The cognitive assessment screening test (CAST) for dementia
Document Type Article
The purpose of this study was to assess the usefulness, the sensitivity, and the specificity of the Cognitive Assessment. Screening Test (CAST), a paper-and-pencil self-administered cognitive test designed to screen elderly people for possible dementia, for use in general physicians' offices, requiring little expertise or staff time. CAST consists of three parts: part A (relatively easy), part B (more demanding), and part C (self-report of concerns). CAST was administered in two studies to: (1) 19 patients known to be mildly to moderately demented versus 24 age-matched normal controls (to establish cutoff standards); and (2) a "real world" sample of 26 elderly patients not known to be demented, attending a general medicine clinic. The sensitivity and specificity of CAST were compared with the Mini-Mental State Examination (MMSE) and the Blessed Dementia Scale cognitive portion (BDS-cog). In study 1, controls were given a detailed neuropsychological battery; in study 2, all patients were given the neuropsychological battery, which served as the "gold standard" to identify individuals with cognitive impairment. In study 1, the cutoff scores for dementia using CAST (Parts A and B) were established. CAST discriminated demented patients from controls with a sensitivity of 95% and a specificity of 88%; the MMSE had a sensitivity of 74% and a specificity of 100%; and the BDS-cog had a sensitivity of 100% and a specificity of 96%. In study 2, CAST discriminated cognitive impairment with a sensitivity of 88% and a specificity of 100%, the MMSE had a sensitivity of 38% and a specificity of 100%; and the BDS-cog had a sensitivity of 50% and a specificity of 94%. Part C was not used to discriminate demented from normal elderly individuals, but to screen for those concerned about their cognitive functioning. CAST is highly useful as a dementia screening test, with sensitivity and specificity equal to or better than the MMSE and BDS-cog, yet requiring minimal examiner time and little training or experience to administer.