Measurement properties of the National Institutes of Health Stroke Scale for people with right- and left-hemisphere lesions: further analysis of the clomethiazole for acute stroke study-ischemic (class-I) trial.

OBJECTIVE

To assess the psychometric properties of the National Institutes of Health Stroke Scale (NIHSS) in people with either left or right acute hemisphere stroke for the purpose of improving the scale's sensitivity in detecting neurologic impairment.

DESIGN

Secondary analysis of data from the Clomethiazole for Acute Stroke Study-Ischemic using the Rasch partial credit model. We evaluated the data's measurement properties using item-total correlations, Rasch item fit statistics, principle component analysis of standardized person and item residuals, differential item functioning, separation reliability, and the separation ratio.

SETTING

Original data were collected in academic and community hospitals as part of a clinical trial.

PARTICIPANTS

People with acute ischemic stroke who were seen within 12 hours of onset: 380 people with left-hemisphere stroke and 347 with right-hemisphere stroke.

INTERVENTIONS

Not applicable.

MAIN OUTCOME MEASURE

The NIHSS.

RESULTS

Items of the NIHSS function differently in the right- and left-hemisphere lesion groups. We constructed for each group separate linear scales consisting of a subset of items of the NIHSS to improve its measurement properties.

CONCLUSIONS

Our findings provide initial support for the use of individual, targeted scales for measurement of impairment after ischemic stroke. Low person separation reliability may be a consequence of the sample, which included only people with large ischemic cortical strokes.