Development and validation of an electronic database-based frailty index to predict mortality and hospitalization in a population-based study of adults with SARS-CoV-2

Abstract

Background: Electronic health databases are used to identify people at risk of poor outcomes. Using electronic regional health databases (e-RHD), we aimed to develop and validate a frailty index (FI), compare it with a clinically based FI, and assess its association with health outcomes in community-dwellers with SARS-CoV-2. Methods: Data retrieved from the Lombardy e-RHD were used to develop a 40-item FI (e-RHD-FI) in adults (i.e., aged ≥18 years) with a positive nasopharyngeal swab polymerase chain reaction test for SARS-CoV-2 by May 20, 2021. The considered deficits referred to the health status before SARS-CoV-2. The e-RHD-FI was validated against a clinically based FI (c-FI) obtained from a cohort of people hospitalized with COVID-19 and in-hospital mortality was evaluated. e-RHD-FI performance was evaluated to predict 30-day mortality, hospitalization, and 60-day COVID-19 WHO clinical progression scale, in Regional Health System beneficiaries with SARS-CoV-2. Results: We calculated the e-RHD-FI in 689,197 adults (51.9% females, median age 52 years). On the clinical cohort, e-RHD-FI correlated with c-FI and was significantly associated with in-hospital mortality. In a multivariable Cox model, adjusted for confounders, each 0.1-point increment of e-RHD-FI was associated with increased 30-day mortality (Hazard Ratio, HR 1.45, 99% Confidence Intervals, CI: 1.42–1.47), 30-day hospitalization (HR per 0.1-point increment = 1.47, 99%CI: 1.46–1.49), and WHO clinical progression scale (Odds Ratio = 1.84 of deteriorating by one category, 99%CI 1.80–1.87). Conclusion: The e-RHD-FI can predict 30-day mortality, 30-day hospitalization, and WHO clinical progression scale in a large population of community-dwellers with SARS-CoV-2 test positivity. Our findings support the need to assess frailty with e-RHD.