Causal Inference using Multivariate Generalized Linear Mixed-Effects Models

Abstract

Dynamic prediction of causal effects under different treatment regimens is an essential problem in precision medicine. It is challenging because the actual mechanisms of treatment assignment and effects are unknown in observational studies. We propose a multivariate generalized linear mixed-effects model and a Bayesian g-computation algorithm to calculate the posterior distribution of subgroup-specific intervention benefits of dynamic treatment regimes. Unmeasured time-invariant factors are included as subject-specific random effects in the assumed joint distribution of outcomes, time-varying confounders, and treatment assignments. We identify a sequential ignorability assumption incorporating treatment assignment heterogeneity, that is analogous to balancing the latent treatment preference due to unmeasured time-invariant factors. We present a simulation study to assess the proposed method's performance. The method is applied to observational clinical data to investigate the efficacy of continuously using mycophenolate in different subgroups of scleroderma patients.

Speaker

I am a postdoctoral researcher at the Biostatistics Department and School of Medicine at the Johns Hopkins University. I work with Dr. Scott Zeger on longitudinal causal inference methodology to account for patient heterogeneity, with applications in Scleroderma and SARS-CoV-2, and Dr. Zheyu Wang on Bayesian latent variable modeling of progression in the Alzheimer's Disease. I received my Ph.D. training in Biostatistics with Dr. Joseph W. Hogan at Brown University.

My research focuses on the development of innovative Bayesian latent variable modeling, statistical ensemble learning, and causal inference methods for electronic health records. Applications involve infectious and chronic diseases such as HIV, Scleroderma, Alzheimer’s disease, and SARS-CoV-2.