You are on page 1of 11

Continuing Medical Education Article

Approach to the comatose patient

Robert D. Stevens, MD; Anish Bhardwaj, MD, FCCM

On completion of this article, the reader should be able to:
1. Describe changes in the level of consciousness.
2. Identify the indicators of prognosis in comatose patients.
3. Use this information in a clinical setting.
Dr. Stevens has disclosed that he is the recipient of direct grant/research funds from the National Institutes of Health, Public
Health Service; Dr. Bhardwaj has disclosed that he was the recipient of direct grant/research funds from the American Heart
Association and that he is/was the recipient of grant/research funds from the National Institutes of Health, Public Health
Wolters Kluwer Health has identified and resolved all faculty conflicts of interests regarding this educational activity.
Visit the Critical Care Medicine Web site ( for information in obtaining continuing medical education

Background: Coma is a medical emergency and may constitute and are associated with a substantial risk of death and disability.
a diagnostic and therapeutic challenge for the intensivist. Management of impaired consciousness includes prompt stabili-
Objective: To review currently available data on the etiology, zation of vital physiologic functions to prevent secondary neuro-
diagnosis, and outcome of coma. To propose an evidence-based logic injury, etiological diagnosis, and the institution of brain-
approach for the clinical management of the comatose patient. directed therapeutic or preventive measures. Neurologic
Data Source: Search of Medline and Cochrane databases; prognosis is determined by the underlying etiology and may be
manual review of bibliographies from selected articles and mono- predicted by the combination of clinical signs and electrophysi-
graphs. ological tests. (Crit Care Med 2006; 34:3141)
Data Synthesis and Conclusions: Coma and other states of KEY WORDS: coma; vegetative state; hypoxic-ischemic enceph-
impaired consciousness are signs of extensive dysfunction or alopathy; traumatic brain injury; neurologic diagnosis; outcome
injury involving the brainstem, diencephalon, or cerebral cortex prediction

oma and other states of im- potentially reversible with the timely in- tive function, and motivation (2). The re-
paired consciousness repre- stitution of medical or surgical therapy. lationship between wakefulness and
sent a severe derangement in Physicians in the emergency and inten- awareness is hierarchical: Awareness can-
cerebral function that may be sive care arena have a central role in not occur without wakefulness, but wake-
structural or nonstructural (toxic-meta- diagnosing and treating comatose pa- fulness may be observed in the absence of
bolic, pharmacologic, seizures) in origin. tients, the principles of which are out- awareness (e.g., the vegetative state; dis-
Many of the underlying processes leading lined in this review. cussed subsequently).
to coma can be both life-threatening and Although many aspects of conscious-
Disorders of Consciousness ness remain unexplained, its neuroana-
Assistant Professor, Division of Neurosciences
tomical underpinnings have been studied
From a clinical perspective, con-
Critical Care, Department of Anesthesiology/Critical extensively. Wakefulness is linked to the
sciousness may be schematized as the
Care Medicine, Neurology and Neurosurgery (RDS), ascending reticular activating system
Vice Chairman, Department of Neurology, Co-Director, product of two closely related cerebral
functions: wakefulness (i.e., arousal, vig- (ARAS), a network of neurons originating
Division of Neurosciences Critical Care, Associate Pro-
fessor of Neurology, Neurological Surgery, Anesthesi- ilance, alertness), and awareness of self or in the tegmentum of the pons and mid-
ology/Critical Care Medicine (AB), The Johns Hopkins of the environment, often referred to as brain and projecting to diencephalic and
University School of Medicine, Baltimore, MD. cortical structures (3, 4). Awareness is
Supported, in part, by grant NS 046379 from the the content of consciousness (1). The
U.S. Public Health Service National Institutes of Health. content of consciousness encompasses, dependent on the integrity of the cerebral
Copyright 2005 by the Society of Critical Care in turn, several other overlapping brain cortex and its subcortical connections
Medicine and Lippincott Williams & Wilkins functions including attention, sensation (5). Biochemical analysis of the ARAS re-
DOI: 10.1097/01.CCM.0000194534.42661.9F and perception, explicit memory, execu- veals cholinergic and glutamatergic (pon-

Crit Care Med 2006 Vol. 34, No. 1 31

Table 1. Global disorders of consciousness

Pattern of Cerebral
Cyclic Respiratory Metabolism
Arousal Awareness Arousal Motor Function Function EEG Activity (% Normal)a

Brain death Absent Absent Absent Absent Absent Electrographic 0

Coma Absent Absent Absent Nonpurposeful Variable; Polymorphic 50
abnormal delta or theta
Vegetative state Present Absent Present Nonpurposeful Present Polymorphic 4060
delta or theta,
slow alpha
Minimally conscious Present Partial Present Intermittently Present Mixed theta and 5060
state purposeful alpha activity
Akinetic mutism Present Partial Present Paucity of Present Diffuse 4080
movement nonspecific
Delirium Present Partial Present Normal Present Diffuse 70100
Locked-in syndrome Present Present Present Quadriplegia, Present Normal 90100
Vertical eye
movement and
blinking only

EEG, electroencephalograph.
As assessed by fluoro-deoxyglucose positron emission tomography.

tomesencephalic origin), adrenergic (lo- pharmacologic, physiologic, and meta- not speak or move spontaneously. They
cus ceruleus), serotonergic and bolic causes of coma should be excluded. do not follow commands, and when pro-
dopaminergic (brainstem), and histamin- Coma. Coma is characterized by the voked by a noxious stimulus their eyes
ergic (hypothalamic) pathways (5). Many total absence of arousal and of awareness. remain closed, vocalization is limited or
of these neurons converge on the thala- As opposed to states of transient uncon- absent, and motor activity is absent or
mus, which then sends projections to the sciousness such as syncope or concus- abnormal and reflexive rather than pur-
cerebral cortex. Additional neurons of the sion, coma must last 1 hr (10). Coma- poseful or defensive (1). Sleep-wake cy-
ARAS project directly to the cerebral cor- tose patients have no eye opening and do cles are lacking. Coma is typically a tran-
tex or to other diencephalic structures sitional state, evolving toward recovery of
such as the hypothalamus. consciousness, the vegetative state, the
Pathologic changes in consciousness Table 2. Glasgow Coma Scale minimally conscious state, or brain death
(Table 1) imply a significant alteration in (Fig. 1). Coma is associated with injury to
the awareness of self and of the environ- Motor response (M) or functional disruption of bilateral cor-
Follows commands 6
ment, with variable degrees of wakeful- Localizes pain 5 tical structures or of the ARAS. Lesions
ness (6). Descriptive terms such as som- Withdraws to pain 4 involving the brainstem portion of the
nolence, stupor, obtundation, and Flexion 3 ARAS frequently coexist with oculomotor
lethargy used to denote different levels of Extension 2 findings and pathologic breathing pat-
None 1
wakefulness are best avoided, given the terns.
Verbal response (V)
lack of uniformity in the way these states Oriented 5 Vegetative State. The vegetative state
are defined in the literature (1, 6) and the Confused speech 4 (VS) is notable for preserved arousal
availability of objective measures such as Inappropriate words 3 mechanisms associated with a complete
the Glasgow Coma Scale (Table 2) (1, 7, Incomprehensible 2 lack of self or environmental awareness
None 1
8). Eye opening (E) (10, 11). Patients in a VS open their eyes
Brain Death. Consciousness disorders Spontaneous 4 spontaneously; however, there is no evi-
must be distinguished from brain death, To command 3 dence of sustained visual pursuit (track-
which is the irreversible loss of all brain To pain 2 ing) or visual fixation. They do not follow
None 1
and brainstem function, clinically diag- commands and do not move in any mean-
nosed by demonstrating absence of con- When assessing record best motor and re- ingful or purposeful manner. They evolve
sciousness, lack of motor response to sponse. Endotracheal tube or tracheostomy in- through temporal cycles of increased and
noxious stimulus, and the disappearance validates the verbal component. Coma defined as decreased arousal akin to a sleep-wake
of brainstem reflexes and respiratory M 4, V 2, E 2 (Glasgow Coma Scale 8). pattern. Cardiovascular regulatory func-
drive (9). Before this determination, Adapted from Teasdale and Jennett (8). tion, breathing patterns, and cranial

32 Crit Care Med 2006 Vol. 34, No. 1

the acute onset and fluctuating level of
consciousness that are characteristic of a
delirium (21). Delirium is frequently seen
in acute toxic, metabolic, or endocrine
derangements but may also result from
more focal injury to the frontal or right
parietal lobes (26). Nonconvulsive seizure
activity and the postictal state may also
mimic delirium.
Locked-In Syndrome. The locked-in
syndrome (LIS) consists of quadriplegia
and anarthria in the setting of preserved
awareness and arousal (1). It not a con-
sciousness disorder per se but may be
clinically confused with one given the
limited expressive capability of these pa-
tients. It is associated with acute injury to
Figure 1. Spectrum of consciousness disorders. the ventral pons, just below the level of
the third nerve nuclei, thus classically
nerves are usually intact. Although some Akinetic Mutism. Akinetic mutism sparing vertical eye movements and
patients who are in a VS regain partial or (AM) is a state of wakefulness with lim- blinking and not interfering with the
complete consciousness, others remain ited objective evidence of awareness (16, more dorsally located ARAS. More rostral
for extended periods without significant 17). Patients with this condition gener- lesions may induce a total LIS, in which
changes in their neurologic state (Fig. 1), ally seem unable to move or speak and even eye and lid movement is lost, pro-
prompting the term persistent vegeta- have cyclic periods of increased arousal as hibiting all communication. Common
tive state (PVS). The Multi-Society Task indicated by eye opening. Some authors etiologies of the LIS are pontine infarc-
Force, an interdisciplinary consensus describe elements of visual pursuit and tion, hemorrhage, and trauma (27). An
group, defined PVS as a VS present 1 an intent gaze suggesting an unfulfilled analogous awake but de-efferented state
month after an acute traumatic or non- promise of speech (18). Contrary to the may occur in patients with severe Guil-
traumatic injury (10). There has been de- MCS, there is no motor responsiveness to lain-Barr syndrome, botulism, and crit-
bate regarding the significance of this verbal, tactile, or noxious stimuli. A dis- ical illness neuropathy and those receiv-
designation, and one expert panel sug- tinguishing feature from the VS is that ing neuromuscular blocking agents
gested that the term PVS be abandoned patients with AM do not have spasticity or without adequate sedation (28).
altogether (12). The VS is caused by wide- abnormal reflexes, suggesting relative Other States of Impaired Conscious-
spread damage to bilateral cerebral hemi- sparing of the corticospinal tract (19). AM ness. Several other states involve or sug-
spheres with sparing of the brainstem; has been associated with bilateral medial gest a global perturbation in conscious-
trauma and hypoxic-ischemic encepha- frontal lobe (e.g., cingulate gyrus) injury ness. Hypersomnia or excessive daytime
lopathy are the most common acute or dysfunction, leading to a profound de- somnolence is an increase in sleeping
causes of VS. ficiency in motivation and an inability to time with preserved sleep-wake cycles,
Minimally Conscious State. The min- plan and initiate activity (executive dys- most often seen in the setting of sleep
imally conscious state (MCS) describes a function) (17). deprivation, sleep-related breathing dis-
subset of patients who do not meet the Delirium. Delirium is synonymous orders, narcolepsy, drug toxicity, meta-
criteria for coma or VS. Patients in MCS with the acute confusional state and with bolic encephalopathy, or damage to the
have a severe alteration in consciousness acute encephalopathy (20, 21). Its is char- ARAS (29). When aroused, patients with
but demonstrate wakefulness and cyclic acterized by an acutely developing im- hypersomnia typically have a normal
arousal and intermittently demonstrate pairment in attention, associated with neurologic examination. Catatonia is a
self or environmental awareness, such as changes in the level of consciousness, complication of psychiatric illnesses such
following of commands, the ability to sig- disorganized thinking, and a fluctuating as severe depression, bipolar disorder, or
nal yes/no (regardless of accuracy), intel- course (22, 23). Additional features in- schizophrenia, in which patients have
ligible speech, or purposeful behavior clude perceptual disturbances, altered open eyes but do not speak or move spon-
(13). Emergence from the MCS to higher sleep-wake cycle, increased or decreased taneously and do not follow commands,
states of consciousness is signaled by the psychomotor activity, and memory im- with an otherwise normal neurologic ex-
ability to communicate reliably or use pairment. Delirium is exceedingly com- amination and an electroencephalogram
objects functionally (13). Although data mon in hospitalized patients and partic- showing low voltage but no slowing (21,
are limited, it is believed that the MCS ularly in the intensive care unit (24). It 30). General anesthesia and barbiturate
represents a greater likelihood of recov- may precede or may evolve from other coma are pharmacologically induced
ery compared with the VS. As in the VS, disorders of consciousness, namely coma states of decreased arousal and aware-
lesions or dysfunction associated with the (Fig. 1) (25). Delirium should be distin- ness, associated with minimal or absent
MCS involve the cerebral hemispheres, guished from dementia. Both disorders responses to noxious (surgical) stimuli
with possibly greater sparing of cortico- are characterized by memory impair- and prominent brainstem dysfunction
cortical and cortico-thalamic connective ment, but patients with dementia alone and respiratory depression (31). General-
fibers (14, 15). are more commonly alert and do not have ized convulsive or complex partial sei-

Crit Care Med 2006 Vol. 34, No. 1 33

Table 3. Etiology of coma and altered consciousness

I. Primary cerebral disorders

Bilateral or diffuse hemispheric disorders
Traumatic brain injury (contusions, diffuse axonal injury)
Ischemic (watershed, cardioembolism, vasculitis, hypercoagulable disorder)
Hemorrhagic (subarachnoid hemorrhage, intraventricular hemorrhage)
Hypoxic-ischemic encephalopathy
Cerebral venous thrombosis
Meningitis; encephalitis
Generalized or complex partial seizures; status epilepticus (convulsive, nonconvulsive)
Hypertensive encephalopathy
Posterior reversible encephalopathy syndrome
Acute disseminated encephalomyelitis
Unilateral hemispheric disorders (with displacement of midline structures)
Traumatic (contusions, subdural hematoma, epidural hematoma)
Large hemispheric ischemic stroke
Primary intracerebral hemorrhage
Cerebral abscess
Brain tumor
Brain stem disorders (pons, midbrain)
Hemorrhage, infarction, tumor, trauma
Central pontine myelinolysis
Compression from cerebellar infarct, hematoma, abscess, tumor
II. Systemic derangements causing coma
Medication overdose/adverse effects (opioids, benzodiazepines, barbiturates, tricyclics, neuroleptics, aspirin, selective serotonin reuptake inhibitors,
acetaminophen, anticonvulsants)
Drugs of abuse (opioids, alcohol, methanol, ethylene glycol, amphetamines, cocaine)
Exposures (carbon monoxide, heavy metals)
Systemic inflammatory response syndrome-sepsis
Hypoxia; hypercapnia
Hypoglycemia; hyperglycemic crises (diabetic ketoacidosis, nonketotic hyperosmolar hyperglycemic state)
Hyponatremia, hypernatremia
Hepatic failure
Renal failure
Wernickes encephalopathy
Adrenal insufficiency
Hypothyroidism; hyperthyroidism

zures or a resulting postictal state may enough to cause displacement of midline ide), alterations in neuronal excitability
also lead to altered consciousness. structures (shift) (1). Brainstem and dien- and signaling (seizures, acidosis, drug
cephalic lesions resulting in coma may be toxicity), or changes in brain volume (hy-
Etiology and Pathogenesis comparatively small; however, they must pernatremia, hyponatremia). The degree
also involve bilateral structures (4). Com- of neurologic impairment is related to
Common causes of coma are traumatic partmental shift of sufficient magnitude the time course of the underlying cere-
brain injury (TBI), hypoxic-ischemic en- will disrupt the structural integrity or func- bral pathology. Thus, an acute hemi-
cephalopathy (HIE), drug overdose, isch- tion of contralateral reticulothalamic or spheric or brainstem hemorrhage with
emic stroke, intracranial hemorrhage, cen- thalamocortical fibers, impairing the ARAS mass effect will be associated with de-
tral nervous system infections, and brain and its projections. Shift may also cause pressed consciousness, whereas a slowly
tumors. From a pathophysiologic stand- central or tentorial herniation with com- developing brain tumor of identical loca-
point, coma may be viewed as the expres- pression of the midbrain, compromising tion and volume may be asymptomatic. A
sion of a) primary insults to the cerebral more proximal elements of the ARAS (32). similar observation can be made for met-
cortex, diencephalic structures, midbrain The pathophysiology of toxic and met- abolic changes, such as acute vs. progres-
or rostral pons; and b) secondary cerebral abolic coma is specific to the underlying sive hyponatremia.
manifestations of systemic toxic, metabolic, cause and in many instances incom-
or endocrine derangements (Table 3) (1, pletely understood (33). In a simplified Approach to the Comatose
32). view, these conditions have been linked Patient
To affect consciousness, lesions of the to an interruption in the delivery or uti-
cerebral cortex must involve both hemi- lization of oxygen or substrate (hypoxia, Disorders of consciousness such as
spheres or must be unilateral lesions large ischemia, hypoglycemia, carbon monox- coma are potentially life threatening and

34 Crit Care Med 2006 Vol. 34, No. 1

Similar observations have been made in
patients with ischemic stroke (42).
Neurologic Evaluation. The goal of
the neurologic examination is to recog-
nize the type of consciousness disorder
and make inferences about its etiology.
Essential neuroanatomically localizing
information may be gleaned by a rela-
tively rapid survey, which should include
the level of consciousness (wakefulness),
cranial nerve examination, motor exami-
nation, and respiratory pattern (1). Addi-
tional clues may be sought by examining
the head and neck (e.g., meningismus),
the optic fundi (e.g., subhyaloid hemor-
rhage in subarachnoid hemorrhage), and
the skin (e.g., purpuric lesions in menin-
gococcal meningitis) (32, 43). An impor-
tant early step is to differentiate patients
who have a structural cause of coma from
those with a metabolic one. Structural
causes are indicated by the presence of
lateralizing deficits and a rostrocaudal
progression of brainstem dysfunction
(Tables 4 and 5). The presence of invol-
Figure 2. Algorithm for initial emergent management of the comatose patient. GCS, Glasgow Coma untary movements (seizures, myoclonus,
Scale; MAP, mean arterial pressure; ICP, intracranial pressure; IV, intravenous; CT, computed tomog- asterixis), especially if generalized, sug-
raphy; EEG, electroencephalograph; MRI, magnetic resonance imaging. Adapted from Reference 122.
gests a metabolic etiology.
Level of Consciousness. The patient
warrant a rapid and structured approach. tose patient suggests elevated intracra- should be inspected for spontaneous body
A basic sequence of steps is outlined here- nial pressure, drug overdose position, motor activity, eye opening, or
after. These include stabilization of vital (amphetamines, cocaine), or hyperten- verbalization. Purposeful movements
physiologic functions, performing a fo- sive encephalopathy, but more frequently (e.g., reaching for endotracheal tube) and
cused neurologic examination, targeted hypertension is a nonspecific hyperadren- comfort postures (e.g., crossing the legs)
diagnostic tests, and when available the ergic response to an acutely evolving in- are signs of cortical integration. The re-
institution of specific therapeutic mea- tracranial or systemic process (35). Hy- sponse to stimuli of graded intensity
sures (Fig. 2). potension and shock in the comatose should then be observed, starting with
Initial Stabilization. As in any medical patient usually reflect nonneurogenic verbal commands, progressing to tactile
or surgical emergency, initial steps mechanisms of circulatory failure but cues and finally to noxious provocation.
should be directed to ensuring adequacy may also be a consequence of severe brain Noxious stimuli should be delivered with-
of airway, breathing, and circulatory damage (neurogenic stunned myocar- out inducing tissue trauma and with re-
function. In patients who are comatose dium) (36). Acute respiratory failure in gard to the possibility of conscious pain
from TBI, or in whom the trauma cannot the comatose patient may occur second- perception; preferable sites are the nail-
be ruled out as an etiological factor, the ary to airway obstruction, pulmonary as- bed and the notch of the supra-orbital
neck should be immobilized until cervi- piration, acute lung injury, or lesions af- nerve.
cal spine instability has been ruled out by fecting the pontine and medullary The Glasgow Coma Scale (GCS) (Table
clinical examination and appropriate im- respiratory centers. Acute central ner- 2) was initially devised for patients with
aging (34). Efforts should be made to vous system insults have been linked to TBI (8) but has gained widespread accep-
swiftly identify the causes of, and correct, neurogenic pulmonary edema, charac- tance as a bedside tool for evaluating the
systemic derangements such as hyperten- terized by severe hypoxemia and protein- level of consciousness in virtually all
sion, hypotension, hypoxemia, anemia, rich diffuse alveolar exudates (37). Coma acutely ill patients (44). It has good in-
acidosis, hypothermia, hyperglycemia, may also represent a consequence of hy- terobserver reliability (45) and is a pow-
and hyperthermia. percapnic or hypoxemic respiratory fail- erful predictor of survival and neurologic
Systemic Derangements. The rela- ure, suggesting a vicious cycle mutually outcomes after head trauma (46 49),
tionship of acute neurologic disorders reinforcing brain and respiratory dys- nontraumatic coma (50), ischemic stroke
with derangements in circulatory and re- function. (51, 52), subarachnoid hemorrhage (53),
spiratory function is complex, and it is The importance of appropriately treat- intracerebral hemorrhage (54, 55), and
frequently not possible at the outset to ing systemic derangements is under- meningitis (56). The GCS is also an inde-
accurately determine whether the sys- scored by studies demonstrating that pendent predictor of survival in the gen-
temic derangement occurred secondary neurologic outcomes are substantially eral critically ill population (57) and has
to coma, was a cause of it, or was inde- worse in patients with TBI who develop been incorporated into a number of
pendent of it. Hypertension in the coma- hypotension and/or hypoxia (38 41). widely used prognostic intensive care

Crit Care Med 2006 Vol. 34, No. 1 35

scoring systems (58 64). Not withstand- range (1315) correlate poorly with out- Unilateral pupillary dilation in the co-
ing, the GCS has several important limi- come and neuroimaging findings (68, matose patient is evidence of oculomotor
tations (44, 65, 66). Subtle alterations in 69). nerve compression from ipsilateral uncal
wakefulness and brainstem findings may Cranial Nerve Examination. Exami- herniation until demonstrated otherwise.
not be captured by the GCS. A full GCS nation of the eyes may yield valuable in- This presentation may also occur in pa-
cannot be obtained in patients who are formation about the level of brainstem tients with posterior communicating ar-
endotracheally intubated, are sedated, or disease causing coma, given the proxim- tery aneurysmal rupture. Bilateral dilated
have craniofacial trauma or in patients ity of centers governing eye movement, pupils that do not react to light are a sign
with dominant hemisphere lesions and pupillary function, and elements of the of extensive midbrain injury, central her-
aphasia. Midrange scores (6 12) may re- ARAS. Completely normal pupillary func- niation, drug intoxication (tricyclic anti-
sult in different combinations of the tion and eye movements suggest that the depressants, anticholinergic agents, am-
three components yielding equivalent to- lesion causing coma is rostral to the mid- phetamines, carbamazepine), or brain
tal scores that do not reflect the same brain (70). Detection of brainstem injury death. Unilateral miosis is suggestive of
degree of unconsciousness (65, 67). Fi- may also aid in prognostication (dis- Horners syndrome due to sympathetic
nally, differences in the higher GCS cussed subsequently). denervation. Bilateral miosis is seen with
lesions in the pontine tegmentum, opioid
Table 4. Signs of intracranial hypertension and associated herniation syndromes overdose, and cholinergic toxicity (or-
ganophosphates and other cholinesterase
Sign Mechanism Type of Herniation inhibitors).
Abnormalities of eye position and
Coma Compression of midbrain tegmentum Uncal, central
Pupillary dilation Compression of ipsilateral third nerve Uncal
movement may be informative. Conju-
Miosis Compression of the midbrain Central gate lateral deviation of the eyes is a sign
Lateral gaze palsy Stretching of the sixth nerves Central either of an ipsilateral hemisphere lesion,
Hemiparesisa Compression of contralateral cerebral Uncal a contralateral hemisphere seizure focus,
peduncle against tentorium or damage involving the contralateral
Decerebrate posturing Compression of the midbrain Central, uncal
Hypertension, Compression of the medulla Central, uncal, cerebellar pontine horizontal gaze center (parapon-
bradycardia (tonsillar) tine reticular formation). Lateral gaze
Abnormal breathing Compression of the pons or medulla Central, uncal, cerebellar palsy may signal central herniation with
patterns (tonsillar) compression of bilateral sixth nerves.
Posterior cerebral artery Vascular compression Uncal Tonic downward deviation of gaze is sug-
Anterior cerebral artery Vascular compression Subfalcine (cingulate)
gestive of injury or compression involv-
infarction ing the thalamus or dorsal midbrain such
as may occur with acute obstructive hy-
Hemiparesis will occur ipsilateral to the hemispheric lesion (false-localizing sign). drocephalus or midline thalamic hemor-

Table 5. Brainstem reflexes in the comatose patient

Examination Normal
Technique Response Afferent Pathway Brainstem Efferent Pathway

Pupils Response to Direct and Retina, optic Edinger-Westphal nucleus Oculomotor

light consensual nerve, chiasm, (midbrain) nerve,
pupillary optic tract sympathetic
constriction fibers
Oculocephalic Turn head from Eyes move Semicircular Vestibular nucleus. Medial Oculomotor and
side to side conjugately canals, longitudinal fasciculus. abducens
in direction vestibular Parapontine reticular nerves
opposite to nerve formation (pons)
Vestibulo- Irrigate external Nystagmus with Semicircular Vestibular nucleus. Medial Oculomotor and
oculocephalic auditory fast canals, longitudinal fasciculus. abducens
canal with component vestibular Parapontine reticular nerves
cold water beating away nerve formation (pons)
Corneal reflex Stimulation of Eyelid closure Trigeminal nerve Trigeminal and facial nuclei Facial nerve
cornea (pons)
Cough reflex Stimulation of Cough Glossopharyngeal Medullary cough center Glossopharyngeal
carina and vagus and vagus
nerves nerves
Gag reflex Stimulation of Symmetric Glossopharyngeal Medulla Glossopharyngeal
soft palate elevation of and vagus and vagus
soft palate nerves nerves

36 Crit Care Med 2006 Vol. 34, No. 1

rhage. Tonic upward gaze has been asso- dal portions of the pons. Ataxic breathing intracranial abnormalities that might
ciated with bilateral hemispheric is irregular in both rate and tidal volume cause brain herniation after removal of
damage. Ocular bobbing, a rapid down- and suggests damage to the medulla. cerebrospinal fluid (82).
ward jerk followed by a slow return to Diagnostic Tests. Patients with coma Electroencephalography (EEG) is fre-
midposition, is indicative of pontine le- should be placed on monitors including quently indicated in patients with unex-
sions. Rapid intermittent horizontal eye pulse oximetry, blood pressure, and elec- plained coma as it may diagnose clinically
movements suggest seizure activity. trocardiogram, and blood should be im- occult seizure activity and in particular
Integrity of the brainstem may be fur- mediately sent for glucose, electrolytes, nonconvulsive status epilepticus. Recent
ther ascertained by eliciting specific re- and arterial blood gas analysis. Routine studies indicate that nonconvulsive sta-
flexes (Table 4). Disappearance of brain- serum tests of liver, renal, thyroid, and tus epilepticus is present in 8 19% of
stem reflexes may also aid in adrenal function and urine toxicology patients who undergo EEG as part of the
prognostication (45) (discussed subse- screens should be obtained without delay. workup for an unexplained decrease in
quently). Computed tomography (CT) of the level of consciousness (83 85). The spec-
Motor Examination. Movements may brain is warranted in virtually all patients ificity of EEG for the etiological diagnosis
be classified as involuntary, reflexic, or with an acute onset of unexplained coma. of nonepileptic causes of coma is low;
purposeful. Involuntary movements in- CT will readily identify intracranial hem- however, certain characteristic abnor-
clude seizures, myoclonus, or tremor orrhage, hydrocephalus, brain edema, malities have been described. In patients
(e.g., toxic-metabolic encephalopathy). and compartmental shift and may sug- with coma of metabolic origin, a generic
Reflexes are stereotypical responses of the gest stroke, abscess, or tumor. CT is un- pattern of diffusely decreased frequency
extremities evoked in patients who lack revealing in most instances of hypoxic- and increased amplitude has been de-
descending hemispheric modulation of ischemic or toxic-metabolic coma. Also, scribed, often in association with tripha-
motor function. Purposeful movements the usefulness of CT in patients who be- sic waves (86). Some comatose patients
(e.g., localization) imply cortical process- come comatose during the course of a have an EEG that superficially resembles
ing of environmental variables. critical illness is unclear. One study an awake (alpha) pattern. This alpha-
Extensor (decerebrate) posturing is found that CT was unlikely to reveal ab- coma is most commonly seen in patients
characterized by adduction, extension, normalities in intensive care unit pa- with extensive damage or dysfunction in-
and pronation of the upper extremities tients who did not have new neurologic volving the brainstem or cerebral cortex
and extension of the lower extremities; it deficits or seizures (72). In each case, the and generally carries a poor prognosis
is indicative of injury to the caudal dien- anticipated benefit of CT in terms of ac- (87). Finally, herpesvirus encephalitis has
cephalon, midbrain, or pons. Flexor tionable information should be carefully been associated with changes such as pe-
(decorticate) posturing involves flexion weighed against the risks of transporting riodic sharp waves or epileptiform dis-
and adduction of arms and wrists with a critically ill patient outside the inten- charges (88, 89). Although the diagnostic
extension of lower extremities; it sug- sive care unit. capabilities of EEG are disappointing, it
gests hemispheric or thalamic damage, Magnetic resonance imaging (MRI) has emerged as a valuable tool in the
with sparing of structures below the di- should be sought in patients with unex- prognostication of comatose patients
encephalons (1). plained coma and normal or equivocal CT (discussed subsequently).
Respiratory Pattern. Coma has been findings. MRI is highly sensitive to acute
associated with disordered breathing pat- ischemic stroke, intracerebral hemor- Prognosis
terns, reflecting injury to brainstem re- rhage, cerebral venous sinus thrombosis,
spiratory centers or interference with su- brain edema, brain tumor, inflammatory Outcomes after coma include death,
prabulbar regulation of these sites (1, 71). processes, and cerebral abscess. MRI is VS, various degrees of functional impair-
Circulatory or respiratory failure, toxic- more sensitive than CT to diffuse axonal ment, and complete neurologic recovery.
metabolic disorders, drugs used for seda- injury, a pattern of damage seen in TBI Such categories have been formalized in
tion or analgesia, and mechanical venti- patients (73, 74). Studies in encephalo- scoring systems such as the Glasgow Out-
lation may also contribute to abnormal pathic or comatose patients with sepsis or come Scale (GOS) (90) (Table 6) and ex-
breathing, limiting the inferential value after cardiac surgery have found that MRI tended GOS (91). There are many other
of specific patterns. Cheyne-Stokes respi- may be highly sensitive for the detection outcomes of coma including nonvegeta-
ration denotes a cyclic pattern of alter- of lesions not suspected by clinical exam- tive disorders of awareness (e.g., MCS,
nating hyperpnea and apnea. It is seen in ination or CT (7578). AM) and gradations of cognitive dysfunc-
patients with a bilateral hemispheric or Lumbar puncture and cerebrospinal tion that are not captured by systems
diencephalic insult but may also be fluid analysis should be considered in co- such as the GOS (92). Nevertheless, the
present in congestive heart failure, matose patients with a suspected infec- GOS is widely used by clinical investiga-
chronic obstructive pulmonary disease, tious or inflammatory central nervous tors in both traumatic and nontraumatic
and sleep apnea. Hyperventilation has system condition (79). However, in im- coma.
been linked to injury in the pontine or munologically competent critically ill pa- The prediction of neurologic outcome
midbrain tegmentum; however, it can tients being worked up for fever and en- is a central concern in the management
also result from respiratory failure, he- cephalopathy who have not undergone a of patients with coma. Surveys indicate
modynamic shock, fever, sepsis, meta- recent neurosurgical procedure, the diag- that many individuals would prefer death
bolic disarray, and psychiatric disease. nostic yield of this test is low (80, 81). to the prospect of living in a vegetative or
Apneustic breathing is characterized by a Patients with alterations in conscious- highly dependent state (93). Prognostic
prolonged pause at the end of inspiration ness should undergo CT of the head be- information, when available, provides a
and indicates lesions of the mid- and cau- fore lumbar puncture to identify occult rational basis for decision making about

Crit Care Med 2006 Vol. 34, No. 1 37

Table 6. Glasgow Outcome Scale Neuroimaging. Neuroimaging is es-
sential to coma diagnosis and may have
1 Death
2 Persistent vegetative state Patient exhibits no obvious cortical function.
prognostic value. In patients with trau-
3 Severe disability (Conscious but disabled). Patient depends on others for matic coma, patterns on CT that have
daily support due to mental or physical disability or been associated with worse neurologic
both. outcome include lesions in the brain-
4 Moderate disability (Disabled but independent). Patient is independent as far as stem, encroachment of the basal cisterns,
daily life is concerned. The disabilities found include and diffuse axonal injury (103108). CT
varying degrees of dysphasia, hemiparesis, or ataxia, as
well as intellectual and memory deficits and personality
findings are also predictive of outcome in
changes comatose patients with intracerebral
5 Good recovery Resumption of normal activities even though there may be hemorrhage (55) and subarachnoid hem-
minor neurological or psychological deficits. orrhage (109). However, when compared
with electrophysiologic and clinical vari-
Adapted from Jennett and Bond (90). ables, the predictive value of CT is re-
ported to be low (110). MRI in comatose
patients may disclose pathologic changes
therapeutic intensity, withdrawal of life somatosensory evoked potentials (SSEP), of prognostic importance (73, 74, 111). In
support, and rehabilitation (94). Prognos- transcranial motor evoked potentials,
a study of patients with posttraumatic
tication of comatose patients is based on brainstem auditory evoked potentials
PVS, Kampfl et al. (112) assessed the re-
a consideration of etiology, clinical signs, (BAEP), and event-related potentials (86,
lationship between MRI findings and out-
electrophysiology, neuroimaging, and 101). In patients with postanoxic coma,
come at 12 months. They noted that le-
biochemical data (95). meta-analysis revealed that EEG with an
sions in the corpus callosum and
Etiology. The prognosis of coma is isoelectric or burst-suppression pattern
determined in large part by the underly- was independently associated with poor dorsolateral brainstem were highly pre-
ing etiology. It is widely believed that outcome (99). A recent systematic review dictive of nonrecovery, whereas initial
toxic/metabolic coma carries a better implied that in patients with an alpha- GCS and pupillary findings were not
prognosis than coma of structural origin; coma after HIE or TBI, mortality rate was (112). A relationship between MRI abnor-
however, direct evidence demonstrating 6190%, whereas mortality rate was 27% malities and outcome is also suggested in
this is limited. On the other hand, the when alpha-coma was metabolic in origin comatose patients following ischemic
better prognosis of traumatic vs. anoxic (87). Although it remains a dependable stroke (113) and hypoxic-ischemic injury
coma is supported in a number of studies tool for epilepsy assessment, EEG is lim- (114).
and two recent systematic reviews (96, ited by its high sensitivity to the electrical Biochemical Markers. The presence of
97). In adults with PVS secondary to TBI environmental noise, its alteration by brain-specific molecules in the blood or
who were reevaluated at 1 yr, the propor- sedative drugs, and its inability to test the cerebrospinal fluid has been postulated to
tion with a good recovery was 7%, mod- brainstem. reflect the severity of brain damage. Mol-
erate disability 17%, severe disability Data from SSEP are strongly predic- ecules that have been studied in the prog-
28%, PVS 15%, and dead 33% (96). In tive of outcome. A systematic review re- nostic evaluation of coma include neuron
patients with nontraumatic PVS (princi- vealed that among comatose patients specific enolase (115118), s100 beta
pally HIE), these outcomes were signifi- with bilaterally absent cortical SSEP, the (118), glial fibrillary acidic protein (117,
cantly worse (respectively 1%, 3%, 11%, incidence of death or VS was 100% after 119), and the BB isoform of creatine ki-
32%, and 53%) (96). Younger age and HIE, 99% after intracranial hemorrhage, nase (120). Although sensitive to brain
better general health of TBI patients may 95% after TBI, and 93% in children and injury, the specificity and predictive value
account for some of this difference. adolescents 18 yrs old with coma of of these tests have been insufficient when
Clinical Signs. Clinical signs that cor- diverse etiologies (102). These results compared with clinical variables and
relate with poor prognosis after coma in- were consistent with the meta-analysis studies such as SSEP (101, 121, 122).
clude the motor component of the GCS that concluded from a review of 33 stud-
(46, 47, 50, 51, 53, 55, 56), the length of ies that bilateral absence of cortical SSEP
time a patient remains in coma (97, 98), within the first week after anoxic coma CONCLUSIONS
and signs of brainstem damage. In two was 100% specific in identifying patients
separate systematic reviews of postanoxic with poor outcome (99). Most studies of The goal of acute care of patients with
coma, absent motor responses (GCS mo- BAEP to assess prognosis after coma were coma is to maximize the likelihood of
tor score 1) on day 3 were highly predic- conducted with patients with TBI (101). A neurologic recovery. Initial resuscitative
tive of poor outcome (death or VS) (45, pooled analysis demonstrated that of 107 steps should be as in any other medical
99). Absent pupillary responses on day 1 patients with absence of peak V (gener- emergency. Subsequent therapeutic in-
(45) or 3 (99) and absent corneal reflexes ated in the upper pons and in the mid- tervention is dependent on the underly-
on day 1 (45) were also strongly corre- brain) on BAEP recorded 159 days after ing etiology, and the clinician must rap-
lated with poor outcome. Clinical signs TBI, 106 were dead or vegetative at 312 idly integrate clinical examination,
appeared to be less accurate in the pre- months follow-up (100). At our institu- laboratory, and imaging data in the for-
diction of outcome after TBI (100). tion, BAEP and SSEP are obtained in mulation of a presumptive diagnosis. Pre-
Electrophysiologic Tests. Electro- selected comatose patients in whom diction of outcome may be accomplished
physiologic tests that have been used for prognosis is not readily predictable from by clinical examination and electrophys-
predicting coma outcome include EEG, clinical parameters alone. iological tests.

38 Crit Care Med 2006 Vol. 34, No. 1

15. Coleman MR, Menon DR, Fryer TD, et al: 33. Kunze K: Metabolic encephalopathies.

Neurometabolic coupling in the vegetative J Neurol 2002; 249:1150 1159
anagement of and minimally conscious states: prelimi- 34. Cervical spine immobilization before ad-
nary findings. J Neurol Neurosurg Psychi- mission to the hospital. Neurosurgery 2002;
impaired con- atry 2005; 76:432 434 50:S7S17
16. Cairns HWB, Whitteridge D: Akinetic mut- 35. Varon J, Marik PE: The diagnosis and man-
sciousness in- ism with an epidermoid cyst of the 3rd agement of hypertensive crises. Chest 2000;
ventricle. Brain 1941; 64:273291 118:214 227
cludes prompt stabilization 17. Ackermann H, Ziegler W: Akinetic mut- 36. Samuels MA: Neurally induced cardiac
ismA review of the literature. Fortschr damage. Definition of the problem. Neurol
of vital physiologic functions
Neurol Psychiatr 1995; 63:59 67 Clin 1993; 11:273292
to prevent secondary neuro- 18. Recommendations for use of uniform no- 37. Colice GL, Matthay MA, Bass E, et al: Neu-
menclature pertinent to patients with se- rogenic pulmonary edema. Am Rev Respir
logic injury, etiological diag- vere alterations in consciousness: American Dis 1984; 130:941948
Congress of Rehabilitation Medicine. Arch 38. Chesnut RM, Marshall LF, Klauber MR, et
nosis, and the institution of Phys Med Rehabil 1995; 76:205209 al: The role of secondary brain injury in
19. Cartlidge N: States related to or confused determining outcome from severe head in-
brain-directed therapeutic or with coma. J Neurol Neurosurg Psychiatry jury. J Trauma 1993; 34:216 222
2001; 71(Suppl 1):i18 i19 39. Fearnside MR, Cook RJ, McDougall P, et al:
preventive measures. 20. Mesulam MM: Principles of Behavioral and The Westmead Head Injury Project out-
Cognitive Neurology. Second Edition. New come in severe head injury. A comparative
York, Oxford, Oxford University Press, 2000 analysis of pre-hospital, clinical and CT
21. American Psychiatric Association: Task variables. Br J Neurosurg 1993; 7:267279
REFERENCES Force on DSM-IV. Diagnostic and Statistical 40. Zygun DA, Kortbeek JB, Fick GH, et al:
Manual of Mental Disorders: DSM-IV-TR. Non-neurologic organ dysfunction in severe
1. Plum F, Posner JB: The Diagnosis of Stupor Fourth Edition. Washington, DC, American traumatic brain injury. Crit Care Med 2005;
and Coma. Contemporary Neurology Series Psychiatric Association, 2000 33:654 660
19. Third Edition. Philadelphia, Davis, 1980 22. American Psychiatric Association: Task 41. Robertson CS, Valadka AB, Hannay HJ, et
2. Young GB, Pigott SE: Neurobiological basis Force on DSM-IV. Diagnostic and Statistical al: Prevention of secondary ischemic insults
of consciousness. Arch Neurol 1999; 56: Manual of Mental Disorders: DSM-IV. after severe head injury. Crit Care Med
153157 Fourth Edition. Washington, DC, American 1999; 27:2086 2095
3. Moruzzi G, Magoun HW: Brain stem retic- Psychiatric Association, 1994 42. Sulte G, Elting JW, Langedijk M, et al: Ad-
ular formation and activation of the EEG. 23. Inouye SK, van Dyck CH, Alessi CA: Clari- mitting acute ischemic stroke patients to a
1949. J Neuropsychiatry Clin Neurosci fying confusion: the confusion assessment stroke care monitoring unit versus a con-
1995; 7:251267
method. A new method for detection of de- ventional stroke unit: A randomized pilot
4. Parvizi J, Damasio AR: Neuroanatomical
lirium. Ann Intern Med 1990; 113:941948 study. Stroke 2003; 34:101104
correlates of brainstem coma. Brain 2003;
24. Ely EW, Shintani A, Truman B, et al: Delir- 43. Ziai WC: Coma and altered consciousness.
126:1524 1536
ium as a predictor of mortality in mechan- In: Handbook of Neurocritical Care. Cur-
5. Zeman A: Consciousness. Brain 2001; 124:
ically ventilated patients in the intensive rent Clinical Neurology. Bhardwaj A, Mirski
care unit. JAMA 2004; 291:17531762 MA, Ulatowski JA (Eds.) Totowa, NJ, Hu-
6. Laureys S, Owen AM, Schiff ND: Brain func-
25. McNicoll L, Pisani MA, Zhang Y, et al: De- mana Press, 2004
tion in coma, vegetative state, and related
lirium in the intensive care unit: Occur- 44. The Brain Trauma Foundation: The Ameri-
disorders. Lancet Neurol 2004; 3:537546
rence and clinical course in older patients. can Association of Neurological Surgeons.
7. Bateman DE: Neurological assessment of
J Am Geriatr Soc 2003; 51:591598 The Joint Section on Neurotrauma and
coma. J Neurol Neurosurg Psychiatry 2001;
26. Burns A, Gallagley A, Byrne J: Delirium. Critical Care. Glasgow Coma Scale Score.
71(Suppl 1):i13i17
8. Teasdale G, Jennett B: Assessment of coma J Neurol Neurosurg Psychiatry 2004; 75: J Neurotrauma 2000; 17:563571
and impaired consciousness. A practical 362367 45. Booth CM, Boone RH, Tomlinson G, et al: Is
scale. Lancet 1974; 2:81 84 27. Smith E, Delargy M: Locked-in syndrome. this patient dead, vegetative, or severely
9. Wijdicks EF: The diagnosis of brain death. BMJ 2005; 330:406 409 neurologically impaired? Assessing out-
N Engl J Med 2001; 344:12151221 28. Vargas F, Hilbert G, Gruson D, et al: Ful- come for comatose survivors of cardiac ar-
10. Medical aspects of the persistent vegetative minant Guillain-Barre syndrome mimick- rest. JAMA 2004; 291:870 879
state (1): The Multi-Society Task Force on ing cerebral death: Case report and litera- 46. Winkler JV, Rosen P, Alfry EJ: Prehospital
PVS. N Engl J Med 1994; 330:1499 1508 ture review. Intensive Care Med 2000; 26: use of the Glasgow Coma Scale in severe
11. Jennett B, Plum F: Persistent vegetative 623 627 head injury. J Emerg Med 1984; 2:1 6
state after brain damage. A syndrome in 29. Douglas NJ: The sleepy patient. J Neurol 47. Baxt WG, Moody P: The impact of advanced
search of a name. Lancet 1972; 1:734 737 Neurosurg Psychiatry 2001; 71(Suppl 1): prehospital emergency care on the mortal-
12. Giacino J, Whyte J: The vegetative and min- i3i6 ity of severely brain-injured patients.
imally conscious States: current knowledge 30. Taylor MA, Fink M: Catatonia in psychiatric J Trauma 1987; 27:365369
and remaining questions. J Head Trauma classification: A home of its own. Am J Psy- 48. Massagli TL, Michaud LJ, Rivara FP: Asso-
Rehabil 2005; 20:30 50 chiatry 2003; 160:12331241 ciation between injury indices and outcome
13. Giacino JT, Ashwal S, Childs N, et al: The 31. Rudolph U, Antkowiak B: Molecular and after severe traumatic brain injury in chil-
minimally conscious state: Definition and neuronal substrates for general anaesthet- dren. Arch Phys Med Rehabil 1996; 77:
diagnostic criteria. Neurology 2002; 58: ics. Nat Rev Neurosci 2004; 5:709 720 125132
349 353 32. Wijdicks EFM: Neurologic Complications of 49. Servadei F, Nasi MT, Cremonini AM, et al:
14. Laureys S, Perrin F, Faymonville ME, et al: Critical Illness. Contemporary Neurology Importance of a reliable admission Glasgow
Cerebral processing in the minimally con- Series 64. Second Edition. Oxford; New Coma Scale score for determining the need
scious state. Neurology 2004; 63:916 918 York, Oxford University Press, 2002 for evacuation of posttraumatic subdural

Crit Care Med 2006 Vol. 34, No. 1 39

hematomas: A prospective study of 65 pa- descriptor of a complex clinical outcome. ture in a hospitalized patient who has not
tients. J Trauma 1998; 44:868 873 Crit Care Med 1995; 23:1638 1652 undergone neurosurgery? Clin Infect Dis
50. Sacco RL, VanGool R, Mohr JP, et al: Non- 65. Segatore M, Way C: The Glasgow Coma 1997; 25:285288
traumatic coma. Glasgow Coma Score and Scale: Time for change. Heart Lung 1992; 82. Hasbun R, Abrahams J, Jekel J, et al: Com-
coma etiology as predictors of 2-week out- 21:548 557 puted tomography of the head before lum-
come. Arch Neurol 1990; 47:11811184 66. Sternbach GL: The Glasgow Coma Scale. bar puncture in adults with suspected men-
51. Rordorf GW, Koroshetz W, Efird JT, et al: J Emerg Med 2000; 19:6771 ingitis. N Engl J Med 2001; 345:17271733
Predictors of mortality in stroke patients 67. Jagger J, Jane JA, Rimel R: The Glasgow 83. Towne AR, Waterhouse EJ, Boggs JG, et al:
admitted to an intensive care unit. Crit Coma Scale: To sum or not to sum? Lancet Prevalence of nonconvulsive status epilep-
Care Med 2000; 28:13011305 1983; 2:97 ticus in comatose patients. Neurology 2000;
52. Steiner T, Mendoza G, De Georgia M, et al: 68. Shackford S, Wald SL, Ross SE, et al: The 54:340 345
Prognosis of stroke patients requiring me- clinical utility of computed tomographic 84. Claassen J, Mayer SA, Kowalski RG, et al:
chanical ventilation in a neurological criti- scanning and neurologic examination in Detection of electrographic seizures with
cal care unit. Stroke 1997; 28:711715 the management of patients with minor continuous EEG monitoring in critically ill
53. Qureshi AI, Sung GY, Razumovsky AY, et al: head injuries. J Trauma 1992; 33:385394 patients. Neurology 2004; 62:17431748
Early identification of patients at risk for 69. Borczuk P: Predictors of intracranial injury 85. Varelas PN, Spanaki MV, Hacein-Bey L, et
symptomatic vasospasm after aneurysmal in patients with mild head trauma. Ann al: Emergent EEG: Indications and diagnos-
subarachnoid hemorrhage. Crit Care Med Emerg Med 1995; 25:731736 tic yield. Neurology 2003; 61:702704
2000; 28:984 990 70. Liu GT: Coma. Neurosurg Clin N Am 1999; 86. Kaplan PW: The EEG in metabolic enceph-
54. Tuhrim S, Dambrosia JM, Price TR, et al: 10:579 586, vii-viii alopathy and coma. J Clin Neurophysiol
Intracerebral hemorrhage: External valida- 71. North JB, Jennett S: Abnormal breathing 2004; 21:307318
tion and extension of a model for prediction patterns associated with acute brain dam- 87. Kaplan PW, Genoud D, Ho TW, et al: Etiol-
of 30-day survival. Ann Neurol 1991; 29: age. Arch Neurol 1974; 31:338 344 ogy, neurologic correlations, and prognosis
658 663 72. Rafanan AL, Kakulavar P, Perl J II, et al: in alpha coma. Clin Neurophysiol 1999;
55. Tuhrim S, Dambrosia JM, Price TR, et al: Head computed tomography in medical in- 110:205213
Prediction of intracerebral hemorrhage sur- tensive care unit patients: clinical indica- 88. Chien LT, Boehm RM, Robinson H, et al:
vival. Ann Neurol 1988; 24:258 263 tions. Crit Care Med 2000; 28:1306 1309 Characteristic early electroencephalo-
56. van de Beek D, de Gans DJ, Spanjaard L, et 73. Wilberger JE Jr, Deeb Z, Rothfus W: Mag- graphic changes in herpes simplex enceph-
netic resonance imaging in cases of severe alitis. Arch Neurol 1977; 34:361364
al: Clinical features and prognostic factors
head injury. Neurosurgery 1987; 20: 89. Smith JB, Westmoreland BF, Reagan TJ, et
in adults with bacterial meningitis. N Engl
571576 al: A distinctive clinical EEG profile in her-
J Med 2004; 351:1849 1859
74. Yokota H, Kurokawa A, Otsuka T, et al: pes simplex encephalitis. Mayo Clin Proc
57. Bastos P, Sun PGX, Wagner DP, et al: Glas-
Significance of magnetic resonance imag- 1975; 50:469 474
gow Coma Scale score in the evaluation of
ing in acute head injury. J Trauma 1991; 90. Jennett B, Bond MR: Assessment of out-
outcome in the intensive care unit: Find-
31:351357 come after severe brain damage. Lancet
ings from the Acute Physiology and Chronic
75. Restrepo L, Wityk RJ, Grega MA, et al: Dif- 1975; 1:480 484
Health Evaluation III study. Crit Care Med
fusion- and perfusion-weighted magnetic 91. Jennett B, Snoek J, Bond MR, et al: Disabil-
1993; 21:1459 1465
resonance imaging of the brain before and ity after severe head injury: Observations on
58. Knaus W, Wagner DP, Draper EA, et al: The
after coronary artery bypass grafting sur- the use of the Glasgow Outcome Scale.
APACHE III prognostic system. Risk predic-
gery. Stroke 2002; 33:2909 2915 J Neurol Neurosurg Psychiatry 1981; 44:
tion of hospital mortality for critically ill
76. Wityk RJ, Goldsborough MA, Hillis A, et al: 285293
hospitalized adults. Chest 1991; 100:
Diffusion- and perfusion-weighted brain 92. Kaye AH, Andrewes D: Glasgow Outcome
1619 1636 magnetic resonance imaging in patients Scale: Research scale or blunt instrument?
59. Knaus WA, Draper EA, Wagner DP, et al: with neurologic complications after cardiac Lancet 2000; 356:1540 1541
APACHE II: A severity of disease classifica- surgery. Arch Neurol 2001; 58:571576 93. Emanuel LL, Barry MJ, Stoeckle JD, et al:
tion system. Crit Care Med 1985; 13: 77. Finelli PF, Uphoff DF: Magnetic resonance Advance directives for medical careA case
818 829 imaging abnormalities with septic enceph- for greater use. N Engl J Med 1991; 324:
60. Knaus WA, Zimmerman JE, Wagner DP, et alopathy. J Neurol Neurosurg Psychiatry 889 895
al: APACHE-acute physiology and chronic 2004; 75:1189 1191 94. Thompson BT, Cox PN, Antonelli M, et al:
health evaluation: A physiologically based 78. Hollinger P, Zurcher R, Schroth G, et al: Challenges in end-of-life care in the ICU:
classification system. Crit Care Med 1981; Diffusion magnetic resonance imaging find- Statement of the 5th International Consen-
9:591597 ings in cerebritis and brain abscesses in a sus Conference in Critical Care: Brussels,
61. Le Gall JR, Loirat P, Alperovitch A, et al: A patient with septic encephalopathy. J Neu- Belgium, April 2003: Executive summary.
simplified acute physiology score for ICU rol 2000; 247:232234 Crit Care Med 2004; 32:17811784
patients. Crit Care Med 1984; 12:975977 79. Practice parameters: Lumbar puncture 95. Bates D: The prognosis of medical coma.
62. Le Gall JR, Lemeshow S, Saulnier F: A new (summary statement). Report of the Quality J Neurol Neurosurg Psychiatry 2001;
Simplified Acute Physiology Score (SAPS II) Standards Subcommittee of the American 71(Suppl 1):i20 i23
based on a European/North American mul- Academy of Neurology. Neurology 1993; 43: 96. Medical aspects of the persistent vegetative
ticenter study. JAMA 1993; 270:29572963 625 627 state (2): The Multi-Society Task Force on
63. Vincent JL, Moreno R, Takala J, et al: The 80. Adelson-Mitty J, Fink MP, Lisbon A: The PVS. N Engl J Med 1994; 330:15721579
SOFA (Sepsis-related Organ Failure Assess- value of lumbar puncture in the evaluation 97. Levy DE, Caronna JJ, Singer BH, et al: Pre-
ment) score to describe organ dysfunction/ of critically ill, non-immunosuppressed, dicting outcome from hypoxic-ischemic
failure. On behalf of the Working Group on surgical patients: A retrospective analysis of coma. JAMA 1985; 253:1420 1426
Sepsis-Related Problems of the European 70 cases. Intensive Care Med 1997; 23: 98. Madl C, Grimm G, Kramer L, et al: Early
Society of Intensive Care Medicine. Inten- 749 752 prediction of individual outcome after car-
sive Care Med 1996; 22:707710 81. Metersky ML, Williams A, Rafanan AL: Ret- diopulmonary resuscitation. Lancet 1993;
64. Marshall JC, Cook DJ, Christou NV, et al: rospective analysis: Are fever and altered 341:855 858
Multiple organ dysfunction score: A reliable mental status indications for lumbar punc- 99. Zandbergen EG, de Haan RJ, Stoutenbeek

40 Crit Care Med 2006 Vol. 34, No. 1

CP, et al: Systematic review of early predic- al: Initial CT findings in 753 patients with outcome after cardiac arrest. Crit Care Med
tion of poor outcome in anoxic-ischaemic severe head injury. A report from the NIH 1997; 25:11331138
coma. Lancet 1998; 352:1808 1812 Traumatic Coma Data Bank. J Neurosurg 116. Schoerkhuber W, Kittler H, Sterz F, et al:
100. Attia J, Cook DJ: Prognosis in anoxic and 1990; 73:688 698 Time course of serum neuron-specific eno-
traumatic coma. Crit Care Clin 1998; 14: 109. Adams HP Jr, Kassell NF, Torner JC: Use- lase. A predictor of neurological outcome in
497511 fulness of computed tomography in predict- patients resuscitated from cardiac arrest.
101. Young GB, Wang JT, Connolly JF: Prognos- ing outcome after aneurysmal subarach- Stroke 1999; 30:1598 1603
tic determination in anoxic-ischemic and noid hemorrhage: A preliminary report of 117. Vos PE, Lamers KJ, Hendriks JC, et al: Glial
traumatic encephalopathies. J Clin Neuro- the Cooperative Aneurysm Study. Neurol- and neuronal proteins in serum predict out-
physiol 2004; 21:379 390 ogy 1985; 35:12631267 come after severe traumatic brain injury.
102. Robinson LR, Micklesen PJ, Tirschwell DL, 110. Cusumano S, Paolin A, Di Paola F, et al: Neurology 2004; 62:13031310
et al: Predictive value of somatosensory Assessing brain function in post-traumatic 118. Rosen H, Sunnerhagen KS, Herlitz J, et al:
evoked potentials for awakening from coma by means of bit-mapped SEPs, BAEPs, Serum levels of the brain-derived proteins
coma. Crit Care Med 2003; 31:960 967 CT, SPET and clinical scores. Prognostic S-100 and NSE predict long-term outcome
103. Gennarelli TA, Spielman GM, Langfitt TW, implications. Electroencephalogr Clin Neu- after cardiac arrest. Resuscitation 2001; 49:
et al: Influence of the type of intracranial rophysiol 1992; 84:499 514 183191
lesion on outcome from severe head injury. 111. Levin HS, Williams D, Crofford MJ, et al: 119. Pelinka LE, Kroepfl A, Leixnering M, et al:
J Neurosurg 1982; 56:26 32 Relationship of depth of brain lesions to GFAP versus S100B in serum after trau-
104. Teasdale E, Cardoso E, Galbraith S, et al: CT consciousness and outcome after closed matic brain injury: Relationship to brain
scan in severe diffuse head injury: Physio- head injury. J Neurosurg 1988; 69:861 866 damage and outcome. J Neurotrauma 2004;
logical and clinical correlations. J Neurol 112. Kampfl A, Schmutzhard E, Franz G, et al: 21:15531561
Neurosurg Psychiatry 1984; 47:600 603 Prediction of recovery from post-traumatic 120. Karkela J, Pasanen M, Kaukinen S, et al:
105. Teasdale G, Teasdale E, Hadley D: Com- vegetative state with cerebral magnetic- Evaluation of hypoxic brain injury with spi-
puted tomographic and magnetic resonance resonance imaging. Lancet 1998; 351: nal fluid enzymes, lactate, and pyruvate.
imaging classification of head injury. J Neu- 17631767 Crit Care Med 1992; 20:378 386
rotrauma 1992; 9(Suppl 1):S249 S257 113. Saunders DE, Clifton AG, Brown MM: Mea- 121. Zandbergen EG, de Haan RJ, Hijdra A: Sys-
106. Hashimoto T, Nakamura N, Richard KE, et surement of infarct size using MRI predicts tematic review of prediction of poor out-
al: Primary brain stem lesions caused by prognosis in middle cerebral artery infarc- come in anoxic-ischaemic coma with bio-
closed head injuries. Neurosurg Rev 1993; tion. Stroke 1995; 26:22722276 chemical markers of brain damage.
16:291298 114. Chalela JA, Wolf RL, Maldjian JA, et al: MRI Intensive Care Med 2001; 27:16611667
107. Ruff RM, Marshall LF, Crouch J, et al: Pre- identification of early white matter injury in 122. Becker KJ, Ulatowski JA: Disorders of con-
dictors of outcome following severe head anoxic-ischemic encephalopathy. Neurol- sciousness. In: Current Therapy in Neuro-
trauma: Follow-up data from the Traumatic ogy 2001; 56:481 485 logic Disease. Current Therapy Series. Fifth
Coma Data Bank. Brain Inj 1993; 7:101111 115. Fogel W, Krieger D, Veith M, et al: Serum Edition. Johnson RT, Griffin JW (Eds). St.
108. Eisenberg HM, Gary HE Jr, Aldrich EF, et neuron-specific enolase as early predictor of Louis, MO, Mosby, 1997

Crit Care Med 2006 Vol. 34, No. 1 41