Primary Image

RehabMeasures Instrument

National Institutes of Health Stroke Scale


  • Measures the severity of symptoms associated with cerebral infarcts; used as a quantitative measure of neurological deficit post stroke.
  • A retrospective scoring algorithm has been found to be reliable for research purposes (Williams et al, 2000).

Link to Instrument

Instrument Details

Acronym NIHSS

Area of Assessment

Vision & Perception

Assessment Type


Administration Mode

Paper & Pencil




  • Stroke Recovery


Key Descriptions

  • The NIHSS is a composite scale derived from the Toronto Stroke Scale, the Oxbury Initial Severity Scale, the Cincinnati Stroke Scale and the Edinburgh-2 Coma Scale.
  • 15 items assessing severity of impairment in LOC, ability to respond to questions and obey simple commands, papillary response, deviation of gaze, extent of hemianopsia, facial palsy, resistance to gravity in the weaker limb, plantar reflexes, limb ataxia, sensory loss, visual neglect, dysarthria and aphasia severity.
  • Items are graded on a 3- or 4-point ordinal scale; 0 means no impairment.
  • Scores range from 0 – 42. Higher scores indicate greater severity.
  • Stroke severity may be stratified on the basis of NIHSS scores as follows (Brott et al, 1989):
    1) Very Severe: >25
    2) Severe: 15 – 24
    3) Mild to Moderately Severe: 5 – 14
    4) Mild: 1 – 5

Number of Items


Time to Administer

6 minutes

Required Training

Training Course

Age Ranges


18 - 64


Instrument Reviewers

Updated by Carmen Capo-Lugo, PT, PhD and Dorian Rose PT, PhD and the Stroke Edge task force in 2016.

ICF Domain

Body Function

Measurement Domain


Professional Association Recommendation

Recommendations for use of the instrument from the Neurology Section of the American Physical Therapy Association’s Multiple Sclerosis Taskforce (MSEDGE), Parkinson’s Taskforce (PD EDGE), Spinal Cord Injury Taskforce (PD EDGE), Stroke Taskforce (StrokEDGE), Traumatic Brain Injury Taskforce (TBI EDGE), and Vestibular Taskforce (VEDGE) are listed below. These recommendations were developed by a panel of research and clinical experts using a modified Delphi process.


For detailed information about how recommendations were made, please visit:




Highly Recommend




Reasonable to use, but limited study in target group / Unable to Recommend


Not Recommended


Recommendations for use based on acuity level of the patient:



(CVA < 2 months post)

(SCI < 1 month post)

(Vestibular < 6 weeks post)


(CVA 2 to 6 months)

(SCI 3 to 6 months)


(> 6 months)






Recommendations based on level of care in which the assessment is taken:


Acute Care

Inpatient Rehabilitation

Skilled Nursing Facility



Home Health








Recommendations for entry-level physical therapy education and use in research:


Students should learn to administer this tool? (Y/N)

Students should be exposed to tool? (Y/N)

Appropriate for use in intervention research studies? (Y/N)

Is additional research warranted for this tool (Y/N)





Not reported


  • The NIHSS may be most useful for early prognostication assessment, whereas the Barthel Index may be more useful for planning rehabilitation interventions (Kasner, 2006)
  • The NIHSS was originally designed to assess differences among clinical trial interventions. However, the NIHSS is increasingly used as an initial assessment tool and for planning post-acute care (Kasner, 2006).
  • 4 items have poorly reliability or are redundant (level of consciousness, facial weakness, ataxia, and dysarthria (Kasner, 2006)

Do you see an error or have a suggestion for this instrument summary? Please e-mail us!


back to Populations

Cut-Off Scores

Stroke severity may be stratified on the basis of NIHSS scores as follows (Brott et al, 1989):

  • Very Severe:  >25

  • Severe: 15 – 24

  • Mild to Moderately Severe: 5 – 14  

  • Mild: 1 – 5 


Acute Stroke: (Schlegel et al, 2003; Rundek et al, 2000)

Outcomes related to NIHSS scores at admission:

  • Scores of <5; 80% of stroke survivors will be discharged to home

  • Score between 6 and 13 typically require acute inpatient rehabilitation

  • Scores of >14 frequently require long-term skilled care

Normative Data

Acute Stroke: (Williams et al, 1999; n = 34, first stroke, patients assessed 1 and 3 months (+/- 1 week) after stroke)

Health-Related Quality of Life (NIHSS mean scores)










1 Month





3 Months




A Lot Worse

A Little Worse


A Lot Worse

A Little Worse









Test/Retest Reliability

Acute Stroke: (Goldstein & Samsa, 1997; 4 patients assessed by 30 physicians and 29 study coordinators; 3 months between assessments)

  • Excellent test-retest reliability; ICC = 0.93

Interrater/Intrarater Reliability

Acute Stroke: (Goldstein & Samsa, 1997)

  • Excellent interrater reliability; ICC = 0.95


Acute Stroke: Interrater Agreement (Goldstein et al, 1989; n = 20 with 2 independent observers)

  • Adequate to Excellent agreement was found for 9 of the 13 items on the NIHSS (Kappa = 0.32 to 0.79); lowest levels of agreement were found for the Facial palsy (Kappa = 0.22) and limb ataxia (Kappa = -0.16) items.

(Oh et al., 2012; n=19 with 21 raters Korean version of the NIHSS (K-NIHSS)).

  • Excellent interrater reliability; ICC=0.998. Lowest levels of agreement were found for Facial paresis (Kappa = 0.439) and dysarthria (Kappa = 0.465) and highest for consciousness commands (0.950) and sensory function (0.911)..

  • Excellent intrarater reliability; ICC=0.969

Hindi version of NIHSS (Prasad et al, 2012; n=107 with 2 raters)

  • Excellent interrater reliability; ICC=0.995

Criterion Validity (Predictive/Concurrent)

Acute Stroke: Predictive validity (Adams et al, 1999; n = 1268)

  • NIHSS scores at baseline predicted outcome at 7 and 90 days

  • An excellent outcome was achieved by nearly two-thirds of the survivors who scored 3 or less at day 7

  • Only a few patients who scored more than 15 at baseline achieved excellent outcomes after 90 days


Acute Stroke: Predictive validity (Baird et al, 2001; n = 66; <48 hours post-stroke)

  • NIHSS combined with Magnetic Resonance Diffusion-Weighted imaging (MR DWI) and volume of ischaemic brain tissue on MR DWI significantly predicted stroke recovery


Acute Stroke: (Bohannon et al, 2002; n = 92, mean age = 70.0 (12.4) years; NIHSS was administered while patients were still in the emergency department, prior to admission)

  • Poor* (but significant) correlation with length of stay (r = 0.276)
  • Adequate* correlation with hospital charges (r = 0.320)
  • Adequate* correlation with discharge destination (home or elsewhere) (r= - 0.355)

*Significant at p < 0.05


Acute Stroke: Concurrent Validity (Fink et al, 2002)

  • Adequate to Excellent correlations with diffusion weighted MRI lesion volumes (= 0.48 right, = 0.58 left); and perfusion-weight hypoperfusion volumes (= 0.62 right, = 0.60 left)

Chronic Stroke: Concurrent Validity (Peters, et al., 2015)

  • Examined the concurrent validity of the NHISS with the Stroke Impact Scale. No association between NIHSS and SIS-physical dimension (Spearman's rho = -.036; p=.666). SIS-overall perception of recovery (Spearman's rho = -.039; p=.640) nor SIS ADL/IADL score (Spearman's rho = -.054; p=.520).

Predictive Validity

  • K-NIHSS at baseline (within 7 days of stroke onset) and modified Rankin Scale at 90-days post-onset was significantly positively correlated (Spearman's ρ=0.600; p<0.001).

  • K-NIHSS significantly negatively correlated with the Barthel Index for the same time period (Spearman's ρ=-0.647; p<0.001).

  • Hindi version (HV_NIHSS; Prasad et al., 2012, n=107 patients with stroke) and

Glasgow Coma Scale at 3 months are highly correlated (Spearman’s rho= -0.863, p<0.001)

  • HV_NIHSS and Barthel Index at 3 months (Spearman’s rho: -0.829, p<0.001)

  • HV_NIHSS and Modified Rankin Scale at 3 months (Spearman’s rho: 0.851, p<0.001)

Construct Validity

Acute Stroke: (Schlegel et al, 2003; Rundek et al, 2000)

Outcomes related to NIHSS scores at admission:

  • Scores of <5; 80% of stroke survivors will be discharged to home

  • Score between 6 and 13 typically require acute inpatient rehabilitation

  • Scores of >14 frequently require long-term skilled care


(Oh et al., 2012 n=207, patients with acute ischemic stroke Korean Version of the NIHSS (K-NIHSS))

Construct Validity determined by comparison with the Glasgow Coma Scale (Spearman ρ=-6.71; p<0.001) and deemed acceptable.

Content Validity

Items are based on components of a standard neurological examination (Kasner, 2006)


(Oh et al., 2012) Korean Version of the NIHSS (K-NIHSS)

Used the Content Validity Index (CVI) which is the proportion of expert raters (n=11) rating an item higher than 3 points on a 4-point ordinal rating scale; a rating of 1 denotes an irrelevant item, and a rating of 4 denotes an extremely relevant item. Ten of the NIHSS items received a 1.0 meaning that all of the expert raters, rated that item higher than a 3. The visual fields item received a CVI of 0.91 as one rater, rated that item a "2" in terms of relevancy. Items with a CVI of at least 0.78 are accepted as valid (Lynn; 1986). The means scores of the CVI for each item ranged from 3.46-3.73.

Floor/Ceiling Effects

Acute Stroke:

  • Floor effects are less commonly reported in the literature to date
  • Ceiling effects (Pickard et al, 2005) A ceiling effect was observed at 6 months with the NIHSS effecting 20% of patients who completed the measure


Acute Stroke: (Brott et al, 1989)

NIHSS scores were compared to infarction size (measured by computed tomography) on 65 patients at 1 week post stroke. 10 items demonstrated an average of 25% change over 7 days. However, changes in limb ataxia and best gaze may have been overstated.


Adams, H. P., Davis, P. H., et al. (1999). "Baseline NIH Stroke Scale score strongly predicts outcome after stroke: A report of the Trial of Org 10172 in Acute Stroke Treatment (TOAST)." Neurology 53: 126-131. Find it on PubMed

Baird, A. E., Dambrosia, J., et al. (2001). "A three-item scale for the early prediction of stroke recovery." Lancet 357: 2095-2099. Find it on PubMed

Bohannon, R. W., Lee, N., et al. (2002). "Postadmission function best predicts acute hospital outcomes after stroke." Am J Phys Med Rehabil 81: 726-730. Find it on PubMed

Brott, T., Adams, H. P., Jr., et al. (1989). "Measurements of acute cerebral infarction: a clinical examination scale." Stroke 20(7): 864-870. Find it on PubMed

Fink, J. N., Selim, M. H., et al. (2002). "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 33: 954-958. Find it on PubMed

Goldstein, L. B., Bertels, C., et al. (1989). "Interrater reliability of the NIH stroke scale." Arch Neurol 46(6): 660-662. Find it on PubMed

Goldstein, L. B. and Samsa, G. P. (1997). "Reliability of the National Institutes of Health Stroke Scale. Extension to non-neurologists in the context of a clinical trial." Stroke 28(2): 307-310. Find it on PubMed

Kasner, S. E. (2006). "Clinical interpretation and use of stroke scales." Lancet Neurol 5(7): 603-612. Find it on PubMed

Peters, H. T., et al. (2015). "The National Institutes of Health Stroke Scale Lacks Validity in ChronicHemiparetic Stroke." J Stroke Cerebrovasc Dis 24(10): 2207-2212. Find it on PubMed.

Oh, M. S., et al. (2012). "Validity and reliability of a korean version of the national institutes of healthstroke scale." J Clin Neurol 8(3): 177-183. Find it on PubMed.

Pickard, A. S., Johnson, J. A., et al. (2005). "Responsiveness of generic health-related quality of life measures in stroke." Qual Life Res 14: 207-219. Find it on PubMed

Prasad, K., et al. (2012). "Validation of the Hindi version of National Institute of Health Stroke Scale."Neurol India 60(1): 40-44. Find it on PubMed.

Rundek, T., Mast, H., et al. (2000). "Predictors of resource use after acute hospitalization: the Northern Manhattan Stroke Study." Neurology 55: 1180-1187. Find it on PubMed

Schlegel, D., Kolb, S. J., et al. (2003). "Utility of the NIH Stroke Scale as a predictor of hospital disposition." Stroke 34: 134-137. Find it on PubMed

Williams, L. S., Yilmaz, E. Y., et al. (2000). "Retrospective assessment of initial stroke severity with the NIH Stroke Scale." Stroke 31: 858-862. Find it on PubMed