Glasgow Coma Scale (GCS): Scoring, the GCS-P Update, and Charting a Non-Testable Component

The Glasgow Coma Scale (GCS) is the standard instrument for documenting a patient’s level of consciousness at the bedside — three components, summed to a single number, repeated on a schedule so the trend, not just the snapshot, is what gets charted. It was developed by Graham Teasdale and Bryan Jennett at Queen Elizabeth Hospital in Glasgow and published in The Lancet in 1974; the original 14-point version was expanded to the current 15-point scale in 1977, and it has since been incorporated into Advanced Trauma Life Support (ATLS) protocols worldwide. This page covers the instrument itself — the exact scoring criteria, how to chart a component that can’t be tested, the 2018 GCS-P refinement, and what a hospital’s neuro-check policy has to account for that the raw number alone doesn’t capture. For the broader deteriorating-patient program (track-and-trigger scores, escalation thresholds, alarm management), see CASRAI’s Early Warning Score Implementation guide, which covers that governance layer in depth but does not walk through GCS scoring mechanics.

The three components and their point values

GCS scores three behavioral responses independently, then sums them. A lower score on any component means less response; the total is the sum of the best response observed in each category, not an average.

Eye opening (E) — scored 1 to 4

• 4 — Spontaneous: eyes open without any stimulus.
• 3 — To command/voice: eyes open when spoken to.
• 2 — To pain: eyes open only in response to a painful stimulus.
• 1 — None: no eye opening regardless of stimulus.

Verbal response (V) — scored 1 to 5

• 5 — Orientated: knows who they are, where they are, and roughly when.
• 4 — Confused: converses but is disoriented.
• 3 — Inappropriate words: intelligible words, but not forming coherent conversation.
• 2 — Incomprehensible sounds: moaning, groaning, no recognizable words.
• 1 — None: no vocalization.

Motor response (M) — scored 1 to 6

• 6 — Obeys commands: follows simple instructions.
• 5 — Localizes to pain: a purposeful movement toward the painful stimulus.
• 4 — Withdrawal from pain: pulls away from the stimulus without purposeful localization.
• 3 — Abnormal flexion: decorticate posturing — arms flex toward the body.
• 2 — Abnormal extension: decerebrate posturing — arms extend and rotate.
• 1 — None: no motor response to pain.

The three components are usually charted individually (E, V, M) as well as summed, so a documented score reads as, for example, “E3 V4 M5 = GCS 12” rather than just “12” — the breakdown is clinically informative in a way the total alone isn’t (see the confounders section below for why).

Total score and severity bands

The three components sum to a total ranging from 3 (the floor — no response in any category) to 15 (fully alert and oriented). The conventional severity bands, most often applied to traumatic brain injury:

• 13–15 — Mild injury or minor impairment.
• 9–12 — Moderate injury.
• 3–8 — Severe injury; a GCS of 8 or less is the widely used threshold for airway protection concerns and is frequently cited as an indication to consider intubation.

These cutoffs were derived from and are most validated in adult traumatic brain injury; pediatric literature generally uses a lower threshold (around 5 or less) to identify severe injury, reflecting differences in baseline verbal and motor development rather than a different injury mechanism. A facility’s neuro-assessment policy for a pediatric or NICU/PICU population should specify which pediatric-adapted scale it uses (the Pediatric GCS modifies the verbal criteria for pre-verbal children) rather than applying the adult verbal criteria directly.

Charting a component that can’t be tested

The scenario every ICU and post-anesthesia unit runs into constantly — and the one most consumer-facing explainers skip — is what to chart when a component genuinely cannot be assessed. An intubated patient cannot produce a verbal response; heavy periorbital swelling after facial trauma can make eye opening untestable regardless of the patient’s actual consciousness; a paralyzed limb from a prior stroke can make the motor score reflect old deficit rather than current status.

The convention, reflected in the clinical literature, is to record the untestable component as “NT” (not testable) rather than scoring it as the lowest value by default — scoring an intubated patient’s verbal response as 1 systematically understates their actual level of consciousness and corrupts trend comparisons over the admission. For an intubated patient specifically, the verbal component is often charted as “V–ET” (endotracheal tube) rather than a numeric score, with the total reported as the eye-plus-motor sum accompanied by that qualifier, not as a false composite number. One published imputation approach (used mainly in research settings where a single comparable total is needed) estimates the missing verbal score from the eye-plus-motor sum — but this is a statistical estimation method for study data, not a substitute for accurate bedside charting, and a hospital’s own neuro-assessment policy should be explicit about which convention its EHR flowsheet uses so nursing staff chart consistently rather than each defaulting to a different workaround.

The GCS-P (pupil reactivity) update

In 2018, Paul Brennan, Gordon Murray, and Graham Teasdale (one of the scale’s original authors) published the GCS-P, which combines the standard GCS with a Pupil Reactivity Score (PRS) to correct for a known limitation: the base GCS does not capture brainstem/pupillary findings at all, and two patients with the same GCS total can have very different underlying severity if one has fixed, unreactive pupils. The PRS is added as follows:

• 0 points — both pupils react to light.
• 1 point — one pupil does not react.
• 2 points — neither pupil reacts.

GCS-P is then calculated as GCS − PRS, producing a combined score from 1 to 15, with low GCS-P values (roughly 1–8) indicating more severe injury than the same raw GCS number would suggest on its own. GCS-P has not replaced standard GCS charting in routine nursing flowsheets — it’s mainly used in trauma/neurosurgical severity scoring and research — but a facility building or auditing a neuro-assessment policy for a trauma or neurosurgical unit should know it exists and decide deliberately whether pupillary findings are captured alongside GCS in that unit’s documentation, rather than assuming the base GCS alone is a complete picture.

Why the same total number can mean different things

This is the piece of the instrument a quality or risk-management audience needs and a purely instructional explainer typically doesn’t cover: GCS is additive, so two patients can land on an identical total via very different component breakdowns, and those breakdowns are not clinically equivalent. A worked example cited in the clinical literature: a total GCS of 4 built from E1 V1 M2 (no eye opening, no verbal response, abnormal extension) carries a materially different reported mortality rate than the same total of 4 built from E2 V1 M1 (eyes open to pain, no verbal response, no motor response) — the motor component tends to dominate outcome prediction more than the total implies. This is the practical argument for charting E/V/M individually rather than only the sum, and for training staff not to treat “GCS 8” as a single fixed clinical picture without looking at which components produced it.

Several factors can also depress a GCS score independent of the underlying neurological status, and a defensible neuro-assessment policy documents how staff are expected to note them rather than charting the number in isolation: alcohol or drug intoxication, sedative or analgesic medication, seizure activity or the post-ictal state, hypothermia, hypoxia, and hypotension can all lower the observed score. A patient’s GCS taken immediately after a sedating medication is not directly comparable to their prior baseline without that context noted in the same entry.

Building GCS into a neuro-assessment policy

The instrument itself is freely available and short; the operational work a patient-safety or quality program actually owns is different:

• Reassessment frequency by acuity. Facilities commonly stratify neuro-check frequency by unit and time since injury or procedure — more frequent (e.g., hourly) immediately post-neurosurgery or in a fresh traumatic brain injury admission, tapering as the patient stabilizes. The specific interval is a local clinical-practice decision, not something GCS itself specifies, and should be written into unit-level policy rather than left to individual judgment.
• Inter-rater reliability. Because motor scoring in particular requires distinguishing localization from withdrawal from abnormal flexion, competency validation for staff who perform neuro checks (new hires, float staff, and periodic refreshers) reduces the kind of scoring drift that makes a trend unreliable across shifts.
• Trending into escalation, not just charting. A GCS that drops by even one or two points from a patient’s own baseline is a recognized deterioration signal, and unrecognized neurological decline is a pattern that shows up in root cause analyses of missed-deterioration events. Where a facility runs a track-and-trigger early warning score program, the neuro-assessment policy should specify how a GCS change interacts with that escalation pathway rather than existing as a separate, unlinked flowsheet entry — see CASRAI’s Early Warning Score Implementation guide for the escalation-design side of that work.
• Sedation and assessability. Units that use scheduled sedation (ICU ventilated patients, procedural sedation) need an explicit policy on when GCS is expected to be assessable at all, and how a sedation-related NT/untestable entry is distinguished from a genuine neurological decline — see CASRAI’s Moderate Sedation Policy Requirements guide for the regulatory side of that overlap.

GCS vs. other bedside assessment tools

GCS measures level of consciousness broadly and is not a stroke-specific, delirium-specific, or sedation-specific instrument, and hospitals commonly run more than one scale on the same patient for different purposes. The NIH Stroke Scale (NIHSS) is a more granular, stroke-specific neurological exam used to quantify deficit severity and guide thrombolytic/thrombectomy decisions; the Richmond Agitation-Sedation Scale (RASS) measures sedation depth and agitation in ICU patients rather than consciousness after injury; the Confusion Assessment Method (CAM/CAM-ICU) screens specifically for delirium. None of these substitute for GCS and GCS does not substitute for them — a policy that conflates “we do neuro checks” with “we screen for delirium” is a real documentation gap, not a redundancy.

Frequently asked questions

What does a GCS of 8 mean?

A total of 8 falls at the boundary of the severe category (3–8) and is the threshold most commonly cited for considering airway protection, since a score this low often means the patient cannot reliably protect their own airway. It should always be read alongside the individual E/V/M breakdown, not the total alone.

Is a GCS of 15 always normal?

A GCS of 15 (E4 V5 M6) reflects full alertness and orientation on this specific instrument, but it does not rule out neurological injury on its own — a patient can have a normal GCS with a significant focal deficit, a subtle bleed, or early signs that a stroke-specific tool like the NIHSS would catch and GCS would not, which is why GCS and a focal neurological exam are typically used together rather than GCS alone.

What is GCS-P and is it required?

GCS-P adds a Pupil Reactivity Score to the standard GCS to account for brainstem findings the base scale misses. It is not a mandatory replacement for standard GCS charting; facilities decide deliberately whether to incorporate pupillary findings into a given unit’s neuro-assessment documentation, most often in trauma and neurosurgical settings.

How do you chart GCS for an intubated patient?

The verbal component is recorded as untestable (“NT” or “V–ET” for endotracheal tube) rather than scored as the lowest possible value, and the eye and motor components are charted normally. Facility policy and the EHR flowsheet should specify the exact notation staff are expected to use so charting is consistent across shifts and comparable over the length of stay.

Is GCS the same as the NIH Stroke Scale?

No. GCS is a general level-of-consciousness instrument used across trauma, neurosurgical, and critical-care settings; the NIHSS is a stroke-specific neurological exam used to quantify deficit severity in a suspected or confirmed stroke and to help guide time-sensitive treatment decisions. They are complementary, not interchangeable.

Back to the CASRAI Patient Safety hub for surveillance definitions, root cause analysis, credentialing, and the rest of the hospital patient-safety and infection-prevention library.