HIV Working Group

Evelyn Zhang, Ph.D.
Research Associate, Department of Biostatistics, Harvard T. H. Chan School of Public Health

“Subgroup analysis for Survival Outcomes: An Individualized Treatment Selection Approach for First line ART”

ABSTRACT: Historically, the main goal for randomized clinical trials is to evaluate the overall treatment effect in efficacy and toxicity, but there has been a growing interest to understand the heterogeneity in treatment effect across participants characterized by their pre treatment characteristics. Much literature exists on the limitations of traditional subgroup analysis and new methods have emerged to address some of the issues, but their implementation to the clinical or public health research is limited, especially when the outcome is time to event type of endpoint. We implemented an individualized treatment selection approach involving kernel estimating procedure to conduct subgroup analyses for ACTG A5257, a phase III randomized trial on first line ART, to evaluate whether the overall treatment effect can be generalized to a more individualized level.