Online Program Home
My Program

Abstract Details

Activity Number: 656 - Subgroup Evaluations in Advancing Drug Development
Type: Topic Contributed
Date/Time: Thursday, August 2, 2018 : 10:30 AM to 12:20 PM
Sponsor: Biopharmaceutical Section
Abstract #328647
Title: Quantitative Assessment of Risk with Subgroup Pursuit in Clinical Trials
Author(s): Xinzhou Guo* and Xuming He
Companies: University of Michigan and University of Michigan
Keywords: subgroup analysis; decision risk; bootstrap; model-free

Subgroup analysis is the analysis of the treatment effect within the population. In some cases, the treatment effect of a new treatment might be only marginally effective for the population of the original study and a highly promising treatment effect is observed in a sub-population. A managerial decision on whether an additional investment should be made for a trial on the most promising sub-population we have observed needs to be made. However, without a careful risk/reward analysis, the decision of subgroup pursuit can be of high risk because the estimated subgroup effect size from the most promising subgroup tends to be overly optimistic. In this research, we propose a bootstrap-based measure of risk in pursuing a subgroup identified from an existing trial and study the statistical properties of the proposed measure. The measure we propose is intuitive, model-free, easy to calculate and has the potential to be generalized to some complicated situations. The proposed measure of risk can be used as a screening method and help a better-informed decision of subgroup pursuit.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program