Identification of Simple Subgroups of Enhanced Treatment Effect from Randomized Clinical Trials Data

• Activity Number: 12
• Type: Topic Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Jeremy Taylor*+ and Wenting Cheng and Jared Foster
• Companies: University of Michigan and University of Michigan and NICHD
• Keywords: subgroups ; personalized medicine ; clinical trials

Abstract

We consider the situation of a randomized clinical trial of a new treatment versus a standard treatment with a moderate number of baseline covariates. The goal is to identify a simple region of the covariate space from the data, such that future patients should receive the new treatment if in the region and the standard treatment if not in the region. We consider a two-stage procedure. In stage 1, nonparametric regression is used to estimate treatment effects for each subject. In the stage 2, a systematic evaluation of many subgroups of a simple, pre-specified form is performed. Using the criterion of maximizing the expected value of the response for future patients the region of the covariate space is estimated. The methodology will be evaluated in a simulation study and compared to alternatives.