Cognitive function testing in comprehensive geriatric assessment. A comparison of cognitive test performance in residential and clinic settings.

Tests of cognitive function are frequently used in geriatric assessment, but the effect of test setting has rarely been explored. To determine the effect of testing site on the performance of elderly patients undergoing a comprehensive geriatric assessment, we administered the Mini-Mental State Exam to 116 geriatric patients in the clinic and at their residence. Their cognitive abilities varied from normal to severely impaired. The patients' scores were 1.5 +/- 3.6 (mean +/- SD) higher at their residence. The clinical importance of a difference in score of 1.5 is not clear. For this reason a second analysis was performed in which a difference in scores of five points or greater between settings was considered clinically meaningful. Twenty-five percent (29 of 116) differed by five points or more. Of these 29 patients, 22 (76%) tested better in the residential setting. These differences were statistically significant (P = .001). We conclude that the testing site may affect test performance and that in-home assessment may reveal the optimal cognitive function of geriatric patients.