Praise

for The Ventilator Book from Amazon readers:

"It is a kitschy, fun read (never thought this was possible for something as dry as changing vent settings) -
would be a great start for an intern, probably more that you would need to know. Reviews setting changes in
a fun, memorable way."
—WonkaTron

"This is one of the best books I have ever read on ventilators. It's like a running commentary. It's concise,
clear and full of realistic examples introduced at the right time (just before the concepts make you to start
scratching your head). Of course as Dr. Owens admits, there are many more detailed books on mechanical
ventilation which you can read for more knowledge. This book is so "down to earth" that any beginner can
make sense out of it and any expert would agree with what I just wrote above."
—Avatar

"Every resident should make their lives easier and get a copy of this book.
I'm an RT and this makes perfect sense. It's simple if you actually know what you're doing...which this book
explains how to do perfectly."
—Sara Elane

"Excellent review of the fundamentals. Great for ICU fellows up at night. An excellent review also for older
attendings 25 years out from their fellowship (yours truly). Good illustrations of ventilator mode variables
and excellent text giving sound reasoning for making choices and adjustments in common disease states."
—xhighbar

"As a surgical resident working in the ICU, this book was an excellent introduction into ventilator
management. Its main strength is in the way it is written. It does not read like a typical textbook but in more
of a personal tone. I've recommended it to all of my junior residents and I would recommend it to anyone
who is looking to improve their understanding of ventilator management."
— SPM88

"Vents finally make sense! Recommend to all medical professionals with any confusion about vents, settings,
etc."
—D
 The Advanced
Ventilator Book

William Owens, MD

First Draught Press

MMXVII
Medicine is an ever-changing discipline and the subject matter of this
book is no exception. While the author has done his best to ensure
that this book reflects contemporary evidence-based practice, new
developments in the field may supersede the material published here.
Only properly trained and licensed practitioners should provide
medical care to patients with respiratory failure. Nothing in this book
should be construed as advice regarding the care of a specific patient
or group .

Copyright © 2017 by William Owens, MD

All Rights Reserved

Cover Design By Lorien Owens

ISBN 978-0-9852965-2-0
This book is dedicated to the fellows, residents, medical students, nurses, and
respiratory therapists whom I have had the privilege to teach over the years.
Medicine is neither art nor science, but rather a craft. It requires a commitment
to excellence from a craftsman. Paying it forward is part of the deal. This work
is my attempt to share what I've learned about critical care medicine with the
next generation.

Writing a book is not an easy task, and neither is being a physician. I could not
do it without the love and support of Lorien, my wife and fellow adventurer.
Table of Contents
Introduction
1. Oxygen Delivery and Consumption
2. Permissive Hypercapnia
3. Seven Rules For Respiratory Failure
4. PEEP, More PEEP, and Optimal PEEP
5. Severe Bronchospasm
6. Prone Positioning and Neuromuscular Blockade
7. Inhaled Pulmonary Vasodilators
8. Veno-Venous ECMO
9. 2 A.M.
References
About the Author
Introduction

The Ventilator Book was written as a guide for students, residents,

nurses, and respiratory therapists. It was written with the goal of being a quick
reference and an easy-to-read overview of mechanical ventilation. Based on
feedback from readers, I believe that it has accomplished its purpose.

The Advanced Ventilator Book aims to take the reader to the next
level, while preserving the same format and structure that makes The Ventilator
Book a useful reference. This is a book designed for clinicians with some
experience in caring for critically ill patients who would like some guidance on
how to manage cases of severe respiratory failure. I have written it with the
assumption that the reader understands the basics of mechanical ventilation and
the pathophysiology of critical illness or injury. The first two chapters get back
to the basics, with an overview of oxygen delivery and the concept of permissive
hypercapnia. Following this are chapters covering the titration of positive end-
expiratory pressure; the management of the patient with severe bronchospasm;
the use of prone positioning and therapeutic neuromuscular blockade; inhaled
nitric oxide and prostacyclin; veno-venous extracorporeal life support; and a
chapter on incorporating all of this into a treatment strategy .

One feature of The Ventilator Book was the emphasis on practical use.
Many textbooks and articles describe the rationale for a particular mode of
ventilation or therapy, but relatively few actually tell the reader how to do it. The
Advanced Ventilator Book provides the same step-by-step guidance to help
clinicians put these principles into practice.

The Advanced Ventilator Book also continues the original book's

emphasis on support and lung protection rather than cure. No magic bullets are
promised, as none exist. Mechanical ventilation for patients with severe
respiratory failure has great potential to harm, and so the avoidance of
preventable injury is stressed with each topic in the book. The bulk of critical
care medicine is supportive in nature, and the treatment of acute respiratory
failure is no exception.
 Chapter 1
Oxygen Delivery and Consumption

Many textbooks on respiratory and critical care medicine begin with

statements like, "Oxygen is the most necessary and basic building block of life."
In clinical training, the early application of high-flow oxygen is taught as a life-
saving maneuver in emergencies. In the emergency department and intensive
care unit, much importance is placed on keeping the pulse oximeter reading over
90% (and usually over 95%); likewise, there is a compulsion to keep the PaO 2 in
the normal range of 90-100 mm Hg.

At first glance, there is nothing wrong with this approach. Oxygen is

indeed necessary for life, and avoiding hypoxemia is a core part of resuscitation.
When treating patients with severe respiratory failure, however, attaining a
normal PaO 2 may be either impossible or only possible by the application of
injurious airway pressures. Therefore, a more complete understanding of oxygen
delivery and consumption is necessary.

Oxygen Content
Each gram of hemoglobin can bind 1.34 mL of oxygen when fully
saturated. A small amount of oxygen is also carried in the plasma in its dissolved
form. This is represented by the PaO 2 . The solubility coefficient for oxygen in
plasma is 0.003. Putting all of this together yields the oxygen content equation:

CaO 2 = 1.34 x Hgb x SaO 2 + [PaO 2 x 0.003]

 With normal hemoglobin of 15 g/dL, SaO 2 of 100%, and a PaO 2 of
100 mm Hg, the oxygen content of arterial blood is 20.4 mL O 2 /dL blood. It is
important to note that the contribution made by the dissolved oxygen (PaO 2 x
0.003) is very small—0.3 mL O 2 /dL blood. The hemoglobin binds 98.5% of the
oxygen content. The fraction contributed by the dissolved oxygen is negligible.
If the FiO 2 on the ventilator were increased to bring the PaO 2 up to 500 mm Hg
(keeping the SaO 2 at 100%), only 1.2 mL O 2 /dL blood would be added to the
oxygen content.

Keeping the PaO 2 elevated beyond what's necessary for adequate

saturation of the hemoglobin is unlikely to be consequential except in cases of
profound anemia (Hgb < 5 g/dL) or hyperbaric conditions. In fact, the PaO 2 can
often be ignored when calculating oxygen content and delivery in order to make
the math easier. This leads us to the first rule of oxygen: The SaO 2 is what
matters, not the PaO 2 .

Oxygen Delivery
Once the arterial blood is loaded with oxygen, it is delivered to the
tissues to be used for metabolism. The amount of blood circulated per minute is
the cardiac output, which is expressed in liters blood per minute. Since the CaO 2
is measured in deciliters, the units are converted by multiplying by 10. This
yields the oxygen delivery equation:

DO 2 = CO x CaO 2 x 10

If a normal cardiac output is 5 L/min, the DO 2 is 1020 mL O 2

/minute. In order to make comparisons among different patients of various
heights and weights, this can be indexed by dividing the DO 2 by the body
surface area. A "typical" body surface area is 1.7 m 2 , so the "typical" DO 2 I
would be 1020/1.7, or 600 mL O 2 /min/m 2 .

The cardiac output has the greatest influence on oxygen delivery.

Even during periods of arterial hypoxemia, an increase in cardiac output can be
sufficient to deliver the necessary amount of oxygen to the tissues. The table
below shows the effect that an increase in cardiac output can have on oxygen
delivery, even with significant anemia or hypoxemia. It also shows that anemia
has a more pronounced effect on oxygen delivery than hypoxemia. For the
purposes of simplifying the calculations, the PaO 2 has been omitted. This leads
us to the second rule of oxygen: An increase in cardiac output can offset
hypoxemia .

Changes In Oxygen Delivery

CO Hgb SaO 2 DO 2

3 L/min 15g/dL 100% 603 mL O2 /min

8 L/min 7 g/dL 100% 750 mL O2 /min

5 L/min 15 g/dL 100% 1005 mL O2 /min

8 L/min 15 g/dL 75% 1206 mL O2 /min

Oxygen Consumption
During periods of rest, the body's consumption of oxygen (VO 2 ) is
approximately 200-250 mL O 2 /minute. Indexed for body surface area, the
resting VO 2 I is 120-150 mL O 2 /min/m 2 . Normal subjects can increase their
VO 2 during peak exercise by a factor of 10, and elite athletes can reach a
maximum VO 2 of 20-25 times their resting consumption. During critical
illnesses like septic shock, multisystem trauma, or burn injury, VO 2 increases
over baseline by approximately 30-50%.

The consumption of oxygen by the tissues (VO 2 ) varies by organ

system. The brain and heart consume the most delivered oxygen, while hair,
bones, and nails consume a negligible amount. This can be further complicated
by the fact that different organ systems receive different amounts of the cardiac
output—the brain consumes the most oxygen, for example, but also receives
15% of the total blood flow. The coronary circulation, on the other hand,
accounts for only 5% of the total cardiac output so the percentage of delivered
oxygen that is consumed is much higher. Fortunately for the clinician, this is not
important because regional monitoring of oxygen delivery and consumption is
practical only in laboratory animals. Measurement of the total body VO 2 , on
the other hand, can be done rather easily with a pulmonary artery catheter (more
accurate) or by using a combination of a noninvasive cardiac output monitor
along with a measurement of central venous oxygen saturation (less accurate).
While this is not as precise as directly measuring the content of oxygen in
expired gas, it is a close enough approximation for clinical use.

By measuring the mixed venous oxygen saturation in the pulmonary

artery, the venous oxygen content can be calculated:

CvO 2 = 1.34 x Hgb x SvO 2 + [PvO 2 x 0.003]

As with the arterial oxygen content equation, the minor contribution

made by the dissolved oxygen (in this case, the PvO 2 ), can be omitted from the
calculation. Thus, for a hemoglobin of 15 g/dL and a normal SvO 2 of 75%, the
venous oxygen content is 15.1 mL O 2 /dL blood. The difference between arterial
and venous oxygen content is normally 3-5 mL O 2 /dL blood .

The VO 2 can then be calculated by multiplying the arterial-venous

oxygen difference by the cardiac output and converting units:

VO 2 = CO x [CaO 2 - CvO 2 ] x 10

Expanded, this equation is:

VO 2 = CO x [(1.34 x Hgb x SaO 2 )–(1.34 x Hgb x SvO 2

)]x 10

Rearranged (and simpler):

 VO 2 = CO x 1.34 x Hgb x (SaO 2 – SvO 2 ) x 10

In this case, with a cardiac output of 5 L/min, the DO2 is 250 mL O2 /minute.
Indexed for a typical body surface area of 1.7 m2 , the DO2 I is 147 mL O2
/min/m2 .

Using The DO 2 and VO 2 Together

Knowing the DO 2 or VO 2 in isolation is not particularly useful. The

clinical question is whether the delivery is adequate to meet the body's
consumption requirements. To answer this, the DO 2 :VO 2 ratio is helpful.
During periods of both rest and exercise, the DO 2 :VO 2 ratio is maintained at
approximately 4:1 to 5:1 by changes in the cardiac output. This provides a
reserve of sorts—after all, it wouldn't be very useful from a survival perspective
to only deliver as much oxygen as the body absolutely needs at any given time.
This lack of a physiologic reserve would mean that a person would have no
ability to withstand a sudden change in circumstances like having to sprint away
from an attacker, or deal with a high fever or pulmonary embolism.

As seen in the following figure, the DO 2 can vary widely as the VO 2

remains constant. This reflects the aforementioned physiologic reserve. As the
DO 2 declines, however, it can reach a point at which further drops in oxygen
delivery cause a drop in consumption. This point is known in physiology as the
hypoxic, or anaerobic, threshold. It is at this point that the reserve is exhausted
and the consumption becomes supply-dependent. A patient at or below this point
for a prolonged period will become severely acidotic and, in most cases, will not
survive.
It would make sense that the anaerobic threshold would occur when
the DO 2 equals the VO 2 . Experimentally, however, it has been shown that the
threshold is closer to the 2:1 mark, and is explained by the variable oxygen
consumption of different organ systems. Cardiac output delivered to hair, teeth,
and bones doesn't contribute much to meet the needs of the more vital organ
systems.
DO 2 :VO 2 Relationship

Mathematically, the DO2 :VO2 ratio looks like this:

Cancelling common factors greatly simplifies the equation:

If the SaO 2 is assumed to be 100%, then the SvO 2 correlates with the
DO 2 :VO 2 ratio:
 DO2 :VO2 SvO2
5:1 80%
4:1 75%
3:1 67%
2:1 50%

This correlation makes clinical estimation of the DO 2 :VO 2

relationship much easier, as the SvO 2 can be measured directly and continuously
by a pulmonary artery catheter. If a pulmonary artery catheter is not present, a
central venous oxygen saturation (ScvO 2 ) can be measured by obtaining a
venous blood gas from a central venous line placed in the internal jugular or
subclavian vein. The ScvO 2 is usually 5-8% higher than the SvO 2 . While not as
accurate as the true mixed venous oxygen saturation obtained with a pulmonary
artery catheter, the ScvO 2 can be used to estimate of the DO 2 :VO 2 relationship.

The SvO 2 , as a surrogate for the DO 2 :VO 2 relationship, can be

used to identify when a patient has insufficient oxygen delivery to meet
consumption requirements. The SvO 2 also has the advantage of not requiring
continuous calculation of the actual DO 2 and VO 2 —any changes in the
relationship between delivery and consumption will be reflected in the SvO 2 .
The SvO 2 drops as oxygen delivery drops relative to consumption. An SvO 2
below 70% should warrant evaluation, and an SvO 2 below 60% is definitely
concerning—it means that the patient is approaching the anaerobic threshold.

Looking back at the DO 2 equation, impaired oxygen delivery is

always due to either low cardiac output, anemia, or hypoxemia. Correction of
these should increase DO 2 , with a resultant increase in SvO 2 . Keep in mind
that the cardiac output has the most significant effect on DO 2 , and conditions
like congestive heart failure, hypovolemia, hemorrhagic shock, and cardiac
tamponade will all reduce cardiac output. This leads us to the third rule of
oxygen: The SvO 2 is low in low-flow states .

Using the SvO 2 With DO 2 and VO 2

 Patients with severe respiratory failure may have uncorrectable
hypoxemia. A reduction in the SaO 2 will lead to a corresponding reduction in
SvO 2 if the DO 2 :VO 2 ratio remains constant. Calculating the oxygen extraction
ratio is a quick way to estimate the balance between oxygen delivery and
consumption even when the SaO 2 is markedly reduced:

For a normal SaO 2 of 100% and SvO 2 of 75%, the O 2 ER is: (1.0–
0.75)/1.0 = 0.25/1.0 = 0.25, or 25%. This means that of the delivered oxygen,
25% was extracted and consumed by the tissues. A normal O 2 ER is 20-25%.

As an example, consider a patient with severe respiratory failure

whose SaO 2 is 84%. His SvO 2 is 60%. According to the above figure, an SvO 2
this low would be concerning. However, the assumption in Figure 2 is that the
SaO 2 is 100%. Calculating the oxygen extraction ratio:

O2 ER = (0.84 – 0.60)/0.84 = 0.24/0.84 = 0.286, or 28.6%.

While this is a bit higher than the normal range of 20-25%, it isn't that much. Put
another way, this indexing of the oxygen extraction would correlate with an
SvO2 of 71.4% (if the SaO2 were 100%).

As a second example, take a patient with severe respiratory failure

with an SaO 2 of 86%. His SvO 2 is 49%. The O 2 ER is (0.86 – 0.49)/0.86, or
43%. This would correlate with an SvO 2 of 57% if the SaO 2 were 100%, and is
certainly concerning for a low cardiac output state. An O 2 ER of 30% or higher
should warrant investigation, and an O 2 ER higher than 40% indicates that the
patient is approaching the anaerobic threshold .

The fourth rule of oxygen: The DO 2 :VO 2 ratio, SvO 2 , and O 2 ER reflect the
balance between delivery and consumption. They don't represent a specific
target for intervention .

So, How Much Oxygen Is Really Needed?

Unfortunately for physiologists and writers of clinical algorithms,
simply saying to keep the SvO 2 over 70% and all will be well doesn't work. This
should come as no surprise to anyone familiar with the medical literature in
critical care medicine—multiple studies proposing one physiologic manipulation
or another have been consistently disproven. The combined processes of oxygen
delivery, oxygen consumption, stress response, and cellular adaptation are far too
complex to be summed up in this chapter, let alone a one-size-fits-all algorithm.

A normal PaO 2 while breathing ambient air at sea level is 90-100 mm

Hg, but humans are able to tolerate much less over prolonged periods of time.
The minimum necessary PaO 2 and SaO 2 is not known, and it is unlikely that any
IRB will grant approval to a study aiming to withhold supplemental oxygen from
critically ill patients. The degree of tolerable hypoxemia is also highly variable,
and depends on factors such as the patient's age, comorbid conditions, living
environment, genetic factors, and ability to cope with physiologic stress. What is
known is that some people are able to survive moderate and even severe
hypoxemia. Keep the following in mind:

• Mitochondrial PO 2 in cardiac and skeletal muscle is normally

between 1 and 5 mm Hg.

• Oxidative phosphorylation in mitochondria doesn't begin to fail

until the PO 2 is between 0.1 and 1 mm Hg .

• Climbers on Mount Everest who obtained femoral arterial samples

from each other had PaO 2 in the 24-28 mm Hg range, and lived to tell
the tale.

• In septic shock, the problem is not inadequate oxygen delivery. It's

the inability of the tissues to properly metabolize the delivered
oxygen. That's why patients die despite having an SvO 2 of 80%. The
reasons for this are (very) incompletely understood.

• In the various ARDSNet trials, a PaO 2 as low as 55 mm Hg (with

an SaO 2 of 88%) was considered acceptable. This is probably the
best we will get as far as prospective evidence on the subject.

• Patients in the ARDSNet trial who received higher tidal volumes

had better oxygenation, but also had a higher mortality rate. This
suggests that preventing lung injury was more important than
improving oxygenation.

• Many interventions have been shown to improve oxygenation in

mechanically ventilated patients, but not to improve survival.

Using lactate levels is an appealing method of determining whether

oxygen delivery is adequate, but it has its limitations as well. Most lactate
production in critical illness is not due to anaerobic metabolism, despite common
assumptions. Instead, it is a product of increased pyruvate production (with
metabolism to lactate) in the setting of impaired or altered glycolysis and
gluconeogenesis. Lactate is the preferred fuel for cardiac myocytes in the setting
of adrenergic stimulation and is produced by aerobic cellular respiration. Thus,
lactate should be viewed as a nonspecific marker of physiologic stress. If the
lactate comes down following intubation, fluid resuscitation, etc., then it simply
indicates that the patient is responding to therapy. It doesn't imply restoration of
aerobic metabolism in previously anaerobic tissues. Likewise, an increasing
lactate may indicate that the patient has a condition that is leading to an increase
in sympathetic tone and cortisol-mediated stress response. Increasing oxygen
delivery may or may not help the situation—it depends on what the underlying
condition is.

This concept leads to the fifth rule of oxygen: SaO 2 , SvO 2 , O 2 ER,
and lactate are all pieces of information and not goals in themselves . They must
be taken into account along with urine output, peripheral perfusion, mentation,
and other clinical information before any treatment decisions can be made.

Oxygen Toxicity
The idea that supplemental oxygen can be toxic, especially in high
doses, is not new. In neonates, high FiO 2 has been associated with retinopathy
and bronchopulmonary dysplasia. In adults, there is evidence of worse outcomes
with hyperoxia in the setting of acute myocardial infarction and following
cardiac arrest. High FiO 2 in adults can cause irritation of the tracheobronchial
tree and absorption atelectasis (due to the oxygen being absorbed without the
stabilizing effect of nitrogen gas, leading to alveolar collapse).

Laboratory studies have demonstrated the increased presence of

reactive oxygen species in the setting of infection, inflammation, and tissue
reperfusion. The clinical significance of this is unclear, as the oxidative burst is a
known component of inflammation and may be a part of the host response to
infection. Reactive oxygen species can cause cellular injury and apoptosis in vitro
but they rapidly combine with chloride and other ions in vivo , mitigating their
effect. The degree to which the PaO 2 itself plays a role is also not fully
understood, and it may be the case that the oxidative burst occurs as a part of
inflammation or reperfusion under any kind of aerobic conditions (and not solely
hyperoxic) .

The degree to which clinically significant oxygen toxicity occurs in

humans is poorly understood, and the role that the PaO 2 itself plays is unclear.
Just because we don't know that there is toxicity, however, doesn't mean that it
isn't occurring. The safest practice, then, is to treat oxygen like any other drug
and to only give the patient as much as he needs. A useful analogy is the
administration of norepinephrine in septic shock. A normal mean arterial
pressure is 93 mm Hg, but organ perfusion is adequate with a mean arterial
pressure of 65 mm Hg. Norepinephrine is titrated to achieve the lower target
since that's all that's necessary. Aiming for the higher, "normal" target would
require higher doses of norepinephrine and expose the patient to the risk of harm
(ischemic fingers and toes, splanchnic vasoconstriction, increased afterload
leading to impaired cardiac function, etc.).

Avoiding hyperoxia is easy, and can be accomplished by reducing the

FiO 2 . Even normoxia may not be necessary, and it may be prudent to tolerate a
degree of permissive hypoxemia in order to avoid exposing the patient to high
FiO 2 or ventilator pressures. Remember that cardiac output has a much more
significant effect on oxygen delivery than the saturation, and focus on signs of
adequate or inadequate oxygen delivery rather than strictly following the SaO 2
and PaO 2 . This approach leads us to the sixth and final rule of oxygen: Give the
patient just as much oxygen as he needs. This may be less than you think .

Six Rules Of Oxygen

1. The SaO 2 is what matters, not the PaO 2 .

2. An increase in cardiac output can offset hypoxemia.

3. The SvO 2 is low in low-flow states

4. The DO 2 :VO 2 ratio, SvO 2 , and O 2 ER reflect the
balance between delivery and consumption. They don't
represent a specific target for intervention.

5. SaO 2 , SvO 2 , O 2 ER, and lactate are all pieces of

information and not goals in themselves. They must be
taken into account along with urine output, peripheral
perfusion, mentation, and other clinical information
before any treatment decisions can be made.
6. Give the patient just as much oxygen as he needs. This
may be less than you think.
 Chapter Two
Permissive Hypercapnia

Permissive hypercapnia is the practice of allowing a mechanically

ventilated patient to develop or remain in a respiratory acidosis rather than
exposing him to the risk of injurious ventilator settings. For the purposes of this
chapter, permissive hypercapnia is defined as a PaCO 2 > 45 mm Hg with a pH <
7.35. Hickling et al. first described this concept in two papers that demonstrated
a survival benefit with lower tidal volumes and elevated PaCO 2 levels. 1 ' 2 This
work was influential on later studies that showed the superiority of low tidal
volume ventilation, including the landmark ARMA study performed by the
ARDS Network investigators. Most of the studies examining this topic have
focused on the benefit of using a lower tidal volume (4-6 mL/kg predicted body
weight) in ARDS. There is less research on the benefits and risks of permissive
hypercapnia itself, but there may be some advantages to permitting a mild to
moderate respiratory acidosis in patients with severe respiratory failure.

Pulmonary Benefits of Permissive Hypercapnia

The primary rationale for hypercapnia is that avoiding iatrogenic
ventilator-induced lung injury is more important that attaining normal gas
exchange. Overdistension of healthy alveoli leads to cellular injury, and is
referred to as volutrauma. This is the primary mechanism of ventilator-induced
lung injury (VILI) and is independent of distending pressures (barotrauma). The
ARMA study demonstrated a reduction in mortality in patients with ARDS when
tidal volumes of 4-6 mL/kg PBW were used, compared with tidal volumes of 12
mL/kg. 3 This benefit was seen despite worsening gas exchange in the low tidal
volume group. In patients with status asthmaticus, using lower tidal volumes and
respiratory rates prevents dynamic hyperinflation, pneumothorax, and
pneumomediastinum, even though it may lead to a respiratory acidosis.
Permissive hypercapnia is considered acceptable because the benefits of
avoiding lung injury are considered far more important than achieving "normal"
alveolar ventilation.

Since current practice emphasizes the use of a low tidal volume in

ARDS, increasing the tidal volume to correct a respiratory acidosis is seldom
done. Instead, the respiratory rate is adjusted to increase or decrease the minute
ventilation. Most of the time, increasing the respiratory rate on the ventilator is
sufficient to blow off CO 2 and normalize the pH. This may not be necessary,
however, as patients are able to tolerate even a significant respiratory acidosis so
long as oxygenation is maintained. 4 In fact, there may be harm with this common
practice. An increase in the frequency of tidal ventilation invariably leads to an
increase in the cyclical opening and closure of vulnerable lung units. A patient
with a set respiratory rate of 20 breaths per minute will have 11,520 more
ventilatory cycles per day than another patient with a respiratory rate of 12
breaths per minute. Each one of those ventilatory cycles has the potential, albeit
small, to contribute to VILI. Laboratory data supports the idea of using a lower
ventilator rate whenever possible; 5 however, prospective studies in humans will
be needed to validate this concept. In the absence of data, though, it is certainly
reasonable to question the necessity of routinely increasing the ventilator rate to
correct mild to moderate acidemia .

Extrapulmonary Benefits of Permissive Hypercapnia

No prospective, randomized human trials examining the
extrapulmonary benefits of permissive hypercapnia have been done. There are
several laboratory studies in animals that have demonstrated a beneficial effect
of hypercapnia on free radical production, myocardial injury, and cerebral
ischemia. 6 This reduction in pro-inflammatory cytokines and oxidative injury
may prove to be helpful in reducing multisystem organ dysfunction, especially
because the majority of patients with ARDS die of multisystem organ failure
rather than of primary respiratory failure.
 In healthy human volunteers, controlled hypercapnia under general
anesthesia was shown to increase both cardiac output and tissue oxygenation . 7
A study of patients with severe ARDS demonstrated an increase in cardiac
output and systemic oxygen delivery with a tidal volume reduction and
hypercapnic acidosis; 8 the same study, however, also showed worsening right
ventricular function and hemodynamics. In a study of patients with subarachnoid
hemorrhage and cerebral vasospasm, controlled hypercapnia led to an increase in
cerebral blood flow without prohibitory elevations in intracranial pressure. 9
While this is not sufficient to justify a change in recommended ventilator
management, these findings do argue against the presumption of harm with
respiratory acidosis during mechanical ventilation.
Buffering
In the ARMA study and subsequent ARDS Network studies,
administration of buffering fluids was permitted to keep the pH ≥ 7.15. Sodium
bicarbonate (NaHCO 3 ) is often used to treat acidemia, but it does have several
drawbacks. Under usual conditions, the bicarbonate anion is converted to CO 2
and H 2 O via carbonic anhydrase:

The elimination of the excess CO2 produced by this reaction is not normally an
issue—one or two breaths are sufficient to clear it. In the setting of severe
respiratory failure, however, elimination of the CO2 may not be possible and the
pH may in fact fall with the administration of sodium bicarbonate. In addition,
CO2 diffuses freely over cell membranes (including in the CSF), but HCO3 -
does not. This has the effect of worsening intracellular acidosis, even if the
systemic pH rises. A transient hemodynamic improvement is often seen when a
bolus of sodium bicarbonate (e.g., an "amp") is given, but this is more likely due
to the loading of sodium than the change in pH—similar effects are seen with
bolus dosing of hypertonic saline. Keep in mind that the NaHCO3 given in a 50
mL ampule is 8.4%, which is a very hypertonic sodium solution.

THAM (tris-hydroxymethyl aminomethane) is a direct H + ion buffer

that does not depend on alveolar ventilation like NaHCO 3 . It also crosses cell
membranes freely and produces intracellular buffering. This may be a more
effective buffer for hypercapnic acidosis, but there are scant clinical data for its
efficacy. At the time of this writing, the point is moot—THAM has been
discontinued by the only manufacturer that was producing it.

The necessity for buffering a respiratory acidosis during mechanical

ventilation is debatable. The purported benefits of permissive hypercapnia
(beyond the prevention of volutrauma) may be lost when the pH is increased.
Administration of sodium bicarbonate may have some adverse effects, as
described above, and there are no available non-bicarbonate buffers available for
clinical use. Additionally, acidemia may confer a protective effect on hepatic and
renal function, and systemic acidosis shifts the oxygen-hemoglobin dissociation
curve rightward, thereby augmenting tissue oxygen delivery. Using buffer
therapy to keep the pH≥ 7.15 is a common, but unproven, practice. Until clinical
studies show a benefit to doing this, it would be prudent to reserve buffering for
situations where the clinician feels that the acidosis is having an adverse effect
on the patient.

Downsides of Permissive Hypercapnia

Despite the aforementioned benefits, there are some clinical
downsides of hypercapnic acidosis in critically ill patients. The most widely
recognized is the correlation between hypercapnia and intracranial hypertension.
Hypercapnia does lead to vasodilatation, including cerebral vasodilatation.
While this may augment cerebral oxygen delivery, 9 it also increases the
intracerebral blood volume. If intracranial compliance is diminished, this can
lead to higher intracranial pressure. This may or may not be dangerous,
depending on the degree of intracranial hypertension, but it certainly bears
consideration. If hypercapnia is unavoidable in a patient with significant brain
injury, then intracranial pressure monitoring should be considered.

In patients with both acute and chronic pulmonary hypertension,

hypercapnia can cause higher pulmonary artery pressures and lead to right
ventricular dysfunction. Much of this is due to the underlying lung disease, but if
there is clinical evidence of impaired hemodynamic function, then lowering the
PaCO 2 may be beneficial.

Other systemic effects of hypercapnia are more related to the

resulting acidosis than from the effect of the PaCO 2 itself. These include
impaired cardiac contractility, prolonged QT interval, decreased systemic
vascular resistance, and hyperkalemia. If these occur, then buffer therapy may be
warranted if other methods of correcting the respiratory acidosis would result in
lung injury.

On a microcellular level, hypercapnia has been associated with

increased tissue nitration and the production of peroxynitrite. This radical is
released during conditions of physiologic stress, and may mediate tissue damage.
10 The significance of this in clinical medicine is yet to be determined.
Neutrophilic activity against bacterial infection is also attenuated with
hypercapnia, but this can be overcome with the administration of antibiotics. 11

Summary and Recommendations

Permissive hypercapnia is a proven strategy for reducing VILI in
patients with severe respiratory failure, be it due to ARDS or obstructive
diseases like asthma or COPD. To state it simply, it's more important to prevent
iatrogenic lung injury than to get "normal" gas exchange. The degree to which
permissive hypercapnia has a clinical benefit beyond preventing volutrauma
remains to be seen, but the existing literature suggests that this may be the case.

Takeaway Points Regarding Permissive Hypercapnia

• Pay attention to tidal volumes. For ARDS, 4-6 mL/kg

PBW is recommended. For obstructive airway diseases, 6-
8 mL/kg PBW is recommended.
Exceeding these thresholds may increase the risk of VILI.
• High respiratory rates may bring the pH up and the
PaCO2 down, but this may be harmful due to repetitive
cycling of vulnerable lung units.

• Acidemia in itself may be beneficial in critical illness by

augmenting blood flow and tissue oxygen delivery.
• Look beyond the pH—consider buffer therapy, high
respiratory rates, and alternative ventilator strategies
only if the acidosis seems to be causing problems like
intracranial hypertension, pulmonary hypertension,
impaired cardiac contractility, refractory hyperkalemia,
and systemic hypotension.
 Chapter Three
Seven Rules For Severe Respiratory
Failure

ONE

Positive Pressure Ventilation is supportive, and may

be therapeutic, but it is not curative .
Without mechanical ventilation, patients with severe respiratory failure will
undoubtedly die. Positive pressure ventilation can reduce shunt, improve gas
exchange, and take over the work of breathing until the patient has recovered.
That doesn't mean, however, that the ventilator can do anything to reverse the
underlying condition or disease process that has led to respiratory failure.

"It is incident to physicians, I am afraid, beyond all other men, to mistake

subsequence for consequence."
—Dr. Samuel Johnson

TWO

Don't hurt the patient any more than you have to .

Ventilator-induced lung injury (VILI) has been recognized as a necessary evil
since the advent of modern critical care medicine, and to altogether eliminate the
risk of any lung injury from the ventilator is not realistic. That said, much VILI
is actually an unnecessary evil, since it occurs in the pursuit of "normal" gas
exchange or "optimized" physiologic parameters. In cases of severe respiratory
failure, the risk of VILI is high and the potential for rewards is small—it isn't
reasonable to injure the patient's lungs in order to increase the PaO2 from 65 to
95, when 65 is sufficient to maintain life. Focusing on doing the minimum
intervention necessary to support the patient is much more likely to be helpful in
the long run.

"As to diseases, make a habit of two things—to help, or at least, to do no harm."

—Hippocrates

THREE

Throw normal values out the window .

Do what's necessary, not whatever is possible. Chasing the ideal of "normal" gas
exchange will inevitably lead to VILI and unnecessary therapeutic interventions,
all of which carry very real (and unwanted) side effects. With severe respiratory
failure in particular, the twin objectives are to sustain the patient and minimize
the risk of further injury. Sustaining the patient is obviously the more important
objective, and there will be times when very high ventilator pressures are
necessary to achieve it; however, anything that exposes the patient to real or
potential harm should be justifiable .

As an aside, this can be the most difficult part of taking care of critically ill
patients. We have all been taught what's "normal," and we all face the temptation
to do things in order to bring things [lab values, physiologic measurements, vital
signs] back into these ranges.

"Preconceived, fixed notions can be more damaging than cannon."

—Barbara Tuchman
 FOUR

Don't be afraid to experiment ....

We use clinical studies and guidelines as a framework for therapy, but what
works for one patient may not necessarily work for another. Additionally, the
volume of evidence for the critical care of the most severely injured or ill
patients is scant. Therefore, it takes a willingness to try different things and an
ability to admit when a particular therapy isn't working. In these cases, protocols
and clinical pathways can be harmful in that they can constrain physicians from
trying a new approach to the problem.

"Most of our assumptions have outlived their uselessness."

—Marshall McLuhan

FIVE

... But don't be afraid to stay the course .

Trying a different approach may be necessary. More often than not, however, the
patient is adequately sustained with his current ventilator settings but the
clinicians are tempted to change course in order to improve the numbers. This
has the potential for harm without much benefit and should be avoided. Any
modifications should be done either to lower the risk of injury or if the current
settings are not providing an acceptable degree of life support. Keep in mind that
the medical literature is full of therapies that improve oxygenation, ventilation,
and vital signs. Very few of these have actually translated into better patient
outcomes.

"Difficulties are just things to overcome, after all."

—Ernest Shackleton
 SIX

Tracheotomize early .
Patients with severe respiratory failure are in it for the long haul. This means that
the chances of improvement in a few days are low and that the need for at least
some mechanical ventilator support for several weeks or months is quite high.
Couple this with the sedation requirements and relative immobility that
accompanies endotracheal intubation, and it's obvious that the sooner the patient
has a tracheostomy, the sooner he can begin some degree of mobilization and
rehabilitation. A tracheostomy is associated with less sedation, more patient
comfort, better mobilization, and fewer days on the ventilator when compared
with the endotracheal tube. Do it as soon as it's safe.

"You were sick, but now you're well again, and there's work to do."
—Kurt Vonnegut

SEVEN

Remain positive .
Most patients with respiratory failure, even severe ARDS, will eventually
recover. Those who survive ARDS will have near-normal lung function after six
to twelve months. Even people with cardiopulmonary or neurologic disease who
ultimately require a long-term tracheostomy can have an acceptable quality of
life. Declaring a patient "ventilator-dependent" or saying that he has "no chance
of recovery" after one or two weeks in the ICU may be premature or even
wrong. Unbridled optimism isn't appropriate, but neither is pessimistic nihilism.

Some conditions are not survivable. Some conditions are survivable, and even
have the potential for some recovery, but will leave the patient with significant
disability and the need for partial or full ventilatory support. Lastly, some
conditions are survivable and will require a prolonged period of critical care and
ventilatory support, but with a chance at a full recovery to independence.
Obviously, nothing is guaranteed, but clinicians caring for patients with
respiratory failure should be able to discern which scenario is most likely and
present this to the patient and his family.

Once a treatment plan is decided upon, it is imperative for the clinician to

maintain a positive outlook. The patient and his family will be looking for
encouragement and guidance, especially when there's a setback or a run of bad
days. Throughout the course, the most important thing is open and honest
communication. There are times when a shift to palliative care or hospice is
appropriate—failure to recover, development of a new and severe complication,
or if the patient is unwilling to continue a therapy with a small chance of success
that is associated with significant discomfort or an unacceptable quality of life.
In these situations, providing the patient and his family with a peaceful,
comfortable death is a vital function of the clinician. There may be other times,
however, when a setback is temporary and reversible, albeit discouraging (for
example, development of pneumonia that requires going back to full ventilator
support until it's adequately treated). Here, the clinician should encourage the
patient and continue to focus on the ultimate goal of therapy, which is recovery
to an acceptable quality of life.

"Attitude is a little thing that makes a big difference."

—Winston Churchill
 Chapter Four
PEEP, More PEEP, and Optimal PEEP

Positive end-expiratory pressure (PEEP) is used during mechanical

ventilation to maintain alveolar patency and to improve oxygenation. When the
ventilator delivers a breath, the airway pressure is raised to the peak airway
pressure (or peak inspiratory pressure). Expiration is passive, but instead of
allowing the lungs to exhale until the pressure is atmospheric (i.e., zero), the
ventilator stops expiratory flow when the pressure reaches the preset end-
expiratory pressure. This is analogous to breathing out while a large fan blows
air into your lungs—it provides a pressurized splint of air, which in turn
maintains airway and alveolar patency when they would otherwise collapse.
During conventional mechanical ventilation, this is known as PEEP. During
spontaneous and noninvasive ventilation, it is called continuous positive airway
pressure (CPAP). Since both have the same physiologic effect, they are
essentially synonymous.

PEEP is most often applied and adjusted to improve a patient's

oxygenation. The primary effect of PEEP in hypoxemic respiratory failure is to
reduce the fraction of intrapulmonary shunting; that is, areas of the lungs that are
perfused but not ventilated. The more alveolar flooding there is (from pulmonary
edema, bleeding, pneumonia, exudate, etc.), the more pressure is needed to open
up collapsed lung units. The patient's chest X-ray can be used as a rough guide
for the initial PEEP.

Initial PEEP Settings in the Emergency Department or ICU

Chest X-Ray Initial PEEP

Clear
 5 cm H2 O
Scattered Infiltrates 10 cm H2 O
Diffuse Infiltrates Dense 15 cm H2 O
Bilateral White Out 20 cm H2 O

The majority of patients with hypoxemic respiratory failure can be

managed rather easily with a PEEP in the range of 5-10 cm H 2 O. In those with
moderate to severe acute respiratory distress syndrome (ARDS), a more
intensive regimen may be required. This is known as trying to set the "best
PEEP" or "optimal PEEP"—that is, the PEEP that attains the best oxygenation
and compliance while minimizing the risk of ventilator-induced lung injury.
Numerous clinical approaches to the problem of finding the "optimal PEEP" for
a patient have been described in the medical literature, and each method has both
adherents and detractors. As you might expect, each method has its strong points
and drawbacks, and no one approach is superior to the rest (otherwise everyone
would use it and ignore the others). These will be discussed in turn.

ARDSNet Tables
The tables used in the ARDSNet studies have the advantage of
simplicity and titratability to oxygenation, which can be measured easily with an
arterial blood gas or a pulse oximeter. Two tables have been published—one that
uses a high-PEEP approach, and one using lower levels of PEEP. The two
methods were compared head-to-head in the ALVEOLI study, 12 which did not
demonstrate improved outcomes from either approach as long as a lung-
protective tidal volume of 4-6 mL/kg PBW was used. This is actually
advantageous to the physician—it suggests that either table can be used,
depending on the patient's condition. A patient who is morbidly obese, or who
has abdominal compartment syndrome, has reduced chest wall compliance and
might benefit from a higher-PEEP strategy. The extrinsic compression of the
lungs, combined with the poor lung compliance due to ARDS, means that a
higher expiratory pressure should be used to prevent alveolar collapse and
derecruitment.
 On the other hand, using a lower level of PEEP might be indicated in
some cases. A patient with a bronchopleural fistula, or one with tenuous
hemodynamics, or someone with one lung significantly more injured than the
other, may get worse with a high-PEEP strategy. Since one table doesn't have
any proven advantage over the other, the physician can pick whichever one
seems to fit the patient better.
Using the ARDSNet PEEP Tables
• PEEP is measured in cm H 2 O
• Go up and down the table as needed to keep the PaO 2 55-80 mm
Hg, or the SpO 2 88-94%

Lower PEEP Table

FiO 2 PEEP
30% 5
40% 5
40% 8
50% 8
50% 10
60% 10
70% 10
70% 12
70% 14
80% 14
90% 14
90% 16
90% 18
100% 18
100% 20
100% 22
100% 24

Higher PEEP Table

FiO 2 PEEP
 30% 5
30% 8
30% 10
30% 12
30% 14
40% 14
40% 16
50% 16
50% 18
50% 20
60% 20
70% 20
80% 20
80% 22
90% 22
100% 22
100% 2 4

Decremental PEEP Trial

The method of the decremental PEEP trial is that the patient's lungs
should be recruited as fully as possible using a CPAP recruitment maneuver,
followed by a stepwise gradual reduction in expiratory pressure until there is a
drop-off in oxygenation, or compliance, or both. This has the advantage of being
easy to perform at the bedside; additionally, monitoring oxygenation is easily
done with a pulse oximeter, and most ventilators will display static and dynamic
respiratory system compliance. *

A decremental PEEP trial is performed as follows. Remember that you are going
to Recruit, Reduce , and Recruit .

• Ensure the patient is adequately sedated. Neuromuscular blockade

is not necessary as long as the patient isn't making a lot of
 spontaneous respiratory effort.

• Set the ventilator to an FiO 2 of 100%.

• Put the ventilator on CPAP 40 cm H 2 O, with no pressure support.

Hold at this level for 40 seconds (40 for 40). This is the recruitment
maneuver.

• After the recruitment maneuver, change the ventilator mode to

either Volume Control with a tidal volume of 6 mL/kg PBW, or
Pressure Control with a driving pressure of 15 cm H 2 O. Set the
PEEP at 20 cm. Note the patient's compliance .

• Reduce the FiO 2 by 10-20% at a time every 5-10 minutes until the
SpO 2 levels off at 88-94%.

• Once the FiO 2 has been reduced, begin dropping the PEEP in 2 cm
increments every 5-10 minutes until the SpO 2 falls below 88%, or
until there's a notable drop in compliance. Either of these would
indicate alveolar derecruitment.

• Repeat the recruitment maneuver (40 for 40), and set the PEEP at 2
cm higher than the level where derecruitment occurred.

The disadvantages of a decremental PEEP trial include the time required to

properly perform the trial, the need for deep sedation, and the possibility of
hemodynamic or respiratory compromise during the recruitment maneuver.
Clinical trials examining the decremental PEEP strategy have found that it may
improve oxygenation and respiratory compliance, but have not proven any
benefit toward survival. It may not be reasonable to perform a trial on every
ventilated patient in the ICU. For those with moderate to severe ARDS,
however, this can be a useful tool for finding an appropriate level of PEEP.

Pressure-Volume Curves
 Using a dynamic pressure-volume loop to determine the optimal
level of PEEP is appealing. Many ventilators can produce the P-V loop for
review, and it seems intuitive that setting the PEEP at or above the point where
pulmonary compliance falls would be useful.

The inspiratory limb of the P-V curve is thought to represent the

change in compliance as the lungs fill with gas. Initially, the compliance (the
slope of the curve) is poor, reflecting the significant inspiratory pressure needed
to open collapsed lung units. Once these lung units open up, they inflate rapidly
and much more easily. This is the steeper part of the inspiratory P-V curve, and it
indicates that the compliance of the respiratory system has improved. The point
where the compliance changes (in other words, where the slope of the curve
changes) is known as the lower inflection point (LIP).

As the lungs continue to fill with gas, they reach a point where
further application of pressure doesn't expand the lungs much at all—this occurs
at the upper inflection point (UIP), and inspiratory pressures beyond this point
are thought to contribute to alveolar overdistension and potential barotrauma.

Upper and Lower Inflection Points

 Theoretically, using the inspiratory P-V curve should tell the clinician
everything he needs to know regarding the PEEP and driving pressure. The
PEEP should be set at or just above the lower inflection point to keep the alveoli
from collapsing during expiration, and the plateau pressure (that is, the alveolar
pressure at end-inspiration) should be kept at or just below the upper inflection
point to minimize overdistension and barotrauma. This would keep the patient
ventilating along the steep part of the compliance curve.

Unfortunately, it's not that easy. To begin with, establishing a true

pressure-volume curve is difficult. The patient can't be breathing spontaneously,
because patient-initiated breathing alters intra- and extrathoracic mechanics.
Neuromuscular blockade and deep sedation are often necessary. Second, the
inspiratory flow must be constant and relatively low—using a decelerating
inspiratory flow, which is the case in pressure control ventilation and pressure-
regulated volume control ventilation, will produce an inaccurate curve. Third,
the PEEP must be at zero during the maneuver, which can be risky in a severely
hypoxemic patient. Fourth, and perhaps most important, is the argument that it
makes little sense to set an expiratory pressure based on inspiratory pulmonary
mechanics.

Clinical data in humans has shown that while there is some rationale
for the lower inflection point, alveolar recruitment tends to continue during the
entire inspiratory cycle. Additionally, the upper inflection point may represent
the end of the recruitment process but not necessarily alveolar overdistension.
During expiration, which is largely passive, an expiratory inflection point occurs
at a pressure much higher than the inspiratory lower inflection point. This would
suggest that alveolar derecruitment begins at a much higher pressure than the
LIP, and that in ARDS this may be as high as 20-22 cm H 2 O. 13 Additionally,
derecruitment is affected by gravity and the position of the patient. The
heterogeneous nature of both ARDS and alveolar recruitment/derecruitment
makes the use of a single pressure-volume relationship difficult when it comes to
setting the PEEP .

Inspiratory and Expiratory Inflection Points

PEEP At Different Inflection Points

Plateau Pressure-Guided Titratio n

The plateau pressure (P PLAT ) is the pressure measured at the end of
inspiration when inspiratory flow is held at zero. This pressure reflects
equilibration of pressures throughout the respiratory tree, and presumably is the
end-inspiratory alveolar pressure. In general, clinicians should aim to keep the
plateau pressure less than 30-35 cm H 2 O, as this is felt to be the upper limit of
alveolar pressure before lung injury occurs * . In the ExPress trial, the tidal
volume was set at 6 mL/kg predicted body weight, and the PEEP was increased
until the P PLAT was 28-30 cm H 2 O. 14 The control group had a PEEP of 5-9 cm
H 2 O. The hypothesis was that this would lead to full alveolar recruitment while
preventing lung injury. The trial did demonstrate an improvement in oxygenation
in the group receiving this intervention; however, there was no difference in
survival.

One drawback of this approach is that patients with less severe

ARDS may actually receive higher levels of PEEP. Take two patients with
ARDS who each have a predicted body weight of 67 kg. For both, the tidal
volume should be 400 mL. If one has less severe ARDS and a respiratory system
compliance of 40 mL/cm H 2 O, then it will take an inspiratory driving pressure
of 10 cm to deliver the tidal volume. Addition of 18 cm PEEP would bring the
plateau pressure up to 28.

In the case of the second patient, assume that his condition is worse
and that his respiratory compliance is 20 mL/cm H 2 O. This requires a driving
pressure of 20 cm to get the tidal volume, and by following this protocol, he
would only get 8-10 cm PEEP to bring the P PLAT up to 28-30 .

This example is simplistic and purposefully ignores the fact that

compliance would change (for either better or worse) with the application of
PEEP, but the point is that targeting one specific number in all patients could be
harmful. It is also worth considering that this method of setting PEEP did not
improve survival when compared with the control group.

Transpulmonary Pressure
The transmural, or transpulmonary, pressure in the lung is defined as
the difference between the pressure inside the alveoli and the pleural pressure. In
other words, Pressure (in) – Pressure (out). Under normal conditions, this value
is quite small—the alveolar pressure is atmospheric, or zero, while breathing
through an open glottis, and the pleural pressure ranges from around - 3 cm H 2 O
at end-expiration to -8 cm at end-inspiration. Since the transpulmonary pressure
is the difference between the two, it ranges from 3 (o - - 3) to 8 (o - - 8) cm H 2 O.
This is what keeps the lungs open and acts as a counterbalance to the elastic
recoil of the lung.

During positive pressure ventilation, the alveolar pressure becomes

positive and ranges between the plateau pressure at end-inspiration and the end-
expiratory pressure (PEEP). The pleural pressure, if unchanged, remains slightly
negative. Under certain conditions, however, the pleural pressure may become
positive. This usually occurs when there is a reduction in chest wall compliance
either due to primary pleural disease or extrinsic compression (increased
abdominal pressure, volume overload, morbid obesity, or a circumferential burn
of the torso). When this occurs, the transpulmonary pressure is reduced .

Take two patients with ARDS who have a PEEP set at 15 cm. The
first patient has no extrinsic chest wall restriction and a pleural pressure of -5
cm. His transpulmonary pressure at end-expiration is 20 (15 - - 5), which serves
to maintain alveolar recruitment in the setting of lung inflammation and edema.

Transpulmonary Pressure = PEEP – Pleural Pressure

Transpulmonary Pressure = 15 – (-5) = 20
Net Pressure Effect Leads To Alveolar Expansion

The second patient, in addition to having ARDS, also has reduced chest wall
compliance due to morbid obesity (BMI 52). His pleural pressure is +18 cm,
which means that his transpulmonary pressure at end-expiration is -3 (15 - +18 ).
The net effect is alveolar collapse at the end of each respiratory cycle.
 Trans pulmonary Pressure = PEEP – Pleural Pressure
Trans pulmonary Pressure = 15 – 18 = -3
Net Pressure Effect Leads To Alveolar Collapse

Direct measurement of the intrapleural pressure in ICU patients isn't

possible, so the esophageal pressure is used as a surrogate. This is by no means
exact—pleural pressure itself varies from the base of the lung to the apex and is
affected by supine or prone positioning, and esophageal pressure is subjected to
the weight of the mediastinal contents. 15 It is, however, useful for titrating PEEP
in patients in whom there is considerable extrinsic reduction of chest compliance
.

In order to measure esophageal pressure (Peso), an air-filled

esophageal balloon catheter must be inserted. These are commercially available
16 and can be connected to a standard pressure monitoring system. The

CareFusion Avea ® ventilator has a port to connect an esophageal pressure probe

and can display the esophageal pressure as well.
 Insertion of the esophageal balloon catheter should be done by a
qualified practitioner in accordance with the manufacturer's instructions. The
depth to which the catheter should be inserted can be estimated by multiplying
the patient's height in centimeters by 0.288. This should, in most people, position
the balloon in the lower third of the esophagus. Partial inflation of the balloon
with 1 mL of air will allow changes in the esophageal pressure to be reflected on
the monitor. The esophageal pressure waveform should slightly increase during
ventilator-delivered breaths and have a negative deflection during patient-
initiated breaths. Gentle pressure applied to the abdomen that leads to an
increase in the pressure reading suggests gastric placement of the balloon, and it
should be withdrawn.

Once the esophageal balloon catheter is in proper position, the end-

expiratory transpulmonary pressure can be calculated:

Transpulmonary Pressure = PEEP – [P eso – 5] *

For a patient with a PEEP of 15 cm and a PESO of 22, his end-expiratory

transpulmonary pressure is -2 cm H2 O. In other words, at end expiration the
compression of his lungs by the increased pleural pressure is leading to alveolar
collapse. In this situation, the PEEP should be increased to a minimum level of
17 to keep the end-expiratory transpulmonary pressure at zero .

One clinical trial examining the effect of transpulmonary pressure

monitoring in patients with ARDS demonstrated a significant improvement in
oxygenation but did not show a survival benefit. 17 As such, this technique is not
recommended for routine use. It may be helpful, though, in determining the
appropriate level of PEEP for patients with intra-abdominal hypertension or
morbid obesity.

Optimal PEEP vs. Good Enough PEEP

In a trial of 51 patients with ARDS, Chiumello and colleagues
examined different methods of setting PEEP (ARDSNet tables, targeting a
plateau pressure a la the ExPress trial, using the time-pressure stress index, and
via transpulmonary pressure using esophageal pressure measurement). 18 All
methods were assessed using CT scanning to determine the change in
recruitability of the lung. Their findings suggested that the only method that
correlated with the degree of whole-lung recruitability and the severity of ARDS
was the use of the PEEP-FiO2 table. The other methods were associated with
more hyperexpansion of normal lung units without a commensurate benefit in
recruitment of collapsed alveoli.

The multiple methods of determining the best, or optimal, level of

PEEP have a few things in common. They tend to be laborious. They tend to
make significant physiologic assumptions that may not be valid—for example,
the assumption that the pressure in the lower esophagus accurately reflects
pleural pressure throughout the patient, or the assumption that lung recruitment
is complete by the lower inflection point of the inspiratory pressure-volume
curve. Lastly, they most often focus on surrogate endpoints that may not be
meaningful. Clinical trials of different maneuvers designed to find optimal PEEP
often report improved oxygenation or compliance when compared with controls,
but none have shown a survival benefit .

Perhaps we need to stop searching for optimal PEEP. The history of

critical care medicine has consistently shown that attempts by clinicians to
optimize different physiologic parameters are often unnecessary and
occasionally harmful. * This may be no different. Luciano Gattinoni, one of the
foremost researchers in the field, has suggested this very thing. A "good enough"
PEEP maintains oxygenation and lung recruitment without compromising
hemodynamic function, and can be based on a combination of the severity of
ARDS and the good sense of the treating clinician.

Good Enough PEEP 19

Degree of ARDS PaO 2 /FiO 2 Ratio PEEP

Mild 201-300 5-10 cm H2 O
Moderate 101-200 10-15 cm H2 O
Severe ≤ 100 15-20 cm H2 O
* Compliance = Volume / Pressure

Dynamic compliance on the ventilator = Tidal Volume / [Peak Inspiratory Pressure – PEEP]

Static compliance on the ventilator = Tidal Volume / [Plateau Pressure – PEEP]

* It is important to keep in mind that no one has established a truly "safe" level of plateau pressure, above
which lung injury is present and below which no injury occurs. Most experts, however, advise keeping the
plateau pressure at or below this range.

* 5 cm is subtracted from the Peso to account for mediastinal weight. This is a crude estimation, not an
exact measurement.

* Perioperative hemodynamic optimization using the pulmonary artery catheter; ScvO2 monitoring in septic
shock; aggressive transfusion strategies in penetrating trauma, GI hemorrhage, and critical illness;
decompressive craniectomy to treat intracranial hypertension; intra-aortic balloon counterpulsation for
cardiogenic shock; high-frequency oscillatory ventilation for ARDS. The beat goes on....
 Chapter Five
Severe Bronchospasm

Mechanical ventilation of patients with severe bronchospasm can be

very difficult. This usually occurs with status asthmaticus, but it can also be seen
with respiratory failure due to chronic obstructive lung disease, inhalation of
toxic fumes, and viral bronchiolitis. Status asthmaticus in particular can be tough
to treat due to the combination of bronchospasm and mucus plugging, which
significantly worsens ventilation-perfusion matching.

Respiratory care of the patient with severe bronchospasm begins with

treatment of the underlying condition. Asthma and COPD exacerbations are
treated with inhaled beta-adrenergic agonists like albuterol; inhaled anti-
muscarinics like ipratropium bromide; and systemic corticosteroids. The dosing
of corticosteroids depends on the disease in question and the underlying
pathophysiology. Asthma has a strong allergic component and is associated with
inflammation and airway hyper-responsiveness. Initial dosing of corticosteroids
in the ICU should be 1-2 mg/kg of prednisone (or equivalent) daily. The
bronchospasm in COPD, on the other hand, is more related to excessive mucus
production and narrowing of airways due to dynamic collapse during forced
exhalation. It has much less of an inflammatory component when compared with
asthma. Lower dosing of prednisone (or equivalent) is appropriate. Most studies
suggest that there is little benefit from a prednisone dose in excess of 40-60 mg
per day for COPD exacerbations, and there is always the potential for harmful
effects from higher doses. Corticosteroids do not have a linear dose-response
relationship in respiratory illness—doubling the steroid dose will not halve the
bronchospasm.

For those patients with bronchospasm requiring mechanical

ventilation, it is important to differentiate high airway resistance from high
alveolar pressure. This can be done by performing an inspiratory pause
maneuver on the ventilator. At the end of inspiration, flow is temporarily stopped
for 0.5 to 1.0 seconds. During this time, pressures will equilibrate across the
respiratory system. The pressure in the endotracheal tube will be the same (or
nearly so) as the pressure in the alveoli. This is known as the plateau pressure (P
PLAT ). The difference between the peak inspiratory pressure (PIP) and the
PPLAT is an estimate of the airway resistance. Normally, the gradient between
the PIP and PPLAT is 5 cm or less. If the gradient is higher, then the resistance to
flow is elevated. This can be due to an issue with the endotracheal tube—if it's a
small-diameter tube, or if it's kinked, or if it's partially plugged with mucus, the
airway resistance will rise. If the tube is functioning well, is properly sized, and
not obstructed, then a high airway resistance usually means that there's
bronchospasm present (even if wheezing can't be auscultated). Treatment with
inhaled beta-agonists and anti-muscarinics may help.

If the PIP and P PLAT are both elevated, especially if the P PLAT is over
30 cm H 2 O, then it indicates increased alveolar pressure. Common situations
that increase the alveolar pressure include mainstem intubation (all the volume
for two lungs is going into one); pneumothorax; pulmonary edema; mucus
plugging leading to atelectasis; dynamic hyperinflation; and increased abdominal
pressure. While all of these can occur in any patient requiring mechanical
ventilation, asthmatics are especially prone to mucus plugging, dynamic
hyperinflation, and pneumothorax. These should definitely be considered in any
ventilated asthmatic who suddenly gets worse .

Dynamic hyperinflation can be diagnosed both on the ventilator and

by physical exam. On examination, the patient will usually appear
uncomfortable. Paradoxical breathing, with the chest and abdomen moving in a
dyssynchronous manner, can be seen. On auscultation, loud wheezing can be
heard right up until the next breath is delivered. The cardiac monitor may display
lower voltage of the QRS complex during inspiration, due to the trapped gas in
the thorax impeding the electrical current. If the patient has an arterial line,
pulsus paradoxus * may be present. The patient's jugular veins may be distended
during expiration but collapse during inspiration, owing to the increased
intrathoracic pressure at the end of expiration.
 On the ventilator, dynamic hyperinflation can be observed by looking
at the expiratory flow-time waveform. Normally, at end-expiration the flow
should be zero. All of the gas has escaped, leaving the lungs at functional
residual capacity. With dynamic hyperinflation, there is usually some expiratory
gas flow ongoing at the time of the next inspiration. This may not always be
present, however, and so an expiratory pause maneuver should be performed if
dynamic hyperinflation is suspected.
Dynamic Hyperinflation

The mechanics behind the expiratory pause maneuver are similar to

the inspiratory pause—when flow stops, pressure equilibrates. This time, the
pressure is equilibrating at the end of expiration, when alveolar pressure should
normally be zero. If any PEEP is set on the ventilator, the alveolar pressure
should equal the PEEP. With dynamic hyperinflation, the actual (or measured)
end-expiratory pressure will be higher than the applied (or set) PEEP. This is
why dynamic hyperinflation is often referred to as "auto-PEEP."

Expiratory Pause Maneuver Demonstrating Auto-PEEP

Ventilator Management
The goals with mechanical ventilation in the patient with severe
bronchospasm are to minimize hyperinflation, allow the respiratory muscles to
rest, and to avoid complications like pneumothorax, pneumomediastinum, and
pulmonary interstitial emphysema. With regard to gas exchange, the goal is
adequate (not perfect) oxygenation and toleration of hypercapnic acidosis.

Dynamic hyperinflation can be minimized by simply giving the

patient enough time to exhale completely. This can be done by lowering the
respiratory rate and/or reducing the inspiratory time (the time it takes for the
breath to be delivered). Of the two, lowering the respiratory rate is the most
effective. Consider a patient with a respiratory rate of 20 and an inspiratory time
(I-time) of l second. With 20 breaths per minute, there are 3 seconds allocated
per breath. The I-time is set at 1.0 seconds, leaving 2.0 seconds for expiration.
The ratio between inspiration and expiration time is thus 1:2. If the clinician
wants to permit more time for expiration, he can shorten the inspiratory time and
prolong the I:E ratio. In this example, if the I-time is lowered to 0.7 seconds,
there are now 2.3 seconds left for expiration. This translates to an I:E ratio of
1:3.3. This is better, but it may not be sufficient to permit adequate expiration
through narrowed, inflamed airways. You will also find that shortening the
inspiratory time by too much can cause a good bit of air hunger and discomfort,
as well as very high peak inspiratory pressures. In turn, this can lead to heavy
sedation and neuromuscular blockade. Try taking in your breaths in 0.5 seconds
—you won't like it for long.

Lowering the respiratory rate is the more effective way of prolonging

the I:E ratio while minimizing patient discomfort. In the example above, if we
lower the respiratory rate to 15, there are now 4 seconds per breath. If the I-time
is kept at 1.0 second, the expiratory time is 3.0 seconds. The I:E ratio is 1:3.
Lower the respiratory rate to 12, and the I:E ratio is 1:4. A rate of 10—the I:E
ratio is 1:5. Most of the time, there isn't much benefit to an I:E ratio any longer
than 1:5.

Low tidal volumes can also work against the patient with severe
bronchospasm, especially if he isn't heavily sedated. Air hunger is a common
symptom of an exacerbation of asthma or COPD, and low tidal volumes can
make the patient very tachypneic. A tidal volume of 6 mL/kg predicted body
weight (PBW) is great for ARDS, but with asthma and COPD, a higher tidal
volume is often needed. 8 mL/kg PBW usually works without causing
overdistension, and it mitigates the tachypnea and air hunger. Higher tidal
volumes, especially > 10 mL/kg PBW, do increase the risk of barotrauma .

Adequate sedation and analgesia is important. The discomfort of the

endotracheal tube, along with the tachypnea associated with an exacerbation of
pulmonary disease, can lead to air trapping and hyperinflation. Narcotic
analgesia (like a fentanyl infusion) minimizes the discomfort of the tube and
other devices and helps take away the subjective feeling of breathlessness.
Titratable sedatives such as propofol and dexmetetomidine may also be helpful.
Use of a sedation scale, such as the Richmond Agitation-Sedation Scale (RASS),
is important for nursing staff to titrate the sedation. In my experience, titration to
a RASS of -1 to -2 is effective in patients with severe bronchospasm.
Neuromuscular blockade can be used if the patient-ventilator dyssynchrony is
severe or if the dynamic hyperinflation is causing significant hemodynamic
instability. It should be used sparingly and only until the patient stabilizes,
however, due to the higher risk of critical illness-associated weakness with these
drugs (especially when corticosteroids are also administered, as they usually
are). Of the neuromuscular blockers available, cistatracurium is preferred due to
its metabolism not being affected by renal or hepatic dysfunction.

Initial Ventilator Settings for Severe Bronchospasm

• Mode: Assist-Control Ventilation

• Rate: 10-14 Breaths/minute
• Adjust Inspiratory Time to keep I:E ratio 1:3—1:5
• Constant inspiratory flow
• Tidal Volume: 8 mL/kg PBW
• PEEP: zero (or ZEEP) *
• FiO2: 100% to start, adjusted downward to keep SpO 2 88-95%

The concept of permissive hypercapnia is very important when

caring for patients with severe bronchospasm. The idea is that preventing lung
injury (in the form of dynamic hyperinflation or barotrauma) is more important
than having normal gas exchange. Maintaining adequate oxygenation is
important—keep the SpO 2 88-95%, and the PaO 2 55-80 mm Hg. The PaCO 2 is
less important, unless there is a significant coexisting condition like increased
intracranial pressure that precludes permissive hypercapnia. In general, as long
as the pH is kept above 7.10, it's all right to allow the PaCO 2 to rise. Buffer
therapy such as sodium bicarbonate infusions can also be used to keep the pH
above this threshold. The adoption of permissive hypercapnia seems to be
associated with a significant decline in the mortality seen with status asthmaticus
and mechanical ventilation. 2 0

Most of the time, the measures described above are sufficient to treat
the patient with severe bronchospasm. Keep the vent rate low; administer
albuterol, ipratropium, and prednisone; don't stress about the PaCO 2 ; and let the
patient get better. This usually works, until it doesn't. When the patient is still
deteriorating, the clinician should consider one or more rescue therapies. These
include Heliox, ketamine infusion, and therapeutic bronchoscopy. Inhalational
anesthetics have been described in the literature, but the cumbersome nature of
anesthesia machines and the potential toxicity of leaking anesthetic gases to
hospital staff make this a less desirable option.

Heliox
Heliox refers to a blend of helium and oxygen, usually in a 70:30
ratio. A gas blender can be used to change this ratio to 60:40. When the fraction
of oxygen exceeds 40%, the potential benefit of Heliox is lost. Therefore, Heliox
should only be used when the patient can be adequately oxygenated with an FiO
2 of 40% or less.

The benefit of Heliox is with the helium having a much lower density
than nitrogen gas. This translates into Heliox improving the tendency of inspired
gas toward laminar flow, which improves gas flow through narrowed proximal
and larger airways. More laminar and less turbulent gas flow in the conducting
airways leads to better gas exchange and aerosol delivery of medications like
albuterol to the respiratory and terminal bronchioles. This can also help reduce
the work of breathing. Heliox can be delivered by facemask, through
noninvasive positive pressure breathing, or via the mechanical ventilator.

Two issues must be addressed when using Heliox through the

ventilator. The first is the mode of ventilation. Most newer ventilators are not
calibrated for a mixture of helium and oxygen. The inspiratory valve may permit
a larger volume of gas to be delivered than what is set in volume-control mode,
and some ventilators will not register an accurate exhaled tidal volume. For this
reason, pressure-control ventilation is preferable. The inspiratory pressure should
be set at a level sufficient to attain rise of the chest and provide acceptable (if not
normal) alveolar ventilation. In order to avoid barotrauma, keeping the
inspiratory pressure less than 30 cm H 2 O is advised, if possible. If volume-
control ventilation is desired, then the clinician should either use a conversion
table for the particular ventilator being used 21 to estimate the true tidal volume,
or else periodically measure the exhaled tidal volume at the level of the
endotracheal tube with a density-independent pneumotachograph. 22 Using
pressure-control ventilation is easier.

The second issue is the FiO 2 . Most ventilators have two gas inlets—
one for 100% oxygen, and one for air (21% oxygen). The gas blender will mix
the two to attain the FiO 2 selected on the ventilator. Administration of helium
through the air inlet can result in a different FiO 2 actually delivered to the
patient. In the case of pure helium going through the air inlet, the delivered FiO 2
may be less than what's selected (normally the air inlet has 21% oxygen, and
with pure helium there's 0% oxygen). More commonly, a premixed Heliox tank
is connected to the air inlet. If the Heliox is, say, 70% helium and 30% oxygen,
and the FiO 2 on the vent is set at 30% oxygen, then the patient will end up
receiving a higher FiO 2 than the selected 30% (due to mixture of the gases). This
may reduce the efficacy of the Heliox, especially if the true FiO 2 exceeds 40%. It
can also increase the volume of gas delivered to the patient and increase the
airway pressures. To make this even more complicated, different ventilators use
different mixing valves and blenders. It's important to know how the particular
ventilators you're using work with Heliox.

There are two ways around the issue of the FiO 2 . The first is to
directly sample the inspired gas with a density-independent measuring device,
which is cumbersome. The second is to connect a premixed tank of Heliox to the
air inlet of the ventilator, and set the FiO 2 on the vent to 21%. This means that no
supplemental oxygen is added to the Heliox mixture and that only the Heliox is
delivered to the patient. The gas blender on the tank, or the mixture of the gas in
the tank itself, can be used to control the "true" FiO 2 —the vent may say the
FiO 2 is 21%, but if the tank is full of 70:30 Heliox, then the patient is actually
getting 30% oxygen. This is the easiest method, but it requires informing the
nursing and respiratory staff that the FiO 2 is not actually 21% (despite what it
says on the ventilator).
Ketamine
Ketamine is a dissociative agent that has marked sedative and
analgesic properties, and it's most commonly used for procedural sedation.
Unlike benzodiazepines, which act on inhibitory GABA receptors, ketamine
blocks excitatory NMDA receptors in the central nervous system. NMDA
receptors are also present in the lungs and appear to play a role in
bronchoconstriction. Ketamine's anti-NMDA effect would therefore make it an
attractive agent to use in conditions of severe bronchospasm like status
asthmaticus. Experimental models have also suggested that ketamine has a
salutary effect on bronchospasm through effects on norepinephrine reuptake and
vagal inhibition. 23

Ketamine is not without side effects. The most commonly reported is

an increase in airway secretions. Psychiatric side effects have also been
described, including disorientation and hallucinations. It should be noted that the
psychiatric side effects are more common with higher doses of ketamine,
especially when it is used for general anesthesia. 24 Benzodiazepines can
ameliorate these side effects. Other adverse reactions to ketamine include
laryngospasm, hypertension, and increased intracranial pressure.

Clinical trials report mixed results, with some demonstrating an

improvement in airway pressures, gas exchange, and bronchospasm. Others have
not been as supportive of using this agent in mechanically ventilated asthmatics.
At this time, no large, randomized, prospective trials have been performed to
support or discourage the use of ketamine in the setting of severe bronchospasm.
As such, it should be considered a rescue agent to be used when conventional
therapy (steroids, bronchodilators, appropriate ventilation settings, and adequate
sedation) has proven ineffective.

The typical dosing for ketamine in status asthmaticus has not been
clearly defined. Clinical trials have used an initial bolus anywhere from 0.1
mg/kg to 2.0 mg/kg, followed by infusions ranging from 0.15 mg/kg/hr to 2.5
mg/kg/hr for up to five days. 23 Because of the lack of clear evidence-based
recommendations, it seems prudent to start with a lower bolus dose like 0.1 to
0.5 mg/kg, followed by an infusion starting at 0.15 to 0.25 mg/kg/hr. The
infusion can be titrated upward until the desired clinical response of adequate
sedation, improvement in gas exchange, and improvement in bronchospasm (as
determined by chest auscultation and lower airway pressures) is attained.
Furthermore, it makes sense to administer benzodiazepines when the ketamine is
being weaned off to prevent emergence reactions.

Therapeutic Bronchoscopy
Therapeutic bronchoscopy is occasionally necessary to clear the
airways of thick mucus plugs and bronchial casts. Occlusion of large conducting
airways can definitely affect lung mechanics and gas exchange, and seems to
limit the delivery of inhaled bronchodilators to smaller airways. One review of
93 cases of fatal asthma showed that airway obstruction by exudative secretions
was a significant cause of death. 25 Early evaluation of the tracheobronchial tree
with fiberoptic bronchoscopy should be considered in patients with severe
bronchospasm, and mucus plugs should be aggressively lavaged. Giving
mucolytics like N-acetylcysteine through the bronchoscope may be helpful, but
this can cause bronchospasm. Instrumentation of the airways themselves with
the bronchoscope can also cause immediate or delayed bronchospasm. For
"routine" cases of asthma, bronchoscopy is probably not warranted and may
cause complications. For severe cases, however, the likelihood of significant
mucus plugging or casting of the airways is higher and the benefits of
bronchoalveolar lavage may outweigh the risk.

Wait It Out
One of the major challenges in the critical care of patients with
severe bronchospasm is that they don't get better right away. It may take several
days or longer for the steroids and bronchodilators to take effect and for the
inflammation and bronchospasm to subside. In the meantime, it is important to
stay focused on the following goals:

• Maintain adequate (but not perfect) oxygenation—an SpO 2 88-95%

is fine. Hyperoxia isn't necessary or helpful.
 • Reduce hyperinflation by giving the lungs plenty of time to exhale.

• Tolerate a respiratory acidosis, especially if "normalizing" the

PaCO 2 and pH means injuring the lungs with high tidal volumes or
dynamic hyperinflation. Prevention of ventilator-induced lung injury
is far more important than a "good" ABG. Let the pH go down as low
as 7.10 if necessary.

• Use therapies like Heliox, ketamine, and therapeutic bronchoscopy

if necessary, but don't let them distract you from the goals listed
above .

• Provide good holistic critical care. That includes nutritional

support, DVT prophylaxis, and mobilization when appropriate.
• Be patient. It takes time to get better. Put the patient on the right
course, monitor for changes, and let your plan work. In an age of
instant gratification, this can be the hardest part!

* A drop in the systolic blood pressure by more than 10 mm Hg during inspiration. The "paradox" is that the
heart is still beating, but the radial pulse may be absent. Pulsus paradoxus can also be seen with cardiac
tamponade, constrictive pericarditis, anaphylaxis, pneumothorax, and other conditions.

* Alveolar flooding and collapse is usually not a problem with status asthmaticus or exacerbation of COPD.
Applying PEEP, especially with status asthmaticus, can worsen air trapping and hyperinflation. There is
occasionally a role for applied PEEP with dynamic airway collapse, like in a COPD exacerbation, but to
start with I recommend zero end-expiratory pressure (ZEEP). This is discussed further in The Ventilator
Book .
 Chapter Six
Prone Positioning and
Neuromuscular Blockade

Two adjunctive strategies for treating severe ARDS that have been
used for years are prone positioning and therapeutic neuromuscular blockade.
These are often used in conjunction with each other, and the clinical rationale is
improved ventilator-perfusion matching and alveolar recruitment. Until recently,
neither treatment had shown an improvement in survival from severe respiratory
failure (although there was proof of improved oxygenation).

In 2013, Guérin and colleagues published a multicenter randomized

trial (PROSEVA) examining the effect of prone positioning for 16 hours,
followed by 8 hours supine, in patients with ARDS. 26 They reported an overall
reduction in mortality of 16.8%. In 2010, Papazian and colleagues published a
multicenter, randomized, double-blind study (ACURASYS) that demonstrated a
reduction in the hazard ratio for death from ARDS when a cisatracurium
infusion was used for 48 hours early in the treatment of moderate-to-severe
ARDS. 27 The publication of these two papers led to the inclusion of these
therapies in professional guidelines and has prompted interest in these, and other,
strategies for the treatment of ARDS .

Despite the enthusiasm that has greeted these findings, it's important
to keep in mind that there are limitations with these and other studies, and that
their findings should not result in wholesale application of prone positioning and
neuromuscular blockade in every patient with ARDS. In this chapter, the pros
and cons of each treatment will be discussed. If this seems like hedging, well,
that's because it is. Both prone positioning and neuromuscular blockade have a
place in the treatment of ARDS, and both have significant risks. Neither is a
magic bullet, and neither is a substitute for lung-protective ventilation and good
supportive critical care. By the time you're reading this, there may be new
developments either supporting or refuting (or both!) these therapies. For now,
the focus should be on identifying the patients who may benefit while
minimizing the risks.

In Favor Of Prone Positioning

In ARDS, dorsal lung units tend to become more consolidated.
Transpulmonary pressures are increased in dorsal alveoli and lower in ventral
alveoli. At the same time, the gravitational effect on pulmonary blood flow leads
to these units being relatively more perfused than the aerated ventral lung units.
This has the effect of increasing the shunt fraction and worsening oxygenation.
The rationale for prone positioning is that flipping the patient onto his abdomen
will improve ventilation-perfusion matching and thereby improve gas exchange.
At the same time, a more even distribution of aeration and transpulmonary
pressure occurs.

There are other benefits as well. Prone positioning improves drainage

of pulmonary secretions from the airways. Allowing the abdominal contents to
be dependent (by padding the patient's chest and pelvis) reduces the pressure on
the diaphragm and improves chest wall compliance. The weight of the heart,
which is normally directed over the left lower lobe, is shifted centrally .

Previous studies of prone positioning were able to show

improvement in pulmonary blood flow and oxygenation, but did not improve
mortality. 28 , 29 This may have been due to using prone positioning in patients
with less-severe ARDS, poorly-defined protocols, and a shorter duration of the
time spent prone. The 2013 Guérin study, on the other hand, was well-defined
and included patients with a PaO 2 /FiO 2 ratio ≤ 150. These patients were
sufficiently ill enough to where a treatment benefit might become evident. This
study also mandated that the time spent prone (16 hours at a time) was long
enough for any physiologic benefit to occur. Earlier studies had prone times of 6-
8 hours.
 The risks of prone positioning include dislodgement of life support
equipment like endotracheal tubes and vascular catheters; pressure injury to the
eyes, face, and extremities; and the need for heavy sedation and often
neuromuscular blockade. Many of these complications can be reduced or
avoided by using a clearly-defined protocol for turning and adequate staff
training. While sedation and neuromuscular blockade have risks, limiting prone
positioning to those patients who have moderate-to-severe ARDS (and stopping
it once the patients begin to recover) should keep the number of days spent
heavily sedated down.

Arguments Against Prone Positioning

The centers that participated in the PROSEVA trial have ample
experience with prone positioning for ARDS. The importance of staff education
and protocols for proning cannot be overstated. The risks to both the patient and
to the staff are highest during the turning process. Intensive care units that wish
to make this a part of how they care for ARDS should develop checklists and
practice doing it so that it can happen seamlessly when actually done for
critically ill patients .

The PROSEVA trial did show a mortality benefit, but it's important to
note that it was the first trial of prone positioning in ARDS to do so after
numerous other clinical trials had failed. This may have been due to
improvement in the process and a refinement of the indications and duration of
treatment. It could also easily reflect the fact that clinical statistics are not exact
and that occasionally a trial can demonstrate a benefit when none actually exists.
The PROSEVA trial was very similar to another clinical study of prone
positioning published in 2009. 30 That study looked at a similar number of
patients (466, vs. 342 in PROSEVA) and included patients with a PaO 2 /FiO 2
ratio ≤ 200. It also used a similar duration of proning (20 hours at a time, vs. 16
hours in PROSEVA). There was the expected improvement in oxygenation, but
no statistically significant difference in mortality. The fact that two very similar
clinical trials reached very different conclusions suggests that a much larger
tiebreaker trial is needed.
 The final issue that clinicians should consider with PROSEVA is the
overall reduction in mortality. There was a nearly 17% absolute risk reduction,
which is huge . No other treatment in critical care medicine has been able to
consistently reduce the risk of death that much. This may be a case of "if it looks
too good to be true, it probably is."

In Favor of Therapeutic Neuromuscular Blockade

Much of the ventilator-induced lung injury seen in ARDS is due to
high transpulmonary pressure in vulnerable alveoli and overdistension of
relatively healthy lung units during mechanical ventilation. Therapeutic
neuromuscular blockade, along with heavy sedation, aims to improve respiratory
system compliance and patient-ventilator dyssynchrony. Inflammatory
biomarkers in both the blood and in bronchoalveolar fluid are also reduced when
neuromuscular blockade is used. 31 These effects putatively lower the risk of
ventilator-induced lung injury and improve survival from ARDS .

The ACURASYS trial used a bolus of cisatracurium, followed by a

continuous infusion, for 48 hours early in the treatment of moderate-to-severe
ARDS (PaO 2 /FiO 2 ≤ 150). This led to an improvement in the hazard ratio of
death at 90 days, along with a reduction in the incidence of pneumothorax (4%
vs. 11.7%). The benefit was most pronounced in patients with a PaO 2 /FiO 2
ratio ≤ 120, suggesting that neuromuscular blockade is most effective in the
sickest patients. Other, smaller clinical trials have also suggested that a
cisatracurium infusion may be beneficial. 31 , 32 Importantly, these trials did not
show a higher risk of prolonged myopathy or ICU-acquired weakness when
compared with the control groups.

Arguments Against Neuromuscular Blockade

The ACURASYS trial claimed a reduction in the 90-day hazard ratio
of death, but the overall reduction in mortality was not statistically significant. In
other words, patients who received cisatracurium lived longer than the control
arm, but a similar number were dead by the 90-day mark. A reduction in the
hazard ratio may be a meaningful outcome in a trial looking at a new treatment
for, say, lung cancer—the five-year mortality may not be different, but the new
drug may prolong life for another year or two. Most people would consider that
successful. Most people would probably not consider an additional week or two
of life in the ICU, intubated and attached to machines, to be a successful result.

There are also concerns regarding the external validity of the study.
The rate of pneumothorax in the control group was nearly 12%, which seems
higher than what's seen in clinical practice. This leads to questions regarding the
ventilator strategy used, and it turns out that the prescribed tidal volume in this
trial was 6-8 mL/kg PBW. This is a higher tidal volume than what's
recommended for moderate-to-severe ARDS. Another trial using a lower tidal
volume strategy would seem to be needed .

Nearly 22% of the control group in the ACURASYS trial received

open-label cisatracurium. This makes interpretation of the results difficult, as the
study wasn't completely blinded. Patients with more pronounced dyssynchrony
with the ventilator would be the ones expected to receive open-label
neuromuscular blockade, and these would also be the ones in the control group.
In an effort to preserve blinding, all patients in the study had to be sedated to the
point of complete unresponsiveness prior to receiving either the cisatracurium
infusion or placebo. Heavy sedation is also known to be associated with higher
risks. These issues should not be discounted.

Lastly, much of the improvement in critical care medicine over the

last 15-20 years has been with the realization that "less is more." Heavy sedation
and routine neuromuscular blockade have given way to daily awakening trials
and analgesia-first sedation strategies. Mobilization of critically ill patients is
becoming more accepted, as is recognition and prevention of delirium.
Implementation of the ACURASYS method could be a step backwards.

Putting It Together
The point of the preceding arguments was not to convince you that
all patients with ARDS should be proned and paralyzed, and it wasn't intended to
say that proning and neuromuscular blockade are worthless. The truth is that
both may have a role in moderate-to-severe ARDS (PaO 2 /FiO 2 ≤ 150), and they
should be considered on a case-by-case basis. The patients most likely to benefit
from prone positioning are those with significant dorsal consolidation as seen on
CT imaging. Patients with more diffuse infiltrates may not see as much of a
response to changes in respiratory mechanics and pulmonary blood flow.
Additionally, the nursing care of other issues (long bone fractures, recent chest
or abdominal surgery, brain injury, etc.) may be adversely affected with proning.
Proper training and drilling of ICU staff, along with the use of a proning
checklist, should minimize the risk of turning to both patients and caregivers.

The patients most likely to benefit from therapeutic neuromuscular

blockade include those with extremely poor respiratory compliance; those with
significant dyssynchrony with the ventilator despite the best efforts of the
clinician; and those with coexisting issues like abdominal compartment
syndrome or intracranial hypertension where an improvement in thoracic
compliance could lead to an overall improvement in hemodynamics and end-
organ perfusion. Neuromuscular blockade should be questioned in patients
receiving high-dose corticosteroids due to the higher risk of ICU-acquired
weakness syndrome.

For this reason as well, cisatracurium (a benzylisoquinolone that is

metabolized in the plasma by Hofmann degradation) is preferred over
aminosteroidal neuromuscular blockers like vecuronium or pancuronium—these
agents have a higher risk of ICU-acquired weakness when concomitantly
administered with steroids. They (the aminosteroidal drugs) also depend on
hepatic and renal metabolism and their effects may be prolonged with hepatic or
renal dysfunction, which are quite common in the ICU. Peripheral nerve
stimulation should be used to monitor the depth of neuromuscular blockade, and
a daily sedation/paralytic holiday should be considered.

Prone Positioning Checklist

Indications For Prone Positioning

Hypoxemic respiratory failure with the following features:
• PaO 2 /FiO 2 ratio ≤ 150
• Diffuse bilateral lung infiltrates
• Dorsal consolidation on CT (if available)

Minimum Necessary Personnel

• 2 respiratory therapists (or other qualified personnel) to control
the airway and ventilator
• 4 turners (may be nurse, physician, patient care tech, respiratory
therapist, or student)
• 1 supervisor, who should not be involved in the proning process
itself

Turning Process

PREPARE
• Apply lubricant to eyes and tape eyelids closed
• Remove any jewelry from the patient's head or neck
• Remove any bite blocks
• Bolus necessary analgesia/sedation/neuromuscular blocker
• Confirm SpO 2 and ETCO 2 monitors are in place and functional

POSITION
• Two turners on either side of the patient (four total)
• Two respiratory therapists at the head of the patient
o One to manage the head, airway, and face pillow
o One to manage ventilator tubing and provide backup
• Supervisor at the foot of the be d

PAD (if going from SUPINE to PRONE)

• Foam face pillow, making sure the endotracheal tube is not
kinked (it may be necessary to cut out some of the foam padding)
• Two pillows each on the chest, lower pelvis, and shins
• Place a sheet over the patient (head to toe) and wrap snugly,
bundling the pillows to the patient

DISCONNECT
• Central lines (after necessary boluses)
• Arterial lines
• Hemodialysis lines
• Cardiac monitor leads
• The endotracheal tube from the ventilator
o Attach a self-inflating bag connected to oxygen
o Adjust the PEEP valve on the bag to the appropriate
level, based on the patient's oxygenation
o Put the ventilator on standby

TURN
Supervisor should read each step aloud, with verbal confirmation
by the team members

o Supervisor confirms that the airway and ventilator tubing are

under control by the respiratory therapists

o Supervisor confirms that all lines and leads have been

disconnected (SpO 2 and ETCO 2 monitors may be left in place,
unless they interfere with the turning process)

o On the supervisor's count, the team will turn the patient onto his
left/right (specify which) side, keeping the pillows tight against
the body using the sheet

o Supervisor confirms that nothing needs to be repositione d

o On the supervisor's count, the team will turn the patient to the
PRONE or SUPINE position, ensuring that the pillows and face
pad are kept in the proper position

o Respiratory therapists confirm to the supervisor that the

endotracheal tube is at the proper depth and that the tube is not
obstructed, with an appropriate ETCO2 waveform

o If PRONE , Turners confirm to the supervisor that the patient is

appropriately padded and that arms and legs are positioned
comfortably

o If SUPINE , Turners remove padding

o Reattach cardiac monitor leads, arterial line, and restart infusions

Prone position should be maintained for 16 hours, followed by 8 hours

in the supine position. Eye and mouth care is essential. Tube feeding in
the prone position is permissible if the tube is post-pyloric; otherwise,
hold tube feeding while prone and increase the rate of feeding while
supine.
 Chapter Seven
Inhaled Pulmonary Vasodilators

Positive airway pressure has a beneficial effect on left ventricular

function by reducing both preload and afterload (the transmural pressure across
the left ventricle). At the same time, however, positive airway pressure can
worsen right ventricular function—the normally low-pressure pulmonary
vascular circuit now is subjected to significant pressure from the ventilator.
Hypoxic pulmonary vasoconstriction also increases the workload on the right
ventricle. Most of the time, this doesn't affect hemodynamics too much, and
fluid loading is sufficient to maintain right ventricular output. In some patients,
though, pulmonary hypertension and right ventricular dysfunction can have a
notably adverse effect on both cardiac and pulmonary function.

Right ventricular (RV) dysfunction and even overt RV failure can be

seen with severe ARDS. It is also seen with massive or submassive pulmonary
embolism, right ventricular infarction, and in patients with preexisting
pulmonary hypertension (chronic obstructive pulmonary disease, obstructive
sleep apnea, connective tissue diseases, primary pulmonary hypertension, etc.).
RV failure can be particularly difficult to treat—the right ventricle is normally a
thin-walled structure that operates best in conditions of low vascular pressure
and resistance. A sudden increase in pulmonary vascular resistance is hard for
the RV to deal with—it just doesn't have the muscle mass of the left ventricle.
Inotropes like milrinone and dobutamine can be used to "whip the heart," but an
increase in cardiac output is often neutralized by a corresponding rise in
myocardial oxygen consumption. In this situation, a selective pulmonary arterial
vasodilator may prove to be helpful.

The most commonly used pulmonary arterial dilator in critical care

medicine is inhaled nitric oxide (iNO). iNO can be delivered by mask or through
the endotracheal tube and has a rapid vasodilatory effect on pulmonary arterioles
and capillaries. One particular advantage of iNO is that it will only cause
vasodilation in the alveolar-capillary beds that it reaches. This has the effect of
improving ventilation-perfusion matching in patients with severe hypoxemia.
Inhaled prostacyclin can also be used and has the same physiologic effect.
Intravenous pulmonary vasodilators like prostacyclin and alprostadil can be
used, but tend to have a much more potent effect on hemodynamic function and
often cause hypotension.

There have not been many clinical trials of inhaled prostacyclin, and
the evidence base is limited. iNO has been studied much more extensively, and
so further discussion will center on the use of iNO. This does not mean that
inhaled prostacyclin is not effective, and it may work just as well as iNO in
similar clinical settings. It is important to note that neither iNO nor inhaled
prostacyclin are FDA-approved for use in adults with ARDS or right ventricular
failure, and any use is off-label.

The allure of inhaled pulmonary vasodilators is that they cause

selective vasodilation only in the lung units that they can reach; they have a
rapid onset and offset; that they have minimal adverse hemodynamic effects; and
that there are no downstream metabolites. For years, this was thought to be the
case. iNO was believed to be inactivated immediately by reacting with
hemoglobin in the pulmonary capillaries. Recent research has shown that this is
not the case. iNO reacts with hemoglobin and leads to formation of nitrite and S-
nitrosohemoglobin. Nitrite can be recycled in downstream tissues to nitric oxide,
which can cause systemic capillary vasodilation. S-nitrosohemoglobin also
induces nitric oxide production, particularly in the setting of tissue hypoxia. This
couples vasodilation and deoxygenation, which may lead to mitochondrial
dysfunction. This has been demonstrated in clinical trials, where use of iNO is
associated with a higher rate of renal failure. 33 Presumably, the toxic effects of
these metabolites are not limited to the kidneys, which means that the
metabolites of iNO could contribute to multisystem organ dysfunction.

iNO and ARDS

 In patients with ARDS, iNO may improve oxygenation via selective
pulmonary vasodilatation. No studies have shown a survival benefit with this
therapy, however, and a recent meta-analysis 34 of nine clinical trials concluded
that, "Nitric oxide does not reduce mortality in adults or children with acute
respiratory distress syndrome, regardless of the degree of hypoxemia." The
reason for the lack of benefit seems to be in line with other therapies that have
been shown to improve oxygenation but not survival—very few patients with
ARDS die of refractory hypoxemia. The majority die of multisystem organ
failure, and the potentially toxic metabolites of iNO may potentiate this.
Therefore, iNO should only be used in ARDS as a true rescue therapy. It may be
helpful in patients with a PaO 2 /FiO 2 ratio less than 55 despite optimal care and
who are not candidates for other rescue therapies that have been proven to be
beneficial (prone positioning, veno-venous ECMO).

iNO and Right Ventricular Failure

Acute right ventricular failure is primarily treated with fluid loading
and inotropic support. Dobutamine and milrinone are inotropes that increase right
ventricular contractility. Milrinone, a phosphodiesterase-III inhibitor, also has
vasodilatory properties on the pulmonary circulation. Levosimendan is another
calcium-sensitizing inodilator, but it is not commercially available in the United
States.

RV failure is often associated with moderate-to-severe hypoxemia

and pulmonary dysfunction. Conventional ventilator strategies that use high
levels of PEEP or increase the mean airway pressure (like APRV) can worsen
right ventricular function and increase pulmonary vascular pressures. Inhaled
nitric oxide or prostacyclin can be used to lower the pulmonary vascular
resistance, thereby improving right ventricular function and improving gas
exchange.

When initiating inhaled pulmonary vasodilators for right ventricular

failure, a pulmonary artery catheter is strongly encouraged. Echocardiography
can also be used to evaluate contractility and to measure pulmonary artery
pressure, but it isn't available continuously and is not ideal for titrating
medications. The pulmonary artery catheter can continuously measure
pulmonary artery pressure, cardiac output, and SvO 2 . It can also be used to
calculate the pulmonary vascular resistance. This is very useful for
differentiating between conditions that cause pulmonary arterial hypertension
and those associated with pulmonary venous hypertension. Selective pulmonary
vasodilators tend to be more effective for the former.

Pulmonary vascular resistance (PVR) can be calculated by measuring

the mean pulmonary artery pressure and the pulmonary artery occlusion pressure
at end-expiration. The difference between the two measurements is then divided
by the cardiac output (in L/min).

PVR = [mean PAP – PAOP] / [CO]

A patient with a normal mean pulmonary artery pressure (20 mm Hg),

pulmonary artery occlusion pressure (10 mm Hg), and cardiac output (5 L/min)
would have a pulmonary vascular resistance of 2 mm Hg-min/L, or Wood units.
Normal PVR is 2-3 Wood units. * Conditions that elevate both the mean PAP
and the PAOP (most commonly left ventricular dysfunction, but also mitral and
aortic valvular disease) are characterized by pulmonary hypertension and a
normal PVR. This is often referred to as pulmonary venous hypertension. The
high pressure in the left atrium leads to high right-sided pressures in order to
keep the blood flowing. Caution should be used with any kind of pulmonary
vasodilator—lowering the mean PAP, while the PAOP remains elevated, often
leads to pulmonary edema.

Consider a patient who has severe systolic CHF. He has a mean PAP
of 40 mm Hg and a PAOP of 30 mm Hg. The PAOP (a.k.a. the wedge pressure)
represents left atrial pressure. Left atrial pressure equals left ventricular pressure
at the end of diastole, when blood stops flowing from the atrium to the ventricle.
The left ventricular end-diastolic pressure is elevated due to severe CHF. The
only way that blood can flow from the right ventricle through the pulmonary
vasculature and into the left ventricle is if the pulmonary artery pressure is
higher than the left ventricular pressure. Now, this patient is started on iNO. The
mean PAP falls, as predicted. iNO is a selective pulmonary vasodilator, which
means that it will not reduce the left ventricular afterload. The left atrial pressure
remains the same. If the mean PAP is now 28, and the left atrial pressure is 30,
you can see where this is going. Blood flow will reverse, leading to pulmonary
edema and hypotension.

Pulmonary arterial hypertension, on the other hand, is characterized

by an imbalance between the mean PAP and the PAOP. Thromboembolic
disease, connective tissue disease, and chronic hypoxemia are common causes.
A patient with a mean PAP of 45 mm Hg, PAOP of 15 mm Hg, and a cardiac
output of 6 L/min has PVR of 5 Wood units, suggesting pulmonary arterial
hypertension. Right ventricular dysfunction with a PVR ≥ 4 Wood units may
improve with a pulmonary vasodilator.

It is important to remember that inhaled nitric oxide or prostacyclin is

an adjunctive therapy, and not a treatment in itself. The underlying condition
leading to right ventricular failure should be treated aggressively. Pulmonary
embolism should be treated with anticoagulation and thrombolysis. Acute chest
syndrome in patients with sickle cell disease should be treated with antibiotics
and exchange transfusion. Acute myocardial infarction should be treated with
reperfusion therapy. Attention to volume status is crucial—while hypovolemia
will certainly lead to hypotension, volume overload will cause bowing of the
interventricular septum and compromise left ventricular filling. Euvolemia,
guided by echocardiography and/or pulmonary artery catheter monitoring,
should be achieved by diuresis or renal replacement therapy.

Administration of iNO and Inhaled Prostacyclin

iNO is available with a commercial delivery system that has a long
track record of reliability and safety. Using an unapproved, self-made delivery
system has the risk of unreliable dosing of iNO and potentially toxic exposure of
the patient and staff to nitrogen dioxide. Use the commercial system!

Inhaled prostacyclin can be reconstituted in saline and delivered by a

jet nebulizer system modified for use with mechanical ventilation. This requires
an aerosol delivery device that can be coordinated with the ventilator's
inspiratory cycle, which has been described in the literature . 3 5
Initial Dosing of iNO
iNO should be started at 20 parts per million (ppm). A successful
response is a reduction in the mean pulmonary artery pressure by at least 10%,
and usually an improvement in the PaO 2 by at least 20 mm Hg. If the patient
does not respond within 5-10 minutes, then a higher concentration (40 ppm, or
even 80 ppm) can be tried. Most patients who are going to respond will do so at
20 ppm. iNO should be stopped in those patients who do not have an initial
response to therapy.

In patients who respond, the dose of iNO should be lowered in 5-10

ppm increments every 15-30 minutes, to a floor of 5 ppm. An increase in the
mean PAP by ≥ 5 mm Hg, or a fall in the SpO 2 by ≥ 5%, should be treated by
increasing the dose of iNO back to the level where it was effective.
iNO Initial Dosing Algorithm
Weaning iNO
Once the patient has begun showing signs of recovery (improvement
in gas exchange, less need for inotropes), iNO can be reduced or stopped
altogether. This should be done slowly, as abrupt discontinuation of the drug can
lead to rebound hypoxemia and pulmonary hypertension.

When the patient is tolerating an iNO dose of 5 ppm, the drug can be
weaned off. The iNO dose should be lowered by 1 ppm every 30 minutes. If the
mean PAP increases by ≥ 5 mm Hg, or the SpO 2 falls by ≥ 5%, the iNO should
be returned to 5 ppm and further attempts at weaning should be postponed for at
least 12 hours. Once the iNO is at 2 ppm, a single dose of sildenafil 20 mg can
be given (if it's not already being administered). This may help prevent rebound
pulmonary hypertension as the drug is weaned off. After administration of the
sildenafil, continue to reduce the iNO by 1 ppm every 30 minutes until it's off.
* PVR is often expressed in dyne-sec-cm-5. This is obtained by multiplying the number of Wood units by
79.9. I'm not sure why this is. I find it easier to use Wood units.
 Chapter Eight
Veno-Venous ECMO

There are times when a patient's lung disease is so severe that

adequate gas exchange is either impossible, or can only be accomplished with
prohibitively high airway pressures and tidal volumes. When this is the case,
veno-venous extracorporeal membrane oxygenation (VV ECMO) should be
considered as a rescue therapy. This chapter is simply an overview of the use of
extracorporeal support and is designed to familiarize the reader with the
rationale for its use. Those desiring to treat patients with ECMO are strongly
encouraged to attend a training program sponsored by the Extracorporeal Life
Support Organization (ELSO).

VV vs. VA
Veno-venous ECMO is considerably different from veno-arterial
ECMO (VA ECMO). VA ECMO is similar to heart-lung bypass. Blood is
drained from the venous side using a cannula placed in the femoral vein. The
blood is pumped through an oxygenator, and the fully oxygenated blood is
returned to the patient via a cannula placed in the femoral or subclavian artery
(Figure 1). In neonates, the carotid artery is often used; in adults, however, the
carotid artery is avoided due to the risk of stroke. With VA ECMO, the
extracorporeal circuit can support both the pulmonary and the cardiac systems.
The flow through the pump can make up for even the most severe heart failure.
In fact, the primary indication for VA ECMO in adults is refractory cardiogenic
shock.
The VA ECMO circuit provides both respiratory and cardiac support by
pumping oxygenated blood directly into the aorta. The pump flow is
sufficient to replace the entire cardiac output, if necessary .
 VV ECMO, on the other hand, provides no cardiac support. Blood is
drained from the inferior vena cava through a cannula placed in the femoral vein.
After being pumped through an oxygenator, the blood is returned to the right
atrium through a cannula placed in the internal jugular vein.
 Dual-lumen cannulas are also available, and function similarly to
dual-lumen hemodialysis catheters (albeit much larger, to accommodate a flow of
4-7 L/min). The dual-lumen cannula is placed via the right internal jugular vein.
It passes through the superior vena cava and into the inferior vena cava. The
siphon, or drainage, ports are at the tip of the cannula in the IVC. Using
transesophageal echocardiography, the cannula is manipulated so that the return
port is directed over the tricuspid valve. This helps reduce the risk of
recirculation.

How VV ECMO Provides Respiratory Support

The best way to visualize VV ECMO is to consider the entire circuit
as an extension of the right atrium. Normally, venous blood returning to the right
atrium has an SvO 2 of 70-80%. In a markedly hypoxemic patient, the SvO 2 is
much lower—50-60% is the norm. The venous blood goes through the
pulmonary vascular system, where oxygen is picked up and CO 2 is unloaded
(and ventilated off). Obviously, if a patient has severe ARDS, then the degree to
which this gas exchange can occur is quite limited. Blood with an SvO 2 of 50%
may only rise to an SaO 2 of 80%, even while the patient is breathing 100%
oxygen and receiving high levels of PEEP.

When VV ECMO is initiated, some (but not all) of the venous blood
is siphoned into the circuit. A pump drives a blood flow of 4-7 L/min through a
membrane oxygenator. As the blood passes through the oxygenator, the
hemoglobin becomes fully saturated and the PaO 2 may rise as high as 400-500
mm Hg. When this blood, with an SaO 2 of 100%, is returned to the right atrium,
it mixes with the remaining venous blood and then proceeds through the
pulmonary circulation. If half of the venous blood has an SvO 2 of 60% and the
other half has an SvO 2 of 100% (thanks to the ECMO circuit), the total venous
return going through the pulmonary circulation has an SvO 2 of roughly 80%.
With higher pump flows, a greater percentage of the venous return is oxygenated
using the ECMO circuit, leading to a higher overall SvO 2 . In most patients, the
ECMO flow can be increased to the point where the total SvO 2 is 85-90% .

Here's where VV ECMO becomes really cool. It's also where you
have to remember the principles of oxygen delivery [see the earlier chapter in
this book]. With a high enough cardiac output and hemoglobin, oxygen delivery
to the tissues can be maintained even with mild-to-moderate hypoxemia. In other
words, an SaO 2 of 80-85% is perfectly fine so long as cardiac output is sufficient
and there is enough hemoglobin to carry the bound oxygen.

If an SaO 2 of 80-85% is enough to get the job done (with adequate

cardiac output and hemoglobin), and if the SvO 2 can be kept at 80-85% with VV
ECMO, then there is no need for pulmonary gas exchange whatsoever . This
is a very important point, and is the cornerstone to understanding why VV
ECMO can be an effective rescue therapy for severe ARDS. If the venous blood
has an SvO 2 of 85% and it flows through lungs that contribute absolutely
nothing, then the blood reaching the left atrium will have an SaO 2 of 85%. Since
we've already established that an SaO 2 of 85% is sufficient if there's adequate
cardiac output and hemoglobin, then there is no need to "beat up the lungs" with
high PEEP, high ventilator rates, or any of the other usual things that are done
for severe respiratory failure. Instead, the ventilator can be put on what are
generally considered "rest settings."

Ventilator Rest Settings on VV ECMO *

• Pressure Control Ventilation

• Rate 10 breaths/minute
• I-time 1.0-2.0 seconds, adjusted for comfort
• P INSP 25 cm H 2 O
• PEEP 10 cm H 2 O
• FiO 2 30%

Gas exchange in the ECMO circuit is a function of blood flow

through the oxygenator and the gas flow of oxygen across the oxygenator's
membrane. The oxygen that is flowing across the membrane is known as the
sweep gas because it "sweeps off" CO 2 from the blood in the membrane. CO 2
has a much higher solubility than oxygen, and so it can be rapidly eliminated by
increasing the sweep gas flow. Oxygenation can be increased by raising the rate
of blood flow through the membrane oxygenator. Put simply, circuit flow
controls oxygenation, while the flow of sweep gas controls ventilation. Sweep
gas is usually 2-6 L/min. The FiO 2 of the sweep gas is initially set at 1.0 in order
to attain the best oxygenation of the blood in the circuit.

Initial VV ECMO Settings

• Circuit blood flow of 50-60 mL/kg
• Adjust circuit flow to keep patient SaO 2 80-85%
• FiO 2 on sweep gas of 100%
• Set the sweep gas flow about the same as the circuit
flow
• Adjust sweep gas flow to keep PaCO 2 35-45

Most of the time, some ventilator support is necessary. This isn't to

provide additional gas exchange support, but instead to improve patient comfort
and to prevent complications. If a patient is put on VV ECMO and then
extubated, the lack of any positive pressure on the lungs will lead to near-
complete alveolar collapse and consolidation. This can cause significant
tachypnea and respiratory distress. Alveolar consolidation also prevents the
normal clearance of secretions from the pulmonary tree, which can lead to
pneumonia or lung abscess.

That said, recent experience with VV ECMO has shown that as

patients recover, they can spend more and more time off the ventilator. This is
important because it means that physical therapy and mobilization can begin
early on, even while on VV ECMO. This is much easier with the dual-lumen
cannula placed through the internal jugular vein. Early tracheostomy should be
done as soon as it's feasible in order to lessen sedation and begin mobilization.
There is nothing like seeing a patient with severe ARDS walk down the hallway,
with the ECMO circuit being pushed behind him .

Weaning VV ECMO
As the patient begins to recover, the FiO 2 of the sweep gas can be
lowered. Circuit flow, once established, should not be lowered—lower blood
flow increases the risk of thrombosis in the circuit. Keep in mind that VV
ECMO is just like a really big right atrium. As the FiO 2 on the sweep gas is
reduced, the blood flowing through the oxygenator will pick up less oxygen.
That means that the proportion of gas exchange that has to occur through the
patient's lungs is increasing. Once the FiO 2 on the sweep gas is 0.21, the ECMO
circuit is contributing nothing at all to the patient's oxygenation—it's all being
done with low-level ventilator support. Blood is simply flowing through the big
right atrium but there's no assistance with oxygenation. If the patient's condition
is acceptable, it's time to come off ECMO.

Patient Selection
This is often the most difficult part of using VV ECMO for severe
acute respiratory failure. For many years, ECMO was used predominantly for
neonates with infant respiratory distress syndrome, meconium aspiration, and
congenital diaphragmatic hernia. In recent years, however, ECMO has become
more popular for older children and adults. The H1N1 influenza pandemic of
2009 accelerated the interest in ECMO, particularly VV ECMO, as a rescue
therapy. The CESAR trial, published in The Lancet in 2009, demonstrated a
survival benefit for patients with severe influenza-related ARDS transferred to
ECMO centers. 36 Only 75% of those randomized to the ECMO arm actually
received ECMO, which is an interesting finding. It may be that the true benefit
was treating patients in high-volume centers with the appropriate expertise and
ability to provide rescue therapies, including ECMO, rather than the provision of
ECMO itself .

Indications for VV-ECMO

• Hypoxic respiratory failure with a predicted mortality risk ≥ 50%
a. PaO 2 /FiO 2 < 150 on FiO 2 > 90% despite optimal care for 6
hours or more
b. Murray Score * ≥ 3 despite optimal care for 6 hours or more
• Hypercapnic respiratory failure refractory to treatment with pH <
 7.15
• Acute onset of a potentially reversible cause of respiratory failure
• Age ≤ 65
• Immediate respiratory collapse that is unresponsive to optimal care
(obstructed airway, etc.)

Contraindications to VV-ECMO +
• Mechanical ventilation at high settings (e.g. FiO 2 ≥ 90%, P PLAT >
30, PEEP ≥ 15) for 7 days or longer
• Contraindication to anticoagulation
• Absolute neutrophil count < 500/mm3
• Major CNS damage or other nonreversible comorbidity
• Age > 65
• Progression of chronic respiratory disease to the point of
respiratory failur e

Prior to initiation of VV-ECMO, the following steps should be taken

to improve the patient's condition. These are listed in order of preference,
although not all are required prior to cannulation for ECMO.

1. Lung protective ventilation using a tidal volume of 4-6 mL/kg

predicted body weight, with PEEP according to the ARDSNet study
protocol
2. Airway pressure release ventilation, with a P HIGH up to 35 cm H 2 O
3. Prone positioning for 16 hours, followed by supine positioning for 8
hours
4. Diuresis or CRRT to within 105% of dry weight, if hemodynamics
permit
5. Bronchoscopy with therapeutic aspiration of the tracheobronchial
tree

If the patient has not improved with the aforementioned therapy, VV ECMO
team should be considered. Additional rescue maneuvers that can be tried
include:
 6. Inhaled nitric oxide
7. High frequency oscillatory ventilation

The Extracorporeal Life Support Organization (ELSO) provides extensive expert

guidelines for patient selection and referral at its website: www.elso.org .

Ventilator Management on Veno-Arterial ECMO

The bulk of this chapter is concerned with VV ECMO as a rescue
therapy for severe respiratory failure. The ventilator management in VV ECMO
is quite easy—let the circuit do the heavy work, and use the ventilator to simply
keep the lungs open without injuring them. It is quite common for the tidal
volume to be less than 100 mL while on the rest settings described earlier, and
that's fine—after all, the whole point of VV ECMO is to let the patient's lungs
rest and recover.

With VA ECMO, on the other hand, it's important to keep in mind

that while the ECMO circuit provides oxygenated blood into the aorta, the blood
in the aortic root is more dependent on the patient's native cardiopulmonary
function. That means that the blood flowing from the pulmonary circulation into
the left atrium and ventricle is delivered preferentially into the coronary ostia
and the aortic arch. In other words, oxygen delivery to the coronary arteries still
depends on mechanical ventilatory support. The rest of the body, particularly the
lower thorax and abdomen, gets its oxygen delivery from the ECMO circuit.

The exact degree to which different areas of the body are perfused
depends on how strong the heart is. Consider a patient with zero native cardiac
function. All of his arterial flow, including the coronary ostia, depends on the
ECMO circuit flow. As he begins to recover cardiac function, the native heart
will begin pumping blood into the aortic arch. As the heart grows stronger, it will
perfuse more and more blood to the coronary arteries and the vessels coming off
the aortic arch. This can lead to a situation where the lower half of the body is
well-oxygenated by the ECMO circuit, while the upper body (including the
brain) is relatively hypoxic. The management of the "blue nose syndrome" is
beyond the scope of this chapter.
 Therefore, simply putting a patient receiving VA ECMO on rest
settings like you would do with VV ECMO risks myocardial hypoxia. A higher
FiO 2 and PEEP may be necessary. Unlike with VV ECMO, ventilator settings
should be adjusted to attain adequate gas exchange. The most accurate site for
arterial blood gas monitoring depends on where the arterial ECMO cannula is
located. In most cases, the cannula is placed in a femoral artery and into the
descending aorta. After the coronary ostia, the next branch off the aortic arch is
the right brachio cephalic artery. A blood gas specimen from a right radial arterial
line will give the most accurate measure of the native heart's oxygen delivery. In
the cases of central subclavian artery cannulation, there is much less discrepancy
in regional oxygen delivery. The arterial line is generally placed in the radial
artery on the opposite side from the subclavian artery cannula.

* Always remember that the whole point of VV ECMO for respiratory failure is to let the lungs rest and
recover. If they look whited-out on the CXR and the tidal volume on these settings is < 100 mL, so be it!
Resist the temptation to use the vent for gas exchange. Let the ECMO circuit do the work.

* http://cesar.lshtm.ac.uk/murrayscorecalculator.htm

+ Contraindications are relative, not absolute; however, the presence of these conditions is associated with a
higher risk of treatment failure.
 Chapter Nine
2 A.M .

Your mother was right—nothing good happens after 2 A.M.

Especially in the ICU. If you get called to the bedside of a critically ill patient in
the middle of the night, it is almost never going to be for good news.

This chapter is designed as a step-by-step approach to the

mechanically ventilated patient who is either not getting better or deteriorating
despite your best efforts. It begins with the things you should check when a
patient has a sudden change in condition. Some of this is also discussed in The
Ventilator Book , but it doesn't hurt to read it again. There is also a stepwise
algorithm for escalating ventilator support for severe ARDS, as well as a guide
to initial vent settings for acute lung injury and obstructive lung disease. Despite
the title of this chapter, this information is useful any time of day.

First Things
Whenever a critically ill patient takes a turn for the worse, the initial
assessment should go back to the ABCs. This is drilled repeatedly in Advanced
Cardiac Life Support and Advanced Trauma Life Support courses, and for good
reason. For a mechanically ventilated patient, think Tube, Sounds , Sats . Make
sure the endotracheal tube is in place and is patent—capnography is very helpful
in this regard. Auscultate the chest to make sure there is bilateral air entry, and
listen for wheezing or rales that might point you toward the reason for the
patient's deterioration. Ensure that the patient is adequately oxygenated—
hypoxemia can be due to a mechanical problem, a pulmonary issue, a
cardiovascular issue, or a combination of any of these.
Another useful mnemonic for evaluating the crashing ventilated patient is
DOPES *:

• D isplacement of the endotracheal tube—assess with capnography to

make sure the tube is still in the trachea. Mainstem intubation can also
make a ventilated patient get worse. The tube usually migrates down the
right mainstem bronchus, so if breath sounds are not equal pull the tube
back a few centimeters and reassess.

• O bstruction of the endotracheal tube—again, capnography can be

helpful. A suction catheter that doesn't easily pass is another clue. Make
sure the tube isn't kinked. Tube obstructions from secretions can sometimes
be cleared with a bronchoscope or a CAM Rescue Cath™. If there's any
doubt, take a look with a laryngoscope and reintubate the patient with a
fresh tube.

• P neumothorax—chest X-ray is usually helpful but may not be

immediately available. Bedside ultrasound can show a lack of pleural
sliding. If there's concern for a tension pneumothorax (hypotension,
hypoxemia, and absent breath sounds), emergent decompression should be
strongly considered.

• E quipment malfunction—the best way to exclude this as a cause is to

disconnect the patient from the ventilator, attach a self-inflating bag, and
manually ventilate while further assessments are performed .

• S tacked breaths—this almost always happens in severe obstructive lung

disease. Auto-PEEP can progress to the point where it causes hypotension
or even pulselessness. A clue is when the patient is very difficult to
manually ventilate and breath sounds are markedly diminished bilaterally.
The treatment is disconnection of the vent or bag—if there's a rush of air
out of the endotracheal tube, followed by hemodynamic improvement, then
auto-PEEP is the culprit. Reconnect the patient to the ventilator with a
lower respiratory rate and ensure that there's enough time for exhalation.
Initial Ventilator Setup
These are general guidelines for initial ventilator settings, divided
between acute lung injury (sepsis, trauma, ARDS, pulmonary edema, etc.) and
obstructive lung disease (asthma, COPD). The specific ventilator management of
these patients must be individualized, and the general principles are described in
more detail elsewhere. The purpose of these guidelines is to provide a quick
reference that is applicable to the majority of patients placed on the ventilator in
the ICU or the Emergency Department.
Acute Lung Injury Ventilator Setup
[ARDS, Sepsis, Trauma, Pneumonitis, Pulmonary Edema]
Obstructive Lung Disease Ventilator Setup
[Status Asthmaticus, COPD, Bronchospasm]
Escalating Therapy for ARDS
So here you are, at 2 A.M., and you're caring for a patient with severe
ARDS. There are three things you need to know—where to start, what goals to
aim for, and what to do if what you're doing isn't working.

The most evidence-based recommendation is to start with a lung-

protective strategy, a la ARDSNet. Specifically, that means using lower tidal
volumes (4-6 mL/kg PBW) and enough PEEP to open up and stabilize
vulnerable lung units. Following the guidelines listed earlier in this chapter for
acute lung injury is appropriate for initial vent settings, and adjustments should
be made as necessary to achieve the listed goals. These goals include adequate
oxygenation, a reasonable (but not perfect) degree of ventilation, and keeping
the alveolar pressure (as represented by the plateau pressure on the vent) less
than 30. If the plateau pressure exceeds 30, lowering the tidal volume to 5 or
even 4 mL/kg PBW is appropriate. The respiratory rate can be increased if
needed, but remember that hypercapnia is not terribly dangerous in most
settings. Adjustment of the PEEP can be done using PEEP-FiO 2 tables, pressure-
volume curves, esophageal pressure monitoring, or any of the other methods
described in the earlier chapter.

If an ARDSNet approach isn't working, you may need to escalate.

First off, however, we should define "isn't working." Most of the time, the
ARDSNet ventilator management strategy is sufficient to achieve adequate
oxygenation and provide lung protection. Rescue therapies like the ones
discussed here do not carry strong evidence-based medicine recommendations,
and should only be considered for moderate-to-severe respiratory failure. Many
of the clinical trials use a PaO 2 /FiO 2 ratio ≤ 150, which seems reasonable. For
the purpose of this algorithm, "isn't working" is defined as a PaO 2 /FiO 2 ratio ≤
150 despite optimal treatment with the listed modality. For example, the
ARDSNet approach would be considered to be "not working" if the PaO 2 /FiO 2
ratio is ≤ 150 despite using a low tidal volume and a PEEP of 15-20. This does
not mean, however, that you must try something else. A mode of ventilation or
support has only failed if it can't provide the minimum necessary oxygenation
for a patient, or if the adverse effects on the lungs or hemodynamics are
intolerable. True refractory hypoxemic respiratory failure would be best defined
as a PaO 2 /FiO 2 ratio < 55 despite optimal treatment.
ARDS Escalation Algorithm

for truly refractory hypoxemia

salvage therapy only

INO
HFO V

Airway Pressure Release Ventilation *

 My go-to rescue mode of ventilation for ARDS is APRV. APRV
works by increasing the mean airway pressure while avoiding excessively high
distending pressure on the alveoli. It does this by going up to an inspiratory
pressure (P HIGH ) and holding it for 3 seconds, 4 seconds, or even longer. Brief
(usually less than one second) releases of airway pressure allows the gas in the
patient's lungs to escape, carrying off CO 2 , and the lungs are then rapidly re-
expanded to the P HIGH .

APRV works very well for diffuse, bilateral lung injury. It does not
work as well when one lung is considerably worse than the other, and it doesn't
work very well in patients with significant obstructive pulmonary disease due to
the air-trapping it creates. Patients with tenuous hemodynamics may also do
poorly with APRV if the distending airway pressure impacts venous return or
pulmonary blood flow. APRV does seem to be well-tolerated in most patients
with ARDS, however, and it has the added benefit of permitting spontaneous
ventilation and not requiring heavy sedation and neuromuscular blockade.
APRV Setup Flowchart
Prone and Paralyze
If a patient with severe ARDS has a contraindication to APRV, or
doesn't do well on APRV, I think there is sufficient evidence to recommend
prone positioning (usually in conjunction with neuromuscular blockade). The
success of prone positioning depends greatly on a well-trained staff and
meticulous avoidance of complications like pressure injuries and dislodgement
of life support devices. Therefore, using a checklist each time the patient is
turned is highly recommended. Regular training of the ICU staff is also
necessary. The ventilator should be kept on ARDSNet-style settings to help
protect the lungs from injury, and the same goals for gas exchange apply.

If neuromuscular blockade is used, cisatracurium is the preferred

agent for the reasons described in the chapter in this book. Daily interruption of
the paralytic drug is advisable to avoid accumulation and prolonged
neuromuscular blockade.

Patients should be prone for 16 hours, followed by 8 hours in the

supine position. For both proning and paralysis, the therapy should be continued
until the patient begins to show signs of recovery. Most of the time, this will
mean a PaO 2 /FiO 2 ratio > 150 while supine and off the paralytic agent.

Concurrent Therapy
While much of the treatment for ARDS focuses on respiratory
support, it's important to recognize that volume overload, excessive pulmonary
secretions, and cardiac dysfunction can also contribute to severe respiratory
failure. In addition to providing optimal ventilator support, the following should
be considered:

• Diuresis or ultrafiltration as tolerated, with a goal of reaching 105% of

the patient's "dry weight." Volume overload is an especially common cause
of persistent hypoxemia in ventilated patients.

• Therapeutic bronchoscopy to clear the tracheobronchial tree. This can

also be diagnostic if the primary cause of respiratory failure is infection or
 alveolar hemorrhage.

• Echocardiography or a pulmonary artery catheter to identify and treat

cardiac dysfunction.

Veno-Venous ECMO
VV ECMO is the ultimate rescue strategy for respiratory failure, and
it works by essentially taking the lungs out of the equation so they can rest and
recover. The indications for VV ECMO are outlined in the chapter in this book.
VV ECMO carries very real risks—the cannulas are very large, and the
anticoagulation necessary for the circuit often leads to significant bleeding and
the need for multiple transfusions. It is also quite resource-intensive and can
only be performed in ECMO centers. Nevertheless, VV ECMO is growing in
popularity as a method of support for adults with severe respiratory failure. If a
patient appears to be heading toward this and is not already at an ECMO center,
early transfer should be arranged if possible.

Other Rescue Therapies

Two rescue therapies for ARDS that are not supported by the medical
literature, at least in adults, are inhaled nitric oxide (iNO) and high frequency
oscillatory ventilation (HFOV). That doesn't mean that they are worthless, but
based on the published data, they should not be included in a general treatment
algorithm .

As discussed previously in the book, iNO can be helpful for the

treatment of acute right ventricular failure. For patients with ARDS, however, no
mortality benefit has been described and some trials have shown an increase in
harm with iNO. For that reason, the use of iNO should be limited to those
patients with demonstrable acute right ventricular failure and pulmonary arterial
hypertension; or, ARDS with truly refractory hypoxemia (PaO 2 /FiO 2 < 55)
when other rescue therapies have either failed or are not an option.
 HFOV was, at one time, a commonly used rescue therapy. The
OSCILLATE trial, published in 2013, was a multicenter trial examining the use
of HFOV early in the treatment of moderate-to-severe ARDS. 37 The
investigators found no evidence of benefit and a trend toward increased in-
hospital mortality. This was validated by the OSCAR trial, another multicenter
trial of HFOV in ARDS that found similar results. 38 For this reason, HFOV
should be limited to those patients who have a specific need like a large
bronchopleural fistula, or those with truly refractory hypoxemia (PaO 2 /FiO 2 <
55) when other rescue therapies have either failed or are not an option.
HFOV Setup Flowchart

Initial HFOV Settings

• Mean Airway Pressure: 2-5
cm H 2 O higher than mean
airway pressure on
conventional ventilation
• Amplitude: PaCO 2 (on last
ABG) + 20
• Frequency: 5 Hz
• Inspiratory time: 33%
• FiO 2 100%
* http://wikem.org/wiki/detenoration_after_intubation

* There's a very good chapter on APRV in The Ventilator Book , if I do say so myself!
References
1 Hickling KG, Henderson SJ, Jackson R. Low mortality associated with low

volume pressure limited ventilation with permissive hypercapnia in severe adult

respiratory distress syndrome. Intensive Care Med 16: 372–377.
2 Hickling KG, Walsh J, Henderson S, Jackson R. Low mortality rate in adult

respiratory distress syndrome using low-volume, pressure-limited ventilation

with permissive hypercapnia: a prospective study. Crit Care Med 22:1568-1578.
3 Ventilation with lower tidal volumes as compared with traditional tidal

volumes for acute lung injury and the acute respiratory distress syndrome. The
Acute Respiratory Distress Syndrome Network. N Engl J Med 342:1301-1308.
4 Frumin MJ, Epstein RM, Cohen G. Apneic oxygenation in man.

Anesthesiology 20(6): 789-798.

5 Hotchkiss JR, Blanch L, Murias G, et al. Effects of decreased respiratory

frequency on ventilator-induced lung injury. Am J Respir Crit Care Med 161:

463-468.
6 Laffey JG, O'Croinin D, McLoughlin P, Kavanagh BP. Permissive hypercapnia

—role in protective lung ventilatory strategies. In Applied Physiology in

Intensive Care Medicine 2 (pp. 111-120). Springer Berlin Heidelberg.
7 Akca O, Doufas AG, Morioka N, et al. Hypercapnia improves tissue

oxygenation. Anesthesiology 97: 801-806.

8 Mekontso Dessap A, Charron C, Devaquet J, et al. Impact of acute hypercapnia

and augmented positive end-expiratory pressure on right ventricle function in

severe acute respiratory distress syndrome. Intensive Care Med 35:1850-1858 .
9 Petridis AK, Doukas A, Kienke S. et al. Acta Neurochir 152: 2143.
10 Beckman JS, Koppenol WH. Nitric oxide, superoxide, and peroxynitrite: the

good, the bad, and ugly. Am J Physiol 271: C1424-C1437.

11 O'Croinin DF, Nichol AD, Hopkins N, et al. Sustained hypercapnic acidosis

during pulmonary infection increases bacterial load and worsens lung injury. Crit
Care Med 36: 2128-2135.
12 Higher versus lower positive end-expiratory pressures in patients with the

acute respiratory distress syndrome. The National Heart, Lung, and Blood
Institute ARDS Clinical Trials Network. N Engl J Med 2004; 351: 327-336.
13 Crotti S, Mascheroni D, Caironi P, et al. Recruitment and derecruitment

during acute respiratory failure: a clinical study. Am J Respir Crit Care Med
2001; 164:131–140.
14 Mercat A, Richard JC, Vielle B, et al. Positive end-expiratory pressure setting

in adults with acute lung injury and acute respiratory distress syndrome: a
randomized controlled trial. JAMA 2008; 299:646–655.
15 Washko GR, O'Donnell CR, Loring SH. Volume-related and volume-

independent effects of posture on esophageal and transpulmonary pressures in

healthy subjects. J Appl Physiol 2006; 100:753–758.
16 http://www.coopersurgical.com/Products/Detail/Esophageal-Balloon-Catheter-

Se t
17 Talmor D, Sarge T, Malhotra A, et al. Mechanical ventilation guided by

esophageal pressure in acute lung injury. N Engl J Med 2008; 359:2095–2104.

18 Chiumello D, Cressoni M, Carlesso E, et al. Bedside selection of positive end

expiratory pressure in mild, moderate, and severe acute respiratory distress

syndrome. Crit Care Med 2014; 42:252–264.
19 Gattinoni L, Carlesso E, Brazzi L, et al. Friday night ventilation: a safety

starting tool kit for mechanically ventilated patients. Minerva Anestesiol 2014;
80:1046–1057.
20 Peters JI, Stupka JE, Singh H, et al. Status asthmaticus in the medical
intensive care unit: a 30-year experience. Respir Med 2012 Mar; 106(3):344-8.
21 Tassaux D, Jolliet P, Thouret JM, et al. Calibration of seven ICU ventilators

for mechanical ventilation with helium-oxygen mixtures. Am J Respir Crit Care

Med 1999;160(1): 22–32.
22 Venkataraman, ST. Heliox during mechanical ventilation. Respir Care 2006;

51(6):632-9.
23 Goyal S, Agrawal A. Ketamine in status asthmaticus: a review. Indian J Crit

Care Med 2013; 17(3): 154-61.

24 Strayer RJ, Nelson LS. Adverse events associated with ketamine for

procedural sedation in adults. Am J Emerg Med 26(9): 985–1028.

25 Kuyper LM, Paré PD, Hogg JC, et al. Characterization of airway plugging in

fatal asthma. Am J Med 2003; 115: 6-11.

26 Guérin C, Reignier J, Richard JC, et al, PROSEVA Study Group. Prone

positioning in severe acute respiratory distress syndrome. N Engl J Med 2013;

368: 2159.
27 Papazian L, Forel JM, Gacouin A, et al. Neuromuscular blockers in early

acute respiratory distress syndrome. N Engl J Med 2010; 363:1107-1116.

28 Gattinoni L, Tognoni G, Pesenti A, et al, Prone-Supine Study Group. Effect of

prone positioning on the survival of patients with acute respiratory failure. N

Engl J Med 2001; 345: 568-573.
29 Guérin C, Gaillard S, Lemasson S, et al. Effects of systematic prone

positioning in hypoxemic acute respiratory failure: a randomized controlled trial.

JAMA 2004; 292: 2379-2387.
30 Taccone P, Pesenti A, Latini R, et al, Prone-Supine II Study Group. Prone

positioning in patients with moderate and severe acute respiratory distress

syndrome: a randomized controlled trial. JAMA 2009; 203:1977-1984.
31 Forel JM, Roch A, Marin V, et al. Neuromuscular blocking agents decrease
inflammatory response in patients presenting with acute respiratory distress
syndrome. Crit Care Med 2006; 34: 2749-2757.
32 Gainnier M, Roch A, Forel JM, et al. Effect of neuromuscular blocking agents

on gas exchange in patients presenting with acute respiratory distress syndrome.

Crit Care Med 2004; 32:113-119.
33 Adhikari NK, Burns KE, Friedrich JO, et al. Effect of nitric oxide on

oxygenation and mortality in acute lung injury: systematic review and meta-
analysis. BMJ 2007; 334(7597): 779 .
34 Adhikari NK, Dellinger RP, Lundin S, et al. Inhaled Nitric Oxide Does Not

Reduce Mortality in Patients With Acute Respiratory Distress Syndrome

Regardless of Severity: Systematic Review and Meta-Analysis. Crit Care Med
2014; 42: 404–12.
35 Siobal MS, Kallet RH, Pittet JF, et al. Description and evaluation of a delivery

system for aerosolized prostacyclin. Respir Care 2003; 48(8): 742-753.

36 Peek GJ, Mugford M, Tiruvoipati R, et al. Efficacy and economic assessment

of conventional ventilatory support versus extracorporeal membrane

oxygenation for severe adult respiratory failure (CESAR): a multicentre
randomised controlled trial. Lancet 2009; 374:1351-1363.
37 Ferguson ND, Cook DJ, Guyatt GH, et al. High-frequency oscillation in early

acute respiratory distress syndrome. N Engl J Med 2013; 368: 795-805.

38 Young D, Lamb SE, Shah S, et al. High-frequency oscillation for acute

respiratory distress syndrome. N Engl J Med 2013; 368: 806-813.

 Also available from First Draught Press

The Ventilator Book

The best-selling guide to the basics of

mechanical ventilation.

Available at Amazon.com .

The Ventilator App

The iOS app designed to be

used in the ICU or ED. It
includes a tidal volume
calculator, a PEEP-FiO2
adjustment tool, & a ventilator
troubleshooting guide for easy
reference.

Available in the App Store.

About The Author

William Owens, MD, is the Director of the Medical Intensive Care Unit at
Palmetto Health Richland, a tertiary referral center in Columbia, SC. He is also
the Division Chief for Pulmonary, Critical Care, and Sleep Medicine in the
Palmetto Health-USC Medical Group and an Associate Professor of Clinical
Medicine with the University of South Carolina. He has also served on the
faculty at the University of Pittsburgh School of Medicine.

Dr. Owens is a graduate of The Citadel and the University of South Carolina
School of Medicine. He trained in Emergency Medicine at the Earl K. Long
Medical Center in Baton Rouge, LA. He did his fellowship training in Critical
Care Medicine at the University of South Florida in Tampa, FL. He is board-
certified in Emergency Medicine, Critical Care Medicine, and Neurocritical Care
Medicine. He has spoken at regional and national conferences and has published
articles in the peer-reviewed medical literature.

Throughout his career, Dr. Owens has been an active clinician and educator. He
enjoys training physicians, nurses, and respiratory therapists in the care of the
most seriously ill and injured patients and is a firm believer in a holistic
approach to critical care medicine. He believes in the rational application of
physiology and in always questioning our assumptions.

Dr. Owens lives in Columbia, SC, with his wife and three free-range children.
He also lives with two large St. Bernards and a beehive with about 60,000 bees.
He enjoys mountain biking, whitewater kayaking, playing lacrosse, and going on
family adventures .