9/30/2016

Post date: Oct 10, 2016 3:25:41 PM

Title: Semiparametric Regression for Composite Endpoints with Different Component Censoring times

Speaker: Guoqing Diao, Department of Statistics, George Mason University

Abstract:

Composite endpoints with censored data are commonly used as study outcomes in clinical trials. For example, progression free survival (PFS) is a widely used composite endpoint with disease progression and death as the two components. PFS is often defined as the time from randomization to the earlier occurrence of disease progression or death from any cause. The censoring times of the two components could be different for patients not experiencing the endpoint event. Conventional approaches, such as taking the minimum of the censoring times of the two components as the censoring time for PFS, may suffer from efficiency loss and could potentially produce biased estimate of the treatment effect. In this article, we propose a new likelihood-based approach that decomposes the endpoints and models time from randomization to progression and time from progression to death separately. In this new approach, the censoring time for different components will be distinguished and linked to individual components. Therefore, this approach makes full use of available information and improves the estimate of the treatment effect on PFS. The proposed method can be readily extended to other composite endpoints in general. Extensive simulation studies demonstrate that the proposed method outperforms several conventional approaches and is also robust to various model misspecifications. An application to a prostate cancer clinical trial is provided.