Ability of a computerized geriatric assessment to predict need for change in living status among elderly living at...

The objective of this study was to determine the ability of a software program used in a primary care physician's office to predict the need for alternate living arrangement in a cohort of community-dwelling elderly. An analysis was conducted involving 124 consecutive patients between February 10, 1990, and December 20, 1991, in my private medical practice. These patients, all older than 65 years, underwent a computer-assisted geriatric assessment. Two scales--the Geriatric Functional Rating Scale (GFRS) and the Functional Assessment Screening Questionnaire (FASQ)--were compared for their accuracy at predicting change in living status during the 12- to 24-month period following the assessment. Similar analysis of the Tinetti gait and balance test was also performed. Ten and one-half percent of subjects (n = 13) required a change in living status during the study period. Ten went to nursing homes and three joined relatives' households. The GFRS was 62% accurate (8/13) and the FASQ was 54% (7/13) accurate in predicting this change. Abnormality in both gait and balance was predictive 77% (10/13) of the time. Combining all three parameters raised the successful prediction rate to 85% (11/13). Neither GFRS, FASQ, nor gait/balance testing was predictive of death. A computer software program designed to facilitate performance of geriatric assessments in primary care physicians' offices has a high rate of predictive capability relative to future need for change in living status among community-dwelling elderly. Further studies comparing this software program with traditional geriatric assessment protocols are suggested.