Evaluation of the Fragire grid using data from the Share survey.

Summary In 2014, the Cnav's Social Action Department decided to acquire a new frailty detection tool to improve the targeting of people at risk of loss of autonomy and to offer them prevention programs better adapted to their needs. This new frailty detection tool, called the Fragire grid, was developed by the Burgundy Franche-Comté interregional gerontology cluster (PGI) based on a clinical cohort. In this study, we want to evaluate to what extent the Fragire score is 1) reproducible in a general population, 2) able to identify frail elderly people using Fried's phenotype as a reference indicator, and 3) predictive of the onset of loss of autonomy (appearance of restrictions in activities of daily living and development of personal assistance). We use panel data from the Share survey (European survey on health, ageing and retirement among people aged 50 and over), representative of the population in ordinary households in metropolitan France, between 2004 and 2010. These data allow us to reconstruct the Fragire grid algorithm according to the methodology developed by the AIP, and to measure the association of the Fragire score with different variables related to loss of autonomy at two and four year intervals. Our results indicate that, without being perfect, this score allows the detection of frail people. It also makes it possible to anticipate the use of assistance and future disabilities. These results underline the importance of this new frailty indicator for developing effective prevention policies to target individuals at risk of loss of autonomy.