Conference Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 267 - Nonparametric Statistics Student Paper Competition Presentations
Type: Topic Contributed
Date/Time: Tuesday, August 9, 2022 : 10:30 AM to 12:20 PM
Sponsor: Section on Nonparametric Statistics
Abstract #323283
Title: Efficient Targeted Learning of Heterogeneous Treatment Effects for Multiple Subgroups in Observational Studies
Author(s): Waverly Wei* and Maya Petersen and Mark Van der Laan and Zeyu Zheng and Chong Wu and Jingshen Wang
Companies: UC Berkeley and UC Berkeley and University of California Berkeley and UC Berkeley and Florida State University and UC Berkeley
Keywords: nonparametrics; causal inference; semiparametrics; precision medicine; heterogeneous treatment effect; observational study
Abstract:

In biomedical science, analyzing treatment effect heterogeneity plays an essential role in assisting personalized medicine. The main goals of analyzing treatment effect heterogeneity include estimating treatment effects in clinically relevant subgroups and to predict whether a patient subpopulation might benefit from a particular treatment. Conventional approaches often evaluate the subgroup treatment effects via parametric modeling and can thus be susceptible to model mis-specifications. In this manuscript, we take a model-free semiparametric perspective and aim to efficiently evaluate the heterogeneous treatment effects of multiple subgroups simultaneously from observational data under the one-step targeted maximum-likelihood estimation (TMLE) framework. We further discuss a variation of the one-step TMLE that can be robust to the presence of small estimated propensity scores in finite samples. From our simulations, the method demonstrates substantial finite sample improvements compared to conventional methods. In case studies, our method unveils the treatment effect heterogeneity of statin usage in increasing type 2 diabetes risk and in decreasing Alzheimer's disease risk.


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

Back to the full JSM 2022 program