Targeting elders for geriatric evaluation and management: reliability, validity, and practicality of a questionnaire

Geriatric evaluation and management (GEM) is most cost-effective when provided to persons at high risk for functional decline or heavy use of health services. Identifying high risk members of elderly populations is, therefore, the first step in conducting successful GEM programs. We have developed and tested a mailed, self-administered, eight-item questionnaire to identify home-dwelling elders at risk for heavy use of hospitals. Scored by a logistic formula, this questionnaire estimates each respondent's probability of repeated admission (Pra) to a hospital within four years. Its primary purpose is to help select elders who are likely to benefit from outpatient GEM. We created this instrument by analyzing data from half the subjects in the Longitudinal Study of Aging (LSOA); its test-retest reliability is high (r = 0.78). In a preliminary test of its predictive validity among the other half of the LSOA subjects, the instrument prospectively identified high-risk elders who went on to use hospitals at twice the rate of their lower-risk peers. In a separate study of its predictive validity among low-income urban elders, the instrument again identified a high-risk group that went on to use hospital days at twice the rate of its lower-risk counterpart. In a pilot study, we used the questionnaire to identify potential recipients of outpatient GEM. The identified elders appeared to be appropriate candidates for GEM. They averaged 9.6 significant medical problems, 6.7 significant prescription medications, and two IADL limitations. We are now using this instrument to identify subjects for a randomized clinical trial of outpatient GEM