Development of an easy prognostic score for frailty outcomes in the aged

BACKGROUND: identification of frailty is recommended in geriatric practice. However, there is a lack of frailty scores combining easy-to-collect predictors from multiple domains. OBJECTIVE: to develop a frailty score including only self-reported information and easy-to-perform standardised measurements recommended in routine geriatric practice. DESIGN: prospective population-based study. SETTING/PARTICIPANTS: included 1,007 Italian subjects aged 65 and over. MEASUREMENTS: seventeen baseline possible mortality predictors from several domains, 4-year risk of mortality and other adverse health outcomes associated with frailty [fractures, hospitalisation, and new and worsening activities of daily living (ADL) disability]. METHODS: a multivariate Cox model was used to identify the best sub-group of independent predictors and to develop a mortality prognostic score, defined as the number of adverse predictors present. Logistic regression was used to verify whether the score also predicted risk of other frailty outcomes in the cohort survivors. RESULTS: nine independent mortality predictors were identified. Among subjects with score > or =3, each one point increase in the score was associated with a doubling in mortality risk and, among survivors, with an increased risk of all the other adverse health outcomes. CONCLUSIONS: nine easy-to-collect predictors may identify aged people at increased risk of adverse health outcomes associated with frailty.