NIH Stroke Scale and Acute Stroke Pathway
All 15 rows with the scale's own rules, then the thrombolysis decision · v11. Patient, Timing and Vitals
2. History Relevant to the Bedside Scores and to Thrombolysis
3. NIH Stroke Scale
Point to note. A patient who cannot answer because of aphasia, or who is stuporous and does not comprehend, scores 2. A patient who cannot answer for a mechanical or linguistic reason not secondary to aphasia, meaning intubation, orotracheal trauma, severe dysarthria of any cause or a language barrier, scores 1. Only the initial answer is graded and there is no partial credit for being close.
Credit an unequivocal attempt defeated by weakness. Substitute another one-step command if the hands cannot be used. Pantomime is permitted here and the pantomimed result is scored. Only the first attempt is scored.
Caloric testing is not done. A conjugate deviation that can be overcome scores 1. An isolated third, fourth or sixth nerve palsy scores 1. Gaze is testable in every aphasic patient, so aphasia is never a reason to leave this row unscored.
Score 1 only for a clear-cut asymmetry, including a quadrantanopia. Blind from any cause scores 3. Double simultaneous stimulation is performed at this point, and visual extinction found here scores 1 on this row and is then carried forward to row 11.
The rule that is most often got wrong. Ataxia is scored only if present out of proportion to weakness. Ataxia is absent, meaning score 0 and not UN, in the patient who is paralysed or who cannot understand the task. This row also has the poorest inter-rater agreement in the whole scale.
Only sensory loss attributable to the stroke is scored. Three paths force a 2: bilateral loss from a brainstem stroke, an unresponsive quadriplegic patient, and coma, meaning row 1a scored 3. Stuporous and aphasic patients will probably score 1 or 0.
The discriminator between 1 and 2 is whether the examiner can identify the picture or naming card content from what the patient says. The intubated patient is asked to write; intubation does not make this row untestable. Coma, meaning row 1a scored 3, forces a 3. A score of 3 is otherwise reserved for a patient who is mute and follows no one-step commands.
Severe aphasia is not a reason to record UN. Rate the clarity of articulation of spontaneous speech instead. Do not tell the patient why the sample is being taken.
Because the abnormality is scored only if it is present, this row is never untestable. Severe visual loss with normal cutaneous stimuli scores 0. Aphasia with attention to both sides scores 0.
4. Imaging
The 2026 AHA/ASA guideline moved this decision from a number to a judgement. A disabling deficit within 4.5 hours is treated regardless of the NIHSS score. A non-disabling deficit is not, and dual antiplatelet therapy is preferred instead. Occupation matters: a hand tremor that ends a surgeon's career is disabling, and the guideline expects the treating physician to say so.
While this is set to Not yet done, no thrombolytic dose is printed and the second output carries the answer. Selecting a CT result opens the fields the treatment decision needs.
Evidence & Clinical Pearls
1. What the NIHSS Measures, and What It Does Not
What it is. Fifteen scored rows across eleven numbered categories, total 0 to 42. It is a measure of neurological deficit at a moment in time. It is not a measure of infarct volume, not a measure of vessel status, and not a diagnosis.
Why it matters. Baseline NIHSS is the single strongest clinical predictor of outcome after ischaemic stroke, and it is also the variable that enters almost every prognostic and eligibility rule that follows. In the Indian IRIS-TNK registry of 1,015 thrombolysed patients, each additional NIHSS point carried an odds ratio of 0.823 for a good outcome, ahead of age at 0.983 per year.
Bedside pearl. The commonly quoted five-band severity scheme, minor 1 to 4, moderate 5 to 15, moderate to severe 16 to 20, severe 21 to 42, could not be traced to any primary NINDS or AHA document. It is teaching shorthand. The bands that do carry a source are from the TOAST cohort of 1,281 patients: NIHSS 16 or above forecasts a high probability of death or severe disability, and 6 or below forecasts good recovery. This tool prints the informal bands because residents are examined on them, and labels them as informal.
2. The Left and Right Asymmetry: Seven Points of Language Against Two of Neglect
What it is. The scale allocates 7 of its 42 points to language, rows 1b, 1c and 9, and only 2 points to neglect, row 11. Language is a dominant hemisphere function. Neglect is the signature non-dominant syndrome.
Why it matters. Woo and colleagues showed in the placebo arm of the NINDS trial that for every 5-point NIHSS band below 20, the median infarct volume of a right-hemisphere stroke was about double that of a left-hemisphere stroke at the same score. At an NIHSS of 16 to 20 the medians were 48 mL on the left against 133 mL on the right. Fink and colleagues found that among patients scoring 0 to 5, the diffusion lesion was 8.8 cm³ on the right against 3.2 cm³ on the left. Across four independent cohorts the gap at matched anatomy is consistently about 4 NIHSS points.
Bedside pearl. Desai and colleagues converted the measurement bias into a demonstrated treatment disparity. Among 211 patients meeting thrombectomy criteria, in the NIHSS 6 to 12 band thrombectomy was performed in 81 per cent of left-hemisphere strokes and 52 per cent of right-hemisphere strokes. Above NIHSS 12 there was no gap at all. The disparity lives exactly where a threshold rule such as "NIHSS 6 or more" does its work. A right-hemisphere patient with an NIHSS of 7 does not have a mild stroke; get the vessel imaged and let the angiogram, not the number, decide.
3. The Rows That Disagree Between Examiners
Reliability is not uniform across the instrument. Lyden and colleagues, scoring certification videos, found limb ataxia the worst row in the scale at a kappa of 0.21. Josephson and colleagues, across 7,405 raters, found best language and facial palsy the two largest contributors to variance in the total, at kappas of 0.60 and 0.65.
Bedside pearl. Certification restores agreement on the total, with intraclass correlations above 0.95 in the TOAST programme, while leaving the individual rows as the residual source of disagreement. If two doctors on the same ward round disagree by two points, the likeliest single row is ataxia, and the second likeliest is language.
4. Untestable Is Not Zero, Except Where It Is
Only six rows carry a UN code: 5a, 5b, 6a, 6b, 7 and 10. On the motor rows UN is permitted only for amputation or joint fusion at the shoulder or the hip. A hand amputation, a wrist fusion, an intravenous line or a plaster cast below the shoulder does not authorise UN. On row 10, UN requires intubation or another physical barrier to producing speech; severe aphasia does not qualify.
Two rows collapse to zero rather than to untestable, and both can only lower the total. Row 7 scores 0 in the patient who is paralysed or cannot comprehend. Row 11 has no UN code at all and states in its own instructions that it is never untestable.
Point to note. The instrument prints no rule for how a UN row enters the total, and prints no maximum. The convention in trials and registries is to treat UN as 0, which is what this tool does, but that convention could not be verified from a primary NINDS document. A total of 18 with an untestable arm is not the same clinical object as a total of 18 with every row scored, which is why this tool prints the UN rows alongside the total and puts them in the EMR text.
5. Bedside Stroke Type Scores in Indian Practice: What the Validation Studies Found
What they are. Clinical scores that attempt to separate supratentorial intracerebral haemorrhage from infarction without a CT. The Siriraj score, developed in Bangkok in 1991, is the one Indian residents are taught. The Greek score, published in 2002, uses neurological deterioration, vomiting, a raised white cell count and depressed consciousness.
Why they matter in India. Only a minority of Indian stroke patients reach a CT scanner within the treatment window, and district hospitals frequently see a patient hours before imaging is possible. A score that is right four times in five is not useless in that setting; it is simply not a licence to treat.
| Indian study | Setting | What it found |
|---|---|---|
| Badam, Kalantri et al, Natl Med J India 2003 | MGIMS Sevagram, rural Wardha, n = 134 | Siriraj sensitivity 78.5 per cent, specificity 71 per cent. Guy's sensitivity 81 per cent, specificity 76.2 per cent. Positive likelihood ratios of 2.7 and 3.4, which move a probability very little |
| Goswami et al, Indian J Med Sci 2013 | IPGMER Kolkata, n = 200 | Greek score had the greatest area under the curve at 0.973, above Siriraj. The authors proposed a Greek cut-off of 1.5 for safe bedside diagnosis of ischaemic stroke, which identified 47 per cent of ischaemic patients |
| Nabirajan et al, Cureus 2025 | Rajiv Gandhi GGH Chennai, n = 156 | Siriraj was equivocal in 17.3 per cent and, among the rest, only 38.7 per cent sensitive for haemorrhage although 96.9 per cent specific. Nineteen of 129 non-equivocal cases were false negatives for haemorrhage |
| Bhardwaj et al, Sci Rep 2025 | AIIMS New Delhi, prospective set of 92 | Siriraj weighted accuracy 60.78 per cent. A machine learning model trained on 2,190 Indian patients reached 82.42 per cent on the same set |
The methodological point that matters most. Almost every published sensitivity for these scores is computed after excluding the equivocal cases. Siriraj is equivocal in roughly 17 to 26 per cent of Indian patients. The headline figures therefore overstate what the score does for an unselected patient walking through the door, and this tool reports the equivocal band as an answer in its own right rather than hiding it.
6. Why This Tool Does Not Compute the Guy's Hospital (Allen) Score
The Guy's Hospital score appears in every Indian validation study alongside Siriraj, so its absence here is deliberate and worth explaining. Allen's 1983 paper in the Quarterly Journal of Medicine is closed access. The individual point weights circulate only in secondary restatements, and those restatements disagree with each other: the constant is given as both −12 and −12.6, and the infarction cut-off is given as both "below 4" and "at or below −4". A score whose weights cannot be traced to its own paper should not be computed on a live clinical page.
There is a second, clinical reason. The Allen score requires the diastolic pressure and the conscious level measured 24 hours after admission. Whatever its accuracy, it cannot contribute to a decision made in the first four and a half hours, which is the decision this tool exists to support. Kochar and colleagues made exactly this point from Bikaner in 2000.
7. The Greek Score Is Not a Door Score
Six of its points ride on neurological deterioration within three hours of admission, and four more on a white cell count. Neither is available at the moment the patient arrives. The Greek score therefore answers a question three hours later than the Siriraj score does, and the two are not interchangeable. This tool computes the Greek score only when the leucocyte count has been entered, and says so plainly when it has not, rather than treating a blank field as a normal count.
8. Tenecteplase in India: Two Doses, One Licence
What it is. A genetically modified variant of alteplase with greater fibrin specificity and a longer half-life, given as a single bolus rather than a bolus and an hour-long infusion. The single bolus is the practical advantage on an Indian ward, because it removes an infusion pump and an hour of nursing attention from the door-to-needle pathway.
The conflict, stated plainly. The 2026 AHA/ASA guideline recommends 0.25 mg/kg to a maximum of 25 mg, Class 1. India licensed tenecteplase for stroke in 2016 at 0.2 mg/kg to a maximum of 20 mg, and the evidence for that dose is IRIS-TNK, a 1,015-patient single-arm safety registry mandated as a condition of approval and funded by the manufacturer. There has been no randomised comparison of 0.2 against 0.25 mg/kg, and the IRIS-TNK authors themselves call for one.
What IRIS-TNK actually showed. Symptomatic intracranial haemorrhage 0.6 per cent, three-month mortality 1.0 per cent, modified Rankin 0 to 2 at three months in 75.1 per cent. By severity, functional independence was reached by 93.4 per cent of those with NIHSS 7 or below, 70.9 per cent at NIHSS 8 to 14, and 46.4 per cent at NIHSS 15 or above. Those are the most directly applicable outcome figures available to an Indian resident, and this tool prints them.
Bedside pearl. The 0.4 mg/kg dose has been tested and should not be used. EXTEND-IA TNK Part 2 found no added benefit over 0.25 mg/kg, and NOR-TEST 2 Part A was stopped early for higher mortality and symptomatic haemorrhage at that dose. Ayushman Bharat PM-JAY carries a separately coded package for tenecteplase thrombolysis in acute ischaemic stroke, which is in practice the strongest national endorsement of the drug in India.
9. The Four-Point Change, and Why It Sits Inside the Instrument's Own Noise
A rise of four or more NIHSS points is the conventional definition of neurological deterioration, and a rise of four or more together with haemorrhage on imaging is the conventional definition of symptomatic intracranial haemorrhage. The DEFUSE 3 protocol uses it in both senses, which is a traceable source for the convention.
Point to note. Josephson and colleagues found that for 7 of 11 videotaped patients, the total NIHSS differed by four points or more between the fifth and ninety-fifth percentile of 7,405 raters. The threshold for deterioration therefore lies inside the between-rater noise of the instrument. A four-point change should trigger a repeat scan and a reassessment, which is what DEFUSE 3 does with it. It does not on its own establish that the patient has deteriorated, particularly when the two scores were recorded by different examiners. Serial NIHSS should be done by the same examiner wherever the roster allows, and the examiner's name should be recorded against each score.
10. ASPECTS: The Ten Regions, and What the Threshold Now Is
What it is. Ten points on a normal scan, one subtracted for each region showing early ischaemic change, so a lower score means a larger infarct. Seven regions are read at the ganglionic level, at the thalamus and basal ganglia: caudate head, lentiform nucleus, internal capsule, insular ribbon, and the cortical regions M1 anterior, M2 lateral to the insula and M3 posterior. Three are read immediately above the ganglia: M4, M5 and M6, anterior, lateral and posterior, roughly 2 cm above M1 to M3.
Why the threshold changed. ASPECTS 6 or more was the thrombectomy gate for most of the last decade, and it is still the figure in the Indian Stroke Association endovascular consensus, whose references stop at 2017. Six large-core trials between 2022 and 2024, RESCUE-Japan LIMIT, SELECT2, ANGEL-ASPECT, TENSION, LASTE and TESLA, moved the line. The 2026 AHA/ASA Class 1 band within six hours is now ASPECTS 3 to 10, with ASPECTS 0 to 2 carrying a Class 2a in patients under 80 without significant mass effect.
Bedside pearl for Indian practice. TESLA is the one large-core trial that selected patients on non-contrast CT alone within 24 hours, which is the imaging reality in most Indian district hospitals, and it did not meet its primary endpoint. TENSION also used non-contrast CT but capped the window at 12 hours and was positive. Where perfusion imaging is not available, the honest position is that the evidence for treating a very large core is weaker than the headline Class 1 recommendation suggests.
11. The Prognostic Scores This Tool Computes, and the Ones It Refuses To
Computed. The DRAGON score, because every one of its six weights is stated in the primary paper and validated externally: dense artery or early infarct signs, pre-stroke Rankin above 1, age, glucose, onset-to-treatment time and baseline NIHSS, area under the curve 0.84 in derivation and 0.80 in validation. SPAN-100, because it is a single arithmetic rule, age plus NIHSS at or above 100, with the authors' own caveat that it must never be used to withhold thrombolysis. And the ICH score, whose five components are quoted verbatim in the primary paper.
Refused, and why. The HAT, SEDAN, SITS-SICH, THRIVE, ASTRAL and iScore instruments all appear in the stroke literature and all would have fitted here. In each case the point weights within the components could not be obtained from the primary paper in preparing this module, only the component names and the score-to-risk mapping. For HAT specifically, the two calculator sites that publish weights render the NIHSS bands as "15 to 20" and "20 or above", which overlap at exactly 20 and cannot both be right. A numeric band with an ambiguous edge is the single defect pattern that has recurred most often in this system, so none of these is computed until its own table has been read.
12. Stroke Mimics, and the Score That Runs the Other Way
Mimics account for close to half of hospital admissions for suspected stroke, and between 22 and 33 per cent of acute stroke activations in telestroke series. The common ones are seizure and the post-ictal state, migraine with aura, functional disorder, metabolic disturbance and above all hypoglycaemia, sepsis and syncope.
The FABS instrument is six items scored one point each: absence of Facial droop, no Atrial fibrillation, Age below 50, systolic Blood pressure below 150, history of Seizures, and isolated Sensory symptoms without weakness. A score of 3 or more identifies a mimic with 90 per cent sensitivity and 91 per cent specificity, area under the curve 0.95. Note that it runs in the opposite direction to every other score here: a high FABS argues against a stroke. It was derived only in patients whose CT was already negative and who presented within 4.5 hours, so this tool raises it as a caution rather than as a diagnosis, and it never withholds treatment.
13. What This Tool Does Not Cover
Paediatric stroke, cerebral venous sinus thrombosis, subarachnoid haemorrhage beyond recognising it and stopping, cervical artery dissection, carotid endarterectomy and stenting decisions, the full aetiological workup for stroke in the young, post-stroke rehabilitation and the driving and occupational advice that follows discharge. It also does not read a scan: every imaging finding here is entered by the clinician, and the tool has no way of knowing whether it was entered correctly.
Algorithm References & Evidence Base
- Prabhakaran S, Gonzalez NR, Zachrison KS, et al. 2026 Guideline for the Early Management of Patients With Acute Ischemic Stroke: A Guideline From the American Heart Association/American Stroke Association. Stroke. 2026;57:e316-e436.
- The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med. 1995;333(24):1581-1587.
- 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.
- Woo D, Broderick JP, Kothari RU, et al. Does the National Institutes of Health Stroke Scale favor left hemisphere strokes? Stroke. 1999;30(11):2355-2359.
- Fink JN, Selim MH, Kumar S, et al. Is the association of National Institutes of Health Stroke Scale scores and acute magnetic resonance imaging stroke volume equal for patients with right- and left-hemisphere ischemic stroke? Stroke. 2002;33(4):954-958.
- Desai SM, Rocha M, Starr M, et al. Laterality is an independent predictor of endovascular thrombectomy in patients with low National Institute of Health Stroke Scale. J Stroke Cerebrovasc Dis. 2018;27(11):3172-3176.
- Josephson SA, Hills NK, Johnston SC. NIH Stroke Scale reliability in ratings from a large sample of clinicians. Cerebrovasc Dis. 2006;22(5-6):389-395.
- Lyden P, Raman R, Liu L, et al. NIHSS training and certification using a new digital video disk is reliable. Stroke. 2005;36(11):2446-2449.
- Poungvarin N, Viriyavejakul A, Komontri C. Siriraj stroke score and validation study to distinguish supratentorial intracerebral haemorrhage from infarction. BMJ. 1991;302(6792):1565-1567.
- Efstathiou SP, Tsioulos DI, Zacharos ID, et al. A new classification tool for clinical differentiation between haemorrhagic and ischaemic stroke. J Intern Med. 2002;252(2):121-129.
- Badam P, Solao V, Pai M, Kalantri SP. Poor accuracy of the Siriraj and Guy's hospital stroke scores in distinguishing haemorrhagic from ischaemic stroke in a rural, tertiary care hospital. Natl Med J India. 2003;16(1):8-12.
- Goswami RP, Karmakar PS, Ghosh A. Bedside utility of clinical scoring systems in classifying stroke. Indian J Med Sci. 2013;67(5-6):137-145.
- Kochar DK, Joshi A, Agarwal N, et al. Poor diagnostic accuracy and applicability of Siriraj stroke score, Allen score and their combination in differentiating acute haemorrhagic and thrombotic stroke. J Assoc Physicians India. 2000;48(6):584-588.
- Prasad K, Dash D, Kumar A. Validation of the Hindi version of National Institute of Health Stroke Scale. Neurol India. 2012;60(1):40-44.
- Baby P, Srijithesh PR, Reddy AV, et al. Transcultural adaptation and validation of Kannada version of the National Institute of Health Stroke Scale. Ann Indian Acad Neurol. 2022;25(2):224-228.
- Sylaja PN, Pandian JD, Kaul S, et al. Indian Registry of Ischemic Stroke Treated With Tenecteplase (IRIS-TNK): safety and outcomes of 0.2 mg/kg tenecteplase. J Am Heart Assoc. 2024;13:e036382.
- Campbell BCV, Mitchell PJ, Churilov L, et al. Tenecteplase versus alteplase before thrombectomy for ischemic stroke. N Engl J Med. 2018;378(17):1573-1582.
- Menon BK, Buck BH, Singh N, et al. Intravenous tenecteplase compared with alteplase for acute ischaemic stroke in Canada (AcT): a pragmatic, registry linked, randomised, open-label, non-inferiority trial. Lancet. 2022;400(10347):161-169.
- Goyal M, Menon BK, van Zwam WH, et al. Endovascular thrombectomy after large-vessel ischaemic stroke: a meta-analysis of individual patient data from five randomised trials (HERMES). Lancet. 2016;387(10029):1723-1731.
- Sarraj A, Hassan AE, Abraham MG, et al. Trial of endovascular thrombectomy for large ischemic strokes (SELECT2). N Engl J Med. 2023;388(14):1259-1271.
- Barber PA, Demchuk AM, Zhang J, Buchan AM. Validity and reliability of a quantitative computed tomography score in predicting outcome of hyperacute stroke before thrombolytic therapy: Alberta Stroke Programme Early CT Score. Lancet. 2000;355(9216):1670-1674.
- Pexman JH, Barber PA, Hill MD, et al. Use of the Alberta Stroke Program Early CT Score (ASPECTS) for assessing CT scans in patients with acute stroke. AJNR Am J Neuroradiol. 2001;22(8):1534-1542.
- Katz BS, McMullan JT, Sucharew H, Adeoye O, Broderick JP. Design and validation of a prehospital scale to predict stroke severity: Cincinnati Prehospital Stroke Severity Scale. Stroke. 2015;46(6):1508-1512.
- Heldner MR, Hsieh K, Broeg-Morvay A, et al. Clinical prediction of large vessel occlusion in anterior circulation stroke: mission impossible? J Neurol. 2016;263(8):1633-1640.
- Strbian D, Meretoja A, Ahlhelm FJ, et al. Predicting outcome of IV thrombolysis-treated ischemic stroke patients: the DRAGON score. Neurology. 2012;78(6):427-432.
- Saposnik G, Guzik AK, Reeves M, Ovbiagele B, Johnston SC. Stroke Prognostication using Age and NIH Stroke Scale: SPAN-100. Neurology. 2013;80(1):21-28.
- Hemphill JC 3rd, Bonovich DC, Besmertis L, Manley GT, Johnston SC. The ICH score: a simple, reliable grading scale for intracerebral hemorrhage. Stroke. 2001;32(4):891-897.
- Puneeth UM, Bhatia R, Sreenivas V, et al. Validation of ICH and ICH-GS scores in an Indian cohort: impact of medical and surgical management. J Stroke Cerebrovasc Dis. 2019;28(8):2213-2220.
- Greenberg SM, Ziai WC, Cordonnier C, et al. 2022 Guideline for the Management of Patients With Spontaneous Intracerebral Hemorrhage. Stroke. 2022;53(7):e282-e361.
- Ma L, Hu X, Song L, et al. The third Intensive Care Bundle with Blood Pressure Reduction in Acute Cerebral Hemorrhage Trial (INTERACT3): an international, stepped wedge cluster randomised controlled trial. Lancet. 2023;402(10395):27-40.
- Gao Y, Chen W, Pan Y, et al. Dual antiplatelet treatment up to 72 hours after ischemic stroke (INSPIRES). N Engl J Med. 2023;389(26):2413-2424.
- Goyal N, Tsivgoulis G, Male S, et al. FABS: an intuitive tool for screening of stroke mimics in the emergency department. Stroke. 2016;47(9):2216-2220.
- Khurana D, Padma MV, Bhatia R, et al. Recommendations for the early management of acute ischemic stroke: a consensus statement for healthcare professionals from the Indian Stroke Association. J Stroke Med. 2018;1(2):79-113.
- Department of Health Research, Ministry of Health and Family Welfare, Government of India. Standard Treatment Workflow: Image Guided Management of Stroke. Indian Council of Medical Research; July 2024.
- Directorate General of Health Services, Ministry of Health and Family Welfare, Government of India. Guidelines for Prevention and Management of Stroke. NPCDCS; 2019.
How to Cite This Tool
AMA Style:
Umakanth S. NIH Stroke Scale and Acute Stroke Pathway. MEDiscuss. Published 2026. Accessed .
Vancouver Style:
Umakanth S. NIH Stroke Scale and Acute Stroke Pathway [Internet]. MEDiscuss.org; 2026 [cited ]. Available from:
Save this calculation
Whatever you type here stays on this device and is not sent anywhere. The server receives only a scrambled code made from it, so nobody with access to the server can tell which patient a saved calculation belongs to. Enter the same nickname the next time to see that patient's earlier values. What is stored
Last revised: 20 August 2026