Effectiveness of preventive in-home geriatric assessment in well functioning, community-dwelling older people: Secondary analysis of a randomized trial

OBJECTIVES: To determine whether preventive in-home comprehensive geriatric assessment (CGA) prevents functional decline in community-dwelling older p ersons with different baseline functional status: (1) without any basic act ivities of daily living (BADL) dependency at baseline; and (2) without any instrumental ADL (IADL) and basic ADL dependency at baseline. DESIGN: Subgroup analyses of a 3-year randomized controlled trial. SETTING: The city of Santa Monica, California. PARTICIPANTS: Participants came from the original population (n = 414) of c ommunity-living older persons aged 75 years and older who participated in a trial testing the effectiveness of annual preventive in-home CGA. For the first subgroup analysis, we excluded subjects (n = 27) who were dependent i n one or more BADL before randomization (final sample size, n = 387); for t he second subgroup analysis, we excluded 93 additional subjects who were de pendent in one or more IADL before randomization (final sample size, n = 29 4). INTERVENTION: Annual preventive in-home CCA, with quarterly home visits by gerontologic nurse practitioners, for 3 years. MEASUREMENTS: Functional status data were collected through yearly in-home interviews by independent observers. Subjects were classified as (1) indepe ndent in both BADL and IADL, (2) dependent in IADL but independent in BADL, or (3) dependent in both IADL and BADL. RESULTS: In both subgroup analyses, there was no difference in survival bet ween intervention and control subjects. In the subgroup with no BADL impair ment at baseline, intervention subjects spent significantly fewer days depe ndent in both BADL and IADL during each year of the study (5 days vs 14 day s, P = .022; 13 vs 33, P = .016; and 19 vs 44, P = .014 for years 1, 2, and 3, respectively) and over all 3 years combined (36 days vs 92 days, P = .0 16) in bivariate analyses. In multivariate analyses, the intervention reduc ed time spent in complete (BADL and IADL) dependency (P = .028). In the sub group of subjects without any IADL or BADL impairment at baseline, no signi ficant differences were apparent in the number of days spent in complete in dependence and days spent in complete dependency. Intervention group subjec ts spent more days in partial dependency during Year 1 (24 days vs 9 days, P = .021), but the difference was not significant during Year 2 (47 vs 29, P = .088), Year 3 (49 vs 41, P = .370), and over all 3 years combined (120 vs 79, P = .123) as well as in multivariate analysis (P = .062). CONCLUSION: These findings support the hypothesis that in-home preventive v isits delay the onset of disability in people without initial BADL impairme nt. Further studies in larger samples are needed to determine optimal inter vention strategies and effectiveness among well functioning older people.