NIHSS Pocket Guide

NIHSS: The NIH Stroke Scale, Scored and Interpreted

A bedside guide to the NIH Stroke Scale - all 15 items, the scoring conventions that trip people up, severity bands, serial use, and the posterior-circulation blind spot.

Zaka Ahmed, MD

Vascular Neurology Fellow · Clinical reference · 9 min read

A 15-item exam, scored 0-42, that turns a stroke into a number you can track and hand off.

→ Higher is worse. Serial NIHSS is the fastest way to catch early deterioration or improvement.

→ A low score does not rule out stroke or a large-vessel occlusion - especially in the posterior circulation.

→ For thrombolysis, the question is whether the deficit is disabling, not whether the number clears a threshold.

The NIH Stroke Scale (NIHSS) is a 15-item bedside neurological examination that quantifies the severity of an acute stroke on a scale from 0 to 42. It was built for reproducibility: a tool that a neurologist, an emergency physician, and a stroke nurse could each apply to the same patient and arrive at nearly the same number.¹ That shared number is what makes the rest of stroke care possible - it travels with the patient from the field to the scanner to the angio suite, and it anchors almost every major reperfusion trial.

The 15 items

The scale walks rostral to caudal through the exam. Each item has a fixed point range; you sum them.

1a, 1b, 1c - Level of consciousness, two orientation questions (month, age), two commands (eyes, grip)
2 - Best gaze (horizontal)
3 - Visual fields
4 - Facial palsy
5a / 5b - Motor arm, left and right (scored separately)
6a / 6b - Motor leg, left and right
7 - Limb ataxia
8 - Sensory
9 - Best language (aphasia), scored 0-3
10 - Dysarthria, scored 0-2
11 - Extinction and inattention (neglect), scored 0-2

Scoring conventions that trip people up

Most NIHSS disagreement comes from a handful of rules, not from the obvious items:

Score what you see, not what you expect. Record the first effort, not the best effort after coaching.
Aphasia (item 9) and dysarthria (item 10) are independent. A globally aphasic patient can still be scored for slurred articulation.
Sensory and commands in aphasia. In an aphasic or non-arousable patient, sensory testing (item 8) uses grimace and withdrawal as proxies.
Ataxia only counts when it is out of proportion to weakness. A plegic limb cannot be ataxic - item 7 is 0.
The comatose patient (1a = 3) scores at or near the maximum across multiple motor and language items by convention, which is why deep coma drives the total toward the 30s.

Interpreting the total

There is no single official cut-table, but the bands most stroke teams use are:

0 - No measurable deficit
1-4 - Minor stroke
5-15 - Moderate stroke
16-20 - Moderate to severe
21-42 - Severe stroke

Baseline NIHSS is one of the strongest single predictors of outcome: in the TOAST cohort, a low admission score predicted a high likelihood of good recovery, while a score above roughly 16 carried a high likelihood of death or severe disability.³ Treat the bands as a shared language, not a prognosis you quote to a family.

Serial NIHSS

A single number is a snapshot; the trend is the signal. Re-scoring at fixed intervals - and any time the nurse calls with a change - is the fastest bedside way to detect early neurological deterioration (clot extension, hemorrhagic transformation, malignant edema) or the improvement that follows reperfusion. Document the components, not just the total, so the next clinician can see what changed.

The posterior-circulation blind spot

The NIHSS was weighted for anterior-circulation strokes and systematically under-scores posterior ones. A complete hemianopia is only 2 points, and the scale gives no credit for vertical gaze palsy, severe vertigo, ataxia out of proportion to weakness, or decreased consciousness from a brainstem lesion. A disabling basilar or PCA stroke can score in the low single digits - "zero on the NIHSS does not equal the absence of stroke."⁴

NIHSS and treatment decisions

Thrombolysis. There is no universal lower NIHSS cutoff for IV thrombolysis. The decision turns on whether the deficit is disabling for that patient - an isolated aphasia or a hemianopia can be low-scoring and still life-altering. A high score is likewise not an automatic exclusion.

Thrombectomy. The 2015 early-window trials generally enrolled patients with NIHSS ≥6 and a proven large-vessel occlusion, and that number still anchors most early-window pathways. But selection is clinical and radiographic together: a low NIHSS does not exclude an LVO, and a salvageable patient should not be turned away on the score alone.

Certification and reliability

Interrater reliability is good when raters are trained and certified, and it degrades when they are not - which is why formal NIHSS certification became standard for stroke-trial sites and stroke centers.² Even certified, the scale has documented quirks and ceiling effects; it is a severity index, not a complete neurological exam, and it should be read as one.⁵

Bedside pearl

When the NIHSS and your gestalt disagree - a "low score" in a patient who looks devastated - believe the patient and image the posterior circulation. The scale is a floor for suspicion, never a ceiling.

Frequently asked questions.

What is the maximum NIHSS score?

42. A score in the high 30s or low 40s generally reflects a large dominant-hemisphere stroke or a comatose patient, because deep coma maxes out several motor and language items by convention.

What NIHSS counts as a severe stroke?

By common convention, 0 is no measurable deficit, 1-4 minor, 5-15 moderate, 16-20 moderate to severe, and 21-42 severe. These bands are a shared shorthand, not an official cut-table, and they vary slightly between centers.

Can a stroke have a low NIHSS or even zero?

Yes. The NIHSS under-scores posterior-circulation strokes, so a disabling brainstem or PCA stroke - or even a large-vessel occlusion - can present with a very low score. A low NIHSS does not rule out stroke or LVO.

Is there a minimum NIHSS to give thrombolysis?

No universal minimum. The decision rests on whether the deficit is disabling for that patient rather than on clearing a numeric threshold, and a high score is not an automatic contraindication.

How often should the NIHSS be repeated?

At intervals set by your protocol and whenever there is a clinical change. The trend across serial exams - not any single total - is what flags early deterioration or reperfusion.

References.

Brott T, Adams HP Jr, Olinger CP, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke. 1989;20(7):864-870. PubMed

Goldstein LB, Samsa GP. Reliability of the National Institutes of Health Stroke Scale. Extension to non-neurologists in the context of a clinical trial. Stroke. 1997;28(2):307-310. PubMed

Adams HP Jr, Davis PH, Leira EC, et al. Baseline NIH Stroke Scale score strongly predicts outcome after stroke: a report of the Trial of Org 10172 in Acute Stroke Treatment (TOAST). Neurology. 1999;53(1):126-131. PubMed

Martin-Schild S, Albright KC, Tanksley J, et al. Zero on the NIHSS does not equal the absence of stroke. Ann Emerg Med. 2011;57(1):42-45. PubMed

Lyden P. Using the National Institutes of Health Stroke Scale: a cautionary tale. Stroke. 2017;48(2):513-519. PubMed

Keep the stroke service moving.

Modified Rankin Scale (mRS) The modified Rankin Scale explained: all seven grades 0-6, the independence and walking boundaries, baseline-to-90-day anchoring, and the mRS 0-2 trial cut-point.

ASPECTS ASPECTS explained: the 10-region MCA score on non-contrast CT, the regions most often missed, historical and large-core thrombectomy thresholds, and automated scoring.

IV Thrombolysis Eligibility IV thrombolysis for acute ischemic stroke: the 4.5-hour and extended windows, tenecteplase and alteplase dosing, key exclusions, and where guidelines diverge.

Door-to-Needle Bundle How stroke teams compress door-to-needle time below 30 minutes: EMS pre-notification, direct-to-CT, parallel processing, a bedside thrombolytic, and drip-and-ship.

All clinical tools The full bedside reference index.