Neurological Assessment Instruments

Neurological instruments are usually described by where they sit on the path from impairment to participation. Impairment measures such as the National Institutes of Health Stroke Scale, the Fugl-Meyer Assessment, the Motricity Index and the Glasgow Coma Scale quantify the neurological deficit itself. Activity and disability measures such as the Modified Rankin Scale, the Rivermead Mobility Index, the Motor Assessment Scale and the Functional Ambulation Categories describe what the person can do as a result. Participation measures such as the Community Integration Questionnaire and the Craig Handicap Assessment describe the life they are able to return to.

Disease-specific scales sit across all three levels. The Expanded Disability Status Scale in multiple sclerosis, the Hoehn and Yahr staging in Parkinson's disease, the Unified Huntington's Disease Rating Scale and the ALS Functional Rating Scale each combine elements rather than measuring one level cleanly.

Most of the widely used scales in this group are clinician-rated, which makes rater training the dominant source of measurement error in multi-site work. Certification for the National Institutes of Health Stroke Scale exists precisely because untrained raters disagree materially, and the same applies to the Expanded Disability Status Scale, where the mid-range is notoriously dependent on ambulation and rater judgement.

Several of these scales are also ordinal and unevenly spaced. A one-point change at one end of the Modified Rankin Scale does not represent the same amount of clinical change as a one-point change at the other, which is why stroke trials frequently analyse the whole distribution rather than dichotomising at a single cut. Choosing the analysis at design time is part of choosing the instrument.

Check whether rater certification is required and budget for it. If a study uses the National Institutes of Health Stroke Scale or the Expanded Disability Status Scale across sites, certification and periodic re-checks are part of the protocol, not an optional extra.

Check the timing convention. Neurological scores change rapidly in the acute phase, and a scale recorded at an unspecified point in the first days after a stroke is not comparable between sites. Define the assessment window precisely.

Check that impairment and disability are not being used interchangeably. Interventions frequently improve impairment measures without changing disability, and reporting one as though it demonstrated the other is a recurring problem in this literature. If the claim is about function, the endpoint should be a function measure.

Does the NIH Stroke Scale need certified raters?

Yes for research use, and it is good practice clinically. Untrained raters disagree materially on several items, and certification with periodic re-checks is standard in stroke trials. Record who rated and when.

Why is the Modified Rankin Scale analysed as a distribution rather than a cut-off?

Because its points are not evenly spaced in clinical terms, and dichotomising discards information. Ordinal analysis of the whole distribution is now common in stroke trials and usually has more statistical power than a single cut.

What is the difference between an impairment and a disability measure?

An impairment measure quantifies the neurological deficit, such as weakness or level of consciousness. A disability measure describes what the person can do in daily life. Treatments often move one without moving the other, so the endpoint has to match the claim being made.

Can neurological scales be scored remotely?

Some patient-reported scales transfer directly. Most clinician-rated neurological examinations do not, because they depend on physical examination. Validated telephone or video adaptations exist for a few instruments and should be reported as the adapted version, not the original.