Evaluation of a frailty index based on a comprehensive geriatric assessment in a population based study of elderly Canadians

Background and aims: Objectives were to develop a frailty index (FI) based on a standard comprehensive geriatric assessment (CGA) derived from a clinical examination; to assess the validity of the FI-CGA and to compare its precision with other frailty measures. Methods: Design: Secondary analysis of a prospective cohort study, with five-year follow-up data. Setting: Second phase of the Canadian Study of Health and Aging (CSHA-2); clinical examinations were performed in clinics, nursing homes, and patients' homes. Participants: People selected (as either cognitively impaired cases or unimpaired controls) to receive the CSHA-2 clinical examination (n=2305; women=1431). Measurements: Clinical and performance-based measures and diagnostic data were extracted to correspond to the 10 impairment domains and the single comorbidity domain of a CGA. The proportion of deficits accumulated in each domain was calculated to yield the FI-CGA. The FI-CGA was validated and its predictive ability compared with other frailty measures. Results: Within the seven grades of fitness/frailty identified, subjects with greater frailty were older, less educated, and more likely to be women. The FI-CGA correlated highly with a previously validated, empirically-derived frailty index (r=0.76). Frailty was associated with higher risk of death (for each increment in frailty, the hazard ratio, adjusted for age, sex and education, was 1.23 (95% CI 1.18-1.29) and institutionalization (HR 1.20; 1.10-1.32). Conclusions: In a population survey, the FI-CGA is a valid means of quantifying frailty from routinely collected data.