Screening for cognitive deficits after stroke: a comparison of three screening tools

To assess the concurrent validity of three screening tests for focal cognitive impairments after stroke.

Comparison of results from the screening tests with those from a more comprehensive neuropsychological battery.

Stroke rehabilitation wards of a general hospital and a rehabilitation hospital.

Forty-nine stroke patients (25-91 years, 35% women).

Screening tests were the Cognistat, the Screening Instrument for Neuropsychological Impairments in Stroke (SINS) and the Clock Drawing Test. Health professionals, blind to the results of the reference method, did the screening. Reference method was a neuropsychological assessment based on the Norwegian Basic Neuropsychological Assessment, classifying the patients as ;impaired' or ;not impaired' within the following cognitive domains: language, visuospatial function, attention and neglect, apraxia, speed in unaffected arm, and memory.

The best sensitivity (95% confidence interval) was achieved for language problems by Cognistat, naming (80%, 44-98); for visuospatial dysfunction, attention deficits and reduced speed, all by SINS visuocognitive (82%, 60-95, 72%, 39-94, and 78%, 56-93, respectively); and for memory problems by Cognistat memory (69%, 52-87). The data were insufficient to assess any subtest for apraxia. Sensitivity in detecting deficits in any domain was 82% (71-94) for the Cognistat composite score, 71% (57-85) for the SINS composite score, and 63% (49-78) for the most sensitive score of the Clock Drawing Test.

The Cognistat and the SINS may be used as screening instruments for cognitive deficits after stroke, but cannot replace a neuropsychological assessment. The Clock Drawing Test added little to the detection of cognitive deficits.