Accounting for death as a competing risk in cancer-associated thrombosis studies.

In recent years, clinical investigation in the field of cancer-associated thrombosis has identified a number of potential biomarkers to predict the risk of developing a thrombotic event. Similarly, large randomized clinical trials have demonstrated the benefit of low molecular weight heparins in the treatment as well as prevention of venous thromboembolic events (VTE) in cancer patients. However, the most common statistical methodology to evaluate the occurrence of VTE has been the Kaplan-Meier approach. When used to estimate the cumulative incidence of VTE in cancer studies, the Kaplan-Meier method over-estimates the cumulative incidence function due to failure to account for death as a competing risk for the development of VTE. A more appropriate statistical estimate of cumulative incidence of VTE in cancer studies is the competing risk model. This review describes the theoretical and mathematical basis for estimating the cumulative incidence function by the competing risk model for the analysis of VTE outcomes in cancer-associated thrombosis.