Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 295 - Causal, Robust, and Machine Learning for Survival Outcomes
Type: Invited
Date/Time: Wednesday, August 11, 2021 : 3:30 PM to 5:20 PM
Sponsor: Lifetime Data Science Section
Abstract #317004
Title: Towards Double Robustness under the Cox Marginal Structural Model
Author(s): Ronghui Xu* and Denise Rava and Jelena Bradic
Companies: University of California at San Diego and Ucsd and University of California, San Diego
Keywords: inverse probability weighting; augmented IPW; machine learning; causal inference

The Cox Marginal Structural Model (MSM) has been widely used to draw causal inference from survival data of observational studies. The typical estimation approach under the MSM is inverse probability weighting (IPW), which is known to be inefficient and also inconsistent when the IP weights are not correctly estimated. Due to the non-collapsibility of the Cox regression model, double robustness (DR) is not a straightforward and possibly an ill-posed problem. We explore the extent to which, both theoretically and empirically, the augmented IPW (AIPW) may achieve DR.

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

Back to the full JSM 2021 program