12/5/2014

Post date: Dec 10, 2014 3:22:46 PM

"Nonparametric Benefit-risk Assessment Using Marker Processes in the Presence of a Terminal Event"

Yifei Sun, Department of Biostatistics, JHU

Benefit-risk assessment is a crucial step in medical decision process. In many biomedical studies, both longitudinal marker measurements and time to a terminal event serve as important endpoints for benefit-risk assessment. The effect of an intervention or a treatment on the longitudinal marker process, however, can be in conflict with that on the time to the terminal event; thus questions arise as to how to properly integrate the two endpoints to provide a conclusive benefit-risk evaluation. In this article, we present a unified framework for benefit-risk assessment using the longitudinal marker processes and the time to a terminal event. We propose a cumulative weighted marker process to synthesize information from the two endpoints, and use its mean function at a pre-specified time point as a benefit-risk summary measure. Two self-contained nonparametric estimators, in the sense that they can be derived using data from the same study, are constructed for the proposed benefit-risk summary measure. Nonparametric test statistics for group comparisons are developed, and the asymptotic null distributions of the statistics are derived. The framework can accommodate the case where benefit-risk assessment is based on a multiple event process with a terminal event. The finite-sample properties of the proposed estimators and test statistics are examined through simulation studies. An analysis of data from an AIDS clinical trial is used to illustrate the proposed approaches.