Validation of a telephone version of the mini-mental state examination.

To assess the construct validity of a telephone-administered version of the Mini-Mental State Examination (MMSE). Validity testing by comparing a telephone version of the MMSE administered first to a face-to-face evaluation done several days later. Outpatient geriatric assessment center. 100 of 175 consecutive referrals. MMSE and a brief neuropsychological screening test (BNPS) face-to-face and a telephone version of the MMSE as part of the Adult Lifestyles and Function Interview (ALFI-MMSE). Test scores of the two MMSE versions correlated strongly for all subjects (Pearson's r = 0.85, P = 0.001) and remained significant for the cognitively intact (P = 0.02) and questionably (P = 0.002), mildly (P = 0.0001), and moderately (P = 0.003) demented. Comparison of the two versions' equivalent 22 items revealed no significant difference for scores of all subjects (P = 0.07) but with a trend toward higher scores in the original version. Diminished hearing, reported either by the subject (P = 0.003) or by the collateral source (P = 0.02) was associated with lower scores on the telephone version. Five individual test items were biased by the route of test administration. Sensitivity and specificity relative to the BNPS were 67% and 100% for the ALFI-MMSE and 68% and 100% for the MMSE, respectively. The scores on the ALFI-MMSE correlated strongly with the scores of the original version given face-to-face in subjects undergoing geriatric assessment. The results indicate that the ALFI-MMSE could be a useful and economical tool to screen for cognitive impairment.