B0819
Title: A Bayesian non-parametric approach for causal mediation with a post-treatment confounder
Authors: Michael Daniels - University of Florida (United States) [presenting]
Woojung Bae - University of Florida (United States)
Abstract: A new Bayesian non-parametric (BNP) method is proposed for estimating the causal effects of mediation in the presence of a post-treatment confounder. An enriched Dirichlet process mixture (EDPM) is specified to model the joint distribution of the observed data (outcome, mediator, post-treatment confounder, treatment, and baseline confounders). For identifiability, the extended version of the standard sequential ignorability is used, as introduced in a prior study. The observed data model and causal identification assumptions enable estimating and identifying the causal effects of mediation, i.e., the natural direct effects (NDE) and indirect effects (NIE). The method enables easy computation of NDE and NIE for a subset of confounding variables and addresses missing data through data augmentation under the assumption of ignorable missingness. Simulation studies are conducted to assess the performance of the proposed method. Furthermore, this approach is applied to evaluate the causal mediation effect in the Rural LITE trial, demonstrating its practical utility in real-world scenarios.