Construction and Validation of a Frailty Index in Primary Care in Italy: The Health-Search Frailty Index

Abstract Recognizing frailty in primary care is important to implement personalized care pathways and for prognostication. The aim of this study was to build and validate a frailty index based on routinely collected primary care data in Italy. We used clinical data from 308,280 Italian primary care patients 60+ with at least 5 years of follow-up, part of the Health Search Database. A heuristic algorithm was used to select the deficits to be included in a highly performant frailty index. The fitness of the index was assessed through the c-statistics derived by survival models. Results were externally validated using the Swedish National Study on Aging and Care in Kungsholmen (SNAC-K). After testing 3.4 million of deficits combinations, 25 deficits were selected to be included in the Health Search Frailty Index (HS-FI). After adjusting by sex, age and geographical area, the HS-FI was associated with 5-year mortality (HR per 0.1 increase 1.99; 95%CI 1.95-2.02) and hospitalization rate (HR per 0.1 increase 1.25; 95%CI 1.23-1.27). In the external validation cohort, HS-FI independently predicted mortality, hospitalization, incident disability, incident dementia, and incident falls. This is the first frailty index built following a data-driven approach, using national representative primary care data. The implementation of such tool – derived by routinely collected data – in primary care software will ease the prompt, comparable and reliable recognition of frailty at the population level.