Clinical Calculators

NIHSS (NIH Stroke Scale)

Neurological severity of stroke as the sum of 15 item scores (0 to 42 points)

15 items left

1a. Level of Consciousness

Pick a response even when an endotracheal tube, a language barrier or trauma or bandages around the mouth stop a full assessment. Give 3 only when there is no movement at all in response to a painful stimulus, apart from reflex posturing

1b. LOC Questions (the month and the patient's age)

An answer that is close but not exact earns no credit. Patients with aphasia or stupor who cannot grasp the questions get 2. Patients who cannot speak for a reason other than aphasia (intubation, mouth trauma, severe dysarthria of any cause, a language barrier) get 1. Grade the first answer only, and give no hints, spoken or silent

1c. LOC Commands (open and close the eyes; grip and release the hand)

Use the hand that is not paretic. If the hands cannot be used, another simple one-step command may stand in. Give credit when the patient clearly tries but weakness prevents completion. If the patient does not respond, demonstrate the task by pantomime and score what follows. Only the first attempt counts

2. Best Gaze

Assess horizontal eye movements only, either voluntary or reflex (oculocephalic); no caloric testing. A conjugate deviation that voluntary or reflex movement can overcome, or an isolated palsy of cranial nerve III, IV or VI, scores 1. Gaze can be tested even in aphasia. When there is eye trauma, bandaging, earlier blindness or another disorder of vision, use reflex movements and your own judgment

3. Visual

Test the upper and lower quadrants by confrontation, with finger counting or a visual threat as suits the patient. The patient may be encouraged; if the eyes turn correctly toward the moving fingers, score normal. With blindness or enucleation of one eye, test the other eye. Give 1 only for a clear asymmetry, quadrantanopia included. Blindness from any cause scores 3. Double simultaneous stimulation is done at this step: extinction scores 1 and the finding is carried to item 11

4. Facial Palsy

Ask the patient, or show by pantomime, to bare the teeth or raise the eyebrows and shut the eyes. In a patient who responds poorly or does not understand, compare the two sides of the grimace to a painful stimulus. If trauma, bandages, an orotracheal tube, tape or another barrier hides the face, remove it as far as possible

5a. Motor Arm: Left Arm

Examine one arm at a time, beginning with the unaffected side. Hold the arm out with the palm down, at 90 degrees when the patient sits or 45 degrees when lying. Count drift if the arm sinks before 10 seconds. With aphasia, encourage with an urgent voice and by demonstration, never with a painful stimulus. Choose UN only for an amputation or a fused joint at the shoulder, and write down why

5b. Motor Arm: Right Arm

Examine one arm at a time, beginning with the unaffected side. Hold the arm out with the palm down, at 90 degrees when the patient sits or 45 degrees when lying. Count drift if the arm sinks before 10 seconds. With aphasia, encourage with an urgent voice and by demonstration, never with a painful stimulus. Choose UN only for an amputation or a fused joint at the shoulder, and write down why

6a. Motor Leg: Left Leg

Examine one leg at a time, beginning with the unaffected side. Always examine lying down, with the leg raised to 30 degrees. Count drift if the leg sinks before 5 seconds. With aphasia, encourage with an urgent voice and by demonstration, never with a painful stimulus. Choose UN only for an amputation or a fused joint at the hip, and write down why

6b. Motor Leg: Right Leg

Examine one leg at a time, beginning with the unaffected side. Always examine lying down, with the leg raised to 30 degrees. Count drift if the leg sinks before 5 seconds. With aphasia, encourage with an urgent voice and by demonstration, never with a painful stimulus. Choose UN only for an amputation or a fused joint at the hip, and write down why

7. Limb Ataxia

The aim is to find a one-sided cerebellar lesion. Test with the eyes open, using the intact visual field if vision is impaired. Do finger-nose-finger and heel-shin on both sides, and count ataxia only when it is greater than the weakness would explain. A patient who cannot understand or who is paralyzed has no ataxia. If the patient is blind, have the patient touch the nose from an outstretched arm. Choose UN only for an amputation or a fused joint, and write down why

8. Sensory

Use pinprick, or in an obtunded or aphasic patient look for a grimace or withdrawal from a noxious stimulus. Count only sensory loss due to the stroke, and test enough areas (arms but not hands, legs, trunk, face) to detect loss on one side. Give 2 only when severe or total loss is clearly shown, so stuporous and aphasic patients most likely get 1 or 0. A brainstem stroke with loss on both sides scores 2, as does an unresponsive quadriplegic patient. A patient in coma (item 1a = 3) is automatically given 2

9. Best Language

Much of the information comes from the earlier examination. Use the standard picture, naming and sentence cards: ask the patient to say what is happening in the picture, name the objects, and read the sentences. Judge comprehension from these and from the commands given earlier. If poor vision prevents this, ask the patient to name an object placed in the hand, to repeat, and to speak. Ask an intubated patient to write. A patient in coma (item 1a = 3) automatically gets 3. For a patient in stupor or with limited cooperation, you must choose a score; use 3 only for a patient who cannot speak and obeys no one-step command

10. Dysarthria

If the patient seems normal, collect enough speech by having the patient read or repeat words from the standard word list. With severe aphasia, judge how clear the spontaneous speech is. Do not explain the purpose of the test. Choose UN only when intubation or another physical barrier prevents speech, and write down why

11. Extinction and Inattention (formerly Neglect)

Earlier testing may already give enough information. If severe visual loss prevents visual double simultaneous stimulation and touch stimuli are normal, score normal. If the patient is aphasic but seems to attend to both sides, score normal. Visual-spatial neglect or anosognosia can also count as abnormal. Only an existing abnormality is scored, so the item can never be untestable

Cautions

  • The NIH Stroke Scale has 15 items (5a and 5b, and 6a and 6b, score the left and right limbs separately). The total runs from 0 to 42. This calculator gives no severity bands, because the NINDS scoring sheet defines none.
  • Score the items in the order listed, and never return to an earlier item to alter its score. A score should reflect what the patient actually does, not what the clinician believes the patient could do. Record the answers during the examination, and do not coach the patient except where an item allows it.
  • UN (untestable) is used only where the item says so (the four limb items, limb ataxia and dysarthria). The NINDS sheet does not say how to total an UN item. This calculator adds 0 points for it, so the total is lower by the points that item could have given. Write down the reason for each UN.
  • Some items apply an automatic score: a patient in coma (item 1a = 3) scores 3 on item 9 (Best Language) and 2 on item 8 (Sensory). Item 11 is never untestable.

References

  1. National Institute of Neurological Disorders and Stroke (NINDS). NIH Stroke Scale. 2003. (Items 1a to 11, instructions and scale definitions, and the UN (untestable) notes)[Original article]
  2. Brott T, Adams HP Jr, Olinger CP, Marler JR, Barsan WG, Biller J, Spilker J, Holleran R, Eberle R, Hertzberg V. Measurements of acute cerebral infarction: a clinical examination scale. Stroke. 1989;20(7):864-870. (original not seen) PMID: 2749846, DOI: 10.1161/01.str.20.7.864[Original article]
  3. Lyden P, Brott T, Tilley B, Welch KM, Mascha EJ, Levine S, Haley EC, Grotta J, Marler J; NINDS TPA Stroke Study Group. Improved reliability of the NIH Stroke Scale using video training. Stroke. 1994;25(11):2220-2226. (original not seen) PMID: 7974549, DOI: 10.1161/01.STR.25.11.2220[Original article]

Checked on: 2026-11-07

Disclaimer

This calculation is intended to support the judgment of healthcare professionals. Results are for reference only; decisions about diagnosis and treatment must be made by the responsible healthcare professional. The values you enter never leave your device.