Coma: A Systematic Approach to Patient Evaluation and Management

Author: Jonathan M. Glauser, MD, FACEP, Attending Physician, Cleveland Clinic Foundation, Department of Emergency Medicine; Faculty, MetroHealth Residency Program in Emergency Medicine, Cleveland, OH.

Peer Reviewers: Andrew D. Perron, MD, FACEP, Assistant Professor of Emergency Medicine and Orthopedic Surgery, Department of Emergency Medicine, University of Virginia, Charlottesville; and Jeffrey J. Bazarian, MD, FACP, Assistant Professor, Department of Emergency Medicine, University of Rochester, NY.

Coma is a presenting symptom in approximately 0.5-1% of emergency department (ED) admissions,1 although the only paper addressing frequency of coma in the ED dates from 1934, citing coma as the presentation in 3% of admissions to the ED.2 A more recent retrospective analysis found alteration of mental status in between 4% and 10% of ED patients.3 Disturbances of level of consciousness may be caused by a wide variety of disorders, ranging from systemic disorders to structural central nervous system (CNS) disorders.The Editor


A diminished level of consciousness, encompassing terms such as drowsiness, stupor, and coma, must be distinguished from clouding of consciousness, or confusional states, which entail content of consciousness. Confusion represents the inability to maintain a coherent sequence of thoughts, usually accompanied by inattention and disorientation. There is reduced memory, awareness, mental clarity, and coherence. This is more a disorder of the content of consciousness. Confusional states may be produced by many of the same medical disorders that produce diminished consciousness.

Definitions are, of necessity, inexact. For purposes of documentation it always is desirable to confirm and document specifics regarding what the patient is able to do at any given time. Establishing a baseline and describing changes in the patient’s condition mandates this. Terms such as stupor, obtundation, and lethargy have been used so inexactly that they often have little meaning. For purposes of discussion, the following terms have been used, although objective assessments of patient capabilities and mental status always are preferable.

Clouding of consciousness represents a disturbance characterized by impaired capacity to think clearly and to perceive, respond to, and remember stimuli.

Delirium represents a state of disturbed consciousness with motor restlessness and disorientation. Delusions and hallucinations may be present. It is an acute confusional state, with maximal duration of symptoms lasting for days. Emotional lability and impaired short-term memory are present, with disturbance of all higher cortical functions. Fever, tachycardia, and hypertension are common.

Obtundation is a state in which a patient is awake but not alert. Psychomotor retardation is present.

Drowsiness or lethargy is a disorder that simulates light sleep. The patient is arousable by touch or noise and can maintain alertness for a period of time.

Stupor is a state in which the patient can be aroused only by vigorous stimuli. Efforts to avoid stimulation are displayed. The patient exhibits little or no spontaneous activity, and shows little motor or verbal activity once aroused.

Coma indicates a state in which the patient is not arousable at all to verbal or physical stimuli, and no attempt is made to avoid painful or noxious stimuli. This has been subdivided further into light coma, in which patients respond to noxious stimuli with a variety of protective reflexes, and deep coma, in which patients do not respond at all.4

Coma Scales. Several methods have been proposed to assess objectively the level or grade of consciousness. Scales that assign points do not correlate with any specific terms cited above, which are subjective by necessity.

The most widely used mathematical scale to assess the patient with altered level of consciousness is the Glasgow Coma Scale (GCS). First described in 1974, it is a simple, reproducible scoring system used in trauma patients to define the level of responsiveness.5 (See Table 1.) A GCS score of 8 or lower has been used as an alternative definition of coma. An advantage of the test is that a patient’s neurologic status can be followed in an objective way. Its reproducibility is variable. It has been criticized because of the absence of brain stem signs in the evaluation.6 However, because of its ease, familiarity, and widespread acceptance, the GCS continues to be the standard coma scoring system in the United States. Other named scales for the assessment of the comatose patient include the Liege coma scale, Swedish Reaction Level Scale, and the Apache II scale.

The Swedish Reaction Level Scale (RLS 85) has been proposed for use in patients whose assessment of eye opening may be difficult because of eye swelling, or whose verbal response may be difficult due to intubation. (See Table 2.) It has been shown to correlate with the GCS.7

The AVPU scale has been utilized by some because of its simplicity:

  • A= alert and aware
  • V= responds to verbal stimuli
  • P= responds to painful stimuli
  • U= unresponsive.


Consciousness has two easily assessable components: arousal and awareness. Most simply put, arousal implies the appearance of being awake. Wakefulness and alertness are maintained by a system of upper brain stem and thalamic neurons, the reticular activating system (RAS), with its connections to the cerebral hemispheres. For this reason, reduced consciousness results from depression of either the RAS or of neuronal activity in both cerebral hemispheres. An intact brain stem is necessary for arousal to occur, and a patient who looks awake generally has an intact brain stem.

The RAS is located ventral to the ventricular system in the pons, and extends continuously up to the posterior hypothalamus and thalamus. Even small lesions in the pons are capable of inducing coma. The RAS acts as a gateway for stimuli to the cerebral cortex and as a trigger for arousal from sleep. Therefore, when the RAS is dysfunctional, the cerebral cortex cannot be aroused and coma occurs.

Alternatively, coma can occur due to failure of cognition in the cerebral cortex. Awareness, as opposed to arousal, arises from brain activity in the cerebral cortex. Awareness constitutes sensation, emotion, volition, and thought. Since awareness cannot be well localized, a unilateral cortical lesion cannot abolish it. There must be diffuse, bilateral, cortical lesions or extensive lesions disrupting traffic between the thalamus and the cortex to produce a vegetative state in which patients appear awake but are unaware of their surroundings. Bilateral cortical disease usually is secondary to toxic/metabolic derangement, and alteration in consciousness is dependent upon the size of injury and the speed with which it progresses. Unilateral hemispheric diseases, therefore, do not cause coma unless the brain stem fails. For these reasons, focal neurologic signs are expected to be absent or not prominent, except occasionally due to hypoglycemia.

Table 3 contrasts findings in patients with coma of brain stem origin with those in patients with coma from hemispheric origin.

The RAS may be disrupted or suppressed anatomically by a variety of mechanisms: Intrinsic brain stem lesions directly may compress or destroy RAS fibers, as with pontine hemorrhage, trauma, degenerative diseases, or tumor. The RAS may be disrupted by torque to the brain stem, as from a sudden blow to the head.

Infratentorial lesions may displace posterior fossa contents through the tentorial notch or the foramen magnum, as with cerebellar hemorrhage or posterior fossa tumors.

Supratentorial pressure may result from lesions that displace tissue and result in decreased perfusion and compression of brain stem structures or of the opposite hemisphere, as with subdural hematoma, intracranial hemorrhage, epidural hematoma, tumor, or brain abscess. Supratentorial lesions are a more common cause of coma than subtentorial lesions. In the uncal herniation syndrome, the most common of the herniation syndromes, the medial temporal lobe shifts to compress the upper brain stem. Typically, the ipsilateral third cranial nerve is compressed. Hemiparesis contralateral to the mass may develop from compression of the descending motor tracts in the ipsilateral cerebral peduncle, prior to their crossing in the medulla.

Any disturbance in the RAS, whether structural or functional, may produce a state resembling physiologic sleep that is not interrupted by the normal sleep-ending stimuli of pain or verbal command. Therefore, in brain stem coma, the patient is unresponsive and appears asleep. In hemispheric coma, the patient is awake and unresponsive.

Initial Assessment of the Comatose Patient

The key question in the diagnostic approach to the patient with diminished level of consciousness (LOC) is whether the patient’s altered LOC is due to toxic/metabolic causes or whether there is structural CNS disease. Structural lesions, in general, are diagnosed more urgently because of the possibility of life-saving surgical intervention. In approximately 15% of cases, coma is caused by structural lesions.4 Conversely, toxic-metabolic causes require supportive care until a definitive diagnosis is made and appropriate medical therapy can be instituted.

History. Clinicians must be able to perform a focused history and rapid neurologic examination that screens for most common abnormalities. Since people in coma cannot answer questions or follow commands, it becomes necessary to eliminate many parts of the neurologic examination and to seek historical data from friends, family, or other sources.

Specific questions should include the rate of onset of neurologic or behavioral changes. Abrupt onset favors the diagnosis of CNS hemorrhage or ischemia, or a cardiac cause inducing abrupt loss of cerebral perfusion. Gradual onset over days favors a metabolic problem such as hypercalcemia, diabetic emergency, or electrolyte disturbance.

Any history of trauma or ongoing medical illness, including prescribed and available medications, should be sought. Over-the-counter medications, salicylates, anticoagulants, hypoglycemic agents, antidepressants, anti-epileptic drugs, opioids, neuroleptics, steroids, or anticholinergic agents are all of potential significance in the patient with altered mental status. Any mydriatic or miotic agent can affect the neurologic examination.8

Suicidal ideation, past attempts at self-harm, and any history of substance abuse are critical considerations.

The setting in which the patient was found may be a clue to exposure to products of combustion or to extremes of temperature. Changes in activities of daily living or recent alteration in neurologic status may be reported from others. The social history may indicate substance abuse or HIV risk.

The patient’s belongings should be searched for pill bottles. The patient may have a Medic Alert bracelet indicating prior seizure disorder, diabetes, or renal failure.

Physical Examination. The focus of the examination is to distinguish metabolic causes for coma from structural ones, since the former generally require supportive medical care, whereas the latter may require urgent surgical intervention. Specific signs of brain stem herniation, including respiratory pattern, motor posturing, pupillary abnormalities, and extra-ocular muscle assessment, are critical to recognize, as rapid diagnosis of intracranial mass lesions may be life-saving.

The Role of Vital Signs in the Comatose Patient. Vital signs should be taken, including an accurate core temperature utilizing a probe that reads under 34°C, if necessary. Hypothermia is suggestive of sepsis, hypothyroidism, hypoglycemia, or environmental exposure. Hyperthermia or hyperpyrexia is suggestive of CNS infection, sepsis, serotonin syndrome, neuroleptic malignant syndrome, heat stroke, thyrotoxicosis, stroke, or exposure to certain toxins, such as salicylates, anticholinergics, or sympathomimetics.8

Systolic blood pressure of over 200 mmHg, with a diastolic of over 130 mmHg, is suggestive of intracranial hemorrhage, thyrotoxicosis, or exposure to sympathomimetic agents. Bradycardia is suggestive of increased intracranial pressure, hypothyroidism, or exposure to toxins such as calcium channel blockers or beta-blocking medications. The patient’s breath may suggest fetor hepaticus, uremia, acetone, or alcohol.

Slow respiratory rate suggests opiate or sedative-hypnotic poisoning. Rapid respiratory rate or hyperventilation may be an indicator of hypoxemia or acidosis. Table 4 gives a differential diagnosis of abnormal respiratory rate in the unresponsive patient.

Certain respiratory patterns classically have been described in comatose states. Cheyne-Stokes Respiration (CSR) is a pattern of periodic breathing in which phases of hyperpnea alternate with apnea. It occurs in some mild drug intoxications, as part of the aging process, and in a variety of states associated with prolonged circulation time, including congestive heart failure, and during sleep in some normal persons. CSR requires generally intact brain stem function. The presence of CSR in the comatose patient implies bilateral dysfunction of the cerebral hemispheres or diencephalon.9

Central neurogenic or primary hyperventilation (CNH) has been described in low midbrain and middle pons lesions. Hyperventilation in a comatose patient usually implies some sort of acidosis, perhaps from diabetic ketoacidosis or lactic acidosis, or from some ingested toxin such as methanol, aspirin, or ethylene glycol. (See Table 4.)

Other brain stem respiratory patterns have been described. Apneustic breathing represents a pause at full inspiration, signifying an upper pons lesion. This long pause, also called inspiratory cramp, with a prolonged exhalation, typically is repeated 5-6 times per minute. Clusters of breaths may follow each other in a disorderly fashion with high medullary or lower pons lesions. Ataxic breathing, or Biot’s respiration (see Table 5), has a completely irregular pattern in which deep and shallow breaths occur randomly, indicative of a medullary disorder. These patterns are unusual; the importance to the emergency physician is to recognize brain stem disorders and to pursue diagnoses aggressively. Patients with apneustic breathing or ataxic breathing require intubation.10

Other Physical Findings. The patient should be fully exposed. Skin findings such as cyanosis, pallor, needle tracks, uremic frost, or icterus should be noted. Petechiae may suggest meningococcemia or Rocky Mountain spotted fever.

Examination of the head should concentrate on evidence for trauma, such as hemotympanum, CSF rhinorrhea or otorrhea, mastoid ecchymosis (Battle’s sign), or depressed skull fracture. The patient may have evidence of shunt placement or prior surgery. Funduscopic examination may show hemorrhage or papilledema. Occasionally, subhyaloid hemorrhages can been seen behind the vitreous humor along the edge of the optic disc. This finding is pathognomonic for subarachnoid hemorrhage.

If the cervical spine can be cleared of injury by imaging or by history, meningeal signs may be assessed for evidence of subarachnoid hemorrhage. The neck should be examined for thyroidectomy scar as a possible clue for myxedema coma.

The abdominal examination may reveal ascites or organomegaly suggestive of hepatic encephalopathy. Table 5 summarizes clues to coma based upon the physical examination.

The Neurologic Examination. Given that the region of the nervous system affected in a comatose patient is in the brain stem or above, the neurologic examination should be focused and will differ from that of the awake, conversant patient.

In the alert patient, the mental status examination evaluates speech flow and content; memory; orientation to person, place, and time; and thought process. The patient may be asked to count backward from 20, then by serial 3s or 7s. Memory may consist of recent recall of three unrelated objects.

In the unresponsive patient, posture should be observed. If the patient is in a comfortable-appearing position, the brain stem probably is functioning well. Spontaneous reflexes should be noted. Coughing, sneezing, or hiccuping carries a less positive prognosis.10

The mental status examination in the comatose patient is an assessment of the patient’s response to auditory, visual, and noxious stimuli. Stimuli may range from verbal stimuli to more noxious stimuli, such as tickling the nasopharynx with a cotton swab.

Two specific patterns of motor activity are notable: Decorticate posturing consists of upper extremity adduction with flexion at the elbows, wrists, and fingers. There is lower extremity extension and internal rotation. This indicates disease of the diencephalon. Decerebrate posturing consists of upper extremity adduction, extension, and pronation. Lower extremity extension is present. There is internal rotation and extension of all four limbs. This posture indicates that the vestibular nuclei in the medulla and the vestibulospinal tracts are intact.

While these postures do not have localizing value in human patients, in general, patients with decorticate posturing in response to pain have a better prognosis than do those with decerebrate posturing.

Formal strength testing is impossible in comatose patients, but deep tendon reflexes and resistance to passive manipulation still can be tested in the usual manner. The degree to which movement is purposeful should be recorded. A purely local withdrawal reflex from noxious stimulation of the lower extremity produces triple flexion: dorsiflexion of the ankle with flexion of the knee and hip. To look for purposeful withdrawal, the stimulus should be applied to a location where triple flexion would be an inappropriate response. For example, pinching the skin of the anterior thigh would elicit hip extension as a purposeful movement, while hip flexion would represent purely reflex withdrawal. In both upper and lower extremities, reflex withdrawal produces limb adduction. Therefore, to differentiate reflex withdrawal from purposeful movement, the noxious stimulus should be applied to the medial aspect of the limb.11 Abduction of a limb is a purposeful mediated cortical response. Triple flexion withdrawal of the lower extremity is a spinal cord-mediated reflex.

The cranial nerve examination of the patient in coma should concentrate on the eyes. Although visual acuity cannot be tested, testing of field defects may be performed by assessing response to visual threat. A finger or small object introduced suddenly into the visual field may elicit a response.

Pupillary reflexes can be tested in the same manner as in an awake person. Pupillary abnormalities or focal neurologic findings suggest an intracranial cause for coma. A unilateral fixed, dilated pupil often is taken to be a sign of brain herniation, with the causative mass usually on the same side as the pupillary abnormality. Consider the possibility of topically applied mydriatic agents in an alert patient or, alternatively, compression of CN III by a posterior communicating artery aneurysm. This finding in a comatose patient suggests the need for urgent computed tomography (CT) to assess for space-occupying lesions such as hemorrhage, abscess, or tumor. Pinpoint pupils may result from opiates other than meperidine or from pontine lesions. Finally, pupillary dilatation may be elicited from a painful stimulus below the neck, as with pinching (ciliospinal reflex).

Eye movements are the cornerstone of the neurologic examination of the comatose patient, as they closely approximate the ascending RAS anatomically. These can be assessed by activating certain specific reflexes. Cranial nerves III, IV, and VI enter and exit the brain stem at the lower pontine to upper midbrain levels, closely associated with the RAS, making testing of extra-ocular movements critical. Extraocular movement testing permits evaluation of the cortex, and of the medial longitudinal fasciculus in the brain stem, as evidenced by the function of cranial nerves III, IV, and VI. If the patient’s eyes rove spontaneously and conjugately, the brain stem probably is intact. A persistently adducted eye indicates paresis of cranial nerve VI, while an abducted eye indicates cranial nerve III dysfunction. Dysconjugate gaze in the horizontal plane may be observed in various sedated states, including alcohol intoxication. Dysconjugate gaze in the vertical plane may be caused by pontine or cerebellar lesions and is termed skew deviation. Conjugate eye movements or horizontal roving imply an intact midbrain and pons. Ocular bobbing may indicate bilateral pontine damage, cerebellar hemorrhage, or metabolic derangement.4 Typical ocular bobbing with pontine lesions is characterized by conjugate downward jerks followed by slow but conjugate return to the midposition.12

The oculocephalic reflex, or doll’s eye reflex, is tested by turning the patient’s head from side to side. This should be performed only if there is no chance of a traumatic injury to the cervical spine. Imaging of the cervical spine may have to be performed to rule out injury. A positive doll’s eye test occurs when the eyes do not turn with the head; in other words, it appears that the patient is maintaining fixation on a single point in space. The eyes appear to be moving relative to the head in the direction opposite to the head movement. This reflex usually is suppressed in conscious patients, and is a normal finding in comatose patients. Absence of this reflex in a comatose patient indicates dysfunction somewhere in the reflex pathway. The abnormality may be in the afferent limb, from the labyrinth and vestibular nerve, or from neck proprioceptors. Alternatively, the lesion may be in the efferent limb, including the medial longitudinal fasciculus (MLF), cranial nerves III and VI, or in the muscles they innervate. The function of the MLF is to coordinate head, neck, and eye movement while receiving input from the horizontal semicircular canals. It extends the eyes’ input from the superior colliculus high in the midbrain into the cervical spinal cord. Finally, the dysfunction may be in the connecting pathways in the pons and medulla. An intact doll’s eye reflex demonstrates the functional integrity of the brain stem, and the cause for coma should be sought elsewhere.

An alternative to doll’s eye reflex testing is the vestibulo-ocular reflex, commonly referred to as cold calorics. This may be performed if trauma is suspected, as the test does not entail moving the patient’s neck. The patient is placed supine, with the head or upper body tilted forward so that the neck forms an angle of 30° with the horizontal. This isolates endolymph movement to the horizontal semicircular canal. The tympanic membrane is examined to verify that it is intact prior to the test. A syringe is filled with 25-50 cc of ice water with a small catheter attached, and the water forcefully is injected against the tympanic membrane. This stimulus should have the same effect on the horizontal semicircular canal as sustained turning of the head in the opposite direction, resulting in sustained deviation of the eyes toward the ear being stimulated. Movement begins typically 10-30 seconds after the ice water is introduced. There are several possible results from the test:

1. Absence of this reflex indicates dysfunction of the pons, medulla, or, less commonly, cranial nerves III, VI, or VIII.

2. Normally, this reflex produces deviation of the eyes toward the stimulated ear. There is nystagmus, with the fast component away from the stimulated ear. Bilateral tonic deviation of the eyes toward the stimulus should last 30-120 seconds and indicates an intact brain stem. Absence of the quick, corrective nystagmus indicates damage to the cerebral hemispheres. A normal response in a comatose patient raises the suspicion of psychogenic coma. While the eyes deviate toward the side of the cold water infusion, a mnemonic for the nystagmus is that the rapid nystagmus is away from the side of the ice water infusion, with the slow tonic deviation toward the stimulated side: COWS (cold opposite, warm same).

3. If both eyes deviate tonically toward the side of the cold-water infusion, the brain stem is intact, but the source of coma may be from cerebral hemisphere dysfunction, failing to correct eye position with rapid nystagmus.

4. Movement of only one eye ipsilateral to the stimulus signifies internuclear ophthalmoplegia—a brain stem structural lesion.

Benzodiazepenes, barbiturates, and alcohol may affect reflex eye movements, although they leave pupillary responsiveness intact.

In the setting of trauma, CN VI palsy may result from a brain stem lesion, peripheral nerve injury with or without basilar skull fracture, or from lateral rectus muscle entrapment.13 Because of the close proximity of these structures, a traumatic lesion to the brain stem at the level of the sixth nerve may be accompanied by pyramidal tract signs, as well as a seventh nerve palsy.14

Cranial nerves V and VII can be assessed in comatose patients by testing corneal reflexes and by observing facial grimacing in response to noxious stimulation, such as supraorbital pressure. Cranial nerves IX and X can be assessed by testing the gag reflex, although this may be absent in 20% of normal subjects.11

Classically, patients with progressive metabolic encephalopathy or transtentorial herniation of the brain stem demonstrate a rostrocaudal deterioration. This may be identifiable through four functional levels: diencephalon, midbrain, pons, and medulla. Those patients who are functioning at the highest level (diencephalon) have the best prognosis. (See Table 6.) A mnemonic for functions tested is SPERM: State of consciousness; Pupils and their reactivity; Eye movements; Respiratory pattern; and Motor response.

Emergency Management of the Comatose Patient

As with any medical emergency, establishment of an airway is paramount. Any patient with a GCS of less than 9 should undergo intubation. Patients should be placed on oxygen and fully undressed. If there is any suspicion of cervical spine (C-spine) injury, the C-spine should be immobilized. The neck cannot be moved until there is unimpeachable historical or radiographic evidence that the patient did not sustain neck trauma.

Most causes for coma are non-traumatic and involve medical management and supportive care with specific care dictated by diagnosis. Ventilation and establishment of circulation are paramount. While detailed trauma management is outside the purview of this article, supportive care for the trauma patient includes airway management, including intubation and ventilation, blood and fluid resuscitation, control of external bleeding, and control of agitation.

Primary brain injury occurs with the initial insult, and is best delineated in the ED by CT. Secondary brain injury is defined as those processes that occur later that contribute to overall traumatic brain injury: from decreased cerebral blood flow, elevated intracranial pressure, hemorrhage, and failure to adequately address shock.15,16 It now is recognized that excessive hyperventilation causing constriction of cerebral blood vessels may be a factor in reducing cerebral blood flow, especially if the pCO2 is decreased below 25 mmHg.17,18 Hyperventilation should be viewed as a temporizing measure only in the face of worsening neurologic status, with close monitoring and return of the pCO2 to 30-35 mmHg as soon as possible.19 Mannitol 0.25-1 g/kg has an onset of action of osmotic diuresis within 30 minutes, and may be viewed as another temporizing measure in the ED. Steroids such as dexamethasone have a role in vasogenic cerebral edema, as from tumor, but no role in the management of head injury.20,21 The mean arterial blood pressure should be maintained at 90 mmHg. If there are signs of herniation, any potentially surgical lesion should be identified as rapidly as possible. Otherwise, treatment for elevated intracranial pressure entails elevation of the head of the bed to 30°, adequate volume resuscitation, and maintenance of arterial oxygenation.

The Coma Cocktail

Exposure to drugs and toxins is reported to occur in approximately 2 million Americans annually.22 As well, one report identified 29 patients with hypoglycemia out of 340 consecutive paramedic calls for patients with altered mental status.23 There has been ongoing interest in the use of the "coma cocktail" in patients with altered consciousness. This has dated from the 1960s when analeptic agents such as doxapram hydrochloride, picrotoxin, nikethimide, caffeine, and physostigmine were administered to combat CNS depression. Analeptics have fallen out of favor since they have been demonstrated to increase morbidity and mortality, largely because of their induction of seizures.24,25 Physostigmine usage empirically is mentioned only to be condemned. Proposed components of a reasonable coma cocktail are reviewed in turn.26

Intravenous Glucose. Early use of 25-50% dextrose should be considered to prevent the sequelae from prolonged neuroglycopenia. A normal or elevated bedside glucose test strip is a valid reason to withhold glucose administration in the comatose patient. The prompt response of coma with dextrose confirms hypoglycemia and may obviate the need for any further diagnostic testing. It is notable that patients with poorly controlled diabetes mellitus may experience symptoms of hypoglycemia at greater glucose concentrations than non-diabetics—at average levels of 77 mg/dL in one report.27 At equilibrium, one ampule of 50% dextrose in water, or 25 grams, should raise the serum glucose of a 70-kg patient acutely by approximately 60 mg/dL if it distributes into total body water prior to any metabolism. A much higher rise in serum glucose has been demonstrated, however, of 165 mg/dL when 25 grams of dextrose was administered to adult patients.28 There have been reports that hyperglycemia adversely affects outcomes after ischemic injury,29-31 perhaps related to elevated lactate levels in ischemic tissue. Since even mild degrees of hyperglycemia may result in accentuated neurologic damage from an ischemic insult, it has been proposed that empiric administration of glucose should be avoided in patients at risk for cerebral ischemia. This may include those with acute stroke, cardiac arrest, or severe hypotension.32 It is logical to rapidly and accurately detect hypoglycemia prior to the administration of 50% dextrose in water to a patient with diminished level of consciousness.33

Thiamine. Thiamine functions as a co-factor in enzymes in the Kreb’s cycle and the pentose phosphate pathway. Thiamine requirements, therefore, are dependent upon glucose intake. It is prudent to administer 100 mg of thiamine intravenous (IV) or intramuscular (IM), as acute Wernicke’s encephalopathy has been described following glucose loading.34 Although Wernicke’s initial case descriptions in 1861 included two alcoholics, this encephalopathy has been recognized in non-alcoholic individuals at risk for malnutrition, including those patients with hyperthyoid state, neoplasia, anorexia nervosa, hyperemesis gravidarum, acquired immunodeficiency syndrome,35,36 and a variety of gastrointestinal disorders. Ataxia, vertical and horizontal nystagmus, confusion, hypotension, and ophthalmoplegia are characteristic, but not all are reversible later with thiamine.37 Thiamine uptake into cells is slower than the entrance of dextrose into cells, so it makes little sense to withhold hypertonic dextrose until thiamine administration.38 Furthermore, the IV route has been demonstrated to be safe.39 In fact, since patients who truly need thiamine tend to be malnourished with little muscle mass, the IM route of administration is best avoided.33

Narcotic Antagonists. Naloxone is a pure narcotic antagonist, competitively blocking sites used by narcotic agents. A rapid response to naloxone may obviate the need for airway management. Dosage should be 0.1 mg/kg in children, 2-4 mg IV in adults up to a total dose of 4-6 mg. Patients who have exposure to long-acting narcotics such as methadone will require hospitalization. An hourly infusion rate should deliver a dose of naloxone that is equal to the initial dose required to produce arousal.40 Patients who have consumed shorter-acting narcotics may benefit from an observation stay, due to the short half-life (20-30 minutes, with duration of effects 45-70 minutes) of naloxone. The occasional cases of pulmonary edema or ventricular dysrhythmias seen after opioid reversal are felt to be due to surges in catecholamine levels that may follow naloxone administration.41 Naloxone has been demonstrated to be safe when used in the pre-hospital setting.42 Because of the risk of precipitating withdrawal, a low initial dose of 0.1-0.3 mg has been proposed initially to reverse respiratory depression.40

Nalmefine is a pure opiate antagonist that is structurally similar to naloxone. A 2-mg IV dose has been shown to prevent respiratory depression from fentanyl challenge in volunteers for up to eight hours after administration.43 It has been demonstrated to be safe in children.44

Flumazenil is a centrally acting benzodiazepene antagonist. It competitively blocks benzodiazepene activation of gamma aminobutyric acid synapses. The empiric use of flumazenil in the comatose patient is to be condemned because of the risk of seizures in patients chronically using benzodiazepenes, as well as patients who may have coingested cyclic antidepressants. It has use in isolated benzodiazepene overdose, in which flumazenil may reverse coma within 1-2 minutes. Fatalities from benzodiazepene overdose are rare and usually associated with aspiration.45,46 Airway management may be preferable to administration of flumazenil in this instance, as seizures may be difficult to control because of the presence of a benzodiazepene antagonist. There may be a subset of patients who safely may receive empiric flumazenil: those whose clinical picture is consistent with isolated benzodiazepene ingestion, with no suggestion of stimulant or antidepressant exposure, no history of underlying seizure disorder, and no long-term benzodiazepene use.47 The dose of flumazenil is 0.3 mg over 1 minute IV, followed by 0.3 mg every minute up to a total dose of 1 mg. So long as this dose is not exceeded, no more than 50% of benzodiazepene receptors will be occupied by this drug.48

Ancillary Diagnostic Testing

Testing in the ED. A bedside glucose level should be obtained for all patients presenting with coma. Hypoglycemia has been associated with a variety of focal deficits and neurologic findings, including hemiplegia49 and decerebrate50 and decorticate51 posturing. Pulse oximetry should be obtained for all patients with disturbed level of consciousness.

Urine easily is tested in the ED for infection, hyperglycemia, or presence of ketones.

An electrocardiogram may give evidence for the cause of coma, if characteristic alterations in electrolytes such as calcium or potassium are present. Specific evidence for hypothermia (Osborn J waves) or tricyclic antidepressant poisoning (widened QRS interval, terminal rightward QRS axis) are present.

Other testing must involve the hospital laboratory. A bicarbonate level or arterial blood gas should be obtained to assess for acidosis and for retention of carbon dioxide. Lactic acidosis from a seizure will normalize within one hour of the seizure, but may give a clue to a postictal state.

Venous or arterial carbon monoxide levels may be obtained, especially if there is a discrepancy between measured and calculated oxygen saturation. A history may indicate whether the patient lives alone, whether other people at home have become sick, or whether the patient had suicidal intent.

Thyroid function tests may be performed if myxedema is a consideration. A sensitive thyroid-stimulating hormone (TSH) test is most critical and should be available within 1-2 hours. Free T4 and T3 levels may be helpful to confirm diagnosis of thyroid dysfunction. Toxicologic screening should be focused, based upon clinical findings and vital signs.

Urinalysis may demonstrate hyperglycemia from hyperosmolar coma. Microscopy may indicate calcium oxalate crystals from ethylene glycol poisoning, or leukocytes, bacteria, and nitrites suggestive of urosepsis.

A head CT scan should be ordered when there is no evidence of metabolic cause for coma. If there is a suspicion for intracranial hemorrhage or for trauma, this study should be ordered early. Magnetic resonance imaging (MRI) may be more sensitive for brain stem lesions, but its utility is limited by availability. In the absence of herniation, the most likely cause for brain stem dysfunction is a toxin. CT can delineate surgical mass lesions, midline shift, and compression of cisterns. (See Table 7.)17

MRI is the most sensitive test for diagnosing diffuse axonal injury (DAI), which occurs in deeper cerebral white matter, where shear forces cause mechanical disruption of axons.17 Although this injury is not amenable to surgery, its presence indicates that a patient is more likely to develop persistent vegetative state or persistent coma.52,53 While MRI may demonstrate punctate areas of bleeding below the level of CT detection, its value in the evaluation of trauma is limited by availability. As well, acute brain injury amenable to surgery is well-defined by CT. MRI has value in diagnosis of extra-cranial and intra-cranial arterial dissections. Non-hemorrhagic brain lesions often are more readily identifiable by MRI. The depth of lesions in white matter, midbrain, and the brain stem can be assessed.54,55

A lumbar puncture (LP) should be considered for diagnosis of infection or subarachnoid hemorrhage. Head CT should, in general, precede LP to diagnose a mass lesion that could predispose the patient to herniation and which might be diagnostic of the patient’s coma. Negative imaging and metabolic work-up in the setting of brain stem failure implies a vascular event, such as basilar artery occlusion. Figure 1 summarizes a suggested overall diagnostic and therapeutic approach to the patient in coma.

Electroencephalography (EEG) is of limited utility in the emergency setting, but occasionally may be useful in the diagnosis of status epilepticus. Electrical seizures without corresponding motor movements make EEG the diagnostic modality of choice in this setting. It may be of prognostic utility when performed 6-72 hours after anoxic injury. Malignant EEG findings indicate a poor prognosis for recovery, including most patients with intact brain stem reflexes.56,57

Somatosensory evoked potentials are most useful in predicting a poor or fatal outcome. They are a more specific marker than the EEG, with few false positives. The bilateral absence of an evoked response after one week is a reliable predictor of failure to regain consciousness.57,58

Causes of Coma

There are numerous causes for coma, which should be classified into two broad categories: toxic-metabolic and structural.

Toxic causes for coma may include the following: alcohols, such as ethyl, isopropyl, and methyl; barbiturates; anti-cholinergics; bromides; carbon monoxide; opiates; anticonvulsants; tranquilizers; cyanide; phenothiazines; cyclic antidepressants; sympathomimetics, including cocaine, amphetamines; sedative-hypnotics; salicylates; hallucinogens; heavy metals.

Metabolic causes for coma include any disorder that induces hypoxia or hypercapnea. Acidosis, hyper- and hyponatremia, and hypo- and hypercalcemia must be considered. Hepatic failure, including Reye’s syndrome, may induce altered consciousness. Endocrine abnormalities, such as myxedema, Addisonian crisis, hypoglycemia, and hyperosmolarity, must be considered. Wernicke’s encephalopathy occurs in a variety of malnourished states. Overwhelming sepsis, meningitis, peritonitis, and encephalitis are among infectious causes to be considered.

Structural causes for coma may be divided into supratentorial disorders and infratentorial. Supratentrorial causes include bilateral cerebral disease, concussion, contusion, postictal states, hypertensive encephalopathy, encephalitis, meningitis, subarachnoid hemorrhage, generalized increased intracranial pressure, and unilateral cerebral disease causing herniation such as infarction, hemorrhage, tumor, and abscess. Infratentorial lesions that may cause coma include pontine hemorrhage, cerebellar hemorrhage, basilar artery occlusion, brain stem tumors, or traumatic hemorrhage in the posterior fossa.

Specific Scenarios Causing Coma that Require Specific Consideration

Myxedema. The hypothyroid crisis of myxedema coma is a life-threatening manifestation of the hypothyroid state. Myxedema coma can be defined as severe thyroid deficiency contributing to a decreased level of consciousness.59 The diagnosis is largely a clinical one made in the context of hypothermia, hypoventilation, stupor, and abnormal TSH and free thyroxine. Altered mental status and a precipitating event such as infection or trauma may be present. Levothyroxine or liothyronine (T3), hydrocortisone, and supportive care in the form of airway management, antibiotics, or rewarming techniques may be life-saving.60,61

Subarachnoid Hemorrhage. Subarachnoid hemorrhage may present as coma. It is important to note that the sensitivity of CT in the diagnosis of SAH approximates 93%, and declines with degree of anemia and time from the initial bleed. Lumbar puncture still is the gold standard for diagnosis if the CT is negative.62 A history of sickle cell disease, polycystic kidney, use of sympathomimetic agents, Ehlers-Danlos disease, or Marfan’s syndrome may yield clues.63 Referral for neurosurgical management is mandatory.

Elevated intracranial pressure may cause brain shift and compression through the tentorial opening, inducing brain herniation. When cardiorespiratory centers in the brain stem are compressed, respiratory arrest and cardiovascular collapse may ensue. Large hematomas must be evacuated surgically. Mannitol 0.5-1.0 gm/kg may serve as a temporizing measure to lower intracranial pressure until more definitive therapy can be accomplished. Neurological outcome was shown to improve and mortality to be reduced in patients with subdural hematomas if operative drainage and control of hemorrhage was achieved within 4 hours of injury.64

Brain edema is typically maximal 24-48 hours after injury.65 Proposed mediators of injury have included excretory amino acids, oxygen radical molecules, and nitrous oxide released from injured vascular endothelium.66 Iron released from extravasated blood promotes oxidant reactions.67 Blood in the interstitial space also triggers an inflammatory immune response regulated by local cytokines.68 Management of cerebral edema was discussed earlier.

Conditions that Mimic Coma

Catatonia is a hypomobile syndrome associated with major psychosis. Patients appear awake, but make no voluntary movements. Characteristically, there is a "waxy flexibility," in which limbs maintain their posture when placed by their examiner. It occurs chiefly in elderly patients as a manifestation of severe or vegetative depression. The catatonic patient usually becomes responsive after an injection of amobarbital, distinguishing this from severe abulia.10

Hysterical or conversion pseudocoma may entail voluntary attempts to appear comatose. Patients actively may resist eyelid opening, or may respond to visual threats. Mild stimulation, such as tickling the nose with cotton, should be employed instead of painful stimuli.

Akinetic mutism or abulic state describes a patient who shows long delays between any stimulus, such as a pinch or a question, and a reaction. Movement and speech are markedly deficient, but spontaneous visual tracking always is intact.69 Abulia denotes a form of this disorder of lesser severity. The slowness applies to verbal output as well as motor response. This is considered to be the opposite of ebullience, and in severe cases is called akinetic mutism.10 This results from frontal lobe disease, and may be caused by underlying disorders such as hydrocephalus, frontal meningioma, cerebral trauma, intracerebral hemorrhages, or cerebral infarctions in the distribution of both anterior cerebral arteries. Investigation includes CT or MRI.

Locked-in syndrome is caused by damage to the corticospinal, corticopontine, and corticobulbar tracts. Since the patient cannot move, this has been misdiagnosed as coma. However, extraocular movements are intact. Testing for this motor function establishes that, in fact, the patient is alert—with implications for what should be said in the patient’s presence. Communication may be via eye blinks or vertical eye movement. Causes include high cervical spine transection, spinal cord contusion, infarction or hemorrhage of the pons.70

Psychogenic unresponsiveness occurs rarely, and should be diagnosed only after organic causes for coma have been ruled out. A conversion reaction is the most common etiology and occurs in patients with depressive states, neuroses, or hysteric personalities. In this dissociative state, psychological stress is translated into neurologic symptomatology, such as loss of vision, loss of movement, or unresponsiveness. Once true coma has been ruled out, there have been strategies adduced to give the patient a way out of the situation while maintaining his or her dignity.10

The Role of Hypothermia in Future Therapy

The induction of moderate hypothermia has improved the outcome of patients successfully resuscitated after a cardiac arrest, even when the patient remains comatose after resuscitation.71 The target temperature is 32-34° C for 24 hours.72 One study compared neurologic outcome in patients subjected to hypothermia within 12 hours of return of spontaneous circulation (ROSC) vs. patients maintained at normothermia after ROSC. Threshold neurologic recovery was defined as discharge to home or to a rehabilitation facility. The hypothermic group had a significantly higher percentage of patients who achieved neurologic recovery.73

Prognosis for Coma

The ability to assess prognosis and predict outcome has value in making decisions about triage, transfer, and resource utilization. A good clinical outcome generally is defined as moderate disability or good recovery. A poor outcome has grouped severe disability, persistent vegetative state, and death. As many as 30-40% of survivors of severe brain injury will remain in prolonged states of severely reduced consciousness.69

In one review of 500 patients who sustained nontraumatic, anoxic brain injury, 16% of patients had a good outcome, 11 % were left with severe disability, and 73% never improved beyond a vegetative state. Of the patients in coma after one week, only 7% improved to a good recovery or moderate disability. None of the patients who were in a coma at two weeks improved beyond severe disability.74 The Glasgow Outcome Coma Scale categorizes patient functioning into five categories, and generally is applied to patients who have sustained traumatic brain injury (TBI). (See Table 8.)75 Therapy does make a difference in the management of coma. From 1966 to 1991, mortality from severe TBI fell from approximately 50% to 25% without a corresponding increase in the number of patients left with severe disability.76,77

A meta-analysis was performed in 1998 of 33 studies, which predicted a poor clinical outcome following anoxic brain injury nearly 100% of the time with:

  • no pupillary reaction after 72 hours;
  • absent motor response to pain after 72 hours;
  • burst suppression or isoelectric pattern on electroencephalogram (EEG) within the first week;
  • absence of early cortical responses to median nerve somatosensory evoked potentials within the first week.78

The presence of confounding factors, such as intoxication or shock, may make accurate and reliable prediction of outcome impossible in the comatose patient in the emergency setting.

Persistent Vegetative State and Brain Death

Patients in a persistent vegetative state represent those patients who have suffered an anoxic brain injury and who have progressed to a state of wakefulness without awareness. First described in 1972, it currently is defined as patients with:

  • no evidence of awareness of self or environment and an inability to interact with others;
  • bowel and bladder incontinence;
  • hypothalamic and brain stem autonomic function preservation to permit survival with medical and nursing care;
  • maintainence of intermittent wakefulness and sleep-wake cycles;
  • no evidence of language comprehension or expression;
  • no evidence of reproducible, purposeful, or sustained behavioral responses to visual, auditory, noxious, or tactile stimuli;
  • possible preservation of cranial nerves and spinal reflexes.79-81

Persistent vegetative state is judged to be permanent after three months if from a non-traumatic source. At any given time, there are as many as 10,000-25,000 adult patients and 4000-10,000 children in this category in this country.82 However, reliable neurologic markers presaging neurologic recovery or failure to recover have not been identified,83 and recovery of consciousness has been reported in 37 patients 12 months postinjury who were vegetative at three months.84

Brain death is defined as the cessation of cerebral and brain stem function. The patient must be normothermic and nondrugged, with no contributing metabolic derangements.81 Spinal reflexes may persist, such as the triple flexion response at the hip, knee, and ankle, but there is no respiratory drive. The Uniform Determination of Death Act in the United States mandates irreversible cessation of all functions of the entire brain and brain stem. It has been accepted by 44 states and the District of Columbia.85 The apnea test requires serial measurement of arterial pCO2, which can be measured at the bedside. The premise is to document apnea using acute hypercarbia to maximally stimulate the respiratory centers. When a determination of death is made according to accepted medical and legal standards, organs can be harvested for transplantation and life support measures removed.86

In some cases patients who meet brain death criteria may be potential organ donors. While death clearly has implications for family members, resource utilization, and organ donation, brain death probably cannot be determined in the ED. Disorders such as hypothermia and barbiturate intoxication can produce a flat EEG. Serum chemistry may give the clinician a sense of poor clinical outcome, for example if serum lactate is greater than 16 mmol/L.87 Serum concentrations of neuron-specific enolase and S-100 protein have shown promise, but cannot replace clinical evaluation at this time.88,89


Disorders of consciousness may be the manifestation of a wide variety of medical and traumatic illnesses. It is incumbent upon the emergency physician to resuscitate these patients aggressively, to identify those with a surgically correctable lesion, to provide supportive medical care, and to rapidly identify those patients who may benefit from surgical intervention.


1. Huff JS. Altered mental status and coma. In: Tintinalli J, ed. Emergency Medicine: A Comprehensive Study Guide, 5th ed. McGraw-Hill, New York 2000:1440-1449.

2. Solomon P, Aring CD. The causes of coma in patients entering a general hospital. Am J Med Sci 1934;188:805.

3. Kanich W, Brady WJ, Huff JS, et al. Altered mental status: Evaluation and etiology in the ED. Am J Emerg Med 2002 Nov 20:613-617.

4. Wolfe RE, Brown DFM. Coma and Depressed Level of Consciousness. In: Rosen P, Barkin R, eds. Emergency Medicine Concepts and Clinical Practice 5th ed. Mosby; St. Louis; 2002:137-144.

5. Teasdale G, Jennett B. Assessment of coma and impaired consciosness. A practical scale. Lancet 1974;2:81.

6. Bhatty GB, Kapoor N. The Glasgow Coma Scale: A Mathematical Critique. Acta Neurochirurgica 1993;120:132-135.

7. Johnstone AJ, Lohlun JC, Miller JD, et al. A comparison of the Glasgow Coma Scale and the Swedish Reaction Level Scale. Brain Inj 1993;7: 501-506.

8. American College of Emergency Physicians. Clinical Policy for the Initial Approach to Patients Presenting with Altered Mental Status. Ann Emerg Med 1999;33:251-281.

9. Plum F, Posner JB. The Diagnosis of Stupor and Coma, 3rd ed. Oxford University Press, New York; 1980.

10. Samuels MA. The evaluation of comatose patients. Hospital Practice, March 15,1993: 165-183.

11. Gelb DJ. The Neurologic Examination. In: Gelb DJ. Introduction to Clinical Neurology Butterworth-Heinemann; Woburn, MA 2000.

12. Drake ME, Erwin CW, Massey EW. Ocular bobbing in metabolic encephalopathy: Clinical, pathologic, and electrophysiologic study. Neurology 1982;32:1029-1031.

13. Advani RM, Baumann MR. Bilateral sixth nerve palsy after head trauma. Ann Emerg Med 2003;41:27-31.

14. Baker RS, Epstein AD. Ocular motor abnormalities from head trauma. Surv Ophthalmol 1991;35:245-267.

15. Doberstein CE, Hovda DA, Becker DP. Clinical considerations in the reduction of secondary brain injury. Ann Emerg Med 1993;22:993-997.

16. Chestnut RM, Marshall LF, Klauber MR, et al. The role of secondary brain injury in determining outcome from severe head injury. J Trauma. 1993;34: 216-222.

17. Zink BJ. Traumatic brain injury outcome: Concepts for emergency care. Ann Emerg Med 2001;37:318-332.

18. Bullock R, Chesnut RM, Clifton G, et al. Guidelines for the Management of Severe Head Injury. New York, NY: Brain Trauma Foundation; 1995.

19. Chestnut RM. Guidelines for the management of severe head injury: What we know and what we think we know. J Trauma 1997;42:S19.

20. Dearden NM, Gibson JS, McDowall DG, et al. Effects of high dose dexamethasone on outcome from severe head injury. J Neurosurg 1986;64:81.

21. Gudeman SK, Miller JD, Becker DP. Failure of high dose steroid therapy to influence ICP in patients with severe head injury. J Neurosurg 1979;51:310.

22. Litovitz TL, Clark LR, Soloway RA. 1993 Annual Report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med 1994;12:546-584.

23. Hoffman JR, Schriger DL, Votey SR, et al. The empiric use of hypertonic dextrose in patients with altered mental status: a reappraisal. Ann Emerg Med 1992;21:20-24.

24. Clemmesen C, Nilsson E. Therapeutic trends in the treatment of barbiturate poisoning: the Scandanavian method. Clin Pharmacol Ther 1961;2:220-229.

25. Rappolt TR, Gay GR, Decker WJ, et al. NAGD regimen for the coma of drug-raleted overdose. Ann Emerg Med 1980;9:357-363.

26. Doyon S, Roberts JR. Reappraisal of the "coma cocktail." Emerg Med Clin N orth Am 1994;12:301.

27. Boyle PJ, Schwartz NS, Shah SD, et al. Plasma glucose concentrations at the onset of hypoglycemic symptoms in patients with poorly controlled diabetes and in non-diabetics. N Engl J Med 1988;318:1487-1492.

28. Adler PM. Serum glucose changes after administration of 50% dextrose solution. Am J Emerg Med 1986;4:504-506.

29. Pulsinelli WA, Levy DE, Sigsbee B, et al. Increased damage after ischemia stroke in patients with hyperglycemia with or without established diabetes mellitus. Am J Med 1983;74:540-544.

30. Woo E, Chan YW, Yu YL, et al. Admission glucose levels in relation to mortality and morbidity outcome in 252 stroke patients. Stroke 1988;19: 185-191.

31. Candelise L, Landi G, Orazio EN, et al. Prognostic significance of hyperglycemia in acute stroke. Arch Neurol 1985;42:661-663.

32. Browning RG, Olson DW, Stueven HA, et al. 50% Dextrose: Antidote or Toxin? Ann Emerg Med 1990;19:683-687.

33. Hoffman RS, Goldfrank LR. The poisoned patient with altered consciousness. Controversies in the use of a "coma cocktail." JAMA 1995; 74: 562-569.

34. Drenick EJ, Joven CB, Swendseid ME. Occurrence of acute Wernicke’s Syndrome during prolonged starvation for the treatment of obesity. N Engl J Med 1966;274:937-939.

35. Reuler JB, Girard DE, Cooney TG. Wernicke’s encephalopathy. N Engl J Med 1985;312:1035-1039.

36. Butterworth RF, Gaudreau C, Vincelette J, et al. Thiamine deficiency and Wernicke’s encephalopathy in AIDS. Metab Brain Dis 1991;6:207-212.

37. Watson AJS, Walker JF, Tomkin GH, et al. Acute Wernicke’s encephalopathy precipitated by glucoes loading. Ir J Med Sci 1981;150:301-303.

38. Tate JR, Nixon PF. Measurement of Michaelis constant for human erythrocyte transketolase and thiamine diphosphate. Anal Biochem 1987;160:78-87.

39. Wrenn KD, Murphy F, Slovis CM. A toxicity study of parenteral thiamine hydrochloride. Ann Emerg Med 1989;18:867-870.

40. Goldfrank L, Weisman RS, Errick JK, et al. A dosing nomogram for continuous infusion intravenous naloxone. Ann Emerg Med 1986;15:566-570.

41. Albertson TE, Dawson A, de Latorre F, et al. Tox-ACLS: Toxicologic-Oriented Advanced Cardiac Life Support. Ann Emerg Med 2001;37:S78-S90.

42. Yealy DM, Paris PM, Kaplan RM, et al. The safety of pre-hospital naloxone administration by paramedics. Ann Emerg Med 1990;19:902-905.

43. Gal TJ, DiFazio CA. Prolonged blockade of opioid effect with oral nalmephene. Clin Pharmacol Ther 1986;40:537-541.

44. Chumpa A, Kaplan RL, Burns MM, et al. Nalmephine for elective reversal of procedural sedation in children. Am J Emerg Med 2001;19:545-548.

45. Finkle BS, McCloskey KL, Goodman LS. Diazepam and drug-associated deaths: A survey in the United States and Canada. JAMA 1979;242:429-434.

46. Serafty M, Masterson G. Fatal poisoning attributed to benzodiazepenes in Britain during the 1980s. Br J Psychiatr 1993;163:386-393.

47. Gueye PN, Hoffman JR, Taboulet P, et al. Empiric use of flumazenil in comatose patients: Limited applicability of criteria to define low risk. Ann Emerg Med 1996;27:730-735.

48. Persson A, Pauli S, Halldin CD et al. Saturation analysis os specific C Ro 15-1788 binding to the human neocortex using positron emission tomography. Hum Psychopharmacol 1989;4:21-31.

49. Wallis WE, Donaldson I, Scott RS, et al. Hypoglycemia masquerading as a cerebrovascular disease (hypoglycemic hemiplegia) Ann Neurol 1985;18: 510-512.

50. Siebert DG. Reversible decerebrate posturing secondary to hypoglycemia. Am J Med 1985;78:1036-1037.

51. Liebaldt GP, Schleip I. Aphallic syndrome following protracted hypoglycemia. Monogr Gesamtgeb Psychiatr 1977;14:37-43.

52. Maxwell WL, Povlishock JT, Graham DL. A mechanistic analysis of nondisruptive axonal injury: a review. J Neurotrauma 1997;14:419-440.

53. Povlishock JT. Pathobiology of traumatically induced axonal injury in animals and man. Ann Emerg Med 1993;22:980-986.

54. Gentry LR. Imaging of closed head injury. Radiology. 1994;191:1-17.

55. Mittl RL, Grossman RI, Hiehle JF, et al. Prevalence of MR evidence of diffuse axonal injury in patients with mild head injury and normal head CT findings. AJNR Am J Neuroradiol 1994;15:1583-1589.

56. Rothstein TL, Thomas EM, Sumi SM. Predicting outcome in hypoxic-ischemic coma: a clinical and electrophysiologic study. Electroencephalogr Clin Neurophysiol 1991;79:101.

57. Chen R, Bolton CF, Young B. Prediction of outcome in patients with anoxic coma: a clinical and electrophysiologic study. Crit Care Med 1996;24:672.

58. Pohlmann-Eden B, Dingethyal K, Bender HJ, et al. How reliable is the predictive value of SEP patterns in severe brain damage with special regard to the bilateral loss of cortical responses? Intensive Care Med 1997; 23:301.

59. Myers L, Hays J. Myxedema Coma. Critical Care Clinics 1991;7:43-56.

60. Pittman CS, Zayed AA. Myxedema Coma. Curr Ther Endocrinol Metab 1997;6:98-101.

61. Wartofsky L. Myxedema Coma. In: Braverman LE, Utiger RD, eds. The Thyroid. Philadelphia. JB Lippincott; 1996; 871-877.

62. Edlow JA, Caplan LR. Avoiding pitfalls in the diagnosis of subarachnoid hemorrhage. N Engl J Med 2000; 342: 29-36.

63. Field AG, Wang E. Evaluation of the patient with nontraumatic headache: An evidence based approach. Emerg Med Clin North Am 1999;17:127-151.

64. Seelig JM, Becker DP, Miller JD, et al. Traumatic acute subdural hematoma: major mortality reduction in comatose patients treated within four hours. N Engl J Med. 1981;304:1511-1518.

65. Betz AL, Crockard A. Brain edema and the blood brain barrier. In: Crockard A, Hayward R, Hoff JT, eds. Neurosurgery: The Scientific Basis of Clinical Practice. 2nd ed. Oxford: Blackwell Scientific; 1992: 353-372.

66. Siesjo BK. Basic mechanisms of traumatic brain damage. Ann Emerg Med. 1993;22:959-969.

67. White BC, Krause GS. Brain injury and repair mechanisms: the potential for pharmacologic therapy in closed-head trauma. Ann Emerg Med 1993;22: 970-979.

68. Ott L, McClain CJ, Gillespie M, et al. Cytokines and metabolic dysfunction after severe head injury. J Neurotrauma. 1994;11:447-472.

69. American Conress of Rehabilitation Medicine. Recommendations for use of uniform nomenclature pertinent to patients with severe alterations of consciousness. Arch Phys Med Rehabil 1995;76:205-209.

70. Carroll WM, Mastaglia FL. "Locked-in Coma" in postinfective polyneuropathy. Arch Neurol 1979;36:46-47.

71. Felberg RA, Krieger DW, Chuang R, et al. Hypothermia after cardiac arrest: Feasibility and safety of an external cooling protocol. Circulation 2001; 104:1799.

72. The Hypothermia After Cardiac Arrest Study Group. Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest. N Engl J Med 2002;346:549-556.

73. Bernard SA, Gray TW, Buist MD, et al. Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia. N Engl J Med 2002;346:557.

74. Levy DE, Bates D, Caronna JJ, et al. Prognosis in nontraumatic coma. Ann Intern Med 1981;94:293.

75. Hannay HJ, Sherer M. Outcome from head injury. In: Narayan RK, Wilberger JE Jr, Povlishock JT, eds. Neurotrauma. New York, NY: McGraw-Hill: 1996: 723-747.

76. Marshall LF, Gautile T, Klauber MR, et al. The outcome of severe closed head injury. J Neurosurg 1991;75 S28-S36.

77. Young B, Runge JW, Waxman KS, et al. Effects of pergorgotein on neurologic outcome of patients with severe head injury. JAMA 1966;276: 538-543.

78. Zandbergen EG, de Haan RJ, Stoutenbeek CP, et al. Systematic review of early prediction of poor outcome in anoxic-ischaemic coma. Lancet 1998;352:1808.

79. Medical Aspects of the Persistent Vegetative State. The Multi-Society Task Force on PVS. N Engl J Med 1994;330:1499.

80. Practice parameters for determining brain death in adults (summary statement). The Quality Standards Subcommittee of the American Academy of Neurology. Neurology 1995;45:1012.

81. Wijdicks EF. Brain death worldwide: Accepted fact but no global consensus in diagnostic criteria. Neurology 2002;58:20.

82. Multi Society Task Force on PVS. Medical aspects of the persistent vegetative state. N Engl J Med 1994;330:1499-1508.

83. Sazbon L, Grosswasser Z. Time-related sequelae of TBI in patients with prolonged post-comatose awareness (PC-U) state. Brain Injury 1991;5:3-8.

84. Choi SC, Barnes TY, Bullock R, et al. Temporal profile of outcomes in severe head injury. J Neurosurg 1994;81:169-173.

85. Uniform Determination of Death Act. Chicago: National Conference of Commissioners on Uniform Statutes, 1981.

86. Schlotzhauer AV, Liang BA. Definitions and implications of death. Hematol Oncol Clin North Am 2002;16:1-12.

87. Mullner M, Sterz F, Domanovits H, et al. The association between blood lactate concentration on admission, duration of cardiac arrest, and functional neurological recovery in patients resuscitated from ventricular fibrillation. Intensive Care Med 1997;23;1138.

88. Fogel W, Krieger D, Veith M, et al. Serum neuron-specific enolase as early predictor of outcome after cardiac arrest. Crit Care Med 1997;25:1133.

89. Rosen H, Rosengren L, Herlitz J, et al. Increased serum levels of the S-100 protein are associated with hypoxic brain damage after cardiac arrest. Stroke 1998;29:473.