Coma & the Brainstem Examination
Few bedside encounters are as time-pressured, or as revealing, as the examination of the comatose patient. With history often unavailable and the patient unable to cooperate, the examiner must let the nervous system answer the questions directly: the pupils, the eyes, the breathing pattern, and the motor response to a noxious stimulus together draw a map of which structures are working and which are not. Done well, this examination localizes the lesion along the neuraxis within minutes and separates the two great categories of coma โ diffuse and structural โ that drive everything that follows. This page walks through the central principle of consciousness, the Glasgow Coma Scale, the brainstem reflexes and the levels they test, and the herniation and brain-death frameworks that flow from them.
The Central Principle: Arousal and Content
Consciousness has two components, and both must be intact. Arousal (wakefulness) depends on the ascending reticular activating system (ARAS), a network running through the paramedian tegmentum of the upper brainstem and projecting through the thalamus to the cortex. Content (awareness, the sum of cognitive and affective function) depends on the cerebral cortex working as a whole. Coma โ a state of unarousable unresponsiveness โ therefore requires one of two things:
- Bilateral, diffuse hemispheric dysfunction โ most commonly toxic-metabolic (drugs, hypoglycemia, hepatic or uremic encephalopathy, hypoxia-ischemia, sepsis). A single hemispheric lesion does not cause coma unless it is large enough to compress the opposite side or the brainstem.
- A lesion of the brainstem ARAS or its thalamic relays โ a relatively small, strategically placed lesion in the upper pons or midbrain tegmentum can abolish arousal entirely.
This dichotomy is the organizing idea of the whole examination. A patient who is comatose with preserved, symmetric brainstem reflexes most likely has a diffuse, often toxic-metabolic process; a patient with focal brainstem signs (asymmetric pupils, lost reflexes at a discrete level, abnormal eye position) has a structural lesion until proven otherwise. The job of the examination is to decide between these and, when the cause is structural, to localize it.
The Glasgow Coma Scale
The Glasgow Coma Scale (GCS) standardizes the depth of impaired consciousness across three independently scored axes, summing to a total from 3 (no response on any axis) to 15 (fully responsive). It is best documented as its components (for example, E3 V4 M5) rather than the sum alone, because the breakdown carries more information than the total.
- Eye opening (E, out of 4): E4 spontaneous ยท E3 to speech ยท E2 to pain ยท E1 none.
- Verbal response (V, out of 5): V5 oriented ยท V4 confused conversation ยท V3 inappropriate words ยท V2 incomprehensible sounds ยท V1 none.
- Motor response (M, out of 6): M6 obeys commands ยท M5 localizes to pain ยท M4 withdraws (normal flexion) ยท M3 abnormal flexion (decorticate) ยท M2 extension (decerebrate) ยท M1 none.
Pearl: the motor score is the strongest single predictor of outcome and the most useful for tracking change over time. Always note confounders โ an intubated patient cannot be scored verbally (record "T"), and sedation or paralytics invalidate the score until they wear off.
The Brainstem Reflexes and What They Localize
The brainstem reflexes are the heart of the coma examination because each is mediated by a known segment. Tested from rostral to caudal, an intact reflex confirms that its segment โ and the tegmentum carrying the ARAS through it โ is functioning. Loss of reflexes in a rostral-to-caudal sequence is the signature of a deteriorating, herniating brain.
- Pupillary light reflex โ midbrain. Afferent CN II, efferent CN III (parasympathetic fibers from the Edinger-Westphal nucleus). An intact, reactive pupil confirms a working midbrain.
- Corneal reflex โ pons. Afferent CN V (V1, nasociliary), efferent CN VII (orbicularis oculi). Touching the cornea should produce a bilateral blink.
- Oculocephalic ("doll's eyes") and oculovestibular (cold caloric) reflexes โ pons to midbrain. These test the integrity of the tegmentum between the vestibular nuclei (pons/medulla) and the oculomotor nucleus (midbrain), integrating CN VIII input with CN III and CN VI output through the medial longitudinal fasciculus (MLF). In the comatose patient with an intact brainstem, turning the head makes the eyes move conjugately in the opposite direction (a positive doll's-eyes response); cold water irrigated against an intact tympanic membrane drives slow tonic deviation of the eyes toward the cold ear. Intact reflex eye movements are powerful evidence of an intact brainstem tegmentum. Never test oculocephalics until the cervical spine is cleared.
- Gag reflex โ medulla. Afferent CN IX (glossopharyngeal), efferent CN X (vagus). The pharyngeal contraction provoked by touching the posterior pharyngeal wall.
- Cough reflex โ medulla. Predominantly vagal (CN X): the afferent limb arises from the larynx and trachea (the internal branch of the superior laryngeal nerve), with vagal and brainstem respiratory motor output driving the cough. These are the lowest reflexes in the sequence; their loss reflects medullary involvement.
| Brainstem reflex | Level tested | Cranial nerves |
|---|---|---|
| Pupillary light reflex | Midbrain | CN II (afferent), CN III (efferent) |
| Corneal reflex | Pons | CN V (afferent), CN VII (efferent) |
| Oculocephalic / oculovestibular (caloric) | Pons to midbrain (via MLF) | CN VIII (afferent), CN III & VI (efferent) |
| Gag reflex | Medulla | CN IX (afferent), CN X (efferent) |
| Cough reflex | Medulla | CN X (vagal โ afferent via internal branch of superior laryngeal nerve, and efferent) |
Reading the Pupils
The pupils deserve their own paragraph because they often settle the structural-versus-metabolic question at a glance. As a rule, most metabolic comas preserve pupillary reactivity even when other functions are deeply depressed โ so reactive pupils in a comatose patient point toward a toxic-metabolic cause (a few exceptions exist, such as opioids producing pinpoint reactive pupils and anticholinergic or glutethimide poisoning producing fixed dilated pupils). Structural lesions, by contrast, tend to produce pupils that localize:
- Pinpoint but reactive pupils suggest a pontine lesion โ classically a pontine hemorrhage interrupting descending sympathetic fibers while sparing the midbrain parasympathetic outflow. (Examine with a magnifier; the reaction may be hard to see.)
- Midposition (about 4โ6 mm), fixed pupils suggest a midbrain lesion, where both sympathetic and parasympathetic pathways are interrupted.
- A unilateral fixed, dilated ("blown") pupil is the hallmark of CN III compression from uncal herniation; the superficially placed pupilloconstrictor fibers are squeezed first, often before the patient loses extraocular movement. This is a neurosurgical emergency.
Motor Responses and Posturing
When the patient does not follow commands, the motor response to a standardized noxious stimulus (such as supraorbital or nailbed pressure) is graded along a hierarchy that descends through the neuraxis. From best to worst:
- Localizing โ the hand reaches toward and attempts to remove the stimulus (M5). Implies substantial preserved cortical and pathway function.
- Withdrawal โ normal flexion away from the stimulus, without purposeful localization (M4).
- Decorticate posturing (abnormal flexion, M3) โ arms flexed and adducted at the elbows with the legs extended. Implies a lesion above the red nucleus in the rostral midbrain, releasing the rubrospinal flexor drive to the upper limbs.
- Decerebrate posturing (abnormal extension, M2) โ arms extended, adducted, and pronated with the legs extended. Implies a lower, more caudal lesion at the level of the midbrain/upper pons, below the red nucleus. Decerebrate posturing is generally the more ominous of the two.
- No response (M1) โ flaccidity, the worst prognostic sign.
Pearl: watch the direction of evolution. A patient who progresses from localizing, to decorticate, to decerebrate, to flaccid is deteriorating caudally and may be herniating โ escalate immediately.
๐ Did You Know?
Decorticate (flexor) and decerebrate (extensor) posturing are separated by the red nucleus in the midbrain. A lesion above the red nucleus releases flexor (decorticate) posturing, while a lesion below it โ but above the vestibular nuclei โ produces extensor (decerebrate) posturing. So a patient who deteriorates from decorticate to decerebrate posturing is telling you, anatomically, that the damage is descending through the brainstem โ that the patient is herniating caudally.
Breathing Patterns
Respiratory rhythm, when not overridden by mechanical ventilation, adds another localizing layer. The classic patterns descend the neuraxis just as the reflexes do:
- Cheyne-Stokes respiration โ a crescendo-decrescendo cycle alternating with apnea, reflecting bilateral hemispheric or diencephalic dysfunction (also common in heart failure).
- Central neurogenic hyperventilation โ sustained, rapid, deep breathing, classically attributed to lesions of the rostral brainstem tegmentum (uncommon and a diagnosis made only after pulmonary and metabolic causes of hyperventilation are excluded).
- Apneustic breathing โ prolonged inspiratory pauses (inspiratory "cramps"), pointing to the pons.
- Ataxic (Biot) breathing โ chaotic, irregular breaths of varying depth, indicating medullary respiratory-center failure and often heralding respiratory arrest.
Herniation Syndromes
Mass effect drives brain tissue across the fixed dural and bony partitions of the skull, and each shift produces a recognizable syndrome:
- Uncal (transtentorial, lateral) herniation โ the medial temporal lobe (uncus) pushes over the tentorial edge and compresses the ipsilateral CN III, producing an early ipsilateral fixed, dilated pupil, then a third-nerve palsy, then progressive depression of consciousness and contralateral hemiparesis (occasionally ipsilateral hemiparesis via the Kernohan notch as a false-localizing sign).
- Central (transtentorial) herniation โ symmetric downward displacement of the diencephalon, producing a rostral-to-caudal deterioration: early small reactive pupils and Cheyne-Stokes breathing, evolving to midposition fixed pupils and decerebrate posturing as the midbrain fails.
- Subfalcine (cingulate) herniation โ the cingulate gyrus is pushed under the falx cerebri; may compress the anterior cerebral artery and cause leg-predominant weakness.
- Tonsillar herniation โ the cerebellar tonsils are forced through the foramen magnum, compressing the medulla and causing cardiorespiratory collapse โ the lethal end of posterior-fossa mass effect.
Brain Death: An Overview
Brain death (death by neurologic criteria) is the irreversible loss of all functions of the entire brain, including the brainstem. Per the 2023 AAN/AAP/CNS/SCCM Brain Death/Death by Neurologic Criteria (BD/DNC) consensus guideline, it is fundamentally a clinical diagnosis resting on three pillars:
- Coma of a known, irreversible cause, established by history, examination, and neuroimaging.
- Absence of all brainstem reflexes โ fixed pupils, absent corneal, oculocephalic and oculovestibular (caloric) responses, absent gag and cough.
- A positive apnea test โ no spontaneous respiratory effort off the ventilator despite a documented rise in arterial COโ to the target threshold: PaCOโ โฅ60 mmHg and โฅ20 mmHg above baseline, with arterial pH <7.30 (exact targets per local policy).
Before the determination can stand, confounders must be excluded: hypothermia, drug or alcohol intoxication and neuromuscular blockade, severe electrolyte, acid-base, or endocrine derangement, and significant hypotension. Local protocols and the number of required examiners vary by jurisdiction, so the bedside clinician should follow institutional policy precisely.
A crucial point of the 2023 guideline concerns ancillary testing. An ancillary test is not a substitute for a complete clinical examination and apnea test; it is used only when the clinical examination cannot be fully completed or the apnea test cannot be safely performed or completed (for example, because of high cervical injury, severe hypoxemia, or an open globe precluding a reflex). When indicated, the only acceptable adult ancillary tests are cerebral blood-flow studies:
- Four-vessel catheter (conventional) cerebral angiography.
- Radionuclide cerebral blood-flow scan.
- Transcranial Doppler (adults only).
Under the 2023 US guideline, EEG, auditory evoked potentials, somatosensory evoked potentials, CT angiography (CTA), MR angiography (MRA), and MRI are NOT acceptable ancillary tests for BD/DNC determination.
Bringing It Together
The comatose patient cannot give a history, but the examination can. Begin by deciding whether the picture is diffuse (reactive pupils, symmetric reflexes, no focal signs โ think toxic-metabolic) or structural (asymmetry, a discrete level of lost reflexes, abnormal eye position). Then localize by walking down the brainstem: pupils for the midbrain, cornea for the pons, reflex eye movements for the tegmentum between them, gag and cough for the medulla โ cross-checked against the breathing pattern and the motor response. Above all, treat the examination as a moving picture rather than a snapshot: a rostral-to-caudal march of failing reflexes and worsening posture is the unmistakable signature of herniation and a call to act before the next level is lost.
References
- Posner JB, Saper CB, Schiff ND, Plum F. Plum and Posner's Diagnosis of Stupor and Coma. 4th ed. Oxford University Press; 2007.
- Ropper AH, Samuels MA, Klein JP, Prasad S. Adams and Victor's Principles of Neurology. 11th ed. McGraw-Hill; 2019.
- Wijdicks EFM, Varelas PN, Gronseth GS, Greer DM. Evidence-based guideline update: determining brain death in adults. Neurology. 2010;74(23):1911-1918.
- Greer DM, Shemie SD, Lewis A, et al. Determination of brain death/death by neurologic criteria: the World Brain Death Project. JAMA. 2020;324(11):1078-1097.
- Greer DM, Kirschen MP, Lewis A, et al. Pediatric and Adult Brain Death/Death by Neurologic Criteria Consensus Guideline. Neurology. 2023.
- Teasdale G, Jennett B. Assessment of coma and impaired consciousness: a practical scale. Lancet. 1974;2(7872):81-84.