Ebook Principles of critical care (4th edition): Part 2

PART 6

/
9
ri 9

Neurologic Disorders h

/: /

a
k

s
tt p

e
s

/r u

.t c

a
t
r/

h

section06.indd 755

1/23/2015 12:55:24 PM


Onset

Acute (hours to days)

Insidious (months to years)

Course

Fluctuating

Progressive

Diagnostic
Features

• Impaired ability to focus, shift
• Memory impairment plus one
or sustain attention
of the following:
• Change in cognition (eg, memory • Aphasia
impairment, disorientation
• Apraxia
or language) or development in
• Agnosia
perceptual disturbances
• Impaired executive functioning
• Fluctuating course

• Impairments must be severe
enough to cause impairments
in social or occupational
functioning and represent
a decline from baseline
• Sleep/wake disturbances
• Extremes in psychomotor activity
• Emotional disturbances (fear,
anxiety, depression, irritability,
euphoria, apathy)

Common
Causes

• Acute medical illness
• Medication/substance/toxin
ingestion or withdrawal
• Multifactorial

.t c

s
tt p







/: /

a
k

1. Disturbance of consciousness, with reduced awareness of the environment and impaired ability to focus, sustain or shift attention.
2. Altered cognition (eg, memory impairment, disorientation, or language disturbance) or the development of a perceptual disturbance
(eg, delusion, hallucination, or illusion) that is not better accounted
for by preexisting or evolving dementia.
3. Disturbance develops over a short period of time (usually hours to
days) and tends to fluctuate during the course of the day.
4. Evidence of an etiological cause, which the DSM-IV uses to classify
delirium as Delirium Due to a General Medical Condition, SubstanceInduced Delirium, Delirium Due to Multiple Etiologies, or Delirium
Not Otherwise Specified.

h

Historically, two words were used to describe acutely confused patients.
The Roman word delirium referred to an agitated and confused person
(ie, hyperactive delirium). The Greek word lethargus was used to describe
a quietly confused person (ie, hypoactive delirium). ICU patients commonly demonstrate both subtypes of delirium as they progress through

different stages of their illness and therapy. In both subtypes, the patient’s
brain is not functioning normally. It therefore makes sense that the
original derivation of delirium comes from the Latin word deliria, which
literally means to “be out of your furrow.” For greater clarity and to avoid
misuse of terms such as dementia and delirium, Table  82-1 lists basic
definitions and clinical characteristics of each syndrome.
Delirium in the ICU has been referred to in the medical literature using
a multitude of terms, including ICU psychosis, ICU syndrome, brain
failure, encephalopathy, postoperative psychosis, acute organic syndrome,

section06.indd 756

­

a
t
r/

­

e
s

/r u

The American Psychological Association’s (APA) Diagnostic and Statistical
Manual of Mental Disorders (DSM)-IV describes delirium as a disturbance in consciousness and cognition that develops over a short period
of time (eg, hours to days) and tends to fluctuate during the course of the
day.1 Specifically, there are four criteria required to diagnose delirium1:

/
9
9

r
i
h

Associated
Features

­

DEFINITION AND TERMINOLOGY

 

 

Dementia

­

Patients in the intensive care unit (ICU) who experience delirium are
exhibiting an under-recognized form of organ dysfunction. Delirium is
extremely common in ICU patients as factors such as comorbidity, the
acute critical illness itself, and iatrogenesis intersect to create a high-risk
setting for delirium. This neurologic complication is often hazardous,
being associated with death, prolonged hospital stays, and long-term
cognitive impairment and institutionalization. Neurologic dysfunction

compromises patients’ ability to be removed from mechanical ventilation or to fully recover and regain independence. Unfortunately, health
care providers in the ICU are unaware of delirium in many circumstances, especially those in which the patient’s delirium is manifesting
predominantly as the hypoactive (quiet) subtype rather than the hyperactive (agitated) subtype. Despite being often overlooked clinically, ICU
delirium has increasingly been the subject of research during the past
decade, which has brought to light the scope of the problem in critically
ill patients and provided clinicians with tools for routinely monitoring
delirium at the bedside. This chapter reviews the definition and salient
features of delirium, its primary risk factors, including drugs associated
with the development of delirium, proposed pathophysiologic mechanisms, validated methods for bedside delirium assessment, and nonpharmacologic and pharmacologic strategies for delirium management.

Delirium

­

INTRODUCTION

Differentiating Delirium From Dementia

• Visuospatial impairment
• Little/no awareness of memory
impairment
• Gait disturbances (falls)
• Anxiety/mood/sleep
disturbances
­

Nathan E. Brummel
Timothy D. Girard

TABLE 82-1

­

82

Delirium in the Intensive
Care Unit

­

CHAPTER

PART 6: Neurologic Disorders

 

756

• Dementia of Alzheimer type
• Vascular dementia
• Chronic medical conditions
(eg, Pick disease, HIV, stroke,
head injury)

Data from American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed.
Text Revision. Washington, D.C.: American Psychiatric Association; 2000.

subacute befuddlement, and toxic confusional state.2-5 Neurologists often
use “encephalopathy” to refer to hypoactive delirium and “delirium” to
describe only hyperactive delirium.6 Among ICU practitioners, “delirium”

is used inconsistently, as evidenced by a recent survey of Canadian intensivists that found respondents were more likely to use the term “delirium”
when no specific underlying etiology could be identified for a patient with
fluctuating mental status with inattention, perceptual changes, and disorganized thinking, whereas alternative terms (eg, hepatic encephalopathy)
were used when the etiology of delirium was obvious.5,7
Increasingly, however, the ICU community is seeking to standardize
delirium terminology to conform to the APA definition, with the hope that
use of “delirium” to describe this syndrome of acute brain dysfunction,
regardless of etiology, will improve cross-talk between specialists with
different medical backgrounds, collaborative research efforts, and
ultimately management of this widely prevalent syndrome.4 Therefore,
the unifying term “delirium” should be applied whenever patients meet
DSM-IV diagnostic criteria for delirium, and the underlying etiology,
when known, can be used as an associated term (eg, “delirium secondary
to sepsis” is preferred over “septic encephalopathy”).

PREVALENCE AND SUBTYPES
Delirium during critical illness occurs in 20% to 80% of ICU patients
depending on the severity of illness of the population studied and methods used to detect delirium.8-16 The prevalence is highest, for example, in
mechanically ventilated ICU patients, with 60% to 80% developing delirium during their ICU stay,8,10,12,14,17 whereas lower prevalence rates are
reported in nonventilated patients and in mixed ICU populations.9,11,18
In general, ICU patients have a higher prevalence of delirium compared
with noncritically ill hospitalized patients.19,20 The prevalence of ICU
delirium will likely increase as the U.S. population ages.
Delirium can be subtyped based on observed changes in motor activity, resulting in hypoactive, hyperactive, and mixed subtypes.21 Peterson
et al reported these delirium subtypes in a cohort of 613 ventilated and
nonventilated ICU patients in whom delirium was monitored for more

1/23/2015 12:55:25 PM

CHAPTER 82: Delirium in the Intensive Care Unit

RISK FACTORS

.t c

 

 

 

Nearly every ICU patient is exposed to one or more risk factors for
delirium; the average patient in one study, in fact, had 11 identifiable
risk factors for delirium.32 These risk factors may be divided into
predisposing (baseline) factors and precipitating (hospitalization-related)
TABLE 82-2

s
tt p

Risk Factors for Delirium

Not modifiable or preventable

h

­

h

a
t
r/

e
s

Host Factors

Factors Relating to Critical Illness

Environmental and Iatrogenic

Age
Hypertension
APOE-4
Preexisting cognitive impairment
Alcohol use
Tobacco use
Depression

High severity of illness
Respiratory disease
Medical illness
Need for mechanical ventilation
Number of infusing medications

Lack of daylight
Isolation

Hearing or vision impairment

Anemia
Acidosis
Hypotension
Infection/sepsis
Metabolic disturbances (eg, hypocalcemia, hyponatremia,
azotemia, transaminitis, hyperamylasemia,
hyperbilirubinemia)
Fever

Lack of visitors
Sedatives/analgesics (eg, benzodiazepines
and opiates)
Immobility
Bladder catheters
Vascular catheters
Gastric tubes
Sleep deprivation

­

­

Potentially modifiable/preventable

/
9
ri 9

/r u

a
k

/: /

factors.33 Patients who are highly vulnerable to developing delirium
(ie, who have multiple predisposing risk factors) may become delirious
with only minor insults, whereas those with low baseline vulnerability
may require a greater insult to become delirious.33 Predisposing risk
factors, those related to patient characteristics or underlying chronic
pathology, are difficult to alter, whereas precipitating factors, such as
those related to the acute illness or the ICU environment, represent areas
of risk that are modifiable or preventable (Table 82-2).
Baseline risk factors that have been identified in both ICU and non-ICU
populations include older age, depression, vision impairment, hearing
impairment, hypertension, history of smoking, history of alcohol use,
living single at home, underlying cognitive impairment or dementia, and
APOE4 polymorphism.9,10,13,34-37 Numerous features of the acute critical
illness have been identified as delirium risk factors in studies specifically
examining ICU patients; these include admission to an ICU for a medical
illness, high severity of illness (indicated by high APACHE II and SAPS II
scores), need for mechanical ventilation, receipt of sedative and/or analgesic medications (particularly when used to induce coma), respiratory
disease, anemia, hypotension, hypocalcemia, hyponatremia, azotemia,
transaminitis, hyperamylasemia, hyperbilirubinemia, acidosis, fever, infection, sepsis, gastric tubes, bladder catheters, arterial lines, and more than
three infusing medications.9,13,17,35-39 Risk factors related to the ICU environment include lack of daylight in the ICU, isolation, lack of visitors, and
sleep disturbances.37,40
Though difficult to accurately measure in ICU patients, sleep deprivation is believed to be nearly universal in the ICU and has long been proposed as a risk factor for delirium. The relationship, however, between
sleep disturbance and delirium in the ICU remains controversial, and

there is significant overlap in the symptoms of both syndromes such
that either may present with inattention, fluctuating mental status and
cognitive dysfunction, making it difficult to ascertain whether sleep
deprivation causes delirium or vice versa.40,41 On average, ICU patients
sleep between 2 and 8 hours in a 24-hour period, often with severe and
frequent disruptions and only a small fraction of “restorative,” rapid eye
movement (REM) sleep.42 In repeated studies, between one-third and
one-half of patients’ sleep in the ICU occurs during daytime hours.42,43
Reasons for poor sleep in this setting are multifactorial. The ICU environment, with its continuous cycle of alarms, lights, and care-related
interruptions interferes with a patient’s sleep cycle and may disrupt
their circadian rhythm.41,43 Acute illness, with symptoms such as nausea,
pain, and fever, may also disrupt sleep. Mechanically ventilated patients
may additionally suffer sleep disruptions due to anxiety, ventilator
dyssynchrony, central apneas, and mode of mechanical ventilation.44
­

­

than 20,000 observations. Among patients who developed delirium, pure
hyperactive delirium was rare (<5%), whereas hypoactive was present
in 45% and the mixed subtype—with alternating periods of hypoactive
and hyperactive delirium—was the predominant manifestation (54%).
Interestingly, hypoactive delirium was significantly more common in
patients over the age of 65. Similarly, in a cohort of 100 surgical and
trauma ICU patients, the prevalence of hypoactive delirium was greater
than 60%.22 The risk factors for, and clinical implications of, these
subtypes are the subject of ongoing investigations.23
Because sedation is commonly used in the ICU, the period surrounding cessation of sedation represents a scenario in the ICU during
which delirium could be easily recognized but is often missed. Delirious
patients emerging from the effects of sedation may do so peacefully or

in a combative manner. The “peaceful” patients are often erroneously
assumed to be thinking clearly. Delirium in this context is referred to
as hypoactive delirium and is characterized by lethargy, drowsiness, and
infrequent spontaneous movement,21 which contributes to delirium
being overlooked unless the patient is specifically screened for its
presence.24-28 Even in the absence of agitation, such delirium can lead
to adverse outcomes such as reintubation, which itself has been shown to
increase the risk of prolonging the ICU stay, transfer to a long-term care
or rehabilitation facility, and death.29 In addition, hypoactive delirium is
associated with immobility in the ICU,30 which itself places patients at
risk for adverse outcomes, including aspiration, pulmonary embolism,
and decubitus ulcers.
In contrast to patients with hypoactive delirium are agitated or combative patients with hyperactive delirium; these patients are at risk not only
for self-extubation and subsequent reintubation but also for pulling out
central venous catheters and even falling out of bed. These hyperactive
patients are often given large doses of sedatives that lead to heavy sedation and prevent timely liberation from mechanical ventilation, placing
patients at risk for remaining delirious or even comatose and on invasive
mechanical ventilation unnecessarily.31 To avoid this difficult and dangerous cycle, health care professionals should minimize use of psychoactive
medications and frequently assess patients for delirium, especially during
the transition from drug-induced or metabolic coma to wakefulness.

757

APOE-4, apolipoprotein E polymorphism.
Note: Risk factors for delirium can relate to the host, those relating to critical illness and those relating to the intensive care unit environment or treatment of critical illness. Within each of these divisions, there are
risk factors that are preventable or potentially modifiable and those that are not preventable or modifiable.

section06.indd 757

1/23/2015 12:55:25 PM

758

PART 6: Neurologic Disorders

Probability of transitioning
to delirium (%)

A

100
90
80
70
60
50

No drug

a
k

/: /

PATHOPHYSIOLOGY

­

­

The pathophysiology of delirium remains incompletely understood.
Leading hypotheses, often drawn from research outside the ICU, propose that delirium results from neurotransmitter imbalances and/or
factors that affect neurotransmitter production, such as availability of
large neutral amino acids, or systemic and central nervous system (CNS)
inflammation. Delirium during critical illness is most likely a consequence of a complementary and interlinked series of events (Fig. 82-2).
Delirium due to Atropa belladonna (a plant known as Deadly Nightshade,
which contains the anticholinergic atropine) and anticholinergic
drugs, such as scopolamine, has been recognized for centuries, an observation that led to the hypothesis that imbalances in the synthesis, release,

% Days delirious

p = 0.014

60
40

Users
Non-users

p = 0.031

20
0

Surgical

Trauma

Daily midazolam use (exc. coma days)

 

 

FIGURE 82-1. Relationship between benzodiazepines and delirium. Multiple studies
have demonstrated the association between benzodiazepines and delirium. As the daily dose
of lorazepam increased in medical ICU patients, the odds of transitioning to delirium increase,
such that patients treated with >20 mg of lorazepam per day universally developed delirium
(A). Reproduced with permission from Girard TD, Pandharipande PP, Ely EW. Delirium in the
intensive care unit. Crit Care. 2008;(12 suppl 3):S3. Similarly, daily midazolam use is associated
with an increase in the proportion of days with delirium in surgical and trauma ICU patients
(B). Reproduced with permission from Pandharipande P, Cotton BA, Shintani A. Prevalence
and risk factors for development of delirium in surgical and trauma intensive care unit
patients. J Trauma. July 2008;65(1):34-41.

and inactivation of neurotransmitters—especially acetylcholine and
dopamine—that control arousal and the sleep-wake cycle are the underlying mechanism leading to delirium.49,50 Studies measuring the amount
of anticholinergic activity in hospitalized patients found higher levels of
serum anticholinergic activity (SAA) were associated with an increased
risk of delirium, even in patients not exposed to medications with anticholinergic properties.51,52 Central cholinergic deficiency can theoretically
result from derangements occurring anywhere along the continuum from
acetylcholine production and release to its action on postsynaptic receptors. In addition to cholinergic deficiency, dopamine excess is thought to
be associated with delirium, likely via its action on central dopamine receptors that regulate acetylcholine production.50-54 Finally, imbalances in the
production, release, and degradation of numerous other neurotransmitters,
such as serotonin, norepinephrine, glutamate, melatonin, and gammaaminobutyric acid (GABA), have also been suspected to play a role in the
development of delirium.49-54
Large neutral amino acids (LNAAs), including leucine, valine, tryptophan, tyrosine, and phenylalanine, are the precursors of several
neurotransmitters that are involved in arousal, attention, and cognition
and are therefore hypothesized to be involved in the pathogenesis of

delirium.52 The synthesis of serotonin and melatonin depend on the
availability of tryptophan, whereas the production of norepinephrine and
dopamine require both tyrosine and phenylalanine. The LNAAs compete
for transfer across the blood-brain barrier, such that an increase in
­

­

s
tt p

.t c

/
9
9

ir

h
a
t
r/

e
s

4+ Log scale
55+ Original scale

Midazolam

­

80

/r u

h

section06.indd 758

0-1
1-2
2-3
3-4
0-2.7 2.7-7.4 7.4-20 20-55
Lorazepam dose (mg)

B 100

 

­

­

­

Finally, medications commonly given to ICU patients, such as sedatives, analgesics, vasopressors, β-agonists, and corticosteroids, disrupt

slow-wave and REM sleep.45 Further study of sleep in the ICU is
necessary to understand the underlying mechanisms for sleep disruption and the relationship between sleep and delirium. Meanwhile,
clinicians should attend to modifiable risk factors by reducing noise
and light at night, minimizing other disruptions in the ICU environment, treating symptoms, and judiciously using sleep-disrupting
medications.
The deliriogenic effects of medications given for sedation and/or
analgesia—drugs used to treat nearly all ICU patients at some time
during their ICU stay—have received specific attention in many
studies, as they represent a potent yet potentially modifiable risk factor for delirium. Though sedative and analgesic medications are prescribed to relieve pain and anxiety and to improve patient tolerance
of treatments during critical illness, these medications have important
side effects. Continuous infusion of sedatives, for example, is associated with prolonged mechanical ventilation,31 whereas interruption
of sedative infusions expedites weaning from mechanical ventilation,
speeds discharge from the ICU and hospital, and improves long-term
survival.12,46
Multiple studies have now clearly demonstrated a link between
benzodiazepines and development of delirium. Lorazepam dose was
found to be an independent risk factor for the delirium in medical ICU
patients, such that each day a patient was treated with the drug, the odds
of being delirious the next day increased by 20%. In fact, patients treated
with greater than 20 mg of lorazepam in a day were nearly all delirious or
comatose the following day.13 Numerous other studies have consistently
found similar links between benzodiazepine administration (whether
lorazepam or midazolam) and delirium in patients in surgical, trauma,
burn, and mixed ICUs (Fig. 82-1).14,15,17,36,38,39,47
Narcotic pain medications present a more complex picture in terms
of their relationship with delirium in the ICU, in that they have been
associated with development of delirium in some studies but not in
others. This is likely due to the differing indications for (or dual effects
of) analgesics in the ICU. Narcotic pain medications are associated with
the development of delirium in populations frequently sedated with

these drugs, such as medical and surgical ICU patients.9,17,37 In these
settings, narcotics are often co-administered with benzodiazepines; in
one study, elderly ICU patients treated with benzodiazepines and opioids had a longer duration of delirium.39 When narcotic medications
are used to induce coma, the odds of developing delirium triple.36 Thus,
clinicians should seek to minimize the use of heavily sedating medications, whether benzodiazepines or narcotics, by using evidenced based
protocols to interrupt continuous sedative infusions12,46 and seek to use
nonbenzodiazepine sedative medications where possible.14,15,48 Patients
more often treated with narcotics because of pain, such as trauma ICU
patients, are found to have a lower risk of the development of delirium
when treated with fentanyl or morphine compared to patients who
were not exposed to these drugs.17 Intravenous opiates and exposure
to methadone was protective against development of delirium in burn
ICU patients.47

1/23/2015 12:55:26 PM


CHAPTER 82: Delirium in the Intensive Care Unit

Medications
Medical illness
Surgical illness

Medications
Alcohol withdrawal

Medications
Stroke

Cholinergic

activation

Benzodiazepine and
alcohol withdrawal

Cholinergic
inhibition
Reduced
GABA activity

Dopamine
activation
Cytokine
excess

GABA
activation

Serotonin
activation

/
9
ri 9

Benzodiazepines
Hepatic failure

Delirium
Glutamate

activation
Serotonin
deficiency

Medications
substance withdrawal

759

Cortisol
excess

h
a
t
r/

Hepatic failure
Alcohol withdrawal

Tryptophan depletion
Phenyalanine elevation

Glucocorticoids
Cushings syndrome
Surgery
Stroke

Surgical illness
Medical illness

e
s

 

FIGURE 82-2. Delirium pathophysiology represents a complex series of interrelated events. Multiple pathways to delirium may be present in a single patient. (Reproduced with permission
from Flacker JM, Lipsitz LA, et al. Neural mechanisms of delirium: current hypotheses and evolving concepts. J Gerontol A Biol Sci Med Sci. June 1999;54(6):B239-B246.)

/r u

a
k

/: /

s
tt p

­

­

section06.indd 759

­

­

­

h

­

­

.t c

neuronal cell death, resulting in a functional disconnection between
anatomical structures leading to the acute neurobehavioral changes
observed in delirium.59 Indeed, recent data indicate that inflammatory
biomarkers, such as procalcitonin, are associated with increased days of
delirium or coma.63 Elevation of these inflammatory markers was not
consistently associated with other organ failures, suggesting that systemic
inflammation may modulate CNS inflammation and may be an important contributor to brain dysfunction in critically ill patients.
­

one LNAA causes a decrease in the entry of other LNAAs into the brain.52
Thus, changes in serum levels of individual LNAAs may directly effect
CNS neurotransmitter concentrations. With this in mind, Flacker and
collegues55 examined LNAA levels in acutely ill elderly medical patients
and found an association between delirium and an elevated plasma
phenylalanine/LNAA ratio. Tryptophan/LNAA ratios are decreased
and phenylalanine/LNAA ratios increased in cardiac surgery patients
who developed delirium.56 Low plasma levels of tryptophan were also
observed in delirious postoperative patients.57 Finally, Pandharipande
and collaborators described both high and low tryptophan/LNAA ratios
and high and low tyrosine/LNAA ratios as independent risk factors for
delirium (with mid-range ratios being low-risk for delirium) in a cohort

of mechanically ventilated ICU patients.58 These studies suggest that
changes in LNAA concentrations with subsequent alterations in CNS
neurotransmitter levels are important in the pathogenesis of delirium.
Delirium is also hypothesized to result from systemic inflammation,
which occurs frequently in critical illness as a result of infection, tissue
destruction, or surgery. Proinflammatory cytokines, such as interleukin-1
beta, tumor necrosis factor-alpha, and interleukin-6, as  well as prostaglandins and bloodborne molecules, such as lipopolysaccharide, communicate with the brain via either direct autonomic neural pathways, active
transport of cytokines across the blood-brain barrier, second messenger
systems in the blood-brain barrier, or via disruption of the blood-brain
barrier.59-61 Recognition of these peripheral inflammatory stimuli initiates
a cascade resulting in astrocyte, microglial, and endothelial activation,
leading to production of additional inflammatory cytokines, reactive
oxygen species, and expansion of the microglia population, culminating
in neuroinflammation and ultimately neuronal damage.59,61 Advanced age,
underlying dementia, and states of chronic inflammation may “prime”
microglial cells, resulting in an exaggerated inflammatory response.59-61
In addition, systemic inflammation results in endothelial damage leading
to thrombin formation and vasoconstriction with resultant microvascular
compromise.62 The combination of neuroinflammation and disruption
of normal CNS perfusion may then impair neurotransmitter synthesis
and release (particularly acetylcholine),50 impair oxidative metabolism, and
deplete neuronal energy stores.52 These processes then may lead to

MONITORING FOR DELIRIUM
Current Society of Critical Care Medicine (SCCM) guidelines recommend
that all critically ill patients be monitored for delirium as well as changes
in level of consciousness.64 Bedside critical care nurses and the rest of the
ICU team should use data obtained from well-validated, reliable but brief
assessment tools to monitor both level (which can change frequently during critical illness) and content of consciousness, with changes in both
components required before delirium is diagnosed. Such neurologic

monitoring can be streamlined in the ICU by using a two-step approach.
The first step in the neurologic assessment of an ICU patient is
to assess that patient’s level of consciousness using an objective tool.
Though the available tools are typically referred to as sedation scales,
they should be used to assess all critically ill patients—whether receiving
sedation or not—and should be viewed as assessments of level of consciousness rather than solely level of sedation. In addition to helping
practitioners avoid oversedation, objective sedation scales provide a
common language for the multidisciplinary team to use when discussing goals and treatments for patients. For decades, the Ramsay Scale was
the instrument most widely used in clinical practice and the published
literature.65,66 The Riker Sedation-Agitation Scale67 and Richmond
Agitation-Sedation Scale,68 however, have been better validated67,69 and
are also being widely used.16,66,70 Chapter 22 includes a thorough discussion of how to manage sedation in the ICU.
The second step in the neurological assessment of an ICU patient—a
step that can only be completed when a patient is not comatose—is to
evaluate that patient for delirium using an objective tool. Over the last

1/23/2015 12:55:27 PM


760

PART 6: Neurologic Disorders

Checklist Item
Altered level of consciousnessa

a
k

A

No response

B

Response to intense and repeated stimulation

C

Response to mild or moderate stimulation

D

Normal wakefulness

E

Exaggerated response to normal stimulation

s
tt p

Inattentiveness

/: /

Difficulty following instructions or easily
distracted

Disorientation

To time, place or person

Hallucination-delusion-psychosis

h

Clinical manifestation or suggestive behavior

Psychomotor agitation or retardation Agitation required use of drugs or restraints or
slowing
Inappropriate speech or mood

Related to events or situation or incoherent
speech
Sleeping <4 h/d, waking at night, sleeping
all day

Symptom fluctuation

Symptoms of above occurring intermittently

Total score (one point for obvious
presence of features above)

0-8

­

 

Sleep/wake cycle disturbance

a

If level of consciousness A or B no other features are assessed that day.

 

The Intensive care delirium screening checklist. This 8-item checklist should be completed using clinical
information gathered over the last 8 or 24 hours. First assess level of consciousness. If level of consciousness is C, D, or E proceed with the remaining items. Patients are given 1 point for having an obvious
manifestation of the item. A score of 4 or greater is considered a positive delirium screen.
Modified with permission from Bergeron N, Dubois MJ, Dumont M, et al. Intensive Care Delirium
Screening Checklist: evaluation of a new screening tool. Intensive Care Med. May 2001;27(5):859-864.

section06.indd 760

­

r
i
h

PROGNOSIS FOLLOWING ICU DELIRIUM

Numerous studies have now confirmed that ICU delirium is associated
with multiple poor clinical outcomes, which can be divided into immediate, short-term, and long-term categories.
Immediate complications associated with delirium include prolonged
mechanical ventilation, use of physical restraints, self-extubation, and
catheter removal.9,79,80 Indeed, in one recent study of 344 medical and surgical ICU patients, delirium independently predicted time to extubation

in a dose-dependent fashion, with additional days of delirium predicting
more time on the ventilator; the number of days a patient was delirious, in
fact, was the most significant predictor of time on mechanical ventilation.80
Short-term outcomes associated with ICU delirium include prolonged
ICU length of stay, prolonged hospitalizations, institutionalization after
hospital discharge, increased hospital costs, and increased ICU and
hospital mortality.32,36,80-82 After controlling for covariates, caring for
patients with ICU delirium is associated with a 39% increase in ICU
costs and a 31% increase in total hospital costs.83 Elderly postoperative patients who develop delirium in the ICU are 7 times more likely
to be discharged to a place other than home.84 Finally, patients with
ICU delirium have a higher ICU mortality16 and at least double the
in-hospital mortality rate of nondelirious patients.16,36,77,81,85,86 The risk
of death following delirium does not end at hospital discharge. Indeed,
delirious patients who survive hospitalization remain at a higher
risk for death in the months after discharge.77,81,85,86 In one study of
275  mechanically ventilated medical ICU patients, those who developed delirium in the ICU were three times more likely to die in the
6 months following hospitalization than those patients who were never
delirious.81 The association between delirium and long-term mortality
also increases the longer a patient is delirious, such that after adjusting
for potential confounders, each additional day of delirium predicts a
10% increase in the hazard of dying in the 6 to 12 months following
hospitalization for critical illness (Fig. 82-4).80,81,86
Although often not observed by ICU clinicians caring for delirious
patients, other long-term outcomes associated with ICU delirium are
often as deleterious as the short-term outcomes. Delirious patients are at
high risk for long-term cognitive impairment, and the longer delirium
persists in the ICU, the more severe these impairments are likely to be.87-89
In a prospective study of ICU survivors who underwent neuropsychological testing, nearly 7 in 10 patients demonstrated signs of cognitive
impairment 1-year following critical illness. After adjusting for covariates, the duration of delirium in the ICU was independently associated
with cognitive impairment.89 These long-term cognitive impairments in

ICU survivors manifest in numerous ways, including memory problems
and executive dysfunction, which can cause difficulty with managing
money, reading a map, and following detailed instructions, among
other effects.87,89,90 These impairments have profound effects on patient’s
lives. Rothenhausler et al, for example, followed survivors of the acute

a
t
r/

e
s

/r u

.t c

Description

/
9
9
­

The Intensive Care Unit Delirium Screening Checklist

Intensive Care Unit Delirium Screening Checklist (ICDSC)

validation studies found the CAM-ICU to have excellent sensitivity
(89%-100%) and specificity (93%-100%) with high inter-rater reliability

(κ = 0.79-0.96), and subsequent studies have found the sensitivity
to range from 47% to 100% and the specificity to range from 88% to
96%.8,18,28,72,75-78 As with the ICDSC, patients who are comatose cannot be
assessed using the CAM-ICU but should be evaluated again frequently,
since patients emerging from coma are high risk for delirium. Patients
who are moderately sedated (ie, have some response to verbal stimuli) or
more alert may be assessed for delirium using the CAM-ICM. The CAMICU assesses for four features of delirium. According to the recently
revised format, which was streamlined to improve efficiency, feature 1 is
the acute onset of mental status changes or a fluctuation in mental status
over the last 24 hours, feature 2 is inattention, feature 3 is altered level of
consciousness, and feature 4 is disorganized thinking. A patient is considered delirious if features 1 and 2 and either feature 3 or feature 4 are
present (Fig. 82-3).8,72 The CAM-ICU tool as well as an in-depth training
manual are available for download at www.icudelirium.org.

­

 

 

TABLE 82-3

 

decade, the development of tools designed especially with the unique
characteristics of critically ill ICU patients in mind has allowed the
clinician to rapidly71 and reliably detect delirium at the bedside.8,11,72
Two assessment tools, the Intensive Care Delirium Screening Checklist
(ICDSC) and the Confusion Assessment Method for the ICU (CAMICU), have been validated extensively against expert psychiatric raters
using DSM-IV criteria for delirium; these tools were widely tested in the

ICU setting on both mechanically ventilated and nonmechanically ventilated patients.8,11,72 Several other tools have been developed and assessed
in validation studies with varying results; these studies suggest the
Nursing Delirium Screening Scale (Nu-DESC) is a promising tool, though
more validation data are needed before it can be widely recommended.71
The ICDSC is an eight-item screening tool (Table 82-3) that is completed using clinical information collected during either the previous
eight or 24 hours (depending on how often the tool is used).11 For each of
the eight items on the checklist, patients are given one point for obvious
manifestations of the item or zero points if there is no manifestation or the
item is not assessable. Before the checklist is completed, level of consciousness is assessed, and the checklist is only completed if the patient is not
comatose or stuporous (ie, their level of consciousness is rated other than
A or B on the ICDSC scale). A score of 4 or more on the ICDSC identifies
delirium with 64% sensitivity and 99% specificity according to the original validation study.11 More recently, studies have found the sensitivity to
range from 43% to 74% and the specificity to range from 75% to 95%.28,73
The CAM-ICU is a four-feature delirium-screening tool adapted from
the Confusion Assessment Method for use in nonverbal, mechanically
ventilated ICU patients.8,72 It has been translated into over 14 languages
and has been implemented across the world in medical, cardiovascular,
surgical, trauma, and burn intensive care units.8,16,18,28,74-76 The original

1/23/2015 12:55:27 PM


CHAPTER 82: Delirium in the Intensive Care Unit

2

Step

761

Confusion assessment method for the ICU (CAM-ICU)

Delirium assessment

1. Acute change or fluctuating course of mental status:
Is there an acute change from mental status baseline? Or
Has the patient’s mental status fluctuated during the past 24 hours?

No

CAM-ICU negative
No delirium

0-2
errors

CAM-ICU negative
No delirium

Yes

/
9
ri 9

2. Inattention:
“Squeeze my hand when i say the letter ‘A’. ”
Read the following sequence of letters: SAVE A HAART
Errors: No squeeze with ‘A’ & squeeze on letter other than ‘A’
If unable to complete letters → pictures

>2 Errors

h
a
t
r/

RASS other
than zero

3. Altered level of consciousness
Current RASS level (think back to sedation assessment in step 1)
RASS = zero
4. Disorganized thinking:

CAM-ICU positive
Delirium present

>1 error

1. Will a stone float on water?
2. Are there fish in the sea?
3. Does one pound weigh more than two?
4. Can you use a hammer to pound a nail?

e
s

Command: “Hold up this many fingers” (Hold up 2 fingers)
“Now do the same thing with the other hand” (Do not demonstrate)

Or “Add one more finger” (If patient unable to move both arms)

/r u

0-1
error

CAM-ICU negative
No delirium

FIGURE 82-3. The CAM-ICU assesses for the four features of delirium. Feature 1 is an acute change in mental status or a fluctuating mental status (first box), feature 2, is inattention,
(second box), feature 3, is altered level of consciousness (third box) and feature 4, is disorganized thinking (fourth box). A patient screens positive for delirium if features 1 and 2 and either
feature 3 or feature 4 are present. (Used with permission of E. Wesley Ely, MD and Vanderbilt University. Copyright © 2002.)

a
k

/: /

s
tt p

1.0
0.9

Survival probability

0.8
0.7

h

0.6
0.5
0.4
0.3

1-2 Days
3-4 Days
5-9 Days

10 + Days

0.1
0.0
0

75

150

225

300

375

450

Time to death (Days)

 

FIGURE 82-4. Survival probability and duration of delirium. The hazard ratio for death
at 1 year is 1.10 (95% CI 1.02-1.18, p <.01), indicating a 10% increase in the risk of mortality
at 1-year for each day a patient is delirious. (Reproduced with permission from Kong SY, Kasl
SV, et al. Days of delirium are associated with 1-year mortality in an older intensive care unit
population. Am J Respir Crit Care Med. December 1, 2009;180(11):1092-1097.)

section06.indd 761

Perhaps the most effective strategy to reduce the adverse outcomes
associated with delirium is to prevent delirium in the first place. In
general, preventive strategies should focus on reducing risk factors for
delirium. To date, successful prevention strategies have utilized multicomponent programs of non-pharmacologic interventions designed to
ameliorate delirium risk factors in non-ICU populations at high risk for
delirium.92 Modification of specific delirium risk factors, such as sleep
deprivation, immobility, visual and hearing impairment, and dehydration, was associated in one landmark trial with a 40% relative reduction in the development of delirium in hospitalized (non-ICU) elderly
patients.93 These interventions, however, were less effective if delirium
was already present, indicating an important role for primary prevention. A second trial explored the utility of early geriatrics consultation
in elderly hip fracture patients undergoing fracture repair. The geriatricians followed a specific protocol and made targeted interventions
aimed at specific risk factors, such as reducing potentially deliriogenic
medications, ensuring adequate oxygenation and blood pressure control,
providing adequate pain control as well as ensuring the presence of eye
glasses and hearing aides. Compared with the usual care group, who
could have received a reactive geriatrics consultation, this proactive
strategy was associated with an 18% absolute reduction in incident delirium during the hospitalization between groups (from 50% to 32%).94
Overall rates of delirium in these non-ICU patient cohorts are much
lower than those observed in critically ill patients, and ICU patients
are exposed to many more risk factors than non-ICU patients, suggesting that delirium in the ICU is likely more complex than that outside
the ICU. Thus, the effectiveness of these nonpharmacologic strategies for preventing delirium observed in non-ICU studies may not be

­

0.2

0 Days

STRATEGIES FOR PREVENTION OF DELIRIUM

­

 

 

.t c

respiratory distress syndrome (ARDS) for a median of six years after
ICU discharge and found that 100% of patients with cognitive impairment were unemployed compared with only 23% of those patients who
were not cognitively impaired.91

1/23/2015 12:55:28 PM


PART 6: Neurologic Disorders

When delirium is diagnosed, potential underlying causes should be
sought immediately, and treatment of suspected causes should be undertaken. Then, if the patient remains delirious and to prevent the harmful
sequelae of persistent delirium, current guidelines recommend treatment with pharmacologic agents.64 To date, there have been only small,
preliminary trials examining pharmacologic treatments for delirium in
the ICU.96-98 Without large, well-designed, adequately powered, placebocontrolled, randomized trials to guide drug use for the prevention or

treatment of delirium in critically ill patients, evidence must be extrapolated from studies of non-ICU populations.
Benzodiazepines are used commonly in the ICU for both sedation
and the treatment of delirium,66 but this class of drugs is not recommended for the management of delirium because of the likelihood of
oversedation, exacerbation of delirium, and other adverse effects (eg,
respiratory suppression). As mentioned in the section on risk factors,
benzodiazepines actually increase the likelihood of developing delirium
for most patients.13,17,39 Benzodiazepines, however, remain the drugs of
choice for the treatment of delirium tremens (and other withdrawal
syndromes) and seizures.
Though a number of medications are frequently used to treat delirium
in the ICU,66 there are currently no drugs approved by the U.S. Food
and Drug Administration for this indication. Expert guidelines from
the Society of Critical Care Medicine,64 the American Psychiatric
Association,99 and other authoritative bodies recommend haloperidol as
the drug of choice for the treatment of delirium, but it is acknowledged
that these recommendations are based on sparse data from nonrandomized case series and anecdotal reports.
Haloperidol, a butyrophenone, “typical” antipsychotic, is the most widely
used neuroleptic agent for delirium.66,100 It works primarily as a dopamine
receptor antagonist by blocking the D2 receptor, which is believed to
treat—borrowing terminology from the schizophrenia literature—positive

s
tt p

h

section06.indd 762

/: /

a
k

 

 

 

/
9
9

a
t
r/

r
i
h

e
s

/r u

.t c

DELIRIUM TREATMENT

symptoms of delirium (eg, hallucinations, unstructured thought patterns,
etc). Haloperidol can also have a sedative effect, though this is variable and,
unlike most sedative agents, does not result in respiratory suppression. In
the non-ICU setting, the recommended starting dose of haloperidol
is 0.5 to 1.0 mg orally or parenterally, with doses repeated every 20 to
30  minutes until the desired effect, which is usually resolution of agitation rather than complete resolution of delirium. In the ICU, alternatively,
higher doses are often recommended, eg, 2 to 5 mg intravenously with
doses repeated every 20 to 30 minutes until the desired effect. Some practitioners use scheduled haloperidol every 6 to 12 hours (intravenously or
orally). No strong data exist indicating the ideal dose, but maximal effective doses are believed to be approximately 20 mg/d based upon data that
this dose is usually adequate to achieve the “theoretically optimal” 60% to
80% D2 receptor blockade while avoiding the complete D2 receptor saturation associated with the adverse effects described below.101,102 Because
extreme agitation in the ICU is an urgent problem, due to the potential for
inadvertent removal of catheters, endotracheal tubes, and other devices,
much larger doses of haloperidol are sometimes used, but this approach
is based upon anecdotal experience and expert opinion and should be
considered unproven until more data are available.
Neither haloperidol nor similar agents (eg, droperidol and chlorpromazine) have been extensively studied in the ICU. In fact, the only
placebo-controlled trial examining the effect of haloperidol on ICU
delirium found no significant improvement with this agent.96 This pilot
study was small, however, and cannot be taken to rule out a beneficial
effect of haloperidol in delirium.
Some observational studies of the use of antipsychotics in non-ICU
patients with delirium have reported improvements in delirium in
patients treated with antipsychotics. Nevertheless, these conclusions are
not supported by randomized controlled trials, therefore it is unknown
if this association is due to the natural history of the disease, treatment
of underlying medical conditions or antipsychotics themselves.103,104
In addition to using antipsychotics to treat delirium once present, one
study explored the use of antipsychotic prophylaxis in elderly hip fracture patients at risk of developing postoperative delirium.105 Low-dose
haloperidol did not reduce the incidence of delirium compared with

placebo, but the duration of delirium was shorter in the haloperidol
group. These data suggest a potential role for antipsychotics in the treatment of delirium, but further studies are needed.
In addition to haloperidol, “atypical” antipsychotic agents (eg, risperidone, ziprasidone, quetiapine, and olanzapine) are also used to treat
delirium in the ICU.96-98 The rationale behind the use of atypical antipsychotics over haloperidol (especially in hypoactive/mixed subtypes of
delirium) is theoretical and arises from the atypical antipsychotics’ effect
not only on dopamine but also on other potentially key neurotransmitters, such as serotonin, acetylcholine, and norepinephrine.106 Results
of prospective studies comparing atypical antipsychotics with placebo
and/or typical antipsychotics in the treatment of delirium have been
mixed.96-98 Though one very small randomized trial found quetiapine
was effective in treating delirium compared with placebo,97 another
small randomized trial found no differences in neurologic outcomes
among patients treated with ziprasidone, haloperidol, or placebo.96 In
aggregate, these trials do not provide strong evidence for use of atypical
antipsychotics over typical antipsychotics.
Adverse effects of both typical and atypical antipsychotics include
hypotension, acute dystonia, extrapyramidal effects, thrombotic complications, oversedation, laryngeal spasm, neuroleptic malignant syndrome,
glucose and lipid dysregulation, and anticholinergic effects, such as dry
mouth, constipation, and urinary retention. One of the most immediately life-threatening adverse effects of antipsychotics is torsades de
pointes,107-109 so these agents should be given to patients with prolonged
QTc intervals only with extreme caution. Outpatients treated with either
typical or atypical antipsychotics for schizophrenia are at an increased
risk of sudden cardiac death,107,109 with this risk increasing as either dose
or duration of antipsychotic therapy increases.107,109 It remains unclear
whether similar risk affects critically ill patients, who typically receive
­

 

 

generalizable to the ICU setting; further investigation of strategies for the
prevention and management of ICU delirium is needed. Nevertheless, in
the ICU, where risk factors for delirium are nearly ubiquitous, management and minimization of known risk factors should take precedent.
Risk-factor management strategies may be easily implemented in the
ICU by frequently reorienting patients, removing restraints and catheters
as quickly as possible, minimizing sleep interruptions and noise during
nighttime hours, implementing early mobilization protocols, and ensuring vision and hearing assist devices (eg, eyeglasses and hearing aides)
are present.44,93,95
Whereas most prevention strategies have focused on the use of
nonpharmacologic methods in non-ICU populations, a notable exception has been studies of the α2-agonist dexmedetomidine as a sedative
for mechanically ventilated ICU patients. Two randomized trials have
compared the use of this novel sedative with benzodiazepine sedatives,
finding lower rates of delirium among patients sedated with dexmedetomidine. The MENDS (maximizing the efficacy of targeted sedation
and reducing neurologic dysfunction) trial randomized 106 patients to
sedation with either dexmedetomidine or lorazepam. Patients who were
sedated with dexmedetomidine had a median of 4 more days alive without delirium or coma than those sedated with lorazepam (7 vs 3 days).14
A second trial, the SEDCOM (safety and efficacy of dexmedetomidine
compared with midazolam) study randomized 375 patients in a 2 : 1
fashion to sedation with dexmedetomidine or midazolam.15 Patients
receiving dexmedetomidine demonstrated a 23% absolute reduction
in delirium prevalence compared with the midazolam group (delirium
prevalence 54% in dexmedetomidine group vs 76.6% in midazolam
group). Taken together, these studies provide evidence that choice of
sedation agent may be associated with a reduction in ICU delirium.
Nevertheless, it remains unclear whether the reduction in ICU delirium
is due to treatment with dexmedetomidine or simply due to the avoidance of benzodiazepines.

­

762

1/23/2015 12:55:28 PM


CHAPTER 83: ICU-Acquired Weakness

•

­

•

•

• Girard TD, Jackson JC, Pandharipande PP, et al. Delirium as a predictor of long-term cognitive impairment in survivors of critical
illness. Crit Care Med. July 2010;38(7):1513-1520.
• Hatta K, Kishi Y, Wada K, Takeuchi T, Odawara T, Usui C,
et  al. Preventive effects of ramelteon on delirium: a randomized
placebo-controlled trial. JAMA Psychiatry. 2014;71:397-403.
• Pandharipande P, Shintani A, Peterson J, et al. Lorazepam is an
independent risk factor for transitioning to delirium in intensive
care unit patients. Anesthesiology. Jan, 2006;104(1):21-26.
• Pandharipande PP, Pun BT, Herr DL, et al. Effect of sedation with
dexmedetomidine vs lorazepam on acute brain dysfunction in
mechanically ventilated patients: the MENDS randomized controlled trial. JAMA. December 12, 2007;298(22):2644-2653.
• Patel SB, Poston JT, Pohlman A, Hall JB, Kress JP. Rapidly reversible, sedation-related delirium versus persistent delirium in the
intensive care unit. Am J Respir Crit Care Med 2014; 189:658-65.
• Pisani MA, Kong SY, Kasl SV, Murphy TE, Araujo KL, Van Ness
PH. Days of delirium are associated with 1-year mortality in an
older intensive care unit population. Am J Respir Crit Care Med.

December 1, 2009;180(11):1092-1097.
• Reade MC, O’Sullivan K, Bates S, Goldsmith D, Ainslie WR, Bellomo
R. Dexmedetomidine vs. haloperidol in delirious, agitated, intubated
patients: a randomised open-label trial. Crit Care. 2009;13(3):R75.
• Riker RR, Shehabi Y, Bokesch PM, et al. Dexmedetomidine vs
midazolam for sedation of critically ill patients: a randomized trial.
JAMA. February 4, 2009;301(5):489-499.
• Skrobik YK, Bergeron N, Dumont M, Gottfried SB. Olanzapine vs
haloperidol: treating delirium in a critical care setting. Intensive
Care Med. March 2004;30(3):444-449.
• van Eijk MM, Roes KC, Honing ML, et al. Effect of rivastigmine as
an adjunct to usual care with haloperidol on duration of delirium
and mortality in critically ill patients: a multicentre, doubleblind, placebo-controlled randomised trial. Lancet. November 27,
2010;376(9755):1829-1837.
• van Eijk MM, van den Boogaard M, van Marum RJ, et al. Routine
use of the confusion assessment method for the intensive care
unit: a multicenter study. Am J Respir Crit Care Med. August 1,
2011;184(3):340-344.
•
•
•

•

REFERENCES
Complete references available online at www.mhprofessional.com/hall

­

­

•

•

­

.t c

CHAPTER

83

ICU-Acquired Weakness
William Schweickert
John P. Kress

­

h

s
tt p

e
s

•

Critically ill patients are at great risk for the development of delirium

in the ICU. However, this form of brain dysfunction is grossly underrecognized and undertreated. Delirium is mistakenly thought to be a
transient and expected outcome in the ICU and of little consequence
(ie, part of the “ICU psychosis”). It is now recognized that delirium is
one of the most frequent complications experienced in the ICU; even
after adjusting for covariates such as age, sex, race, and severity of illness,
delirium is an independent risk factor for prolonged length of stay
and higher 6-month mortality rates. In addition, many ICU survivors
demonstrate persistent cognitive deficits at follow-up testing months to
years later. It is essential for health care professionals to be able to recognize delirium readily at the bedside. The CAM-ICU is a valid, reliable,
quick, and easy-to-use serial assessment tool for monitoring delirium in
ventilated and nonventilated ICU patients. Delirium is a multifactorial
problem for ICU patients that demands an interdisciplinary approach
for assessment, management, and treatment. Critical care nurses and
physicians should assume a position of leadership in the ICU with
regard to delirium monitoring because they are the best-suited members
of the ICU team to successfully implement this essential component of
patient management, which is recommended by the SCCM clinical
practice guidelines. Although ongoing trials may elucidate the optimal
ways to treat delirium, standard pharmacologic and nonpharmacologic
management strategies have been reviewed.

/: /

h
a
t
r/

/r u

SUMMARY OF KEY POINTS ON ICU DELIRIUM

a
k

/
9
ri 9

•

 

these medications for much shorter periods of time. Nevertheless, ICU
patients treated with antipsychotics should be monitored closely with
electrocardiography, and the medications should be avoided for patients
with a baseline QTc >450 to 500 ms or a prolongation of 25% or greater
from baseline.
The role of novel agents, such as dexmedetomidine and rivastigmine,
in delirium treatment has recently been investigated. As described in the
delirium prevention section above, use of the α2-agonist dexmedetomidine as a sedative for mechanically ventilated ICU patients is associated
with lower rates of ICU delirium when compared with benzodiazepines.
Dexmedetomidine has also been compared with haloperidol as a
treatment for agitated delirium in a small pilot study of mechanically
ventilated patients.110 Patients treated with dexmedetomidine were
more quickly extubated than those patients whose agitation was treated
with haloperidol. Though delirium prevalence at baseline was similar
between the two groups, patients treated with dexmedetomidine may
have had more rapid resolution of delirium though these results were
not significantly different between groups. Although further study is

required, this pilot study suggests dexmedetomidine may have a role not
only in preventing delirium among mechanically ventilated patients but
also treating delirium in this population.
van Eijk explored the use of a cholinesterase inhibitor, rivastigmine,
as an adjuvant treatment for ICU delirium in a population of ICU
patients.111 The trial was stopped prematurely after differences in the
mortality rate between the rivastigmine group (22%) and placebo (8%)
met the predefined stopping criteria. Further, the rivastigmine group
also demonstrated a trend toward longer duration of delirium compared
with placebo. These results do not support the use of cholinesterase
inhibitors for the treatment of delirium in the ICU.

763

KEY POINTS

•

•

• Bergeron N, Dubois MJ, Dumont M, Dial S, Skrobik Y. Intensive
Care Delirium Screening Checklist: evaluation of a new screening
tool. Intensive Care Med. May 2001;27(5):859-864.
• Ely EW, Shintani A, Truman B, et al. Delirium as a predictor of
mortality in mechanically ventilated patients in the intensive care
unit. JAMA. April 14, 2004;291(14):1753-1762.

section06.indd 763

•

• ICU-acquired weakness designates clinically detected weakness in
critically ill patients in whom there is no plausible etiology other
than critical illness. Patients can be labeled with this diagnosis with
a suggestive history and when they can participate in a comprehensive bedside neuromuscular examination.
• Electrophysiology testing, direct muscle stimulation, and biopsy may
be necessary to characterize neuromuscular injury in the patient who
•

KEY REFERENCES

1/23/2015 12:55:29 PM


764

PART 6: Neurologic Disorders

•

­

•

­

•

is unable to participate in a comprehensive neuromuscular examination, is failing to improve function despite weeks of therapy, or for the
patient with asymmetric weakness.

• When conducted, advanced testing, particularly electrophysiology tests,
can characterize the specific phenotype of ICU-AW including critical
illness polyneuropathy, critical illness myopathy, a combination of the
two (polyneuromyopathy), or prolonged neuromuscular blockade.
• The exact epidemiology of ICUAW is unknown. Studies show that
46% of patients with sepsis, multiorgan failure, or prolonged mechanical ventilation are diagnosed with ICUAW. In patients undergoing
mechanical ventilation for 7 days or more, 25% develop ICUAW.
• Factors associated with the diagnosis of ICUAW include the presence
of multisystem organ dysfunction, sepsis, SIRS, and hyperglycemia
and the duration of mechanical ventilation. The only known therapy
to prevent ICUAW has been strict glycemic control with insulin; however, adverse events with this therapy have prevented its utilization.

INTRODUCTION
Many patients admitted to the intensive care unit (ICU) develop a syndrome of neuromuscular dysfunction characterized by generalized muscle
weakness and an inability to be liberated from mechanical ventilation.
Since this syndrome occurs in the absence of preexisting neuromuscular
disease, it is believed to reflect illnesses or treatments occurring in the ICU.
Early reports described two categories of acute, acquired neuromuscular
dysfunction: polyneuropathy (during sepsis and multisystem organ failure)1,2 and myopathy (particularly in patients with acute respiratory failure
who received glucocorticoids and/or neuromuscular blocking agents).3,4
Decades of research on this acquired nerve and muscle injury has characterized specific phenotypes via comprehensive physical examination,
electrophysiologic testing, and histopathology. Overall, the spectrum of
neuromuscular disorders acquired in the ICU is now collectively referred
to as “ICU-acquired weakness” (ICUAW) (Fig. 83-1).5
The rising incidence and societal burden of critical illness—such as
sepsis and the acute respiratory distress syndrome6-8—coupled with
declining case fatality rates and an aging population,9,10 suggests that
the number of patients with ICUAW and its sequelae may be substantial
and likely to grow. Accordingly, intensivists must have familiarity with
the presentation of ICUAW, recognize when to conduct advanced testing, and understand the diagnostic tests involved. Although currently

limited in scope, measures designed to prevent or attenuate ICUAW
must be considered and implemented.
­

s
tt p

/: /

CRITICAL CARE SURVIVORSHIP AND ICUAW

h

Critical care outcomes research has demonstrated substantial morbidity
in survivors. Injuries include general deconditioning, muscle weakness,
dyspnea, depression, anxiety, and reduced health-related quality of life.11

CIM

ICU-acquired weakness

CIPNM

CIP

Prolonged
NMJ
blockade

 

FIGURE 83-1. Classification of intensive care unit-acquired weakness. CIM, critical illness
myopathy; CINM, critical illness polyneuromyopathy; CIP, critical illness polyneuropathy; NMJ,
neuromuscular junction.

section06.indd 764

a
t
r/

/
9
9

r
i
h

e
s

/r u

.t c

a
k

One widely cited catalyst for attention to the burden of neuromuscular

weakness was the comprehensive observations of a cohort of survivors
of acute respiratory distress syndrome (ARDS) published in 2003.12
These 109 survivors were young (median age, 45 years), had few preexisting comorbidities, and were severely ill (median APACHE II
score, 23). Their critical illness was marked by prolonged mechanical
ventilation (median duration, 21 days) and ICU and hospital lengths
of stay (median duration, 25 and 48 days, respectively). Despite severe
acute lung injury, serial follow-up examination during the first year
after ICU discharge demonstrated restoration of lung function. Lung
volumes and spirometry normalized by 6 months and carbon monoxide
diffusion capacity improved to 72% predicted at 12 months. In contrast,
all 109 patients reported poor function attributed to the loss of muscle
bulk, proximal weakness, and fatigue. One year after ICU discharge, the
median distance walked in 6 minutes was 66% of predicted and only
49% of patients had returned to work.
More recently, the same cohort was characterized at 5 years after ICU
discharge.13 All patients reported subjective weakness and decreased
exercise capacity when compared to function before ICU admission.
Although there was no evidence of clinical weakness on examination,
the median distance walked in 6 minutes remained lower than expected
based on age and sex (76% predicted). By the fifth year, 77% of patients
had returned to work; however, patients often required a modified
work schedule, gradual transition back to work, or job retraining. In
addition, patients were plagued with the psychological ramifications of
their severe illness; more than half of survivors experienced at least one
episode of physician-confirmed depression or anxiety.
Others have reported similar findings of post-ARDS debilitation.
Specifically, an observational trial of 112 ARDS survivors without
baseline impaired physical function noted a 66% cumulative incidence
of physical impairment during 2 year follow-up.14 This impairment,
defined as the acquisition of two or more dependencies in instrumental

activities of daily living, had greatest incidence by 3 months after
discharge and was associated with longer ICU stay and prior depressive
symptoms. More recently, a comprehensive 1 year follow-up of patients
enrolled in a randomized controlled trial of nutritional strategies
in patient with ARDS demonstrated that survivors, regardless of
nutritional strategy, experienced substantial impairments in endurance
(as defined by six minute walk test) and cognitive function.15
Acquired neuromuscular weakness and loss of function have been
measured in other contexts of critical illness, including severe sepsis
and mechanical ventilation in the elderly. To determine the impact of a
hospitalization for severe sepsis, Iwashyna and colleagues utilized The
Health and Retirement Study, a cohort of Americans over age 50 undergoing biennial surveys of physical and cognitive function.16 Participants
were stratified into those surviving a hospitalization for severe sepsis
(n = 516) versus controls (survivors of a nonsepsis hospitalization,
n = 4517). Among patients with no functional limitations at baseline,
severe sepsis was associated with the development of 1.57 new limitations (95% CI: 0.99-2.15), as well as a more rapid rate of development
of functional limitations after hospitalization (0.51 new limitations per
year, p = 0.007 compared with baseline). The study also found that the
incidence of severe sepsis was highly associated with progression to
moderate to severe cognitive impairment.
In a similar design, Barnato et al used a longitudinal cohort study
of Medicare recipients to investigate the association of mechanical
ventilation and disability.17 Community dwelling patients over age 65
completed quarterly interviews of physical function for four years.
Survivors of hospitalization with or without mechanical ventilation had
similar levels of disability from each other, but significantly more than
those who were never hospitalized. There was a substantial increase in
disability in both groups after hospitalization, greater among survivors
of mechanical ventilation than in those hospitalized without mechanical
ventilation. In adjusted analyses, mechanical ventilation was associated

with a 30% greater disability in activities of daily living (ADLs) and a
14% greater disability in mobility.

1/23/2015 12:55:29 PM


CHAPTER 83: ICU-Acquired Weakness

These studies show that decrements in physical function occur across
the spectrum of critical illness. Although these outcomes may be influenced by other factors—such as age, preexisting comorbidities, acquired
psychological and cognitive dysfunction, and social support—it is clear
that ICUAW needs to be recognized early to enable preventive interventions. However, the recognition of ICUAW has often been hindered
by challenges with various diagnostic testing approaches and complex
nomenclature.

CLINICAL PRESENTATION OF ICUAW
The clinical approach is based on the recognition of generalized weakness in the appropriate setting, the exclusion of causes extrinsic to
critical illness, and the measurement of muscle strength.5 The historian
should carefully review the time course of neuromuscular symptoms
as they relate to the underlying critical illness. Potential risk factors for
ICUAW should be identified—including sepsis, multiple organ failure,
mechanical ventilation, hyperglycemia, and exposure to pharmacologic agents like glucocorticoids and neuromuscular blocking agents
(NMBAs). Neurologic examination evaluates key functional domains
including consciousness, cognitive function, cranial nerves, motor and
sensory systems, deep tendon reflexes, and coordination. Motor assessment should include tone and bulk in addition to strength.
Physical examination of patients for ICUAW is dependent on the
cooperation and maximal effort of the patient—an aspect of bedside
assessment that can be confounded by sedation and delirium. When
a reliable motor examination is possible, affected patients will exhibit
diffuse, generally symmetrical motor deficits in all limbs, ranging from

paresis to true quadriplegia.18 Weakness affects the extremities and
diaphragm yet often spares the cranial nerves; accordingly, pupillary
and  oculomotor function and facial grimace are usually preserved.
Patients often have concurrent respiratory failure with protracted
dependence on mechanical ventilation.
An early clue for isolated myopathy (without neuropathy) is that painful stimulation—such as pressure upon the nail bed—results in a limited
to absent limb response, yet normal grimacing. For patients with ability to
undergo a reliable sensory examination, deficits to light touch and pin
prick may implicate the presence of polyneuropathy. Reflexes are usually
diminished to absent, but normal reflexes do not rule out the diagnosis.
The most commonly reported test of muscle strength in critically
ill patients is manual muscle testing. A standardized bedside muscle
exam can be utilized to evaluate individual muscle groups. The Medical
Research Council (MRC) Score grades the strength of functional muscle
groups in each extremity on a scale from zero to five (Table 83-1).19
Individual MRC scores obtained from predefined muscle groups can be
combined into a sum score, yielding a global estimate of motor function.
The usual standard is to combine three muscle group scores for each
  TABLE 83-1    Medical Research Council (MRC) Neuromuscular Examination
Functions Assessed:
Upper extremity: Wrist flexion, forearm flexion, shoulder abduction
Lower extremity: Ankle dorsiflexion, knee extension, hip flexion
Score for Each Movement:
0—No visible contraction
1—Visible muscle contraction, but no limb movement
2—Active movement, but not against gravity
3—Active movement against gravity
4—Active movement against gravity and resistance
5—Active movement against full resistance
Maximum score: 60 (4 limbs, maximum 15 points per limb)—Normal

Minimum score: 0—Quadriplegia

section06.indd 765

765

  TABLE 83-2    Diagnostic Features of ICUAW
1. Weakness is diffuse, symmetric, and often spares the cranial nerves
2. Causes of weakness other than those from the underlying illness have been excluded
3. Alert patient who can follow simple commands and participate in neuromuscular e­ xamination
4. MRC sumscore <48 or mean MRC <4 in all testable muscle groups
MRC, Medical research Council.
Modified with permission from Stevens et al. A framework for diagnosing and classifying intensive care
unit-acquired weakness. Crit Care Med. October 2009;37(suppl 10):S299-S308.

limb; therefore, sum scores span from zero (complete paralysis) to 60
(full strength). This scoring has demonstrated excellent inter-rater reliability and can be utilized to document the extent of disease and track
serial changes over time.20,21
To better characterize the incidence of acquired neuromuscular
disorders in the ICU and to validate the bedside muscle strength examination, De Jonghe and colleagues prospectively evaluated 95 patients
without preexisting neuromuscular injury that had undergone mechanical ventilation for greater than 7 days.22 The first day a patient was awake
and following commands was considered day 1. On the seventh day
after awakening, patients underwent MRC muscle strength testing to
determine a sum score. A priori, they labeled patients with a sum score
of less than 48 to have “ICU-acquired paresis.” To confirm the peripheral
neuromuscular origin of the clinical weakness, all patients underwent
an electrophysiologic examination exam at day 7 and persistently weak
patients underwent muscle biopsy at day 14. All patients with ICUacquired paresis demonstrated sensory-motor axonopathy. Histological
features of primary myopathic changes were observed in all patients
with paresis persisting 1 week after the initial diagnosis.

Since this landmark trial, leaders in the field have established use of the
term “ICU-acquired weakness” to characterize clinically detected weakness in critically ill patients in whom there is no plausible etiology other
than critical illness.5 ICUAW, synonymous with ICU-acquired paresis, is
defined by the MRC muscle strength sum score <48 in a patient that
is awake and able to follow commands (Table 83-2). Since this delineation, this diagnosis has been repeatedly applied as a ­secondary end point
in prospective clinical trials to crudely assess for the presence of muscle
injury and weakness.23,24
The intensivist should remain cognizant of the limitations of the MRC
strength examination. It requires a patient who is awake, cooperative,
and capable of contracting muscle with maximal force. Scores also can
be affected by patient positioning, the number of limbs available for
assessment (pain, dressings, amputation), and, most importantly, timing.
Experts have lamented other limits including the omission of distal lower
extremity function and poor ability to detect subtle changes over time.

DIFFERENTIAL DIAGNOSIS
Generalized weakness may result from injury to the brain or brainstem,
myelopathies, anterior horn cell disorders, polyneuropathies, neuromuscular junction disorders, and muscle disorders (Table 83-3). The
weakness may represent the exacerbation or unmasking of a chronic
underlying neuromuscular disease. Alternatively, the weakness may
represent the acute neuromuscular condition. In cases of uncertainty,
additional tests should be performed, including neuroimaging of the
brain, brainstem, or spinal cord; infectious and immunologic serologies,
cerebrospinal fluid analysis, and electrophysiology (EP) studies.
With a good history and physical examination, many of the differential diagnoses can be excluded with confidence. Given that some of the
other diagnoses are treatable, ICUAW should be regarded as a diagnosis
of exclusion. The weakness must follow the onset of the critical illness;
symptoms prior to admission should direct attention to other etiologies.
The inability to interview the patient, due either to intubation or delirium
may limit historical detail and proper physical examination. The presence of delirium should not dissuade the search for a neuromuscular

disorder, especially when cognition is improving and the weakness is not.

1/23/2015 12:55:30 PM


766

PART 6: Neurologic Disorders

ADJUNCTIVE TESTING FOR NEUROMUSCULAR INJURY

junction testing. Peripheral motor nerve stimulation elicits a compound
muscle action potential (CMAP), which represents the summated
response of all stimulated muscle fibers. Alternatively, a sensory nerve
may be stimulated at separate points to measure the sensory nerve action
potential (SNAP), which represents the summated response of all
stimulated sensory fibers. Nerve conduction velocity is calculated by
measuring the time between nerve stimulation and recording at two
sites separated by a known distance. Taken together, the information
can diagnose an axonal sensory-motor polyneuropathy, such as CIP,
in which decreased CMAP and SNAP amplitudes are measured while
nerve conduction velocity is normal. In contrast, a demyelinating
sensory-motor polyneuropathy, like Guillain-Barré syndrome, will
exhibit preserved CMAP and SNAP amplitudes with markedly reduced
conduction velocities.
Awake and cooperative patients can undergo needle EMG. Recordings
are conducted during muscle rest, mild contraction, and with i­ ncreasing
or maximal voluntary muscle contraction. Fibrillation potentials and
sharp waves at rest suggest recent denervation or muscle necrosis.
Motor unit potentials (MUPs) are recorded during voluntary contraction. Myopathy is suggested when MUPs are of short duration and low

amplitude. With maximal contraction, early recruitment of MUPs may
occur. In contrast, long-duration, polyphasic, high-amplitude MUPs
may suggest neuropathy. For the patient with persistent respiratory
failure, phrenic nerve conduction studies and needle EMG of the diaphragm can be performed.
Assessment of the neuromuscular junction is accomplished via
repetitive nerve stimulation and/or single-fiber EMG. In repetitive nerve
stimulation, a series of supramaximal stimuli are applied at 2 to 3 Hz.
Decreases in CMAP amplitude of greater than 10% between the first
and fourth responses indicate a postsynaptic defect in neuromuscular
transmission, such as myasthenia gravis or prolonged NMBA effect (see
below). When the patient is able to contract muscle voluntarily, singlefiber EMG is possible. This test records the time interval between action
potentials in two muscle fibers that are parts of the same motor unit.
Variable inter-spike intervals, termed jitter, and absence of the second
spike (blocking) are consistent with neuromuscular dysfunction.
Limitations of EP testing include falsely dampened measurements
from tissue edema, electrical interference from other ICU equipment,
the inability for patients to voluntarily contract muscles, and the need for
specialists well-versed in the complexities of interpretation. Importantly,
competing illnesses may cause preexisting axonal polyneuropathy,
including diabetes and effects of chemotherapeutic agents.
To overcome the challenges of patient cooperation, direct m
­ uscle
stimulation can be conducted to distinguish polyneuropathy and
­myopathy.25,26 Theoretically, denervated muscle (as in CIP) should retain
electrical excitability; therefore, direct muscle stimulation CMAP amplitude should be normal. In contrast, patients with myopathy exhibit loss
of electrical excitability; therefore, both nerve and direct muscle stimulated CMAPs are diminished. To accomplish this, a stimulating needle
or surface electrode is placed just proximal to the tendon i­ nsertion. After
obtaining a muscle twitch, a recording needle electrode is placed in the
center of the muscle proximal to the site of stimulation, and the maximal muscle-stimulated CMAP (mCMAP) is recorded. Using the same
recording electrode, the appropriate nerve undergoes surface stimulation to elicit a nerve-evoked CMAP (nCMAP). The nCMAP to mCMAP

ratio is calculated; a value >0.5 suggests impaired muscle membrane
excitability.27,28

Methods to confirm ICUAW and identify its subcategories include EP
studies, direct muscle stimulation, and morphologic analysis of muscle or
nerve tissue. These tests help to exclude other differential diagnoses and
can help to characterize the specific subcategory of ICUAW: neuropathy,
myopathy, neuromyopathy, or prolonged neuromuscular junction blockade. This ­section details the application of each test. Figure 83-2 provides
an algorithm for the work-up of a patient exhibiting weakness or inactivity.
Electrophysiologic studies used to evaluate the peripheral nervous system include nerve conduction studies, needle EMG, and neuromuscular

Nerve histology in patients with electrophysiologically defined CIP
demonstrates distal axonal degeneration involving both sensory and
motor fibers with no evidence of demyelination or inflammation.
Muscle biopsies have demonstrated denervation changes and ­commonly
have myopathy. In contrast, muscle biopsy in CIM demonstrates
acute necrosis, regeneration, type II fiber atrophy, and selective loss
of thick filaments (myosin).29 This last feature is proven by the loss of

  TABLE 83-3    Acute Generalized Weakness Syndromes in Critically Ill Patients
Bilateral or paramedian brain or brainstem lesionsa
1.Trauma
2.Infarction
3.Hemorrhage
4. Infectious and noninfectious encephalitides
5.Abscess
6. Central pontine myelinolysis
Spinal cord disordersa
1.Trauma
2. Nontraumatic compressive myelopathies

3. Spinal cord infarction
4. Immune-mediated myelopathies (transverse myelitis, neuromyelitis optica)
5. Infective myelopathies (eg, HIV, West Nile virus)
Anterior horn cell disorders
1. Motor neuron disease
2.Poliomyelitis
3. West Nile virus infection
4. Hopkins syndrome (acute postasthmatic amyotrophy)
Polyradiculopathies
1.Carcinomatous
2.HIV-associated
Peripheral nervous disorders
1. Guillain-Barré syndromeb
2. Diphtheritic neuropathy
3. Lymphoma-associated neuropathy
4. Vasculitic neuropathy
5. Porphyric neuropathy
6. Paraneoplastic neuropathy
7. Critical illness polyneuropathy
Neuromuscular junction disorders
1. Myasthenia gravis
2. Lambert-Eaton myasthenic syndrome
3. Neuromuscular-blocking drugs
4.Botulism
Muscle disorders
1.Rhabdomyolysis
2. Disuse myopathy
3.Cachexia
4. Infectious and inflammatory myopathiesc
5. Mitochondrial myopathies

6. Drug-induced and toxic myopathies
7. Critical illness myopathy
8.Decompensation of congenital myopathies (eg, myotonic dystrophy, Duchenne
­muscular ­dystrophy, adult onset acid maltase deficiency)
a

Upper motor neuron signs (increased tone, hyperreflexia) may be absent in the acute setting.

Includes acute inflammatory demyelinating polyneuropathy, acute motor axonal neuropathy, acute
motor, and sensory axonal neuropathy.
b

Includes polymyositis, dermatomyositis, pyomyositis.

c

HIV, human immunodeficiency virus.

section06.indd 766

■■BIOPSY

1/23/2015 12:55:30 PM


CHAPTER 83: ICU-Acquired Weakness

767

Critically ill patient exhibits weakness

or inactivity
Awake and able to participate in
neuromuscular exam (MRC exam)?
No

Yes

Asymmetric
weakness

ICUAW

Not weak

EP Studies
EMG, NCS
Repetitive nerve stimulation

Observation
Fails to
improve

CIP, CIM

Prolonged
NMJ blockade

Other
diagnosis

Gradual
improvement
No further
testing

Confirmatory
test needed?
DMS, muscle
biopsy

FIGURE 83-2.  Diagnostic algorithm for weakness in the ICU. CIM, critical illness myopathy; CIP, critical illness polyneuropathy; DMS, direct muscle stimulation; EMG, electromyography;
EP, electrophysiology; ICUAW, intensive care unit-acquired weakness; MRC, Medical Research Council; NCS, nerve conduction studies; NMJ, neuromuscular junction.

myofibrillar adenosine triphosphate staining on electron microscopic
imaging. Although biopsies have provided valuable insight into the
mechanism of injury, the role of nerve and muscle biopsies in clinical
practice is controversial. The prognostic value of histologic findings
remains poorly explored.

■■OTHER DIAGNOSTIC TESTS: BIOMARKERS AND IMAGING

Increased serum creatine kinase has been reported in patients with
acquired myopathy, particularly those with necrotizing myopathy. There
is simultaneous interest in the use of ultrasound to image muscle to
infer muscle bulk.29,30 Decrease in muscle thickness over time has been
documented in measurements of the anterior thigh, forearm, and biceps.
For example, linear array, high frequency probes can be used to measure
quadriceps bulk at a specified point. Validation studies of this tool as a
marker of muscle bulk and injury are ongoing.

OF ICUAW: CIP, CIM, CIPNM AND PROLONGED
■■SUBCATEGORIES
NEUROMUSCULAR JUNCTION BLOCKADE

Given the complex testing options, a comprehensive diagnostic
nomenclature and classification has been generated (Fig. 83-1). As
described above, the term ICU-acquired weakness designates clinically detected weakness in critically ill patients in whom there is no
plausible etiology other than critical illness. When advanced testing
is conducted, more specific phenotypes (or subcategories of ICUAW)
can be described. Critical illness polyneuropathy refers to patients with
ICUAW who have electrophysiological evidence of an axonal polyneuropathy. Critical illness myopathy indicates patients with ICUAW
who have electrophysiologic and/or histologic defined myopathy. The
term critical illness neuromyopathy (CINM) is reserved for patients
who have electrophysiologic and/or histologic findings of coexisting
CIP and CIM. Finally, a rare entity of prolonged neuromuscular junction blockade exists with overlapping clinical features of ICUAW and
distinct EP features.

section06.indd 767

■■CRITICAL ILLNESS POLYNEUROPATHY (see Table 83-4)

Critical illness polyneuropathy (CIP) is a distal axonal sensory-motor
polyneuropathy affecting both limb and respiratory muscles. As in all
cases of ICUAW, it is usually discovered in patients with prolonged critical illness, particularly mechanical ventilation, and affected patients have
limb muscle weakness—particularly distal weakness—with reduced or
absent deep tendon reflexes.27 When measurable, patients have loss of
peripheral sensation to light touch and pin prick, yet preserved cranial
nerve function.
Given the limitations of the sensory examination in critically ill
patients, EP studies have generally been relied upon to establish the

diagnosis. EP studies show a reduction in amplitude of CMAPs and
SNAPs reflecting the underlying axonal loss.31 Nerve conduction velocity is normal or mildly reduced. Over time, fibrillation potentials will
be evident on electromyography (EMG) needle examination. In severe
cases with ventilatory failure, phrenic motor amplitudes are commonly

  TABLE 83-4    Major Diagnostic Features of CIP
1. Evidence for ICUAW
2. Abnormal sensory examination (when possible)
3. Electrophysiologic evidence of axonal motor and sensory polyneuropathy
• Sensory and motor nerve amplitudes <80% of the lower limit of normal in two or
more nerves
• Normal or near normal conduction velocities without conduction block
• Absence of a decremental response on repetitive nerve stimulation
Other supportive findings:
• Needle EMG with reduced recruitment of normal motor unit potentials (early finding)
• Needle EMG with fibrillation potentials and reduced recruitment of long-duration,
high-amplitude MUPs (late finding)
• Normal CSF protein
• Normal serum creatine kinase

1/23/2015 12:55:31 PM


768

PART 6: Neurologic Disorders

reduced or absent. In patients with CIP, direct needle stimulation of
muscle elicits a relatively higher amplitude response compared with the
response recorded from muscle after nerve stimulation.

The serum CK level is normal, and, when performed, cerebrospinal
fluid protein levels are usually normal. Muscle biopsy findings are those
of neurogenic atrophy. Nerve histology in patients with electrophysiologic-defined CIP demonstrates distal axonal degeneration involving
both sensory and motor fibers with no evidence of demyelination or
inflammation.
Prior investigations have commonly associated CIP with severe
sepsis and experts suspect it represents a neurologic manifestation of
the systemic inflammatory response syndrome (SIRS).28,32,33 There is
some correlation with elevations in blood glucose and reductions in
serum albumin.34 The mechanism of axonal injury in CIP is unknown.
However, injury to the microcirculation of distal nerves, causing ischemia and axonal degeneration, is speculated.33,35 During the early stages
of sepsis, electrical inexcitability due to sodium channel inactivation
may be present in otherwise intact nerves.

■■CRITICAL ILLNESS MYOPATHY (see Table 83-5)

The most common form of intensive care unit (ICU)-acquired myopathy is critical illness myopathy (CIM).36 The most common presenting
features of CIM are flaccid quadriparesis that may have a different pattern than CIP. Whereas CIP exhibits a length-related pattern (ie, distal
muscles are weakest), CIM usually affects proximal muscles either
equally or more pronounced than distal muscles. Facial muscle weakness can occur, but extraocular muscle weakness is rare. Like other
entities, patients often repeatedly fail to wean from mechanical ventilation. Although not always assessable, sensation should be normal. For
example, these patients often grimace to painful stimuli even during
periods of delirium.
In retrospective series of patients with CIM, approximately one-half
had elevations in CK.37 In patients with appropriate clinical features, the
diagnosis of CIM can be confirmed by electrophysiologic testing with
nerve conduction studies (NCS) and electromyography (EMG). Muscle
biopsy establishes the diagnosis, but is rarely performed unless another
treatable condition, such as an inflammatory myopathy, is in the differential diagnosis.
The major nerve conduction findings of CIM are normal to low

motor amplitudes with occasional broadening of the CMAP.38,39 Phrenic
motor amplitudes may also be low. Sensory responses are normal or
only mildly reduced, unless there is a coexisting polyneuropathy. Needle
examination frequently demonstrates fibrillation potential activity
implicating recent denervation or muscle necrosis.31 Observation of the
recruitment of motor unit potentials (MUPs) may not be possible in
advanced weakness. When feasible, recruitment tends to be early. MUPs

  TABLE 83-5    Major Diagnostic Features of CIM
1. Evidence for ICUAW
2. Intact sensory examination (when possible)
3. Electrophysiologic evidence of myopathy without neuropathy
a.Needle EMG with short-duration, low-amplitude MUPs with early or normal full
­recruitment, with or without fibrillation potentials in 2 or more muscle groups
b. Absence of other nerve injury
i.Sensory nerve amplitudes >80% of the lower limit of normal in two or more
nerves on nerve conduction studies
ii. Absence of a decremental response on repetitive nerve stimulation
4. Muscle inexcitability on direct muscle stimulation
a. Muscle-stimulated CMAP/nerve-evoked CMAP ratio >0.5 in 2 or more muscles
5. Muscle histopathologic findings of myopathy with myosin loss
Other supportive findings:
1. Motor amplitudes <80% of the lower limit of normal in two or more nerves
2. Elevated serum creatine kinase

section06.indd 768

are short in duration, low in amplitude, and may be polyphasic.40 In
contrast, long-duration, polyphasic, high amplitude MUPs may suggest
neuropathy.

Direct muscle stimulation can be conducted to assess for electrical
inexcitability and may help to differentiate CIM from motor axonopathy.26 However, this modality is often limited to those patients with a
coexisting peripheral neuropathy.
Alternatively, CIM may be established with muscle biopsy. The major
histopathologic finding is the selective loss of myosin, identified as a
lack of reactivity to myosin ATPase in non-necrotic fibers. This finding
can be confirmed with immunohistochemic studies for myosin and by
utilizing electron microscopy to identify loss of thick filaments. There
is usually atrophy of myofibers (type 2 more than type 1), evidence of
myofibrillar disorganization, and occasional necrosis.41,42
Several processes may be involved in the pathogenesis of CIM, including upregulation of calpain, an increase in muscle apoptosis, activation
of the proteasome ubiquitin-degradative system, and upregulation of
the transforming growth factor-beta/mitogen-activated protein kinase
pathway.43 Oxidative stress may also play a role. Observation of the loss
of sarcolemmal nitric oxide synthase isoform 1 may lead to muscle fiber
inexcitability.44
A steroid-denervation animal model reproduces the histopathologic
and electrophysiologic findings of CIM observed in humans.45 This
model suggests that a deleterious interaction between glucocorticoids
and denervation leads to depletion of the mRNA for myosin and results
in muscle atrophy.46 Finally, muscle sodium channel properties have
also been implicated using a chronic sepsis animal model. Patch clamp
technique revealed decreased sodium current that could lead to muscle
inexcitability.47

■■CRITICAL ILLNESS POLYNEUROMYOPATHY

More recent investigations have proven that a reasonable proportion of
patients have features of combined CIM and CIP, termed critical illness
polyneuromyopathy.48 The commonality of this entity was illustrated by

a prospective longitudinal cohort study of 48 patients who had baseline
neurologic examinations and nerve conduction studies (NCS) within
72 hours of developing severe sepsis.49 Electromyography was performed
on patients who developed clinical weakness or had 30% or greater
reduction in nerve conduction response amplitudes. Clinical and electrophysiologic examinations were repeated weekly for the duration of
the ICU stay. Abnormal NCS were present at baseline in 63% of patients,
and an abnormality on baseline NCS was significantly associated with
hospital mortality compared with a normal baseline NCS (55% vs 0%,
respectively). In 20 patients who remained in the ICU long enough to
have serial NCS, neuromuscular dysfunction developed in 10 patients
(50%). Electrophysiologic evidence of both CIM and CIP was present
in 8 of 10 patients with neuromuscular dysfunction. The investigators
hypothesized that sepsis may be a common pathologic mechanism
underlying the development of both CIM and CIP.

■■PROLONGED NEUROMUSCULAR JUNCTION BLOCKADE

Prolonged neuromuscular junction (NMJ) blockade is a rare disorder
occurring in patients who receive non-depolarizing NMBAs who experience persistent generalized weakness and respiratory failure despite drug
cessation.50 These paralytic agents inhibit neuromuscular transmission
via reversible binding to acetylcholine receptors on the motor end-plates
of NMJs. However, specific drugs ­requiring end organ function for
clearance may have persistent effects, particularly when infused for prolonged periods. For example, aminosteroid blocking agents, such as pancuronium and vecuronium, undergo metabolism by the liver and result
in functionally active 3-hydroxy metabolites. In situations of advanced
liver or kidney injury (­creatinine clearance <30 mL/min), these drugs
can accumulate for prolonged effect. Other reported contexts include
hypermagnesemia or metabolic acidosis.
Examination is notable for flaccid quadriplegia, arreflexia, and
involvement of the cranial nerves, including ptosis, ophthalmoparesis,

1/23/2015 12:55:31 PM


CHAPTER 83: ICU-Acquired Weakness

and facial weakness. Train-of-four stimulation with a peripheral
nerve stimulator measures the decremental response semiquantitatively and may detect a major neuromuscular junction defect. Formal
testing with repetitive nerve stimulation is the confirmatory test. In
this test, a series of supramaximal stimuli are applied with a nerve
stimulator. Decreases in CMAP amplitude greater than 10% from the
first to fourth response indicate a postsynaptic defect in neuromuscular transmission, such as myasthenia gravis or prolonged NMBA
effect. Transient improvement in muscle strength after administration of an anticholinesterase reversing agent, such as pyridostigmine,
supports prolonged neuromuscular junction blockade as a cause
of weakness.
The condition is reversible and recovery of motor function is
observed over a period of 2 to 10 days. Weakness beyond this duration should prompt consideration for alternative diagnoses, especially
other neuromuscular junction diseases such as myasthenia gravis.
Prolonged neuromuscular blockade can be prevented by avoiding
aminosteroid blocking drugs in favor of benzylisoquinoline agents,
such as cisatracurium, which has no dependence on end organ
function (metabolized by rapid nonenzymatic degradation in the
bloodstream, Hofmann elimination). Indeed, the routine use of cisatracurium for neuromuscular blockade in the ICU has largely eliminated this problem.

■■ICUAW EPIDEMIOLOGY AND RISK FACTORS

Several studies have attempted to establish the prevalence of ICUAW
and its associated risk factors. Given the history of reliance upon
advanced testing to delineate phenotypes, large-scale epidemiology
studies have not been conducted. In this context, the best summary
data is a systematic review of 24 published studies that included both

clinical and electrophysiologic examination.51 Their end point was
abnormal EP test findings (including CIP, CIM, and CIPNM), which
they termed critical illness neuromuscular abnormalities (CINMAs), a
label now interchangeable with ICUAW. Of the 1421 total patients with
sepsis, multiorgan failure, or prolonged mechanical ventilation, 46% had
ICUAW. The risk of ICUAW was associated with hyperglycemia (and
inversely associated with tight glycemic control), the systemic inflammatory response syndrome (SIRS), sepsis, multiple organ dysfunction,
renal replacement therapy, and catecholamine administration. Across
studies, there was no consistent relationship between ICUAW and
patient age, gender, severity of illness, or exposure to glucocorticoids,
neuromuscular blockers, aminoglycosides or midazolam. Unadjusted
mortality was not increased in the majority of patients with ICUAW, but
mechanical ventilation and ICU LOS were prolonged.
The cohort study that established the validity of the physical examination
for ICUAW had both complementary and different findings.18 For example,
in the 95 ICU patients who underwent mechanical ventilation for 7 days
or more, independent predictors of ICUAW included the number of days
with dysfunction of two or more organs (OR: 1.28, 95% CI: 1.11-1.49) and
the duration of mechanical ventilation (OR: 1.10, 95% CI: 1.00-1.22). In
contrast to the systematic review, female sex (OR: 4.66, 95% CI: 1.19-18.30)
and administration of corticosteroids (OR: 14.90, 95% CI: 3.20-69.80) were
strong predictors.
In search of potentially modifiable risk factors for ICUAW, many
investigations have focused on exposure to corticosteroids and NMBAs.
These agents have been both implicated in animal research and
­observational trials in humans.52 Results have not been consistent or
conclusive, likely due to methodological limitations of these investigations. More recently, randomized controlled trials have included
secondary analyses for evidence of ICUAW to bypass the problem of
confounding by indication.23
Although corticosteroids inhibit protein synthesis in type II muscle

fibers and contribute to severe protein catabolism, the relationship
between corticosteroids and ICUAW has been inconsistent. In a secondary analysis of a multicenter study of patients with severe and persistent
ARDS randomized to methylprednisolone or placebo, 34% developed

section06.indd 769

769

ICUAW as detected by chart review.52 There was no statistically significant association of ICUAW with randomization to methylprednisolone;
however, intervention patients were more likely to have evidence of
ICUAW in the first 28 days of the study, and were more likely to be
clinically diagnosed with myopathy. It is plausible that some benefits
of corticosteroid treatment on lung function were offset by the adverse
effects on strength.
The association of neuromuscular weakness with prolonged effect
of neuromuscular blocking agents (NMBAs) has long been recognized
and was the most prominent reason for a shift away from NMBA use
in the critically ill. A typical scenario involves patients with severe
acute asthma and ventilatory failure who undergo treatment with highdose corticosteroids in combination with NMBAs. These patients may
exhibit severe and protracted myopathy.37,53,54 However, this relationship
has not borne out in the general adult ICU population.36,42 Concerns
about NMBA use have been reduced by a recent multicenter RCT testing the benefit of early neuromuscular blockade for severe ARDS.23
Randomization to cisatracurium versus placebo significantly decreased
90-day mortality from 40.7% to 31.6%. Investigators included ICUAW as
a secondary outcome. At ICU discharge, there were neither differences
in average muscle strength among patients tested nor any difference in
proportion of patients with ICUAW. These findings are a substantial
contribution, challenging the commonly held belief about the causal role
of neuromuscular blockers in ICUAW. However, there are some important limitations, including the use of manual muscle strength testing as
the gold standard for investigating nerve and muscle function in the

ICU and lack of follow-up testing to answer questions about lingering
impairment.

PREVENTION AND TREATMENT
Data supporting specific approaches to prevent or treat ICUAW are
limited. A Cochrane review identified only one successful intervention: insulin therapy with strict glycemic control.55 This evidence for
prevention comes from two trials studying “intensive” insulin therapy
(defined as maintenance of a blood glucose level between 80 and
110 mg/dL) in critically ill patients who remain in the ICU for 7 or
more days. The first trial, focused on surgical patients, demonstrated
a mortality benefit and a secondary end point of fewer cases of CIP
detected by routine electrophysiologic testing after day 7 (29% vs 52%,
p < 0.001).56 The same investigators studied the effect of intensive
insulin therapy in medically critically ill patients. The prospective
subanalysis demonstrated a significant reduction in the incidence of
critical illness polyneuropathy and myopathy (51% vs 39%, p = 0.02)
when similarly screened by weekly EP studies. Unfortunately, despite
this protective effect on the development of ICUAW, intensive insulin
therapy has been associated with an increased risk of severe hypoglycemia and either increased mortality or had no effect on mortality
when compared to more permissive blood glucose ranges (such as
140-180 mg/dL and 180-200 mg/dL).57,58 Furthermore, because more
recent data suggest an increased mortality with aggressive insulin
therapy,58 this treatment option cannot be recommended as a means
to prevent ICUAW.
For all forms of ICUAW, care is supportive. Measures to avoid secondary injury must be undertaken. These practices may span mechanical
ventilation (low tidal volume ventilation for ARDS; protocols to guide
ventilator readiness testing and liberation), stress ulcer and venous
thrombosis prophylaxis, titrated sedation and analgesia therapy, and
efforts to avoid nosocomial infection (head of bed elevation, early
discontinuation of central venous and urinary catheters). Because

­prolonged immobilization and bed rest have been shown to accelerate muscle loss, which may exacerbate ICUAW, mobility therapy has
emerged as a potential preventive measure.59
A new framework for early mobilization during critical illness
has evolved. Rather than delay rehabilitation until the patient has
left the ICU, studies of progressively earlier exercise have repeatedly

1/23/2015 12:55:31 PM


770

PART 6: Neurologic Disorders

demonstrated safety and benefit.59 Specifically, early mobilization can be
safely implemented during mechanical ventilation via an endotracheal
tube, during infusions of vasopressors and relatively higher levels of
oxygen need, and in patients with multiple critical care devices.60-63 These
studies, spanning physical and occupation therapy services to bedside
cycle ergometer use, are detailed extensively in Chap. 24. Overall, these
studies demonstrate improved patient physical function and shorter
durations of ICU and hospital lengths of stay.59

SUMMARY
ICUAW is a common morbidity of critical illness, represents an
­important patient-centered outcome, and has substantial implications
on quality of life and patients’ ability to return to prior health and lifestyle. The ability to measure the presence of ICUAW in a reproducible
fashion via history and physical examination has yielded significant
improvements in global awareness of neuromuscular dysfunction. The
practicing clinician needs to be aware when the presentation is atypical
and more advanced diagnostic testing is needed. For the research environment, longer term outcomes focusing on neuromuscular strength

and patient functional autonomy need to be considered when evaluating
the effect of new interventions. Although it seems doubtful that a single
therapy might prevent weakness in varied populations, the meticulous
application of multidisciplinary care—including early patient engagement and mobilization—may help to improve strength and function in
survivors of critical illness.

KEY REFERENCES
•• Batt J, dos Santos CC, Cameron JI, Herridge MS. Intensive care
unit-acquired weakness: clinical phenotypes and molecular mechanisms. Am J Respir Crit Care Med. 2012;187:238-246.
•• De Jonghe B, Sharshar T, Lefaucheur JP, et al. Paresis acquired in
the intensive care unit: a prospective multicenter study. JAMA.
2002;288:2859-2867.
•• Hermans G, De Jonghe B, Bruyninckx F, Van den Berghe G.
Interventions for preventing critical illness polyneuropathy and
critical illness myopathy. Cochrane Database of Systematic Reviews
(Online). 2009:CD006832.
•• Herridge MS, Batt J, Hopkins RO. The pathophysiology of longterm neuromuscular and cognitive outcomes following critical
illness. Crit Care Clin. 2008;24:179-199, x.
•• Kress JP, Hall JB. ICU-acquired weakness and recovery from critical illness. N Engl J Med. 2014; 370:1626-35.
•• Lacomis D. Electrophysiology of neuromuscular disorders in critical illness. Muscle Nerve. 2013;47:452-463.
•• Latronico N, Peli E, Botteri M. Critical illness myopathy and neuropathy. Curr Opin Crit Care. 2005;11:126-132.
•• Puthucheary ZA, Rawal J, McPhail M, et al. Acute skeletal muscle
wasting in critical illness. JAMA. 2013;310:1591-1600.
•• Stevens RD, Dowdy DW, Michaels RK, Mendez-Tellez PA,
Pronovost PJ, Needham DM. Neuromuscular dysfunction
acquired in critical illness: a systematic review. Intensive Care
Med. 2007;33:1876-1891.
•• Stevens RD, Marshall SA, Cornblath DR, et al. A framework for
diagnosing and classifying intensive care unit-acquired weakness.
Crit Care Med. 2009;37:S299-S308.

•• Stiller K. Physiotherapy in intensive care: an updated systematic
review. Chest. 2013;144:825-847.

REFERENCES
Complete references available online at www.mhprofessional.com/hall

section06.indd 770

CHAPTER

84

Cerebrovascular Disease
William J. Powers
Dedrick Jordan

KEY POINTS
•• Etiology
•• Cardioembolic and other nonarteriosclerotic causes of cerebral
infarction occur more commonly in patients admitted to the ICU
and should be carefully sought by appropriate diagnostic tests.
•• In hypertensive patients with hemispheric lobar hemorrhages
and in patients without hypertension, causes for intracerebral
hemorrhage such as coagulopathies, arteriovenous malformations, or saccular aneurysms should be sought.
•• Nontraumatic spontaneous subarachnoid hemorrhage is almost
always due to a ruptured saccular aneurysm and should be
evaluated by arteriography.
•• Clinical and Laboratory Diagnosis
•• X-ray computed tomography (CT) is the diagnostic neuroimaging test of choice for patients with acute stoke. It is rapid, can be
performed easily on acutely ill patients and acute intracerebral

or subarachnoid hemorrhage are easily identified.
•• Lumbar puncture is the most sensitive test for detection of SAH;
it should be performed when there is a strong clinical suspicion
and a negative CT scan, or when CT is not available or feasible.
•• In suspected ischemic stroke, diffusion-weighted MRI can be
helpful for improving diagnostic certainty when there is no
clear history of an abrupt onset or the localization of the neurological findings is confusing. MRI has not been shown to be of
value in selecting patients for thrombolytic therapy.
•• Early electrocardiographic (ECG) monitoring detects previously unsuspected atrial fibrillation in 3% to 5% of patients
with acute cerebral ischemia.
•• Patients with transient ischemic attacks (TIAs) or mild stroke who
are good surgical candidates for carotid endarterectomy should
be evaluated for symptomatic carotid stenosis immediately since
the risk of stroke can be as high as 1 in 20 within the first 2 days.
•• Treatment of Cerebral Infarction
The following statements can be made based on good clinical trial data.
•• Routine use of supplemental oxygen does not reduce mortality.
•• Early treatment of hyperglycemia to achieve levels <7 mmol/L
does not improve outcome.
•• In patients with systolic blood pressures of 160 to 200 mm Hg,
pharmacological reduction of systolic pressure by 20 to 25 mm Hg
within the first 24 hours is safe, but does not improve outcome.
•• In hemiplegic patients, subcutaneous low-dose heparin or
enoxaparin reduces deep venous thrombosis.
•• Intravenously administered t-PA improves outcome in carefully
selected patients with acute ischemic stroke when instituted
within 4.5 hours of onset.
•• The clinical value of any intra-arterial pharmacological or
mechanical revascularization therapy for acute ischemic stroke
has not been demonstrated.

•• Aspirin 160 or 300 mg/d of aspirin begun within 48 hours of
the onset of ischemic stroke results in a net decrease in further
stroke or death.
•• Full anticoagulation with heparin or similar drugs in patients
with acute ischemic stroke provides no clinical benefit in general

1/23/2015 12:55:32 PM


CHAPTER 84: Cerebrovascular Disease

or in any subgroup, including those with atrial fibrillation or
other cardioembolic sources.
•• Hemicraniectomy reduces mortality in patients with large
hemispheric infarcts and depressed level of consciousness who
are ­operated within 48 hours of stroke onset.
•• Treatment of Intracerebral Hemorrhage
The following statements can be made based on good clinical trial data
•• Prophylaxis for deep venous thrombosis with low-dose subcutaneous heparin or heparinoids may be instituted safely on the
second day after the hemorrhage and reduces subsequent deep
venous thrombosis if begun before day 4.
•• In patients with systolic blood pressure of 150 to 220 mm Hg,
rapid pharmacological reduction of systolic pressure by 27 mm Hg
within the first hour is safe but does not improve outcome.
•• Craniotomy and clot evacuation in patients with supratentorial
ICH, either superficial or deep, is of no benefit.
•• Treatment of Subarachnoid Hemorrhage
The following statements can be made based on good clinical trial data
•• Oral nimodipine at a dose 60 mg every 4 hours for 21 days after
­hemorrhage reduces poor outcome.

•• Early definitive treatment reduces the risk of rebleeding.
•• For aneurysms amenable to both endovascular coiling and
surgical clipping, endovascular treatment is beneficial.
•• Intravascular volume contraction should be avoided.

ETIOLOGY
Cerebrovascular diseases can be divided into three categories: cerebral
ischemia and infarction, intracerebral hemorrhage, and subarachnoid
hemorrhage. Cerebral ischemia and infarction are caused by processes
that reduce cerebral blood flow. Reductions in whole brain blood flow
due to systemic hypotension or increased intracranial pressure (ICP) may
produce infarction in the distal territories or border zones of the major
cerebral arteries. More prolonged global reductions cause diffuse hemispheric damage without localizing findings or, at its most severe, produce
brain death. Prolonged regional reductions can lead to focal brain infarctions. Local arterial vascular disease accounts for approximately 65% to
70% of all focal brain infarctions. In most cases, arterial disease serves
as a nidus for local thrombus formation with or without subsequent
distal embolization. Focal arterial stenosis in combination with systemic
hypotension is a very rare cause of focal brain infarction. Atherosclerosis
is the most common cause of local disease in the large arteries supplying
the brain. Disease of smaller penetrating arteries may cause small deep
(lacunar) infarcts. While emboli arising from the heart cause approximately 30% of all cerebral infarcts in a general population, they assume
more importance in ICU patients.1 Atrial fibrillation is the most common
of these causes. Atherosclerotic emboli following heart surgery, infective endocarditis, nonbacterial thrombotic endocarditis, and ventricular
mural thrombus secondary to acute myocardial infarction or cardiomyopathy should all be considered in the appropriate circumstances.
More rare causes of cerebral infarction must also be considered in the
ICU. These include dissections of the carotid or vertebral artery (after
direct neck trauma, “whiplash” injuries or forced hyperextension ­during
endotracheal intubation), intracranial arterial or venous thrombosis
secondary to meningeal or parameningeal infections, and paradoxical
embolization from venous thrombosis via a patent foramen ovale.1

Hemorrhage into the basal ganglia, thalamus, and cerebellum in
middle-aged patients with long-standing hypertension is the most
­common type of intracerebral hemorrhage. In hypertensive patients

section06.indd 771

771

with hemispheric lobar hemorrhages and patients without hypertension,
other causes should be sought, such as arteriovenous malformations
or saccular aneurysms.2 Amyloid angiopathy becomes increasingly
important in patients in the seventh, eighth, and ninth decades. These
hemorrhages usually occur in the subcortical hemispheric white matter
and may be multiple. Previous microhemorrhages in parietal and occipital lobes are often visible on magnetic resonance images. Hemorrhage
due to anticoagulant and thrombolytic drugs may affect any part of the
brain. Rarer causes of intracerebral hemorrhage occurring in patients
with other systemic diseases include thrombocytopenia, hemophilia,
and disseminated intravascular coagulation. Primary or metastatic brain
tumors will rarely present as ICH.
Nontraumatic spontaneous subarachnoid hemorrhage (SAH) is
almost always due to a ruptured saccular aneurysm. Aneurysms may
also rupture into the brain parenchyma, producing intracerebral hemorrhage as well. Saccular aneurysms are most commonly located on
the large arteries at the base of the brain. Both congenital and acquired
factors appear to play a role in the postnatal development of aneurysms. Acquired factors include atherosclerosis, hypertension, and
hemodynamic stress. In patients with infective endocarditis, mycotic
aneurysms of more distal arteries may form and sometimes rupture.
Other causes of SAH include ruptured arteriovenous malformations
(cerebral and spinal) and fistulae, cocaine abuse, pituitary apoplexy,
and intracranial arterial dissection.3 In some cases, particularly SAH
ventral to the midbrain or restricted to cortical sulci, the cause cannot

be determined.

CLINICAL AND LABORATORY DIAGNOSIS
The initial diagnostic evaluation of the patient with suspected stroke
serves (1) to determine whether neurologic symptoms are due to cerebrovascular disease or to some other condition, such as peripheral nerve
injury, intracranial infection, tumor, subdural hematoma, multiple sclerosis, epilepsy, or hypoglycemia; and (2) to distinguish among different
types of cerebrovascular disease that require different treatments. The
clinical history and examination remains the cornerstone of this process.
Cerebrovascular disease typically produces focal brain dysfunction of sudden onset in a single location. The primary exception to this is aneurysmal
SAH, which usually presents as a sudden onset of severe headache, with or
without nausea, vomiting, or loss of consciousness. In some cases, a less
severe aneurysmal hemorrhage may present as a headache of moderate
intensity, neck pain, and nonspecific symptoms. A high index of suspicion
is needed in order to avoid missing the diagnosis of SAH. Focal brain
dysfunction may not always cause an obvious hemiparesis. Neurologic
­deficits such as neglect, agnosia, aphasia, visual field defects, or amnesia may be the only manifestations of brain infarction or hemorrhage.
Multiple small brain infarcts may produce impaired consciousness with
minimal or no focal neurologic deficits, mimicking metabolic, or toxic
encephalopathy. The clinical distinction between cerebral infarction and
intracerebral hemorrhage is unreliable as both produce sudden focal deficits. Large hemorrhages may produce vomiting or unconsciousness, but
so may infarcts in the vertebrobasilar circulation. The initial neurologic
examination provides a baseline for monitoring the subsequent clinical
course. A thorough medical evaluation is necessary to detect systemic
diseases that may be the cause of the cerebrovascular problem. Careful
evaluation of the heart is imperative to detect conditions that might predispose to embolization, particularly atrial fibrillation, recent myocardial
infarction, and more rarely, infective endocarditis.
X-ray computed tomography (CT) is the diagnostic neuroimaging test
of choice for patients with acute stoke. It is rapid and can be performed
easily on acutely ill patients. Acute intracerebral hemorrhage is easily
identified by noncontrast CT. Cerebral infarction may not be demonstrated by CT for several days. If the infarct is small enough, it may never

be apparent. Magnetic resonance diffusion weighted imaging is more
sensitive than CT for lesion detection in the early period following ischemic infarction. Due to its higher resolution, magnetic resonance imaging (MRI) is also superior for detecting small infarcts (especially those in

1/23/2015 12:55:32 PM


772

PART 6: Neurologic Disorders

the posterior fossa) at any time. However, MRI is more cumbersome to
perform in acutely ill patients because of longer imaging times, the need
for special nonferromagnetic life support equipment, and the necessity of
putting the entire body in the scanner. Demonstration of cerebral infarction by neuroimaging is rarely necessary, since the diagnosis often can be
made reliably by the clinical presentation of the sudden onset of a focal
brain deficit together with a negative CT scan to exclude hemorrhage and
other conditions. MRI can be helpful for improving diagnostic certainty
when there is no clear history of an abrupt onset or the localization of the
neurological findings is confusing. Intravenous contrast administration
increases the sensitivity for detecting diseases that may mimic stroke,
such as tumor, chronic subdural hematoma, and abscess.
Diagnosis of border zone infarction due to systemic arterial hypotension is almost entirely dependent on the pattern of infarction shown
by CT or MRI. Border zone infarctions are often asymmetrical and
patchy; rarely is the entire border zone territory between the middle
cerebral artery and posterior or anterior cerebral artery involved.
Furthermore, the actual location of the border zone varies from person
to person.4 When more than one area of acute infarction has occurred
and all infarcted areas are within the border zones, systemic hypotension
should be considered as a cause of infarction.
MRI has no advantage over CT in the demonstration of acute intracerebral hemorrhage, but it does have superior sensitivity for detecting

subacute or chronic hemorrhage. MRI with contrast is the most sensitive
way to detect a tumor underlying an ICH. Noncontrast CT has a sensitivity of >90% for detecting SAH when performed within 24 hours of
hemorrhage. There is no role for standard MRI in the initial diagnosis of
acute SAH since it is difficult to perform in an acutely ill agitated patient
and it does not increase the likelihood of detecting SAH.
In the patient who is awake and alert with acute focal brain dysfunction and in whom noncerebrovascular causes can be excluded, the immediate distinction between cerebral infarction and cerebral hemorrhage
may not be necessary if no emergent treatment of the stroke is planned.
In certain situations, however, differentiation between infarction and
hemorrhage may be critical. Patients with ischemic stroke whose time
of onset can be determined to be less than 4.5 hours earlier and whose
other medical problems do not preclude thrombolytic therapy, will
benefit from treatment with intravenous tissue plasminogen activator
(t-PA).5,6 In this circumstance, emergency CT to exclude cerebral hemorrhage is imperative (see the section on treatment below). In the patient
with decreased consciousness and a focal neurologic deficit, emergency
CT may be critically important in identifying an intracranial tumor or
subdural hematoma that requires emergency neurosurgical intervention.
Except in patients with cerebral venous thrombosis, hematologic evaluation of patients with ischemic stroke is rarely of value. Antiphospholipid
antibodies are found in a high percentage of patients with arterial stroke,
but they confer neither a worse prognosis nor is there a benefit of longterm anticoagulation.7 Acquired or hereditary hypercoagulable disorders
have not been clearly linked to arterial ischemic stroke, whereas they are
clearly of etiologic importance in cerebral venous thrombosis. In patients
with intracranial hemorrhage, especially in the ICU, acquired hemorrhagic diatheses (eg, anticoagulant or thrombolytic drugs, thrombocytopenia) should always be considered and should be sought by appropriate
laboratory testing when clinical suspicion indicates.
Lumbar puncture with cerebrospinal fluid (CSF) examination can
be an extremely important test in the evaluation of the patient with
apparent stroke, especially in patients with acquired immune deficiency
syndrome (AIDS) or when there is infection elsewhere. Meningitis may
cause stroke by producing thrombosis of arteries or cortical veins. CSF
pleocytosis is common following septic embolism from infective endocarditis and can serve as a valuable clue to its presence. Lumbar puncture
is the most sensitive test for detection of SAH; it should be performed

when there is a strong clinical suspicion and a negative CT scan, or when
CT is not available or feasible. CSF xanthochromia, which begins to
develop after 4 hours and is reliably present at 12 to 24 hours, can help
differentiate SAH from traumatic lumbar puncture.8,9

section06.indd 772

Early electrocardiographic (ECG) monitoring detects previously
unsuspected atrial fibrillation in 3% to 5% of patients with acute cerebral
ischemia.10-12 This information is clinically useful since the superiority of
oral anticoagulation over aspirin for long-term secondary stroke prevention in this circumstance has been demonstrated.13 There is, however, no
benefit for immediate anticoagulation in these patients.14 Transthoracic
echocardiography can provide evidence of poor left ventricular function
and, rarely, left ventricular thrombi. In patients without clinical cardiac
disease (no previous history or signs or symptoms of cardiac disease, no
ECG abnormalities, and normal cardiac silhouette on chest x-ray), left
ventricular thrombi are vanishingly rare. Transesophageal echocardiography has made it possible to identify left atrial thrombi and atherosclerosis of the ascending aorta. Large aortic arch lesions are associated with
an increased risk of stroke. The most common lesion detected by echocardiography in patients with stroke who have no other evidence of heart
disease is patent foramen ovale with or without atrial septal aneurysm.
Treatment implications are problematic (see below). ECG abnormalities are extremely common in patients with SAH. However, the clinical
relevance of these abnormalities is questionable since they often do not
correlate with echocardiographic abnormalities, histopathologic abnormalities, or serum markers of cardiac injury. Approximately 20% of
patients with SAH have elevated serum troponin-I levels. Patients with
elevated troponin-I levels should undergo echocardiography, as elevated
troponin-I levels have been shown to be 100% sensitive and 86% specific
for the detection of left ventricular dysfunction by echocardiography.15
Cerebral arteriography provides high-resolution images of both extracranial and intracranial vessels, which may be useful occasionally in the
identification of causes of cerebral infarction such as arterial dissection.
It is of little value for the diagnosis of isolated cerebral vasculitis due to
the high prevalence of both false-positive and false-negative findings.16

Magnetic resonance arteriography (MRA), often overestimates the
degree of stenosis, sometimes even portraying normal vessels as
abnormal. In addition, MRA lacks the high resolution of conventional
arteriography and cannot be used to exclude small aneurysms or abnormalities in distal arterial branches. In contrast, magnetic resonance
venography has supplanted conventional catheter angiography for the
detection of sagittal and lateral sinus venous thrombosis. In hypertensive
patients with lobar intracerebral hemorrhage and in nonhypertensive
patients with intracerebral hemorrhage in any location, arteriography
may demonstrate vascular malformations or aneurysms.2 CT angiography is almost as sensitive as arteriography for detecting causes of intracerebral hemorrhage but will occasionally miss a small arteriovenous
malformation or fistula.17-19 Cerebral arteriography plays an important
role in the evaluation of the patient with SAH by confirming the existence of an aneurysm and providing the necessary information to plan a
surgical approach. If CT or lumbar puncture demonstrates SAH, a fourvessel angiogram should be performed as soon as possible. A complete
study is necessary to look for multiple aneurysms. If arteriography does
not reveal a cause for SAH, it should be repeated in 1 to 2 weeks.
Doppler ultrasound of the carotid arteries is useful to screen for severe
carotid stenosis at the cervical bifurcation in patients who are candidates
for carotid endarterectomy. It is important to remember that the reliability of this technique varies from center to center. Patients with transient
ischemic attacks (TIAs) or mild stroke who are good surgical candidates
should be evaluated immediately since the risk of stroke following TIA
can be as high as 1 in 20 within the first 2 days.20 On the other hand, in
patients with a completed stroke, there is usually no urgency in obtaining this information since carotid endarterectomy does not play a role
in the management of acute stroke. Transcranial Doppler (TCD) studies
can detect stenosis of intracranial vessels, but the value of this information in management decisions remains to be demonstrated.21 TCD can
also detect increases in flow velocity in most patients with arteriographic
vasospasm following SAH (see below).
The value of regional cerebral blood flow (CBF) measurements with
positron emission tomography (PET), single photon emission computed tomography (SPECT), CT, or MRI in the diagnosis and treatment

1/23/2015 12:55:32 PM

CHAPTER 84: Cerebrovascular Disease

of patients with cerebrovascular disease remains to be demonstrated.
Since the diagnosis of cerebral infarction can be made reliably by means
of the clinical picture and a CT scan, it is rarely if ever necessary to
demonstrate a defect on a CBF study. Furthermore, other conditions
also may produce focal regional reductions of CBF. CBF measurement
as an adjunct in deciding the appropriate therapeutic intervention in
patients with stroke has not been shown to result in improved outcome.
The combination of diffusion weighted imaging (DWI) and perfusion
weighted imaging (PWI) in patients with acute ischemic stroke often
reveals a central area of restricted diffusion surrounded by a larger area
of low perfusion. The diffusion abnormality increases with time and its
final boundaries correspond closely to the eventual infarct. These observations have led to the hypothesis that the area of perfusion-diffusion
mismatch indicates tissue destined for infarction that may be salvaged
by thrombolytic therapy. As of 2012, several clinical trials all have failed
to demonstrate that treatment that decisions based on DWI-PWI magnetic resonance scans lead to better patient outcome.22

TREATMENT

■■CEREBRAL INFARCTION

Immediate supportive care of the patient with cerebral infarction requires
attention to the patient’s airway, breathing, and circulation. Although
most patients have preserved pharyngeal reflexes, those with brain stem
infarction or depressed consciousness may require intubation for airway
protection. Coexisting heart and lung disease are common. Respiratory
and cardiac function should be assessed fully, and appropriate interventions performed to maintain perfusion and oxygenation. The use of supplemental inspired oxygen is rational only if the arterial oxygen content of
the blood is decreased; routine use does not reduce mortality.23 At the time

of hospital admission, some patients may have mild intravascular volume
depletion. In addition to normal maintenance requirements, careful fluid
supplementation may be required. The composition of intravenous fluid
(normal saline solution, one-half normal saline solution, or 5% glucose)
makes no difference as long as serum electrolyte concentrations remain
normal. Care should be taken to avoid hypo-osmolarity, which potentially could exacerbate brain edema. Early treatment of hyperglycemia to
achieve levels <7 mmol/L does not improve outcome.24 Systemic arterial
hypertension is common following acute ischemic stroke. In most cases,
blood pressure returns to baseline levels without treatment in a few days.
There are no known hazards to the brain from this spontaneous transient
elevation in systemic blood pressure. The value of treatment, if any, is
unknown. Case reports describe sudden neurological deterioration when
blood pressure is pharmacologically reduced.25 In patients with systolic
blood pressures of 160 to 200, a randomized trial has demonstrated that
pharmacological reduction of systolic pressure by 20 to 25 mm Hg within
the first 24 hours is safe as it did not cause more early neurological deterioration when compared to the natural decrease of 10 to 15 mm Hg, but
neither did it improve death or dependency at 2 weeks.26 There are insufficient data to permit designation of any target blood pressure levels as
effective.27,28 Continuing or stopping preexisting antihypertensive therapy
for 2 weeks after acute ischemic stroke does not affect outcome.29 When
systemic hypertension causes organ damage elsewhere (eg, myocardial
ischemia, congestive heart failure, or dissecting aortic aneurysm), careful
and judicious lowering of the blood pressure with constant monitoring of
neurologic status is indicated.
No clinical evidence or pathophysiologic rationale supports routine
restriction to bedrest for patients with acute brain infarction. Prolonged
immobility carries an increased risk of iliofemoral venous thrombosis,
pulmonary embolism, and pneumonia. Patients should be out of bed
and walking as soon as possible after a stroke. Occasionally, orthostatic
hypotension with worsening of neurologic deficits will occur. In these
cases, a more gradual program of ambulation should be instituted. In

hemiplegic patients, subcutaneous low-dose heparin or enoxaparin
should be administered to prevent iliofemoral venous thrombosis.30

section06.indd 773

773

Alternating pressure antithrombotic stockings may provide benefit as
well. In the case of pulmonary embolism or deep venous thrombosis, full
anticoagulation with heparin or heparin-like drugs may be instituted.
Fever may occur due to infection or other systemic causes. Central
fevers due to hypothalamic disease are an exceedingly uncommon event
and the search for other causes should be vigorously pursued. Animal
studies have shown that even minor elevations in temperature of a few
degrees poststroke can lead to worse brain damage. Maintaining normothermia through the use of antipyretics and cooling blankets makes
good sense but is of unproven value. Trials of induced hypothermia with
both external and internal cooling are now underway. It is important
to remember that dysphagia occurs commonly, even with unilateral
hemispheric lesions. Before oral feeding is instituted, each patient’s
ability to swallow should be carefully checked. Institutions with formal
dysphagia screening protocols have a reduced incidence of pneumonia.31
Incontinence is also common following acute stroke but the use of Foley
catheters should be kept to a minimum because of the attendant increase
in urinary tract infections. Careful attention must be given to the prevention of decubitus ulcers in bedridden patients.
Intravenously administered t-PA improves outcome in carefully
selected patients with acute ischemic stroke when instituted within
4.5 hours of onset.5,6 These findings were demonstrated in two separate
studies: the NINDS Trial comprising patients within 0 to 3 hours of
onset and the ECASS III Trial comprising patients within 3 to 4.5 hours
of onset. Inclusion and exclusion criteria used in these trials were different and are listed in Tables 84-1 and 84-2. In both trials, patients

  TABLE 84-1    Inclusion and Exclusion Criteria From the NINDS t-PA Stroke Trial
Inclusion criteria
1. Age 18 through 80 years.
2.Clinical diagnosis of ischemic stroke causing a measurable neurologic deficit, defined as
impairment of language, motor function, cognition, and/or gaze or vision, or neglect.
Ischemic stroke is defined as an event characterized by the sudden onset of an acute
focal neurologic deficit presumed to be due to cerebral ischemia after computed
tomography (CT) has excluded h­ emorrhage.
3. Time of onset well established to be less than 180 minutes before treatment would begin.
4.Prior to treatment, the following must be known or obtained: complete blood cell count,
platelet count, prothrombin time (if the patient has a history of oral anticoagulant therapy in
the week prior to treatment initiation), partial thromboplastin time (if the patient has received
heparin within 48 hours of treatment initiation), blood glucose, and CT scan (noncontrast).
Exclusion criteria
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
1 1.
12.
13.
1 4.
15.

Minor stroke symptoms or major symptoms that are improving rapidly.
Evidence of intracranial hemorrhage on CT scan.
Clinical presentation that suggests subarachnoid hemorrhage even if initial CT scan is normal.
Female patient who is lactating or known or suspected to be pregnant.
Platelet count less than 100,000/µL; prothrombin time greater than 15 seconds;
heparin has been given within 48 hours and partial thromboplastin time is greater
than the upper limit of normal for laboratory; anticoagulants currently being given.
Major surgery or serious trauma, excluding head trauma, in the previous 14 days, or
head trauma within the previous 3 months.
History of gastrointestinal or urinary tract hemorrhage in the previous 21 days.
Arterial puncture at a noncompressible site or a lumbar puncture within the previous 7 days.
On repeated measurement, systolic blood pressure >185 mm Hg or diastolic blood pressure
>110 mm Hg at the time treatment is to begin, or patient requires aggressive treatment
to reduce blood pressure to within these limits.
Patient has had a stroke in the previous 3 months or has ever had an intracranial hemorrhage
considered to put the patient at an increased risk for intracranial hemorrhage.
Serious medical illness likely to interfere with this trial.
Abnormal blood glucose (<50 or >400 mg/dL).
Clinical presentation consistent with acute myocardial infarction or suggesting
­postmyocardial infarction pericarditis.
Patient cannot, in the judgment of the investigator, be followed for 3 months.
Seizure occurred at onset of stroke.

1/23/2015 12:55:32 PM


774

PART 6: Neurologic Disorders

  TABLE 84-2    Inclusion and Exclusion Criteria From ECASS III
Inclusion criteria
1. Acute ischemic stroke.
2. Age 18 to 80 years.
3. Onset of stroke symptoms 3 to 4.5 hours before initiation of study-drug administration.
4.Stroke symptoms present for at least 30 minutes with no significant improvement
before treatment.
Exclusion criteria
1.
2.
3.
4.
5.
6.
7.
8.
9.
1 0.
1 1.
12.
13.
14.
15.

Intracranial hemorrhage.
Time of symptom onset unknown.
Symptoms rapidly improving or only minor before start of infusion.
Severe stroke as assesses clinically (NIHSS >25) or by imaging (involving more than
one-third of middle cerebral artery territory).

Seizure at the onset of stroke.
Stroke or serious head trauma, within the previous 3 months.
Combination of previous stroke and diabetes mellitus.
Administration of heparin within the 48 hours preceding the onset of stroke, with
activated partial thromboplastin time at presentation exceeding the upper limit of the
normal range.
Platelet count of less than 100,000/mm3.
Systolic pressure greater than 185 mm Hg or diastolic pressure greater than
110 mm Hg, or aggressive treatment (intravenous medication) necessary to reduce
blood pressure to these limits.
Blood glucose less than <50 mg/dL or >400 mg/dL.
Symptoms suggestive of subarachnoid hemorrhage even if CT scan was normal.
Oral anticoagulant treatment.
Major surgery or severe trauma within the previous 3 months.
Other major disorders associated with an increased risk of bleeding.

received 0.9 mg/kg (90 mg maximum) of alteplase, 10% given as an
initial bolus over 1 minute, followed by a continuous intravenous infusion of the remainder over 60 minutes. The infusion was discontinued
if intracranial hemorrhage was suspected. In the NINDS 0- to 3-hour
trial, all patients were admitted to a neurology special care area or ICU.
Anticoagulant or antiplatelet drugs were not allowed for 24 hours.
Nasogastric tubes and Foley catheters were avoided for 24 hours if
­possible. Blood pressure was monitored every 15 minutes for 2 hours,
every 30 minutes for 6 hours, and then every 60 minutes for 16 hours.
Blood pressure was kept below 180/105 mm Hg with labetalol or sodium
nitroprusside. Symptomatic cerebral hemorrhage occurred more commonly in the group treated with t-PA (6%) than in the control group
(<1%). Recommended treatment of symptomatic intracerebral hemorrhage included cryoprecipitate and platelet transfusion.32 In spite of
this treatment, mortality at 3 months from ICH after t-PA was 75% in
the NINDS trial.33 Even taking into account the increased risk of intracerebral hemorrhage, there was no difference in mortality, and more
t-PA-treated patients demonstrated an excellent neurologic outcome at

3 months by each of four separate outcome scales. The odds ratio for
a favorable outcome due to treatment was 1.7. In the ECASS III 3- to
4.5-hour trial, anticoagulant or antiplatelet drugs were also not allowed
for 24 hours with the exception that subcutaneous heparin (≤10,000 IU)
or equivalent doses of low-molecular weight heparin was permitted for
prophylaxis against deep-vein thrombosis. The odds ratio for a favorable
outcome due to treatment was 1.3. Supporting evidence for these two
pivotal trials is provided by retrospective analyses of small subgroups of
patients enrolled <4.5 hours postevent in other trials.34,35
Even though efficacy of IV t-PA has been demonstrated out to
4.5  hours, eligible patients should be treated as soon as possible since
the benefit is time-dependent.36 For patients who awaken from sleep
with a stroke, the time of onset must be taken to be the last time they
were awake and known to be in their premorbid state, not the time of
awakening. If the time of stroke onset cannot accurately be established
to be less than 4.5 hours, intravenous t-PA should not be given. Several

section06.indd 774

controlled clinical trials failed to demonstrate a benefit of intravenous
t-PA after 4.5 hours, even when magnetic imaging criteria are used to
select patients.37-40
The clinical value of any intra-arterial pharmacological or mechanical
revascularization therapy for acute ischemic stroke has not been demonstrated. A trial of intra-arterial pro-urokinase plus intravenous heparin
within 0 to 6 hours after onset in patients with middle cerebral artery
occlusion showed a barely statistically significant benefit over intravenous
heparin alone. These data were not sufficient proof for the drug to be
approved for use in the United States.41 A trial of intra-arterial urokinase
within 0 to 6 hours of onset in middle cerebral artery occlusion showed
no benefit.42 In neither of these studies was intravenous t-PA administered to any of the estimated 70% of the control groups who could have

received it within 4.5 hours after onset.43 Consequently, the superiority of
the intra-arterial to the intravenous approach in those who are eligible for
IV t-PA within 4.5 hours has not been shown. Data to show efficacy for
those who are ineligible for IV t-PA has not been published. There are no
controlled clinical trials of intra-arterial therapy with other thrombolytic
drugs, including t-PA. Several mechanical devices have been approved
by the United States Food and Drug Administration for intra-arterial use
in acute ischemic stroke based on trials that showed at least equivalent
performance to previous devices in removing thrombus and restoring
­arterial patency. Although these devices were tested in patients up to
8 hours after stroke onset, no trials included a medical control group so
clinical benefit has never been demonstrated.44
Two large studies have shown that 160 or 300 mg/d of aspirin begun
within 48 hours of the onset of ischemic stroke results in a net decrease
in further stroke or death of 9/1000.45 Data from many randomized
controlled trials have shown that full anticoagulation with heparin,
low-molecular-weight heparins, or heparinoids in patients with acute
ischemic stroke provides no net short- or long-term benefit in general
or in any subgroup, including those with atrial fibrillation or other cardioembolic sources.14,30,46-48 Ticlopidine, clopidogrel, and the combination of low-dose aspirin and extended-release dipyridamole (Aggrenox)
all have been demonstrated to be modestly effective in the long-term
prevention of recurrent ischemic stroke, but there are no data regarding
their value during the acute period.49 Many drugs aimed at ameliorating
ischemic neuronal damage in patients with acute stroke have undergone
clinical trials with none showing a benefit. Physicians treating patients
with acute ischemic stroke should be aware of the results of these trials
on an ongoing basis.
Cerebral edema is the major cause of early mortality following cerebral infarction. Mannitol and hyperventilation can temporarily reduce
intracranial pressure. They may be of value to the patient with brain
stem compression from an edematous cerebellar infarct for which
craniotomy and removal of the edematous tissue may be lifesaving.

Hyperosmolar therapy (mannitol or hypertonic saline), hypothermia,
and hemicraniectomy are sometimes used to treat massive edema
from hemispheric infarction. The value of the first two treatments is
unproven. Recent studies have shown that hemicraniectomy can significantly reduce mortality in patients with large hemispheric infarcts and
depressed level of consciousness who are operated on within 48 hours
of stroke onset.50,51
Specific causes of cerebral infarction may require specific definitive
treatments, such as exchange transfusions for cerebral infarction due to
sickle cell anemia. Cerebral venous thrombosis can present a particularly
difficult situation because of the presence of hemorrhage. While two
small controlled trials have demonstrated that anticoagulation is safe
even in patients with hemorrhagic infarction, design issue preclude any
conclusions about efficacy.52,53 Patent foramen ovale (PFO) is detected
commonly in patients with ischemic stroke and is often the only abnormality found. Based on this finding, it is often concluded that the cause
of stroke is paradoxical embolization from deep venous thrombosis.
However, in contrast to pulmonary embolization, it is unusual to find a
deep venous source in these patients. The risk of recurrent stroke is low
and anticoagulation with warfarin does not reduce the risk of long-term

1/23/2015 12:55:33 PM


775

CHAPTER 84: Cerebrovascular Disease

recurrence.54,55 Studies of acute anticoagulation are not available. Acute
anticoagulation of spontaneous or traumatic dissections of the carotid or
vertebral arteries is often recommended. Data to support this approach
are derived only from small nonrandomized, nonblinded studies, and

even these data are weak.56

■■INTRACEREBRAL HEMORRHAGE

Supportive care of patients with primary intracerebral hemorrhage
(ICH) requires attention to the same basic factors as for patients
with cerebral infarction. Any underlying coagulopathy should be corrected as rapidly as possible. No randomized trials on management of
warfarin-associated ICH have been carried out. Prothrombin complex
concentrates, recombinant factor VIIa, and fresh frozen plasma alone
or in combination have all been recommended.57 Fresh frozen plasma
administration may cause pulmonary edema.58 Early use of Factor VIIa
in patients with normal hemostasis resulted in a small reduction in clot
expansion but no difference in clinical outcome.59 Prophylaxis for deep
venous thrombosis with low-dose subcutaneous heparin or heparinoids
may be instituted safely on or after the second day posthemorrhage and
reduces subsequent deep venous thrombosis if begun before day 4.60,61
Systemic blood pressure is often elevated acutely, sometimes to very
high levels. In patients with systolic blood pressure of 150 to 220 mm Hg,
a randomized trial has demonstrated that rapid pharmacological reduction of systolic pressure by 27 mm Hg within the first hour was safe in
that it resulted in equivalent clinical outcomes when compared to a
lesser decrease of 13 mm Hg.62 There are insufficient data to permit designation of any target blood pressure levels as effective.27,63
Clinically evident seizures are more common with lobar ICH compared to basal ganglia hemorrhage.64 Prolonged electroencephalographic
monitoring shows electrical epileptiform events without motor convulsions in 20% to 30% of patients with acute ICH.65,66 The value of treating
the electrographic events is under study. Prophylactic anticonvulsant
treatment does not prevent seizures and may worsen outcome.67,68
The value of ICP monitoring and treatment remains unknown.
Neither mannitol nor corticosteroids reduce morbidity and mortality.69
Although the area of perihematomal edema on CT or MRI increases in
the several weeks following ICH, this growth is not associated with early
clinical deterioration or worse eventual outcome.70-72 Ventriculostomy

is of unproven value as observational studies have shown no benefit.73,74
The efficacy of ventriculostomy in combination with instillation of
thrombolytic drugs is currently under study in patients with intraventricular hemorrhage.75
The value of surgery is best accepted for cerebellar hemorrhages
resulting in brain stem compression, although no data other than anecdotal reports are available. Ideally such surgical intervention should be
undertaken before brain stem damage occurs. Patients with small cerebellar hematomas (<2 cm) may do well without surgical intervention, or
simply with ventricular drainage for hydrocephalus. Those with larger
cerebellar hematomas usually undergo surgical evacuation, although
no prospectively validated criteria for the necessity and the timing of
cerebellar hematoma evacuation are available. Multiple randomized
controlled trials of patients with supratentorial ICH, either superficial
or deep, have shown no benefit from craniotomy and clot evacuation.76

HEMORRHAGE DUE TO RUPTURED
■■SUBARACHNOID
INTRACRANIAL ANEURYSM

Aneurysmal SAH remains a devastating neurologic problem, with a
mortality rate of up to 45% within the first 30 days. Of those patients
that survive, more than half are left with neurologic deficits as a result
of the initial hemorrhage or delayed complications. SAH presents the
intensivist with a unique and challenging series of management issues.
SAH usually presents as an acute neurologic event that is frequently
followed by a series of processes leading to delayed central nervous
system and systemic complications. Patients who are minimally
affected by the initial hemorrhage can, over the course of hours to

section06.indd 775

weeks, deteriorate due to rebleeding, hydrocephalus, or delayed ischemic deficits caused by vasospasm. Management can be complicated

by spontaneous volume contraction, cardiac and pulmonary dysfunction, electrolyte abnormalities, infections, and a catabolic state. The
treatment team should include neurosurgeons, radiologists, anesthesiologists, intensivists, and nurses experienced in the management of
SAH patients. Because of the complicated nature of their surgical and
medical management, SAH patients are best cared for in centers that
specialize in this care.
The management of patients following rupture of intracranial aneurysms has changed significantly over the past decades. The calcium
channel blocker nimodipine is now routinely used to reduce the impact
of vasospasm. Attempts at early obliteration of the ruptured aneurysm
with surgical clipping or endovascular placement of detachable coils
within the aneurysm have become routine. Hemodynamic augmentation is now the cornerstone of the management of vasospasm with
adjunctive endovascular treatment employed in selected cases. New
and promising therapies that specifically target the underlying cause or
direct effects of cerebral vasospasm are currently under investigation.77
Initial Stabilization and Evaluation:  Initial evaluation should assess airway, breathing, circulation, and neurologic function. Patients with a
diminished level of consciousness often have impaired airway reflexes.
In general, patients with a Glasgow Coma Scale score of 8 or less should
be intubated. This should be performed under controlled conditions by
experienced personnel using a rapid sequence protocol. Premedication
with short-acting agents such as propofol or etomidate should be used
to prevent elevations in blood pressure (BP) with tracheal stimulation in
order to minimize the risk of rebleeding.
As soon as the patient is stabilized, a complete neurologic examination, head CT, and, if indicated, lumbar puncture should be performed.
Patients are graded on the basis of clinical and radiographic criteria. The
two common clinical grading scales that are predictive of outcome are
the Hunt-Hess scale and the World Federation of Neurological Surgeons
scale (Table 84-3). The Fisher Scale is based on the amount of blood
­visible on CT scan and is predictive of cerebral vasospasm.78

  TABLE 84-3   The Hunt-Hess, the World Federation of Neurologic Surgeons,
and the Fisher Scales

Hunt-Hess Scale
Grade

Criteria

I

Asymptomatic or mild headache

II

Moderate to severe headache, nuchal rigidity, with or without cranial nerve deficits

III

Confusion, lethargy, or mild focal symptoms

IV

Stupor and/or hemiparesis

V

Comatose and/or extensor posturing

World Federation of Neurologic Surgeons Scale
Grade

Glasgow Coma Scale Score

Motor Deficits

I

15

Absent

II

14-13

Absent

III

14-13

Present

IV

12-7

Present or absent

V

6-3

Present or absent

Fisher Scale (Based on Initial CT Appearance and Quantification of Subarachnoid Blood)
1. No subarachnoid hemorrhage on computed tomography
2. Broad diffusion of subarachnoid blood, no clots and no layers of blood greater than 1 mm thick
3. Either localized blood clots in the subarachnoid space or layers of blood greater than 1 mm thick
4. Intraventricular and intracerebral blood present, in absence of significant subarachnoid blood

1/23/2015 12:55:33 PM


776

PART 6: Neurologic Disorders

Routine management of SAH patients frequently include anticonvulsants and prophylaxis against deep vein thrombosis (DVT) in addition
to close neurologic and cardiopulmonary monitoring to detect the early
complications of hypertension, rebleeding, acute hydrocephalus, pulmonary edema, cardiac arrhythmias, and left ventricular dysfunction.
Seizures can be a presenting symptom of SAH; however, the incidence
of recurrent or new events after hospitalization is low.79 It remains
unclear as to the effect of seizures on the clinical course of patients with
aneurysmal subarachnoid hemorrhage although the use of prophylactic
anticonvulsants has been associated with poor neurological and cognitive outcomes.80 The use of anticonvulsants in the perioperative period
or until definitive treatment of the aneurysm is supported albeit by
retrospective data.81,82 DVT is common in patients with SAH, therefore
prophylaxis is mandatory. The use of pneumatic compression devices is
preferred initially because of the risk of intracranial bleeding, however,
once the aneurysm has been treated, prophylaxis with subcutaneous
heparin or low-molecular weight heparin is generally considered safe.
Routine treatments aimed at reducing the risk of ischemic stroke

secondary to vasospasm include preventing hypovolemia and administering nimodipine. Patients should be hydrated with isotonic saline
at 1.5 to 2 mL/kg/h and indicators of volume status should be monitored closely (clinical exam, fluid balance, daily weights, laboratory
values, and in select cases invasive hemodynamic measurements).
Prophylactic hypervolemia should be avoided as it has not been
shown to be beneficial and may in fact lead to increased medical
­complications.83,84 Several large, prospective, placebo-controlled studies have demonstrated that nimodipine reduces the incidence and
severity of delayed ischemic deficits and improves outcome in SAH.85
It remains uncertain whether this drug acts by causing vasodilation or
by exerting direct neuroprotective effects. The recommended dose is
60 mg every 4 hours for 21 days from the time of hemorrhage. At this
dose, nimodipine can sometimes reduce systemic BP, an effect that is
undesirable in patients with vasospasm (see below). This effect can
be ameliorated by increasing fluid administration and by altering the
dose to 30 mg every 2 hours; however, pharmacologic blood pressure
­support is necessary in some patients.
Early Complications 
Hypertension  Elevated BP often initially accompanies acute SAH. Several
factors may contribute to an increase in BP, including headache, elevated
ICP in patients with hydrocephalus, increased sympathetic nervous
system activity, and preexisting hypertension. The rationale for treating
hypertension is to reduce the risk of aneurysmal rebleeding. There are
few compelling reasons not to treat the elevated BP before the onset of
vasospasm. As definitive data on optimal BP are lacking, it seems prudent to take the patient’s usual BP as a target. When the patient’s usual
BP is not known, it is probably better to overtreat than to undertreat.
There is one important exception—comatose patients in whom CT
shows marked hydrocephalus. In such cases BP should be treated very
cautiously until the ICP is known, to avoid causing a critical reduction
in cerebral perfusion pressure. In patients who present several days after
hemorrhage and are at risk for vasospasm, the appropriate management
of hypertension is less clear. The benefit of preventing rebleeding must

be weighed against the risk of worsening neurologic symptoms by lowering BP in the presence of vasospasm.
The first step in treating elevated BP is to administer a short-acting
analgesic such as fentanyl as pain can be the sole cause of BP elevation.
Patients are routinely given nimodipine to prevent vasospasm, and it
alone may be adequate to control BP. Otherwise, short-acting agents
are preferred, since BP may be labile. Labetalol administered as intermittent intravenous boluses is frequently used, since it appears to have
little effect on ICP and is easily titrated. Other useful agents include
intravenous hydralazine and enalapril. If frequent intravenous boluses
are required, one should consider starting a continuous intravenous
infusion of an antihypertensive agent. Nicardipine is ideal as it is short
acting, can be titrated every 5 to 15 minutes, does not require invasive hemodynamic monitoring, and has been shown to be safe in this

section06.indd 776

patient population.86 Sodium nitroprusside is usually avoided because
of its tendency to increase ICP and thus reduce the cerebral perfusion
pressure.
Rebleeding  Rebleeding is most common in the first 24 hours after the
initial hemorrhage. The cumulative risk after 1 week is ~20%, and the
risk remains elevated for several weeks.87 About one-half of patients
who rebleed will die. Measures employed in the hope of preventing
rebleeding include avoidance of hypertension, cough, the Valsalva
maneuver, and excessive stimulation. Treatment may involve the administration of antitussives, stool softeners, and sedatives when indicated.
Antifibrinolytic medications can reduce the risk of rebleeding, but do
so at the cost of an increased incidence of cerebral ischemia.88 With the
increasingly wide use of early surgery, the use of antifibrinolytics has
largely been abandoned.
The timing of surgical obliteration of the aneurysm has changed
considerably. Up to the 1970s, surgery was routinely delayed because
of reluctance to operate on an edematous brain. Several factors have

resulted in a shift to early surgery (days 1-3) for patients who have a
grade of I to III on the Hunt-Hess scale. These include improved surgical
techniques, better results with early surgery in North America,89 and the
necessity that the aneurysm be clipped before hypertensive therapy for
vasospasm is administered. The timing of surgery in poor-grade patients
(Hunt-Hess grades IV or V) remains controversial, but early surgery is
routinely performed in some centers.90
In the past decade, the role of endovascular repair of amenable
ruptured and unruptured aneurysms has become widespread and the
standard of care at many institutions. Electrolytically detachable coils
can be placed directly in the aneurysm, where they induce thrombosis.
In a recent multicenter randomized trial, 20% of all assessed patients
had a ruptured aneurysm that was considered to be amenable to treatment with either surgical clipping or endovascular coiling. Among this
subgroup of patients (predominantly of good clinical grade with small
ruptured aneurysms of the anterior circulation), the risk of death or
dependency at 1 year was significantly lower with endovascular coiling.91
Follow-up data for an average of 9 years have demonstrated the continued efficacy in this patient population. There was a small increased
rerupture rate among the patients treated with coiling; however, the risk
of death remained significantly lower at 5 years.92
Acute Hydrocephalus  Acute hydrocephalus can develop very quickly after
SAH. It is most common in patients with a poor neurological grade on
admission and higher Fisher Scale scores. The hallmark of symptomatic hydrocephalus is a diminished level of consciousness, sometimes
accompanied by downward deviation of the eyes and poorly reactive
pupils. The diagnostic evaluation can be complicated if the patient
has received sedative drugs; it is important that analgesics be administered in doses that provide adequate relief from pain, but not excessive
­sedation. If sedatives are required for agitated patients, judicious administration of short-acting agents is prudent.
Hydrocephalus can be diagnosed reliably with CT and treated
effectively with external ventricular drainage. Since less than half of
patients with CT evidence of hydrocephalus will deteriorate clinically,
ventriculostomy is typically reserved for patients with a diminished level

of consciousness.
Cardiac Complications  Cardiac arrhythmias and electrocardiographic abnormalities are common in the first 24 to 48 hours after SAH. Most arrhythmias are benign and include atrial fibrillation and atrial flutter. More
serious arrhythmias include supraventricular and rarely ventricular
tachycardia and are associated with electrolyte abnormalities such as
hypokalemia. Mild elevations in cardiac enzymes also commonly occur
however the significance of these elevations is not clear.
A significant number of patients will have some degree of ventricular
dysfunction; however, those at highest risk for neurogenic stunned myocardium are of a high clinical grade. Neurogenic stunned myocardium
is characterized by diffuse T-wave inversions, moderate elevations in
troponin-I, pulmonary edema, cardiogenic shock, and reversible left

1/23/2015 12:55:34 PM


CHAPTER 84: Cerebrovascular Disease

ventricular wall abnormalities. The management of these patients typically requires invasive hemodynamic monitoring and treatment with
ionotropes such as dobutamine. If clinical vasospasm develops in these
patients prior to the resolution of cardiogenic shock, management can
become very difficult.
Pulmonary Complications  Pulmonary complications are seen in almost onefourth of all patients with SAH.93 They include pneumonia (arising from
acute or subacute aspiration, commonly with nosocomial ­organisms),
cardiogenic pulmonary edema, neurogenic pulmonary edema, and
pulmonary embolism. Management of severe pulmonary edema with
refractory hypoxia usually involves positive pressure ventilation and
diuretics; however, diuretics may not be appropriate for neurogenic
pulmonary edema if there is relative intravascular volume depletion.94 In
these cases, hemodynamic monitoring via a pulmonary artery catheter
or via transpulmonary thermodilution may be warranted.
Postoperative Management:  Knowledge of the intraoperative surgical

and anesthetic course facilitates the postoperative care of SAH patients.
Large doses of mannitol may have been administered to shrink the
brain and facilitate retraction. This measure can result in postoperative
hypovolemia. If temporary clipping of cerebral vessels was required,
hypothermia and/or large doses of barbiturates may have been employed
and the risk of focal ischemia exists. These maneuvers may also delay
emergence from anesthesia and add to the systemic complications of
hypothermia. The decision to extubate a postoperative patient must
take these factors into consideration with the understanding that keeping the patient on mechanical ventilation further increases their risk for
medical complications including ventilator-associated pneumonia. If the
aneurysm is successfully treated, many practitioners will accept higher
blood pressures in the postoperative period in anticipation of vasospasm
(see below).
Hyponatremia and Intravascular Volume Contraction:  A total of 30% to
50% of SAH patients develop intravascular volume contraction and a
negative sodium balance (referred to as cerebral salt wasting) when given
volumes of fluids intended to meet maintenance needs. Low intravascular
volume is associated with symptomatic vasospasm and must be avoided.
Hyponatremia develops in 10% to 34% of patients following SAH.
Administration of large volumes (5-8 L/d) of isotonic saline prevents
hypovolemia, but patients may still develop hyponatremia. The degree of
hyponatremia appears to be related to the tonicity rather than the volume
of fluids administered.95 Thus, administration of large ­volumes of isotonic
saline and restriction of free water are usually effective at limiting hyponatremia and preventing hypovolemia. In SAH patients with hyponatremia,
the volume of fluids should never be restricted; instead only free water
intake should be limited. Hypertonic saline solutions and fludrocortisone
may be required in severe or refractory cases.
Vasospasm:  The term vasospasm was originally used to refer to segmental or diffuse narrowing of large conducting cerebral vessels. Recently,
this term has taken on multiple meanings. It may refer to angiographic
findings, to increased transcranial Doppler velocities, or to delayed

ischemic deficits. Angiographic and transcranial Doppler vasospasm
occurs in 60% to 80% of patients, whereas clinical vasospasm (or delayed
ischemic deficit) occurs in 20% to 40% of patients.
The pathogenesis of vasospasm is complex. Several molecular
mechanisms that are involved in the development of vasospasm have
been described in animal models and confirmed in human samples
including inflammation, the presence of degradation blood products,
nitric oxide signaling, and calcium signaling.96 All of these mechanisms
appear to be time-dependent as these pathological changes develop in
a delayed fashion after exposure to subarachnoid blood and are selflimited. In addition to changes in the large conducting cerebral vessels
that traverse the subarachnoid space, small-vessel reactivity may be
impaired as well.
Monitoring for Vasospasm  Serial neurologic assessments are essential in monitoring for vasospasm. These must be performed frequently by physicians

section06.indd 777

777

and nurses well-versed in the neurologic examination and recognition
of subtle deficits. The patients with the highest incidence of vasospasm
are those with Hunt-Hess grades III through V and Fisher Scale of 3.
These patients are often monitored in the ICU (days 3-10). Clinically
vasospasm presents as a decline in the global level of function or a focal
neurologic deficit. Patients may initially appear “less bright” and then
become progressively less alert and finally comatose. The focal deficits
mimic those seen in ischemic stroke. Middle cerebral artery vasospasm
can produce hemiparesis, and if left-sided, aphasia or if right-sided,
neglect. Anterior cerebral artery vasospasm often manifests as abulia or
lower extremity weakness. The focal deficits wax and wane and therefore are not reported by all observers. The symptoms are exacerbated by
hypovolemia or hypotension.

Transcranial Doppler ultrasonography detects changes in the blood
flow velocity in the proximal portion of the major cerebral vessels. Very
high flow velocities (>200 cm/s) in the middle cerebral and intracranial
carotid arteries are closely correlated with angiographic vasospasm,
while low flow velocities (<120 cm/s) suggest a low likelihood of vasospasm. Furthermore, a Lindegaard ratio (MCA/extracranial ICA mean
velocity ratio) which is greater than 6 is also highly predictive of severe
vasospasm.97 Patients with rapidly rising velocities are considered to
be at highest risk for developing clinical vasospasm; therefore, a trend
is frequently more useful than isolated values. Transcranial Doppler
has several limitations. High-flow velocities can be due to increased
blood flow rather than narrowing of the blood vessel; however, this can
be c­orrected for by calculating the Lindegaard ratio instead of using
velocities. Distal segments of the major arteries cannot be evaluated.
The technique is also operator dependent and adequate “acoustic windows” are required. Therefore, transcranial Doppler velocities should
not be used in isolation as an indication for the initiation of aggressive
treatments—the clinical course must be considered as well. Given the
limitations of transcranial Doppler, other imaging modalities have been
explored and further developed. These include CT angiography and CT
perfusion as a recent meta-analysis suggests that these techniques offer
a high diagnostic accuracy.98 The major limitation though is the inability
to intervene which conventional angiography may provide (see below).
Treatment of Vasospasm 
Hemodynamic Augmentation  Hemodynamic augmentation for the treatment
of vasospasm has been referred to as hemodilution hypervolemic hypertensive therapy (“triple H therapy”) or as hypervolemic hypertensive
­therapy (HHT). The pathophysiologic rationale is based on the high
rate of spontaneous hypovolemia, the association of hypovolemia with
delayed ischemic deficits, and the loss of autoregulation of cerebral blood
flow in this population.
Most centers continue aggressive hydration during the period of
vasospasm risk. Some will increase the rate of fluid administration if

transcranial Doppler velocities are rising. The indication for starting
aggressive hemodynamic augmentation is usually the onset of clinical
symptoms of delayed ischemic deficit. Early descriptions of this therapy
emphasized the role of volume expansion, as many of these patients had
not been aggressively hydrated before the onset of symptoms. However, if
intravascular volume has been maintained before the onset of s­ ymptoms,
further volume expansion may not be helpful.83 The optimal intravascular
volume is unknown, and achieving cardiac filling ­pressures that optimize
cardiac output has been advocated.
When symptoms persist despite optimal intravascular volume,
vasoactive drugs are administered, with a goal of either raising mean
arterial pressure (MAP) or augmenting cardiac output in order to
improve cerebral perfusion. In most cases, patients will require monitoring via an arterial line and with either pulmonary artery catheter or
transpulmonary thermodilution hemodynamic monitoring. The most
commonly used agents to increase blood pressure are norepinephrine,
dopamine, and phenylephrine. Caution must be employed when using
dopamine alone, because of a high incidence of tachyarrhythmias.
When using phenylephrine one must be aware that it tends to decrease
cardiac output, especially in those patients with impaired cardiac

1/23/2015 12:55:34 PM


778

PART 6: Neurologic Disorders

function. For augmentation of cardiac output dobutamine titrated to
a goal cardiac index (CI) can augment cerebral perfusion and reverse
neurological deficits.99

When therapy is initiated, the MAP should be raised to 15% to 20%
above baseline rather than to an arbitrary value. If after 1 to 2 hours
the delayed ischemic deficit has not resolved, the MAP should be
raised further. The MAP is increased progressively until the neurologic deficit is completely resolved or the risk of systemic toxicity
becomes unacceptable. Some patients may require a MAP of 150 to
160 mm Hg to completely reverse the neurologic symptoms. For cardiac output augmentation, dobutamine should be titrated to a goal CI
of at least 3.5 L/min/m2 and titrated further as needed to reverse the
neurological deficits.100 The neurologic status should be reevaluated
several times a day to determine MAP or CI goals. Both approaches
are reported to produce neurologic improvement. It has not yet been
determined whether the optimal therapy is to enhance cardiac output,
MAP, or both.
Once instituted, the therapy is generally continued for 3 to 4 days
before attempts are made to wean the patient from it. Weaning should
be done gradually, with very close monitoring of neurologic status. If
the initial attempt at weaning is unsuccessful, a second attempt should
be made after 1 to 2 days. The patient usually is weaned from vasoactive
drugs first, aggressive hydration being continued for several more days.
Hemodynamic augmentation is not without complications. Early
reports indicated high rates of fluid overload, heart failure, and myocardial ischemia; however, when administered in a closely monitored
setting, even in patients with preexisting cardiac disease it can be done
safely.101 Cardiovascular monitoring should include continuous display
of the electrocardiogram, peripheral oxygen saturation, MAP, and
frequent measurements of cardiac filling pressures and cardiac output.
In patients with a history of ischemic heart disease, daily electrocardiograms and cardiac enzyme measurements may be helpful. Close
monitoring of potassium, magnesium, and phosphate levels is important
because of large losses in the urine.
Endovascular Therapies: Percutaneous Transluminal Angioplasty and Direct Intra-Arterial
Vasodilators  Balloon angioplasty can be used to dilate proximal segments
of intracranial vessels, but it is not well suited for use in the distal vasculature. The dilation achieved appears to be long-lasting. Complications that

have been reported include artery rupture and displacement of aneurysm
clips. In most cases there is clear-cut angiographic improvement, but the
clinical efficacy of angioplasty has not been clearly established.
Direct intra-arterial injection of vasodilators into the vessel affected
by vasospasm has become routine in many centers. The most commonly
used agents currently used include verapamil and nicardipine. While
the radiographic improvement is usually evident, the clinical effect has
been less clear. There have not been any randomized controlled trials
demonstrating a benefit on patient outcome. These therapies are usually
reserved for patients who do not tolerate or do not respond to hemodynamic augmentation.
Other Potential Therapies  Prevention rather than treatment of the consequences of vasospasm would significantly reduce the morbidity, mortality, and cost of SAH. Intracisternal instillation of thrombolytic agents
has been employed in an attempt to dissolve clots around the circle of
Willis and thereby decrease vasospasm. A multicenter, randomized,
blinded, placebo-controlled study found trends toward reduction of
angiographic vasospasm, reduced delayed neurologic worsening, lower
14-day mortality, and improved 3-month outcome that did not achieve
statistical significance in patients treated with intracisternal t-PA.
Patients with thick subarachnoid clots had a significant reduction in the
incidence of severe vasospasm with intracisternal t-PA.102
The degradation of blood deposited during an SAH involves the
conversion of oxyhemoglobin to methemoglobin, which releases an activated form of oxygen that catalyzes free radical reactions, including lipid
peroxide formation. The 21-aminosteroid, tirilazad mesylate, a potent
scavenger of oxygen free radicals, inhibits lipid peroxidation and reduces
vasospasm in animal models. A European-Australian multicenter study

section06.indd 778

showed that tirilazad was associated with better outcomes compared
to control patients, but this was not confirmed in a subsequent North
American study.103,104 In a multicenter, randomized, double-blind,

placebo-controlled trial, nicaraven, a hydroxyl radical scavenger, significantly reduced the incidence of severe vasospasm and poor outcome
at 1 month but not at 3 months.105 Ebselen, another lipid peroxidation
inhibitor, did not lower the incidence of symptomatic vasospasm in a
controlled study.106 Clazosentan, an endothelin receptor antagonist, is
one of the more promising medical treatment options currently in phase
3 clinical trials. A phase 2 study demonstrated a reduction in moderate
to severe vasospasm and clazosentan appeared safe.107 Other potential
therapies being studied include statins, magnesium infusions, nitric
oxide donors, and albumin infusions.108

KEY REFERENCES
•• Anderson CS, Huang Y, Arima H, et al. Effects of early intensive
blood pressure-lowering treatment on the growth of hematoma and perihematomal edema in acute intracerebral hemorrhage: the Intensive Blood Pressure Reduction in Acute Cerebral
Haemorrhage Trial (INTERACT). Stroke. 2010;41:307-312.
•• Dorhout Mees SM, Rinkel GJE, Feigin VL, et al. Calcium antagonists for aneurysmal subarachnoid haemorrhage. Cochrane
Database Syst Rev. 2007:CD000277.
•• Fisher CM, Kistler JP, Davis JM. Relation of cerebral vasospasm
to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery. 1980;6:1-9.
•• Hacke W, Kaste M, Bluhmki E, et al. Thrombolysis with alteplase
3 to 4.5 hours after acute ischemic stroke. N Engl J Med.
2008;359:1317-1329.
•• Jüttler E, Unterberg A, Woitzik J, et al. Hemicraniectomy in older
patients with extensive middle-cerebral-artery stroke. N Engl J
Med. 2014;370:1091-1100.
•• Molyneux A, Kerr R, Stratton I, et al. International Subarachnoid
Aneurysm Trial (ISAT) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: a randomised trial. Lancet. 2002;360:1267-1274.
•• Potter JF, Robinson TG, Ford GA, et al. Controlling hypertension and hypotension immediately post-stroke (CHHIPS): a
randomised, placebo-controlled, double-blind pilot trial. Lancet
Neurol. 2009;8:48-56.
•• Robinson TG, Potter JF, Ford GA, et al. Effects of antihypertensive

treatment after acute stroke in the Continue or Stop Post-Stroke
Antihypertensives Collaborative Study (COSSACS): a prospective, randomised, open, blinded-endpoint trial. Lancet Neurol.
2010;9:767-775.
•• Sandercock PAG, Counsell C, Tseng M-C. Low-molecular-weight
heparins or heparinoids versus standard unfractionated heparin for acute ischaemic stroke. Cochrane Database Syst Rev.
2008:CD000119.
•• The National Institute of Neurological Disorders and Stroke rt-PA
Stroke Study Group: tissue plasminogen activator for acute ischemic stroke. N Engl J Med. 1995;333:1581-1587.
•• Vahedi K, Hofmeijer J, Juettler E, et al. Early decompressive
surgery in malignant infarction of the middle cerebral artery:
a pooled analysis of three randomised controlled trials. Lancet
Neurol. 2007;6:215-222.

REFERENCES
Complete references available online at www.mhprofessional.com/hall

1/23/2015 12:55:35 PM


CHAPTER 85: Seizures in the Intensive Care Unit

CHAPTER

85

Seizures in the Intensive
Care Unit
Katharina M. Busl
Thomas P. Bleck

KEY POINTS
•• Seizures are a relatively common occurrence in the intensive care
unit (ICU), but may be difficult to recognize.
•• Seizures that persist longer than 5 to 7 minutes should be treated
to prevent progression to status epilepticus.
•• Three major factors determine outcome in status epilepticus: type
of seizure, cause, and duration.
•• Electroencephalographic (EEG) monitoring to titrate therapy
should be implemented in seizing patients who do not awaken
promptly after institution of antiepileptics, even if tonic-clonic
motor activity resolves.
•• Lorazepam is a preferred agent for initial treatment, followed by
consideration of additional agents for long-term management or
to “break” status epilepticus.
•• Patients with refractory status epilepticus require intubation,
mechanical ventilation, and aggressive treatment with antiepileptics titrated to the EEG.
•• The underlying cause of the seizure disorder must be sought in
tandem with treatment of the seizure disorder itself.

Seizures are a relatively common occurrence in the ICU, complicating
the course of about 3% of adult ICU patients admitted for nonneurologic
conditions.1 Status epilepticus (SE) may be the primary indication for
admission, or it may occur in any ICU patient during a critical illness.
Seizures are second to metabolic encephalopathy as a cause of neurological complications (28.1%).1 A seizure may be the first indication of a
central nervous system (CNS) complication or the result of overwhelming systemic disease. Seizures in the setting of critical illness are often
difficult to recognize and require a complex diagnostic and management
strategy. Delay in recognition and treatment of seizures is associated
with increased mortality,2 thus the rapid diagnosis of this disorder is
mandatory. Conventionally, status epilepticus referred to a protracted
seizure episode or multiple frequent seizures lasting 30 minutes or

longer. However more recently, revised definitions have suggested to
consider seizures lasting for 5 minutes or longer as status epilepticus,3-5
and newer guidelines define status epilepticus as five minutes or more
of either continuous clinical and/or electrographic seizure activity, or
recurrent seizure activity without recovery between seizures.6
While most seizures will terminate spontaneously within a few
­minutes,5 only half of seizure episodes lasting 10 to 29 minutes will
stop spontaneously7 and aggressive treatment should be administered to
prevent ongoing SE.8

EPIDEMIOLOGY AND OUTCOME
Limited data are available on the epidemiology of seizures in the ICU. A
10-year retrospective study of all ICU patients with seizures at the Mayo
Clinic revealed that 7 patients had seizures per 1000 ICU admissions.8
Our 2-year prospective study of medical ICU patients identified 35 with
seizures per 1000 admissions.1 The incidence of generalized convulsive
SE (GCSE) in the United States is estimated to be up to 195,000 episodes
per year,9 but it is unknown how many of these patients require care in
an ICU. The incidence of SE in the elderly is almost twice that of the
general population.10 Nonconvulsive seizures and NCSE are present
in a large proportion of comatose patients with traumatic brain injury,

section06.indd 779

779

intracranial hemorrhage, sepsis, cardiac arrest, or CNS infection.11-15 In
one series, 8% of hospitalized comatose patients were found to be in
electrographic status epilepticus,15 up to 34% of patients in neurological
ICUs,15 and other series of patients with altered mental status found 37%

to have nonconvulsive seizures.16 Of all patients with status epilepticus,
about 80% have nonconvulsive status epilepticus.17 Seizures are probably
even more frequent in the pediatric ICU, as children in the first year of
life have the highest incidence of SE of any age group studied.8
Table 85-1 summarizes the most common causes of SE in adults and
children in the community. An analysis of 204 cases of SE in Virginia
revealed that the primary etiology in children was infection with fever,
followed by remote symptomatic epilepsy, and subtherapeutic levels of
anticonvulsant drugs. In adults, cerebrovascular disease and low antiepileptic drug levels were the most prevalent causes.8 A recent study
from Brazil found anticonvulsant noncompliance to be the main cause
of SE in patients with a prior history of epilepsy, and CNS infection,
stroke, and metabolic disturbances predominated in the group without
previous seizures.18 A prospective study of neurologic complications in
medical ICU patients determined that two-thirds of patients had a
­vascular, infectious, or neoplastic explanation for their seizures1; metabolic and toxic etiologies are common in the ICU as well. A review of
100 cases of nonconvulsive SE (NCSE) demonstrated that 14% were due
to acute neurologic events, 28% due to acute systemic causes, and 31% due
to epilepsy, with the remainder due to multiple causes or a cryptogenic
­etiology,19 and among patients with NCSE in a comatose state, hypoxia
(42%) and stroke (22%) were the most common etiologies.15 In medical
ICU patients, electrographic seizures or periodic epileptiform discharges
were detected in 22% of patients, with the predominant underlying disease
state being sepsis.13 It is important to realize that the frequency of diagnosing NCSE will rise with implementation of continuous EEG monitoring by
6% to 8% accounting for the increment of investigations.20
A prospective study of neurologic complications in medical ICU
patients showed that having one seizure in the ICU doubled mortality.1 At least 20% of patients with status epilepticus die,21,22 and up to
61% of patients developing SE during hospitalization do not survive.23
SE in and of itself confers a mortality rate of 26% to adults older than
16 years and 38% to those 60 years and older.8 Multiple reports corroborate an especially poor outcome in the elderly.15,24 The mortality rate of
SE in children is 3% in the general population and 6% in the ICU,25 and

  TABLE 85-1    Causes of Status Epilepticus Presenting From the Community
Adults
Prior Seizures

Children

No Prior Seizures

Prior Seizures

No Prior Seizures

Subtherapeutic
­anticonvulsant

Ethanol-related

Subtherapeutic
anticonvulsant

Febrile seizures

Ethanol-related

Drug toxicity

Intractable epilepsy

CNS infection

Intractable epilepsy

CNS infection

Common causes

Head trauma

Head trauma
CNS tumor
Less common causes
CNS infection

Metabolic ­aberration Anoxic brain injury

CNS infection

Metabolic aberration Stroke

Head trauma

Drug toxicity

Metabolic aberration Metabolic aberration

Intractable epilepsy

Stroke
CNS tumor
Head trauma

CNS, central nervous system.
Adapted with permission from Bleck TP, Dunatov CJ. Seizures in critically ill patients. In: Shoemaker WC, Ayres
SM, Grenvik A, Holbrook PR, eds. Textbook of Critical Care. 4th ed. Philadelphia, PA: WB Saunders; 2000.

1/23/2015 12:55:35 PM