The HFFI, a new frailty index for assessing long-term outcomes in heart failure

Abstract

Abstract Introduction Patients hospitalized due to heart failure (HF) compose a heterogeneous population whose prognosis is difficult to forecast. Frailty is a well-recognized prognostic marker in multiple chronic diseases, including HF; however, frailty evaluation is often subjective and standardized and objective prediction models are lacking. We aimed to evaluate if an objective and simple index – Heart Failure Frailty Index (HFFI)– can predict long-term outcomes in this population. Methods A retrospective analysis of 258 patients admitted to a Cardiology ward due to HF was performed. The variables albumin, C-reactive protein levels, age and body mass index (BMI) were selected for frailty assessment. After attributing points for each variable, according to odds ratio on univariate analysis, the HFFI was calculated (range 0–8), resulting from the sum of the points attributed to each variable. Kaplan-Meyer and Cox-regression analysis were performed to evaluate HFFI association with 24-month mortality (24MM) and 24-month composite endpoint of rehospitalization or death (24MH). Results Mean patient age was 75 (±11) years; 51% were men. 45.7% had atrial fibrillation, 15.9% hat history of acute myocardial infarction, 67.8% had hypertension. Mean LVEF was 47% (±17). A LVEF <40% was present in 40% of patients. 24MM was 11.5% and 24 MH was 58%. Patients were considered frail if they had an HFFI ≥3. Kaplan-Meyer curve analysis revealed a significantly lower median time to 24MM in frail patients, as assessed by HFFI, comparing to non-frail patients (585±33 days vs 697±12 days, mortality rate: 25.4% vs 5.7%, χ2=18.156, p<0.001). There was also a significantly lower median time to 24MH in frail patients (336±34 days vs 449±24 days, combined endpoint rate: 76.1% vs 50.3%, χ2=10.884, p=0.001). Cox regression analysis demonstrated that HFFI independently predicts 24MM (HR: 1.364, p=0.002) and 24MH (HR: 1.106, p=0.035), even after adjustment for other prognostic markers, such as history of atrial fibrillation, previous myocardial infarction, diabetes and natriuretic peptides serum level at index event. Conclusion HFFI is a simple and objective frailty index correlated with 24MM and 24MH, being an independent prognostic marker in this population. Its use may help to identify patients with a high risk of mortality or readmission, in need of specialized care. Funding Acknowledgement Type of funding sources: None.