant, the pelvic scan should be performed. This view is tion can be assessed to determine whether this may be
obtained by placing the transducer about 1–2 cm contributing to hemodynamic compromise.
superior to the pubic symphysis. The probe is first The subcostal view readily detects the presence of
placed in the transverse plane with the probe marker significant pericardial effusion.1–3,22 The pericardium
toward the patient’s right side. Subsequently the normally appears as a brightly echogenic line directly
probe is placed in the sagittal plane with the probe adjacent to the myocardium. Pericardial fluid will
marker toward the patient’s head.3 Again, fanning of appear as a hypoechoic or anechoic stripe between
the probe to visualize all areas of the pelvis is impor- the myocardium and the pericardium.20 The echo
tant to be sure that free fluid is not overlooked. Care texture of the fluid can appear complex when clot
must be taken not to mistake normal fluid collections begins to form in the pericardial space (Figure 9.3).28
or anatomy in the pelvis as pathological. Normal While formal diagnosis of pericardial tamponade,
observations that potentially may be misinterpreted pulmonary embolism, and fluid responsiveness is
as pathological include the prostate gland and seminal beyond the scope of the traditional FAST exam,
94
diameter IVC with lack of respiratory variability indi-

cates unlikely fluid responsiveness. The IVC can be
located by rotating the probe in a counterclockwise
fashion, while keeping the right atrium in sight from
the subcostal view. Because of this, an IVC evaluation
is sometimes done as an extension of the FAST exam.
In critically ill patients, IVC findings often do not
completely explain the clinical picture, and other stu-
dies may be needed to fully assess volume status and
cardiovascular function.31
Limitations of the FAST Examination

While the FAST examination can provide rapid diag-
Figure 9.3 Massive pericardial effusion: note the black nosis of free fluid and expedite care, it has limitations
(hypoechoic) area below the liver, between the pericardium, and the and pitfalls. First, there is a chance of a false negative
extremely compressed right ventricle. This patient had sustained a study. This may be due to sonographer error or insuf-
knife wound to the anterior chest and required emergency cardiac
surgery. ficient time after injury for enough fluid to accumu-
late in order to be imaged.2,9 Free fluid in the
anesthesiologists should be familiar with their ultra- abdomen can usually be discovered when as little as
sound appearance. Pericardial tamponade should be 200 ml is present.9 If an initial FAST examination is
in the differential diagnosis if a significant pericardial negative, repeat exams over the course of the patient’s
effusion is detected during a state of shock. Right care, particularly if there are significant clinical
atrial collapse (during ventricular systole) followed changes or high suspicion, may eventually reveal free
by right ventricular collapse (during ventricular dia- fluid in the abdomen.2,20 False negative results can
stole) typically is the first ultrasound sign of cardiac also be influenced by other severe injuries. Studies
tamponade.28 In severe states of tamponade, the left have shown that in patients with minor solid organ
atrium and left ventricle will show signs of collapse. injuries, those with comorbid severe head injuries had
These findings generally appear late in the course of increased rates of false negative results on FAST
pericardial fluid accumulation, just prior to a cardiac examination.6,32 False negative results can lead to
arrest.28 poor patient outcomes, as the undetected pathology
Pulmonary embolism is a common cause of acute may lead to continued deterioration and a missed
hemodynamic collapse in critically ill patients. The opportunity for early intervention.
subxiphoid view can assess right ventricular size, False positive findings can arise from many
function, and systolic pressure. Since right ventricular factors and can lead to unnecessary procedures
dilation and dysfunction can be associated with other and testing. Physiologic fluid collections may be
pathologies, these must be ruled out.28 McConnell’s misinterpreted as traumatic free fluid. These situa-
sign is an echocardiographic finding associated with tions arise due to ovarian cyst rupture, ventriculo-
acute pulmonary embolism, consisting of normal peritoneal (VP) shunts, peritoneal dialysis, and
function of the right ventricular apex and right ven- preexisting ascites.3 It can be difficult to discern
tricular free wall hypokinesis or akinesis.29,30 Overall, between new intra-abdominal bleeding and ascites,
echocardiography has a low sensitivity to diagnose as they appear similar on ultrasound.20 Pericardial
acute pulmonary embolism, but its sensitivity for fat pad or preexisting pericardial effusion may be
massive pulmonary embolism is much higher.29 described as new free fluid and may lead to unne-
Ultrasound to assess fluid responsiveness is cessary pericardiocentesis and potential patient
described in greater detail in Chapter 8. Subcostal harm.3
imaging of the inferior vena cava (IVC) comprises Technical difficulty can lead to poor imaging and
one of the most useful modalities in the evaluation inability to accurately assess free fluid. This can be
of fluid responsiveness. In general, a small diameter secondary to patient body habitus, as morbid obesity
IVC with marked respiratory variability in diameter can make it difficult to obtain adequate visualization.
indicates likely fluid responsiveness, while a large Traumatic injury itself can make ultrasonography
difficult, due to subcutaneous emphysema, patient

position, slow rate of bleeding, bowel gas, and
intra-abdominal adhesions.3 Retroperitoneal hema-
toma and hemorrhage are very difficult to assess with
ultrasonography as well.1,3,9 The FAST exam is also
very poor in determining hollow viscus or solid
organ injury that is not associated with free fluid in
the abdomen, such as early bowel injury or pancrea-
tic insults.9,33
Abdominal Aortic Ultrasound

Ultrasound to evaluate the aorta in patients with
undifferentiated hypotension can be a useful tool to
assess for possible aneurysmal disease as a causa- Figure 9.4 Abdominal aortic aneurysm: this patient presented to
tive factor.20 Abdominal ultrasonography has a the emergency department with abdominal pain and hypotension.
The transverse view of the aorta was obtained just cephalad to the
high sensitivity and specificity for detecting umbilicus. The abdominal aortic aneurysm is greater than 5 cm in
abdominal aortic aneurysms (AAA).1 It can also diameter.
be used to assess for possible aortic dissection in
cases where the patient presents with symptoms of
limb ischemia and/or acute severe abdominal measuring the diameter of the aorta: from inner wall
pain.34 Since it is critical that these diagnoses are to inner wall (ITI) and outer wall to outer wall
made rapidly, the portability and speed of ultra- (OTO).35,36 The ITI measurement has greater inter-
sound makes it an optimal initial exam. observer reliability than the OTO measurement.35
The abdominal aortic ultrasound scanning tech- The anteroposterior diameter is more accurate and
nique is similar to that used for the FAST exam- reproducible than the transverse diameter in the mea-
ination. The curvilinear probe is best suited for this surement of abdominal aortic aneurysm.35,37 If mural
study, as the aorta is a deep lying structure hematoma is noted, then OTO measurements should
within the abdomen. With the patient in the supine be utilized.20 Measurements greater than 5.5 cm are
position, the probe is initially placed in the mid- concerning in a hypotensive patient, as the risk for
epigastrium of the subcostal area in a transverse rupture is high.20 If an aneurysm is found, consulta-
plane.20 Firm pressure should be applied to the tion with a vascular surgeon is prudent and further
probe in order to displace any bowel gas that may evaluation with a CT scan is often needed for surgical
overlie the abdominal aorta, which is seen just to planning (Figure 9.4).
the left of the lumbar spine.20 The aorta can then Ultrasound can also be utilized to detect aortic
be scanned from the epigastrium down to the dissection. These patients may present with symp-
bifurcation, at approximately the level of the umbi- toms of obstruction of the branch arteries from the
licus. The takeoffs of the celiac, superior mesen- aorta, including mesenteric ischemia, limb ischemia,
teric, and sometimes inferior mesenteric arteries and paraplegia.34 Ultrasound findings that correlate
can be visualized and assessed during this portion with aortic dissection include visualization of an inti-
of the exam. The transducer should then be turned mal flap and intramural thrombus.34 The intimal flap
clockwise 90 degrees to view the aorta in a long- can sometimes originate in the thoracic aorta and
itudinal plane, as it is often easier to appreciate the extend into the abdominal aorta, affecting the branch
extent of an aortic aneurysm when viewed in this arteries, including the celiac, superior mesenteric, and
plane. inferior mesenteric. Rapid recognition of this pathol-
Once the aorta is identified, care should be taken ogy can lead to quick surgical intervention and better
not only to follow it through its entire course through patient outcomes. Abdominal aortic ultrasound
the abdomen but also to get accurate measurement of decreases the time to initial identification of aortic
its diameter and determine the possible presence of an dissection.34,38 Further evaluation by CT imaging,
aneurysm. The normal diameter of the abdominal however, is usually needed for definitive treatment
aorta is less than 3 cm.20 There are two methods for planning.
96
Gallbladder Ultrasound expert consultation and CT or magnetic resonance

cholangiopancreatography may be necessary for defi-
Disease of the biliary system is a problem that critical
nitive diagnosis.
care physicians commonly encounter, and ultrasound
is the primary modality for assessment of infectious
processes and obstruction.1,2,28 Signs of gallbladder Detection of Pneumoperitoneum
disease include distention, wall thickening, perichole- Perforation of hollow viscus in the abdominal cavity
cystic fluid, biliary sludge, and sonographic Murphy’s can cause critical illness and rapid clinical deteriora-
sign. tion. Quick identification of this abnormality in a
The gallbladder examination is performed by pla- patient with clinical signs consistent with pneumo-
cing the ultrasound probe just inferior to the xiphoid peritoneum allows for swift and appropriate medical
process with the probe indicator directed cephalad, and surgical intervention to reduce patient morbidity.
directing the ultrasound beam into the liver parench- Free air in the abdomen has a few distinctive features
yma. The probe is then moved toward the right side of that sonographers must be aware of. Key findings are
the patient’s body until the anechoic gallbladder is enhancement of the peritoneal stripe and reverbera-
located.22 This examination can also be performed tion artifact from the peritoneum.1,22 Other findings
during the FAST exam of the RUQ. such as enhanced echogenic free fluid in the abdomen
Interrogation of the common bile duct should be due to spillage of enteric contents can also be
performed when assessing for possible obstructive appreciated.22
disease, particularly if stones are discovered. The Identification of free air is typically discovered
common bile duct can be differentiated from vessels incidentally during other ultrasound examinations,
of the portal system with the aid of color Doppler. An such as the FAST examination or ultrasound imaging
anechoic structure leaving the gallbladder with no of the aorta or gallbladder.22 Air collections from
flow will positively identify this structure.22 pneumoperitoneum are typically superficial and can
Common bile duct measurements greater than 4 cm be detected by any transducer that can image a depth
are suspicious for obstructive disease. greater than the thickness of the abdominal wall. The
Several errors may occur when performing ultra- key to recognizing air on an abdominal ultrasound
sound assessment of the biliary system. False positives study is that it appears as A-lines, a repeating pattern
may instigate further unnecessary testing and inter- of horizontal lines generated by reverberation arti-
ventions. Generalized edema, pre-eclampsia, cardiac fact, the same as it appears on thoracic ultrasound
failure, and renal failure often are associated with (see Chapter 7, “Point-of-Care Ultrasound of the
gallbladder wall thickening and may be mistaken for Lung”). Visualization of free air between the liver
pathologic changes suggestive of cholecystitis.28 Care and abdominal wall reduces the chance of a false
should be taken to reduce these errors, but ultimately positive.22
FAST exam key points
Views Main components Notes

Right Upper Quadrant (perihepatic) R subphrenic Easiest to perform most sensitive to detect
Hepato-renal (Morrison’s pouch) fluid
Inferior pole kidney
Left Upper Quadrant (perisplenic) Spleno-renal Most difficult to perform
L subphrenic
L parecolic gutter
Pelvic Bladder Avoid mistaking normal findings for
Rectovesical piuch pathology
Rectouterine space (Douglas’s pouch)
Subcostal Cardiac activity Covered in more detail in other chapters
Pericardial effusion
10. Ollerton JE, Sugrue M, Balogh Z, et al. Prospective

Conclusions study to evaluate the influence of FAST on trauma
Abdominal ultrasonography is a powerful tool in the patient management. J Trauma 2006;60:785–91.
armamentarium of anesthesiologists, critical care 11. Jehle D, Guarino J, Karamanoukian H. Emergency
physicians, and emergency room physicians. Use of department ultrasound in the evaluation of blunt
ultrasound to evaluate for possible abdominal pathol- abdominal trauma. Am J Emerg Med 1993;11
ogy following trauma and abdominal surgical proce- (4):342–6.
dures in unstable patients can provide useful 12. Moore CL, Copel JA. Point-of-care ultrasonography. N
information to diagnose potential life-threatening Engl J Med. 2011;364:749–57.
conditions. Alone, it can guide rapid medical and 13. Noble VE. Think ultrasound when evaluating for
interventional treatments and improve patient out- pneumothorax. J Ultrasound Med 2012;31(3):501–4.
comes. It may also reduce the need for other imaging 14. Dammers D, El Moumni M, Hoogland II, Veeger N,
studies (e.g., CT scan) and possibly prevent unneces- Ter Avest E. Should we perform a FAST exam in
sary procedures (e.g., diagnostic peritoneal lavage). haemodynamically stable patients presenting after
blunt abdominal injury: a retrospective cohort study.
Scand J Trauma Resusc Emerg Med 2017;25:1.
References
15. Melniker LA, Leibner E, McKenney MG, et al.
1. Kameda T, Taniguchi N. Overview of point-of-care
Randomized controlled clinical trial of point-of-care,
abdominal ultrasound in emergency and critical care. J
limited ultrasonography for trauma in the emergency
Intensive Care 2016;4:53.
department: the first sonography outcomes
2. Pace J, Arntfield R. Focused assessment with assessment program trial. Ann Emerg Med
sonography in trauma: a review of concepts and 2006;48:227–35.
considerations for anesthesiology. Can J Anaesth
16. Boulanger BR, McLellan BA, Brenneman FD, Ochoa J,
2018;65:360–70.
Kirkpatrick AW. Prospective evidence of the
3. American Institute of Ultrasound in Medicine. superiority of a sonography-based algorithm in the
AIUM practice guideline for the performance of assessment of blunt abdominal injury. J Trauma
the focused assessment with sonography for 1999;47:632–7.
trauma (FAST) examination. J Ultrasound Med
17. Arrillaga A, Graham R, York JW, Miller RS. Increased
2014;33:2047–56.
efficiency and cost-effectiveness in the evaluation of the
4. McGahan JP, Rose J, Coates TL, Wisner DH, blunt abdominal trauma patient with the use of
Newberry P. Use of ultrasonography in the patient ultrasound. Am Surg 1999;65:31–5.
with acute abdominal trauma. J Ultrasound Med
18. Rozycki GS, Newman PG. Surgeon-performed
1997;12:653–62.
ultrasound for the assessment of abdominal injuries.
5. Laing F, Federle M, Jeffery B, Brown TW. Ultrasonic Adv Surg 1999;33:243–59.
evaluation of patients with acute right upper quadrant
19. Rozycki GS, Ballard RB, Feliciano DV, Schmidt JA,
pain. Radiology 1981;140:449.
Pennington SD. Surgeon-performed ultrasound for
6. Savatmongkorngul S, Wongwaisayawan S, Kaewlai R. the assessment of truncal injuries: lessons learned from
Focused assessment with sonography for trauma: 1540 patients. Ann Surg 1998;228:557–67.
current perspectives. Open Access Emergency Medicine
20. Jehle D, Stiller G, Wagner D. Sensitivity in detecting
2017;9:57–62.
free intraperitoneal fluid with the pelvic views of the
7. Elbaih AH, Abu-Elela ST. Predictive value of focused FAST exam. Am J Emerg Med 2003;21:476–8.
assessment with sonography for trauma (FAST) for
21. O’Brien KM, Stolz LA, Amini R, et al. Focused
laparotomy in unstable polytrauma Egyptian patients.
assessment with sonography for trauma examination:
Chin J of Traumatol 2017;20:323–28.
re-examining the importance of the left upper
8. Sheng AY, Dalziel P, Liteplo AS, Fagenholz P, Noble quadrant view. J Ultrasound Med 2015;34:1429–34.
VE. Focused assessment with sonography in trauma
22. Boniface KS, Calabrese KY. Intensive care ultrasound:
and abdominal computed tomography utilization in
IV. Abdominal Ultrasound in Critical Care. Ann Am
adult trauma patients: trends over the last decade.
Thorac Soc 2013;10:713–24.
Emerg Med Int 2013;2013:678380.
23. Lobo V, Hunter-Behrend M, Cullnan E, et al. Caudal
9. Montoya J, Stawicki SP, Evans DC, et al. From FAST to
edge of the liver in the right upper quadrant (RUQ)
E-FAST: an overview of the evolution of ultrasound-
view is the most sensitive area for free fluid on the
based traumatic injury assessment. Eur J Trauma Surg
FAST exam. West J Emerg Med 2017;18:270–80.
2016;42:119–26.
98
24. Abrams BJ, Sukumvanich P, Seibel R, Moscati R, Jehle responsiveness in spontaneously breathing patients?
D. Ultrasound for the detection of intraperitoneal Crit Care 2015;19:400.
fluid: the role of Trendelenberg positioning. Am J 32. Tayal VS, Nielsen A, Jones AE, et al. Accuracy of
Emerg Med 1999;17:117–20. trauma ultrasound in major pelvic injury. J Trauma
25. Ma OJ, Kefer MP, Mateer JR, Thoma B. Evaluation of 2006;61:1453–57.
hemoperitoneum using a single-view vs multiple-view 33. Gaarder C, Kroepelien CF, Loekke R, et al.
ultrasonographic examination. Acad Emerg Med Ultrasound performed by radiologists-confirming
1995;2:581–6. the truth about FAST in trauma. J Trauma
26. Rozycki G, Ochsner M, Feliciano D, et al. Early 2009;67:323–27.
detection of hemoperitoneum by ultrasound 34. Chenkin J. Diagnosis of aortic dissection presenting as
examination of the right upper quadrant: a multicenter ST-elevation myocardial infarction using point-of-care
study. J Trauma 1998;45:878–83. ultrasound. J Emerg Med 2017;53:880–4.
27. Blackbourne LH, Soffer D, McKenney M, et al. 35. Hartshorne TC, McCollum CN, Earnshaw JJ, Morris J,
Secondary ultrasound examination increases the Nasim A. Ultrasound measurement of aortic diameter
sensitivity of the FAST exam in blunt trauma. J in a national screening programme. Eur J Vasc
Trauma 2004;57:934–8. Endovasc Surg 2011;42:195–9.
28. Blanco P, Volpicelli G. Common pitfalls in point-of- 36. Thapar A, Cheal D, Hopkins T, et al. Internal or
care ultrasound: a practical guide for emergency and external wall diameter for abdominal aortic aneurysm
critical care physicians. Crit Ultrasound J 2016;8:15. screening? Ann R Coll Surg Engl 2010;92(6):503–5.
29. Sosland RP, Gupta K. McConnell’s sign. Circulation 37. Ellis M, Powell JT, Greenhalgh RM. Limitations
2008;118:e517–e518. of ultrasonography in surveillance of small
30. McConnell MV, Solomon SD, Rayan ME, et al. abdominal aortic aneurysms. Br J Surg 1991;78
Regional right ventricular dysfunction detected by (5):614–6.
echocardiography in acute pulmonary embolism. Am J 38. Pare JR, Liu R, Moore CL, et al. Emergency physician
Cardiol 1996;78:469–73. focused cardiac ultrasound improves diagnosis of
31. Airapetian N, Maizel J, Alyamani O, et al. Does inferior ascending aortic dissection. Am J Emerg Med
vena cava respiratory variability predict fluid 2016;34:486–92.
Chapter
Noninvasive Measurement of Cardiac
10 Output
Samuel Gilliland, Robert H. Thiele, and Karsten Bartels
Introduction meters, as well as measurement of CO obtained via the

Fick method. Cardiac output monitoring modalities
Adolf Eugen Fick (1829–1901) described the relation-
described here are assessed comparatively to these
ship between cardiac output (CO), oxygen uptake
standards.
(VO2), and the difference between arterial and venous
blood oxygen content (CaO2–CvO2) as follows:1
Thermodilution
CO = VO2/(CaO2–CvO2 ) Injection of an indicator with subsequent measure-
ment on a time-concentration curve form the basis for
The “Fick method” involves measurement of mixed
Stewart’s “indicator-dilution” method of determining
venous oxygen content and arterial oxygen content,
cardiac output.9 Flow is calculated with a derivation
with oxygen uptake measured using a Douglas bag.
of the Conservation of Indicator Principle, the
The equation is then solved for cardiac output. While
“Stewart−Hamilton equation.”
the Fick technique is too cumbersome to be used ð∞
clinically, it has allowed researchers to develop and
m ¼ Q cðtÞdt
validate other more practical means of measuring 0
cardiac output.
The principle upon which the Fick technique is In this equation, m is the mass of the indicator
based (conservation of mass and energy) is also used injected, Q is flow, and c(t) the change in concentra-
in applications where an indicator (e.g., cold fluid) is tion over time.10 Using temperature as the indicator,
injected upstream and then measured downstream, as Fegler adapted this technique.4 Additional factors
in thermodilution. The advent of the Swan−Ganz must be considered to apply this to the Swan thermo-
catheter in the 1970s, combining thermodilution dilution measurements, and thus CO is calculated as
with the balloon-tipped pulmonary artery catheter ð∞
(PAC), allowed for the measurement and optimiza- CO ¼ VI ðTB TI Þ K1 K2 = ΔTB ðtÞdt
0
tion of cardiac output to become a clinical reality.2–4
Unfortunately, multiple large prospective rando- In this equation VI is injectate volume, TB blood
mized-controlled trials failed to show improvement temperature, TI injectate temperature, K1 density fac-
of mortality in critically ill patients with its use.5–7 tor defined as specific heat multiplied by specific grav-
These failures were attributed, in part, to the counter- ity of the injectate divided by the product of the
balancing adverse effects of such an invasive monitor, specific heat and gravity of blood, and K2 is
frequent misinterpretation of the data, and perhaps a computation constant taking into account the dead
ineffective or harmful treatment based on the data. In space of the catheter, head exchange in transit, and
response, cardiac output monitoring and “optimiza- injection rate. This is then divided byÐ the change in
∞
tion” were not abandoned, but instead less invasive blood temperature over time ( 0 ΔTB ðtÞdt).11
monitoring strategies have been pursued.8 Assumptions made by these calculations are
As the PAC was the clinical reference standard at important to note, as they may affect interpretation
the time of these monitors’ initial development, many of data or result in inaccurate measurements in
have erroneously interpreted it as the experimental particular clinical scenarios. These assumptions
reference standard as well. Additional reference stan- are as follows: (1) that the injectate temperature
dards include electromagnetic and transit time flow will return to baseline in a finite period of time (as
100
Noninvasive Measurement of Cardiac Output
opposed to infinity), (2) that the injectate and the and may be outweighed by the noninvasive nature of
bloodstream are perfectly mixed, and (3) that the these devices.
measurement of temperature difference is accurate. An advantage of Doppler-based devices such as
Clinical scenarios that may invalidate these the CardioQ (Deltex Medical, Chichester, West
assumptions include tricuspid regurgitation,12 fre- Sussex, UK) is ease of use without the requirement
quent repeated measurements,13 low flow states,14 for advanced ultrasound skills. The esophageal
and rapid temperature changes following cardio- Doppler probe resembles a gastric tube in size and
pulmonary bypass.15,16 Additionally, as PAC ther- shape, with a small dedicated probe at the end. It may
modilution relies on measurements taken at the be inserted orally or nasally and is positioned to image
right atrium and pulmonary artery, this measure- the descending thoracic aorta. Because of this posi-
ment is of right-sided cardiac output and ignores tioning, it fails to capture the 30% of the cardiac out-
intracardiac shunting. put that is diverted proximally and so also fails to
The PAC cardiac output, as validated in a large directly measure CO. However, assessment of CO by
number and variety of experimental models, is esophageal Doppler as compared to Fick methods
accurate and precise, with an error of only 13% reveals only a slightly lower correlation than that of
for triplicate readings.17 This, along with its “first the PAC.23–25 Of note, the algorithm developed to
to market” position,2 led to its adoption as the calculate stroke volume from esophageal Doppler
clinical reference standard. Continuous thermodi- devices is derived from a Doppler/thermodilution
lution catheters, which use a heating element as database and thus cannot be more accurate than ther-
opposed to injectate, seem to offer similar clinical modilution itself.26
accuracy, though they do lag behind real-time mea- Additional information provided by esophageal
surements by up to 5 minutes.18–20 The additional Doppler may include stroke volume variation and
benefit of the catheter is the ability to measure corrected flow time (FTc), the latter being time spent
pulmonary artery pressures (PAP) and pulmonary in systole corrected for heart rate. Both have been
artery occlusion pressures (PAOP), though the used as a measure of fluid responsiveness and incor-
ability of clinicians to reliably interpret this infor- porated into goal-directed fluid management
mation has been questioned.21 strategies.
Ultrasound-Based Techniques Arterial Waveform Devices

By relating the frequency of a returning ultrasound Arterial waveform analysis is available commercially
beam (Δf) with the frequency of the initial beam (f0), in two classes of devices, generally split into those
the speed of sound in tissue (c), and the angle of requiring calibration and those that do not (i.e.,
incidence to a moving reflector (θ), the Doppler equa- uncalibrated). The underlying principle of both
tion can calculate the velocity of that reflector (v), or classes is the same and based on Windkessel’s
in the case of ultrasound-based techniques of CO model of blood flow.27 This assumes the volume of
monitoring, blood velocity. To calculate CO, that blood entering a vessel of infinite length must equal
velocity is integrated over time (velocity time integral, the volume of blood leaving a vessel over the period
or VTI) and then multiplied by a cross-sectional area. of cardiac contraction. It also assumes that the vessel
This assumes that the measured velocity is equal at all will expand during systole and contract during dia-
points in the vessel being studied, and that cross- stole. In this model, the arterioles serve as resistors,
sectional change of the vessel is a negligible source while the aorta is a capacitor.28 Impedance, or resis-
of error.22 tance to pulsatile flow, is calculated using a variety of
techniques, including those based on physical mod-
v = cΔf /(2f0cos(θ))
els (e.g., the characteristic impedance [cZ] techni-
Validation of Doppler-based devices has been per- que) and models that are empirically derived (e.g.,
formed both in ex vivo models and animal and FloTrac/Vigileo).29 All of these estimates are imper-
human experimental reference standards, similar to fect, therefore devices utilizing calibration might
the PAC. The accuracy of Doppler compared to ther- have some advantage.
modilution-derived cardiac output is slightly reduced, Pulse contour devices, similar to esophageal
though this difference may not be clinically relevant Doppler devices, additionally calculate stroke
volume (SV) variation and pulse pressure variation. Uncalibrated Arterial Waveform Devices
In some devices, ventricular contractility assess-
ment (rate of left ventricular pressure rise early in Empiric Approach (FloTrac)
systole, or dP/dtmax), extravascular lung water, and
The FloTrac’s (Edwards Lifesciences, Irvine, CA) use
other derived variables may also be determined.
of an empirically derived mathematical model, as
Caution must be taken, however, in those patients
opposed to a physical model, is a significant departure
with a pathology that alters the arterial waveform
in arterial waveform analysis.46 As the device is unca-
(e.g., aortic insufficiency).
librated, accuracy is reduced.47–55
Calibrated Arterial Waveform Devices Pressure Recording Analytical Method

Transpulmonary Thermodilution (PiCCO) The Pressure Recording Analytical Method (PRAM)
(Hewlett Packard, Andover, MA, USA) operates on
By analyzing the arterial waveform distal to the dicrotic
the principle that volumetric changes in blood vessels
notch, the PiCCO system (Maquet, Rastatt, Germany)
are based on the interplay between arterial compli-
treats compliance as a dynamic variable dependent on
ance, left ventricular ejection force, impedance, and
pressure (C(p)) and incorporates instantaneous pres-
peripheral resistance to flow secondary to wave reflec-
sure changes (dP/dt) into its stroke volume estimate.
tion from vessel bifurcations as well as changing
This is meant to better account for the ventricular
diameter.56 It also relies on the assumption that
output fraction that is stored in capacitance vessels.
these changes occur in the radial direction. Though
Furthermore, to account for changes in impedance
the PRAM method compares favorably against
that occur over time, the PiCCO utilizes intermittent
Doppler-based devices in animal models,57 experi-
cardiac output calibration with a proprietary tempera-
mental reference standards,56 and conventional
ture-sensing central venous catheter used to perform
thermodilution,58,59 further validation is needed as
transpulmonary thermodilution:
the majority of these data come from a single investi-
SV = k end-diastole∫end-systole [P(t)/SVR + C(p) × dP/dt]dt gatory group.
where SVR = systemic vascular resistance. Calibrated versus Uncalibrated Arterial

Transpulmonary thermodilution has been shown to Waveform Devices
be non-inferior to conventional thermodilution both In a study comparing LiDCO, PiCCO, and
by experimental reference standards and in multiple FloTRAC monitors to intermittent thermodilution,
animal models.30–38 the calibrated devices (LiDCO, PiCCO) had nar-
rower limits of agreement as compared to the
Lithium Dilution (LiDCO Plus) uncalibrated device (FloTRAC).60 These results
The PulseCO algorithm is used by LiDCO (LiDCO have been replicated in later studies. Though the
Ltd, Lake Villa, IL, USA) monitors to estimate aortic benefits of calibration have not been demonstrated
blood volume by using a transfer function to deter- universally,61 the majority of data favor calibrated
mine characteristic impedance.39 The LiDCO Plus can devices over uncalibrated devices for accuracy and
be calibrated using a lithium dilution curve (LiDCO); precision.62–64 Additionally, the accuracy of
alternatively it may be used as an uncalibrated device. PulseCO, PiCCO, and FloTrac measurements
Cardiac output measurements using lithium appears to decrease during periods of hemody-
injected centrally then sampled via a femoral namic instability, making them potentially less use-
catheter are more accurate than conventional thermo- ful in acute situations.
dilution according to data from experimental studies
conducted in animals; if the lithium is injected per- Photoplethysmography
ipherally, however, accuracy is reduced.40–42
The photoplethysmography (PPG) waveform
Although comparison studies are limited, the avail-
resembles that of an invasively derived waveform
able evidence indicates that this calibrated device is
in a number of ways. They have both systolic-
reasonably accurate compared to conventional
forward and diastolic-backward components,65,66
thermodilution.43–45
102
and they are affected by both changes in vascular Bioimpedance and Bioreactance
tone and aortic elasticity.65,67–69 Thus, the plethys-
The relationship between intrathoracic blood volume
mographic waveform may be similar to an arterial
and thoracic electrical resistance is the basis for thor-
line waveform used to noninvasively characterize
acic electrical bioimpedance (TEB). The maximum
cardiac output.
impedance change over time is assumed to be related
Changes in PPG baseline have been found to be
to peak aortic flow rate. Using EKG tracings to deter-
correlated with arterial systolic pulse pressure varia-
mine ventricular ejection time, the peak aortic flow
tion in multiple human studies and may have a similar
rate is averaged over this period to approximate mean
role in determining fluid responsiveness, especially in
aortic flow rate. It is also assumed that any change in
mechanically ventilated patients.70,71 Similar results
thoracic electrical resistance is solely attributed to
have been demonstrated in patients undergoing gen-
a change in intrathoracic blood volume. Given these
eral anesthesia.72
assumptions, CO measurement with TEB may be sig-
By more closely examining the PPG waveform
nificantly altered by pulmonary edema, electrode
itself with devices such as a reflection mode infra-
positioning, and electrical noise. Comparisons to
red finger probe (ADInstruments, Sydney,
other cardiac output monitors, as well as animal and
Australia), additional information may be gleaned.
human experimental reference standards, suggest that
Left ventricular ejection time (LVET) is the time
it is less accurate than thermodilution and Doppler-
between the start of systolic upstroke and the
based devices. This lack of accuracy is exacerbated by
dicrotic notch.73 LVET measured by Doppler flow
the potential for electrical interference in the operat-
of the aorta correlates with that measured by PPG
ing room, specifically electrocautery.
and decreases proportionally to hypovolemia.74 By
Electrical velocimetry attempts to improve the
examining the pre-ejection period (i.e., the interval
accuracy of these older devices and operates under
between the ECG R wave and the radial artery
the assumption that aortic blood flow velocity, not
pulse pressure upstroke), a clinician may be able
thoracic blood volume, is related to thoracic conduc-
to determine fluid responsiveness.75 A prolonged
tivity. Comparison data against thermodilution and
pre-ejection period positively correlates with
Doppler-based estimates of CO are limited.
improvement in cardiac index after fluid
Bioreactance focuses on the electrical capacitance
challenge.76
and induction properties of intrathoracic blood in
Left ventricular outflow impedance is affected by
order to limit error related to electrode positioning,
vascular tone and pulse wave velocity.65 With vaso-
humidity, body size, and temperature. These proper-
constriction, the faster pulse wave returns to the left
ties relate stroke volume to phase shift between
ventricle earlier, which then increases outflow
applied and received voltage.89 The NICOM
impedance.77 This is reflected qualitatively by the
(Cheetah Medical, Vancouver, WA) measures
PPG contour,78 and quantitatively by the PPG ampli-
CO using this technique and has been compared
tude, reflective index (RI), and PPG-derived stiffness
to both thermodilution47,89,90 and Doppler-based
index. A decrease in systemic vascular resistance not
methods.91 Few studies have been conducted using
only increases the amplitude of the PPG waveform
this new device, precluding a meaningful assessment –
but also shortens the width.78–84 The dicrotic notch
more investigation is needed.
moves leftward with vasoconstriction.85,86 The RI is
determined by the ratio of the amplitudes of the back-
ward and forward waves, which is indirectly related to Conclusions
vascular impedance.87 A stiffness index depends on Several methods of cardiac output determination have
the presence of the dicrotic notch (not always readily been devised to obtain similar data to reference stan-
apparent in the waveform) and is related to the elas- dards such as the Fick method. Their relative features
ticity of the vascular system.87,88 are outlined in Table 10.1. Doppler-based techniques
As noted in other methodologies discussed here, have similar accuracy as PACs with the advantage of
these measurable variables may be incorporated into being less invasive. Calibrated pulse contour devices
algorithms extrapolating cardiac output and may be are more accurate than their uncalibrated counter-
useful in the future as a real-time, noninvasive CO parts but require intermittent operator intervention
monitor. to retain that accuracy. Photoplethysmography is
Table 10.1 Cardiac output monitoring modalities characterized qualitatively.
Characteristic Thermodilution Doppler Calibrated Uncalibrated Photoplethys Bioimpedance Bioreactance Partial
Pulse Pulse mography Rebreathing
Contour Contour
Accuracy +++ ++ ++ + + + + +
Response Time ++ +++ +++ +++ +++ +++ +++ +
Convenience + ++ ++ +++ +++ +++ +++ ++
Note: + = below average, ++ = average, +++ = better than average.
Source: Thiele RH, Bartels K, Gan TJ. Cardiac output monitoring: a contemporary assessment and review. Crit Care Med 2015;43(1):177–85.
a developing and potentially promising modality but thermodilution in patients with tricuspid
requires further refinement and investigation before regurgitation. Am J Med 1989;86(4):417–20.
being ready for widespread clinical use. Bioimpedance 13. Maruschak GF, Potter AM, Schauble JF, Rogers MC.
devices do not correlate as well as either thermodilu- Overestimation of pediatric cardiac output by thermal
tion or Doppler, and bioreactance devices require indicator loss. Circulation 1982;65(2):380–3.
further study in order to make meaningful 14. van Grondelle A, Ditchey RV, Groves BM,
comparisons. Wagner WW Jr, Reeves JT. Thermodilution method
overestimates low cardiac output in humans. Am
J Physiol 1983;245(4):H690–2.
References 15. Bazaral MG, Petre J, Novoa R. Errors in
1. Geerts BF, Aarts LP, Jansen JR. Methods in thermodilution cardiac output measurements caused
pharmacology: measurement of cardiac output. Br by rapid pulmonary artery temperature decreases after
J Clin Pharmacol 2011;71(3):316–30. cardiopulmonary bypass. Anesthesiology 1992;77
2. Ganz W, Donoso R, Marcus HS, Forrester JS, Swan HJ. (1):31–7.
A new technique for measurement of cardiac output by 16. Latson TW, Whitten CW, O’Flaherty D. Ventilation,
thermodilution in man. Am J Cardiol 1971;27 thermal noise, and errors in cardiac output
(4):392–6. measurements after cardiopulmonary bypass.
3. Swan HJ, Ganz W, Forrester J, et al. Catheterization of Anesthesiology 1993;79(6):1233–43.
the heart in man with use of a flow-directed 17. Yang XX, Critchley LA, Joynt GM. Determination of
balloon-tipped catheter. N Engl J Med 1970;283 the precision error of the pulmonary artery
(9):447–51. thermodilution catheter using an in vitro continuous
4. Fegler G. Measurement of cardiac output in flow test rig. Anesth Analg 2011;112(1):70–7.
anaesthetized animals by a thermodilution 18. Philip JH, Long MC, Quinn MD, Newbower RS.
method. Q J Exp Physiol Cogn Med Sci 1954;39 Continuous thermal measurement of cardiac output.
(3):153–64. IEEE Trans Biomed Eng 1984;31(5):393–400.
5. Sandham JD, Hull RD, Brant RF, et al. A randomized, 19. Aranda M, Mihm FG, Garrett S, Mihm MN, Pearl RG.
controlled trial of the use of pulmonary-artery Continuous cardiac output catheters: delay in in vitro
catheters in high-risk surgical patients. N Engl J Med response time after controlled flow changes.
2003;348(1):5–14. Anesthesiology 1998;89(6):1592–5.
6. Richard C, Warszawski J, Anguel N, et al. Early use of 20. Goldstein LJ. Response time of the Opti-Q continuous
the pulmonary artery catheter and outcomes in cardiac output pulmonary artery catheter in the urgent
patients with shock and acute respiratory distress mode to a step change in cardiac output. J Clin Monit
syndrome: a randomized controlled trial. JAMA Comput 1999;15(7–8):435–9.
2003;290(20):2713–20.
21. Iberti TJ, Fischer EP, Leibowitz AB, et al. A multicenter
7. Harvey S, Harrison DA, Singer M, et al. Assessment of study of physicians’ knowledge of the pulmonary
the clinical effectiveness of pulmonary artery catheters artery catheter. Pulmonary Artery Catheter Study
in management of patients in intensive care Group. JAMA 1990;264(22):2928–32.
(PAC-Man): a randomised controlled trial. Lancet
2005;366(9484):472–7. 22. Schuster AH, Nanda NC. Doppler echocardiographic
measurement of cardiac output: comparison with a
8. Porhomayon J, El-Solh A, Papadakos P, Nader ND. non-golden standard. Am J Cardiol. 1984;53(1):257–9.
Cardiac output monitoring devices: an analytic review.
Intern Emerg Med 2012;7(2):163–71. 23. Smith HJ, Grottum P, Simonsen S. Doppler flowmetry
in the lower thoracic aorta. An indirect estimation of
9. Stewart G. The output of the heart in dogs. Am cardiac output. Acta Radiol Diagn (Stockh) 1985;26
J Physiol 1921;57:27–50. (3):257–63.
10. Profant M VK, Eckhardt U. The Stewart-Hamilton 24. Davies JN, Allen DR, Chant AD. Non-invasive
equations and the indicator dilution method. SIAM Doppler-derived cardiac output: a validation study
J Appl Math 1978;34:666–75. comparing this technique with thermodilution and
11. Nishikawa T, Dohi S. Errors in the measurement of Fick methods. Eur J Vasc Surg. 1991;5(5):497–500.
cardiac output by thermodilution. Can J Anaesth 25. Valtier B, Cholley BP, Belot JP, et al. Noninvasive
1993;40(2):142–53. monitoring of cardiac output in critically ill patients
12. Cigarroa RG, Lange RA, Williams RH, Bedotto JB, using transesophageal Doppler. Am J Respir Crit Care
Hillis LD. Underestimation of cardiac output by Med 1998;158(1):77–83.
26. Lowe GC, BM; Philpot, EJ, et al. Oesophageal Doppler invasive direct coronary artery bypass grafting.
Monitor (ODM) guided individualised goal directed J Cardiothorac Vasc Anesth 1999;13(4):437–40.
fluid management (iGDFM) in surgery – a technical 38. Gust R, Gottschalk A, Bauer H, et al. Cardiac output
review. Deltex Medical Technical Review; 2010. measurement by transpulmonary versus conventional
27. Kouchoukos NT, Sheppard LC, McDonald DA. thermodilution technique in intensive care patients
Estimation of stroke volume in the dog by a pulse after coronary artery bypass grafting. J Cardiothorac
contour method. Circ Res 1970;26(5):611–23. Vasc Anesth 1998;12(5):519–22.
28. Essler S, Schroeder MJ, Phaniraj V, et al. Fast 39. Linton NW, Linton RA. Estimation of changes in
estimation of arterial vascular parameters for transient cardiac output from the arterial blood pressure
and steady beats with application to hemodynamic waveform in the upper limb. Br J Anaesth 2001;86
state under variant gravitational conditions. Ann (4):486–96.
Biomed Eng 1999;27(4):486–97. 40. Linton RA, Young LE, Marlin DJ, et al. Cardiac output
29. Wesseling K, de Witt B, Weber A. A simple device for measured by lithium dilution, thermodilution, and
the continuous measurement of cardiac output. Adv transesophageal Doppler echocardiography in
Cardiovasc Phys 1983;5:16–52. anesthetized horses. Am J Vet Res 2000;61(7):731–7.
30. Bajorat J, Hofmockel R, Vagts DA, et al. Comparison 41. Kurita T, Morita K, Kato S, et al. Lithium dilution
of invasive and less-invasive techniques of cardiac cardiac output measurements using a peripheral
output measurement under different haemodynamic injection site comparison with central injection
conditions in a pig model. Eur J Anaesthesiol 2006;23 technique and thermodilution. J Clin Monit Comput
(1):23–30. 1999;15(5):279–85.
31. Marx G, Schuerholz T, Sumpelmann R, Simon T, 42. Kurita T, Morita K, Kato S, et al. Comparison of the
Leuwer M. Comparison of cardiac output accuracy of the lithium dilution technique with the
measurements by arterial trans-cardiopulmonary and thermodilution technique for measurement of cardiac
pulmonary arterial thermodilution with direct Fick in output. Br J Anaesth 1997;79(6):770–5.
septic shock. Eur J Anaesthesiol 2005;22(2):129–34. 43. Linton RA, Band DM, Haire KM. A new method of
32. Pauli C, Fakler U, Genz T, et al. Cardiac output measuring cardiac output in man using lithium
determination in children: equivalence of the dilution. Br J Anaesth 1993;71(2):262–6.
transpulmonary thermodilution method to the direct 44. Costa MG, Della Rocca G, Chiarandini P, et al.
Fick principle. Intensive Care Med 2002;28(7):947–52. Continuous and intermittent cardiac output
33. Goedje O, Hoeke K, Lichtwarck-Aschoff M, et al. measurement in hyperdynamic conditions: pulmonary
Continuous cardiac output by femoral arterial artery catheter vs. lithium dilution technique. Intensive
thermodilution calibrated pulse contour analysis: Care Med 2008;34(2):257–63.
comparison with pulmonary arterial thermodilution. 45. Mora B, Ince I, Birkenberg B, et al. Validation of
Crit Care Med 1999;27(11):2407–12. cardiac output measurement with the LiDCO pulse
34. Sakka SG, Kozieras J, Thuemer O, van Hout N. contour system in patients with impaired left
Measurement of cardiac output: a comparison between ventricular function after cardiac surgery. Anaesthesia
transpulmonary thermodilution and uncalibrated 2011;66(8):675–81.
pulse contour analysis. Br J Anaesth 2007;99 46. Pratt B, Roteliuk L, Hatib F, Frazier J, Wallen RD.
(3):337–42. Calculating arterial pressure-based cardiac output
35. Della Rocca G, Costa MG, Pompei L, Coccia C, using a novel measurement and analysis method.
Pietropaoli P. Continuous and intermittent cardiac Biomed Instrum Technol 2007;41(5):403–11.
output measurement: pulmonary artery catheter 47. Marque S, Cariou A, Chiche JD, Squara P. Comparison
versus aortic transpulmonary technique. Br J Anaesth between Flotrac-Vigileo and Bioreactance, a totally
2002;88(3):350–6. noninvasive method for cardiac output monitoring.
36. Hamzaoui O, Monnet X, Richard C, et al. Effects of Crit Care 2009;13(3):R73.
changes in vascular tone on the agreement between 48. Opdam HI, Wan L, Bellomo R. A pilot assessment of
pulse contour and transpulmonary thermodilution the FloTrac cardiac output monitoring system.
cardiac output measurements within an up to 6-hour Intensive Care Med 2007;33(2):344–9.
calibration-free period. Crit Care Med 2008;36
(2):434–40. 49. Cannesson M, Attof Y, Rosamel P, et al. Comparison of
FloTrac cardiac output monitoring system in patients
37. Buhre W, Weyland A, Kazmaier S, et al. Comparison of undergoing coronary artery bypass grafting with
cardiac output assessed by pulse-contour analysis and pulmonary artery cardiac output measurements. Eur
thermodilution in patients undergoing minimally J Anaesthesiol 2007;24(10):832–9.
106
50. Breukers RM, Sepehrkhouy S, Spiegelenberg SR, 62. Johansson A, Chew M. Reliability of continuous pulse
Groeneveld AB. Cardiac output measured by a new contour cardiac output measurement during
arterial pressure waveform analysis method without hemodynamic instability. J Clin Monit Comput 2007;21
calibration compared with thermodilution after (4):237–42.
cardiac surgery. J Cardiothorac Vasc Anesth 2007;21 63. Bein B, Meybohm P, Cavus E, et al. The reliability of
(5):632–5. pulse contour-derived cardiac output during
51. Biais M, Nouette-Gaulain K, Cottenceau V, et al. hemorrhage and after vasopressor administration.
Cardiac output measurement in patients undergoing Anesth Analg 2007;105(1):107–13.
liver transplantation: pulmonary artery catheter versus 64. Zollner C, Haller M, Weis M, et al. Beat-to-beat
uncalibrated arterial pressure waveform analysis. measurement of cardiac output by intravascular pulse
Anesth Analg 2008;106(5):1480–6. contour analysis: a prospective criterion standard
52. Biancofiore G, Critchley LA, Lee A, et al. Evaluation of study in patients after cardiac surgery. J Cardiothorac
a new software version of the FloTrac/Vigileo (version Vasc Anesth 2000;14(2):125–9.
3.02) and a comparison with previous data in cirrhotic 65. O’Rourke MF, Yaginuma T, Avolio AP. Physiological
patients undergoing liver transplant surgery. Anesth and pathophysiological implications of ventricular/
Analg 2011;113(3):515–22. vascular coupling. Ann Biomed Eng 1984;12(2):119–34.
53. Saraceni E, Rossi S, Persona P, et al. Comparison of two 66. O’Rourke MF. Pressure and flow waves in systemic
methods for cardiac output measurement in critically arteries and the anatomical design of the arterial
ill patients. Br J Anaesth 2011;106(5):690–4. system. J Appl Physiol 1967;23(2):139–49.
54. Eleftheriadis S, Galatoudis Z, Didilis V, et al. 67. Wisely NA, Cook LB. Arterial flow waveforms from
Variations in arterial blood pressure are associated pulse oximetry compared with measured Doppler flow
with parallel changes in FlowTrac/Vigileo-derived waveforms apparatus. Anaesthesia 2001;56(6):556–61.
cardiac output measurements: a prospective
comparison study. Crit Care 2009;13(6):R179. 68. Millasseau SC, Guigui FG, Kelly RP, et al. Noninvasive
assessment of the digital volume pulse: comparison
55. Sander M, Spies CD, Grubitzsch H, et al. Comparison with the peripheral pressure pulse. Hypertension
of uncalibrated arterial waveform analysis in cardiac 2000;36(6):952–6.
surgery patients with thermodilution cardiac output
measurements. Crit Care 2006;10(6):R164. 69. Almond NE, Jones DP, Cooke ED. Noninvasive
measurement of the human peripheral circulation:
56. Romano SM, Pistolesi M. Assessment of cardiac output relationship between laser Doppler flowmeter and
from systemic arterial pressure in humans. Crit Care photoplethysmograph signals from the finger.
Med 2002;30(8):1834–41. Angiology 1988;39(9):819–29.
57. Romagnoli S, Romano SM, Bevilacqua S, et al. Cardiac 70. Natalini G, Rosano A, Taranto M, et al. Arterial versus
output by arterial pulse contour: reliability under plethysmographic dynamic indices to test
hemodynamic derangements. Interact Cardiovasc responsiveness for testing fluid administration in
Thorac Surg 2009;8(6):642–6. hypotensive patients: a clinical trial. Anesth Analg
58. Franchi F, Silvestri R, Cubattoli L, et al. Comparison 2006;103(6):1478–84.
between an uncalibrated pulse contour method and 71. Cannesson M, Besnard C, Durand PG, Bohe J,
thermodilution technique for cardiac output Jacques D. Relation between respiratory variations in
estimation in septic patients. Br J Anaesth 2011;107 pulse oximetry plethysmographic waveform amplitude
(2):202–8. and arterial pulse pressure in ventilated patients. Crit
59. Scolletta S, Franchi F, Taccone FS, et al. An Care 2005;9(5):R562–8.
uncalibrated pulse contour method to measure cardiac 72. Cannesson M, Attof Y, Rosamel P, et al. Respiratory
output during aortic counterpulsation. Anesth Analg variations in pulse oximetry plethysmographic
2011;113(6):1389–95. waveform amplitude to predict fluid responsiveness in
60. Hadian M, Kim HK, Severyn DA, Pinsky MR. Cross- the operating room. Anesthesiology 2007;106
comparison of cardiac output trending accuracy of (6):1105–11.
LiDCO, PiCCO, FloTrac and pulmonary artery 73. Chan GS, Middleton PM, Celler BG, Wang L,
catheters. Crit Care 2010;14(6):R212. Lovell NH. Automatic detection of left ventricular
61. Krejci V, Vannucci A, Abbas A, Chapman W, ejection time from a finger photoplethysmographic
Kangrga IM. Comparison of calibrated and pulse oximetry waveform: comparison with Doppler
uncalibrated arterial pressure-based cardiac output aortic measurement. Physiol Meas 2007;28(4):439–52.
monitors during orthotopic liver transplantation. Liver 74. Geeraerts T, Albaladejo P, Declere AD, et al. Decrease
Transpl 2010;16(6):773–82. in left ventricular ejection time on digital arterial
waveform during simulated hypovolemia in normal waveform and systemic vascular resistance. J Clin
humans. J Trauma 2004;56(4):845–9. Monit Comput 2007;21(6):365–72.
75. Bendjelid K, Suter PM, Romand JA. The respiratory 84. Lee QY, Chan GS, Redmond SJ, et al. Multivariate
change in preejection period: a new method to predict classification of systemic vascular resistance using
fluid responsiveness. J Appl Physiol (1985) 2004;96 photoplethysmography. Physiol Meas 2011;32
(1):337–42. (8):1117–32.
76. Feissel M, Badie J, Merlani PG, Faller JP, Bendjelid K. 85. Lax H, Feinberg AW, Cohen BM. Studies of the arterial
Pre-ejection period variations predict the fluid pulse wave: I. The normal pulse wave and its
responsiveness of septic ventilated patients. Crit Care modification in the presence of human arteriosclerosis.
Med 2005;33(11):2534–9. J Chronic Dis 1956;3(6):618–31.
77. Chowienczyk PJ, Kelly RP, MacCallum H, et al. 86. Dawber TR, Thomas HE, Jr., McNamara PM.
Photoplethysmographic assessment of pulse wave Characteristics of the dicrotic notch of the arterial
reflection: blunted response to pulse wave in coronary heart disease. Angiology
endothelium-dependent beta2-adrenergic vasodilation 1973;24(4):244–55.
in type II diabetes mellitus. J Am Coll Cardiol 1999;34 87. Millasseau SC, Kelly RP, Ritter JM, Chowienczyk PJ.
(7):2007–14. Determination of age-related increases in large artery
78. Murray WB, Foster PA. The peripheral pulse wave: stiffness by digital pulse contour analysis. Clin Sci
information overlooked. J Clin Monit 1996;12 (Lond) 2002;103(4):371–7.
(5):365–77. 88. Arnett DK, Evans GW, Riley WA. Arterial stiffness:
79. Lund F. Digital pulse plethysmography (DPG) in a new cardiovascular risk factor? Am J Epidemiol
studies of the hemodynamic response to nitrates–a 1994;140(8):669–82.
survey of recording methods and principles of analysis. 89. Keren H, Burkhoff D, Squara P. Evaluation of
Acta Pharmacol Toxicol (Copenh) 1986;59 Suppl a noninvasive continuous cardiac output monitoring
6:79–96. system based on thoracic bioreactance. Am J Physiol
80. Ezri T, Steinmetz A, Geva D, Szmuk P. Skin vasomotor Heart Circ Physiol 2007;293(1):H583-9.
reflex as a measure of depth of anesthesia. 90. Raval NY, Squara P, Cleman M, et al. Multicenter
Anesthesiology 1998;89(5):1281–2. evaluation of noninvasive cardiac output measurement
81. Zhang XY, Zhang YT. The effect of local mild cold by bioreactance technique. J Clin Monit Comput
exposure on pulse transit time. Physiol Meas 2006;27 2008;22(2):113–9.
(7):649–60. 91. Weisz DE, Jain A, McNamara PJ, A EL-K. Non-
82. Awad AA, Ghobashy MA, Ouda W, et al. Different invasive cardiac output monitoring in neonates using
responses of ear and finger pulse oximeter wave bioreactance: a comparison with echocardiography.
form to cold pressor test. Anesth Analg 2001;92 Neonatology 2012;102(1):61–7.
(6):1483–6. 92. Thiele RH, Bartels K, Gan TJ. Cardiac output
83. Awad AA, Haddadin AS, Tantawy H, et al. The monitoring: a contemporary assessment and review.
relationship between the photoplethysmographic Crit Care Med 2015;43(1):177–85.
108
Chapter
Assessing Intravascular Volume Status
11 and Fluid Responsiveness: A

Non-Ultrasound Approach
David S. Beebe
Introduction have a positive response to fluid administration.

Furthermore, patients with right or left ventricular
One of the clinical decisions that anesthesiologists
heart failure may be harmed by fluid administra-
routinely face is “when” and “how much” fluid to
tion that does not result in hemodynamic
give patients in the perioperative period in order to
improvement, as fluid overload has been linked
maintain or restore circulating blood volume and
to delirium, congestive heart failure, cardiac con-
pressure. Intravascular volume is commonly
duction abnormalities, non-cardiogenic pulmon-
affected in the perioperative period by multiple
ary edema, hepatic congestion, ileus, acute
factors, including dehydration due to preoperative
kidney injury, peripheral edema, wound dehis-
medical instruction (i.e., fasting), preoperative
cence, decubitus ulcers, deep vein thrombosis,
bowel preparation, possible vomiting and diarrhea,
and pulmonary embolism.
diuretic administration, third-spacing (i.e., move-
A priori identification of patients who are
ment of fluid from the intravascular space to the
likely to respond to fluid administration with an
interstitial space), and intraoperative blood loss.
increase in cardiac output and blood pressure
Further, cardiac depression and systemic vasodila-
would be the optimal approach to resuscitation.
tion due to anesthesia often cause hypotension,
Invasive methods used to guide resuscitation for
which is frequently and erroneously assumed to
more than 50 years, including central venous
be due to hypovolemia.
pressure and pulmonary capillary wedge pressure
Traditionally, anesthesiologists have relied on
measurement, have repeatedly been shown to not
standard fluid replacement strategies based upon
predict fluid responsiveness.
patient weight, length of time the patient was fasting,
Within the past 15–20 years the focus on pre-
blood loss, and type of surgery, in conjunction with
dicting fluid responsiveness has shifted to mea-
heart rate, blood pressure, urine output measurement,
suring the dynamic changes in arterial blood
and personal “expertise.” This approach may be ade-
pressure that occur during positive pressure ven-
quate in most circumstances, but given that hypovo-
tilation. Assessment of these dynamic changes is
lemia and hypervolemia are both associated with
the most sensitive and specific means of predict-
increased morbidity and mortality, more definitive
ing fluid responsiveness and recently has been
and objective assessment of intravascular volume
incorporated into bedside devices. Despite this,
would be beneficial.
determination and utilization of dynamic changes
Accurate assessment of intravascular volume is
in arterial blood pressure remain inconsistently
especially important in certain high-risk patient
applied. This chapter (1) reviews the physiological
groups that either do not respond to or may be
principles underlying the dynamic changes in
harmed by fluid administration. Patients with
arterial blood pressure with mechanical
severe sepsis and hypotension are routinely resus-
ventilation that form the basis for systolic
citated by administration of a fluid challenge (e.g.,
pressure variation, pulse pressure variation,
10–30 ml/kg of crystalloid) to increase preload,
stroke volume variation, and plethysmography
cardiac output, and blood pressure. Heart rate,
monitoring; and (2) provides an overview of
blood pressure, and urine output are measured
monitors that incorporate these principles,
as surrogate markers of the cardiac output
including their applications, contraindications,
response. Only 50% of these patients, however,
and pitfalls.
Assessing Volume Status and Fluid Responsiveness
Physiological Basis of Dynamic Fluid variables the SPV is the simplest to measure. The PPV
and SVV have slightly greater sensitivity and specificity,
Responsiveness Monitors but all three indices have an area under the curve (i.e.,
a sensitivity and specificity) approaching the 0.90 range.
Background Positive pressure ventilation creates the intrathor-
Blood pressure varies over the course of acic physiology that results in these measures’ utility
a spontaneous respiratory cycle, decreasing with in the determination of fluid responsiveness. Positive
inspiration and increasing during expiration. In the pressure ventilation squeezes blood out of the lungs
1970s Rick and Burke observed that with mechanical and into the left heart, resulting in increased left
positive pressure ventilation, blood pressure tempora- ventricular preload and SV. Simultaneously, as
rily increases with inspiration and decreases during intrathoracic pressure rises, the vena cava is com-
expiration; because this is the opposite of what occurs pressed, right atrial pressure increases, and venous
during spontaneous ventilation, the phenomenon was return to the heart decreases. After a few heartbeats
termed “blood pressure respirator paradox.”1 Rick this reduces the amount of blood returning to the
and Burke also observed that there was right ventricle and causes a decreased left ventricular
a relationship between volume status and the range preload and SV. Assuming that the compliance of the
of blood pressure change over the respiratory cycle, systemic vasculature does not change over the course
yet the facts that hypovolemic patients have greater of the respiratory cycle, the SPV and PPV may be
variation in blood pressure during the respiratory calculated from systolic pressure (SP) and pulse pres-
cycle than normovolemic patients and that fluid sure (PP) measurements over one respiratory cycle,
administration reduces this variation were not respectively, and expressed as percentages.
regarded as having clinical significance for more
than 20 years.2 SPV% = (Maximum SP−Minimum SP)/Mean SP
PPV% = (Maximum PP−Minimum PP)/Mean PP

Blood Pressure and Stroke Volume
Variation with Positive Pressure Ventilation The SVV may be determined from the area under the
tracing of the arterial pressure waveform used to
Stroke volume (SV) is the amount of blood
determine the SV over one respiratory cycle, also
pumped by each ventricle with each heartbeat.
expressed as a percentage.
Arterial blood pressure is directly proportional to
the SV. Every single blood pressure waveform may SVV% = (Maximum SV−Minimum SV)/Mean SV
be used to directly determine if a change in SV
occurred by comparing it to the waveforms that The efficacy of SPV, PPV, and SVV measures in pre-
came immediately before and after. As SV varies dicting fluid responsiveness has been well investi-
over the respiratory cycle, so does the arterial pres- gated. Every study has shown that they are highly
sure waveform. The most common elements used predictive of fluid responsiveness to such a degree
to infer the SV change from the arterial pressure that they should be incorporated into routine clinical
waveform are (1) the integral of the pressure over practice as a standard of care. A recent meta-analysis
time (sometimes simply referred to as the area by Yang and Du3 of 22 studies totaling 807 mechani-
under the arterial pressure curve), (2) the systolic cally ventilated patients, 58% of who responded to
pressure in isolation, and (3) the pulse pressure a fluid challenge, showed that PPV predicted fluid
(i.e., the difference between the systolic and the responsiveness with a pooled sensitivity of 0.88 and
diastolic blood pressure). specificity of 0.89. The median threshold PPV value to
The area under the arterial pressure curve may be predict fluid responsiveness was 12% (interquartile
used to calculate the SV and determine its change over range 10% to 13%).3 An earlier systematic review of
time, the stroke volume variation (SVV). Other indica- 29 studies that measured both PPV and SVV in a total
tors of SVV are the systolic pressure variation (SPV) of 685 patients, 56% of who responded to a fluid
and the pulse pressure variation (PPV). The SVV, SPV, challenge, showed similar threshold values for pre-
and PPV may all be used to predict fluid responsiveness dicting fluid responsiveness (PPV = 12.5 ± 1.6% and
and can be used interchangeably. Of these three SVV = 11.6 ± 1.9%).4
110
For patients without an arterial line, recent fluid administered to the patient has a 90% chance of
advances in pulse oximetry monitoring are promising increasing the SV and blood pressure and only a 10%
for predicting fluid responsiveness. An oximetry- chance of being ineffective. Conversely, with lower
derived perfusion index (Pi) reflects the amplitude of SPV, PPV, SVV, and PVi values, the more likely it is
the pulse oximetry waveform (PV) calculated from that the patient is not significantly hypovolemic and
the pulsatile and non-pulsatile infrared signals. that fluid administration will not result in a decrease in
Dynamic Pi change, the pleth variability index (PVi), these indices or an increase in SV and blood pressure.
is calculated over the course of a respiratory cycle as:5 Most of the literature indicates that SPV, PPV, SVV,
and PVi values less than 9% should trigger a more
PVi = (Maximum PV−Minimum PV)/Maximum PV cautious approach to fluid administration because the
specificity is greater than 0.9; fluid administration has
Note that in this formula the denominator is the
a 90% chance of not increasing the SV and blood
maximum PV value, not the mean value as in the
pressure, and has only a 10% chance of being effective.
SPV, PPV, and SVV calculations, thus the PVi is not
Patients with abnormal ventricular function have
a percentage.
unpredictable or no fluid responsiveness, because the
ventricle’s response to increased preload may be vari-
Assessment of Fluid Responsiveness Using able or absent.4,6 The SPV, PPV, SVV, and PVi there-
SPV, PPV, SVV, and PVi fore cannot be used with any degree of certainty in
In patients with normal ventricular contractility, as patients with abnormal ventricular function.
hypovolemia develops (e.g., due to blood loss, diur-
esis, inadequate fluid administration), positive pres- Monitors that Display Pulse Pressure
sure ventilation causes an increase in the SPV, PPV, Variation, Stroke Volume Variation,
SVV, and PVi, which increase further as hypovolemia
becomes more severe. The increase in these indicators and the Pleth Variability Index
is driven by several factors that occur with hypovole-
mia: (1) the vena cava contains less volume and is Overview
more easily compressed or even collapsed during There are now several monitors available for deter-
inspiration, resulting in decreased venous return and mining some combination or permutation of SPV,
subsequent decreased left ventricular preload and SV; PPV, and SVV. However, the SPV and PPV can be
(2) transmission of the pleural pressure to the right determined manually from a hard copy of the arterial
atrium reduces venous return; (3) pulmonary West tracing or by use of the electronic cursor function
zone 1 conditions (i.e., pulmonary artery pressure built into most monitoring screens. Further, their
is less than alveolar pressure) predominate as lower automatic determination and display have been hard-
pulmonary artery pressures, increased right ventricu- wired into standard bedside monitoring units to
lar afterload, and reduced forward flow to the left accompany the basic electrocardiogram, automated
ventricle result in a decreased preload and SV. blood pressure, and pulse oximetry functions. More
By contrast, these physiologic changes are not seen sophisticated determination of SV, SVV, and other
in normovolemia when the vena cava and atria are calculated variables requires arterial pressure wave-
relatively incompressible and West zone 1 conditions form analysis performed by stand-alone monitors
are less prominent, so the net effect of a respiratory dedicated to this task that also can display SPV and/
cycle is minimal change in the SV of both the right or PPV. These monitors may require initial and inter-
and the left ventricles. mittent calibration to a known SV or cardiac output,
The greater the SPV, PPV, SVV, and PVi, the more or rely on proprietary waveform analysis and
likely it is that the patient is significantly hypovolemic algorithms without a need to calibrate. There is only
and that fluid administration will result in a decrease in one pulse oximeter that has the PVi feature.5 The
these indices and an increase in SV and blood pressure. utility of these monitors, however, is limited to
Most of the literature indicates that SPV, PPV, SVV patients with underlying normal physiologic
values that are greater than 13% and or a PVi greater responses; they are less accurate during intense per-
than 14 should trigger fluid administration, as the ipheral vasoconstriction, irregular heart rhythms,
sensitivity is near greater than 0.9. In other words, aortic regurgitation, and cardiac hypocontractility.
While use of these devices has been slowly increas- monitoring,” “+ add non-invasive monitoring,” and
ing, none are currently considered a standard of care, “+ add depth of anesthesia module.” The LiDCO
and they are utilized in only a minority of hospitals. RapidTM system, like the Flo-Trac, does not require
Nevertheless, anesthesiologists should have a basic external calibration. Unlike the Flo-Trac, however,
knowledge of them to understand their data and any type of arterial catheter at any site can be utilized
how they might assist in patient management. with this monitor. Long-term trends in the cardiac
output, SV, SVV, and PPV are also calculated and
Monitors Requiring an Arterial Line and displayed. There are several variations of this device
Calibration to a “Known” Cardiac Output that have undergone frequent reinvention and name
changes. Most recently, a noninvasive arterial pres-
The PiCCO system (MAQUET Holding B.V. & Co., sure methodology (CNAPTM) that utilizes a dual fin-
Germany) uses a simple “pulse contour” analysis of ger cuff to derive beat-to-beat hemodynamic data has
the arterial wave form (i.e., integration of the area been marketed.8
under the arterial waveform), and intermittent cali-
bration with a transpulmonary thermodilution tech- Noninvasive Monitors (No Arterial Line
nique that requires a proprietary thermistor-tipped
arterial line and a central line. It continuously moni- Required)
tors the SV, SVV, and PPV. In addition, it has the The ClearSight System (Edwards Lifesciences
capability for more sophisticated measurement of Corporation, Irvine, CA) is a noninvasive monitor
global end-diastolic volume, intrathoracic blood of arterial blood pressure that uses a small blood
volume, and extravascular lung water via mathemati- pressure cuff placed on the finger and an infrared
cal models utilizing physiologic assumptions, light source with a sensor on the other side of the
although the clinical utility of these parameters is finger to measure the blood volume under the cuff,
less well established than the SVV.7 which changes with pulsation. The device increases
The original LiDCO system (LiDCO, United and decreases the cuff pressure to keep the measured
Kingdom) also analyzes the arterial waveform to blood volume constant via a complex algorithm, and
determine the SV, SVV, and PPV. Calibration relies calculates the blood pressure as well as the SV and
on a lithium dilution technique whereby lithium is SVV continuously.9 LiDCO uses a dual finger cuff that
injected into a peripheral vein and an ion-selective appears to net similar information as the ClearSight
electrode attached to the arterial line senses the system, which has been available for a longer period of
lithium concentration and plots it over time to calcu- time.
late the SV and cardiac output, which are then used to NICOMTM (Cheetah Medical, Vancouver, WA,
calibrate the measurement derived from the arterial USA) uses a unique, noninvasive technology termed
waveform.7 “bioreactance” that is related to bioimpedance.
Bioimpedance, the degree to which the body impedes
Monitors That Require an Arterial Line electric current flow, measures the amplitude of the
voltage change that accompanies changes in fluid
and No Calibration volume. Bioreactance measures the change in phases
The FloTrac system (Edwards Lifesciences, Irvine, of the electrical current via four electrodes applied to
CA) also known as the Vigileo, relies on a special the chest in order to detect SV. Studies measuring the
blood flow sensor connected to a standard arterial cardiac output, SV, and SVV using this device have
line. It uses a proprietary algorithm to integrate the shown a high level of agreement with data from
sensor data, arterial waveform analysis, and patient PiCCO, Flo-Trac, and pulmonary artery catheters.10
demographic data to determine the SV, PPV, and Prior devices utilizing bioimpedance were not widely
SVV. This algorithm has undergone many variations accepted, and of all the devices relying on electrical
over the past several years to achieve greater current, the NICOMTM technology is commercially
accuracy.7 dominant.
LiDCO has expanded its product line over time Cardio Q (Deltex Medical Limited, Chichester,
and on its most recent website under “Build your UK) is a small Doppler probe placed trans-nasally
hemodynamic monitor” now lists customizable into the esophagus and positioned facing the adja-
options, including “standard minimally invasive cent descending aorta. It measures SV and cardiac
112
output and is used to determine the impact of a fluid Goal-Directed Therapy: Principles of Fluid
challenge on these variables; a lack of increase in
response to a fluid challenge may help guide the Management and Outcomes
clinician to avoid further fluid administration. Goal-directed therapy protocols were originally
Unlike all of the other devices described in this developed to treat patients in septic shock14 but have
chapter, this device requires training and skill in been applied to other clinical situations, including
placing the probe properly in order to obtain ade- high-risk patients undergoing major operations. The
quate signals. In the United Kingdom this device has “goal” is to optimize oxygen delivery, primarily by the
been subject to several large randomized trials, pri- administration of fluid, until the SV is maximized.
marily in the context of bowel surgery, that have When the stroke volume is on the flat portion of the
demonstrated its impact on decreasing length of Starling curve, the SV is maximized, and additional
stay and postoperative complications. This device fluid will not increase the cardiac output any further.
has been approved for use by the National Institute At this point more fluid will increase the risk of
for Health Care Excellence.11 hypervolemia, and the only way to further increase
The Masimo-Radical 7TM pulse oximeter (Masimo cardiac output is to administer inotropes. While this
Co, Irvine, California) is a unique monitor that calcu- approach has not been clearly associated with
lates the PVi using the plethysmographic waveform. improved outcomes in septic shock patients, it has
To reiterate, the PVi can be used to determine fluid been incorporated into many enhanced recovery
responsiveness, although it does not determine the after surgery protocols.15,16
SV.5 As this device is completely noninvasive, it is Attempts to maximize SV and minimize SPV,
promising and is undergoing investigation as PPV, SVV, and PVi, however, may inadvertently
a means to determine volume responsiveness in result in more fluid administered than necessary.
children.12 Certain surgeries (e.g., head and neck surgery, pneu-
monectomy) have better outcomes with mild hypo-
Applications volemia; administration of more fluid than necessary
may be detrimental.17
Intraoperative Fluid Management Goal-directed fluid therapy was originally
achieved by utilizing some combination of central
The main goal of advanced hemodynamic and
venous pressure, cardiac output measured from
volumetric monitoring is to objectively guide
a pulmonary artery catheter, and mixed venous oxy-
fluid management. Use of this monitoring is
gen saturation. More recently, the types of monitors
mostly applicable in the intensive care unit but is
discussed herein have been used in several large
increasingly used in the perioperative management
studies.3,5,16 The typical approach is to determine
of high-risk patients, since most of these patients
the SPV, PPV, SVV, and/or PVi, and if it is “normal”
are already on positive pressure ventilation and
(i.e., < 9%) only administer maintenance fluid.
monitored with an arterial line, and because they
However, if the index measure is greater than 13%,
have high morbidity and mortality that may be
a fluid challenge would be administered, the response
reduced with better fluid management. This mon-
tracked, and the process repeated until the index
itoring allows for “just in time” fluid administra-
measure decreases to the normal range. Further
tion, initiated when the direct and indirect SV
volume may be administered if indicated by other
measures indicate significant hypovolemia, but
determinants (e.g., increased heart rate, decreased
prior to development of hypotension. Fluid man-
end tidal CO2, decreased urine output) by maintain-
agement that is guided by advanced hemodynamic
ing a holistic view.
and volumetric monitors is associated with better
There are two relatively large studies of this
outcomes than standard care based on static vari-
approach in patients undergoing major intestinal sur-
ables and intuition, even if there is no difference
gery. The OPTIMISE trial was a multicenter random-
in the total fluid administered. Furthermore, if the
ized trial of 734 patients age greater than 50 assigned
SV indices suggest that a hypotensive patient will
to either standard therapy or a goal-directed fluid
not be fluid-responsive, then vasoactive agents
therapy protocol utilizing the calibrated LiDCO
should be administered and fluid overload may
monitor, intravenous albumin to maximize the stroke
be avoided.5,13
volume, and a constant intravenous infusion of dopex- evidence supporting goal-directed therapy for con-
amine (a rarely used inotrope that is not available in ditions other than sepsis. It is sobering that it took
the United States). There was no significant difference almost 15 years to disprove the most widely
in primary adverse outcomes (i.e., composite of 30-day embraced original indications for goal-directed ther-
moderate or major complications and mortality) or apy, and that related approaches are now being
secondary outcomes (e.g., infection, need for ICU incorporated into enhanced surgery protocols with
admission, length of stay). A meta-analysis of the nearly no evidence supporting their use. Also con-
study data combined with that of 37 other studies of cerning is that once a therapy is incorporated into
lesser quality, however, showed a significant reduction the mainstream, even when it is shown to be useless
in complication rates, including infection and length of or even harmful, removing that therapy is difficult
stay, but no reduction in mortality in the hospital, at and takes a long period of time.
30 days, or at longest follow-up.15
The second trial examining goal-directed fluid Pitfalls and Limitations of PPV, SVV, and
therapy was the POEMAS study, a multicenter rando-
mized trial that used the NICOM monitor in 142 PVi Monitoring
patients who were randomized to standard therapy Two absolute requirements for using PPV, SVV,
or a goal-directed therapy protocol utilizing colloids and PVi measurements are the need for a sinus
and inotropic agents to maximize SV. The patients in rhythm and tidal volumes large enough to recreate
the treatment group received more colloids, blood, the necessary physiological challenge. Typically,
and dobutamine than those receiving standard ther- tidal volumes of 10 ml/kg are required, which are
apy. There was no difference in complication rates or larger than currently recommended in many situa-
length of stay between the two groups.16 tions. Successful monitoring may still occur by
Inevitably, goal-directed therapy will be attempted using smaller tidal volumes chronically, and inter-
in a wider range of patient groups and surgical pro- mittently use 10 ml/kg for 30–60 seconds to
cedures. It is important to note, however, that opera- measure these variables.
tive interventions that may impact hemodynamics Additional limitations are very high respiratory
(e.g., laparoscopy) or require special positioning will rates and abnormal ventricular function. High
result in unpredictable SV measurements and respiratory rates interfere with complete left ven-
responses. For example, changing from a supine to tricle filling and do not allow observation of these
prone position increases the SVV by as much as 25%. phenomena. Abnormal ventricular function inter-
This was considered a normal response in one study, feres with interpretation of these measurements
and the trigger for a fluid challenge in the prone and cannot be detected or quantified by them;
position was therefore set to a 20% change from the therefore, echocardiography may be a better mon-
baseline prone SVV.18 These are unexpected findings itoring adjunct in unstable patients with concomi-
and the approach is very different than in a supine tant cardiac disease.4 In one study, the majority of
patient. ICU patients had one or more conditions (e.g.,
Historically, goal-directed therapy originated in non-sinus rhythm, absolute need for large tidal
the context of emergency room treatment of sepsis, volumes or high respiratory rates, abnormal right
based on a well-known publication by Rivers et al.14 or left ventricular function) that rendered the mon-
in 2001. More recently, however, three large pro- itoring useless.19
spective randomized trials (ProCESS, ARISE, and Although discussed previously, it bears repeat-
ProMISe) performed in patients with sepsis found ing that intense peripheral vasoconstriction is an
no benefit, and a meta-analysis of patients in these impediment to using these kinds of data and
three trials has cast doubt on this approach.20 Some devices, particularly in patients who also have per-
of the failure to reproduce the Rivers results in the ipheral vascular disease. This limiting factor espe-
newer trials may be due to a change in resuscitation cially applies to the Clearsight and LiDCO
of the control groups to be more similar to the CNAPTM monitors which use blood pressure mea-
intervention group in the original Rivers study, as surements from the finger(s), and the Masimo PVi
clinicians over time incorporated the initial findings which is dependent on the plethysmograph and
into their everyday management. There is much less pulse oximetry signal.5,8,9
114
Table 11.1 Summary of nonultrasound monitors to assess intravascular volume status and fluid responsiveness.
Name Manufacturer Key Aspects

Calibrated Arterial Waveform Analysis
PiCCO MAQUET Holding B.V. & Co., Germany Requires central line and arterial line
Calibrated by thermodilution
LiDCO LiDCO, London, UK Requires an arterial line Calibrated by
lithium dilution
Uncalibrated Arterial Waveform Analysis
FloTrac Edwards Lifesciences, Irvine, California, USA Requires an arterial line
LIDCO Rapid LIDCO, London, UK Requires an arterial line
ClearSight Edwards Lifesciences, Irvine, California, USA Non-invasive Finger pressure device
CNAP LiDCO, London, UK Non-invasive Finger pressure device
Other Methodologies
NICOM Cheetah Medical, Vancouver, WA, USA Bioreactance
Cardio Q Deltex Medical Limited, Chichester, UK Esophageal Doppler
Masimo-Radical 7 Masimo Co, Irvine California, USA Pulse oximeter-based plethysmography
Intraoperative use of these measurements may not References

be useful to predict volume responsiveness in patients
1. Rick J, Burke S. Respirator paradox. South Med J
undergoing thoracotomy or laparoscopy.21,22 Their 1978;71:1376–8.
use in children in any setting has not been well
2. Michard F. Changes in arterial pressure during
studied.23 mechanical ventilation. Anesthesiology
2005;103:419–28.
Conclusions 3. Yang X, Du B. Does pulse pressure variation predict
Several hemodynamic and volumetric monitoring fluid responsiveness in critically ill patients?
A systematic review and meta-analysis. Crit Care
devices that can predict fluid responsiveness are
2014;18:650–63.
commercially available (see Table 11.1), but even
simple observation of an arterial line tracing may 4. Marik P, Cavallazzi R, Vasu T, Hirani A. Dynamic
changes in arterial waveform derived variables and fluid
be extremely informative. This approach can
responsiveness in mechanically ventilated patients:
improve patient care by helping to objectively a systematic review of the literature. Crit Care Med
answer the perennial “when” and “how much” ques- 2009;37:2642–7.
tions of fluid administration. In theory, these devices 5. Forget P, Lois F, de Kock M. Goal-directed fluid
should facilitate rapid diagnosis and treatment of management based on the pulse oximeter-derived pleth
hypovolemia, may foster pre-emptive strategies, variability index reduces lactate levels and improves
and assist in avoiding excessive fluid administration. fluid management. Anesth Analg 2010;111:910–4.
All of the devices described above have undergone 6. Carsetti A, Cecconi M, Rhodes A. Fluid bolus therapy:
numerous iterations over time and are still develop- monitoring and predicting fluid responsiveness. Curr
ing with respect to their underlying data acquisition Opin Crit Care 2015;21:388–94.
technology, proprietary algorithms, interface, dis- 7. Porhomayon J, Zadell G, Congello S, Nader N.
play, and pricing. In practice, however, there is no Applications of minimally invasive cardiac output
convincing evidence that employing these devices monitors. Intl J Emerg Med 2012;5:18.
for these purposes has any impact on outcome. 8. LiDCO product information retrieved at www
Nevertheless, as these monitors become less inva- .lidco.com on January 24, 2019.
sive, more sophisticated, and easier to use, they are 9. Raggi E, Sakai T. Update on finger-application-type
increasingly utilized in intensive care units and noninvasive continuous hemodynamic monitors
operating rooms. (CNAP and ccNexfin): physical principles, validation,
and clinical use. Semin Cardiothorac Vasc Anesth abdominal surgery: a prospective, randomized,
2017;21:321–9. multicenter, pragmatic trial: POEMAS Study
10. Waldron N, Miller T, Thacker J, et al. A prospective (PerOperative goal-directed thErapy in Major
comparison of a noninvasive cardiac output monitor Abdominal Surgery). Anesth Analg 2014;119:579–87.
versus esophageal Doppler monitor for goal-directed 17. Chau E, Slinger P. Perioperative fluid management for
fluid therapy in colorectal surgery patients. Anesth pulmonary resection surgery and esophagectomy.
Analg 2014;118:966–75. Semin Cardiothorac Vasc Anesth 2014;18:36–44.
11. Conway D, Hussain A, Gall I. A comparison of 18. Bacchin M, Ceria C, Giannone S, et al. Goal-
noninvasive bioreactance with oesophageal Doppler directed fluid therapy based on stroke volume
estimation of stroke volume during open abdominal variation in patients undergoing major spine
surgery: an observational study. Eur J Anaesthesiol surgery in the prone position: a cohort study. Spine
2013;30:501–8. 2016;41:E1131–E1137.
12. Chandler J, Cook E, Petersen C, et al. Pulse oximeter 19. Mair S, Tschirdewahn J, Gotz S, et al. Applicability of
plethysmograph variation and its relationship to the stroke volume variation in patients of a general
arterial waveform in mechanically ventilated children. intensive care unit: a longitudinal observational study.
J Clin Monit Comput 2012;26:145–51. J Clin Monit Comput 2017;31:1177–87.
13. Guerin L, Monnet X, Teboul J. Monitoring volume 20. Rowan K, Angus D, Baily M, et al. Early goal directed
and fluid responsiveness: from static to dynamic therapy for septic shock-a patient level meta-analysis.
indicators. Best Pract Res Clin Anaesthesiol N Engl J Med 2017;376:2223–34.
2013;27:177–85. 21. Jeong D, Ahn H, Park H, et al. Stroke volume variation
14. Rivers E, Nguyen B, Havstad S, et al. Early goal directed and pulse pressure variation are not useful for
therapy in the treatment of severe sepsis and septic predicting fluid responsiveness in thoracic surgery.
shock. N Engl J Med 2001;345:1368–77. Anesth Analg 2017;125:1158–65.
15. Pearse R, Harrison D, MacDonald N, et al. Effect of 22. Liu F, Zhu S, Ji Q, et al. The impact of intra-abdominal
a perioperative, cardiac output-guided hemodynamic pressure on the stroke volume variation and
therapy algorithm on outcomes following major plethysmographic variability index in patients
gastrointestinal surgery: a randomized clinical trial and undergoing laparoscopic cholecystectomy. Biosci
systematic review. JAMA 2014;311:2181–90. Trends 2015;9:129–33.
16. Pestana D, Espinosa E, Eden A, et al. Perioperative 23. Gan H, Cannesson M, Chandler J, Ansermino J.
goal-directed hemodynamic optimization using Predicting fluid responsiveness in children:
noninvasive cardiac output monitoring in major a systematic review. Anesth Analg 2013;117:1380–92.
116
Chapter
Assessment of Extravascular Lung Water
12 Torsten Loop
Introduction πc = plasma protein oncotic pressure in the capillary

Increased extravascular lung water (EVLW) is caused πev = plasma protein oncotic pressure in the
by abnormal movement of fluid across the pulmonary extravascular tissue/interstitium
capillary barrier. It is comprised of all fluids within the lymphatic flow = the rate of removal of fluid
extravascular compartment of the lung (i.e., outside of While the oncotic gradient is primarily determined by
the pulmonary circulation), including intracellular, protein concentration, the key variable under normal
interstitial, alveolar, lymphatic, and surfactant, but circumstances in the absence of a lung injury is the
pleural fluid is not considered extravascular lung hydrostatic pressure in the pulmonary capillary
water. An increase in EVLW may be caused by two microvasculature, which depends on pulmonary
primary physiologic derangements: (1) an increased blood flow and resistance (equation 2):
hydrostatic intravascular pressure, as seen in cardio-
genic pulmonary edema due to congestive heart fail- Pc = Pla + (PVR × CO)
ure; and (2) a decrease in the effectiveness of the
pulmonary capillary barrier causing an increase in Pc = pulmonary capillary pressure
its permeability, as seen in non-cardiogenic pulmon- Pla = left atrial pressure
ary edema due to acute respiratory distress syndrome
PVR = pulmonary vascular resistance
(ARDS), septic shock, and other conditions that are
associated with less severe or collateral acute lung Under normal circumstances, the small volume of
injuries.1 EVLW resulting from the movement of fluid across
the capillary membrane is whisked away by the
Physiology of Extravascular Lung lymphatic system and deposited back into the venous
Water Accumulation system via the thoracic duct, and hence no
accumulation normally occurs. Under pathologic
Homeostatic mechanisms prevent appreciable trans-
conditions the net movement exceeds the
udation of intravascular fluid into the air-filled alveoli
lymphatic system’s ability to remove it and EVLW
and other extravascular pulmonary tissue. The rela-
increases.
tionship between fluid flow and hydrostatic pressure
Very basically, cardiogenic pulmonary edema is
(P) is described by the classic Starling equation that
caused by increased pressure within the vascula-
describes the balance of hydrostatic and colloidal
ture. Non-cardiogenic pulmonary edema is caused
pressures calculated as (equation 1):2
by an increase in the permeability of the mem-
EVLW = Kf × [(Pc – Pev) – (πc – πev)] – lymphatic brane separating the extra- and intravascular com-
flow partments and perhaps, in very rare cases, by an
exceedingly low oncotic pressure. Cardiogenic pul-
Kf = capillary membrane permeability and filtration monary edema is usually ruled out by the absence
coefficient of heart failure determined by history, echocardio-
Pc = hydrostatic pressure in the capillary graphy, or measurement of the left atrial pressure,
Pev = hydrostatic pressure in the extravascular or its surrogate, the pulmonary capillary wedge
tissue/interstitium pressure.
The tissue and molecular components that control venous pressure (CVP) port of the pulmonary artery
permeability of the membrane separating the extra- catheter; standard thermodilution cardiac output was
and intravascular compartments include the endothe- measured using the pulmonary artery catheter and
lium, the epithelium, an intervening fused basement indocyanine green concentration was measured via a
membrane (i.e., proteoglycan), elastic fibers, and col- femoral artery sample collected via an arterial catheter
lagen I and III. These structures strongly control the continuously sampling blood. This double-indicator
permeability of the pores or channels through which trans-pulmonary thermodilution was reported cap-
fluid moves.3 able of detecting changes in EVLW of 20% or more
The lung is resistant to developing an increased and correlated well with gravimetry.14
EVLW, as the lymphatic drainage is highly effective
under normal conditions. The normal value of the Principles of Transpulmonary
EVLW indexed to body weight (EVLWI) is 7.3 ± Thermodilution
2.8 mL/kg, with normal values of EVLWI considered
The main advance in the measurement of EVLW has
to be less than 10 mL/kg.4,5 Two recent studies have
been the ability to utilize a practical single transpul-
suggested that EVLW should be indexed to height
monary thermodilution (TPTD) technique rather than
only, since indexing EVLW to body weight under-
a cumbersome experimental technique requiring non-
estimates EVLW in patients who are overweight, as
standard processes, including continuous withdrawing
the lungs do not increase in size commensurate with
of blood via a femoral arterial line, and specialized
an increase in adipose tissue.6,7
laboratory equipment to measure the concentration
of the green dye. This single injection technique uses
Measurement of Extravascular Lung one injection of a cold indicator, usually iced saline,
Water into the superior vena cava via a central venous cathe-
Clinical diagnosis of EVLW is based on patient exam- ter, and detection of temperature change over time via
ination, particularly auscultation focused on the pre- a femoral arterial catheter equipped with a thermistor.
sence or absence of rales, chest radiograph, and most This is done by applying a series of assumptions and
recently, lung ultrasound focused on the presence or calculations (see Figure 12.1).15
absence of B-lines (see Chapter 7).8 Various experi- The traditional thermodilution curve used to deter-
mental imaging methods, such as computer tomogra- mine cardiac output is a plot of temperature change on
phy, magnetic resonance imaging, and positron the y axis and time on the x axis. Typically, the curve is
emission tomography, have all been used to quantify asymmetrical, corresponding to the primary dilution
EVLW; however, gravimetry as used in experimental and secondly to recirculation. To quantify the specific
investigation of post-morbid specimens is considered intrathoracic thermal volume (ITTV) and the pulmon-
to be the gold standard.9,10 This ex vivo method ary thermal volume (PTV), the effects of recirculation
consists of measuring the difference in the weight of must be eliminated, and mean transit times must be
the lungs before and after desiccation, which obviously calculated. The mean transit time (MTt) represents the
is not possible in clinical practice.9 While pulmonary time when half of the indicator passes the detection in
ultrasound to detect B-lines helps assess the presence the central artery. It is predominantly determined from
or absence of increased EVLW, the degree and change the area under the thermodilution curve directly after
of EVLW is not measureable using this technique.11,12 the drop of the peak. The exponential downslope time
One of the first studies to demonstrate the clinical (DSt) represents the wash-out function of the indica-
value of measuring EVLW was performed by Sibbald tor. It is calculated from the downslope part of the
et al. in a broad range of critically ill patients with both thermodilution curve. Both mean transit time and
acute cardiogenic and non-cardiogenic pulmonary exponential downslope time serve as the basis for
edema using a standard pulmonary artery catheter calculation of the volumes explained in Figure 12.1.
with a double-indicator thermodilution technique ITTV = CO × MTt
using indocyanine green and temperature change.13
PTV = CO × DSt
To quantify the fluid in the pulmonary capillary bed,
EVLW was measured by injecting 10 ml of iced saline GEDV = ITTV – PTV
solution containing 5 mg of indocyanine green dye ITBV = GEDV × 1.25
into the central circulation as a bolus using the central EVLW = ITTV – ITBV
118
ITTV Figure 12.1 Pulmonary thermodilution

measurement characteristics. RAEDV = right
atrial end-diastolic volume; RVEDV = right
ITBV
ventricular end-diastolic volume;
PBV = pulmonary blood volume;
Cold indicator EVLW Thermal detection EVLW = extravascular lung water;
LAEDV = left atrial end-diastolic volume;
LVEDV = left ventricular end-diastolic
volume; GEDV = cardiac global end-diastolic
volume; ITBV = intrathoracic blood volume;
RAEDV RVEDV PBV LAEDV LVEDV ITTV = intrathoracic thermal volume.
Right heart Left heart
GEDV
ITTV = intrathoracic thermal volume parameters are based on the Stewart−Hamilton and
CO = cardiac output Newman principles.19,20 Both devices use the same
MTT = mean transit time algorithm and provide comparable measurements.21,22
The measurement of EVLWI requires the global
PTV = pulmonary thermal volume
end-diastolic volume (GEDV), which is the cardiac
EDT = exponential decay time volume minus the total pulmonary volume from
GEDV = global end-diastolic volume ITTV (see Figure 12.1). Furthermore, the intrathor-
ITBV = intrathoracic blood volume acic blood volume (ITBV) is estimated from the
EVLW = extravascular lung water GEDV according to the above-mentioned equation
(ITBV = GEDV × 1.25).15 EVLWI is obtained by
The ITBV calculation assumes that the relation- subtracting ITBV from ITTV. The main drawbacks
ship between GEDV and ITBV is linear.15 Potential in measuring EVLW are unreliable data in the case of
limitations in the accuracy of EVLW measurement pulmonary embolism, lung resection, or large pleural
include those common to trans-cardiopulmonary effusions.23 In addition, the EVLW level is indexed to
thermodilution calculation of cardiac output as well the predicted body weight, not the actual body weight,
as tricuspid regurgitation, mitral regurgitation, lung in order to avoid underestimation of EVLW.24
resection, obstruction of major pulmonary vessels and
large increases in positive end-expiratory pressure.16 Acute Respiratory Distress Syndrome
From a pathophysiological point of view, measurement
Clinical Practice of EVLWI or PVPI may be helpful in defining ARDS,
Two commercially available devices that utilize trans- by equating the severity of the pathologic increase in
pulmonary thermodilution, the PiCCO2 (Maquet, pulmonary vascular permeability and alveolar damage
Munich, Germany) and the VolumeView/EV 1000® with the abnormal increase of EVLWI.5,25–29 A retro-
(Edwards Lifesciences, Irvine, CA, USA), may be used spective analysis of 373 critically ill patients demon-
to determine EVLW. The estimation of the EVLW and strated that patients with an EVLW greater than
the pulmonary blood volume also allows the calcula- 15 mL/kg had a mortality rate of 65%, while patients
tion of a pulmonary vascular permeability index with an EVLW less than 10 ml/kg had a mortality rate
(PVPI) equal to the ratio of the extravascular lung of only 33%.27 Specifically, the maximum EVLW was
water index and the pulmonary blood volume. An significantly higher in non-survivors than in survivors,
increased PVPI is consistent with a lung injury, and a with a median of 14.3 mL/kg vs. 10.2 mL/kg, respec-
lower ratio is more suggestive of a hydrostatic-driven tively. Further, an increase in EVLW may be an indi-
congestive heart failure-type increase in EVLW.17,18 cator of early ARDS and becomes apparent before a
Calculations of the PVPI and other hemodynamic large A-a gradient or radiographic change develops.
Septic Shock An EVLWI greater than 11 mL/kg was correlated with

a higher percentage of ALI (70% versus 34%), higher
In patients with septic shock, EVLW measurements
tidal volumes (8.8 ± 1.9 versus 7.8 ± 1.4 mL/kg), and a
are moderately correlated with markers of acute lung
trend to higher mortality (61% versus 44%).36
injury such as lung compliance, PaO2/FiO2 ratio,
EVLW and the associated measurements these
radiographic opacification, lung injury score, and out-
devices record have been investigated in a small
come.30,31 EVLWI and permeability indexes were sig-
study of patients with subarachnoid hemorrhage
nificantly increased in non-survivors at day 3,
who had surgical clipping, and appeared to lower the
compared to scores in survivors. Thus, EVLW might
incidence of vasospasm and cardiopulmonary com-
be of value as an indicator of prognosis and severity of
plications compared with those managed with stan-
sepsis-induced acute lung injury.
dard therapy.37 Its utility in various cardiac and
pulmonary procedures has in some studies shown
Goal-Directed Therapy Targeting benefits, but these studies also have been relatively
Extravascular Lung Water small, had heterogeneous study groups, did not mea-
The quantity of EVLW may be a relevant therapeutic sure the same treatment outcomes, and were incon-
target in the fluid management of critically ill patients. sistent in their findings.35,38,39 EVLW reduction by
Increased positive fluid balance is associated with continuous renal replacement therapy has also been
worse outcomes in acute lung injury (ALI) and demonstrated, but EVLW-guided continuous renal
ARDS, and by inference EVLW and increased positive replacement therapy is not a standard.40
fluid balance are linked.32 The reduction of EVLW via
a goal-directed approach is an appealing strategy, with Application of and Weaning
preliminary evidence supporting its potential incor- from Mechanical Ventilation
poration into other hemodynamic strategies. In a
Determination of EVLW seems intuitively helpful in
small clinical study of 16 patients with septic shock
the setting of high levels of mechanical ventilator
requiring mechanical ventilation, measurement of
support and subsequent weaning plans. Several
ITBVI and TEDVI was shown to be a more useful
experimental studies performed in animals reveal
indicator of EVLW than standard hemodynamic vari-
the protective role of positive end-expiratory pressure
ables, CVP, and pulmonary wedge pressure.33 A 1992
application to reduce EVLW.41,42 Monnet et al.
randomized, prospective study in critically ill ICU
demonstrated that an increase in EVLW during a
patients with pulmonary edema found that fluid man-
spontaneous breathing trial was diagnostic of wean-
agement, primarily by diuresis and fluid restriction,
ing-induced pulmonary edema, with good accuracy
directed by EVLW measurement versus pulmonary
and in particular 100% specificity.43 However, a
artery catheter-directed care, resulted in EVLW
PubMed search performed on March 5, 2018 using
decrease and was associated with fewer ventilator
the search words “wean, extravascular, lung, water”
days and shorter ICU length of stay.34 Thus, driving
identified only 19 publications, none of which used
fluid management by EVLW measurements utilizing
EVLW as a treatment goal or readiness for extubation
a bedside monitor might be advantageous.
parameter.
Cardiac index, PVPI, and EVLWI measured during
the postoperative period after single-lung transplanta-
tion, demonstrated an EVLWI of 12 mL/kg and a PVPI Conclusion
of 2.3, consistent with hydrostatic pulmonary edema.35 Extravascular lung water is a physiologic variable that
EVLWI was significantly increased in patients with a can be determined using commercially available
lower PaO2/FiO2 ratio, 17 mL/kg in the group with devices that utilize a transpulmonary thermodilution
PaO2/FiO2 less than 150 mmHg versus 12 mL/Kg in technique, and may be subject to targeted therapy
the group with PaO2/FiO2 greater than 150 mmHg. including diuresis and treatment of hypotension
Coldermans et al. assessed the prognostic value of with vasopressors instead of fluids. The quantity and
an increased EVLWI and the fluid balance in critically change in EVLW may be prognostic and can be used
ill patients.36 The best predictor for mortality was an in combination with standard vital signs (e.g., blood
EVLWI of greater than 11 mL/kg, showing a 60% pressure), pulse pressure variation or stroke volume
sensitivity and a 57% specificity with a PPV of 61%. variation, and cardiac output, to guide therapy in
120
cases where EVLW minimization itself is a reasonable thermal dye technique in critically ill patients. Chest
goal, especially in patients with ARDS. However, large 1983;83:725–31.
or multicenter outcome studies including EVLW into 14. Mihm FG, Feeley TW, Jamieson SW. Thermal dye
a goal-directed therapeutic strategy are lacking and double indicator dilution measurement of lung water
the approach is not widely practiced. in man: comparison with gravimetric measurements.
Thorax 1987;42:72–6.
15. Sakka SG, Ruhl CC, Pfeiffer UJ, et al. Assessment of
References cardiac preload and extravascular lung water by single
1. Ware LB, Matthay MA. Clinical practice: acute transpulmonary thermodilution. Intensive Care Med
pulmonary edema. N Engl J Med 2005;353:2788–96. 2000;26:180–7.
2. Starling EH. On the absorption of fluids from the 16. Michard F, Schachtrupp A, Toens C. Factors
connective tissue spaces. J Physiol 1896;19:312–26. influencing the estimation of extravascular lung water
3. Miserocchi G. Mechanisms controlling the volume of by transpulmonary thermodilution in critically ill
pleural fluid and extravascular lung water. Eur Respir patients. Crit Care Med 2005;33:1243–7.
Rev 2009;18:244–52. 17. Kushimoto S, Taira Y, Kitazawa Y, et al. The clinical
4. Tagami T, Kushimoto S, Yamamoto Y, et al. usefulness of extravascular lung water and pulmonary
Validation of extravascular lung water measurement vascular permeability index to diagnose and
by single transpulmonary thermodilution: human characterize pulmonary edema: a prospective
autopsy study. Crit Care 2010;14:R162. multicenter study on the quantitative differential
diagnostic definition for acute lung injury/acute
5. Tagami T, Sawabe M, Kushimoto S, et al. Quantitative respiratory distress syndrome. Crit Care 2012;16:R232.
diagnosis of diffuse alveolar damage using
extravascular lung water. Crit Care Med 2013;41:2144– 18. Monnet X, Anguel N, Osman D, et al. Assessing
50. pulmonary permeability by transpulmonary
thermodilution allows differentiation of hydrostatic
6. Huber W, Mair S, Gotz SQ, et al. Extravascular lung pulmonary edema from ALI/ARDS. Intensive Care
water and its association with weight, height, age, and Med 2007;33:448–53.
gender: a study in intensive care unit patients. Intensive
Care Med 2013;39:146–50. 19. Jozwiak M, Teboul JL, Monnet X. Extravascular lung
water in critical care: recent advances and clinical
7. Phillips CR, Chesnutt MS, Smith SM. Extravascular applications. Ann Intensive Care 2015;5:38.
lung water in sepsis-associated acute respiratory
distress syndrome: indexing with predicted body 20. Brown LM, Liu KD, Matthay MA. Measurement of
weight improves correlation with severity of illness extravascular lung water using the single indicator
and survival. Crit Care Med 2008;36:69–73. method in patients: research and potential clinical
value. Am J Physiol Lung Cell Mol Physiol 2009;297:
8. Saugel B, Ringmaier S, Holzapfel K, et al. Physical L547–58.
examination, central venous pressure, and chest
radiography for the prediction of transpulmonary 21. Bendjelid K, Giraud R, Siegenthaler N, Michard F.
thermodilution-derived hemodynamic parameters in Validation of a new transpulmonary thermodilution
critically ill patients: a prospective trial. J Crit Care system to assess global end-diastolic volume and
2011;26:402–10. extravascular lung water. Crit Care 2010;14:R209.
9. Lange NR, Schuster DP. The measurement of lung 22. Kiefer N, Hofer CK, Marx G, et al. Clinical validation
water. Crit Care 1999;3:R19–R24. of a new thermodilution system for the assessment of
cardiac output and volumetric parameters. Crit Care
10. Pearce ML, Yamashita J, Beazell J. Measurement of 2012;16:R98.
pulmonary edema. Circ Res 1965;16:482–8.
23. Monnet X, Teboul JL. Transpulmonary
11. Volpicelli G, Elbarbary M, Blaivas M, et al. thermodilution: advantages and limits. Crit Care
International evidence-based recommendations for 2017;21:147.
point-of-care lung ultrasound. Intensive Care Med
2012;38:577–91. 24. Berkowitz DM, Danai PA, Eaton S, Moss M, Martin
GS. Accurate characterization of extravascular lung
12. Enghard P, Rademacher S, Nee J, et al. Simplified lung water in acute respiratory distress syndrome. Crit Care
ultrasound protocol shows excellent prediction of Med 2008;36:1803–9.
extravascular lung water in ventilated intensive care
patients. Crit Care 2015;19:36. 25. LeTourneau JL, Pinney J, Phillips CR. Extravascular
lung water predicts progression to acute lung injury in
13. Sibbald WJ, Warshawski FJ, Short AK, et al. Clinical patients with increased risk. Crit Care Med
studies of measuring extravascular lung water by the 2012;40:847–54.
26. Kushimoto S, Endo T, Yamanouchi S, et al; PiCCO PiCCO2 device. J Cardiothorac Vasc Anesth 2018;32
pulmonary edema study group. Relationship between (4):1731–5.
extravascular lung water and severity categories of 36. Cordemans C, De Laet I, Van Regenmortel N, et al.
acute respiratory distress syndrome by the Berlin Fluid management in critically ill patients: the role of
definition. Crit Care 2013;17:R132. extravascular lung water, abdominal hypertension,
27. Sakka SG. Extravascular lung water in ARDS patients. capillary leak, and fluid balance. Ann Intensive Care
Minerva Anestesiol 2013;79:274–84. 2012;2:S1.
28. Schuster DP. Identifying patients with ARDS: time for 37. Mutoh T, Kazumata K, Ishikawa T, Terasaka S.
a different approach. Intensive Care Med Performance of bedside transpulmonary
1997;23:1197–203. thermodilution monitoring for goal-directed
29. Perel A. Extravascular lung water and the pulmonary hemodynamic management after subarachnoid
vascular permeability index may improve the hemorrhage. Stroke 2009;40:2368–74.
definition of ARDS. Crit Care 2013;17:108. 38. Lenkin AI, Kirov MY, Kuzkov VV, et al. Comparison
30. Kuzkov VV, Kirov MY, Sovershaev MA, et al. of goal-directed hemodynamic optimization using
Extravascular lung water determined with single pulmonary artery catheter and transpulmonary
transpulmonary thermodilution correlates with the thermodilution in combined valve repair: a
severity of sepsis-induced acute lung injury. Crit Care randomized clinical trial. Crit Care Res Pract
Med 2006;34:1647–53. 2012;2012:821218.
31. Wang H, Cui N, Su L, et al. Prognostic value of 39. Kapoor PM, Magoon R, Rawat RS, et al. Goal-directed
extravascular lung water and its potential role in therapy improves the outcome of high-risk cardiac
guiding fluid therapy in septic shock after initial patients undergoing off-pump coronary artery bypass.
resuscitation. J Crit Care 2016;33:106–13. Ann Card Anaesth 2017;20:83–9.
32. Sakr Y, Vincent JL, Reinhart K, et al. Sepsis occurrence 40. Compton F, Hoffmann C, Zidek W, Schmidt S,
in acutely ill patients investigators: high tidal volume Schaefer JH. Volumetric hemodynamic parameters to
and positive fluid balance are associated with worse guide fluid removal on hemodialysis in the intensive
outcome in acute lung injury. Chest 2005;128:3098– care unit. Hemodial Int 2007;11:231–7.
108. 41. Fernández Mondéjar E, Vazquez Mata G, Cárdnas A,
33. Boussat S, Jacques T, Levy B, et al. Intravascular et al. Ventilation with positive end-expiratory pressure
volume monitoring and extravascular lung water in reduces extravascular lung water and increases
septic patients with pulmonary edema. Intensive Care lymphatic flow in hydrostatic pulmonary edema. Crit
Med 2002;28:712–8. Care Med 1996;24:1562–7.
34. Mitchell JP, Schuller D, Calandrino FS, Schuster 42. Colmenero-Ruiz M, Fernández-Mondéjar E,
DP. Improved outcome based on fluid Fernández-Sacristán MA, Rivera-Fernández R,
management in critically ill patients requiring Vazquez-Mata G. PEEP and low tidal volume
pulmonary artery catheterization. Am Rev Respir ventilation reduce lung water in porcine pulmonary
Dis 1992;145:990–8. edema. Am J Respir Crit Care Med 1997;155:964–70.
35. Tran-Dinh A, Augustin P, Dufour G, et al. 43. Dres M, Teboul JL, Anguel N, et al. Extravascular lung
Evaluation of cardiac index and extravascular lung water, B-type natriuretic peptide, and blood volume
water after single-lung transplantation using the contraction enable diagnosis of weaning-induced
transpulmonary thermodilution technique by the pulmonary edema. Crit Care Med 2014;42:1882–9.
122
Chapter
Point-of-Care Hematology
13 Jacob Raphael, Liza Enriquez, Lindsay Regali,

and Linda Shore-Lesserson
Introduction The Sonoclot® (Sonoclot Coagulation & Platelet

Function Analyzer, Sienco Inc., Arvada, CO) pro-
Perioperative management of hemostasis and coagu-
vides both qualitative and quantitative information
lopathy is a complex, time-sensitive task for the
on the entire coagulation process, from clot initia-
anesthesiologist caring for patients undergoing sur-
tion to fibrinolysis.4 It has been used in obstetric,
gery. The combination of anticoagulant medications
cardiac, liver, vascular, and trauma surgery to
and possible inherent bleeding disorders makes the
identify platelet dysfunction and predict the risk
ability to diagnose potential causes and risks of bleed-
of bleeding (discussed in more detail later in this
ing and guide therapy critically important.
chapter). Studies have also shown Sonoclot to be
Point-of-care (POC) testing is an essential tool
beneficial in managing heparin therapy during car-
that has been used in clinical practice for decades
diac surgery and monitoring for residual heparin
and provides rapid results at the bedside. Recent tech-
after cardiac surgery.5,6 The test uses a kaolin-
nology has progressed, and several new instruments
based activated clotting time (kACT) to monitor
are commercially available to assist the practitioner in
high-dose heparin and a glass bead-activated test
guiding therapy, reducing the administration of
(gbACT) to monitor baseline and post-protamine
unnecessary blood products, and improving patient
levels (low heparin levels).
outcomes.
The most commonly used viscoelastic POC device
is the Thromboelastogram® (TEG).7 The TEG is used
Monitoring Anticoagulation to guide blood product administration and can reduce
The activated clotting time (ACT) is a classic POC test unnecessary transfusions by using a patient-directed
used to guide high-dose heparin therapy in patients approach to specific transfusion therapy. Thus, it is
undergoing cardiopulmonary bypass (CPB), extracor- considered an essential tool to monitor patients who
poreal membrane oxygenation (ECMO), vascular are at high risk for bleeding. In reducing transfusions,
interventions, hemodialysis, and cardiac catheteriza- TEG has been shown to decrease costs and risks
tion. First described by Hattersley in 1966, the ACT associated with transfusion.8,9 Its use in goal-directed
uses whole blood and measures clot formation via the therapy transfusion algorithms is described in more
intrinsic pathway using surface activators such as detail later in this chapter.
celite, kaolin, or glass beads.1 Devices that easily fit TEG has been studied in the monitoring of
in the palm of hand, use disposable cartridges, and heparinization during cardiac surgery. It has
have a test turnaround time of less than 5 minutes shown good correlations with the “gold standard”
have made this test a mainstay of care for more than ACT; however, there are not enough data support-
25 years. It is, however, influenced by several patient- ing the safety of TEG monitoring as a substitute for
related and technology-dependent factors other than ACT monitoring of heparinization during CPB.
the effects of heparin, and thus is highly variable and Current guidelines, therefore, do not make any
subject to artifact.2,3 Devices that assess the viscoelas- recommendations for its use in this setting.10 At
tic properties of whole blood are an advance over present, modifications have been made where the
basic ACT testing, increasing the breadth of hemo- use of heparinase, an enzyme that cleaves heparin,
static abnormalities that can be detected and adding a can be used to detect any residual heparin effect
versatility well beyond simply guiding heparinization. after reversal with protamine.
Monitoring Direct Thrombin Inhibitors Point-of-Care Monitoring of Direct Oral

Direct thrombin inhibitors (DTIs) are a relatively new Anticoagulants
class of anticoagulants that act by inhibiting throm- Direct oral anticoagulants (DOACs) are increasingly
bin. They are used primarily as an alternative to
used for antithrombotic prophylaxis in patients with
heparin for patients with heparin-induced thrombo-
non-valvular atrial fibrillation and at high risk for
cytopenia (HIT) or acute coronary syndrome, and for
venous thromboembolism. This new class of medica-
stroke prevention in patients with atrial fibrillation.11
tions includes the direct thrombin inhibitor dabiga-
There are currently four parenterally administered
tran and the factor Xa inhibitors rivaroxaban,
DTIs (lepirudin, desirudin, bivalirudin, and argatro-
apixaban, and edoxaban. When compared to vitamin
ban) and one orally administered DTI (dabigatran)
K antagonists (VKA), DOACs have similar thrombo-
available in the United States.12,13 prophylaxis efficacy and fewer bleeding complica-
Routine plasma-based coagulation testing (i.e.,
tions.23,24 Perhaps most importantly, DOACs have a
activated partial thromboplastin time) results poorly
more predictable pharmacodynamic profile com-
correlate with DTI anticoagulation and drug concen-
pared to VKAs, hence routine anticoagulation mon-
trations.14,15 The ecarin clotting time (ECT) is a useful
itoring is unnecessary. Nevertheless, knowing the
test in measuring thrombin inhibition. Ecarin, a pro-
exact anticoagulant effect of these medications may
teolytic activator, cleaves prothrombin to form mei-
be important in patients presenting for emergent or
zothrombin, a procoagulant that is directly inhibited
urgent surgical procedures, in posttraumatic hemor-
by DTIs but not by heparin. Thus, the prolongation of rhage, or in anticipation of specific procedures such as
the ECT is specific to inhibition by DTIs and has a
neuroaxial anesthesia. Important pharmacokinetic
direct linear relationship with drug levels.16–19
and pharmacodynamic features of these medications
A modification of the ECT is the ecarin chromo-
are summarized in Table 13.1.25 Standard laboratory
genic assay (ECA). In this assay meizothrombin
tests do not accurately assess the anticoagulation
breakdown of a chromogenic substrate is measured
effect of DOACs,26 and tests that more specifically
in the presence of DTIs. This technique has been
assess the factor Xa activity or the thrombin time are
shown to be more accurate than the standard ECT,
needed. However, tests such as a diluted thrombin
but it is not widely clinically available.20 time or anti-factor Xa activity27,28 are time-consum-
A dilute thrombin time assay is available for
ing and may not be widely available in the context of
research purposes to monitor the qualitative effects
emergency clinical conditions. Thus, whole blood
of dabigatran in plasma. It is currently produced by
POC coagulation tests would be ideal to assess the
Hyphen Biomed and marketed in North America by
presence of DOAC-induced anticoagulation.
Aniara. Pre-diluted patient plasma is mixed with nor-
mal pooled plasma, and purified α-thrombin is added
to initiate clotting. Clotting time is recorded and Dabigatran
graphed linearly compared to DTI calibrators. The Several studies have reported that in the presence of
dilute thrombin time assay has been shown to have a therapeutic plasma levels of dabigatran the ACT is
direct relationship with dabigatran concentration.21,22 prolonged in a concentration-dependent manner.19,28,29
Table 13.1 Pharmacokinetic and pharmacodynamic features of direct oral anticoagulants.
Dabigatran Rivaroxaban Apixaban Edoxaban

a
Half Life 12–18 h 5–13 h 12 h 9–11 h
Offset of Action 24–96 h 24–48 h 24–48 h No Data
Mode of Elimination Renal: 80% Hepatobiliary: 66% Hepatobiliary: 73% Hepatobiliary: 50%
Hepatobiliary: 20% Renal: 33% Renal: 27% Renal: 50%
Elimination by Yes No No No
Dialysis
a
Assuming normal creatinine clearance
Source: Ruff CT, Giugliano RP, Braunwald E, et al. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients
with atrial fibrillation: a meta-analysis of randomised trials. Lancet 2014;383:955–62.
124
Using viscoelastic tests, dabigatran-spiked blood (taken bearing, reducing and arresting blood flow. The time
from healthy volunteers) exhibits significant prolonga- between two consecutive drops is measured as the
tion of the reaction (R-time) in both the kaolin-activated occlusion time (OT) or the time to form a thrombus.
TEG® and the RapidTEG® assays.30–32 Similarly, prolon- The machine also measures thrombolysis using a
gation of the clotting time (CT) of the INTEM® and parameter called the clot lysis time (LT).
EXTEM® assays was reported in dabigatran-treated The GTT has been studied in patients with cardi-
patients.33–36 At this time though, no POC test can be ovascular risk factors,41 end-stage renal disease,45 and
used to assess the effect of anti-thrombin inhibitors with risk for dabigatran-related bleeding complica-
the necessary accuracy and reliability required to guide tions,46,47 but it has not been studied in the periopera-
clinical decision-making. tive setting.
Factor Xa Inhibitors IMPACT-RTM

The ability to detect the effects of factor Xa inhi- The original cone and plate(let) analyzer (CPA) is a
bitors using POC coagulation tests is less predict- POC device that uses a cone and plate apparatus
able than for the direct thrombin inhibitors. and tests platelet function based on the adhesion of
Standard TEG® and ROTEM® intrinsic coagulation platelets to extracellular matrix (ECM) under near
pathway assays and ACT do not correlate linearly physiologic flow conditions.48 The latest CPA mar-
with rivaroxaban concentration29,30 and in fact keted is the IMPACT-RTM (Matis Medical, Beersel,
may fail to demonstrate a residual anticoagulant Belgium). This device assesses all aspects of platelet
effect of rivaroxaban.37 The effects of rivaroxaban function, from primary hemostasis to ultimate pla-
on the extrinsic coagulation pathway assay of telet aggregation. Using IMPACT-R, platelet adhe-
ROTEM® (EXTEM) have been studied more extension, activation, and aggregation are measured
sively, and the results are inconsistent. Eller and under arterial flow conditions. Citrated whole
colleagues have demonstrated prolongation of blood is placed in a polystyrene well and placed
EXTEM only with extremely high plasma levels of under shear stress by the spinning of a cone on a
rivaroxaban,29 whereas others found prolongation plate at a speed of 1,800 revolutions/sec. Primary
at clinically relevant plasma levels.38–40 Current hemostasis causes platelets to adhere to the well,
evidence suggests that viscoelastic POC tests are while “platelet-to-platelet” aggregation also occurs.
also not reliable for detecting the anticoagulation These adherent platelet clumps are then washed,
effect of apixaban.37 stained, and quantified by an image analyzer. The
surface area covered (SC) by the platelets repre-
sents platelet adhesion, and the average size (AS)
Novel Tests of Global Hemostasis of the platelet groups represents platelet aggrega-
tion.49–51
The Global Thrombosis Test IMPACT-RTM has been used to study platelet
The global thrombosis test (GTT) (Thromboquest function in cardiac surgical patients, monitor
Ltd, London, UK) is a novel POC instrument that thrombocytopenic patients, test for congenital
assesses platelet function using non-anticoagulated and acquired platelet defects, and monitor
whole blood.41–44 Platelets are exposed to high shear response to antiplatelet medications.52–54 The
stress under physiological conditions without the device is available for research only and is not
need for different platelet agonists. In this new system, yet commercially available.
whole blood (WB) flows into a conical test tube in
which two ceramic balls are placed (Figure 13.1). The
inner surface of the tube has flat segments creating
Viscoelastic Point-of-Care Coagulation
narrow gaps. When WB flows by gravity through the Tests
narrow gaps and the upper ball bearing, high shear Conventional laboratory tests have limited value in
stress occurs, and platelets are activated. In the space the perioperative management of patients with coa-
between the ball bearings, platelet aggregation occurs, gulopathic bleeding (e.g., CPB-induced coagulopa-
and thrombin is generated. As clot forms, the gaps are thy, trauma-associated coagulopathy). The
gradually occluded at the level of the lower ball prothrombin time (PT, also known as the
Figure 13.1 Schematic showing principle of the global thrombosis test.
international normalized ratio or INR) and the Thromboelastography® (TEG®,

activated partial thromboplastin time (aPTT) were
designed to manage warfarin and heparin antic- Haemonetics, Braintree, MA, USA) and
oagulation, respectively, as well as to evaluate Thromboelastometry (ROTEM®, TEM
patients with hemophilia and other bleeding dis-
orders. As these tests are performed on plasma and International GmbH, Munich, Germany)
only represent the time to initiation of clot forma- Both TEG® and ROTEM® are viscoelastic POC whole
tion, they do not provide data on the platelet−fibri- blood coagulation analyzers that provide global infor-
nogen interaction in clot formation and mation on the dynamics of clot formation, stabiliza-
stabilization. Furthermore, these tests are per- tion, and breakdown, as well as accurately reflect in
formed in a central laboratory and have a long vivo hemostasis (see Figure 13.2). Results are obtained
turnover time (up to 60 minutes), and therefore in less than 30 minutes, making these devices optimal
are not suitable for prediction or management of for the perioperative setting. In their early utilization,
perioperative hemorrhage.55,56 Given these limita- both systems used a semiautomatic pipette to place
tions, the use of viscoelastic POC coagulation citrated whole blood and different activators in a
assays to predict excessive bleeding and guide plastic cup. The modern devices are predicated on a
hemostatic therapies in suspected coagulopathic similar hemostasis testing principle but utilize a car-
patients has significantly increased over the last tridge-based system into which the blood is placed;
two decades and has been incorporated into thus, the test can be run without using a pipette. In
numerous patient blood management guide- TEG® the cup rotates around a pin and a torsion wire
lines.57–61 Several viscoelastic coagulation devices translates the resistance to movement into an electri-
are available for clinical use. cal signal. In ROTEM® the pin rotates while the cup is
126
Clothing Figure 13.2 Schematic TEG (upper part)/

ROTEM (lower part) trace indicating the
commonly reported variables reaction time
Time Kinetics Strength Lysis (R)/ clotting time (CT), clot formation time
(CFT), alpha angle (α), maximum amplitude
(MA)/ maximal clot firmness (MCF) and lysis
α (Ly)/clot lysis (CL). Johansson PI, et al.
Thromboelastography and
K
thromboelastometry in assessing
coagulopathy in trauma. Scand J Trauma
TEG Resusc Emerg Med. 2009;17:45.
R MA Ly
MCF CL
CT
ROTEM
α
CFT
stationary, and the movement of the pin is detected The use of POC-based transfusion algorithms
and displayed graphically. As clot develops, the blood using TEG® or ROTEM® have resulted in significant
becomes more gelatinous and the motion of the cup/ reduction of allogeneic blood product transfusion in
pin is impeded; this is translated graphically as a curve high-risk clinical settings such as cardiovascular sur-
depicting clot kinetics (on the x axis), strength (on the gery,62–65 trauma,66–69 and obstetric-related bleed-
y axis), and dissolution (combined x and y axis). The ing.70,71 A large prospective study of more than
two devices have different names for the tests per- 7,000 cardiac surgery patients from 17 Canadian cen-
formed, but both measure the integrity of the intrinsic ters analyzed transfusion rates and postoperative
and extrinsic coagulation pathways, platelet function, bleeding endpoints, before and after implementation
fibrinogen function, and fibrinolysis. of the transfusion algorithm using ROTEM® and a
After re-calcification, coagulation is initiated by platelet POC analyzer (PlateletWorks®). The use of a
tissue factor to evaluate the extrinsic coagulation POC-based transfusion algorithm resulted in a signif-
pathway, or by a contact activator to evaluate the icant decrease in red blood cell (RBC) and platelet
intrinsic coagulation pathway. Once thrombin is transfusions.72 Furthermore, a prospective study
formed, platelets are activated and interact with fibri- using an algorithm based on ROTEM® and POC pla-
nogen, causing linkage and development of a dense telet analysis (Multiplate® aggregometry) demon-
network of fibrin polymerization. To measure coagu- strated reduced RBC and plasma transfusion,
lation parameters in anti-coagulated, heparinized compared to a conventional laboratory test-based
patients, lyophilized heparinase can be added to the algorithm. Six-month mortality also was significantly
sample. For evaluation of fibrinogen contribution to lower in the POC-based group. Many studies that
clot firmness, a platelet inhibitor may be added to the incorporate viscoelastic-based transfusion algorithms
sample to solely elicit the fibrinogen component of the and demonstrate reduced transfusion therapy empha-
platelet−fibrinogen interaction. Finally, evaluation of size that early use of pro-hemostatic factor concen-
fibrinolysis reveals a very distinctive and easily recog- trates, including fibrinogen concentrates and
nizable pattern. The commercially available TEG® and prothrombin complex concentrates, is key to the
ROTEM® assays and the normal reference ranges for intervention.73–76 The early use of coagulation factor
their analyzed variables are summarized in Tables concentrates in bleeding cardiac surgery patients,
13.2 and 13.3, respectively. The latest models TEG® however, has been questioned,77,78 and whether this
6s and ROTEM® Sigma have been incorporated into practice improves patient outcomes beyond the
clinical practice. Both devices allow multiple assays reported reduction in transfusion of allogeneic blood
using a single cartridge. As these devices are slightly products is still debated.79,80
different from the older models, studies are needed to In conclusion, POC coagulation analysis using
demonstrate the validity of the results obtained by the TEG® and ROTEM® transfusion algorithms to guide
newer models. blood product management has increased
Table 13.2 Commercially available assays for TEG and ROTEM and their clinical indication.
Assay Activator/Additive Purpose of Activator/ Clinical relevance

Additive
TEG
Calcium Reversal of citrate Used for samples collected in citrate
anticoagulation
KTEG Kaolin Intrinsic pathway activation Evaluation of the intrinsic
coagulation pathway
RapidTEG Tissue Factor + Kaolin Extrinsic pathway activation Evaluation of the extrinsic
coagulation pathway
HTEG Heparinase + Kaolin Neutralize heparin Diagnose incomplete heparin
reversal
Functional abciximab Platelet inhibition Evaluation of fibrinogen
Fibrinogen contribution to clot firmness
Platelet Mapping Activator F (reptilase + factor Platelet activation with and Assessment of platelet function
XIII) + ADP without the contribution of
thrombin
ROTEM
Calcium Reversal of citrate Used for samples collected in citrate
anticoagulation
INTEM Ellagic acid + Phospholipid Intrinsic pathway activation Evaluation of the intrinsic
coagulation pathway
EXTEM Tissue Factor + Phospholipid Extrinsic pathway activation Evaluation of the intrinsic
coagulation pathway
HEPTEM Heparinase + Ellagic acid + Diagnose incomplete heparin
Phospholipid reversal
FIBTEM Cytochalasin D Platelet inhibition Evaluation of fibrinogen
contribution to clot firmness
APTEM Aprotinin Inhibition of fibrinolysis Diagnosis of hyperfibrinolysis
ROTEM platelet ADP, Arachidonic acid, TRAP Platelet activation Assessment of platelet function
ADP – adenosine diphosphate; TRAP – thrombin receptor activating peptide.
significantly over the last two decades. These coagula- probe is altered as the blood in the sample clots and
tion tests are performed on whole blood and more becomes more gelatinous. Results are provided both
accurately reflect in vivo hemostasis compared to graphically (Sonoclot signature) and numerically. The
conventional coagulation tests. Furthermore, the ACT represents the time from activation of the reac-
results obtained by TEG® and ROTEM® are available tion to the beginning of fibrin formation. The kinetics
much more quickly than conventional laboratory stu- of clot formation and platelet function can also be
dies. Recent studies with TEG® and ROTEM® have assessed. Recent studies have demonstrated that
demonstrated early detection and a high positive pre- Sonoclot parameters are predictive for post-CPB
dictive value for coagulopathic bleeding as well.81–83 bleeding in adults5 and children84 undergoing cardiac
surgery.
Sonoclot (Sieno Inc. Arvada, CO, USA) Another novel use for the Sonoclot analyzer has
The mechanism of measurement of the Sonoclot POC been reported in heparin-induced thrombocytope-
device involves the insertion of a plastic probe that is nia (HIT) antibody detection. When HIT mono-
vibrating at an ultrasonic frequency into a sample of clonal antibody was added to the Sonoclot sample,
whole blood in a glass cuvette. The vibration of the the anticoagulant activity of heparin was reduced
128
Table 13.3 Nomenclature of analyzed clot variables and normal reference values in TEG and ROTEM.
TEG ROTEM
Clotting time (period to 2 mm amplitude) R (reaction time) CT (clotting time)
WB: 4–8 min INTEM: 122–208 sec
Kaolin: 3–8 min EXTEM: 43–80 sec
Clot kinetics (period from 2 mm-20 mm K (kinetics) CFT (clot formation time)
amplitude) WB: 1–4 min INTEM: 45–110 sec
Kaolin: 1-3 min EXTEM 48–127 sec
Clot strengthening (α angle) α (slope between R and K) α (slope of tangent at 2 mm amplitude)
WB: 47°–74° INTEM: 70°–12°
Kaolin: 55°–78° EXTEM: 65°–80°
Amplitude (at set time points after CT) A10, A30 A5, A10, A20
Maximal clot strength MA (maximal amplitude) MCF (maximal clot firmness)
WB: 55–73 mm INTEM: 51–72 mm
Kaolin: 51–69 mm EXTEM: 52–70 mm
FIBTEM: 9–24 mm
Clot lysis (at set time points) CL (clot lysis) 30, 60 LY 30, 60
Maximal clot lysis - ML (maximal lysis)
WB: whole blood; Kaolin: kaolin-activated re-calcified blood; A5 – amplitude at 5 min; A10 – amplitude at 10 min; A20 – amplitude at 20 min;
A30 – amplitude at 30 min.
in a dose-dependent manner, much like in the in Point-of-Care Platelet Function Tests

vivo disease.85 Though intriguing, this test has not
Platelets play crucial roles in hemostasis and patholo-
yet been tested in HIT patients and is not in clin-
gical thromboembolic conditions.91,92 Dual anti-pla-
ical use.
telet therapy (DAPT) with aspirin and a P2Y12
receptor inhibitor has become the main anti-throm-
Quantra Hemostasis Analyzer (HemoSonics botic treatment in patients with cardiovascular patho-
logical conditions. DAPT is used to treat acute
LLC, Charlottesville, VA, USA) coronary syndrome (ACS), myocardial infarction,
The Quantra Hemostasis Analyzer is a novel viscoe- stroke, and peripheral arterial disease, and it is pre-
lastic POC analyzer that uses Sonic Estimation of scribed after percutaneous coronary interventions
Elasticity via Resonance (SEER) Sonorheometry, an (PCI), coronary artery bypass surgery, and other
ultrasound-based technology that uses high-fre- interventional vascular procedures.
quency ultrasound pulses to quantify the stiffness of Variability in response to antiplatelet medications
a blood sample during the process of coagulation ex makes it essential that the response to antiplatelet drug
vivo.86 Several components of hemostasis can be ana- therapy should be monitored. Drug resistance, mani-
lyzed, including clot time, heparinase clot time, clot fest as normal platelet reactivity (HPR) despite admin-
stiffness, and platelet and fibrinogen contribution to istration of anti-platelet medication, has been
clot stiffness.87 Recent studies have shown good cor- associated with thrombosis and increased incidence
relation between results obtained by the Quantra of major adverse cardiac events (MACE) after
Hemostasis Analyzer and ROTEM® in cardiac and PCI93,94 and raised the importance of platelet function
high-risk spine surgery patients;88,89 however, discre- testing. Additionally, decreased platelet activity exists
pancies have been reported when comparing the after CPB and severe trauma and is one of the most
Quantra Hemostasis Analyzer to TEG®.90 The clinical important contributing factors to perioperative non-
experience with this device is very limited; it has been surgical micro-vascular bleeding,95–98 emphasizing the
recently approved in the European Union for clinical importance of platelet function testing in surgical
use but its approval in the United States by the Food patients with excessive bleeding. Platelets are not only
and Drug Administration is pending. crucial for primary hemostasis, but they also augment
thrombin generation and promote fibrin polymeriza- 180–376, and a therapeutic response to P2Y12 inhi-
tion via GPIIb/IIIa receptors. Therefore, decreased bition is defined as PRU less than 180. HPR in
platelet function may also result in decreased fibrin clopidogrel-treated patients is defined as PRU
polymerization and reduced thrombin generation, greater than 208.110 It is important to note that
further exacerbating coagulopathy and bleeding.99,100 low hematocrit values spuriously increase PRU
In patients treated with antiplatelet therapy, pre- readings,111 probably because of a confounding
operative assessment of platelet function has been influence on light transmittance. In a retrospective
shown to predict bleeding complications in cardiac cohort of CABG patients, clopidogrel-treated
surgery.101–103 Thus, platelet function testing may be patients with low agglutination and low PRU were
used to individualize and optimize timing of surgery more likely to bleed postoperatively and require
in ACS patients treated with DAPT. The use of plate- transfusion of pro-hemostatic factors.112 However,
let function analyses, in combination with viscoelastic in a small group of low-risk CABG patients, neither
coagulation tests as part of goal-directed transfusion the VerifyNow nor any other platelet POC test cor-
management, contributes to decreased perioperative related well with blood loss, as assessed by post-
bleeding and reduced transfusion requirements after operative chest tube output or reduction in
cardiac surgery9,65,74,104–106 and trauma.66,69,107,108 hematocrit value.113 Further research is needed to
Light transmission aggregometry (LTA) using pla- establish the role of this platelet function assay in
telet-rich plasma is the gold standard for platelet predicting peri-procedural bleeding.
function testing, against which many POC monitors
are compared. This in vitro technique is technically Innovance® Platelet Function Analyzer
cumbersome, time-consuming, and performed under
non-physiological conditions,109 thus it is not widely (PFA-100/200, Siemens Healthineers)
utilized in the perioperative setting. Several whole The PFA-100 original device was a cartridge-based
blood POC methods for platelet function testing are assay that mimics the bleeding time in an ex vivo
commercially available (see Table 13.4). Each device test. A small volume of blood is aspirated through
has different features, including sample processing, an aperture in a membrane coated with platelet
type and potency of platelet agonists, and platform agonists until the aperture is completely occluded
for detecting platelet activity. by a platelet plug. The time to complete occlusion
is reported as “closure time.” A result greater than
VerifyNow System (Accriva Diagnostics, San 175 seconds is considered abnormal. The newer
device, Innovance® PFA-200, has added a P2Y
Diego, CA, USA) assay to the preexisting cartridges of the earlier
VerifyNow (formerly Ultegra, Accumetrics) is a model. The Innovance® P2Y cartridge was devel-
fully automated whole blood optical-based analy- oped in view of the relative insensitivity of pre-
zer that uses agglutination and light transmission vious cartridges to P2Y12 receptor antagonists.
to measure P2Y12 platelet reactivity units (PRUs). This global test of platelet function is easy to
Citrated blood is mixed with fibrinogen-coated use, rapid, requires a small volume of blood
polystyrene beads and activated by arachidonic (0.8 mL per cartridge), and does not require sub-
acid (aspirin cartridge), ADP and prostaglandin stantial training. Yet as this is a global platelet
E1 (P2Y12 cartridge), or thrombin receptor activat- adhesion assay, the closure time may be influenced
ing peptide (GPIIb/IIIa cartridge). Agglutination by platelet count, von Willebrand factor levels,
between activated platelets and the fibrinogen- and hematocrit. Results need to be carefully inter-
coated beads causes the beads to fall out of sus- preted in patients with a platelet count less than
pension, thereby resulting in an increase in light 50,000/μL and anemic patients with hematocrit
transmission through the sample. This increase in below 25%.114 It has been reported that patients
light transmittance is converted to a value (i.e., with blood type O have longer closure times com-
PRU) that is related to the ability of the platelets pared to non-O blood groups.115 A normal closure
to be activated or to agglutinate. As the degree of time may be useful in ruling out a significant
platelet inhibition increases, agglutination platelet defect (high negative predictive value);
decreases, as do PRUs. The normal PRU range is however, the positive predictive value of closure
130
Table 13.4 POC platelet function monitoring devices.
Device Principle Available Assays Advantages Limitations

PFA-100/ Mimics bleeding time Col/ADP (CADP) Whole blood test Dependent on platelet count,
200 Measures time clot time Col/EPI (CEPI) Quick hematocrit level and von
related to high-shear Innovance P2Y assay Requires small Willebrand factor levels
platelet adhesion and sampling volume Weak predictive value for
aggregation periprocedural bleeding116
Multiplate Measures changes in ADP Arachidonic acid Whole blood test Dependent on platelet count
impedance caused by Thrombin receptor Run multi channels and hematocrit values
platelet aggregation after activating protein simultaneously
exposure to agonists (TRAP) Widely used
Good predictive value
for bleeding in cardiac
surgery
VerifyNow Agglutination between PRU (P2Y12 receptor Whole blood test Dependent on hematocrit
activated platelets and inhibition) aspirin test Simple to use values
fibrin-coated beads GPIIb/IIIa test Three antiplatelet Low predictive value in
Light transmission through medication assays postoperative bleeding
sample is measured exist
TEG Use of kaolin-activated Platelet activation by Whole blood test Incomplete inhibition of
Platelet thromboelastography to arachidonic acid (TEG Good correlation with thrombin-mediated platelet
mapping estimate platelet function PMAA) optically monitored activation may result in
Platelet activation by platelet underestimation of ADP or AA
ADP (TEG PMADP) aggregation129 inhibition
Can be combined with Results may be affected by
viscoelastic clot fibrinogen and hematocrit
assessment levels
ROTEM Impedance aggregometry ADP-TEM Whole blood test Significantly different results
Platelet ARA-TEM Can be combined with when compared with
TRAP-TEM viscoelastic clot multiplatelet
assessment aggregometry132
Limited publications
regarding clinical use
AA – arachidonic acid; ADP adenosine diphosphate; GPIIb/IIIA – glycoprotein IIb/IIIa; PRU – P2Y12 reaction units; TEG –
thromboelastography.
time and peri-procedural bleeding endpoints has transmittance aggregometry (LTA) results has been
been relatively weak.116,117 reported.119 ADP is used as an agonist for detection of
P2Y12 receptor inhibition (by clopidogrel or other
Multiple Electrode Aggregometry (The thienopyridines), and thrombin receptor-activating
peptide (TRAP) is used for assessment of GPIIb/IIIa
Multiplate® System, Roche Diagnostics, inhibition.120 Results are reported as area under the
Rotkreuz, Switzerland) aggregation curve, and for each reagent there is a
This assay is based on whole blood impedance aggre- different normal range. For example, the normal
gometry. It is widely used in DAPT-treated patients to ADP response is 57–113 aggregation units, and values
assess platelet function before surgical interventions. less than 57 aggregation units represent impaired
Whole blood is collected in a hirudin-containing tube. platelet function.
Each sample is inserted into a test well containing two Several studies have attempted to identify a cutoff
independent sensor units (electrode wires) and mixed level of ADP-initiated platelet aggregation in order
with normal saline and a platelet agonist. Electrical to predict major bleeding complications in DAPT-
impedance changes are monitored and recorded over treated patients who require surgical interventions.
six minutes, while activated platelets adhere to the It has been suggested that for ADP-initiated aggre-
electrodes.118 Significant agreement with light gation a cutoff of greater than 22 aggregation units
represents an acceptable bleeding risk in clopido- function and monitor platelet inhibition by aspirin or
grel-treated cardiac surgery patients103 and ticagre- P2Y12 inhibitors. The standard kaolin TEG is used as a
lor-treated cardiac surgery patients.101 It is reference for thrombin-mediated platelet activation,
important to consider that several factors such as which is considered maximal activation. In the plate-
age, presence of diabetes,121 hematocrit level, platelet let mapping assay, heparinized whole blood is mixed
count, and the delay between blood sample collec- with Activator F (reptilase and activated FXIII), which
tion and platelet function testing are all known to induces fibrin polymerization independent of platelet
influence whole blood platelet aggregometry activation. In another sample, ADP is added, which
values.122–124 Despite this fact, whole blood aggrego- causes platelet activation if platelets are responsive to
metry has served as an important guide for platelet ADP. ADP-stimulated maximum amplitude
transfusions in many goal-directed transfusion (MAADP) and fibrin-specific MA (MAFib) are used to
algorithms.74,125 calculate platelet inhibition in reference to thrombin-
mediated MA (MAThrombin) using the following
PlateletWorks® (Helena Laboratories, formula:
Beaumont, TX, USA) % Platelet inhibition = 100 − [(MAADP − MAFib)
This device uses the principle of the platelet count ÷ (MAThrombin − MAFib) × 100]
ratio to assess platelet reactivity. The platelet count is
measured in a standard EDTA tube and tubes into Alternatively, the area under the curve of ADP-stimu-
which a platelet agonist (e.g., ADP) has been added. In lated TEG tracing at 15 minutes (AUC15) can be used
patients with normal platelet function, the presence of for quicker results (short TEG). HPR under clopido-
an agonist significantly reduces the platelet count, due grel treatment is defined as less than 30% reduction in
to their aggregation into clumps, thus rendering them the AUC15 in the ADP-activated TEG compared to
“unmeasured as platelets.” In contrast, the platelet the kaolin TEG.128
count in the control sample, without an agonist, Comparison of results obtained by TEG® platelet
remains unchanged. In patients with platelet dysfunc- mapping with optically monitored platelet aggregation
tion, platelets will be refractory to the agonist and will demonstrated good correlation,129 although incom-
not aggregate, thus reducing the difference between plete thrombin inhibition by heparin may lead to
the activated platelet count and the control condition. thrombin-mediated platelet activation and hence
The ratio of the activated platelet count to the non- underestimation of the degree of platelet inhibition.130
activated (control) platelet count is inversely related ROTEM® has a platelet function analyzer that is
to platelet reactivity and reflective of platelet function. based on impedance aggregometry. In the ROTEM®
Although a study that examined the correlation Delta analyzer two channels are available for whole
between preoperative platelet dysfunction (as assessed blood platelet analysis, in addition to the standard
by PlateletWorks®) and postoperative bleeding risk126 four channels that are used for viscoelastic clot
showed no relationship, this analyzer is still valued for analysis. The results are presented by three differ-
its use at the POC during cardiac surgery, and many ent parameters: (1) the amplitude at 6 minutes in
have successfully incorporated PlateletWorks® into Ohm (A6), (2) maximum slope (MS) of the aggre-
their POC-based transfusion algorithms.72,105,127 gation graph in Ohm/min, and (3) area under the
Further research with this device is needed to validate curve (AUC) in Ohm x min. Using three types of
its predictive power for hemmorrhage, determine if it reagents the effect of different platelet inhibitors
can be used to define a safe level of platelet activity to can be evaluated: ADP-TEM for detection of
proceed with interventions despite being on therapy, P2Y12 inhibitors, ARA-TEM (arachidonic acid-
and determine its cost-effectiveness in transfusion TEM) for detection of aspirin and other cycloox-
algorithms. ygenase inhibitors, and TRAP-TEM (thrombin
receptor agonist peptide-TEM) for measuring pla-
telet inhibition mediated by GPIIb/IIIa antagonists.
Platelet Testing Using TEG and ROTEM There are very limited data regarding incorpora-
Both TEG® and ROTEM® have platforms for POC tion of this analyzer into perioperative POC-based
platelet function analysis. The TEG platelet mapping transfusion algorithms for patients with excessive
platform uses thromboelastography to assess platelet bleeding.67,131
132
Cardiopulmonary Bypass. Ann Thorac Surg

Conclusion 2018;105:650–62.
The perioperative management of patients receiving
11. Di Nisio M, Middeldorp S, Buller HR. Direct thrombin
anticoagulant medication or developing intraoperative inhibitors. N Engl J Med 2005;353:1028–40.
coagulopathy relies heavily on the ability to rapidly and
12. Stangier J, Rathgen K, Stahle H, Gansser D, Roth W.
accurately detect hemostasis abnormalities and guide
The pharmacokinetics, pharmacodynamics and
therapy. Point-of-care devices aimed at coagulation tolerability of dabigatran etexilate, a new oral direct
testing have progressed, and there are several commer- thrombin inhibitor, in healthy male subjects. Br J Clin
cially available devices that allow for detection of a wide Pharmacol 2007;64:292–303.
variety of specific defects. With increased utilization of 13. Stangier J, Stahle H, Rathgen K, Fuhr R.
these tests and the implementation of goal-directed Pharmacokinetics and pharmacodynamics of the
blood management protocols, blood product adminis- direct oral thrombin inhibitor dabigatran in healthy
tration will be precise and rational, and total product elderly subjects. Clin Pharmacokinet 2008;47:47–59.
administration should decrease. 14. Nielsen VG, Steenwyk BL, Gurley WQ, Pereira SJ, Lell
WA, Kirklin JK. Argatroban, bivalirudin, and
lepirudin do not decrease clot propagation and
References strength as effectively as heparin-activated
1. Hattersley PG. Activated coagulation time of whole antithrombin in vitro. J Heart Lung Transplant
blood. JAMA 1966;196:436–40. 2006;25:653–63.
2. Garvin S, FitzGerald DC, Despotis G, Shekar P, Body 15. Samama MM, Guinet C. Laboratory assessment of new
SC. Heparin concentration-based anticoagulation for anticoagulants. Clin Chem Lab Med 2011;49:761–72.
cardiac surgery fails to reliably predict heparin bolus
16. Nowak G. The ecarin clotting time, a universal method
dose requirements. Anesth Analg 2010;111:849–55.
to quantify direct thrombin inhibitors. Pathophysiol
3. Gravlee GP, Arora S, Lavender SW, et al. Predictive Haemost Thromb 2003;33:173–83.
value of blood clotting tests in cardiac surgical
17. Siegmund R, Boer K, Poeschel K, Wolf G, Deufel T,
patients. Ann Thorac Surg 1994;58:216–21.
Kiehntopf M. Comparison of the ecarin chromogenic
4. Hett DA, Walker D, Pilkington SN, Smith DC. assay and different aPTT assays for the measurement
Sonoclot analysis. Br J Anaesth 1995;75:771–6. of argatroban concentrations in plasma from healthy
5. Bischof DB, Ganter MT, Shore-Lesserson L, et al. individuals and from coagulation factor deficient
Viscoelastic blood coagulation measurement with patients. Thromb Res 2008;123:159–65.
Sonoclot predicts postoperative bleeding in cardiac 18. Vaishnava P, Eagle KA. Coronary stents and risk for
surgery after heparin reversal. J Cardiothorac Vasc noncardiac surgery: much ado about something,
Anesth 2015;29:715–22. nothing, or DAPT? J Am Coll Cardiol 2016;67:1050–2.
6. Dzemali O, Ganter MT, Zientara A, et al. Evaluation of 19. van Ryn J, Stangier J, Haertter S, et al. Dabigatran
a new Sonoclot device for heparin management in etexilate–a novel, reversible, oral direct thrombin
cardiac surgery. Clin Appl Thromb Hemost inhibitor: interpretation of coagulation assays and
2017;23:20–6. reversal of anticoagulant activity. Thromb Haemost
7. Luddington RJ. Thromboelastography/ 2010;103:1116–27.
thromboelastometry. Clin Lab Haematol 2005;27:81–90. 20. Lange U, Nowak G, Bucha E. Ecarin chromogenic
8. Ak K, Isbir CS, Tetik S, et al. Thromboelastography- assay–a new method for quantitative determination of
based transfusion algorithm reduces blood product use direct thrombin inhibitors like hirudin. Pathophysiol
after elective CABG: a prospective randomized study. J Haemost Thromb 2003;33:184–91.
Card Surg 2009;24:404–10. 21. Avecilla ST, Ferrell C, Chandler WL, Reyes M. Plasma-
9. Shore-Lesserson L, Manspeizer HE, DePerio M, diluted thrombin time to measure dabigatran
Francis S, Vela-Cantos F, Ergin MA. concentrations during dabigatran etexilate therapy.
Thromboelastography-guided transfusion algorithm Am J Clin Pathol 2012;137:572–4.
reduces transfusions in complex cardiac surgery. 22. Love JE, Ferrell C, Chandler WL. Monitoring direct
Anesth Analg 1999;88:312–9. thrombin inhibitors with a plasma diluted thrombin
10. Shore-Lesserson L, Baker RA, Ferraris VA, et al. The time. Thromb Haemost 2007;98:234–42.
Society of Thoracic Surgeons, The Society of 23. Ufer M. Comparative efficacy and safety of the novel
Cardiovascular Anesthesiologists, and The American oral anticoagulants dabigatran, rivaroxaban and
Society of ExtraCorporeal Technology: Clinical apixaban in preclinical and clinical development.
Practice Guidelines-Anticoagulation During Thromb Haemost 2010;103:572–85.
24. Ruff CT, Giugliano RP, Braunwald E, et al. 37. Iapichino GE, Bianchi P, Ranucci M, Baryshnikova E.
Comparison of the efficacy and safety of new oral Point-of-care coagulation tests monitoring of direct
anticoagulants with warfarin in patients with atrial oral anticoagulants and their reversal therapy: state of
fibrillation: a meta-analysis of randomised trials. the art. Semin Thromb Hemost 2017;43:423–32.
Lancet 2014;383:955–62. 38. Escolar G, Arellano-Rodrigo E, Lopez-Vilchez I, et al.
25. Levy JH, Spyropoulos AC, Samama CM, Douketis J. Reversal of rivaroxaban-induced alterations on
Direct oral anticoagulants: new drugs and new hemostasis by different coagulation factor concentrates
concepts. JACC Cardiovasc Interv 2014;7:1333–51. – in vitro studies with steady and circulating human
26. Favaloro EJ, Lippi G. Laboratory testing in the era of blood. Circ J 2015;79:331–8.
direct or non-vitamin K antagonist oral 39. Korber MK, Langer E, Ziemer S, et al. Measurement
anticoagulants: a practical guide to measuring their and reversal of prophylactic and therapeutic peak
activity and avoiding diagnostic errors. Semin Thromb levels of rivaroxaban: an in vitro study. Clin Appl
Hemost 2015;41:208–27. Thromb Hemost 2014;20:735–40.
27. Douxfils J, Mullier F, Robert S, et al. Impact of 40. Perzborn E, Heitmeier S, Laux V, Buchmuller A.
dabigatran on a large panel of routine or specific Reversal of rivaroxaban-induced anticoagulation with
coagulation assays. Laboratory recommendations for prothrombin complex concentrate, activated
monitoring of dabigatran etexilate. Thromb Haemost prothrombin complex concentrate and recombinant
2012;107:985–97. activated factor VII in vitro. Thromb Res
28. Douxfils J, Tamigniau A, Chatelain B, et al. 2014;133:671–81.
Measurement of non-VKA oral anticoagulants versus 41. Saraf S, Christopoulos C, Salha IB, Stott DJ, Gorog DA.
classic ones: the appropriate use of hemostasis assays. Impaired endogenous thrombolysis in acute coronary
Thromb J 2014;12:24. syndrome patients predicts cardiovascular death and
29. Eller T, Busse J, Dittrich M, et al. Dabigatran, nonfatal myocardial infarction. J Am Coll Cardiol
rivaroxaban, apixaban, argatroban and fondaparinux 2010;55:2107–15.
and their effects on coagulation POC and platelet 42. Saraf S, Wellsted D, Sharma S, Gorog DA. Shear-
function tests. Clin Chem Lab Med 2014;52:835–44. induced global thrombosis test of native blood: pivotal
30. Dias JD, Norem K, Doorneweerd DD, et al. Use of role of ADP allows monitoring of P2Y12 antagonist
thromboelastography (TEG) for detection of new oral therapy. Thromb Res 2009;124:447–51.
anticoagulants. Arch Pathol Lab Med 2015;139:665–73. 43. Yamamoto J, Inoue N, Otsui K, Ishii H, Gorog DA.
31. Solbeck S, Meyer MA, Johansson PI, et al. Monitoring Global Thrombosis Test (GTT) can detect major
of dabigatran anticoagulation and its reversal in vitro determinants of haemostasis including platelet
by thromboelastography. Int J Cardiol 2014;176: reactivity, endogenous fibrinolytic and thrombin
794–9. generating potential. Thromb Res 2014;133:919–26.
32. Solbeck S, Ostrowski SR, Stensballe J, Johansson PI. 44. Yamamoto J, Yamashita T, Ikarugi H, et al. Gorog
Thromboelastography detects dabigatran at thrombosis test: a global in-vitro test of platelet
therapeutic concentrations in vitro to the same extent function and thrombolysis. Blood Coagul Fibrinolysis
as gold-standard tests. Int J Cardiol 2016;208:14–8. 2003;14:31–9.
33. Cotton BA, McCarthy JJ, Holcomb JB. Acutely injured 45. Sharma S, Farrington K, Kozarski R, et al. Impaired
patients on dabigatran. N Engl J Med 2011;365:2039– thrombolysis: a novel cardiovascular risk factor in end-
40. stage renal disease. Eur Heart J 2013;34:354–63.
34. Davis PK, Musunuru H, Walsh M, et al. The ex vivo 46. Otsui K, Gorog DA, Yamamoto J, et al. Global
reversibility of dabigatran-induced whole-blood thrombosis test – a possible monitoring system for the
coagulopathy as monitored by thromboelastography: effects and safety of dabigatran. Thromb J 2015;13:39.
mechanistic implications for clinical medicine. 47. Rosser G, Tricoci P, Morrow D, et al. PAR-1 antagonist
Thromb Haemost 2012;108:586–8. vorapaxar favorably improves global thrombotic status
35. Herrmann R, Thom J, Wood A, et al. Thrombin in patients with coronary disease. J Thromb
generation using the calibrated automated Thrombolysis 2014;38:423–9.
thrombinoscope to assess reversibility of dabigatran 48. Kenet G, Lubetsky A, Shenkman B, et al. Cone and
and rivaroxaban. Thromb Haemost 2014;111:989–95. platelet analyser (CPA): a new test for the prediction of
36. Stein P, Bosshart M, Brand B, et al. Dabigatran bleeding among thrombocytopenic patients. Br J
anticoagulation and Stanford type A aortic dissection: Haematol 1998;101:255–9.
lethal coincidence: case report with literature review. 49. Gerrah R, Snir E, Brill A, Varon D. Platelet function
Acta Anaesthesiol Scand 2014;58:630–7. changes as monitored by cone and plate(let) analyzer
134
during beating heart surgery. Heart Surg Forum 2004;7: 61. Shaylor R, Weiniger CF, Austin N, et al. National and
E191–5. international guidelines for patient blood management
50. Spectre G, Brill A, Gural A, et al. A new point-of-care in obstetrics: a qualitative review. Anesth Analg
method for monitoring anti-platelet therapy: 2017;124:216–32.
application of the cone and plate(let) analyzer. Platelets 62. Shore-Lesserson L, Manspeizer HE, DePerio M, et al.
2005;16:293–9. Thromboelastography-guided transfusion algorithm
51. Varon D, Lashevski I, Brenner B, et al. Cone and plate reduces transfusions in complex cardiac surgery.
(let) analyzer: monitoring glycoprotein IIb/IIIa Anesth Analg 1999;88:312–9.
antagonists and von Willebrand disease replacement 63. Rahe-Meyer N, Solomon C, Winterhalter M, et al.
therapy by testing platelet deposition under flow Thromboelastometry-guided administration of
conditions. Am Heart J 1998;135:S187–93. fibrinogen concentrate for the treatment of excessive
52. Gerrah R, Brill A, Tshori S, et al. Using cone and plate intraoperative bleeding in thoracoabdominal aortic
(let) analyzer to predict bleeding in cardiac surgery. aneurysm surgery. J Thorac Cardiovasc Surg
Asian Cardiovasc Thorac Ann 2006;14:310–5. 2009;138:694–702.
53. Shenkman B, Schneiderman J, Tamarin I, et al. Testing 64. Ranucci M, Baryshnikova E, Ranucci M, Silvetti S,
the effect of GPIIb-IIIa antagonist in patients Surgical and Clinical Outcome Research (SCORE)
undergoing carotid stenting: correlation between Group. Fibrinogen levels compensation of
standard aggregometry, flow cytometry and the cone thrombocytopenia-induced bleeding following
and plate(let) analyzer (CPA) methods. Thromb Res cardiac surgery. Int J Cardiol 2017;249:96–100.
2001;102:311–7. 65. Weber CF, Gorlinger K, Meininger D, et al. Point-of-
54. Shenkman B, Matetzky S, Fefer P, et al. Variable care testing: a prospective, randomized clinical trial of
responsiveness to clopidogrel and aspirin among efficacy in coagulopathic cardiac surgery patients.
patients with acute coronary syndrome as assessed by Anesthesiology 2012;117:531–47.
platelet function tests. Thromb Res 2008;122:336–45. 66. Spahn DR. TEG(R)- or ROTEM(R)-based
55. Toulon P, Ozier Y, Ankri A, et al. Point-of-care versus individualized goal-directed coagulation algorithms:
central laboratory coagulation testing during don’t wait–act now! Crit Care 2014;18:637.
haemorrhagic surgery. A multicenter study. Thromb 67. Stein P, Kaserer A, Spahn GH, Spahn DR. Point-of-
Haemost 2009;101:394–401. care coagulation monitoring in trauma patients. Semin
56. Haas T, Fries D, Tanaka KA, et al. Usefulness of Thromb Hemost 2017;43:367–74.
standard plasma coagulation tests in the management 68. Schochl H, Nienaber U, Maegele M, et al. Transfusion
of perioperative coagulopathic bleeding: is there any in trauma: thromboelastometry-guided coagulation
evidence? Br J Anaesth 2015;114:217–24. factor concentrate-based therapy versus standard fresh
57. Society of Thoracic Surgeons Blood Conservation frozen plasma-based therapy. Crit Care 2011;15:R83.
Guideline Task F, Ferraris VA, Brown JR, et al. 2011 69. Gonzalez E, Moore EE, Moore HB, et al. Goal-directed
update to the Society of Thoracic Surgeons and the hemostatic resuscitation of trauma-induced
Society of Cardiovascular Anesthesiologists blood coagulopathy: a pragmatic randomized clinical trial
conservation clinical practice guidelines. Ann Thorac comparing a viscoelastic assay to conventional
Surg 2011;91:944–82. coagulation assays. Ann Surg 2016;263:1051–9.
58. Task Force on Patient Blood Management for Adult 70. Snegovskikh D, Souza D, Walton Z, et al. Point-of-care
Cardiac Surgery of the European Association for viscoelastic testing improves the outcome of
Cardio-Thoracic S, the European Association of pregnancies complicated by severe postpartum
Cardiothoracic Anaesthesiology (EACTA), Boer C, hemorrhage. J Clin Anesth 2018;44:50–6.
Meesters MI, et al. 2017 EACTS/EACTA Guidelines on 71. Butwick AJ, Goodnough LT. Transfusion and
patient blood management for adult cardiac surgery. J coagulation management in major obstetric
Cardiothorac Vasc Anesth 2018;32(1):88–120. hemorrhage. Curr Opin Anaesthesiol 2015;28:275–84.
59. Carson JL, Guyatt G, Heddle NM, et al. Clinical 72. Karkouti K, Callum J, Wijeysundera DN, et al. Point-
practice guidelines from the AABB: red blood cell of-care hemostatic testing in cardiac surgery: a
transfusion thresholds and storage. JAMA stepped-wedge clustered randomized controlled trial.
2016;316:2025–35. Circulation 2016;134:1152–62.
60. Rossaint R, Bouillon B, Cerny V, et al. The European 73. Ranucci M, Baryshnikova E, Crapelli GB, et al.
guideline on management of major bleeding and Randomized, double-blinded, placebo-controlled trial
coagulopathy following trauma: fourth edition. Crit of fibrinogen concentrate supplementation after
Care 2016;20:100.
complex cardiac surgery. J Am Heart Assoc 2015;4: bleeding in pediatric patients undergoing cardiac
e002066. surgery for congenital cyanotic heart disease: a
74. Gorlinger K, Dirkmann D, Hanke AA, et al. First-line prospective observational study. J Cardiothorac Vasc
therapy with coagulation factor concentrates Anesth 2017;31:901–8.
combined with point-of-care coagulation testing is 85. Wanaka K, Asada R, Miyashita K, et al. Novel HIT
associated with decreased allogeneic blood transfusion antibody detection method using sonoclot(R)
in cardiovascular surgery: a retrospective, single-center coagulation analyzer. Thromb Res 2015;135:127–9.
cohort study. Anesthesiology 2011;115:1179–91. 86. Corey FS, Walker WF. Sonic estimation of elasticity via
75. Tanaka KA, Esper S, Bolliger D. Perioperative factor resonance: a new method of assessing hemostasis. Ann
concentrate therapy. Br J Anaesth 2013;111 Suppl 1: Biomed Eng 2016;44:1405–24.
i35-49. 87. Ferrante EA, Blasier KR, Givens TB, et al. A novel
76. Da Luz LT, Nascimento B, Shankarakutty AK, Rizoli S, device for the evaluation of hemostatic function in
Adhikari NK. Effect of thromboelastography (TEG(R)) critical care settings. Anesth Analg 2016;123:1372–9.
and rotational thromboelastometry (ROTEM(R)) on 88. Huffmyer JL, Fernandez LG, Haghighian C, Terkawi
diagnosis of coagulopathy, transfusion guidance and AS, Groves DS. Comparison of SEER sonorheometry
mortality in trauma: descriptive systematic review. Crit with rotational thromboelastometry and laboratory
Care 2014;18:518. parameters in cardiac surgery. Anesth Analg
77. Rahe-Meyer N, Levy JH, Mazer CD, et al. Randomized 2016;123:1390–9.
evaluation of fibrinogen vs placebo in complex 89. Naik BI, Durieux ME, Knisely A, et al. SEER
cardiovascular surgery (REPLACE): a double-blind sonorheometry versus rotational thromboelastometry
phase III study of haemostatic therapy. Br J Anaesth in large volume blood loss spine surgery. Anesth Analg
2016;117:41–51. 2016;123:1380–9.
78. Bilecen S, de Groot JA, Kalkman CJ, et al. Effect of 90. Reynolds PS, Middleton P, McCarthy H, Spiess BD. A
fibrinogen concentrate on intraoperative blood loss comparison of a new ultrasound-based whole blood
among patients with intraoperative bleeding during viscoelastic test (SEER Sonorheometry) versus
high-risk cardiac surgery: a randomized clinical trial. thromboelastography in cardiac surgery. Anesth Analg
JAMA 2017;317:738–47. 2016;123:1400–7.
79. Afshari A, Wikkelso A, Brok J, Moller AM, Wetterslev J. 91. Bhatt DL. Role of antiplatelet therapy across the
Thromboelastography (TEG) or thromboelastometry spectrum of patients with coronary artery disease. Am J
(ROTEM) to monitor haemotherapy versus usual care Cardiol 2009;103:11A–9A.
in patients with massive transfusion. Cochrane Database
Syst Rev 2011:CD007871. 92. Jennings LK. Role of platelets in atherothrombosis. Am
J Cardiol 2009;103:4A–10A.
80. Deppe AC, Weber C, Zimmermann J, et al. Point-of-
care thromboelastography/thromboelastometry-based 93. Breet NJ, van Werkum JW, Bouman HJ, et al. High on-
coagulation management in cardiac surgery: a meta- aspirin platelet reactivity as measured with
analysis of 8332 patients. J Surg Res 2016;203:424–33. aggregation-based, cyclooxygenase-1 inhibition
sensitive platelet function tests is associated with the
81. Laursen TH, Meyer MAS, Meyer ASP, et al. occurrence of atherothrombotic events. J Thromb
Thromboelastography early amplitudes in bleeding Haemost 2010;8:2140–8.
and coagulopathic trauma patients: results from a
multicenter study. J Trauma Acute Care Surg 94. Breet NJ, van Werkum JW, Bouman HJ, Ten Berg JM,
2018;84:334–41. Hackeng CM. Platelet function tests for the monitoring
of P2Y12 inhibitors. Expert Opin Med Diagn
82. Dirkmann D, Gorlinger K, Dusse F, Kottenberg E, 2010;4:251–65.
Peters J. Early thromboelastometric variables reliably
predict maximum clot firmness in patients undergoing 95. Orlov D, McCluskey SA, Selby R, et al. Platelet
cardiac surgery: a step towards earlier decision making. dysfunction as measured by a point-of-care monitor is
Acta Anaesthesiol Scand 2013;57:594–603. an independent predictor of high blood loss in cardiac
surgery. Anesth Analg 2014;118:257–63.
83. Song JG, Jeong SM, Jun IG, Lee HM, Hwang GS. Five-
minute parameter of thromboelastometry is sufficient 96. Zaffar N, Joseph A, Mazer CD, et al. The rationale for
to detect thrombocytopenia and hypofibrinogenaemia platelet transfusion during cardiopulmonary bypass:
in patients undergoing liver transplantation. Br J an observational study. Can J Anaesth 2013;60:345–54.
Anaesth 2014;112:290–7. 97. Wohlauer MV, Moore EE, Thomas S, et al. Early platelet
84. Rajkumar V, Kumar B, Dutta V, Mishra AK, Puri GD. dysfunction: an unrecognized role in the acute
Utility of Sonoclot in prediction of postoperative coagulopathy of trauma. J Am Coll Surg 2012;214:739–46.
136
98. Nekludov M, Bellander BM, Blomback M, Wallen clopidogrel to prasugrel in Japanese patients with
HN. Platelet dysfunction in patients with severe stable coronary artery disease. Circ J 2015;79:2439–
traumatic brain injury. J Neurotrauma 2007;24:1699– 44.
706. 111. Kakouros N, Kickler TS, Laws KM, Rade JJ.
99. Bolliger D, Tanaka KA. Point-of-care coagulation Hematocrit alters VerifyNow P2Y12 assay results
testing in cardiac surgery. Semin Thromb Hemost independently of intrinsic platelet reactivity and
2017;43:386–96. clopidogrel responsiveness. J Thromb Haemost
100. Tanaka KA, Bolliger D, Guzzetta NA. Clinical and 2013;11:1814–22.
practical aspects of restoring thrombin generation in 112. Rosengart TK, Romeiser JL, White LJ, et al. Platelet
acute coagulopathic bleeding. Anesth Analg activity measured by a rapid turnaround assay
2017;124:701. identifies coronary artery bypass grafting patients at
101. Malm CJ, Hansson EC, Akesson J, et al. Preoperative increased risk for bleeding and transfusion
platelet function predicts perioperative bleeding complications after clopidogrel administration. J
complications in ticagrelor-treated cardiac surgery Thorac Cardiovasc Surg 2013;146:1259–66, 66 e1;
patients: a prospective observational study. Br J discussion 66.
Anaesth 2016;117:309–15. 113. Berger PB, Kirchner HL, Wagner ES, et al. Does
102. Mahla E, Prueller F, Farzi S, et al. Does platelet preoperative platelet function predict bleeding in
reactivity predict bleeding in patients needing urgent patients undergoing off pump coronary artery bypass
coronary artery bypass grafting during dual surgery? J Interv Cardiol 2015;28:223–32.
antiplatelet therapy? Ann Thorac Surg 114. Carcao MD, Blanchette VS, Stephens D, et al.
2016;102:2010–7. Assessment of thrombocytopenic disorders using the
103. Ranucci M, Colella D, Baryshnikova E, Di Dedda U, platelet function analyzer (PFA-100). Br J Haematol
Surgical, Clinical Outcome Research Group. Effect of 2002;117:961–4.
preoperative P2Y12 and thrombin platelet receptor 115. Cho YU, Jang S, Park CJ, Chi HS. Variables that affect
inhibition on bleeding after cardiac surgery. Br J platelet function analyzer-100 (PFA-100) closure
Anaesth 2014;113:970–6. times and establishment of reference intervals in
104. Kane LC, Woodward CS, Husain SA, Frei-Jones MJ. Korean adults. Ann Clin Lab Sci 2008;38:247–53.
Thromboelastography–does it impact blood 116. Fattorutto M, Pradier O, Schmartz D, Ickx B, Barvais
component transfusion in pediatric heart surgery? J L. Does the platelet function analyser (PFA-100)
Surg Res 2016;200:21–7. predict blood loss after cardiopulmonary bypass? Br J
105. Karkouti K, McCluskey SA, Callum J, et al. Evaluation Anaesth 2003;90:692–3.
of a novel transfusion algorithm employing point-of- 117. Ng KF, Lawmin JC, Tsang SF, Tang WM, Chiu KY.
care coagulation assays in cardiac surgery: a Value of a single preoperative PFA-100 measurement
retrospective cohort study with interrupted time- in assessing the risk of bleeding in patients taking
series analysis. Anesthesiology 2015;122:560–70. cyclooxygenase inhibitors and undergoing total knee
106. Ranucci M, Baryshnikova E, Pistuddi V, et al. The replacement. Br J Anaesth 2009;102:779–84.
effectiveness of 10 years of interventions to control 118. Bolliger D, Seeberger MD, Tanaka KA, et al. Pre-
postoperative bleeding in adult cardiac surgery. analytical effects of pneumatic tube transport on
Interact Cardiovasc Thorac Surg 2017;24:196–202. impedance platelet aggregometry. Platelets
107. Schochl H, Nienaber U, Hofer G, et al. Goal-directed 2009;20:458–65.
coagulation management of major trauma patients 119. Paniccia R, Antonucci E, Maggini N, et al. Assessment
using thromboelastometry (ROTEM)-guided of platelet function on whole blood by multiple
administration of fibrinogen concentrate and electrode aggregometry in high-risk patients with
prothrombin complex concentrate. Crit Care 2010;14: coronary artery disease receiving antiplatelet therapy.
R55. Am J Clin Pathol 2009;131:834–42.
108. Theusinger OM, Stein P, Levy JH. Point of care and 120. Mazzeffi MA, Lee K, Taylor B, Tanaka KA.
factor concentrate-based coagulation algorithms. Perioperative management and monitoring of
Transfus Med Hemother 2015;42:115–21. antiplatelet agents: a focused review on aspirin and
109. Born GV, Cross MJ. Effect of adenosine diphosphate P2Y12 inhibitors. Korean J Anesthesiol 2017;70:379–
on the concentration of platelets in circulating blood. 89.
Nature 1963;197:974–6. 121. Bolliger D, Filipovic M, Matt P, et al. Reduced aspirin
110. Nishi T, Ariyoshi N, Nakayama T, et al. Increased responsiveness as assessed by impedance
platelet inhibition after switching from maintenance aggregometry is not associated with adverse outcome
after cardiac surgery in a small low-risk cohort. testing and bleeding after coronary artery surgery.
Platelets 2016;27:254–61. Scand Cardiovasc J 2012;46:32–8.
122. Kaiser AF, Neubauer H, Franken CC, et al. Which is 128. Sambu N, Hobson A, Curzen N. “Short”
the best anticoagulant for whole blood aggregometry thromboelastography as a test of platelet reactivity in
platelet function testing? Comparison of six response to antiplatelet therapy: validation and
anticoagulants and diverse storage conditions. reproducibility. Platelets 2011;22:210–6.
Platelets 2012;23:359–67. 129. Craft RM, Chavez JJ, Snider CC, Muenchen RA,
123. Rubak P, Villadsen K, Hvas AM. Reference intervals Carroll RC. Comparison of modified
for platelet aggregation assessed by multiple electrode thrombelastograph and plateletworks whole blood
platelet aggregometry. Thromb Res 2012;130:420–3. assays to optical platelet aggregation for monitoring
124. Stissing T, Dridi NP, Ostrowski SR, Bochsen L, reversal of clopidogrel inhibition in elective surgery
Johansson PI. The influence of low platelet count on patients. J Lab Clin Med 2005;145:309–15.
whole blood aggregometry assessed by multiplate. 130. Tantry US, Bliden KP, Gurbel PA. Overestimation
Clin Appl Thromb Hemost 2011;17:E211–7. of platelet aspirin resistance detection by
125. Rahe-Meyer N, Winterhalter M, Boden A, et al. thrombelastograph platelet mapping and
Platelet concentrates transfusion in cardiac surgery validation by conventional aggregometry using
and platelet function assessment by multiple arachidonic acid stimulation. J Am Coll Cardiol
electrode aggregometry. Acta Anaesthesiol Scand 2005;46:1705–9.
2009;53:168–75. 131. Fabbro M, 2nd, Winkler AM, Levy JH. Technology: is
126. Carroll RC, Chavez JJ, Snider CC, Meyer DS, there sufficient evidence to change practice in point-
Muenchen RA. Correlation of perioperative platelet of-care management of coagulopathy? J Cardiothorac
function and coagulation tests with bleeding after Vasc Anesth 2017;31:1849–56.
cardiopulmonary bypass surgery. J Lab Clin Med 132. Ranucci M, Baryshnikova E, Crapelli GB, et al.
2006;147:197–204. Electric impedance platelet aggregometry in cardiac
127. Dalen M, van der Linden J, Lindvall G, Ivert T. surgery patients: a comparative study of two
Correlation between point-of-care platelet function technologies. Platelets 2016;27:185–90.
138
Chapter
Assessment of Intraoperative Blood Loss
14 Kyle James Riley and Daniel Katz
Overview surgical sponges, clot size and shape, and suction

canister volumes or visual assessment of the blood
Accurate intraoperative estimation of blood loss is
in the surgical field
important to guide fluid administration, assess the
• The use of more accurate volumetric and
need to transfuse packed red blood cells and other
gravimetric measurement techniques may reduce
blood products, and anticipate the severity of post-
the risk of underestimating blood loss
operative anemia. Avoidance of transfusion is a gen-
eral goal as blood products are expensive, there is a In the survey of consultant expert and ASA member
limited and sometimes inadequate supply, and blood opinions, both groups strongly agreed that there is a
transfusion is associated with increased risk for a need to:8
variety of postoperative complications.1–6 Despite • Assess the presence of bleeding by a periodic
the need to accurately quantify surgical blood loss visual assessment of the surgical field
for these purposes, surprisingly there is no gold stan-
• Perform quantitative measurement of blood loss,
dard. A variety of methods exist, each with its own
including evaluation of surgical sponges and
advantages and disadvantages.
surgical drains, using standard methods
Monitoring Standard and Practice AABB Guidelines

Guidelines The AABB, formerly known as the “American
Association of Blood Banks,” is an international,
American Society of Anesthesiologists not-for-profit association of individuals and institu-
Standards and Guidelines tions involved in the fields of cellular therapies and
transfusion medicine.9,10 The AABB clinical guide-
The American Society of Anesthesiologists (ASA)
lines for red blood cell (RBC) transfusion thresholds,
Standards for Basic Anesthetic Monitoring sets the
based on a review of randomized clinical trials, made
most basic standards of anesthetic monitoring.
recommendations for appropriate use of a restrictive
These standards set the requirement for continual
transfusion strategy, but provided no recommenda-
(“repeated regularly and frequently in steady rapid
tions on how to estimate blood loss and only weakly
succession”) monitoring of oxygenation, ventilation,
recommended using hemoglobin concentration (Hb)
circulation, and temperature.7 ASA practice guide-
to guide transfusion decision-making.11–14
lines for perioperative blood management have been
developed to assist in decisions regarding periopera-
tive blood transfusions and adjuvant therapies.8 They Blood Loss Assessment Methods
include recommendations to optimize conditions to
reduce potential blood loss, and advocate the use of Visual Assessment (Subjective
transfusion algorithms, specialized testing, and Measurement)
restrictive strategies. With regard to determining the In the latest ASA Practice Guidelines survey,8 91.6%
extent of blood loss, however, the guidelines simply of the consultant group and 92.7% of the ASA
state the following:8 membership responded as agreeing or strongly
• The literature is inadequate to evaluate the efficacy agreeing with the need to “periodically perform a
of measuring blood loss through observation of visual assessment of the surgical field with the
surgeon to assess the presence of excessive microvas- For blood collected in suction canisters:
cular (i.e., coagulopathy) or surgical bleeding.” Because
1. Gently remix effluent
of its convenience and efficiency, or because other meth-
ods may not be available or practical, a visual assessment 2. Use a plasma/low Hb analyzer or plasma
of blood loss by the anesthesiologist and/or surgeon is spectrometer to determine the canister Hb
commonly used and often referred to as “subjective (HbCanister)
assessment.”15,16 The literature, however, is replete 3. Determine mass of suction canister liquid
with studies that demonstrate the inaccuracies of visual (MCanister in grams)
blood loss assessment.1,17–20 There is a tendency to 4. Volume of canister liquid is:
overestimate blood loss volume when blood loss is low VCanister = MCanister /(100 g/dL)
and underestimate blood loss volume when blood loss is 5. MCanisterHb = HbCanister × VCanister
high,19 and the degree of inaccuracy of blood loss esti- 6. VCanisterBloodLoss = MCanisterHb/HbPreop
mation increases as blood loss itself increases.20 The total volume of blood loss is:
It was demonstrated that a brief didactic training
course can significantly improve the visual blood loss VTotal = VSpongeBloodLoss + VCanisterBloodLoss
assessment.19,21 The improvement after training,
however, was transient, as there was a significant Studies may also determine a baseline assay yield
decay in estimation accuracy by 9 months after train- rate by applying banked blood of a known quan-
ing was completed. The median error in blood loss tity to sponges and determining an expected
estimation was −47.8% pre-training, improved to recovery rate for the specific extraction and ana-
−13.5% immediately following training, but worsened lysis procedure being used.1,22 These yield rates
to −34.6% at the 9-month follow-up.21 (YR) range from 89.5% using manual compres-
sion, to 98.99% using a centrifuge to extract the
rinse liquid.
Assay Method Applying a YR, the equation for total volume of
The assay method, also called “photometric analy- blood loss becomes:
sis,” is a set of reliable techniques for measuring
blood loss.16 It is often used in studies as the refer- VTotal = (VSpongeBloodLoss /YR) + VCanisterBloodLoss
ence standard against which other blood loss deter-
mination methods are compared. The basic where YR is in decimal % (e.g., 0.895 used for 89.5%).
procedure is as follows, with all Hb values always
expressed in g/dL:1 Gravimetric Blood Loss Determination
Determine the patient’s preoperative Hb
Because of issues with visual estimation, the use of the
(Hbpreop). For surgical sponges:
gravimetric method, also called the “quantitative
1. Collect all soiled sponges method,” is recommended to determine blood loss
2. Rinse each sponge with saline solution and by national organizations, including the Association
compress to remove and collect all rinse liquid; of Women’s Health, Obstetric and Neonatal Nurses
repeat 3–4 times (AWHONN), the California Maternal Quality Care
3. Use a plasma/low Hb photometer or plasma Collaborative (CMQCC), and the Council on Patient
spectrometer to determine rinse Hb (HbRinse) Safety in Women’s Healthcare.1
4. Determine mass of product rinse liquid (MRinse) in In a study analyzing the accuracy of different
grams blood loss determination methods during cesarean
5. Volume of rinse liquid is: delivery procedures, Doctorvaladan and colleagues1
VRinse = MRinse/100 g/dL, where 100 g/dL is density used the following gravimetric procedure and formu-
used for liquids las to determine blood loss.
6. MSpongeHb = HbRinse × VRinse, where MSpongeHb is
the mass of Hb loss in grams Prior to procedure:
7. VSpongeBloodLoss = MSpongeHb/HbPreop, where 1. Record weights of empty suction canisters
VSpongeBloodLoss is the volume of sponge blood loss 2. Record weights of 3 packs of 5 sponges each and
in dL determine average sponge weight (MDry)
140
During procedure: The application then uses the stored dry weights
1. At uterine incision, record the canister volume and soiled weights to calculate and track blood
using graduated markings (VCanister1) loss.24,25
2. After all amniotic fluid is aspirated, record the
second measurement of canister volume Blood Loss Formulas, Hemoglobin
(VCanister2)
3. VAmniotic = VCanister2 – VCanister1
Monitoring, and Hematocrit
4. Record total amount of irrigation fluid used Monitoring
(VIrrigation)
Blood Loss Formula
Studies dating back to 197426–29 developed mathema-
Post-procedure:
tical formulas that use changes in hematocrit (Hct) to
1. Record the weight of each sponge (MWet)
calculate an estimated blood loss or an allowable
2. Record the weight of each canister effluent blood loss limit during surgical procedures to deter-
(MCanisterEffluent) mine transfusion triggers. The theoretical dilution
3. Quantitative blood loss (QBL) is calculated using equation that models the problem of isovolemic
the following formulas (fluid density is 1.0 g/mL): hemodilution is the differential equation:27
VSpongeQBL = (MWet – MDry)/(1.0 g/mL) dHct/Hct = dVBloodLoss/V
VCanisterQBL = MCanisterEffluent/(1.0 g/mL) where Hct = hematocrit, dHct is the change in Hct,
dVBloodLoss = blood loss (change in blood volume),
VTotalQBL = ΣVSpongeQBL + ΣVCanisterQBL – VAmniotic
and V = patient’s total blood volume.
– VIrrigation
Integration over the limits from initial to final
VTotalQBL is the total quantitative blood loss calculated yields:
as the sum of all the quantitative blood loss of each
VBloodLoss = V × [ln(HctInitial ) – ln(HctFinal )]
sponge, plus the sum of the quantitative blood loss of
each canister, minus the volume of amniotic fluid and or
minus the volume of irrigation fluid used.
Limitations of the gravimetric method include the VBloodLoss = V × [ln(HctInitial/HctFinal )]
following:1,16,22,23
with ln(X) being the natural log of X.
• It is time-consuming and impractical for real-time Bourke and Smith27 replaced the natural log func-
intraoperative use tion by an approximation using a Taylor series with
• The presence of non-sanguineous fluids and other the higher-ordered terms dropped:
substances, including amniotic fluid, ascites,
saline, and other tissues, decrease its accuracy VBloodLoss = V × (HctInitial – HctFinal ) × (3 – HctAvg )
• The assumption that the patient’s blood
where HctAvg = (HctInitial + HctFinal )/2
hemoglobin concentration stays constant is
incorrect if administered intravenous (IV) fluids Gross29 proposed the following simpler equation
which progressively dilute the patient’s blood23 that more closely approximates the plot of the loga-
New technology has been introduced that may rithmic formula:
help reduce the effort and time associated with the
VBloodLoss = V × ((HInitial – HFinal)/HAvg )
gravimetric blood loss method. The Triton L&D
(Gauss Surgical Inc., Los Altos, CA) is a blood loss where H can be either the Hct or Hb. Hct-based blood
monitor for labor and delivery that uses a tablet-based loss equations have also been used to evaluate treat-
application containing a customizable list of blood ments to reduce bleeding. One such study performed
absorbent items with pre-determined dry weights. A by Barrachina and colleagues assessed total blood loss
scale, connected to the tablet using Bluetooth, is used 48 hours after hip replacement surgery with different
to weigh individual or batches of blood-soiled items. regimens of tranexamic acid (TXA).2 This study used
the formula proposed by Camarasa et al.30 to calculate calculated as the number of RBCs per the mean cell
blood loss volumes across the different TXA volume, or simply as the product of the Hb × 2.941.
treatments. Maslow and colleagues34 compared three POC
In a comparison of hematocrit-based blood loss testing devices to the hospital reference analyzer for
equations, Lopez-Picado et al.31 determined whether cardiac surgery patients. The devices were the GEM
applying different blood loss formulas to the data 4000 (Instrumentation Laboratory, Bedford, MA), the
from Barrachina’s study2 would have affected the i-STAT (Abbot Point of Care, Princeton, NJ), and the
findings. They compared the four most frequently Radical-7 Pulse CO-Oximeter (Masimo, Irvine, CA).
used formulas of Bourke and Smith,27 Gross,29
Mercuriali et al.32 and Camarasa,30 as well as their Device Overview
own equation that combined Camarasa’s blood loss • The GEM 4000 uses CO-oximetry measurement
formula with the International Council for the absorption of multi-wavelengths of light by the
Standardization in Haematology’s (ICSH) total different Hb structures. The Hct value is calculated
blood volume calculation (see Appendix 1 for a as the Hb concentration × 3. Cellular reflectors
further description of the equations). In evaluating that cause light to scatter may affect the results of
the five blood loss equations, Lopez-Picado et al.31 CO-oximetry based devices.
found that: • The i-STAT device uses conductivity to measure
the Hct based on an electrical current; a higher Hct
• Use of the formulas based on anthropometric and
results in less current passing through the blood.
laboratory parameters to calculate patient blood
The Hct × 0.34 is used to calculate the Hb
loss would not have affected the clinical
concentration. Hemodilution, reduced serum
conclusions of the Barrachina et al. study
protein levels, and heparinization affect sample
• Use of Gross’s equation would negate the
conductivity and hence the accuracy of
significance of some of the Barrachina et al. study
conductivity-calculated Hct.
findings, probably because it does not take
• The Radical-7 Pulse CO-Oximeter uses a non-
transfusions into account
invasive percutaneous sensor that measures Hb
• There were no significant differences in the total
using multiple wavelengths of light. It provides
patient blood volumes calculated using the three
continuous, real-time monitoring of Hb
total blood volume formulas
concentrations. Digital perfusion, sensor position,
• There was a very low level of agreement of their movement, and external light interference can
chosen reference method and the other formulas affect the sensor results.
The authors concluded that, in some cases, the
differences in the blood loss amounts calculated by Device Performance
the different formulas could affect clinical decisions. • The GEM 4000 data correlated with the lab
standard (r = 0.97), but it had a positive bias
(overestimation) that was stable over the phases of
Hemoglobin and Hematocrit Monitoring surgery and range of Hb/Hct data.
Hb and Hct are diagnostic parameters that can be • The i-STAT data correlated with the lab standard
obtained from an automated hematology analyzer in (r = 0.97) and had the lowest mean bias, but that
a satellite clinical laboratory or from a point-of-care bias switched from overestimation before
(POC) device.33,34 The determination of Hb in a clin- cardiopulmonary bypass to underestimation after
ical laboratory using an automated analyzer is con- cardiopulmonary bypass. Also, the bias was
sidered the gold standard for Hb/Hct measurement, positive at higher Hb values and negative at lower
the major drawback being the delay in obtaining test Hb values – so higher Hb concentrations were
results.35 These machines lyse the RBCs and then use overestimated and lower ones underestimated.
spectrophotometry (CO-oximetry) to determine the • The Radical-7 Pulse CO-Oximeter data correlated
concentration of Hb. The wavelengths of light used with the lab standard (r = 0.84), but not as highly
are not always stated by the manufacturers, but as the other two devices. It consistently
525 nm is utilized by at least one of them. Once the overestimated the lab standard, but with a smaller
Hb concentration is determined, the Hct may be bias prior to surgery that increased during surgery
142
and peaked after protamine administration. The SpHb monitoring and started receiving intermittent
bias was positive for all Hb values and the positive blood sampling starting when their EBL reached 15%
bias increased as Hb values decreased – so lower or more. In both groups, a transfusion was initiated
Hb concentrations were overestimated to a larger when Hb or SpHb was ≤ 10 g/dL and continued until
degree than higher Hb concentrations. the EBL was replaced and the Hb was > 10 g/dL. The
control and SpHb-monitored groups were similar in
Transfusion Decisions surgical procedures and demographics (except
In a small study of 24 consecutive elective adult car- that the SpHb group was lower in weight).
diac surgery patients,34 if these POC devices had been Preoperatively, the only significant clinical group
used in conjunction with a restricted transfusion trig- difference was that the SpHb group had an average
ger value of Hb 7 g/dL or Hct 21, then: baseline Hb about 1 g/dL lower than the control
group.
• Using the GEM 4000, no patients would have
This study found no difference in the percentage
undergone an unnecessary transfusion
of patients transfused; however, in the SpHb group,
• Using the i-STAT, three patients would have
transfused patients received an average of 1.6 units
undergone an unnecessary transfusion
fewer units (p < 0.01), the percent of patients receiving
• Using the Radical-7, no patients would have more than 3 RBC units was reduced from 73% to 32%
undergone an unnecessary transfusion (p < 0.01), the post-transfusion Hb values averaged
Continuous Real-Time Hemoglobin Monitoring 0.5 g/dL less (p < 0.01), and the average delay in time
to transfusion was reduced from 50.2 to 9.2 minutes
Two studies looked at the effect on blood transfusions (p < 0.001). The study also found that in the SpHb
of continuous and noninvasive spectrophotometry group the Radical-7 Pulse CO-Oximeter values had a
hemoglobin (SpHb) monitoring during surgery very small bias compared to the laboratory Hb values
using the Radical-7 Pulse CO-Oximeter. (0.0 ± 0.8 g/dL) and the limits of agreement were −1.6
Ehrenfeld and colleagues6 conducted a study in to 1.5 g/dL.
elective orthopedic surgery patients with a moderate These studies demonstrated that continuous real-
overall risk of requiring transfusion, randomized to time monitoring of Hb concentrations can affect
have continuous intraoperative SpHb monitoring intraoperative blood transfusion decision-making,
versus standard care without SpHb monitoring; reduce decision delay, and decrease the number of
there was no standardized transfusion protocol. In patients and total quantity of blood transfused.
the SpHb group, 0.6% (1 out of 170) of patients Everything else being equal, restrictive transfusion
received transfusions, while in the standard care strategies require better information about the
group, 4.5% (7 out of 157) of patients received trans- patient’s Hb in addition to blood loss. Continuous
fusions (risk difference is −0.04; 95% CI: −0.007, monitoring of SpHb concentrations is a promising
−0.004). The groups did not differ with respect to method that may promote more timely decision-
preoperative patient characteristics, outcomes after making.
surgery, or the incidence of postoperative
complications at 28 days. The only effect of SpHb
monitoring in this study was to reduce the rate of Imaging Analysis
transfusions by 87%, from 4.5% to 0.6%, a 3.9% abso- The Triton OR System (Gauss Surgical Inc., Los Altos,
lute reduction. But this was a relatively small study in CA) uses photo-imaging technology and cloud-based
which even the control group transfusion rate was machine-learning (ML) algorithms to determine
relatively low. hemoglobin mass either on sponges or in suction
Awada and colleagues5 evaluated the impact of canisters. The images are taken using a tablet compu-
continuous SpHb monitoring by Radical-7 Pulse ter, encrypted, and transferred to a remote server. For
CO-Oximeter on blood transfusions in 86 high blood-soaked sponges, Gauss Feature Extraction
blood-loss neurosurgery patients in a prospective Technology, similar to facial recognition software,
cohort study. Information recorded included esti- processes the photo by identifying relevant areas of
mated blood loss (EBL), Hb, and SpHb values, units the image, makes adjustments for differences in light-
transfused, timing of each blood draw, and the start of ing conditions, and filters out the effects of non-san-
each transfusion. The control group did not have guineous fluids. Proprietary ML computational
models then determine the mass of the Hb from the hemorrhage was identified more frequently by the
image and the result is signalled back to the device Triton system than by visual estimation and that,
within seconds.22–23 possibly due to earlier recognition and treatment,
Studies have shown that the Triton OR System use of the device reduced blood transfusions.
data strongly correlate with an R= 0.91 to 0.95 for
Hb and estimated blood loss volumes to multiple Conclusion
reference methods.1,15,22–23,36 It also demonstrated
Table 14.1 summarizes the major advantages and dis-
low levels of bias in blood loss estimates by sponge
advantages of each methodology used to track or
Hb measurement (.01 to 0.7 g Hb per sponge),22–23
quantify blood loss.
and suction canister content (50 mL).1 In all of the
While an accurate assessment of Hb/Hct and
blood loss determination comparison studies, the
intraoperative blood loss is critical in optimization
Triton OR system performed significantly better
of resuscitation and transfusion, their determination
than visual and gravimetric estimations of blood
in the operating room environment in real time
loss. It generally had a higher correlation with the
remains challenging. Noninvasive strategies are
studies’ reference assay methods, lower biases, and a
ideal, but all currently available devices suffer from a
narrower range of biases.1,15,22,36
variety of practical and statistical problems. A combi-
In a retrospective analysis of cesarean delivery
nation of these strategies seems to significantly reduce
data,37 clinical outcomes were evaluated for 2,025
unnecessary transfusions in select populations, and
patients whose blood loss was visually determined by
when used by an educated provider with a support
consensus between obstetrician and anesthesiologist
system or decision tree, can affect outcomes.
at procedure conclusion (traditional group), and 756
patients whose blood loss was determined by scan-
ning all surgical sponges as the procedure progressed Appendix I
and the surgical canisters, typically at procedure end, Source and overview of the five blood loss calculation
using the Triton OR system (device group). All methods used by Lopez-Picado et al.:31
patients were pre-procedurally evaluated for postpar-
tum hemorrhage (PPH) risk as being low, medium or 1. Bourke and Smith:27 this method uses the
high, and separate analyses of outcome measures were approximation for natural log with the product of 3
performed for each PPH risk group. Overall, the char- minus the mean hematocrit value. Applied Nadler’s
acteristics of traditional and device groups were simi- formula for total blood volume of the patient.
lar, with the device group having more multiparous 2. Gross:29 this method uses the formula
(78.7% versus 67.9%; p < 0.01) and slightly older simplification that calculates blood loss utilizing
patients (32.2 versus 31.7 years of age, p < 0.01). The the initial hematocrit, final hematocrit, and the
average measured blood loss, both overall and for the mean of these two values. Volumes transfused are
high PPH risk group, was lower in the device group at not taken into account. Applied Moore’s formula
555.8 mL and 622.9 mL respectively than in the tradi- for total blood volume of the patient.
tional group at 662.1 mL and 683.2 mL (p < 0.0001 for 3. Mercuriali et al.:32 this method uses the
both comparisons). Blood loss greater than 1000 mL hematocrit values preoperatively and at
was more frequently identified in the device group 5 days postoperatively and estimates the
than the traditional group (14.1% versus 3.5%, p < volume of red cells. Takes into account the
0.0001). Additionally, the blood loss measurement in total volume of blood transfused. Applied
the device group demonstrated higher correlation to Nadler’s formula for total blood volume
the PPH risk scores than the traditional group (linear of the patient.
trend difference p = 0.0024). The percent of patients 4. Camarasa et al.:30 this formula calculates blood
receiving transfusions in both groups was not signifi- loss using the preoperative hematocrit value and
cantly different, but the device group had fewer units the hematocrit at the time of performing the
transfused (1.83 versus 2.56, p < 0.038). The device calculation. It takes into consideration
group patients also had a shorter length of stay (aver- transfusions, differentiating between blood from
age 4.0 days versus 4.4 days, p < 0.0006). The study autologous and homologous transfusions, as well
authors concluded that clinically significant as blood recovery systems. Applied a simplified
144
Table 14.1 Major advantages/disadvantages of different technology to monitor or quantify blood loss.
Methodology Advantages Disadvantages

Assay Method high accuracy expensive
performed only post-procedure
involves a laboratory procedure
wide adoption is not practical
Visual Estimation easy low accuracy
perform as needed high variability
Gravimetric quantitative relatively high effort
more practical than assay method performed only post-procedure
Automated Hematology Analyzer high accuracy invasive
perform as needed slow turnaround time
does not measure losses
measures Hb/Hct at one point in time only
Invasive POC Hb/Hct Analyzer (e.g. good accuracy invasive
GEM 4000, i-STAT) perform as needed accuracy can be situational (bias different
at high and low end of spectrum)
Noninvasive Hb analyzer (Radical-7) good accuracy accuracy can be situational (bias different
perform as needed at high and low end of spectrum)
noninvasive
real time
Hb/Hct Formulas moderately easy requires total blood volume estimate
perform as needed no gold standard formula
inherits the accuracy and inaccuracy of the
required Hb/Hct determination
Imaging Analysis (Triton) easy small positive bias over the assay method
near real time limited to sponges and canisters
highly accurate
low bias/variability
version of Moore’s formula that uses weight and 2. Nadler:

sex to calculate the total blood volume of the i. Women: ETBVðmlÞ ¼ 183 þ ð0:000356
patient. heightÞ3 þ ð33 weightÞ
5. Lopez-Picado et al.:31 this method uses Camarasa’s ii. Men: ETBVðmlÞ ¼ 604 þ ð0:0003668
blood loss calculation but applied the ICSH total heightÞ3 þ ð32:2 weightÞ
blood volume calculation. 3. ICSH
All of the formulas include a calculation of the i. Women: ETBVðmlÞ ¼ plasma volumeðmlÞ
patient’s total blood volume based on different þ red cell volumeðmlÞ
variables as follows:31 ¼ ðweight 0:425 height 0:725 Þ ð0:007184 2:217Þ
þ ðageðyearsÞ 1:06Þ
1. Moore: weight (kg), build (values for obese, thin, ii. Men: ETBVðmlÞ ¼ plasma volumeðmlÞ
normal, muscular builds), and sex29 þ red cell volumeðmlÞ
2. Nadler: weight (kg), height (cm), and sex
¼ ðweight 0:425 height 0:725 Þ
3. ICSH: body surface area and sex ð0:007184 3:064Þ 825
Total Blood Volume and Blood Loss Formulas used in • Estimation of blood loss volume (EBLV)
Lopez-Picado et al.31
1. Bourke: EBLVðmlÞ ¼ ETBV MooreðmlÞ
• Estimation of total blood volume of the patient ðinitial hematocrit final hematocritÞ
(ETBV) ð3 mean hematocritÞ
1. Moore: 2. Gross: EBLVðmlÞ ¼ ETBV MooreðmlÞ
i. Women: ETBVðmlÞ ¼ weight 65 þ ðinitial hematocrit final hematocritÞ=
ii. Men: ETBVðmlÞ ¼ weight 70 ðmean hematocritÞ
3. Mercuriali: EBLVðmlÞ ¼ ETBV NadlerðmlÞ .org/quality-and-practice-management/practice-gui

ðinitial hematocrit hematocrit day5Þ dance-resource-documents/standards-for-basic-ane
þ transfused red cell volumeðmlÞ sthetic-monitoring accessed: 11/11/2017.
4. Camarasa: 8. Practice Guidelines for Perioperative Blood
Management. Anesthesiology 2015;122:241–75.
EBLV (ml) = ETBV Moore × (initial hematocrit – 9. www.aabb.org/ accessed: 11/12/2017
final hematocrit) + transfused red cell volume / 10. Wiegmann TL, Mintz PD. The growing role of AABB
mean hematocrit clinical practice guidelines in improving patient care.
Transfusion 2015;55:935–6.
1 U packed homologous blood = 170 ml 11. Tobian AAR, Heddle NM, Wiegmann TL, Carson JL.
1 U packed autologous blood = 140 ml Red blood cell transfusion: 2016 clinical practice
guidelines from AABB. Transfusion 2016;56:2627–30.
100 ml of recovered blood = 54 ml
12. Yazer MH, Triulzi DJ, DI S. AABB red blood cell
5. New formula: transfusion guidelines. JAMA 2016;316:1984–5.
13. Carson JL, Guyatt G, Heddle NM, et al. Clinical
EBLV (ml) = ETBV ICSH × (initial hematocrit – practice guidelines from the AABB. JAMA
final hematocrit) + transfused red cell volume / 2016;316:2025–35.
mean hematocrit 14. Carson JL, Grossman BJ, Kleinman S, et al. Red blood
cell transfusion: a clinical practice guideline from the
1 U packed homologous blood = 450 ml × hemato- AABB*. Ann Intern Med 2012;157:49–58.
crit of the blood transfused 15. Sharareh B, Woolwine S, Satish S, Abraham P,
1 U packed autologous blood = 450 ml × hemato- Schwarzkopf R. Real time intraoperative monitoring of
crit in the pre-surgical anesthesia consult blood loss with a novel tablet application. Open Orthop
J 2015;9:422–6.
16. Schorn MN. Measurement of blood loss: review of the
References literature. J Midwifery Womens Health 2010;55:20–7.
1. Doctorvaladan S, Jelks A, Hsieh E, et al. Accuracy of
17. Stafford I, Dildy GA, Clark SL, Belfort MA. Visually
blood loss measurement during cesarean delivery. Am J
estimated and calculated blood loss in vaginal and
Perinatol Reports 2017;7:e93–100.
cesarean delivery. Am J Obstet Gynecol 2008;199:519.e1-7.
2. Barrachina B, Lopez-Picado A, Remon M, et al.
18. Bose P, Regan F, Paterson-Brown S. Improving the
Tranexamic acid compared with placebo for reducing
accuracy of estimated blood loss at obstetric
total blood loss in hip replacement surgery. Anesth
haemorrhage using clinical reconstructions. BJOG
Analg 2016;122:986–95.
2006;113:919–24.
3. Frank SM, Savage WJ, Rothschild JA, et al. Variability in
19. Dildy GA, Paine AR, George NC, Velasco C.
blood and blood component utilization as assessed by
Estimating blood loss: can teaching significantly
an anesthesia information management system.
improve visual estimation? Obstet Gynecol
Anesthesiology 2012;117:99–106.
2004;104:601–6.
4. Guinn NR, Broomer BW, White W, Richardson W, Hill
20. Eipe N, Ponniah M. Perioperative blood loss
SE. Comparison of visually estimated blood loss with
assessment – How accurate? Indian J Anaesth
direct hemoglobin measurement in multilevel spine
2006;50:35–8.
surgery. Transfusion 2013;53:2790–4.
21. Toledo P, Eosakul ST, Goetz K, Wong CA, Grobman
5. Awada WN, Mohmoued MF, Radwan TM, Hussien GZ,
WA. Decay in blood loss estimation skills after web-
Elkady HW. Continuous and noninvasive hemoglobin
based didactic training. Simul Healthc 2012;7:18–21.
monitoring reduces red blood cell transfusion during
neurosurgery: a prospective cohort study. J Clin Monit 22. Holmes AA, Konig G, Ting V, et al. Clinical evaluation
Comput 2015;29:733–40. of a novel system for monitoring surgical hemoglobin
loss. Anesth Analg 2014;119:588–94.
6. Ehrenfeld JM, Henneman JP, Bulka CM, Sandberg WS.
Continuous non-invasive hemoglobin monitoring 23. Konig G, Holmes AA, Garcia R, et al. In vitro
during orthopedic surgery: a randomized trial. J Blood evaluation of a novel system for monitoring surgical
Disord Transfus 2014;5(237):1–5. hemoglobin loss. Anesth Analg 2014;119:595–600.
7. Standards for Basic Anesthetic Monitoring – 24. Gauss S. Triton L&D – Gauss Surgical, www.gausssur
American Society of Anesthesiologists., www.asahq gical.com/tritonld accessed:12/04/2019
146
25. Gauss S. Gauss Surgical Receives FDA Clearance for 32. Mercuriali F, Inghilleri G. Proposal of an algorithm to
Second Generation Triton for Real-Time Monitoring help the choice of the best transfusion strategy. Curr
of Surgical Blood Loss – Gauss Surgical., www.gauss Med Res Opin 1996;13:465–78.
surgical.com/news/2017/6/22/gauss-surgical-receives- 33. Vos JJ, Kalmar AF, Struys MM, et al. Accuracy of
fda-clearance-for-second-generation-triton-for-real-ti non-invasive measurement of haemoglobin
me-monitoring-of-surgical-blood-loss accessed:12/04/ concentration by pulse co-oximetry during steady-
2019. state and dynamic conditions in liver surgery. Br J
26. Kallos T, Smith TC. Replacement for Anaesth 2012;109:522–8.
intraoperative blood loss. Anesthesiology 34. Maslow A, Bert A, Singh A, Sweeney J. Point-of-care
1974;41:293–5. hemoglobin/hematocrit testing: comparison of
27. Bourke DL, Smith TC. Estimating allowable methodology and technology. J Cardiothorac Vasc
hemodilution. Anesthesiology 1974;41:609–11. Anesth 2016;30:352–62.
28. Ward CF, Meathe EA, Benumof JL, Trousdale F. A 35. Lamhaut L, Apriotesei R, Combes X, et al. Comparison
computer nomogram for loss replacement. of the accuracy of noninvasive hemoglobin monitoring
Anesthesiology 1980;53:S126. by spectrophotometry (SpHb) and HemoCue® with
29. Gross JB. Estimating allowable blood loss: corrected automated laboratory hemoglobin measurement.
for dilution. Anesthesiology 1983;58:277–80. Anesthesiology 2011;115:548–54.
30. Camarasa MA, Ollé G, Serra-Prat M, et al. Efficacy of 36. Konig G, Waters JH, Javidroozi M, et al. Real-time
aminocaproic, tranexamic acids in the control of evaluation of an image analysis system for monitoring
bleeding during total knee replacement: a surgical hemoglobin loss. J Clin Monit Comput
randomized clinical trial. Br J Anaesth 2006;96: 2018;32:303–10.
576–82. 37. Rubenstein A, Zamudio S, Al-Khan A, et al. Clinical
31. Lopez-Picado A, Albinarrate A, Barrachina B. experience with the implementation of accurate
Determination of perioperative blood loss: measurement of blood loss during cesarean delivery:
accuracy or approximation? Anesth Analg influences on hemorrhage recognition and allogeneic
2017;125:280–6. transfusion. Am J Perinatol 2018;35:655–9.
Chapter
Respiratory Monitoring in Low-Intensity
15 Settings
Andrew B. Leibowitz and Adel Bassily-Marcus
Introduction saturation (SpO2) less than 85% was 27.1% in the no

Respiratory failure is a common occurrence in low- OSA group, 39.5% in the known OSA group, and
intensity hospital settings (i.e., settings other than 39.9% in the positive OSA screening group.
operating rooms, post-anesthesia care units Highlighting the risk of undiagnosed OSA, the posi-
(PACUs), or intensive care units (ICUs)), and tive OSA screening group had significantly higher
accounts for a large percentage of intra-hospital trans- rates of reintubation, ICU admission, prolonged hos-
fers to ICUs. Traditional monitoring of respiratory pitalization, and all-cause 30-day mortality, compared
function in low-intensity settings is challenging and to the other groups.
has limitations. It relies on clinical observations of Another group of patients who are at high risk for
skin color appearance and counting the respiratory respiratory compromise in low-intensity hospital set-
rate, both of which are unreliable, time-consuming, tings are those treated with opioids. An American
and only performed intermittently.1 It is difficult to Society of Anesthesiologists (ASA) Closed Claims
detect cyanosis in dark-skinned individuals and ane- database study that reviewed 92 malpractice claims
mic patients. Cyanosis is a very late sign of respiratory related to postoperative opioid-induced respiratory
insufficiency, usually only developing just prior to a depression determined that 97% of the events were
cardiopulmonary arrest when there are limited preventable had there been better monitoring and
opportunities for intervening. Additionally, clinical response.4 In the PRODIGY (PRediction of Opioid-
observation cannot determine the depth of respira- induced respiratory Depression In patients moni-
tion or the underlying pathology interfering with tored by capnoGraphyY) study 1,496 adult patients
effective gas exchange. who were on a general hospital ward and received
Patients with morbid obesity and obstructive sleep parenteral opioid therapy for post-surgical and non-
apnea (OSA) are especially at risk for respiratory surgical pain as the primary analgesic modality were
compromise in low-intensity hospital settings. In a monitored with capnography and pulse oximetry.5
meta-analysis of 17 studies involving 7,162 postopera- Impressively, 46% of patients experienced opioid-
tive patients, OSA was associated with an increased induced respiratory depression. These data indicate
risk of respiratory failure (odds ratio 2.42) and cardiac that there is tremendous unmet need for better mon-
events (odds ratio is 1.63).2 The prevalence of known itoring of respiratory function in low-intensity settings.
OSA in hospitalized patients has increased signifi- Respiratory function monitors that are commonly
cantly over the years, yet it remains underdiagnosed used in procedural and ICU settings are designed for
despite widespread use of screening tools such as the use in immobile patients for brief periods of time in
STOP-BANG Questionnaire, Berlin Questionnaire, areas where the staff-to-patient ratio is high, such as
and Sleep Apnea Clinical Score. A study of 28,921 PACUs and ICUs where the ratio is 1:1 or 1:2, or
nonemergent inpatient surgical patients who were “stepdown” and “intermediate care units” where the
classified as having no OSA, known OSA, or positive ratio usually does not exceed 1:4. While these moni-
screening for OSA (i.e., patients who denied a history tors may be used in low-intensity settings, the
of OSA but had a STOP-BANG score greater than 3), required wires and tubes and the potential for motion
found that adverse respiratory events (primarily artifact make their use in low-intensity settings chal-
hypoxemia) were much more common in the two lenging. Patients find the devices uncomfortable to
OSA groups.3 The incidence of postoperative episodes wear, they impede patient mobility, and they are fre-
of hypoxemia identified using pulse oximetry oxygen quently disconnected either intentionally or
148
Respiratory Monitoring in Low-Intensity Settings
accidentally. As a result, there is a high frequency of Given the availability and low cost of pulse oxime-
false positive alarms and long periods of time with no try, it is surprising how little investigation there has
monitoring. been of its use in low-intensity settings. A recent
Ideally, respiratory monitoring in low-intensity review of four studies comparing pulse oximetry to
settings would reliably identify patients with respira- usual nursing care in surgical patients prescribed
tory compromise early enough to trigger effective opioids in the postoperative period revealed a pooled
interventions (e.g., naloxone administration, incen- trend toward decreased ICU transfers of 34% (p =
tive spirometry, pulmonary toilet, continuous positive 0.06), but the four studies were heterogeneous in
airway pressure therapy, high flow nasal oxygen) that both their design and outcomes.6 A Cochrane review
decrease the need for ICU transfer and tracheal intu- of studies in which 22,992 patients were randomized
bation and improve outcomes by reducing mortality. to usual care versus the addition of pulse oximetry
The respiratory function monitoring modalities that monitoring found that hypoxemia and related events
are currently utilized in low-intensity hospital settings were detected more frequently in the pulse oximetry
are continuous pulse oximetry, expired carbon diox- monitored patients. There was, however, no evidence
ide (CO2) measurement, photoplethysmography that use of pulse oximetry significantly affected any
(PPG), bioimpedance-based respiratory minute outcome variable, including ICU transfers and
volume monitoring, and acoustic monitoring, mortality.7
although none are widely used or accepted as a stan- There are multiple limitations that must be taken
dard of care. This chapter reviews these approaches, into consideration when interpreting this literature.
their advantages, and their limitations. There is a need to identify at-risk patients who might
benefit from monitoring and as a result have fewer
Pulse Oximetry ICU transfers and at least a decrease in interventions
ranging from minor (e.g., naloxone administration
Pulse oximetry (reviewed in Chapter 4) is widely used
and non-invasive ventilator support) to major (e.g.,
in emergency rooms, operating rooms, and ICUs to
tracheal intubation). Alarm settings have not been
monitor oxygen saturation as an indicator of respira-
standardized, and there is no widely accepted lower
tory function. It is noninvasive, accurate, inexpensive,
limit that has an acceptable high positive predictive
simple to use, and requires minimal training.
value and low false alarm rate. Notification systems
In brief, pulse oximeters use small diodes that
vary widely in clinical practice and across studies.
emit light in the 660 nm and 940 nm wavelengths
Differences in method of signaling to hospital staff,
aimed through a body part (i.e., fingertip or ear-
such as devices that emit an auditory signal that
lobe). These wavelengths are absorbed in different
require staff to be within hearing range versus devices
amounts by oxygenated and deoxygenated blood,
with automatic notification via a pager-type technol-
and the resultant transmitted light is detected by a
ogy, would expectedly yield different results.
sensor on the other side of the body part. An
One of the greatest limitations of this literature is
internal processor uses the data to determine the
that many studies have failed to account for supple-
SpO2. These devices are accurate to within ± 2% of
mental oxygen use. Increasing the FiO2 to near 0.3 via
the arterial blood oxygen saturation (SaO2) deter-
nasal cannula or face mask oxygen supplementation
mined from an arterial blood sample using a CO-
allows the pCO2 to rise to 90 mmHg before the SpO2
oximeter utilizing spectrophotometry, a much
would decrease to less than 94%. Further, the PaO2
more sophisticated device located in the main
may decrease from approximately 600 mmHg (e.g.,
laboratory of most hospitals.
on high FiO2 via facemask) to as low as 75 mmHg, a
Pulse oximeters are commonly used in low-inten-
dramatic increase in the alveolar-arterial gradient,
sity settings, but there are no accepted standards for
before a noticeable decrease in the SpO2 from 100%
indications, required staffing ratio of patients
to less than 94%. Desaturation noted when the SpO2
assigned to this level of monitoring, alarm settings,
changes from near 100% to less than 94% will then
or staff alarm notification. New advances have focused
progress rapidly as the steep portion of the oxygen
on automatic notification systems to pagers, phones,
hemoglobin dissociation curve is approached.
and other devices via a wireless network (e.g., Masimo
Some historical context may help put this into
Patient SafetyNet System), but this approach is not
perspective. Pulse oximetry became an operating
widespread.
room standard in 1987, before there was definitive perfusion, such as during advanced cardiac life sup-
evidence that it impacted outcomes. Even as late as port, and could provide insight into the percentage of
1991, the lack of evidence was troubling, but as stated dead space ventilation and metabolic abnormalities.
in a prescient review, “the ‘proof’ that a given device is The absence of expired CO2 indicates lack of ventila-
efficacious is often difficult: to define its effectiveness tion and/or perfusion; a significant decline in EtCO2
in large populations may be impossible … it would be urgently prompts the clinician to reassess ventilation
foolish to ignore [a monitor’s] potential value for lack and perfusion.
of ‘proof’ for such may never be forthcoming.”8 This Trending the EtCO2 and respiratory rate are accu-
statement also sums up the current circumstance in rate indicators of respiratory compromise, but their
low-intensity settings – high-risk patients are experi- routine use in large groups of patients has not been
encing morbidity and mortality, several noninvasive clearly associated with improved outcomes. In a study
and relatively inexpensive monitors should in theory of 133 patients monitored with minute-by-minute
provide a substantive measure of safety, and it is best vital signs, pulse oximetry, and EtCO2, a subset of
to proceed with their deployment for this purpose patients (including those with OSA) benefited from
without the proof. EtCO2 monitoring, but overall, 84% of the alarms
triggered by an EtCO2 value less than 20 or greater
Expired CO2 Monitoring than 50 mm were false.9 As discussed in Chapter 18,
There are three main methods of monitoring expired alarm fatigue is a significant issue and high rates of
CO2. The most commonly employed method uses false alarms inevitably lead to the device’s alarms to go
sensors that measure absorption of infrared light by unheeded, rendering it useless and possibly even a
CO2. These sensors may be mainstream, placed medical-legal risk.
directly into the breathing circuit (e.g., Masimo A review of five studies compared capnography
EMMA™ Mainstream Capnometer, United States), with or without pulse oximetry, to usual nursing care
or sidestream requiring aspiration of gas through tub- in surgical patients prescribed opioids in the post-
ing attached to the device containing the sensor (e.g., operative period.6 It revealed a trend toward increased
Philips Respironics LoFLo Side-Stream CO2 Sensor recognition of respiratory depression, although none
Module, Netherlands). A second method uses chemi- of the studies examined the impact on rescue team
cal indicators (Medtronic Nellcor™ Easy Cap, United activation, ICU transfers, or mortality. This review
States). These are most commonly used for very short also commented on capnography’s disadvantages,
periods of time to confirm tracheal intubation by a including false alarms and alarm fatigue.
change in color of the indicator. The third method by A more recent study of 80 women with a STOP-
Mass or Raman spectroscopy is an older cumbersome Bang score greater than 3 who underwent cesarean
technology that is usually performed in a centralized delivery with intrathecal morphine and were moni-
location, although it is now rarely used. tored with continuous capnography and pulse oxime-
Capnometry is the measurement of the partial pres- try for 24 hours revealed that 53% experienced an
sure (concentration) of CO2 in respiratory gases. A apnea event (i.e., no breath for 30–120 seconds).10
capnogram is a graphic display of the partial pressure Standard nursing monitoring did not detect a single
of expired CO2 over time, which has a waveform con- one of these events, although no clinically relevant
figuration. End-tidal CO2 (EtCO2) is the CO2 at the end events occurred in any patient. Eighty-two percent
of an expiratory cycle and is normally 35–45 mmHg. of the capnography patients reported itchy nose, nau-
Capnography is an operating room standard to assess sea, interference with nursing, and overall inconveni-
the presence and adequacy of ventilation. The detec- ence. These discomforts and the frequent alerts may
tion of expired CO2 is the “gold standard” to verify limit capnography application after cesarean delivery.
tracheal intubation. In many ICUs capnography is The main limitation of capnography in low-
routinely used in mechanically ventilated patients and intensity environments is a high false alarm rate.
is sometimes used in non-intubated spontaneously This problem, as noted in every investigation and
breathing patients, most commonly for a period of in clinical practice, is mostly due to patient
time immediately after extubation, and on occasion noncompliance with wearing the device and wiring
in patients at high risk of respiratory arrest. The or tubing. Other common causes of false alarms
expired CO2 can also be used to assess adequacy of include mouth breathing rendering nasal cannula
150
sampling useless, clogging of the sampling tubing in data synthesis of respiratory rate, SpO2, and heart
sidestream devices, and interference of mainstream rate, as well as time-trend analyses of these variables,
sensors by respiratory debris and humidity. and an algorithmically calculated risk alert triggering
A novel technology related to expired CO2 mon- an alarm.
itoring is transcutaneous CO2 monitoring (tcCO2).
This technology may be limited because detection of Bioimpedance-Based Respiratory
apnea is not really possible, but tcCO2 will rise with
hypoventilation, and in low-intensity settings this
Volume Monitoring
detection may be adequate. There is little literature Until recently, the use of bioimpedance for clinical
about this method other than confirming its correla- purposes has focused on monitoring cardiac output
tion to the arterial blood gas pCO2.11 (e.g., Cheetah NOCOM Bioreactance® Cheetah
Medical, MA, USA), and its utility for monitoring
respiration was overlooked. Because inspiration and
Photoplethysmography expiration expand and contract the lungs’ and chest’s
Photoplethysmography (PPG) is an optical measure- volume, impedance to electrical current will change
ment technique used in pulse oximetry that is used to over the respiratory cycle, and this change may be
detect blood volume changes in the microvasculature. used to detect the rate and depth of respiration. Use
All pulse oximeters utilize this technology to display of simple bioimpedance measured via EKG leads built
the pulsatile signal. As reviewed in Chapter 11, into standard bedside monitors (e.g., GE Dash 300,
respiration impacts cardiac stroke volume, which in General Electric, United States) to measure only
turn affects the blood volume within the microvascu- respiratory rate has been a simple feature often
lature. Because PPG detects blood volume changes taken for granted. Respiratory volume monitoring
within the microvasculature, it can be used to deter- (RVM) is a new approach that provides an absolute
mine respiratory rate, in combination with the stan- and graphic measurement of respiratory rate, tidal
dard pulse oximetry indicators of SpO2 and heart volume, and minute ventilation (Expiron™ 1Xi
rate.12 Respiratory Motion, Inc, United States). Uniquely,
As this technology is relatively new, investigation the depth of respiration and the respiratory rate may
has focused simply on its accuracy and not on its be trended, and alarm settings may reveal very early
impact. One study compared PPG respiratory rate respiratory compromise (e.g., change of only 20%
monitoring using the Medtronic Nellcor™ device to from baseline), before the patient becomes hypercar-
EtCO2 monitoring in 79 healthy subjects and patients bic and long before they may become hypoxemic.
in low-intensity settings who had a respiratory rate Thus, this method has distinct advantages over pulse
range of 4–34 breaths per minute.13 There was oximetry, EtCO2, and PPG monitoring.
excellent correlation and limits of agreement between Bioimpedance-based RVM has been mostly stu-
the two methods, suggesting they could be used died in the immediate postoperative period. It is
interchangeably. highly dependent on baseline calibration, does not
The obvious benefit to PPG respiratory monitor- lend itself to use in patients that are fairly mobile,
ing is that accurate continuous determination of and because it is a very new technology compared to
respiratory rate combined with SpO2 data provides pulse oximetry, EtCO2, and PPG, education and train-
measures of both respiratory rate and efficacy using a ing of staff are critical.
noninvasive sensor that is well tolerated. The com- The available data, though limited, are encoura-
mercial availability of PPG respiratory rate monitor- ging. Bioimpedance-based RVM was studied in 50
ing is relatively new, and its impact on patient patients who received opioids in a PACU, 18 of
outcomes has not been reported. It seems to have whom were classified as high-risk for respiratory
the potential benefits of capnography without the depression based on initial minute ventilation mea-
patient acceptance issues, and at very least might surement being less than 80% of that predicted.14 The
allow for automated respiratory rate acquisition in a RVM monitoring showed that 13 of the 18 (72%) high-
continuous fashion and eliminate the need for manual risk patients had a significant decline in the minute
counting by staff. ventilation, versus only 1 of the 32 (3%) opioid-
This pulse oximetry functionality may be able to receiving patients not at high risk. RVM monitoring
detect respiratory compromise using a quantitative also detected apnea, likely of no consequence, in 12 of
the 82 patients who received no opioids, a false alarm respiratory rate. Three, the number of rescue events,
rate that may be unacceptable. This very exacting pro- care escalations, ICU transfers, and opioid reversals
tocol that identified at-risk patients as those with a were not impacted by RRa monitoring. Surprisingly,
baseline MV less than 80% of the predicted value had even in patients with severe oxygen desaturation,
a sensitivity of 93% (13 true positives and 1 false respiratory rates were frequently in the normal range.
negative) and specificity of 86% (31 true negatives
and 5 false positives), an encouraging finding. Conclusion
In another study of 48 healthy volunteers
Currently there is no single respiratory monitoring
coached to change their respiratory rate, RVM mon-
modality that is widely utilized in low-intensity set-
itoring reached detection of the new steady state
tings. Furthermore, there is no monitor that has the
more rapidly that the EtCO2 monitoring.15 This sug-
requisite sensitivity, specificity, and patient accep-
gests that RVM monitoring can detect respiratory
tance for use in low-intensity settings, though newer
changes early, but evidence of improved outcomes
monitors are coming closer to this goal. A significant
is still lacking.
limitation of these technologies is the high rate of false
alarms and lack of evidence of improved patient out-
Acoustic Respiratory Rate Monitoring comes, particularly with regard to reducing the need
Acoustic Respiratory Rate (RRa®) monitoring is to transfer patients to an ICU and reducing mortality.
essentially detection of respiratory vibrations origi- It is highly likely that a monitor will be developed that
nating in the walls of the large airways during breath- combines SpO2 with one or more of these other mod-
ing that are transformed into electrical signals and alities, utilizes an algorithmic analysis to determine a
generate a displayed respiratory rate. As this technol- composite variable that identifies respiratory compro-
ogy is proprietary to Masimo, it is usually paired with mise at an early stage, and does not have a high rate of
some combination of their pulse oximeter and other false alarms that precludes widespread adaptation.
technologies.
Early iterations of this technology were plagued by References
problems with ambient noise and motion artifacts,
1. Lynn LL, Curry JP. Patterns of unexpected in-hospital
but these have mostly been eliminated by iterative deaths: a root cause analysis. Patient Safety in Surgery
technologic advances in the sensor and algorithm. 2011; 5(3):1–24.
The main advantage of RRa is that minimal training
2. Hai, F, Porhomayon J, Vermont L, et al. Postoperative
is required, the technology is easily understood, it is complications in patients with obstructive sleep apnea: a
easy to apply as it is well tolerated, and does not have meta-analysis. J Clin Anesth 2014; 26(8):591–600.
to be removed and reapplied in ambulating patients. 3. Fernandez-Bustamante A, Bartels K, Clavijo C, et al.
In a study that included 14 trauma patients that Preoperatively screened obstructive sleep apnea is
had their respiratory rate monitored with RRa, simple associated with worse postoperative outcomes than
bioimpedance, and EtCO2 as the standard, RRa had a previously diagnosed obstructive sleep apnea. Anesth
much lower bias and better agreement than simple Analg 2017;125(2):593–602.
biopedance.16 In another study of 62 children ages 2– 4. Lee LA, Caplan RA, Stephens LS, et al. Postoperative
16, acoustic respiratory rate monitoring was well tol- opioid-induced respiratory depression: a closed claims
erated 87% of the total monitoring time, and was analysis. Anesthesiology 2015;122(3):659–65.
much more accurate and had fewer false alarms than 5. Khanna AK, Overdyk FJ, Greening C, Di Stefano P,
the simple impedance method.17 Buhre WF. Respiratory depression in low acuity
A much larger investigation performed at hospital settings – seeking answers from the PRODIGY
Dartmouth-Hitchcock Medical Center entailing trial. J Crit Care 2018;47:80–7.
8,712 days of monitoring revealed three main find- 6. Lam T, Nagappa M, Wong J, et al. Continuous pulse
ings. One, there was a high rate (22.7%) of patients oximetry and capnography monitoring for
that refused the monitoring. Two, with a respiratory postoperative respiratory depression and adverse
events: a systemic review and meta-analysis. Anesth
rate alarm threshold set to ≤ 6 and ≥ 40 breaths per Analg 2017;125:2019–29.
minute, and a notably low threshold for SpO2 less
7. Pedersen T, Nicholson A, Hovhannisyan K, et al. Pulse
than 80%, the majority of alarms (43%) were due to
oximetry for perioperative monitoring. Cochrane
low oxygen saturation, and 21% were due to abnormal
152
Database of Systematic Reviews 2014, Issue 3. Art. No.: comparison with capnography. Anesth Analg 2017; 124
CD002013. (4):1153–9.
8. Duncan PG, Cohen MM. Pulse oximetry and 14. Voscopoulos C, MacNabb CM, Freeman J, et al.
capnography in anaesthetic practice: an Continuous noninvasive respiratory volume
epidemiological appraisal. Can J Anaesth 1991;38 monitoring for the identification of patients at risk for
(5):619–25. opioid-induced respiration and obstructive breathing
9. Blankush JM, Freeman R, McIlvaine J, et al. patterns. J Trauma Acute Care Surg. 2014;77:(s208–
Implementation of a novel postoperative monitoring 215).
system using automated Modified Early Warning 15. Williams GW II, George CA, Harvey BC, Free JE. A
Scores (MEWS) incorporating end-tidal capnography. comparison of measurements of changes in respiratory
J Clin Monit Comput 2017;31(5):1081–92. status in spontaneously breathing volunteers by the
10. Weiniger C, Akdagli S, Turvall E, et al. Prospective expiron noninvasive respiratory volume monitor
observational investigation of capnography and pulse versus the capnostream capnometer. Anesth Analg
oximetry monitoring after cesarean delivery with 2017;124:120–6.
intrathecal morphine. Anesth Analg 2019;128(3):513– 16. Menner A, Hu P, Stansbury L, et al. Acoustic sensor
22. versus electrocardiographically derived respiratory
11. Chhajed PN, Gehrer S, Pandey KV. Utility of rate in unstable trauma patients. J Clin Monit Comput
transcutaneous capnography for optimization of non- 2017;31:765–72.
invasive ventilation pressure. J Clin Diagn Res 2016;10 17. Patino M, Kalin M, Griffin A, et al. Comparison of
(9):OC06–OC09. postoperative respiratory monitoring by acoustic and
12. Charlton PH, Bonnici T, Tarassenko L, et al. As transthoracic impedance technologies in pediatric
assessment of algorithms to estimate respiratory from patients at risk or respiratory depression. Anesth Analg
the electrocardiogram and photoplethysmogram. 2017;124(6)1937–42.
Physiol Meas 2016;37(4):610–26. 18. McGrath SP, Pyke J, Taenzer AH. Assessment of
13. Bergese SD, Mestek ML, Kelley SD, et al. Multicenter continuous acoustic respiratory rate monitoring as an
study validating accuracy of a continuous respiratory addition to a pulse oximetry-based patient surveillance
rate measurement derived from pulse oximetry: a system. J Clin Monit Comput 2017;31(3):561–91.
Chapter
The Electronic Health Record as a Monitor
16 for Performance Improvement

David B. Wax
Introduction former are highly customized to meet local needs, but

they may lack interoperability with other systems. The
The first consistent use of anesthesia records is
latter may not fit perfectly into local workflows, but
thought to have been in 1895 when Cushing and
they get closer to the goal of a universal health record.
Codman kept “ether charts” as part of a wager to see
It is projected that 84% of academic anesthesiol-
who could most improve their anesthesia skills.
ogy departments (and an unknown number of com-
Cushing later remarked that “we both became very
munity practices) in the United States will utilize an
much more skillful … particularly due to the detailed
AIMS by 2020.4 These systems create a wealth of data
attention which we had to put upon the patient by the
that can be used for clinical, administrative, training,
careful recording of the pulse rate throughout the
legal, research, compliance, and other purposes. Such
operation.” By 1920, Silk remarked on the “impor-
“big data” are the fuel for countless research efforts
tance of observing the variations in blood pressure of
that will eventually lead to improvements in practice.5
a patient while under an anaesthetic.”1 Thus, there
However, the focus of this chapter will be on more
was early recognition that measuring vital signs in
immediate uses of AIMS data to inform decisions and
order to record them, and then observing trends in
improve our practice in real time as other bedside
those recordings, could lead to improved care. This
patient monitors do.
elevated anesthesia records to the status of patient
monitors rather than just banal transcripts.
Nearly a century later, innovation took a wrong Electronic Anesthesia Records
turn when an automated anesthesia record-keeping At its most fundamental level, an electronic anesthesia
system was reported that simply used a video camera record is simply a digital (and more legible) version of
pointed at the anesthesia workstation for the entire a traditional paper anesthesia record. The primary
case.2 While this provided a high fidelity record of value of these records is the same now as it was
everything the practitioner saw and did, and might be described long ago: measurement and documentation
ideal for a courtroom, there was no way to utilize any of physiologic data, and observation of their trends,
of the recorded data other than to replay the footage. can help practitioners provide better care. As a visual
Around the same time, however, computerized representation of data aggregated from various other
anesthesia record-keepers were being developed that patient monitors over time, the AARK itself is a
stored information electronically as discrete data, patient monitor that provides unique insight beyond
allowing them to be later retrieved and analyzed for the standalone monitors that feed it.
myriad purposes.3 These systems were initially called There was initial concern that moving from a
AARKS (automated anesthesia record keepers) and paper record to an electronic record could actually
then AIMS (anesthesia information management sys- decrease its effectiveness as a clinical tool. The theory
tems), as their scope of use widened beyond just was that, with the computer automatically recording
automatically creating a paper intraoperative record. vital signs, practitioners could/would be less attentive
Some AIMS are homegrown systems, others are to the patient’s state. There was also fear that mon-
standalone commercial products, and still others are itoring artifacts or non-reassuring data might be per-
modules integrated into a comprehensive EHR/EMR manently recorded in the record and possibly increase
(electronic health/medical record) that encompasses malpractice exposure. Some of these fears were
all aspects of care rather than just anesthesia care. The allayed in various studies on these matters.
154
Regarding vigilance, two studies showed that use of an Decision Support Systems
AARK did not decrease practitioner awareness of
The computers running AIMS no longer just cap-
their patient’s state.6,7 A later study found that
ture, display, and store clinical data for review.
where internet surfing was available as a potential
Their processing power is now utilized to make
distraction on the same workstation as the AARK,
use of those data for performance improvement
practitioners did spend substantial amounts of time
with integrated decision support systems
looking at something other than the AARK, but this
(DSS).24–27 These algorithms take physiologic
did not adversely impact the incidence of abnormal
data from monitors, clinical information entered
patient hemodynamics.8 Other studies showed that
by practitioners, laboratory data, and other inputs
paper records were incomplete and inaccurate, while
and algorithmically generate a recommendation
AARK records better reflected reality than the “rail-
that is provided real time or near real time to the
road tracks” of vital signs on paper records.9–12
practitioner to guide care. The algorithms them-
Regarding malpractice claims, one study showed
selves can be based on standards, guidelines, reg-
that using an AIMS helped result in dismissal, settle-
ulations, checklists, study protocols, corrective
ment, or successful litigation of cases, and did not
actions, or presumed best practices. They need to
hinder the defense.13 AIMS metadata have also been
be designed so as not to trigger inappropriately
utilized to discourage attending anesthesiologists
and cause alert fatigue. In this way, the anesthesia
from improperly attesting to clinical events that have
record becomes more than just a “dumb” monitor
not yet occurred.14
mirroring patient state – it becomes an accessory
Beyond the basic “ether chart,” the AIMS
brain for the practitioner to interpret the data and
record may include pre-anesthesia notes, procedure
act appropriately upon it. This type of assistance
notes, post-anesthesia notes, medication and nur-
becomes increasingly important as more and more
sing orders, and administrative data. As part of an
research generates more and more predictive algo-
enterprise EHR, this documentation is potentially
rithms, scoring systems, and evidence-based
viewed by other anesthesia care teams as well as
recommendations that individual practitioners
other specialists, administrators, and payers.15 It
cannot reasonably be expected to know and utilize
can also be viewed remotely when supervising resi-
all of without assistance. The aggregation, synth-
dent/CRNA cases.
esis, and dissemination of knowledge by DSS help
As more centers adopt them, concerns about pit-
overcome this barrier to adoption of evidence-
falls of EHRs extend to anesthesia and critical care
based medicine.28 This is not without controversy,
practices.16–19 In fact, in a 2013 poll of healthcare
however, as legitimate criticism of evidence-based
leaders, health information technology (HIT) was
medicine can decrease acceptance of DSS.29,30
voted the “safety hazard of the year.”20 The most
commonly cited concerns are those of copy/paste There is a multitude of EHR-based DSS. Overall,
functionality and automated note templates which they appear to have positive or neutral impacts on
lead to “note bloat” and repetition of irrelevant or patient outcomes.31 There is currently little to no
erroneous information that can fatigue or mislead regulatory oversight of these systems as there is for
subsequent caregivers.21 Other concerns include other medical devices, leaving developers and users to
excessive time spent on ever-increasing documenta- ensure the quality and safety of their own systems. It
tion, patient dissatisfaction with practitioner compu- remains to be seen if these systems will become a
ter use during consultations, alert fatigue, privacy and significant source of adverse events or malpractice
security risks, potential for documentation in the litigation.
wrong patient’s record, and alienation of senior prac- Most DSS are directed at improving patient care,
titioners not accustomed to computers. Despite these though some focus on practitioner issues and others
concerns, EHR adoption has accelerated due to reg- focus on administrative issues. Unfortunately, only
ulatory and financial incentives to utilize HIT.22,23 those DSS that are formally published, anecdotally
Anesthesiology EHRs may be less subject to some of disclosed, or built into commercial EHRs are known
these concerns, as the majority of the record is not and will be described here. This likely leaves many
subject to copy and paste, and patients are generally other systems in use that are known only to their
unaware of the time spent on recordkeeping. developers and users.
Antibiotic Administration cardiopulmonary bypass ends and the AIMS does

not detect vital signs associated with resumption of
Some of the earliest AIMS-based interventions were
ventilation. A “rule of thumb” DSS is exemplified by a
directed at improving on-time administration of anti-
“rule of fours” alert that is triggered to raise awareness
biotics in an effort to follow guidelines for prevention
if the AIMS data indicate over 4 units of blood trans-
of surgical site infections.32–36 These systems utilized
fused, 4 liters of fluids, base deficit of 4, and 4 hours of
some combination of on-screen reminders, real-time
surgery. Abnormal point-of-care lab results have also
alerts (pager/text messaging), and performance
been used to trigger alerts to the attending anesthe-
reports to ensure pre-incision antibiotic administra-
siologist who may be out of the room and otherwise
tion and intraoperative re-dosing as needed.
unaware of such results.44 Retained guidewires have
led to electronic checklist items that confirm guide-
Postoperative Nausea/Vomiting wire removal when a central line is placed. Electronic
Prophylaxis procedure notes for intrathecal opioid injections or
Risk factors for postoperative nausea and vomiting neuraxial/peripheral nerve catheter placements have
(PONV) have been known for some time, but antie- also been used to trigger notifications to a pain man-
metic prophylaxis is still not always tailored to indi- agement service to ensure continuity of care.
vidual patients. DSS has been utilized to change
this.37,38 The systems work by using data from the Chart Review
preoperative evaluation (e.g., age, gender) and intrao- Self-reporting of adverse events is imperfect due to
perative record (e.g., opioids, inhaled anesthetics, various human and technical factors. The literature
procedure type) to identify risk factors for PONV indicates that electronic queries of AIMS records are
and then make real-time recommendations to the significantly more reliable than manual reporting of
practitioner about anti-emetic agents to administer. incidents.45–47 Additionally, the volume of clinical
documentation in an EHR often exceeds an amount
Blood Pressure that can be efficiently or reasonably reviewed manu-
Since both hypotension and hypertension have been ally prior to rendering care. Based on this, a DSS was
associated with adverse outcomes, some DSS have created that screens old AIMS records for both dis-
focused on limiting blood pressure extremes. One DSS crete data and narrative documentation suggestive of
monitored for hypotension in the setting of high levels of critical intraoperative events (e.g., difficult intubation,
inhaled anesthetic, as well as hypertension in the setting laryngospasm, bronchospasm, difficult IV access).
of vasoconstrictor infusion, and alerted practitioners of Cases with such events detected are flagged so that,
the conditions.39 This system led to changes in practi- if the same patient returns for another procedure, the
tioner behavior, but patient outcomes were not reported. practitioner is forewarned of the potential problem
Another system monitored for a “double low” state of and can anticipate or prevent it from recurring.48
hypotension and low bispectral index and suggested
intervention.40 Unfortunately, this seemed to have little Crisis Avoidance
effect on practitioner behavior or short-term outcomes. Another exciting area of DSS are those systems that
attempt to identify impending crises in advance so
Corrective Actions that early intervention can be initiated. This is made
The root cause analyses and peer review processes that possible by an enterprise-wide EHR that aggregates
follow near misses or adverse events can identify physiologic data from various patient areas and
causative or contributory factors that are amenable automatically screens them using evidence-based
to preventive measures based on DSS. One of the criteria of impending clinical deterioration. This
earliest reports of this involved liability partly result- process occurs in the background so it does not
ing from a gap in vital sign recording (and alleged gap require the patient’s chart to be already accessed/
in patient monitoring). DSS was subsequently utilized open to trigger the monitoring/screening and alerts.
to alert practitioners when the AIMS vital sign stream Such systems have been used to identify patients
is lost so they can intervene promptly.41–43 Another developing SIRS/sepsis, respiratory failure, and
example is an apnea alert that is triggered if acute kidney injury.49–52 These alerts (and the
156
protocols employed once they are triggered) are in and reminders), and combines best-practice evidence
their infancy, with most just deployed in the last 5 from a variety of areas conveniently in one place (see
years, so the outcome data are preliminary. The Figure 16.1).
“weak link” seems to be provider response and
adherence to protocol, perhaps as a result of alert Risk Reduction
fatigue, false alarms, and the overall embrace of pro-
Various other EHR-based systems have been imple-
tocols, bundles, care maps, and algorithms that make
mented in the anesthesia, pain management,
adherence to any one of these more difficult.
and critical care realms. These include surgical “time
outs,” anesthesia workstation preparation,
DSS Packages beta blocker administration, opioid prescribing, cen-
Just as guidelines are now coming in packages (e.g., tral line-associated bloodstream infection (CLABSI)
ERAS protocols), DSS are also being packaged together reduction practices, venous thromboembolism (VTE)
to provide optimal care.53–56 Some of these systems are prophylaxis, antibiotic stewardship, urinary catheter
layered on top of an EHR and aggregate data to present avoidance, glucose control, ventilator management,
them to the clinician in a more visually/graphically or blood transfusion restriction, and rational lab and
animated way than the standalone (mostly numeric) radiology testing.57–69
monitors they originated from. They also have triggers
more complex than the simple low/high limits on most Drug Diversion
standalone monitors. Decision support in these sys-
Anesthesia practitioners are at risk for substance use
tems may be implicit (e.g., an image of lungs turning
disorder, and early detection and intervention can be
from green to red with hypercarbia) or explicit (alerts
life-saving. To this end, DSS have been used to
Back Surgical Incision Active Alerts

MAC* = 0.7
with Propofol Hypotension: MAP = 52
Patient Drake
146729674 Ventilated Cumulative time for MAP <
Age: 68 RR = 12 bpm 55 = 22 minutes.
Weight (kg): 88.5 PIP = 24 cm H20
Ideal Weight: 48.0 PEEP = 6.0 mm H20
TV act = 350 ml
U-OR 05
ASA 4 CO2 SpO2
Total Colectomy 100
Case Duration: 4:00 50 Pa 90
ce
r
30
Progress 80% 40.0 96%

mmHg
I/O Balance: –1153
Infectious Diseases HR = 90 bpm [ekg]

BP Art Line [7s]
SPV = 6 [8s]
67 / 45 MAP = 52
Normal Volume
PreOp Pain Score: 4
Temp INR Plat Hct Est K+ Glu
Infusions 200
1.5 450 40 5.0
37 30 140 CAUTION: Check monitors and
Low MAP Alert < 55 36 1.2 4.0
35 1.0 150 20 3.0 medical record before making
70
medical decisions.
36.8 1.0 90 32 28 4.0 120
Creat = 1.55 500 ml
[2d] [4h] [3h] [2d] [45m]
[4h] 150 ml/hr More Info Emergency Refs
2 ml/kg/hr
Figure 16.1 An integrated, graphical, anesthesia monitoring and clinical decision support system package. (A black and white version of this
figure will appear in some formats. For the color version, please refer to the plate section).
Kheterpal S, Shanks A, Tremper KK. Impact of a Novel Multiparameter Decision Support System on Intraoperative Processes of Care and
Postoperative Outcomes. Anesthesiology. 2018; 128:272–282.
Source: Kruger GH, Tremper KK. Advanced integrated real-time clinical displays. Anesthesiol Clin 2011; 29: 487–504.
monitor AIMS drug administration records and phar- to clinical coordinators so they can schedule residents
macy records and identify discrepancies and patterns as necessary to meet their ACGME minimum case
that have been associated with drug diversion.70,71 requirements.74 Another effort to assign residents to
cases of appropriate acuity involved using informa-
Relief Equity tion from the EHR to calculate case risk stratification
scores for use during resident scheduling.75 Another
Inequitable work hours can be a source of job dissatis-
use is facilitating staff evaluations mandated by the
faction and burnout in healthcare workers. Two DSS
ACGME. AIMS-based reminders can identify all the
systems have been described that utilize AIMS case
faculty/resident dyads for each clinical day and send
information to determine recent departure times for
reminders to encourage completion of mutual evalua-
attendings and residents and provide this information
tions in a timely fashion.76,77 An (unsuccessful) effort
to the on-call team (see Figure 16.2).72,73 This helps
was even made to use AIMS data about intraoperative
balance work hours by prioritizing early relief for those
blood pressure management by residents to predict
who have recently worked late. This has traditionally
their performance on later competency exams.78
been accomplished with maintenance of paper or
grease-board “diaries” and “group think,” and is
prone to human error and explicit and implicit bias. Cost Containment
Pharmaceutical costs can be a significant component
Resident Monitoring of anesthesia practice expenses. In order to reduce
spending, AIMS data have been leveraged to monitor
DSS have also been applied to monitoring of residents
patterns of medication selection by individual practi-
for training and evaluation purposes. One area of use
tioners and provide negative feedback to those
is to provide AIMS-based case data for each resident
Figure 16.2 Decision support tool for anesthesiologist work hour monitoring and end-of-shift relief equity with suggested relief order ranked
by descending average of departure times on prior five weekdays (Mount Sinai Hospital, New York, NY USA).
158
utilizing more costly agents than their peers.79 standards) require reporting of data to surveyors,
Excessive fresh gas flow when using inhaled anes- payers, and registries in order to qualify for incentive
thetics can also increase costs (and air pollution). payments or avoid penalties or citations. Missing data
Efforts have been made using AIMS data and DSS to can thus result in sub-maximal revenue or accredita-
monitor fresh gas flow rates and encourage practi- tion issues. In order to ensure complete data for
tioners to reduce them if they exceed norms.80,81 reporting, EHRs can be designed with reminders
and warnings that prevent practitioners from omit-
Billing ting items that are mandatory for reporting compli-
ance. This can be a simple reminder that a particular
Much of the return-on-investment for EHRs is pur-
mandatory item in a preoperative note is incomplete
ported to result from improved revenue collection by
or a postoperative note is missing, or a more compli-
monitoring documentation and billing processes to
cated DSS that, for example, alerts the anesthesia team
ensure rapid clean bill submission (i.e., minimal
that no patient temperature has been recorded in an
charge lag) and maximal payment (i.e., gross collec-
ongoing general or neuraxial anesthetic case of greater
tion rate) for services. Early efforts in this regard were
than one hour.
aimed at identifying a minimum dataset necessary to
have a bill that would be paid without delay or denial
due to incomplete items. A background process Conclusions
would screen AIMS cases prior to claim submission Electronic anesthesia records and decision support
and alert practitioners if deficiencies were found so systems have become an integral part of the modern
they could be remedied.82,83 Maximizing reimburse- monitoring armamentarium. Together they act as
ment was the goal of another system that would check both historian and consultant. These systems are still
the AIMS data for invasive blood pressure data and in their infancy and will continue to grow in both
send an alert if no procedure note was filed for the number and complexity for the benefit of patients,
placement of the arterial catheter (thus precluding practitioners, and systems.
billing for it).84 Other algorithms have been used to
avoid lost bills by verifying that every case on the OR References
schedule on a given day results in a billing entry. Still
1. Zeitlin GL. History of Anesthesia Records. American
other efforts have used AIMS data and DSS in mon- Society of Anesthesiologists Newsletter. www.woodli
itoring operating room utilization to minimize delays brarymuseum.org/news/pdf/Zeitlin.pdf
and maximize efficiency.85,86
2. Piepenbrink JC, Cullen JI Jr, Stafford TJ. The use of
video in anesthesia record keeping. Biomed Instrum
Compensation Technol 1990;24:19–24.
Physician compensation can also be driven with 3. Kenney GNC. Implementation of computerized
AIMS data. A modified fee-for-service system has anaesthetic records. Baillière’s Clinical Anaesthesiology
been described in which productivity pay is allocated 1990;4:1–6.
among the faculty based on case complexity and dura- 4. Stol IS, Ehrenfeld JM, Epstein RH. Technology diffusion
tion as documented in the AIMS case records.87 In of anesthesia information management systems into
complex multisite practices, this may be an effective academic anesthesia departments in the United States.
Anesth Analg 2014;118:644–50.
means of adjudicating relative “value” of individuals’
work based on real-time knowledge of the work bur- 5. Levin MA, Wanderer JP, Ehrenfeld JM. Data, big data,
den, assuring compensation fairness and encouraging and metadata in anesthesiology. Anesth Analg
2015;121:1661–7.
productivity.
6. Loeb RG. Manual record keeping is not necessary for
anesthesia vigilance. J Clin Monit 1995;11:9–13.
Compliance 7. Allard J, Dzwonczyk R, Yablok D, Block FE Jr,
Various quality measures and regulations (e.g., EHR McDonald JS. Effect of automatic record keeping on
meaningful use [HITECH-MU], Surgical Care vigilance and record keeping time. Br J Anaesth
Improvement Project [SCIP], Merit-Based Incentive 1995;74:619–26.
Payment System [MIPS], Department of Health 8. Wax DB, Lin HM, Reich DL. Intraoperative non-
[DOH] regulations, Joint Commission [JC] record-keeping usage of anesthesia information
management system workstations and associated 22. Gálvez JA, Rothman BS, Doyle CA, Morgan S, Simpao
hemodynamic variability and aberrancies. AF, Rehman MA. A narrative review of meaningful use
Anesthesiology 2012;117:1184–9. and anesthesia information management systems.
9. Thrush DN. Are automated anesthesia records better? Anesth Analg 2015;121:693–706.
J Clin Anesth 1992;4:386–9. 23. McWilliams JM. MACRA: big fix or big Problem? Ann
10. Lerou JG, Dirksen R, van Daele M, Nijhuis GM, Crul Intern Med 2017;167:122–4.
JF. Automated charting of physiological variables in 24. Epstein RH, Dexter F, Patel N. Influencing anesthesia
anesthesia: a quantitative comparison of automated provider behavior using anesthesia information
versus handwritten anesthesia records. J Clin Monit management system data for near real-time alerts and
1988;4:37–47. post hoc reports. Anesth Analg 2015;121:678–92.
11. Devitt JH, Rapanos T, Kurrek M, Cohen MM, Shaw M. 25. Simpao AF, Tan JM, Lingappan AM, Gálvez JA,
The anesthetic record: accuracy and completeness. Can Morgan SE, Krall MA. A systematic review of near real-
J Anaesth 1999;46:122–8. time and point-of-care clinical decision support in
12. Reich DL, Wood RK Jr, Mattar R, et al. Arterial blood anesthesia information management systems. J Clin
pressure and heart rate discrepancies between Monit Comput 2017;31:885–94.
handwritten and computerized anesthesia records. 26. Nair BG, Gabel E, Hofer I, Schwid HA, Cannesson M.
Anesth Analg 2000;91:612–16. Intraoperative clinical decision support for anesthesia:
13. Feldman JM. Do anesthesia information systems a narrative review of available systems. Anesth Analg
increase malpractice exposure? Results of a survey. 2017;124:603–17.
Anesth Analg 2004;99:840–3. 27. Belard A, Buchman T, Forsberg J, et al. Precision
14. Vigoda MM, Lubarsky DA. The medicolegal diagnosis: a view of the clinical decision support
importance of enhancing timeliness of documentation systems (CDSS) landscape through the lens of critical
when using an anesthesia information system and the care. J Clin Monit Comput 2017;31:261–71.
response to automated feedback in an academic 28. Sadeghi-Bazargani H, Tabrizi JS, Azami-Aghdash S.
practice. Anesth Analg 2006;103:131–6. Barriers to evidence-based medicine: a systematic
15. Wanderer JP, Gruss CL, Ehrenfeld JM. Using visual review. J Eval Clin Pract 2014;20:793–802.
analytics to determine the utilization of preoperative 29. Kumanan W. Evidence-based medicine. The good the
anesthesia assessments. Appl Clin Inform 2015;6:629– bad and the ugly. A clinician’s perspective. J Eval Clin
37. Pract 2010;16:398–400.
16. Bowman S. Impact of electronic health record systems 30. Prasad V, Cifu A. Medical reversal: why we must raise
on information integrity: quality and safety the bar before adopting new technologies. Yale J Biol
implications. Perspect Health Inf Manag 2013;10:1c. Med 2011;84:471–8.
17. Meeks DW, Smith MW, Taylor L, Sittig DF, Scott JM, 31. Varghese J, Kleine M, Gessner SI, Sandmann S, Dugas
Singh H. An analysis of electronic health record- M. Effects of computerized decision support system
related patient safety concerns. J Am Med Inform Assoc implementations on patient outcomes in inpatient
2014;21:1053–9. care: a systematic review. J Am Med Inform Assoc
18. Shaarani I, Taleb R, Antoun J. Effect of computer use 2018;25(5):593–602.
on physician-patient communication using a validated 32. O’Reilly M, Talsma A, VanRiper S, Kheterpal S,
instrument: patient perspective. Int J Med Inform Burney R. An anesthesia information system designed
2017;108:152–7. to provide physician-specific feedback improves timely
19. Blijleven V, Koelemeijer K, Jaspers M. Identifying and administration of prophylactic antibiotics. Anesth
eliminating inefficiencies in information system usage: Analg 2006;103:908–12.
a lean perspective. Int J Med Inform 2017;107:40–7. 33. Wax DB, Beilin Y, Levin M, Chadha N, Krol M, Reich
20. Denham CR, Classen DC, Swenson SJ, Henderson MJ, DL. The effect of an interactive visual reminder in an
Zeltner T, Bates DW. Safe use of electronic health anesthesia information management system on
records and health information technology systems: timeliness of prophylactic antibiotic administration.
trust but verify. J Patient Saf 2013;9:177–89. Anesth Analg 2007;104:1462–6.
21. Tsou AY, Lehmann CU, Michel J, Solomon R, 34. Nair BG, Newman SF, Peterson GN, Wu WY, Schwid
Possanza L, Gandhi T. Safe practices for copy and paste HA. Feedback mechanisms including real-time
in the EHR. Systematic review, recommendations, and electronic alerts to achieve near 100% timely
novel model for health IT collaboration. Appl Clin prophylactic antibiotic administration in surgical
Inform 2017;8:12–34. cases. Anesth Analg 2010;111:1293–300.
160
35. Nair BG, Newman SF, Peterson GN, Schwid HA. quality assurance program. J Clin Monit Comput
Automated electronic reminders to improve redosing 2000;16:211–17.
of antibiotics during surgical cases: comparison of two 47. Grant C, Ludbrook G, Hampson EA, Semenov R,
approaches. Surg Infect 2011;12:57–63. Willis R. Adverse physiological events under
36. Schwann NM, Bretz KA, Eid S, et al. Point-of-care anaesthesia and sedation: a pilot audit of electronic
electronic prompts: an effective means of increasing patient records. Anaesth Intensive Care 2008;36:222–9.
compliance, demonstrating quality, and improving 48. Wax DB, McCormick PJ, Joseph TT, Levin MA. An
outcome. Anesth Analg 2011;113:869–7. automated critical event screening and notification
37. Kooij FO, Klok T, Hollmann MW, Kal JE. Decision system to facilitate preanesthesia record review. Anesth
support increases guideline adherence for prescribing Analg 2018;126:606–10.
postoperative nausea and vomiting prophylaxis. 49. Pedersen NE, Rasmussen LS, Petersen JA, Gerds TA,
Anesth Analg 2008;106:893–8. Østergaard D, Lippert A. A critical assessment of early
38. Kappen TH, Vergouwe Y, van Wolfswinkel L, warning score records in 168,000 patients. J Clin Monit
Kalkman CJ, Moons KG, van Klei WA. Impact of Comput 2018;32:109–16.
adding therapeutic recommendations to risk 50. Blankush JM, Freeman R, McIlvaine J, Tran T, Nassani
assessments from a prediction model for S, Leitman IM. Implementation of a novel
postoperative nausea and vomiting. Br J Anaesth postoperative monitoring system using automated
2015;114:252–60. Modified Early Warning Scores (MEWS)
39. Nair BG, Horibe M, Newman SF, Wu WY, Peterson incorporating end-tidal capnography. J Clin Monit
GN, Schwid HA. Anesthesia information management Comput 2017;31:1081–92.
system-based near real-time decision support to 51. Breighner CM, Kashani KB. Impact of e-alert systems
manage intraoperative hypotension and hypertension. on the care of patients with acute kidney injury. Best
Anesth Analg 2014;118:206–14. Pract Res Clin Anaesthesiol 2017;31:353–9.
40. McCormick PJ, Levin MA, Lin HM, Sessler DI, Reich 52. Schmidt PE, Meredith P, Prytherch DR, Watson D,
DL. Effectiveness of an electronic alert for hypotension Watson V. Impact of introducing an electronic
and low bispectral index on 90-day postoperative physiological surveillance system on hospital
mortality: a prospective, randomized trial. mortality. BMJ Qual Saf 2015;24:10–20.
53. Ljungqvist O, Scott M, Fearon KC. Enhanced recovery
41. Vigoda MM, Lubarsky DA. Failure to recognize loss of after surgery: a review. JAMA Surg 2017;152:292–8.
incoming data in an anesthesia record-keeping system
may have increased medical liability. Anesth Analg 54. Kruger GH, Tremper KK. Advanced integrated real-
2006;102:1798–802. time clinical displays. Anesthesiol Clin 2011;29:487–
504.
42. Ehrenfeld JM, Epstein RH, Bader S, Kheterpal S,
Sandberg WS. Automatic notifications mediated by 55. Nair BG, Newman SF, Peterson GN, Schwid HA.
anesthesia information management systems reduce Smart Anesthesia Manager™ (SAM) – a real-time
the frequency of prolonged gaps in blood pressure decision support system for anesthesia care during
documentation. Anesth Analg 2011;113:356–63. surgery. IEEE Trans Biomed Eng 2013;60:207–10.
43. Nair BG, Horibe M, Newman SF, Wu WY, Schwid H. 56. Kheterpal S, Shanks A, Tremper KK. Impact of a novel
Near real-time notification of gaps in cuff blood multiparameter decision support system on
pressure recordings for improved patient monitoring. J intraoperative processes of care and postoperative
Clin Monit Comput 2013;27:265–71. outcomes. Anesthesiology 2018;128:272–82.
44. Freundlich RE, Grondin L, Tremper KK, Saran KA, 57. Wetmore D, Goldberg A, Gandhi N, Spivack J,
Kheterpal S. Automated electronic reminders to McCormick P, DeMaria S Jr. An embedded checklist in
prevent miscommunication among primary medical, the anesthesia information management system
surgical and anaesthesia providers: a root cause improves pre-anaesthetic induction setup: a
analysis. BMJ Qual Saf 2012;21:850–4. randomised controlled trial in a simulation setting.
BMJ Qual Saf 2016;25:739–46.
45. Sanborn KV, Castro J, Kuroda M, Thys DM. Detection
of intraoperative incidents by electronic scanning of 58. Shear T, Deshur M, Avram MJ, et al. Procedural
computerized anesthesia records. Comparison with timeout compliance is improved with real-time clinical
voluntary reporting. Anesthesiology 1996;85:977–87. decision support. J Patient Saf 2018;14(3):148–52.
46. Benson M, Junger A, Fuchs C, et al. Using an 59. Patel S, Carmichael JM, Taylor JM, Bounthavong M,
anesthesia information management system to prove a Higgins DT. Evaluating the impact of a clinical
deficit in voluntary reporting of adverse events in a decision support tool to reduce chronic opioid dose
and decrease risk classification in a veteran population. on an analysis of atypical drug transactions. Anesth
Ann Pharmacother 2018;52:325–31. Analg 2007;105:1053–60.
60. Quan KA, Cousins SM, Porter DD, et al. Electronic 71. Epstein RH, Gratch DM, McNulty S, Grunwald Z.
health record solutions to reduce central line- Validation of a system to detect scheduled drug
associated bloodstream infections by enhancing diversion by anesthesia care providers. Anesth Analg
documentation of central line insertion practices, line 2011;113:160–4.
days, and daily line necessity. Am J Infect Control 72. Wax DB, McCormick PJ. A real-time decision support
2016;44:438–43. system for anesthesiologist end-of-shift relief. Anesth
61. Kahn SR, Morrison DR, Cohen JM, et al. Interventions Analg 2017;124:599–602.
for implementation of thromboprophylaxis in 73. Bhutiani M, Jablonski PM, Ehrenfeld JM, McEvoy MD,
hospitalized medical and surgical patients at risk for Fowler LC, Wanderer JP. Decision support tool
venous thromboembolism. Cochrane Database Syst improves real and perceived anesthesiology resident
Rev 2013;7:CD008201. relief. Anesth Analg 2018;127(2):513–19.
62. Jenkins I, Doucet JJ, Clay B, et al. Transfusing wisely: 74. Wanderer JP, Charnin J, Driscoll WD, Bailin MT,
clinical decision support improves blood transfusion Baker K. Decision support using anesthesia
practices. Jt Comm J Qual Patient Saf 2017;43:389–95. information management system records and
63. Nair BG, Grunzweig K, Peterson G, et al. accreditation council for graduate medical education
Intraoperative blood glucose management: impact of a case logs for resident operating room assignments.
real-time decision support system on adherence to Anesth Analg 2013;117:494–9.
institutional protocol. J Clin Monit Comput 75. Was A, Wanderer J. Matching clinicians to operative
2016;30:301–12. cases: a novel application of a patient acuity score. Appl
64. Schulz L, Osterby K, Fox B. The use of best practice Clin Inform 2013;4:445–53.
alerts with the development of an antimicrobial 76. Blum JM, Kheterpal S, Tremper KK. A comparison of
stewardship navigator to promote antibiotic de- anesthesiology resident and faculty electronic
escalation in the electronic medical record. Infect evaluations before and after implementation of
Control Hosp Epidemiol 2013;34:1259–65. automated electronic reminders. J Clin Anesth
65. Chen YY, Chi MM, Chen YC, Chan YJ, Chou SS, Wang 2006;18:264–7.
FD. Using a criteria-based reminder to reduce use of 77. Rusa R, Klatil F, Fu R, Swide CE. Impact of faculty-
indwelling urinary catheters and decrease urinary tract specific electronic reminders on faculty compliance
infections. Am J Crit Care 2013;22:105–14. with daily resident evaluations: a retrospective study. J
66. Karbing DS, Allerød C, Thomsen LP, et al. Clin Anesth 2009;21:159–64.
Retrospective evaluation of a decision support system 78. Sessler DI, Makarova N, Riveros-Perez R, Brown DL,
for controlled mechanical ventilation. Med Biol Eng Kimatian S. Lack of association between blood
Comput 2012;50:43–51. pressure management by anesthesia residents and
67. Nair BG, Peterson GN, Newman SF, Wu WY, Kolios- competence committee evaluations or in-training
Morris V, Schwid HA. Improving documentation of a exam performance: a cohort analysis. Anesthesiology
beta-blocker quality measure through an anesthesia 2016;124:473–82.
information management system and real-time 79. Lubarsky DA, Sanderson IC, Gilbert WC, et al. Using
notification of documentation errors. Jt Comm J Qual an anesthesia information management system as a
Patient Saf 2012;38:283–8. cost containment tool. Description and validation.
68. Krasowski MD, Chudzik D, Dolezal A, et al. Promoting Anesthesiology 1997;86:1161–9.
improved utilization of laboratory testing through 80. Nair BG, Peterson GN, Neradilek MB, Newman SF,
changes in an electronic medical record: experience at Huang EY, Schwid HA. Reducing wastage of
an academic medical center. BMC Med Inform Decis inhalation anesthetics using real-time decision support
Mak 2015 February 22;15:11. to notify of excessive fresh gas flow. Anesthesiology
69. Schneider E, Zelenka S, Grooff P, Alexa D, Bullen J, 2013;118:874–84.
Obuchowski NA. Radiology order decision support: 81. Wax DB, Hill B, Levin MA. Ventilator data extraction
examination-indication appropriateness assessed with a video display image capture and processing
using 2 electronic systems. J Am Coll Radiol system. J Med Syst 2017;41:101.
2015;12:349–57.
82. Reich DL, Kahn RA, Wax D, Palvia T, Galati M, Krol
70. Epstein RH, Gratch DM, Grunwald Z. Development of M. Development of a module for point-of-care charge
a scheduled drug diversion surveillance system based capture and submission using an anesthesia
162
information management system. Anesthesiology 85. Foglia RP, Alder AC, Ruiz G. Improving perioperative
2006;105:179–86. performance: the use of operations management and
83. Spring SF, Sandberg WS, Anupama S, Walsh JL, the electronic health record. J Pediatr Surg 2013;48:95–
Driscoll WD, Raines DE. Automated 8.
documentation error detection and notification 86. Van Winkle RA, Champagne MT, Gilman-Mays M,
improves anesthesia billing performance. Aucoin J. Operating room delays: meaningful use in
Anesthesiology 2007;106:157–63. electronic health record. Comput Inform Nurs
84. Kheterpal S, Gupta R, Blum JM, Tremper KK, O’Reilly 2016;34:247–53.
M, Kazanjian PE. Electronic reminders improve 87. Reich DL, Galati M, Krol M, Bodian CA, Kahn RA. A
procedure documentation compliance and mission-based productivity compensation model for
professional fee reimbursement. Anesth Analg an academic anesthesiology department. Anesth Analg
2007;104:592–7. 2008;107:1981–8.
Chapter
Future Monitoring Technologies:
17 Wireless, Wearable, and Nano

Ira S. Hofer and Myro Figura
be deployed on a secure and robust infrastructure

Introduction with reporting and monitoring features. Some chal-
Advances in technology permit monitoring of an lenges that face wired technologies are amplified, and
increasing array of physiological parameters in near new ones are introduced once connectivity and device
real-time in diverse environments, by healthcare pro- interoperability is incorporated. As solutions to these
viders, their patients, and even consumers interested challenges are being developed, what remains to be
in self-monitoring. Miniaturization and nanotechnol- answered is how clinicians will access and utilize the
ogy engender monitors that are more portable and wide breadth of this information. Outstanding ques-
allow for the measurement of previously unobtainable tions include: How will monitoring data from wear-
physiological parameters. These advances allow us to ables and remote sensors interface with current
glean much more physiological data, and when com- systems? How secure can and will this communica-
bined with improved data analytic capabilities, are on tion be? Will the data from wireless sensors be accu-
the verge of creating a system that gives providers the rate and reliable enough to be the basis for remotely
ability to monitor patients accurately and continu- ordered medical interventions? And ultimately, will
ously, without geographic limitations. this newly acquired information reduce morbidity
Sophisticated monitoring is no longer restricted and mortality, or at least improve the quality of life?
to hospitalized patients. Advances in sensors, bat- One thing is for certain, clinicians will begin to
tery technology, and connectivity of previously acquire insight into patients’ daily lives, communicate
hospital-based “professional” monitors may now and connect with them in new ways, and for the first
be packaged into consumer products. The fitness time obtain data that are not limited to a hospital stay
and wellness industry has already showcased or brief office visit.
devices that can be purchased anywhere and quan-
tify heart rate, caloric expenditure, cardiac rhythm, Wireless Technologies in the
pulse oximetry, and other measures with an accu-
racy that approaches that of medical grade devices. Healthcare Sector
Further, startups and large companies alike are The most basic advantage of wireless monitors is
delving into the current gap between consumer the untethering of patients from the walls and
and healthcare products, developing noninvasive electrical outlets of clinical settings. Wires, cables,
and minimally invasive technologies and applica- and tubing form a web around patients that clin-
tions for home telemetry, glucose monitoring, icians often need to untangle, taking time and
control of infusion pumps, medication adherence, attention away from more critical patient care
and symptom tracking for various medical diseases activities.1 In high-stake locations like the operat-
(e.g., shortness of breath in lung disease). As these ing room or intensive care unit (ICU), this diverts
technologies emerge, the volume of data generated precious resources away from heuristic critical
is growing exponentially, and we are beginning to tasks. This web, sometimes also referred to as
see the early integration of these technologies into “spaghetti”, may pose potential physical hazards.
existing health systems’ care delivery. Tripping on tangled wires may remove intravenous
Implementation of these technologies faces a (IV) catheters and result in lost access, blood loss
number of new challenges. They not only have to be and air embolus, damage the monitors themselves,
dependable, reliable, and accurate, but they must also or harm patients and providers due to heavy falling
164
Future Monitoring Technologies
objects. Wireless technology should reduce these as AliveCor and ECG Check) for very short-term
hazards. (seconds to minutes) monitoring. These address the
Wireless technology may also reduce the risk of power consumption issue by using a second device,
complications that come with immobility (e.g., the smartphone, for data storage and transmission.
thromboembolism, delirium). The ability to ambulate Systems like AliveCor and ECG Check are able to
and not be subject to sedentary care due to monitor- record single lead ECG by sensing electrical activity
ing is associated with improved patient outcomes.2 In from a separate small recording device that connects
the ambulatory setting, wireless technology allows wirelessly to the smartphone. The user places the
patients to get out of bed and ambulate, and perform fingers of both hands on the recording device, which
activities of daily living while being monitored con- then displays the single lead ECG on the smartphone
tinually without interruption. screen. The device then interfaces with the smart-
Further development of wireless technology may phone and is able to provide a snapshot of the
eventually enable remote monitoring of patients from patient’s cardiac rhythm viewable on screen. The
their homes via monitors that communicate with a AliveCor and ECG Check applications are currently
telemedicine center through an app on patients’ able to detect atrial fibrillation without human need
smartphones. This, however, will depend on a num- for interpretation in real-time.
ber of advances in the monitors themselves and the Adhesive ECG patch devices are more complex and
supportive infrastructure. are composed of a sensor system, a microelectronic
circuit with a recorder and memory storage, and an
Cardiac Telemetry internal battery embedded in a relatively flexible syn-
Noninvasive electrocardiography (ECG) monitoring thetic matrix, resin, or other material. They are usually
devices use electrodes attached to the skin with elec- intended for medium-term use ranging from days to
tro-conductive gel. The electrodes record signals that several weeks. AECG system patches are leadless, mini-
are sent to a main console, which processes and optimally intrusive to daily activities, water-resistant, and
mizes the data and displays them on a screen in real- are designed for single use only. These systems have the
time, sometimes with a suggested interpretation. capability to detect clinically relevant arrhythmias and
Several wireless iterations of the ECG already exist, conduction system abnormalities and may soon sup-
and most involve communications between a fixed port analysis of the QT intervals and ST segment
base station and multiple mobile stations located changes. A limitation of these devices, however, is
within a coverage area. Battery-powered portable that they do not process data in real-time. Clinicians
wireless consoles perform most of the signal proces- and patients ultimately depend on the device company
sing locally and transmit results remotely. This pro- to process the data collected from these devices, and
cessing and transmission normally consume a lot of this may take weeks after data collection is complete.
power, relative to what can be supplied by current Some systems, such as the ZIO® XT system, actually
battery sources. Current research is geared toward require that the user return the device in a postage-paid
addressing tradeoffs between range limitations, envelope upon study completion, further adding to the
power consumption, and immobility of the local lag time between data collection, analysis, and physi-
fixed home station. cian interpretation. Such barriers make these devices
Recent years have introduced several wearable unsuitable for real-time monitoring.
ECG monitoring devices into mainstream use. There Further research is geared toward wearable mon-
are Adhesive ECG patches (AECG) with integrated itoring systems (WMS) such as textile-based smart
microelectronics, such as the ZIO® XT Patch and systems that collect ECG data from a wearable gar-
those incorporated into the NUVANTTM and ment. Thus far, ECG data have been collected using
SEEQTM mobile cardiac telemetry (MCT) systems. smart shirts, harnesses, and other body wear. These
These systems are indicated for short- to medium- systems are being optimized to record ECG signals
term (days to weeks) monitoring. They solve the pro- without the use of a gel, and reduce baseline noise and
blem of power consumption by storing the data motion artifacts thorough hardware-implemented
internally and only transmitting it at set intervals. high-pass, low-pass, and notch filters.3
There also are handheld smartphone-enabled systems Improvements in the sensors utilized in these systems
with electrode-embedded attachment modules (such are often coupled with software that harnesses the
processing power and communications capabilities of data, as most approaches to ECG acquisition require
existing mobile devices. One advantage of these sys- the user to touch an electrode on the wearable device.
tems is that by spreading the “leads” across a garment Additionally, in order to maintain accuracy, all exist-
the overall size of the device is increased, allowing ing models require intermittent calibration with stan-
more space to place the necessary batteries, transmit- dard blood pressure measurement.13,14 The variance
ters, and processors. of blood pressure due to other factors, such as changes
These systems are changing the landscape of car- in vasomotor tone and heart rate, require additional
diac diagnosis. Truly wireless mobile ECG monitor- parameters to be incorporated into the algorithm to
ing systems with real-time transmission may achieve accuracy.15,16 Currently, the accuracy of these
eventually address convenience and comfort, reduce technologies measured via the regression coefficient
cost and travel time, and enable immediate medical (R2) is low, with significant variations even at the
assistance in case of emergency. To meet a medical same activity level within the same subject.17 Future
standard as a clinically accepted monitoring system, advances will likely overcome these limitations and
these systems will have to incorporate high-quality produce the higher accuracy and precision necessary
real-time data acquisition, early detection of abnor- for a medical-grade device.
mal conditions, and accurate decision support. In There is also promise for invasive and minimally
order to be accepted by consumers, they must also invasive blood pressure monitoring devices.
be user-friendly and easy to wear. They must also have Microelectromechanical system (MEMS)-based pres-
fast processing, low power consumption, and small sure sensors have been implanted into the live canine
size, and therefore likely will be smartphone based. aorta and coupled with magnetic telemetry. Rozenman
et al. (2007) implanted a miniature device into both
Blood Pressure Monitoring animals and a patient cohort.18 The sensor was based
on an acoustically powered piezoelectric transducer
Wireless blood pressure monitoring is already in
with a custom-built, low-power control chip able to
mainstream use and automatic portable blood pres-
transmit pressure measurements continuously for 5–
sure cuffs are readily available to consumers. These
10s. Readings were simultaneously taken with a cathe-
battery powered devices often provide results com-
ter tip transducer advanced to the same location. Good
parable to those of hospital-grade devices and classic
agreement was found between the two methods, with a
auscultation. Data transmission to smartphones and
maximum deviation of less than 5 mmHg.
computer-based applications is relatively easy and
A similar protocol was used by Verdejo et al.
monitoring of large groups of patients in this inter-
(2007) to evaluate the accuracy of the
mittent manner is viable. Although chronic monitor-
CardioMEMSTMTM heart failure sensor that is also
ing of blood pressure in ambulatory patients provides
based on a MEMS pressure-sensitive capacitor.19 The
useful telehealth and alarm functionality, blood pres-
sensor was electromagnetically coupled to an external
sure may vary widely and continuous blood pressure
antenna that both powers the device and captures a
measurements, possibly using minimally invasive
resonant frequency that reflects the arterial pressure.
implants, might present a new opportunity to
The method was validated against standard care, and
improve management and outcomes.
in a randomized controlled trial proved to result in a
Wearable cuffless blood pressure monitoring
significant and large reduction in the rate of heart-
devices are currently under development. The under-
failure-related hospitalizations at 6 months.20
lying principle of this technology is based on the time
Although these implantable devices show promise,
it takes for a volume of blood to travel from the heart
they are quite invasive and therefore carry risk. A
to the measurement location, such as wrist.4–8
proposed alternative to intra-arterial pressure mea-
Algorithmic models predict blood pressure from the
surement is extra-arterial blood pressure monitoring
time delay, usually obtained from the cardiac electri-
using similar technology.21 This technology, however,
cal signal acquired by ECG and the recording device at
is quite nascent today.
the peripheral site such as a pulse oximeter.9–12 There
are several practical limitations, however, that must
be overcome before wearable cuffless blood pressure Pulse Oximetry
monitoring devices are adopted for clinical practice. Pulse oximetry is a robust tool that allows for the
Many existing systems do not provide continuous simultaneous measurement of perfusion, oxygenation,
166
Processor
Circuit board Antenna
Batteries
Charging
coil
Optodes
Figure 17.1 The Moxy Monitor System muscle oxygenation monitor consists of a spectrometer (light emitting diodes and photo detectors),
a battery for power, a microprocessor for operating the spectrometer, a memory chip for storing data, and a radio transmitter for sending live
data to other devices. From www.moxymonitor.com/device/.
heart rate, and heart rhythm. Wireless pulse oximeters power, which today has prohibitied it’s use in a small
have long been available to consumers and can be device such as a watch. Two companies, BSX Athletics
purchased for less than $20.00. Some of these devices (Austin, TX) and Moxy Monitor (Hutchinson, MN),
can be connected to smartphones and thereby allow are producing slightly larger wearable monitors using
transmission of data. near-infrared wavelengths to measure muscle oxyge-
The trend in pulse oximetry has been the conver- nation for athletes (Figure 17.1). Further, Apple and
sion of a stand-alone monitor into a wearable contin- Garmin are believed to be designing smaller sensors to
uous fitness tracker. Beginning in 2012, watches with integrate into their wearables. As this technology
optical heart rate monitors (OHRMs) have been develops it likely will become reliable enough for
directly available to consumers. Since then the indus- healthcare use.
try’s leading companies (i.e., Apple, Fitbit, and
Garmin) have steadily moved toward the addition of
pulse oximetry to their devices. Many of these moni- Consumer Wearables
tors currently use the principle of photoplethysmo- Until recently, cardiac telemetry and blood pres-
graphy (PPG). The PPG signal may contain valuable sure and pulse oximeter monitors were only avail-
information about the cardiovascular system, such as able to the medical community; however, these
oxygen saturation (SpO2), heart rate, blood pressure, modalities are increasingly available to consumers.
and respiratory rate. Devices utilizing PPG currently Rapid advances in semiconductor technology are
are able to measure heart rate; however, the green shrinking the gap between consumer and medical
light wavelength (typically 540 nanometers) used in grade devices in terms of diversity and quality.
PPG is readily absorbed by human tissue and is In fact, one of the biggest changes in the
affected by hydration, hemoglobin concentration, healthcare market has been the proliferation of
the ambient environment, and other factors. These direct-to-consumer marketing of wearable devices
limitations have reduced the ability to provide health- that currently range from simple fitness trackers to
care-grade continuous home oxygen saturation mon- more advanced devices such as continuous ECG
itoring by OHRMs. and pulse oximeter devices. Micro-sensors may be
Traditional wavelengths used in pulse oximetry, seamlessly integrated into textiles, and consumer
red light and infra-red light (600 and 900nm, respec- electronics are being embedded in clothes, watches,
tively), penetrate tissue with much more ease than belts, eye-glasses, and contact lenses. This direct-
green light wavelengths. However, continous device to-consumer monitoring device industry may soon
use at these wavelength requires a large supply of surpass that geared toward hospitals.
Smartwatches edema utilizing ankle circumference and tissue elasti-

city sensors. The circumference sensor is composed of
Smartwatches are currently the dominant wearable in
two magnets and an electromagnet that respond to
consumer health. To accurately monitor motion, 3-
the difference in position of these sensors as a result of
axis accelerometers, magnetometers and gyroscope
change in circumference. The tissue elasticity sensor
sensors obtain data, and in combination have the
requires the user to push on a small lever in the sock
capability to sense in 9 degrees of freedom.
that compresses the skin, and then measures elasticity
Although very small in size, each monitor contains
based on the applied force and produced tissue
several parts that typically use piezoelectric and capa-
displacement.24
citance sensors that utilize microscopic crystal struc-
The basic requirements of e-textile technologies
tures to generate a voltage current in response to
are that they must be washable and durable, connect
accelerative forces. Algorithms based on combined
to information repositories such as smartphones, and
motion-sensing are able to measure step count and
be highly reliable in order to be fit for medical grade
sometimes approximate distance traveled.
use. Ultra-low power advanced reduced instruction
There are a number of such products currently on
set computing machine microcontrollers (ARM
the market, and research has demonstrated these track-
MCUs) seem to be ideal for this emerging technology.
ers to be accurate with a step count error rate of less
Their architecture combines ultralow power, high-
than 1.3%.22 Many consumer wearables additionally
signal processing functionality, and ease of use.
measure heart rate utilizing the PPG signal with accep-
Fiberoptics are ideal, as they can connect multiple
table results for consumer use. As previously stated,
sensors and carry no risk of electroshock. Sensors,
PPG is based on reflection or transmission of light that
including light, heat, oximetry, pressure, and bioim-
changes with blood volume in the microvascular tissue.
pedance, are sufficiently developed in terms of func-
For example, Fitbit demonstrates a moderately strong
tionality, but must be adapted to reduce size and
correlation with ECG for heart rate (r = 0.83, p < 0.001),
power consumption to make them practical. These
with an average mean bias of −8.8 beats per minute
challenges are not insurmountable and smart clothing
(95% LoA 24.2, −41.8). These technologies, however,
could completely replace bedside monitoring in hos-
tend to be less accurate and reliable at heart rates greater
pitals, with shirts that track heart rate, temperature,
than 110 beats per minute.23 Further research is geared
position, blood pressure, and oxygen saturation.
toward inclusion of additional sensors to monitor SpO2,
estimate VO2 (typically via a variation of Cooper run
test or Rockport walk test), and heart rate variability. The
Ocular Devices
accuracy and reliability of these monitoring technolo- Other wearables in development have focused on
gies, however, must improve in order to gain acceptance incorporation of sensors into eyeglasses and contact
for healthcare use. lenses. Contact lens sensors have already been devel-
oped for analysis of glucose composition of tears as a
surrogate for blood glucose, and for the diagnosis of
E-textiles glaucoma by measuring intraocular pressure. With
E-textiles are a futuristic wearable modality. Special further advances in electronics, connectivity, polymer
weaving can be used to incorporate electrically con- synthesis, and micro-nanofabrication, contact lenses
ductive yarn into a fabric that becomes a “wearable with sensors to monitor concentrations of other ocu-
motherboard” and can connect multiple sensors on lar fluid contents such as electrolytes, lactate, urea,
the body (e.g., ECG electrodes and pulse oximetry) to and albumin are in development.25
signal acquisition bases. Many uses for e-textiles have There are a number of challenges to overcome
been proposed and some have entered into clinical before contact lens monitoring technologies become
trials. Products have been developed that integrate mainstream. The relationship between the concentra-
pressure sensors into a bedsheet or mattress in order tion of several analytes in tear fluid and blood has
to manage and prevent bed sores, bioimpedance sen- been difficult to establish and may not be consistent.
sors into a vest measuring extravascular lung water to Rates of tear production and type of tear production
track congestive heart failure, and smart-lights into can vary. The low sample volumes and low solute
therapy blankets for infants with neonatal jaundice. concentrations require high sensitivity for analysis.26
Washable stockings that measure lower extremity Wearing contact lenses itself alters the fluid profile in
168
a complex but consistent manner.27,28 Although these complications related to immobility such as pressure
challenges seem insurmountable, the prospect of a ulcers.
contact lens sensor that is powered externally and In addition to the creation of new types of moni-
permits wireless reporting using an auxiliary device tors, previously unconnected monitors are leveraging
has generated significant interest from Google, smartphones and other devices to become connected
Novartis, Microsoft, and other companies, and it is to the “internet of things.” Many glucometers, weight
an active area of research. scales, fitness trackers, and symptom trackers now
Sensors incorporated into eyeglasses and contact connect to iOS and Android systems, allowing these
lenses to track eye and head movements with accel- disparate data to be pooled and thus generating new
erometers and cameras are also under development. insights from their analysis. For example, Propeller
These wearables may not only have monitoring (Propeller Health, Madison, WI) is an attachment to
potential (e.g., hemi-neglect, seizures) but may also an inhaler that serves as a symptom tracker for asthma
project images to the wearer that can be used for and chronic obstructive pulmonary disease. The
patient education, therapy, or health notifications. Propeller device attaches to an already existing inha-
Overall, consumer wearables have the potential to ler, tracks its use, and then sends that information to
incorporate vital information about our patients into the application installed on a smartphone. Cloud
the current decision-making processes. These analytics track inhaler use (i.e., frequency, time of
technologies provide the opportunity to monitor day, and context), set medication reminders, and pro-
patients after discharge, perhaps reducing length of vide analytics to patients and providers. This allows
stay and alerting providers of complications before real-time tracking of disease exacerbation, and sim-
they are clinically apparent. This increase in data plifies previously complicated associations such as
collection may guide differential diagnosis, aid in frequency of inhaler use and the amount and type of
risk stratification, and assist in algorithm-based man- pollen predominant in a particular geographical loca-
agement in the hospital. Lastly, this new database of tion. This kind of tracking and analytics allows
information may be mined, and analytics developed Propeller’s system to reduce short-acting bronchodi-
that allow for development of better predictive mod- lator use,29 healthcare system utilization,30 and
els. For example, a simple Fitbit may be used to count asthma symptoms.31
steps per postoperative day, and those data may be Another example is a platform called Livongo
used to develop an independent marker of readiness (Mountainview, CA), a combination of a cellular-
for discharge or to set target activity levels for connected blood-sugar monitor and artificial intelli-
patients. gence–driven data analytics and coaching. Depending
on the blood glucose reading, the patient receives
Insight from Analytics and Technology smartphone-enabled suggestions tailored to the patient
and the situation, such as drinking a glass of fruit juice
Convergence if hypoglycemic or adjusting insulin dosing if a parti-
A powerful and rapidly evolving trend in monitoring cular time of the day always yields a high value. If the
technology is the application of analytic techniques to values stray too much from the norm, Livongo con-
data that are already being obtained via smartphones, nects the user to a live coach or refers them to specialist.
wearables, and other devices. One example is Leaf This integrated approach is improving patient care.
(Leaf Healthcare, Pleasonton, CA), a small, light- After one year, Livongo members show an average
weight, waterproof patch equipped with acceler- 18.4% decrease in the likelihood of having a day with
ometers that are very similar to those used in hypoglycemia (BG < 70 mg/dL) and an average 16.4%
smartphones. The patch has the capability of deter- decrease in hyperglycemia (BG > 180 mg/dL).32 It is
mining the position and movement of the patient. The also estimated to save as much as $100 per month per
data are transferred to a dashboard that provides patient to payers in healthcare utilization.33 Similar
patient positioning analytics and presents continuous platforms are in various stages of development for a
results to providers and caretakers in acute and wide variety of other chronic diseases.
chronic care facilities, alerting them to patients who Other approaches have focused on continuous
have been immobile for an excessive period of time. collection of data that were previously only available
The device aims to increase mobility and avoid as intermittent measurements, and then applying a
layer of data analytics. One disease that has been physical danger to a patient. An increasingly common
particularly amenable to this approach is diabetes concern is that pacemakers and insulin pumps con-
mellitus. Implantable sensors coupled with insulin nected to a wireless network can potentially be hacked
pumps have been developed and combined with and reprogrammed to change treatment or reported
advanced analytics to predict times of hypo- and results and cause harm. Thus, integration of future
hyperglycemia. The combined monitoring-delivery devices into the “internet of things” requires advanced
devices can then dose the insulin, as well as generate security protocols to ensure that the devices cannot be
nutrition and exercise recommendations that can help modified without appropriate approval.
prevent such episodes. Many systems are currently in Another potential challenge in implementing
development, with MiniMedTM by Medtronic these systems is their reliability. Dead zones and inter-
(Minneapolis, MN) approved by the FDA in 2017, ference zones will cause monitoring or at least data
and devices by BigFoot Medical (Milpitas, CA), reporting failures. Even the most widely used electro-
Tandem (San Diego, CA), Insulet (San Diego, CA), nic medical record in the United States has a frequent
and Beta Bionics (Boston, MA) currently undergoing downtime requirement, during which storage of data
clinical trials with planned approval and commercial is interrupted and paper orders and record keeping
availability in the next few years. In the meantime, a need to be implemented. A blackout or an internet
number of motivated patients have adapted delivery outage could impact tens of thousands of patients all
systems utilizing OpenAPS and Loop frameworks that at once. Thus, successful development of these tech-
enable patient-programmed closed-loop systems or nologies requires appropriate fail-safes and backup
automatic adjustment of output (insulin administra- systems that may include household batteries and
tion) based on inputted glucose readings. generators.
Altogether, consistently recorded measurements Additional challenges relate to integration of var-
drive analytic platforms to determine associations ious monitoring technologies and protocol standardi-
and insights that providers and patients previously zation. Most medical devices are simply not
did not consider. The power of these technologies interoperable because they provide output that is pro-
will be amplified when used in concert (e.g., weight prietary. Even publicly available healthcare taxonomies
scale, glucometer, calorie counter). used by these devices such as Systematized
Improved monitoring and advanced analytics will Nomenclature of Medicine Clinical Terms
drive telehealth and the shift from patient hospitaliza- (SNOMED), Logical Observation Identifiers Names
tion toward ambulatory care, and will ward off read- and Codes (LOINC), International Organization for
mission. Close follow up with simple remote video Standardization (ISO), and others are not designed to
monitoring session programs are already reducing be interoperable with one another. A single parameter,
hospital readmission for heart failure patients.34 In such as a heart rate or temperature, may be represented
select patient populations, total joint replacements are differently in each system, and thus each proprietary
being performed as same-day surgeries. Tele-rehabi- system must be translated into a common format such
litation programs34 have been implemented for total as HL7. Although widely accepted as a standard, HL7 is
knee replacement patients who have difficulty attend- also inherently inflexible, and each data system even in
ing in-person therapy sessions. This trend will con- a single hospital may still use a slightly different stan-
tinue and expand to many procedures and medical dard. Thus, all HL7 messages may need to pass through
diagnosis that previously required hospitalization. an interface engine that facilitates exchange, transla-
tion, and sharing of data, in order to integrate into the
Challenges electronic health record. All of these processes prevent
In recent years there have been great strides in the sharing of high-resolution data and hinder real-time
development of new sensors and wireless technolo- applications and advances.
gies. How these technologies will be used and inte- Initiatives such as Integrating the Healthcare
grated, however, remains to be seen. There are many Enterprise’s (IHE) Patient Care Device (PCD) and
potential barriers to adoption, including concerns the Fast Healthcare Interoperability Resources
about privacy and security. Wireless technologies cre- (FHIR) are advocating for new standards in intero-
ate a potential door for intruders which, in addition to perability and device communication. FHIR specifica-
posing privacy concerns, can actually result in tions are evolving, with significant changes proposed
170
to accompany each new version’s release. Vendors physiological signals including blood pressure.
and hospitals are therefore postponing elective adop- Conf Proc IEEE Eng Med Biol Soc 2014;2014:
tion of FHIR standards until the industry firmly com- 2286–9.
mits to a more permanent standard. 5. Kim J, Park J, Kim K, et al. Development of a
nonintrusive blood pressure estimation system for
computer users. Telemed J E Heal [Internet] 2007;13
Conclusions (1):57–64.
New healthcare monitoring technologies are devel- 6. Kim JS, Chee YJ, Park JW, Choi JW, Park KS. A new
oping at an increasingly rapid pace, mirroring the approach for non-intrusive monitoring of blood
larger overall technology revolution. New problems pressure on a toilet seat. Physiol Meas 2006;27(2):203–
including integration, stability, and hacking have 11.
accompanied these developments. A new set of 7. Baek HJ, Lee HB, Kim JS, et al. Nonintrusive biological
associated ethical issues is arising too. For example, signal monitoring in a car to evaluate a driver’s stress
if a patient is monitored at home several questions and health state. Telemed e-HEALTH 183 2009;15
quickly arise: Who should have access to the data – (2):182–9.
physician, healthcare system, employer, and/or the 8. Gu WB, Poon CCY, Leung HK, et al. A novel
health insurance company? What is the patient– method for the contactless and continuous
physician contract for acting on abnormal data? measurement of arterial blood pressure on a
sleeping bed. Conf Proc IEEE Eng Med Biol Soc
Does the patient have the right to share data that
2009;2009(c):6084–6.
are generated by (and at some expense to) one
provider and health system with other providers 9. Nye ER. The effect of blood pressure alteration on the
pulse wave velocity. Br Heart J 1964;26(2):261–5.
and health systems?
New technologies have been developed to include 10. Gribbin B, Steptoe A, Sleight P. Pulse wave velocity as a
sensors in medication tablets that monitor medication measure of blood pressure change. Psychophysiology
1976;13(1):86–90.
compliance that is known to be inadequate in many
medical diseases and in specific subsets of patients. 11. Ahmad S, Chen S, Soueidan K, et al.
Electrocardiogram-assisted blood pressure estimation.
With the viability of this technology, what will be the
IEEE Trans Biomed Eng 2012;59(3):608–18.
repercussions for non-compliant patients? Can a
health system participating in value-based care dis- 12. Lass J, Meigas K, Karai D, et al. Continuous blood
pressure monitoring during exercise using pulse wave
miss the patient from their practice? Can the insurer
transit time measurement. Conf Proc IEEE Eng Med
raise their rates or perhaps change their co-pay Biol Soc 2004;3:2239–42.
requirements for complications related to non-
13. McCarthy BM, O’Flynn B, Mathewson A. An
compliance? investigation of pulse transit time as a non-invasive
What is certain is that these devices will be com- blood pressure measurement method. J Phys Conf Ser
monplace in the very near future, their sophistication 2011;307(1): 012060.
will likely exceed most of our imaginations, the quan- 14. Singh RB, Cornélissen G, Weydahl A, et al. Circadian
tity of data generated will be enormous, and a new heart rate and blood pressure variability considered for
medical paradigm will be required. research and patient care. Int J Cardiol 2003;87(1):9–
28.
References 15. Liu Q, Yan BP, Yu CM, Zhang YT, Poon CC.
Attenuation of systolic blood pressure and pulse
1. Kalisch BJ, Lee S, Dabney BW. Outcomes of inpatient
transit time hysteresis during exercise and
mobilization: a literature review. J Clin Nurs 2014;23
recovery in cardiovascular patients. IEEE Trans
(11–12):1486–501.
Biomed Eng [Internet]. 2014;61(2):346–52.
2. Hofer I, Cannesson M. Is wireless the future of Available from: http://ieeexplore.ieee.org/docu
monitoring? Anesth Analg 2016;122(2):305–6. ment/6645391/
3. Baig MM, Gholamhosseini H, Connolly MJ. A 16. Wong YM, Zhang YT. The effects of exercises on the
comprehensive survey of wearable and wireless ECG relationship between pulse transit time and arterial
monitoring systems for older adults. Med Biol Eng blood pressure. 2005 IEEE Eng Med Biol 27th Annu
Comput 2013;51(5):485–95. Conf [Internet]. 2005;5:5576–8. Available from: http://
4. Thomas SS, Nathan V, Chengzhi, et al. BioWatch – ieeexplore.ieee.org/lpdocs/epic03/wrapper.htm?
a wrist watch based signal acquisition system for arnumber=1615748
17. Wong MY, Pickwell-MacPherson E, Zhang YT. The 26. Fullard RJ, Snyder C. Protein levels in
acute effects of running on blood pressure estimation nonstimulated and stimulated tears of normal
using pulse transit time in normotensive subjects. Eur J human subjects. Investig Ophthalmol Vis Sci
Appl Physiol [Internet]. 2009;107(2):169–75. Available 1990;31(6):1119–26.
from: http://link.springer.com/10.1007/s00421-009–1 27. Grus FH, Sabuncuo P, Augustin AJ. Quantitative
112-8 Analyse der Tränenproteinmuster bei
18. Rozenman Y, Schwartz RS, Shah H, Parikh KH. weichen Kontaktlinsen – Klinische Studie 1.
Wireless acoustic communication with a miniature 2001;239–42.
pressure sensor in the pulmonary artery for disease 28. Mann A, Tighe B. Contact lens interactions with the tear
surveillance and therapy of patients with congestive film. Exp Eye Res [Internet]. 2013;117:88–98.
heart failure. J Am Coll Cardiol 2007;49(7):784–9.
29. Merchant RK, Inamdar R, Quade RC. Effectiveness of
19. Verdejo HE, Castro PF, Concepción R, et al. population health management using the Propeller
Comparison of a radiofrequency-based wireless health asthma platform: a randomized clinical trial. J
pressure sensor to Swan-Ganz catheter and Allergy Clin Immunol Pract [Internet]. 2016;4(3):
echocardiography for ambulatory assessment of 455–63.
pulmonary artery pressure in heart failure. J Am Coll
Cardiol 2007;50(25):2375–82. 30. Merchant R, Tuffli M, Barrett M, Hogg C, Van Sickle
D. Interim results of the impact of a digital health
20. Abraham WT, Adamson PB, Bourge RC, et al. intervention on asthma healthcare utilization. J Allergy
Wireless pulmonary artery haemodynamic monitoring Clin Immunol [Internet]. 2017;139(2):AB250. Available
in chronic heart failure: a randomised controlled trial. from: http://linkinghub.elsevier.com/retrieve/pii/
Lancet 2011;377(9766):658–66. S0091674916323235
21. Potkay JA. Long term, implantable blood pressure 31. Barrett MA, Humblet O, Marcus JE, et al. Effect of a
monitoring systems. Biomed Microdevices 2008;10 mobile health, sensor-driven asthma management
(3):379–92. platform on asthma control. Ann Allergy, Asthma
22. Alharbi M, Bauman A, Neubeck L, Gallagher R. Immunol [Internet]. 2017;119(5):415–421.e1. Available
Validation of Fitbit-Flex as a measure of free-living from: http://linkinghub.elsevier.com/retrieve/pii/
physical activity in a community-based phase III S1081120617306415
cardiac rehabilitation population. Eur J Prev Cardiol 32. Downing J, Bollyky J, Schneider J. Use of a connected
2016;23(14):1476–85. glucose meter and certified diabetes educator coaching
23. Jo E, Lewis K, Directo D, Kim MJ, Dolezal BA. to decrease the likelihood of abnormal blood glucose
Validation of biofeedback wearables for excursions: the Livongo for Diabetes Program. J Med
photoplethysmographic heart rate tracking. J Sport Sci Internet Res 2017;19(7):1–6.
Med 2016;15(3):540–7. 33. June P, Health L. Livongo Clinical and Financial
24. De Rossi DE, Paradiso R. Future direction: E-textiles. Outcomes Report Diabetes to Live a Better Life. 2016;
In: Bonfiglio A, De Rossi D (eds) Wearable Monitoring (June).
Systems. Boston, MA: Springer, 2011. 34. Jiang S, Xiang J, Gao X, Guo K, Liu B. The comparison
25. Farandos NM, Yetisen AK, Monteiro MJ, Lowe of telerehabilitation and face-to-face rehabilitation
CR, Yun SH. Contact lens sensors in after total knee arthroplasty: a systematic review and
ocular diagnostics. Adv Healthc Mater 2015;4 meta-analysis. J Telemed Telecare [Internet]. 2016
(6):792–810. December 27;1357633X16686748.
172
Chapter
Downside and Risks of Digital Distractions
18 Peter J. Papadakos and Albert Yu
Introduction The practice of medicine in general has similarly

evolved to the point where the physician−patient
This chapter will focus on two issues that have arisen
relationship may never even occur in-person or
in our newly electronicized workplace: (1) digital dis-
require physical presence in the same city, state, or
traction and (2) alarm fatigue. Digital distraction due
country. Telemedicine allows for the care of our sick-
to personal electronic devices (PEDS) and, paradoxi-
est patients in intensive care units by doctors sitting in
cally, electronic medical records can undermine our
their own homes, or in monitoring centers, some-
ability to pay attention to one task for more than a few
times thousands of miles away. When there is intent
minutes. Alarm fatigue refers to a desensitization to
to provide in-person care and have an “old-fashioned”
alarm stimuli that clinicians experience in the context
physician−patient relationship, over the course of a
of frequent alarms and alerts, causing a delayed or
workday, an individual provider will interact and
missed response to an alarm. The combination of
respond to hundreds of alarms, emails, texts, and
digital distraction and alarm fatigue is particularly
alerts from their AIMS and Electronic Medical
disruptive and potentially injurious to patients.
Records (EMRs), as well as PEDs such as phones,
There has been a marked evolution in the modern
tablets, watches, and fitness trackers. From these
operating room (OR) over the last 20 years. Not so
devices a plethora of information related to patient
long ago there was genuine one-on-one engagement
care and personal life is constantly intermingled.
with patients. Anesthesiologists were literally tethered
These changes have occurred very rapidly over the
to their patient with a custom molded earpiece, which
past ten years and have had a dramatic effect on work-
itself was almost a rite of passage into the specialty.
place behavior, yet there has been little effort to quan-
Now, the term “earpiece” in the context of monitoring
tify and understand their impact on patient care or
respiration and breath sounds is likely unknown to
establish an appropriate “code of conduct.”
any graduate from this millennium, and would be
Anesthesiologists traditionally consider vigi-
more likely mistaken for an earbud. Patients were
lance as their primary responsibility. The word
routinely touched without gloves to assess tempera-
“Vigilance” appears over the picture of a light-
ture and the presence or absence of diaphoresis, and
house in the emblem of the American Society of
describing the patient as “cold and clammy” immedi-
Anesthesiologists, signifying the central element of
ately conveyed serious concern. With the typical
the specialty. Anesthesiologists, therefore, must
latex-free gloves worn to suppress transmission of
take the lead in understanding digital distraction
infection, this assessment is now near impossible.
and alarm fatigue, minimize their impact on
Written paper records in triplicate form required
patient care, and establish an appropriate culture
entry of vital signs, medications, notation of processes
of electronic etiquette. This requires that a new
and procedures, and other critical information. Now
educational experience be incorporated into the
nearly 80% of larger anesthesiology practices utilize
curriculum of all health personnel that would
Anesthesia Information Management Systems
likely need to evolve over time as new digital
(AIMS) that automatically download and chart vital
distractions emerge.
signs, allow quickly charted groups of phrases, and
Computers and smartphones (which are really just
use dozens to hundreds of click boxes, resulting in
computers more powerful than those that were used
charts that are exponentially larger (especially in
to land man on the moon shrunken into a handheld
intensive care units) but may actually convey less
device) give us a wide array of information that was
information.
simply not available as recently as 2008, when the first mortality. This behavior and the possible risk that
iPhone was released and the US government pushed it poses have given rise to the term distracted
for EMRs, resulting in the widespread appearance of doctoring.
computer workstations everywhere that patients are
cared for. This has resulted in information pushed to Distracted Doctoring
us and pulled by us, arriving in our virtual in-boxes
There is a growing literature that addiction to technol-
faster than was ever imaginable a decade ago. Pushed
ogy is endemic and as a result the physician−patient
amber alerts, weather warnings, breaking news, sales
relationship has decayed. In his 2011 New York Times
advertisements, journal article e-releases, hospital
front-page medical exposé, Matt Richtel wrote:7
announcements, and departmental updates are com-
ing in all day, every day. Our reliance on pulled Hospitals and doctors’ offices, hoping to curb med-
information, including drug dosages, guidelines, pro- ical error, have invested heavily to put computers,
tocols, scoring systems, and remote viewing of the smartphones and other devices into the hands of
EMR, is so total that without PEDs providing access medical staff for instant access to patient data,
drug information and case studies. But like many
to this information, current medical care might come
cures, this solution has come with an unintended
to a standstill. In addition to pure information, text side effect: doctors and nurses can be focused on the
messaging, Skype, Instagram, Snapchat, Facebook, screen and not the patient, even during moments of
and other social media designed to connect us critical care.
instantly to our family, friends, colleagues, and some-
times patients have also weaved their way into the Distracted doctoring has greatly affected the care of
work environment. patients. Until recently, the focus was on the impact of
Use of these applications is addictive, may result in EMRs and entry of data infringing on the patient
the creation and need to maintain a personal avatar (a −doctor relationship. Failure to make eye contact,
virtual digital identity), can interfere with job perfor- sitting with one’s back to the patient, and copying
mance, and may even promote dangerous levels of and pasting of prior notes all seemed to dramatically
inattentiveness and a complete lapse in the vigilance change the typical office and bedside encounter.
that anesthesiologists uphold as their main virtue. Doctors distracted by their electronic record keeping
PED “dependency” is a genuine problem. In 2017, abandon traditional verbal and body language cues
the average time spent on social media worldwide and often physical examination as well. PEDs, FOMO,
was reported to be 135 minutes per day, and young and the need to participate in social media in order to
adults have been reported to use their phones an market one’s practice and maintain a brand have
average of 5 hours a day, or almost one-third of all become part and parcel of daily practice.
their awake time! The current zeitgeist of device speed In 2012, 37% of residents and 12% of faculty
and dependence is termed “hyperculture,”1 to reflect admitted that during hospital rounds they read and
the rapid change in technology’s impact on society; responded to personal texts and e-mails.8 Even more
this of course also applies to the microcosm of concerning, 19% of residents and 12% of attendings
healthcare. believed they had missed important information
Physical ailments secondary to PED use like because of distraction from their smartphones. In
“blackberry thumb” seem antediluvian. This depen- 2020, it is highly likely these percentages have at
dency on PEDs has spawned a whole new set of least doubled.9 If critical care physicians and residents
psychological phenomena, including “nomophobia” are inappropriately focused on their PEDs while in the
(the fear of having no mobile phone handy),2,3 “phan- group setting of patient rounds, then lone anesthesia
tom vibration and ringing syndrome” (the sensation providers without peer pressure or group surveillance
that a phone has vibrated in our pocket or rung when are likely more affected. Anesthesia training would
in fact it has not),4 and “FOMO” (the fear of missing seem to require some education in the risk of dis-
out).5 While these new words and acronyms may be tracted doctoring and applicable mitigation strate-
amusing, it is possible that healthcare workers, gies.9 Another study published in 2012 reported that
including anesthesiologists, are “addicted”6 to their use of the anesthesia workstation for non-patient care
PEDs and that the addiction interferes with vigilance activity (e.g., internet surfing) occurred for 29% of the
and increases medical errors, morbidity, and total case time in cases with a duration of more than
174
4 hours.10 Again, in 2020, the combination of PEDs Changing the Culture

and workstation use for nonmedical activity has likely
Since PED technology is here to stay, an organization-
driven the proportion of time spent in this behavior
wide and profession engagement model can help pro-
much higher.
mote a culture of patient safety through improved
An environment that is laden with distractions is
focus on patient care. Widespread education on how
conducive to medical error.11 A good analogy is auto
technology can be addictive and detract from patient
safety – imagine driving at night on a rainy night
care must be introduced in school and in training
while at the same time your beeper is going off,
programs.9 Such training should incorporate simula-
someone is calling on your cell phone, the navigation
tion scenarios that illustrate to trainees the power of
system is informing you to turn, and your tire pres-
PEDs to distract and affect their ability to focus on the
sure alarm alerts you to the possibility of a flat tire.
task at hand. Electronic distractions need to be
Are you more prone to have an accident then, com-
included in all clinical scenarios so that trainees can
pared to driving on a sunny day with no electronic
see how they affect the vigilance and response time.
signals at all? That is not too dissimilar from the state
For example, a simulation scenario in which a care-
of affairs in our operating rooms today. A 2011 study
giver is receiving a barrage of text and social media
described distractions that were present during criti-
alerts as they are caring for a patient who is slowly
cal phases of anesthesia12 and included auditory
declining. Post-simulation analysis of how rapidly the
alarms, conversations, music, and other equipment
caregiver reacted to the changing clinical climate
sounds. Notably, conversation unrelated to the pro-
would be a very clear educational tool to point out
cedure occurred during 28 of 30 emergences, and loud
how PEDs affect focus and attention to the environ-
noise (> 70 dB) occurred during emergence in almost
ment around them. Through such education we can
every case! These distractions’ potential negative
train healthcare workers to use PED technology in a
impact on communication, concentration, and situa-
way that enhances their professional performance,
tional awareness was noted by the authors who cited
rather than diminishing their performance by distrac-
the aviation literature and suggested a “sterile cock-
tion. Healthcare institutions, facilities, departments,
pit” strategy to eliminate them. In the current envir-
professional groups, and insurers also need to address
onment there is an even wider array of distractions
this growing issue of patient safety, professionalism,
because of the ubiquitous presence of PEDs and the
legal implications, and costs – the caregiver is not the
stream of e-mail, texts, news alerts, tweets, and social
only stakeholder. Similar to the development of the
media that compete for attention.
“time out” in a past era, there is a need for increased
This focus on PEDs decreases reaction time and
awareness, education, and guidelines that cross pro-
degrades performance of tasks requiring attention,
fessional boundaries, with the overriding goal of
concentration, and decision-making.12 Furthermore,
patient safety.
distractions from PEDs can degrade the entire team’s
performance. According to Attri et al., smartphones
impair short-term memory, vigilance, and other Alarm Fatigue
aspects of cognitive performance.13 Electronic distractions not only come from devices
The problem of the distracted anesthesiologist is that we bring on our person into the clinical setting,
not new, but it has escalated over the years. Early but also alarms that have always had a place in the care
concerns revolved around reading and use of anesthe- of patients. The increased digitalization of medicine
sia OR computer stations for non-patient care activ- has led to an increase in monitors, LED screens,
ities. The ubiquitous presence of PEDs in the OR robots, and other devices in the operating room, all
raises greater concerns, as their portability presents of which compete for our attention. These electronic
an ever-present source of distraction as users are devices are an integral part of patient care, especially
typically in a state of readiness to respond quickly to in operating rooms and ICUs, but every advance is
every incoming signal. The nearly constant stream of accompanied by a new risk. This increase in the vari-
sounds and/or vibrations signaling incoming mes- ety and quantity of alarms, however, results in “alarm
sages demands the individual to attend to the PED fatigue,” desensitization to alarm stimuli that results
screen14–16 and distracts them from attending to the in delayed and missed responses. Alarms have been
patient and monitors. identified as a factor that increases the risk of adverse
patient events, due to auditory overload and its dis- physiologic and technical alarms can issue true and
tracting from other “important” alarms and duties. false positive alarms.21 False alarms can also occur
Clinicians may be exposed to hundreds of when the monitor receives signals that it interprets
alarms per patient per day, with 74–99% being “non- as pathological when the clinical condition of the
actionable,”17 meaning that there is no mechanical or patient is normal (e.g., a tachycardia alarm when
clinical response required of the clinician to resolve electrocautery is used during surgery). Nuisance
the alarming. As the awareness of alarm fatigue and alarms are nonactionable alarms that occur when a
distraction is increasing, however, so is the number of physiologic parameter exceeds set levels but does not
devices and alarms in our environment that we must require any action on the part of the clinician because
attend to, and the task to minimize their potential for it self-resolves in a clinically insignificant amount of
harm seems Sisyphean. While there is no gold stan- time (e.g., when a healthy patient’s heart rate transi-
dard for measuring alarm fatigue, most studies have ently dips below a preset of 60 BPM).22
used surrogate measures such as total alarm count or It is normal human behavior to deprioritize or
response time to alarm (i.e., alarm duration). There even disregard signals that in the past proved to be
are no studies, however, that have quantified the false. A clinician’s response to an alarm is related to
combined burden of digital distraction by PEDs and the perceived likelihood that the alarm represents a
EMRs coupled with alarm fatigue on clinical genuine urgency to be addressed.23,24 The positive
performance. predictive value (PPV) of an alarm is the probability
The Association for the Advancement of Medical that there is a true clinical event associated with the
Instrumentation (AAMI) Alarm Standards alarm: PPV = true positive alarms/(true positive
Committee defines alarm fatigue as a condition that alarms + false positive alarms).25 If the alarm has a
occurs when a user is desensitized by the presence of low PPV due to a large percentage of false alarms,
excessive alarm signals, many of which are non- clinicians will be desensitized over time and respond
actionable or in some cases false, and no or delayed less frequently and with decreased speed to these
response to the alarm signal occurs and harm to the alarms. An observational study performed in a post-
patient could result. Factors contributing to alarm operative care unit of nurse response times to pulse
fatigue can include false alarms, nonactionable oximetry desaturation alarms found that response
alarms, confusing alarms or inability to discriminate times increased (i.e., worsened) incrementally as the
alarm sounds from background noise, distressing number of nonactionable alarms increased in the pre-
sounds, and more.18 vious 120 minutes.26
In 2013, the Joint Commission identified the In the operating room multiple mental and physi-
recognition of alarm fatigue and the need for clinical cal tasks need to be performed simultaneously.
alarm management as a National Patient Safety Goal Primary tasks include continuous physical assess-
by 2016.19 This call for action was in response to 98 ment, operating room table optimization, draping,
alarm-related sentinel events reported from 2009 to administration of medications and fluids, and perfor-
2012, including 8 deaths and 13 patients with perma- mance of procedures. Secondary tasks include docu-
nent loss of function. The US Food and Drug mentation, responding to alarms, and team
Administration has reported over 566 deaths where communication. It has been shown that as primary
alarm fatigue was a contributing factor in just a five- workload increases secondary task performance
year time span.19 The Emergency Care Research decreases, and multitasking simply results in less
Institute has identified missed alarms as one of the attention paid to each individual task, especially
top health technology hazards since 2012 and con- when the PPV of an alarm is low. This is appropriate
tinues to track it as it remains in the top ten. as time and effort spent on responding to alarms
Clinical alarms usually indicate a derangement in would detract from the primary tasks.27,28 Alarms,
the patient’s physiologic state that requires attention however, are designed to be disturbing and get the
by a provider.20 Most physiologic alarms have their attention of the provider. Simply recognizing (e.g.,
threshold set to be sensitive in order to not miss a hearing or seeing), evaluating (e.g., what exactly is
clinically significant physiologic derangement. alarming – blood pressure monitor, respirator, pulse
Technical alarms indicate that something related to oximeter), and responding to an alarm can interfere
our biomedical equipment requires attention. Both with the primary tasks of the clinician and add to
176
workload, even if there is nothing required to be done abnormal and potentially dangerous zone. This will
other than attending to the alarm and resetting it. increase the PPV but potentially decrease the sensi-
Interrupted tasks also increase the chance for error. tivity of the alarm, as sensitivity = true positive
If a clinician is interrupted during a medication- alarms/(true positive alarms + false negative alarms);
related task, the chance of making an error after thus, a true actionable event becomes more likely to
resuming the task approaches 25%.25 Newer alarms not alarm. One should consider that risk really equals
with easily differentiable signals utilizing volume and probability multiplied by consequence in order to
pitch variation help stratify the alarm’s importance determine the setting of alarm thresholds, because
and alert the clinician when immediate attention is all “actionable” events are not equal. If the conse-
absolutely necessary. quence of an event not triggering an alarm is minimal
Optimizing appropriate use of alarms can of (e.g., an alarm designed to trigger if the urine output is
course potentially mitigate alarm fatigue. Many less than .5 ml/kg over one hour) this strategy is
newer monitors and devices have preprogrammed acceptable, but if the consequence is potentially lethal,
settings for specific clinical situations (e.g., pediatric even if the probability is low (e.g., abnormal electro-
profiles); however, many times the profiles are out- cardiogram may be interference or ventricular fibril-
dated, not selected, or not adjusted appropriately to lation), then this strategy is not acceptable. A
the patient’s baseline status. Alarm technology can randomized trial showed that reducing a pulse oxi-
leverage its sophistication to generate signals based meter alarm parameter from 90% to 85% resulted in
on physiological trends and/or multiple variables, 61% fewer alarms, but tripled the number of patients
reducing the false alarm rate, and increasing the clin- who indeed desaturated to less than 85%.29
ician’s response rate to alarms. For example, a pulse Other potential solutions to reduce the number of
oximetry waveform displayed in addition to the digi- nonactionable alarms is to apply a delay, so that self-
tal readout can easily clue the observer into the data’s resolving conditions are not audibly alarmed at all,
validity. while non-latching alarms are programmed to stop
Setting alarm parameters individualized for the signaling when the alarmed parameter returns to nor-
patient’s condition is one of the most frequently mal without requiring manual resetting. These poten-
described methods for decreasing alarm fatigue; it tial solutions may be used when the alarmed
decreases the total number of alarms.26 Allowing clin- parameter is expected to have a large proportion of
icians to individualize alarm settings to their patient is time-limited, clinically insignificant derangements. In
another possible means of decreasing the risk of alarm a prospective investigation of pulse oximetry, the
fatigue. In a systemic review published in 2016, the impact of lowering the SpO2 alarm trigger from
proportion of alarms requiring intervention ranged ≤ 90% to ≤ 85% and adding a time delay was investi-
from < 1% to 36%, with rates of < 1%–26% in adult gated. Simply lowering the trigger of course reduced
ICU settings and 20–36% in adult ward settings.27 A alarms, but patients experience triple the incidence of
study from the 1990s showed the PPV of a pulse SpO2 ≤ 85% and a six-fold increase at SpO2 ≤ 80%;
oximetry alarm and a ventilator alarm to be 1% and review of the data revealed that a 15-second alarm
3%, respectively, and demonstrated that less than 10% delay for SpO2 ≤ 90% might be an effective means of
of all alarms in a pediatric ICU setting influenced reducing false alarms without increasing the risk of
clinical care.27 Intelligently and consciously adjusting significant hypoxemia.29
alarm parameters for individual patients may lead to The next generation of monitors and alarms are
less total alarms and fewer false or nuisance alarms. being integrated into multi-parameter formats that
Examples of over-monitoring that can lead to diluting allow for programmed cross-checking of physiologic
the PPV of alarms include arrhythmia monitoring on parameters and decrease the number of false, non-
every floor patient and high oxygen saturation alarms actionable alarms. Intelligent alarms that are trig-
in adults on room air. The oxygen saturation alarms gered by trends rather than a single value and can
for premature cyanotic patients with congenital heart potentially decrease the number of nonactionable
disease need to be significantly different than that of a alarms (see Chapter 4 describing the Oxygen
healthy adult. Another potential means to decrease Reserve Index), especially when combined with pro-
alarm fatigue is to move the alarm parameter further grammed patient history, are being introduced into
away from the normal zone and more clearly into the new anesthesia systems. Other systems may use
novel means of notification such as voice alerts, as Reference

well as differential routing of alarm notifications
1. Bertman S. Hyperculture: The Human Cost of Speed,
whereby if an alarm is not acted upon by the first- Westport: Greenwood, 1998, p. 267.
tier responder it can be sent to a backup responder.
2. Farooqui IA, Pore P, Gothankar J. Nomophobia: an
Another option is to prioritize alarm signals by emerging issue in medical institutions? J Ment Health
assigning soft alerts to low-priority self-limiting 2017;22:1–4.
conditions and loud noxious alarms for significant
3. Bragazzi NL, Del Puente G. A proposal for including
conditions only. nomophobia in the new DSM-V. Psychol Res Behav
Optimizing the technology for the desired result Manag 2014;7:155–60.
is another way to decrease total alarms and false
4. Deb A. Phantom vibration and phantom ringing
alarms. For example, there are several ways among mobile phone users: a systematic review of
to monitor for respiratory insufficiency (see literature. Asia Pac Psychiatry 2015;7(3):231–9.
Chapter 16), including visual examination, end 5. Dossey L. FOMO, digital dementia and our dangerous
tidal CO2 detection, bioimpedance of the chest experiment. Explore 2015;10(2):69–73.
wall, respiratory-induced movement of the bed sur-
6. Roberts J, Pullig C, Manolis C. I need my smartphone:
face, pulse oximetry, and continuous acoustic a hierarchical model of personality and cell-phone
respiratory rate monitoring. In one study, acoustic addiction. Personal Individ Differ 2015;79:13–19.
monitoring produced fewer false alarms and
7. Richtel M. As doctors use more devices, potential for
detected more apnea events than did capnogra- distraction grows. New York Times 2011 December
phy,29 but good data comparing all of these poten- 15:A1,4. Available from www.NYtimes.com
tial monitoring modalities are lacking. 8. Katz-Sidlow RJ, Ludwig A, Miller S, Sidlow R.
One basic intervention to decrease nonaction- Smartphone use during inpatient attending rounds:
able alarms is to optimize physiologic monitors prevalence, patterns and potential for distraction. J
and check sensor placement. Changing the EKG Hosp Med 2012;7(8):595–9.
electrodes prior to the start of a procedure will 9. Papadakos PJ. Electronic etiquette: a curriculum for
potentially decrease artifacts from motion, electro- health professionals. In: Papadakos PJ, Bertman S.,
magnetic interference, and poor electrode contact. eds. Distracted Doctoring: Returning to Patient-
Use of disposable pulse oximeter probes that stick Centered Care in the Digital Age. Cham, Switzerland:
on the finger significantly reduces probe displace- Springer International Publishing AG, 2017;219–27.
ment and time spent trouble-shooting simple sen- 10. Wax DB, Lin HM, Reich DL. Intraoperative non-
sor disconnects. Disposable ECG wires for single- record-keeping usage of anesthesia information
patient use have been shown to lead to a signifi- management system workstations and associated
hemodynamic variability and aberrancies.
cant decrease in the total number of alarms, with-
out negative implications for patient safety or cost-
11. Kalisch BJ, Aebersold M. Interruptions and
effectiveness.31
multitasking in nursing care. Jt Comm J Qual Patient
Saf 2010;36(3):126–32.
Conclusion 12. Broom MA, Capek AL, Carachi P, Akeroyd MA,
The information and communications technology Hildutch G. Critical phase distractions in anesthesia
revolution has brought many advances that have and the sterile cockpit concept. Anaesthesia
led to improved patient care, but like many 2011;66:175–9.
advances in the past, they have come with 13. Attri J, Kheptarpal R, Chatrath V, Kaur J. Concerns
unforeseen consequences. Digital distractions are about usage of smartphones in the operating room
here to stay; EMRs and PEDs are not going and critical care scenario. Saudi J Anesth 2016;10
(1):87–94.
away, and the array of digital physiological moni-
tors and their associated alarms is only growing. 14. Piscotty R, Voepel-Lewis T, Lee SH, et al. To tweet or
The impact of digital distractions on physician not to tweet? Nurses, social media, and patient care.
Nurs Manag 2013;44(5):52–3.
behavior and the physician−patient relationship
and the risks posed by digital distraction and 15. Cohn TA, Shappell SA, Reeves ST, Boquet AJ.
alarm fatigue require further investigation, innova- Distracted doctoring: the role of personal
electronic devices in the operating room.
tion, and education.
178
Perioperative Care and Operating Room 23. Bliss JP, Dunn MC. Behavioral implications of alarm
Management 2018;10:10–13. mistrust as a function of task workload. Ergonomics
16. Papadakos PJ. Electronic distractions of the respiratory 2002;43(9):1283–300.
therapist and their impact on patient safety. Respir 24. Edworthy J, Hellier E. Alarms and human behavior:
Care 2014;59(8):1306–9. implications for medical alarms. Br J Anaesth 2006;97
17. Paine CW, Goel VV, Ely E, et al. Systematic review of (1):12–17.
physiologic monitor alarm characteristics and 25. Westbrook JI, Coiera E, Dundmuir WE. The impact of
pragmatic interventions to reduce alarm frequency. J interruptions on clinical task completion. Qual Saf
Hosp Med 2016;11(2):136–44. Healthcare 2010;19:284–9.
18. Winters BD, Cvach MM, Bonafide CP. 26. Graham KC, Cvach M. Monitor alarm fatigue:
Technologic distractions (part 2): a summary of standardizing use of physiologic monitors and decreasing
approaches to manage clinical alarms with intent nuisance alarms. Am J Crit Care 2010;19:28–34.
to reduce alarm fatigue. Crit Care Med 2018;46 27. Paine CW, Goel VV, Ely E, et al. Systematic review of
(1):130–7. physiologic monitor alarm characteristics and
19. The Joint Commission Sentinel Event Alert. Medical pragmatic interventions to reduce alarm frequency. J
device alarm safety in hospitals. Hosp Med 2016;11(2):136–44.
www.jointcommission.org/sea_issue_50/ 28. Lawless ST. Crying wolf: false alarms in a pediatric
20. Ruskin KJ, Hueske-Kraus D. Alarm fatigue: impacts intensive care unit. Crit Care Med 1994;22(6):981–5.
on patient safety. Curr Opin Anaesthesiol 2015;28 29. Rheineck-Leyssius AT, Kalkman CJ. Influence of pulse
(6):685–90. oximeter lower limit on the incidence of hypoxemia in
21. Manzey D, Gerals N, Wiczorek R. Decision-making the recovery room. Br J Anaesth 1997;79(4):460–4.
and response strategies interaction with alarms: the 30. Tanaka PP, Tanaka M, Drover DR. Detection of
impact of alarm reliability, availability of alarm validity respiratory compromise by acoustic monitoring,
information and workload. Ergonomics 2014;57 capnography and brain function monitoring during
(12):1833–55. monitored anesthesia care. J Clin Monit Comput
22. Bonafide CP, Lin R, Zander M, et al. Association 2014;28:561–6.
between exposure to nonactionable physiologic 31. Albert NM, Murry T, Bena JF, et al. Differences in
monitor alarms and response times in a alarm events between disposable and reusable
children’s hospital. J Hosp Med 2015;10 electrocardiography lead wires. Am J Crit Care 2015;24
(6):981–5. (1):67–74.
Index
AAA. See abdominal aortic aneurysm mid-esophageal aortic valve short- apnea, ultrasound, 72
AABB, clinical guidelines for RBC axis view, 38 ARDS. See acute respiratory distress
transfusion thresholds, 139 mid-esophageal ascending aortic syndrome
AAMI. See Association for the long-axis view, 38 arterial waveform analysis, 101–102,
Advancement of Medical mid-esophageal ascending aortic 104
Instrumentation short-axis view, 38 calibrated devices, 102
AARKS. See automated anesthesia mid-esophageal bi-caval view, 38 fluid responsiveness monitors,
record keepers mid-esophageal four-chamber view, 112, 115
abdominal aortic aneurysm (AAA), 96 37–38 lithium dilution, 102
abdominal ultrasound, 92, 98 mid-esophageal long-axis view, 38 transpulmonary thermodilution,
aortic, 96 mid-esophageal right ventricular 102
FAST exam, 92–93 inflow-outflow view, 38 uncalibrated devices, 102
key points, 97 mid-esophageal two-chamber empiric approach, 102
left upper quadrant (perisplenic) view, 38 fluid responsiveness monitors,
view, 93–94 transgastric mid-papillary short-axis 112, 115
limitations of, 95–96 view, 36 Pressure Recording Analytical
pelvic view, 94 analytics, technology convergence Method, 102
right upper quadrant with, 169–170 artifacts, lung ultrasound, 68–69, 70
(perihepatic) view, 93 Anesthesia Information Management atelectasis, 72, 74
subcostal view, 94–95 Systems (AIMS) consolidation, 72, 74
gallbladder, 97 digital distraction due to, 173 increased water, 71–73
pneumoperitoneum detection on, 97 performance improvement normal, 69–70
transducer selection, 92 using, 159 pleural effusions, 72, 73
ACLS. See advanced cardiac life decision support systems, 155, pneumothorax, 70–71, 72, 73
support 156–159 ASA. See American Society of
acoustic respiratory rate (RRa®) electronic anesthesia records, Anesthesiologists
monitoring, 152 154–155 assay method, blood loss assessment,
activated clotting time (ACT), 123 history of, 154 140, 145
acute respiratory distress syndrome anesthesiology Association for the Advancement of
(ARDS) digital distraction in, 173–174, 178 Medical Instrumentation
EVLW measurement in, 119 alarm fatigue, 173, 175–178 (AAMI), Alarm Standards
TTE of, 63 culture change to address, 175 Committee, 176
ultrasound, 72 distracted doctoring, 174–175 asthma, ultrasound, 72
advanced cardiac life support (ACLS), recent evolution in, 173 atelectasis, ultrasound, 72, 74
NIRS monitoring in, 26 vigilance in, 173 automated anesthesia record keepers
AI. See aortic insufficiency antibiotic administration, DSS for, 156 (AARKS), 154–155
AIMS. See Anesthesia Information anticoagulation monitoring, 123 autoregulation. See cerebral
Management Systems aorta, ultrasound evaluation, 96 autoregulation
alarm fatigue, 173, 175–178 aortic disease, TEE of, 51–52
American Society of Anesthesiologists aortic dissection, 96 BAT. See blunt abdominal trauma
(ASA), 173 aortic insufficiency (AI), 44, 45, 46 Bernoulli principle, 41
Standards for Basic Anesthetic aortic valve, TEE evaluation, 41–42 bias, 2–4
Monitoring, 139 aortic regurgitation, 44, 45, 46 “big data,”, 15–16
American Society of aortic stenosis, 42–43 billing, DSS for, 159
Echocardiography, basic TEE aortic valve velocity time integral bioimpedance
views, 36, 37 (VTIAV), 86–88 cardiac output measurement, 103,
descending aortic apical four- and five-chamber 104
long-axis view, 39 views, TTE, 56, 58–61, 62 respiratory volume monitoring,
descending aortic short-axis view, 39 apixaban, 124, 125 151–152
180
Index
bioreactance CCVmax. See common carotid maximal contractility, 81

cardiac output measurement, 103, velocity COPD. See chronic obstructive
104 CEA. See carotid endarterectomy pulmonary disease
fluid responsiveness, 112, 115 cEEG. See continuous cor pulmonale, 63, 64
blackberry thumb, 174 electroencephalography coronary ischemia, 41, 42
Bland–Altman analysis, 2–4, 6–8 central venous pressure (CVP), fluid corrective actions, DSS for, 156
blood flow. See cerebral blood flow administration in shock, 80–81 cost containment, DSS for, 158–159
blood loss. See intraoperative blood cerebral autoregulation, 10–11 COx. See cerebral oximetry index
loss monitoring of, 14 CPA. See cone and plate(let) analyzer
blood pressure NIRS monitoring index for, 25 CPB. See cardiopulmonary bypass
DSS for, 156 cerebral blood flow (CBF), 10–11, 12 CPP. See cerebral perfusion pressure
variation with positive pressure invasive monitoring, 12 crisis avoidance, DSS for, 156–157
ventilation, 110–111 noninvasive monitoring, 12–13 culture of patient safety, 175
variation with respiration, 110 cerebral microdialysis (MD), 14–15 CVP. See central venous pressure
wireless monitoring, 166 cerebral oximetry, 14, 20, 26 cyanosis, 148
blunt abdominal trauma (BAT), FAST in clinical practice, 24
exam, 92, 93 cardiopulmonary bypass, 24 dabigatran, 124–125
brain metabolism monitoring, 14–15 cardiopulmonary resuscitation, 26 data collection, 2
brain oxygen monitoring, 13. See also carotid endarterectomy, 24 decision support systems (DSS), 155
cerebral oximetry cerebral autoregulation index, 25 antibiotic administration, 156
invasive, 13–14 general surgery, 25 billing, 159
noninvasive, 14 orthopedic surgery, 24–25 blood pressure, 156
brain physiology, 10–11 postoperative cognitive decline, chart review, 156
25–26 compensation, 159
capnography, 150–151 traumatic brain injury, 26 compliance, 159
cardiac arrest, 64 interpretation, 22–23 corrective actions, 156
cardiac output (CO), fluid NIRS principles, 20–21, 22 cost containment, 158–159
responsiveness, 81 cerebral oximetry index (COx), 25 crisis avoidance, 156–157
cardiac output (CO) monitoring. See cerebral perfusion pressure (CPP), drug diversion, 157–158
also noninvasive cardiac output 10–11, 14 packages, 157
monitoring chart review, DSS for, 156 postoperative nausea and vomiting
validation studies chronic obstructive pulmonary disease prophylaxis, 156
background, 2–3 (COPD), 72 relief equity, 158
Bland–Altman analysis, 2–4, ClearSight System, 112, 115 resident monitoring, 158
6–8 clinical alarms, 176. See also alarm risk reduction, 157
measurement of output, 1–2 fatigue dependency, PED, 174
protocol design and data clinical examination, neurological descending aortic long-axis view, TEE,
collection, 2 injuries, 10 39
report of validation study data, CNAPTM, 112, 115 descending aortic short-axis view,
6–8 CO. See cardiac output TEE, 39
time plots and regression analysis, coagulation tests. See hematology desensitization, to alarm stimuli, 173,
5–6 Cochrane reviews, 1 175–178
trend analysis, 2–3, 4–5 common bile duct, 97 diabetes mellitus, 169–170
cardiac tamponade, 51, 63 common carotid artery (CCA), 88–89 digital distraction, 173–174,
cardiac telemetry, 165–166 common carotid maximal velocity 178
Cardio Q, 112–113, 115 (CCVmax), 88 alarm fatigue, 173, 175–178
cardiogenic shock, 63 compensation, DSS for, 159 culture change to address, 175
CardioMEMSTMTM heart failure compliance, DSS for, 159 distracted doctoring, 174–175
sensor, 166 computer stations, digital distraction dilute thrombin time assay, 124
cardiopulmonary bypass (CPB), 24 due to, 173–175 direct oral anticoagulants (DOACs),
cardiopulmonary resuscitation concordance analysis, 4, 5 124–125
NIRS monitoring in, 26 cone and plate(let) analyzer (CPA), 125 direct thrombin inhibitors (DTI),
TTE in, 64 consolidation, pulmonary, 72, 74 124–125
carotid endarterectomy (CEA), 24 consumer wearables, 167 distracted doctoring, 174–175
carotid velocity time integrals, 88 e-textiles, 168 distraction. See digital distraction
heart–lung interaction model, 88 ocular devices, 168–169 distributive shock, 63
image acquisition, 88, 89 smartwatches, 168 DOACs. See direct oral anticoagulants
passive leg raise test, 88–89 contact lens monitoring, 168–169 drug diversion, DSS for, 157–158
CBF. See cerebral blood flow continuous electroencephalography DSS. See decision support systems
CCA. See common carotid artery (cEEG), 15 DTI. See direct thrombin inhibitors
Index
dynamic fluid responsiveness estimated blood loss (EBL), formulas, noninvasive, 112–113, 115
monitors, 109, 111–112, 115 141–142, 145, 146 physiological basis of, 110–111
applications of e-textiles, 168 pitfalls and limitations of,
goal-directed therapy, 113–114 EVD. See external ventricular drain 114–115
intraoperative fluid management, evidence-based approach, to heart–lung interaction model, 80,
113 monitoring, 1 81–83
arterial line with calibration, 112, EVLW. See extravascular lung water passive leg raise test, 80, 83–84
115 evoked potentials (EPs), 15 in shock and hypotension, 80–81,
arterial line without calibration, 112, expired CO2 monitoring, 150–151 109
115 EXTEM. See extrinsic coagulation ultrasound techniques, 84, 89
noninvasive, 112–113, pathway assay of ROTEM® carotid velocity time integrals,
115 external ventricular drain (EVD), 11 88–89
physiological basis of, 110 extravascular lung water (EVLW), 117, FAST exam, 95
blood pressure and stroke volume 120–121 inferior vena cava diameter
variation, 110–111 measurement of, 118 evaluation, 84–86
fluid responsiveness assessment in ARDS, 119 left ventricular outflow tract/
using SPV, PPV, SVV, and PVi, in goal-directed therapy, 120 aortic valve velocity time
111 in septic shock, 120 integral evaluation, 86–88
pitfalls and limitations of, 114–115 transpulmonary thermodilution, focused assessment with sonography in
118–120 trauma (FAST), 92–93
Early Goal-Directed Therapy (EGDT) in weaning from mechanical key points, 97
trial, 80–81 ventilation, 120 left upper quadrant (perisplenic)
EBL. See estimated blood loss physiology of accumulation of, view, 93–94
ecarin clotting time (ECT), 124 117–118 limitations of, 95–96
ECG. See electrocardiography extrinsic coagulation pathway assay of pelvic view, 94
echocardiography. See transesophageal ROTEM® (EXTEM), 125 right upper quadrant (perihepatic)
echocardiography; eyeglass sensors, 168–169 view, 93
transthoracic subcostal view, 94–95
echocardiography FAC. See fractional area change FOMO, 174
ECT. See ecarin clotting time factor Xa inhibitors, 124, 125 formulas, blood loss calculations,
edoxaban, 124 false alarms, 176 141–142, 145, 146
EEG. See electroencephalography FAST. See focused assessment with four-quadrant plot, 4, 5
EGDT trial. See Early Goal-Directed sonography in trauma fractional area change (FAC)
Therapy trial Fick, Adolf Eugen, 100 left ventricular, 39, 40
electrical velocimetry, cardiac output Fick method, cardiac output right ventricular, 39–40, 41
measurement, 103 measurement with, 1, 100 Frank Starling Curve (FSC), 81
electrocardiography (ECG), wireless, Fitbit, 168 future monitoring technologies, 164,
165–166 FloTrac, 102, 112, 115 171
electroencephalography (EEG), 15 fluid analytics and technology
electronic anesthesia records, 154–155. abdominal, FAST exam, 92–94, convergence, 169–170
See also Anesthesia 95–96 challenges with, 170–171
Information Management lung consumer wearables, 167
Systems EVLW, 117–121 e-textiles, 168
electronic distractions. See digital ultrasound of, 71–73 ocular devices, 168–169
distraction fluid administration smartwatches, 168
electronic medical records (EMRs) dynamic fluid responsiveness wireless, 164–165
digital distraction due to, 173–175 monitors, 113–114 blood pressure monitoring, 166
performance improvement using, for shock, 80–81, 109 cardiac telemetry, 165–166
159 fluid responsiveness pulse oximetry, 166–167
decision support systems, 155, definition of, 81
156–159 dynamic monitors, 109, 111–112, gallbladder, ultrasound, 97
electronic anesthesia records, 115 GEM 4000, 142–143
154–155 arterial line with calibration, 112, general surgery, NIRS monitoring in,
history of, 154 115 25
electrophysiology, 15 arterial line without calibration, global thrombosis test (GTT), 125,
EMRs. See electronic medical records 112, 115 126
endogenous fluid challenge model, 80, goal-directed therapy with, goal-directed therapy
83–84 113–114 dynamic fluid responsiveness
EPs. See evoked potentials intraoperative fluid management monitors, 113–114
esophageal intubation, ultrasound, 72 with, 113 EVLW measurement in, 120
182
Index
gravimetric blood loss determination, DSS for, 156 IVC CI. See inferior vena cava
140–141, 145 fluid responsiveness in, 80–81, 109 collapsibility index
GTT. See global thrombosis test hypovolemic shock, 63 IVC DI. See inferior vena cava
distensibility index
Hb. See hemoglobin ICP. See intracranial pressure IVC diameter evaluation. See inferior
Hct. See hematocrit IMPACT-RTM, 125 vena cava diameter evaluation
heart–lung interaction model (HLIM) inferior vena cava collapsibility index
carotid velocity time integrals, 88 (IVC CI), 85–86 Joint Commission, National Patient
fluid responsiveness, 80, 81–83 inferior vena cava (IVC) diameter Safety Goals, 176
VTILVOT variation, 87–88 evaluation, 84 jugular venous oxygen saturation
hematocrit (Hct), blood loss collapsibility index, 85–86 (SvjO2), 13
calculations, 141–142, 145, distensibility index, 85, 86
146 image acquisition, 84–85 Leaf, 169
hematocrit (Hct) monitoring, 142, 145 inferior vena cava distensibility index left upper quadrant (LUQ), FAST
continuous real time, 143 (IVC DI), 85, 86 exam, 93–94
device overview, 142 Innovance® Platelet Function Analyzer left ventricle (LV), TEE evaluation, 39,
device performance, 142–143 (PFA-100/200), 130–131 40
transfusion decisions, 143 intracranial pressure (ICP), 10–11 left ventricular outflow tract velocity
hematology point-of-care tests, 123, intracranial pressure (ICP) time integral (VTILVOT), 86
133 monitoring, 11 heart–lung interaction model, 87–88
anticoagulation monitoring, 123 invasive, 11 image acquisition, 86–87
direct oral anticoagulant noninvasive, 12 mini-fluid bolus, 88
monitoring, 124–125 intraoperative blood loss, 139, 144 passive leg raise test, 88
direct thrombin inhibitor assay method, 140, 145 LiDCO system, 102, 112, 115
monitoring, 124–125 formulas, 141–142, 145, 146 limits of agreement, 2–3
global thrombosis test, 125, 126 gravimetric method, 140–141, 145 lithium dilution
IMPACT-RTM, 125 hemoglobin and hematocrit cardiac output measurement, 102
platelet function tests, 129–130 monitoring, 142, 145 fluid analysis, 112, 115
Innovance® PFA-100/200,130–131 continuous real time, 143 Livongo, 169
Multiplate®, 131–132 device overview, 142 lung ultrasound, 67, 77
PlateletWorks®, 132 device performance, 142–143 artifacts, 68–69, 70
TEG® and ROTEM®, 131, 132 transfusion decisions, 143 atelectasis, 72, 74
VerifyNow system, 130, 131 imaging analysis, 143–144 consolidation, 72, 74
viscoelastic coagulation tests, monitoring standard and practice increased water, 71–73
125–126 guidelines normal, 69–70
Quantra Hemostasis Analyzer, AABB, 139 pleural effusions, 72, 73
129 ASA, 139 pneumothorax, 70–71, 72, 73
ROTEM®, 126–128, 129 visual method, 139–140, 145 in perioperative care
Sonoclot, 128–129 intraoperative exam, lung ultrasound intraoperative exam, 75–76
TEG®, 126–128, 129 in, 75–76 postoperative exam, 76–77
hemodynamic monitoring. See intraoperative fluid management, 113 preoperative exam, 74–75
dynamic fluid responsiveness intraparenchymal catheters, 13–14 sonography for intubation, 72, 75
monitors intravascular volume status, 109, technique, 67–68, 69
hemodynamic resuscitation, 80–81, 111–112, 115 lung water, 71–73. See also
109 arterial line with calibration, 112, extravascular lung water
hemoglobin (Hb), blood loss 115 LUQ. See left upper quadrant
calculations, 141–142, 145, 146 arterial line without calibration, 112, LV. See left ventricle
hemoglobin (Hb) monitoring, 142, 145 115
continuous real time, 143 goal-directed therapy with, 113–114 mainstem intubation, ultrasound,
device overview, 142 intraoperative fluid management 72
device performance, 142–143 with, 113 MAP. See mean arterial pressure
transfusion decisions, 143 noninvasive, 112–113, 115 Masimo-Radical 7TM pulse oximeter,
hemorrhage. See intraoperative blood physiological basis of, 110–111 113, 115
loss pitfalls and limitations of, 114–115 McConnell’s sign, 95
hemostasis tests. See hematology intubation, lung ultrasound, 72, 75 MD. See cerebral microdialysis
heparin therapy, monitoring, 123 invasive brain oxygen monitoring, mean arterial pressure (MAP), 10–11
HLIM. See heart–lung interaction 13–14 mechanical ventilation, EVLW
model invasive CBF monitoring, 12 measurement in weaning from,
hypertension, DSS for, 156 invasive ICP monitoring, 11 120
hypotension i-STAT, 142–143 MEPs. See motor-evoked potentials
Index
metabolism, brain, 14–15 in clinical practice, 24 opioids, respiratory monitoring and,

mid-esophageal aortic valve short-axis cardiopulmonary bypass, 24 148
view, TEE, 38 cardiopulmonary resuscitation, 26 OR. See operating room
mid-esophageal ascending aortic long- carotid endarterectomy, 24 ORi. See Oxygen Reserve Index
axis view, TEE, 38 cerebral autoregulation index, 25 orthopedic surgery, NIRS monitoring
mid-esophageal ascending aortic general surgery, 25 in, 24–25
short-axis view, TEE, 38 orthopedic surgery, 24–25 ORx. See Oxygen Reactivity Index
mid-esophageal bi-caval view, TEE, 38 postoperative cognitive decline, OSA. See obstructive sleep apnea
mid-esophageal four-chamber view, 25–26 oximetry. See cerebral oximetry
TEE, 37–38 traumatic brain injury, 26 Oxygen Reactivity Index (ORx), 14
mid-esophageal long-axis view, TEE, interpretation, 22–23 Oxygen Reserve Index (ORi), 31–34
38 principles, 20–21, 22 oxygen saturation. See also regional
mid-esophageal right ventricular neurological injuries, monitoring cerebral oxygen saturation
inflow-outflow view, TEE, 38 paradigm for, 10 jugular venous, 13
mid-esophageal two-chamber view, neurological monitoring. See pulse, 30–34
TEE, 38 multimodal neurological
mitral valve, TEE evaluation, 44 monitoring PAC. See pulmonary artery catheter
mitral regurgitation, 45–48 NICE. See National Institute for parasternal long-axis view, TTE,
mitral stenosis, 44–45, 46 Clinical Excellence 55–57, 58
monitoring NICOMTM, 112, 115 parasternal short-axis view, TTE, 56,
digital distraction in, 173–174, 178 NIRS. See near-infrared spectroscopy 57–58
alarm fatigue, 173, 175–178 nomophobia, 174 partial pressure of oxygen (PbtO2), in
culture change to address, 175 noninvasive brain oxygen monitoring, brain, 13–14
distracted doctoring, 174–175 14. See also cerebral oximetry passive leg raise (PLR)
evidence-based approach to, 1 noninvasive cardiac output carotid velocity time integrals, 88–89
future technologies in, 164, 171 monitoring, 100, 103–105 fluid responsiveness, 80, 83–84
analytics and technology arterial waveform devices, 101–102, VTILVOT variation, 88
convergence, 169–170 104 PbtO2. See partial pressure of oxygen
challenges with, 170–171 calibrated versus uncalibrated, 102 PE. See percentage error; pulmonary
consumer wearables, 167–169 empiric approach, 102 embolism
wireless monitors, 164–167 lithium dilution, 102 PEDS. See personal electronic devices
recent evolution in, 173 Pressure Recording Analytical pelvic view, FAST, 94
vigilance in, 173 Method, 102 percentage error (PE), 2–4
Monroe-Kellie doctrine, 10 transpulmonary thermodilution, performance improvement, EMRs in,
motor-evoked potentials (MEPs), 15 102 159
Moxy Monitor System, 167 bioimpedance and bioreactance, decision support systems, 155,
multimodal neurological monitoring, 103, 104 156–159
10 photoplethysmography, 102–103, electronic anesthesia records,
“big data” in, 15–16 104 154–155
brain metabolism monitoring, thermodilution, 1–2, 100–101, history of, 154
14–15 104 pericardial disease, TEE of, 50
brain oxygen monitoring, 13–14 ultrasound-based techniques, 101, cardiac tamponade, 51
brain physiology, 10–11 104 pericardial effusion, 50–51
CBF monitoring, 12–13 noninvasive CBF monitoring, 12–13 pericardial effusion
cerebral autoregulation monitoring, noninvasive dynamic fluid FAST exam, 94, 95
14 responsiveness monitors, TEE of, 50–51
electrophysiology, 15 112–113, 115 TTE of, 63
ICP monitoring, 11–12 noninvasive ICP monitoring, 12 pericardial tamponade, 94–95
Multiple Electrode Aggregometry nuisance alarms, 176 perihepatic view, FAST, 93
(Multiplate®), 131–132 perioperative care, lung ultrasound in
myocardial ischemia obesity, respiratory monitoring and, intraoperative exam, 75–76
TEE assessment of, 41, 42 148 postoperative exam, 76–77
TTE assessment, 63 obstructive shock, 63, 64 preoperative exam, 74–75
obstructive sleep apnea (OSA), 148 sonography for intubation, 72,
National Institute for Clinical ocular devices, consumer wearable, 75
Excellence (NICE), 1 168–169 perisplenic view, FAST, 93–94
National Patient Safety Goals, Joint operating room (OR) personal electronic devices (PEDS),
Commission, 176 alarms in, 175–178 173–174
near-infrared spectroscopy (NIRS), 14, distracted doctoring in, 174–175 culture change to address, 175
20, 26 evolution of monitoring in, 173 distracted doctoring and, 174–175
184
Index
PFA-100/200,. See Innovance® Platelet positive predictive value (PPV), of noninvasive, 112–113, 115
Function Analyzer alarms, 176, 177 pitfalls and limitations of, 114–115
phantom vibration and ringing positive pressure ventilation (PPV) PVi. See pleth variability index
syndrome, 174 blood pressure and stroke volume PVPI. See pulmonary vascular
photometric analysis, blood loss, 140, variation with, 110–111 permeability index
145 heart–lung interaction with, 80,
photoplethysmography (PPG) 81–83 Quantra Hemostasis Analyzer, 129
cardiac output measurement, postoperative cognitive decline
102–103, 104 (POCD), 25–26 Radical-7 Pulse CO-Oximeter,
respiratory monitoring, 151 postoperative exam, lung ultrasound 142–143
phrenic nerve palsy, 72 in, 76–77 radionuclide imaging studies, cardiac
physician–patient relationship postoperative nausea and vomiting output measurement with, 1
distracted doctoring effects on, (PONV) prophylaxis, 156 red blood cell (RBC) transfusion, 139
174–175 PPG. See photoplethysmography reference method, cardiac output
recent evolution in, 173 PPV. See positive predictive value; monitoring, 1–2
physiologic alarms, 176 positive pressure ventilation; regional cerebral oxygen saturation
PiCCO system, 102, 112, 115 pulse pressure variation (rScO2)
PiCCO 2,119 PRAM. See Pressure Recording causes of decreased, 22
platelet function tests, 129–130 Analytical Method causes of increased, 22
Innovance® PFA-100/200,130–131 preload, SV response to, 81 regional cerebral oxygen saturation
Multiplate®, 131–132 preload responsiveness. See fluid (rScO2) monitoring, 20, 26
PlateletWorks®, 132 responsiveness in clinical practice, 24
TEG® and ROTEM®, 131, 132 preoperative exam, lung ultrasound in, cardiopulmonary bypass, 24
VerifyNow system, 130, 131 74–75 cardiopulmonary resuscitation, 26
platelet monitoring. See hematology pressure and flow assessment, TEE, 41 carotid endarterectomy, 24
PlateletWorks®, 132 aortic valve evaluation, 41–44, 45, 46 cerebral autoregulation index, 25
pleth variability index (PVi), 109, mitral valve evaluation, 44–48 general surgery, 25
111 tricuspid valve evaluation, 48–50 orthopedic surgery, 24–25
fluid responsiveness assessment Pressure Reactivity Index (PRx), 14 postoperative cognitive decline,
using, 111 Pressure Recording Analytical Method 25–26
goal-directed therapy using, (PRAM), 102 traumatic brain injury, 26
113–114 Propeller, 169 interpretation, 22–23
monitors displaying, 111–112 protocol design, 2 principles, 20–21, 22
arterial line with calibration, 112, PRx. See Pressure Reactivity Index regional wall motion abnormalities
115 pulmonary artery catheter (PAC), 1–2, (RWMA)
arterial line without calibration, 100–101 TEE assessment, 41, 42
112, 115 pulmonary bleb, 72 TTE assessment, 63
noninvasive, 112–113, 115 pulmonary consolidation, 72, 74 regression analysis, 5–6
pitfalls and limitations of, 114–115 pulmonary embolism (PE), 63, 64, 95 relief equity, DSS for, 158
pleural adhesions, 72 pulmonary vascular permeability report, of validation study data, 6–8
pleural effusions, 72, 73 index (PVPI), 119 resident monitoring, DSS for, 158
PLR. See passive leg raise pulse contour devices, 101–102, 104 respiration, blood pressure variation
pneumoperitoneum, 97 pulse oximetry, 30–31 with, 110
pneumothorax, 70–71, 72, 73 respiratory monitoring, 149–150 respiratory failure, 148
POCD. See postoperative cognitive wireless, 166–167 respiratory monitoring, 148–149, 152
decline pulse oxygen saturation (SpO2) acoustic respiratory rate monitoring,
point-of-care abdominal ultrasound. monitoring 152
See abdominal ultrasound Oxygen Reserve Index, 31–34 bioimpedance-based respiratory
point-of-care hematology. See pulse oximetry, 30–31 volume, 151–152
hematology point-of-care tests pulse pressure variation (PPV), 109, expired CO2, 150–151
point-of-care lung ultrasound. See lung 110–111 photoplethysmography, 151
ultrasound fluid responsiveness assessment pulse oximetry, 149–150
point-of-care TEE. See transesophageal using, 111 recent evolution in, 173
echocardiography goal-directed therapy using, respiratory variation model, 80, 81–83
point-of-care TTE. See transthoracic 113–114 respiratory volume monitoring
echocardiography monitors displaying, 111–112 (RVM), 151–152
polar plots, 2–3, 4–5 arterial line with calibration, 112, Resuscitation Council, 1
PONV prophylaxis. See postoperative 115 Richtel, Matt, 174
nausea and vomiting arterial line without calibration, right upper quadrant (RUQ), FAST
prophylaxis 112, 115 exam, 93
Index
right ventricle (RV) arterial line without calibration, platelet testing, 131, 132
TEE evaluation, 39–40, 41 112, 115 viscoelastic coagulation testing,
TTE evaluation, 63–64 noninvasive, 112–113, 115 126–128, 129
risk reduction, DSS for, 157 pitfalls and limitations of, 114–115 time plots, 5–6
rivaroxaban, 124, 125 with positive pressure ventilation, TPTD. See transpulmonary
ROTEM®. See thromboelastometry 110–111 thermodilution
RRa® monitoring. See acoustic study design, 2 transcranial Doppler (TCD), 12–13
respiratory rate monitoring subarachnoid hemorrhage (SAH), 11, transcutaneous CO2 monitoring
rScO2. See regional cerebral oxygen 12–14 (tcCO2), 151
saturation subcostal long- and short-axis views, transducer, ultrasound, 92
RUQ. See right upper quadrant TTE, 56, 60–62 transesophageal echocardiography
RV. See right ventricle subcostal view, FAST, 94–95 (TEE), 35, 52
RVM. See respiratory volume SV. See stroke volume American Society of
monitoring SvjO2. See jugular venous oxygen Echocardiography basic views,
RWMA. See regional wall motion saturation 36, 37
abnormalities SVV. See stroke volume variation descending aortic long-axis view,
systolic pressure variation (SPV), 109, 39
SAH. See subarachnoid hemorrhage 110–111 descending aortic short-axis view,
sample size, 2 fluid responsiveness assessment 39
sensitivity, of alarms, 177 using, 111 mid-esophageal aortic valve short-
septic shock, 80–81, 120 goal-directed therapy using, axis view, 38
shock 113–114 mid-esophageal ascending aortic
EVLW measurement in, 120 monitors displaying, 111–112 long-axis view, 38
fluid responsiveness in, 80–81, 109 arterial line with calibration, 112, mid-esophageal ascending aortic
TTE evaluation, 63, 64 115 short-axis view, 38
single bolus thermodilution, 1–2 arterial line without calibration, mid-esophageal bi-caval
smartphones, digital distraction due to, 112, 115 view, 38
173–174 noninvasive, 112–113, 115 mid-esophageal four-chamber
smartwatches, 168 pitfalls and limitations of, 114–115 view, 37–38
somatosensory-evoked potentials mid-esophageal long-axis view, 38
(SSEPs), 15 TBI. See traumatic brain injury mid-esophageal right ventricular
Sonoclot®, 123, 128–129 tcCO2. See transcutaneous CO2 inflow-outflow view, 38
SpO2 monitoring. See pulse oxygen monitoring mid-esophageal two-chamber
saturation monitoring TCD. See transcranial Doppler view, 38
SPV. See systolic pressure variation TDF. See thermal diffusion flowmetry transgastric mid-papillary short-
SSEPs. See somatosensory-evoked TEB. See thoracic electrical axis view, 36
potentials bioimpedance aortic disease on, 51–52
Standards for Basic Anesthetic technical alarms, 176 complications and
Monitoring, ASA, 139 technology. See digital distraction; contraindications, 36
statistics. See validation statistics future monitoring technologies coronary ischemia assessment, 41,
Stewart–Hamilton equation, 100–101 TEE. See transesophageal 42
stroke volume (SV), fluid echocardiography pericardial disease on, 50
responsiveness, 81 TEG. See thromboelastogram® cardiac tamponade, 51
carotid velocity time integrals, 88–89 telemedicine, 170, 173 pericardial effusion, 50–51
heart–lung interaction model, 80, thermal diffusion flowmetry (TDF), 12 pressures and flow assessment, 41
81–83 thermodilution. See also aortic valve evaluation, 41–44, 45,
inferior vena cava diameter transpulmonary 46
evaluation, 84–86 thermodilution mitral valve evaluation, 44–48
left ventricular outflow tract/aortic cardiac output measurement, 1–2, tricuspid valve evaluation, 48–50
valve velocity time integral 100–101, 104 technical concepts, 35–36
evaluation, 86–88 fluid responsiveness, 112, 115 ventricular evaluation
passive leg raise test, 80, 83–84 thoracic electrical bioimpedance left ventricular size and function,
stroke volume variation (SVV), 109 (TEB), CO measurement, 103, 39, 40
fluid responsiveness assessment 104 right ventricular size and
using, 111 thromboelastogram® (TEG) function, 39–40, 41
goal-directed therapy using, anticoagulation monitoring, 123 transfusion decisions
113–114 platelet testing, 131, 132 clinical guidelines for RBC
monitors displaying, 111–112 viscoelastic coagulation testing, thresholds, 139
arterial line with calibration, 112, 126–128, 129 hemoglobin and hematocrit
115 thromboelastometry (ROTEM®) monitoring, 143
186
Index
transgastric mid-papillary short-axis Cardio Q probe, 112–113, 115 non-cardiac output studies, 8
view, TEE, 36 fluid responsiveness assessment, 84, ventricles
transpulmonary thermodilution 89 TEE evaluation, 39–40, 41
(TPTD), 102 carotid velocity time integrals, TTE evaluation, 63–64
cardiac output measurement, 1–2 88–89 VerifyNow system, 130,
EVLW measurement FAST exam, 95 131
in clinical practice, 119–120 inferior vena cava diameter vigilance, 173
principles of, 118–119 evaluation, 84–86 Vigileo. See FloTrac
transthoracic echocardiography left ventricular outflow tract/ viscoelastic coagulation tests, 125–126
(TTE), 55 aortic valve velocity time Quantra Hemostasis Analyzer, 129
views and possible assessments, 55 integral evaluation, 86–88 ROTEM®, 126–128, 129
apical four- and five-chamber noninvasive CO measurement, 101, Sonoclot, 128–129
views, 56, 58–61, 62 104 TEG®, 126–128, 129
parasternal long-axis view, 55–57, noninvasive ICP and CBF visual assessment, blood loss, 139–140,
58 monitoring, 12–13 145
parasternal short-axis view, 56, urinary bladder, FAST exam, 94 volumetric monitoring. See dynamic
57–58 fluid responsiveness monitors
subcostal long- and short-axis validation statistics, 1 VolumeView/EV 1000®, 119
views, 56, 60–62 cardiac output monitoring VTIAV. See aortic valve velocity time
traumatic brain injury (TBI), 11, background, 2–3 integral
13–14, 26 Bland–Altman analysis, 2–4, 6–8 VTILVOT. See left ventricular outflow
trend analysis, 2–3, 4–5 measurement of output, 1–2 tract velocity time integral
tricuspid valve, TEE evaluation, 48–50 protocol design and data
Triton L&D, 141 collection, 2 wearables. See consumer wearables
Triton OR System, 143–144 report of validation study data, wireless monitoring technologies,
TTE. See transthoracic 6–8 164–165
echocardiography time plots and regression analysis, blood pressure monitoring, 166
5–6 cardiac telemetry, 165–166
ultrasound. See also specific types trend analysis, 2–3, 4–5 pulse oximetry, 166–167
Figure 3.2 CASMED FORESIGHT™ monitor, cables, and oximetry
probes. *Printed with permission from the manufacturer.
Figure 3.3 Somanetics INVOS™ cerebral oximetry monitor.

*Printed with permission from the manufacturer.
Figure 3.5 Nonin EQUANOX™ cerebral oximetry sensors. *Printed

with permission from the manufacturer.
Deep detector
Shallow
detector
emitter
~2.5 cm Depth of penetration
Figure 3.4 CASMED FORESIGHT™ oximetry probe and depth of

penetration. *Printed with permission from the manufacturer.
Segmental model of the left ventricle
Four-chamber view Two-chamber view Long-axis view Mid-short axis view Basal short-axis view
4
10
11 3è 5
9
12
6 1 8 2 6
3 4
7
2 1
10 7 5
9 12
RCA RCA or CX
13 11 8
14 16 15 LAD LAD or CX
LCX RCA or LAD
17 17
15 or 16 17 13 or 14
1. Basal anterior 7. Mid-anterior 13. Apical anterior

2. Basal anteroseptal 8. Mid-anteroseptal 14. Apical septal
3. Basal inferoseptal 9. Mid-inferoseptal 15. Apical inferior
4. Basal inferior 10. Mid-inferior 16. Apical lateral
5. Basal inferolateral 11. Mid-inferolaterale 17. Apex
6. Basal anterolateral 12. Mid-anterolateral
Figure 5.4 was taken from Badano LP, Picano E. (2015) Standardized Myocardial Segmentation of the Left Ventricle. In: Stress
Echocardiography. Springer, Cham.
Figure 5.7 Vena contra aortic valve. Mid-esophageal aortic valve long axis. The vena contracta may be used to differentiate among degrees
of aortic regurgitation. The image on the left has a narrow vena contracta, which is associated with a mild degree of aortic regurgitation. By
contrast, the large vena contracta on the right is associated with a severe degree of aortic regurgitation.
Figure 5.8 Top row: Mid-esophageal four-chamber view. The mitral valve is severely stenotic with severe calcification of the annulus and
leaflets. Bottom row: Transmitral Doppler spectrum. The panel on the left is the normal transmitral Doppler flow and the panel on the right is
the transmitral flow in the presence of mitral stenosis.
Figure 5.9 Mitral regurgitation. The top row illustrates a Barlow’s mitral valve with multiple billowing and prolapsed segments. The second
row illustrates a prolapsed posterior leaflet. The posterior leaflet is above the level of the mitral annulus. By contrast, a flail segment is illustrated
in the bottom row, with the mitral valve segment pointing toward the atrium.
Figure 5.10 Functional tricuspid
regurgitation. A mid-esophageal four-
chamber view demonstrates a large
functional tricuspid regurgitant jet. The
vena contracta is measured as the jet
emanates from the right ventricle.
RV
IVS
E A Anterior leaflet
Poterior leaflet
LVPW
Figure 6.1 E-point septal separation (EPSS) in the parasternal long-axis view of a patient with severe left ventricular systolic dysfunction. EPSS
> 8 mm is associated with systolic dysfunction. IVS, interventricular septum; LVPW, left ventricular posterior wall; RV, right ventricle. Reprinted
with permission from McKaigney CJ, Krantz MJ, La Rocque CL, et al. E-point septal separation: a bedside tool for emergency physician
assessment of left ventricular ejection fraction. Am J Emerg Med 2014;32:493–7.
Figure 6.2(b) Apical five-chamber view
showing pulsed-wave Doppler (PWD) at the
left ventricular outflow tract (LVOT). The
velocity-time integral (VTI) function on the
ultrasound machine traces the area under
the PWD curve, which, when multiplied by
the area of the LVOT obtained in the
parasternal long-axis view, allows
calculation of left ventricular stroke volume.
Stroke volume multiplied by heart rate
determines cardiac output.

with pulsed-wave Doppler (PWD) across the
mitral valve showing biphasic diastolic flow.
The E wave represents early passive filling,
whereas the A wave shows atrial
contraction during late diastole. Peak E- and
A-wave velocities are used to calculate the
E/A ratio.
with tissue Doppler imaging (TDI) of the
mitral annulus showing early (e’) and late (a’)
excursion of the mitral annulus during
diastole. The E/e’ ratio utilizing peak e’
velocity is another measure of left
ventricular diastolic function.
Figure 6.5 Continuous-wave Doppler

(CWD) of tricuspid regurgitation (TR) in the
apical four-chamber view. Maximal TR jet
velocity can be used to calculate right
ventricular systolic pressure using the
simplified Bernoulli equation and right atrial
pressure.
Figure 8.3 Illustration of the heart–lung interaction with positive
pressure ventilation in a hypovolemic patient. (1) As the pressure
increases with the initiation of a mechanical breath, the flow to the RA
through the vena cava decreases significantly. (2) With a minimal delay
the flow in the main pulmonary artery decreases as well, secondary to
both an afterload increase and preload decrease. (3) The LV SV initially
increases as the pulmonary capillaries are drained by the increased
intrathoracic pressure and increase the LV preload transiently, however,
thereafter preload drops significantly and LV SV drops (red bars).
Source: Michard F. Changes in arterial pressure during mechanical
ventilation. Anesthesiology 2005; 103(2):419–28.
Figure 8.5 Subcostal view required for measurement of the IVC diameter. IVC= inferior vena cava, RA = right atrium, HV = hepatic vein, L =
Liver, Ao = aorta, ST = soft tissue layer. (a) Subcostal long axis view of the IVC: The IVC is directly attached to the liver parenchyma, receives
blood from the hepatic vein (blue arrow) just distal to the diaphragm, and clearly drains into the right atrium; the measurement of the diameter
is performed within 3 cm distal to the diaphragm and distal to the HV inlet. (b) It is important to differentiate the IVC (blue dot) from the
descending aorta (red dot). This is done by tilting the ultrasound beam to the patient’s right. (c) The resulting view is a subcostal long-axis view
of the descending aorta. The aorta is clearly separated from the liver by a soft tissue layer and is not connected to a cardiac chamber at this
level.
Figure 8.7 Acquisition of the LVOT VTI/ Vmax. (a) The apical five-chamber view is obtained with the LVOT (*) positioned in the center of the
image. (b) The pulsed wave Doppler function is activated and the sample volume (SVol) is placed into the LVOT just above the aortic valve. (c)
The LVOT is represented by the yellow quadrangle and the aortic valve is highlighted in blue. Once in pulsed wave Doppler mode, the
envelopes below the time axis (red dotted line) can be highlighted and the area under the curve can be measured at the point of expiration
and inspiration.
Figure 8.8 Acquisition of the CCA Vmax. As
indicated in the 2D image at the top of the
image, the sample volume is placed in the
middle of the artery and at an angle less
than 60 degrees to the blood flow. Once the
Doppler signal has been recorded the
maximal and minimal flow velocities are
measured, in this case 114,5 cm/s and
132 cm/s27.
Back Surgical Incision Active Alerts
MAC* = 0.7
with Propofol Hypotension: MAP = 52
Patient Drake
146729674 Ventilated Cumulative time for MAP <
Age: 68 RR = 12 bpm 55 = 22 minutes.
Weight (kg): 88.5 PIP = 24 cm H20
Ideal Weight: 48.0 PEEP = 6.0 mm H20
TV act = 350 ml
U-OR 05
ASA 4 CO2 SpO2
Total Colectomy 100
Case Duration: 4:00 50 Pa 90
ce
r
30
Progress 80% 40.0 96%

mmHg
I/O Balance: –1153
Infectious Diseases HR = 90 bpm [ekg]

BP Art Line [7s]
SPV = 6 [8s]
67 / 45 MAP = 52
Normal Volume
PreOp Pain Score: 4
Temp INR Plat Hct Est K+ Glu
Infusions 200
1.5 450 40 5.0
37 30 140 CAUTION: Check monitors and
Low MAP Alert < 55 36 1.2 4.0
35 1.0 150 20 3.0 medical record before making
70
medical decisions.
36.8 1.0 90 32 28 4.0 120
Creat = 1.55 500 ml
[2d] [4h] [3h] [2d] [45m]
[4h] 150 ml/hr More Info Emergency Refs
2 ml/kg/hr
Figure 16.1 An integrated, graphical, anesthesia monitoring and clinical decision support system package.
Source: Kruger GH, Tremper KK. Advanced integrated real-time clinical displays. Anesthesiol Clin 2011;29: 487–504.
Kheterpal S, Shanks A, Tremper KK. Impact of a Novel Multiparameter Decision Support System on Intraoperative Processes of Care and
Postoperative Outcomes. Anesthesiology. 2018; 128:272–282.

Modern Monitoring in Anesthesiology and Perioperative Care 2020

Uploaded by

Copyright:

Available Formats

Modern Monitoring in Anesthesiology and Perioperative Care 2020

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

Modern Monitoring in Anesthesiology and Perioperative Care 2020

Uploaded by

Copyright:

Available Formats

Freemedicalbooks4download

Suzan Uysal, PhD

Cambridge University Press is part of the University of Cambridge.

Neha S. Dangayach, MD, MSCR Ronald A. Kahn, MD

College of Physicians & Surgeons of Columbia Department of Anesthesiology and Perioperative

Torsten Loop, MD Linda Shore-Lesserson, MD, FAHA, FASE

David B. Wax, MD University of Rochester

1 and Monitoring Applications

Introduction Instead, at-a-distance (e.g., transthoracic Doppler)

Scatter plot Bland–Altman plot

Four-quadrant plot 90 Polar plot – half moon

ΔCl (litre.min–1 min–2) [i.e., mean of ΔCluscom & ΔCOcardioq]

–0.5 0.5 –15

2 Samson Zarbiv, Erica Fagleman, and Neha S. Dangayach

Introduction The Monroe–Kellie Doctrine

is accomplished by vasoconstriction or vasodilation of as intracranial hemorrhage, seizures, and catheter-

Noninvasive Intracranial Pressure Invasive Cerebral Blood Flow Monitoring

3 Benjamin Salter and Elvera Baron

Introduction Cerebral Oximetry

Figure 3.2 CASMED FORESIGHT™ monitor, cables, and oximetry

BOX 3.1 Common Causes of Increased rScO2

BOX 3.2 Common Causes of Decreased rScO2

Head position Unilateral Reposition or replace

Proposed algorithm for the treatment fo cerebral desaturations

Figure 3.6 Proposed algorithm for the treatment of cerebral desaturations.

curve in the PaO2 of 100–200 mmHg range provides a

80 no usable absorbance signals have been detected.

Figure 4.2 The ORi and SpO2 values at

50 100 150 200 240

Introduction Ultrasound waves are reﬂections when the

(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

Figure 5.2 Calculation of fractional area

Assessment of Coronary Ischemia ventricle and classifying it as normal, hypokinetic,

Segmental model of the left ventricle

1. Basal anterior 7. Mid-anterior 13. Apical anterior

Table 5.1 Summary of aortic stenosis.

Aortic sclerosis Mild Moderate Severe

Figure 5.6 Doppler spectrum through a

Aortic Regurgitation The vena contracta is the narrowest portion of a

Table 5.2 Quantiﬁcation of aortic regurgitation.

Mild Moderate Severe

* At Nyquist limits of 50–60 cm/sec

Table 5.3 Quantiﬁcation of mitral stenosis.

Mild Moderate Severe

Table 5.4 Summary of mitral regurgitation (data from reference 17).

Mild Moderate Severe

* At Nyquist limits of 50–60 cm/sec.

Table 5.5 Quantiﬁcation of tricuspid regurgitation.

Mild Moderate Severe

* At Nyquist limits of 50–60 cm/sec.

Figure 5.10 Functional tricuspid

Figure 5.11 Transgastric short-axis view

Figure 5.12 Aortic dissection. An aortic

Cardiovascular Computed Tomography. J Am Soc Society of Cardiovascular Anesthesiologists; Society

Introduction blood ﬂow at a speciﬁc site, but it is limited by its

PLAX PSAX A4 C A5 C Subcostal IVC

Figure 6.2(a) Close-up of the left

Figure 6.2(b) Apical ﬁve-chamber view