Speaker:
Subgroup analyses are frequently conducted in comparative effectiveness research and randomized clinical trials to assess evidence of heterogeneous treatment effect across patient subpopulations. This dissertation develops and extends propensity score methods for causal subgroup analysis with continuous and survival outcomes. In observational studies, we propose to use the overlap weighting method to achieve exact balance within subgroups, and combine overlap weighting and LASSO, to balance the bias-variance tradeoff in subgroup analysis. In randomized trials, we extend the propensity score weighting method for covariate adjustment to improve the precision and power of subgroup analysis. We also design a new diagnostic graph, the Connect-S plot, for visualizing the subgroup covariate balance. The proposed methods are applied to real data examples to evaluate the subgroup causal effects.
Mentor: Laine Thomas, PhD
Zoom Link: Please contact ladonna.huseman@duke.edu for details on how to join.