# Neurological Assessment in the ICU

## Introduction

Critically ill neurosurgical patients require frequent, systematic neurological assessment to detect changes that may require urgent intervention. The neurological examination in the ICU is adapted for patients who may be intubated, sedated, or pharmacologically paralyzed. Neurosurgeons and neurocritical care teams must integrate clinical findings with multimodal monitoring data to guide management decisions including surgical intervention, osmotherapy, and prognostication.

## The Focused Neurosurgical ICU Examination

### Level of Consciousness

The Glasgow Coma Scale is the universal tool for quantifying consciousness, comprising Eye response scored 1 to 4, Verbal response scored 1 to 5, and Motor response scored 1 to 6. The GCS Motor score is the most prognostically significant component. GCS must be assessed off sedation when possible, with documentation of the time since the last sedative dose. The FOUR Score, or Full Outline of UnResponsiveness, was developed specifically for intubated patients and assesses Eye, Motor, Brainstem reflexes, and Respiration, each scored 0 to 4, eliminating the need for a verbal response. Trends are more important than single assessments, and a decline of 2 or more GCS points is considered clinically significant.

### Pupillary Examination

The pupillary examination assesses size, shape, symmetry, and reactivity using a standardized bright light source. A unilateral fixed dilated pupil indicates ipsilateral uncal herniation until proven otherwise and represents an emergency requiring CT and potential surgical intervention. Bilateral fixed dilated pupils suggest severe brainstem injury, bilateral herniation, drug effects from atropine or epinephrine, or hypothermia. Bilateral miotic reactive pupils suggest metabolic encephalopathy, opioid effect, or a pontine lesion. Automated pupillometry using the Neurological Pupil index provides objective, reproducible measurements, with an NPi below 3 considered abnormal, eliminating interobserver variability.

### Motor Examination

The motor examination assesses spontaneous and stimulus-evoked movements in all four extremities. Localizing refers to purposeful movement crossing midline toward the stimulus and is prognostically favorable. Withdrawal is stereotyped flexion away from the stimulus. Flexor posturing, also called decorticate posturing, consists of arms flexed with legs extended and suggests a hemispheric or upper brainstem lesion. Extensor posturing, also called decerebrate posturing, consists of arms and legs extended with internal rotation and suggests a lower brainstem or midbrain lesion. Asymmetry in motor response suggests focal lateralized pathology.

### Brainstem Reflexes

The corneal reflex tests cranial nerve V as the afferent limb and cranial nerve VII as the efferent limb; bilateral absence indicates severe brainstem injury. The oculocephalic reflex, or doll's eyes maneuver, is contraindicated if the cervical spine is not cleared and tests brainstem integrity. The vestibulo-ocular reflex is tested with cold caloric irrigation of ice water in the ear canal; absent response indicates brainstem dysfunction. The cough and gag reflexes test cranial nerves IX and X and should be assessed prior to extubation consideration. The pupillary light reflex tests cranial nerve II as the afferent limb and cranial nerve III as the efferent limb.

## Impact of Sedation and Pharmacology

Propofol is short-acting and allows neurological assessments during sedation holds, with pupillary reflexes typically preserved. Midazolam is longer acting, accumulates, and confounds neurological assessment. Opioids cause miosis but preserve pupillary reactivity. Neuromuscular blocking agents eliminate motor assessment entirely and should be avoided during active neurological monitoring unless absolutely necessary. Sedation holds, also called neurological windows, are essential for serial examinations and require coordination with nursing. Hypothermia slows reflexes and reduces consciousness, representing an additional confounding factor.

## Multimodal Neuromonitoring

### Intracranial Pressure Monitoring

ICP monitors include the external ventricular drain with transducer, which is the gold standard and also allows CSF drainage, and intraparenchymal monitors such as the Codman or Camino devices. Normal ICP ranges from 5 to 15 mmHg in adults, with treatment indicated if ICP is sustained above 22 mmHg per Brain Trauma Foundation guidelines. Cerebral perfusion pressure is calculated as MAP minus ICP, with a target of 60 to 70 mmHg. ICP waveform analysis showing P2 greater than P1, representing loss of normal waveform compliance, suggests decreased intracranial compliance.

### Brain Tissue Oxygenation

PbtO2 monitoring uses an invasive probe to measure regional brain tissue oxygen tension. Normal PbtO2 ranges from 25 to 35 mmHg, and treatment is indicated if values fall below 20 mmHg, which indicates cerebral hypoxia. This monitoring guides optimization of CPP, FiO2, hemoglobin, and ventilation parameters.

### Continuous EEG

Continuous EEG monitoring detects non-convulsive seizures, which are present in 10 to 35 percent of neurocritical care patients. It also monitors for delayed cerebral ischemia after subarachnoid hemorrhage. Quantitative EEG trends aid in detecting subclinical changes. Indications include unexplained altered consciousness, post-status epilepticus monitoring, subarachnoid hemorrhage, and TBI with seizure risk.

### Cerebral Microdialysis

Cerebral microdialysis measures metabolic markers including lactate, pyruvate, glucose, glutamate, and glycerol in brain interstitial fluid. A lactate-to-pyruvate ratio greater than 40 indicates metabolic crisis from ischemia or mitochondrial dysfunction. This modality is used primarily in research settings but is increasingly available in specialized neurocritical care units.

### Transcranial Doppler (TCD)

Transcranial Doppler provides non-invasive assessment of cerebral blood flow velocities. For vasospasm detection, a mean flow velocity greater than 120 centimeters per second in the middle cerebral artery suggests vasospasm, and a Lindegaard ratio greater than 3 confirms the diagnosis. An elevated pulsatility index suggests increased ICP or decreased distal vessel compliance. TCD is also used for cerebral circulatory arrest confirmation in brain death evaluation.

## Herniation Syndromes

Uncal or transtentorial herniation presents with ipsilateral cranial nerve III palsy producing a dilated pupil, followed by contralateral hemiparesis and then bilateral posturing, and represents the most common surgical emergency. Central or downward herniation produces bilateral pupillary dilation, loss of upward gaze, and bilateral posturing. Subfalcine herniation causes midline shift with ipsilateral anterior cerebral artery compression, resulting in contralateral leg weakness. Tonsillar or cerebellar herniation occurs when the cerebellar tonsils herniate through the foramen magnum, causing sudden respiratory arrest and neck stiffness. Upward or ascending herniation occurs when a posterior fossa mass herniates upward through the tentorium and may be precipitated by rapid CSF drainage from an EVD.

## Clinical Pearls

A decline of 2 or more GCS points requires urgent investigation, typically with emergent CT imaging. A unilateral fixed dilated pupil in a neurosurgical patient is uncal herniation until proven otherwise and demands immediate action. The FOUR Score is superior to GCS for intubated patients because it does not require a verbal response and includes brainstem reflex assessment. Sedation holds are essential for accurate neurological assessment, and the examination performed during a sedation hold is the most reliable. Non-convulsive seizures are common in neurocritical care patients, and continuous EEG monitoring should be considered for unexplained altered consciousness.

## References
1. Wijdicks EF, Bamlet WR, Maramattom BV, et al. Validation of a new coma scale: The FOUR Score. Ann Neurol. 2005;58(4):585-593.
2. Hawryluk GWJ, Aguilera S, Buki A, et al. A management algorithm for patients with intracranial pressure monitoring (BTF 4th edition). Neurosurgery. 2019;84(5):E272-E276.
3. Claassen J, Mayer SA, Kowalski RG, et al. Detection of electrographic seizures with continuous EEG monitoring in critically ill patients. Neurology. 2004;62(10):1743-1748.
4. Le Roux P, Menon DK, Citerio G, et al. Consensus summary statement of the International Multidisciplinary Consensus Conference on Multimodality Monitoring. Neurocrit Care. 2014;21(Suppl 2):S1-S26.