FOUR Score vs the Glasgow Coma Scale: Why the ICU Needed a Better Coma Score

Part of the ICU Emergencies Hub — browse every related guide in one place.

Updated July 2026  |  More ICU clinical guides →

The Glasgow Coma Scale has been the default coma score for decades, and for good reason — it is fast, universally known, and reasonably reproducible. But every ICU nurse eventually hits its blind spots. The verbal component collapses the moment a patient is intubated, forcing the awkward "T" workaround that throws away information. And the GCS says nothing about the brainstem, so a patient can hold a stable GCS while their pupils fix and their breathing falls apart. The FOUR score (Full Outline of UnResponsiveness) was designed to close exactly those gaps. This guide compares the two and shows how to score the FOUR reliably at the bedside.

Where the GCS runs out of road

The GCS's weaknesses are structural, not fixable by scoring harder. The verbal score is meaningless in an intubated or aphasic patient — you either default to a number or annotate "T," and either way the total loses fidelity. Worse, the GCS has no window into the brainstem: pupillary and corneal reflexes, cough, and respiratory pattern are all invisible to it. A patient sliding toward herniation can hold the same GCS while their brainstem is failing underneath, because the scale simply does not look there. The FOUR score was built by neurointensivists specifically to add that missing brainstem and respiratory dimension while staying quick enough for hourly bedside use.

The four components, and what each one catches

Two design choices are worth noting. The eye component uses voluntary tracking and command blinking, which are cortical and often decline early, giving the score sensitivity at the top end where GCS eye-opening is already maxed out. And the respiration component scores the pattern — including whether an intubated patient is breathing above the ventilator rate or has stopped triggering — turning the vent from a scoring obstacle into a source of data.

Scoring it without drifting

The same discipline that keeps a GCS honest applies here: score the best response you can genuinely elicit, use a real noxious stimulus when needed, and score what you see rather than what you expect. A few FOUR-specific cautions. For the eye component, absence of tracking must be tested in more than one direction before you call it lost. For motor, the hand-signal commands (thumbs-up, fist, peace sign) are the standard prompt — a patient who follows any of them scores full marks on M. For the brainstem component, one fixed pupil and one reactive is scored differently from both fixed, so document each eye. And the respiration component requires you to actually watch the pattern and, in a vented patient, compare the patient's rate to the set rate rather than assuming.

Which score to reach for

In practice, many units use both: GCS for its universality and rapid shared shorthand, and the FOUR score when the patient is intubated, when brainstem function is the question, or when a neuro-ICU wants finer resolution near the severe end of the spectrum. The FOUR score's real payoff is in the intubated, deteriorating patient — precisely the population where the GCS is weakest — because it keeps generating meaningful, comparable numbers when the verbal component would otherwise go dark. If your unit is trending a patient hour to hour toward possible herniation, the FOUR score plus quantitative pupillometry gives you a far richer trend than a GCS with a "T" annotation ever could.

Where to go from here

Ground this in the neuro-checks and GCS guide, add objective pupil data with automated pupillometry, and know the herniation red flags that any coma score is ultimately trying to catch before it is too late.