Frailty and mortality risk in patients with severe covid-19: prognosis beyond the age criterion

OBJECTIVE: To investigate the association between frailty and death in hospitalized patients with COVID-19. METHODOLOGY: Prospective cohort study with patients ≥ 50 years hospitalized with COVID-19. Frailty was assessed using the Clinical Frailty Scale and the frailty index. Patients with a Clinical Frailty Scale score ≥ 5 were considered frail. The primary endpoints were mortality at 30 and 100 days after hospital admission. We used Cox proportional hazard models to investigate the association between frailty and mortality. We also explored whether frailty predicted different mortality levels among patients within strata of similar age and acute disease severity (Sequential Organ Failure Assessment score). RESULTS: A total of 1,830 patients were included (mean age 66 years; 58% men; 27% frail according to Clinical Frailty Scale score). The mortality risk at 30 days (adjusted hazard ratio = 1.7; 95% CI 1.4 - 2.1; p <0.001) and 100 days (adjusted hazard ratio = 1.7; 95% CI 1.4 - 2.1; p <0.001) was almost double for frail patients. The Clinical Frailty Scale also predicted different mortality levels among patients within strata of similar age and acute disease severity. Frailty intensified the effect of acute disease severity on the risk of death (p for interaction = 0.01). Of note, the Clinical Frailty Scale achieved outstanding accuracy to identify frailty according to the frailty index (area under the ROC curve = 0.94; 95% CI 0.93 - 0.95). CONCLUSIONS: Our results encourage the use of the Clinical Frailty Scale in association with measures of acute disease severity to determine prognosis and promote adequate resource allocation in hospitalized patients with COVID-19.