Application of competing risks analysis improved prognostic assessment of patients with decompensated chronic heart failure and reduced left ventricular ejection fraction.

Objective: The Kaplan-Meier method may overestimate absolute mortality risk (AMR) in the presence of competing risks. Urgent heart transplantation (UHT) and ventricular assist device implantation (VADi) are important competing events in heart failure. We sought to quantify the extent of bias of the Kaplan-Meier method in estimating AMR in the presence of competing events and to analyze the effect of covariates on the hazard for death and competing events in the clinical model of decompensated chronic heart failure with reduced ejection fraction (DCHFrEF).
Study Design and Setting: We studied 683 patients. We used the cumulative incidence function (CIF) to estimate the AMR at 1 year. CIF estimate was compared with the Kaplan-Meier estimate. The Fine-Gray subdistribution hazard analysis was used to assess the effect of covariates on the hazard for death and UHT/VADi.
Results: The Kaplan-Meier estimate of the AMR was 0.272, whereas the CIF estimate was 0.246. The difference was more pronounced in the patient subgroup with advanced DCHF (0.424 vs. 0.338). The Fine-Gray subdistribution hazard analysis revealed that established risk markers have qualitatively different effects on the incidence of death or UHT/VADi.
Conclusion: Competing risks analysis allows more accurately estimating AMR and better understanding the association between covariates and major outcomes in DCHFrEF.
(Copyright © 2018 Elsevier Inc. All rights reserved.)