Bayesian Sensitivity Analysis for Causal Estimation with Time-varying Unmeasured Confounding

TL;DR

This paper develops Bayesian sensitivity analysis methods to assess the impact of unmeasured confounding on causal estimates in longitudinal studies with time-varying confounders, providing practical tools for observational data analysis.

Contribution

It introduces extended Bayesian sensitivity analysis approaches for time-varying unmeasured confounding, enhancing causal inference in longitudinal observational studies.

Findings

Methods perform well in simulation studies

Applied to pediatric disease registry data

Guidance on implementation of sensitivity analysis

Abstract

Causal inference relies on the untestable assumption of no unmeasured confounding. Sensitivity analysis can be used to quantify the impact of unmeasured confounding on causal estimates. Among sensitivity analysis methods proposed in the literature for unmeasured confounding, the latent confounder approach is favoured for its intuitive interpretation via the use of bias parameters to specify the relationship between the observed and unobserved variables and the sensitivity function approach directly characterizes the net causal effect of the unmeasured confounding without explicitly introducing latent variables to the causal models. In this paper, we developed and extended two sensitivity analysis approaches, namely the Bayesian sensitivity analysis with latent confounding variables and the Bayesian sensitivity function approach for the estimation of time-varying treatment effects with longitudinal observational data subjected to time-varying unmeasured confounding. We investigated the performance of these methods in a series of simulation studies and applied them to a multi-center pediatric disease registry data to provide practical guidance on their implementation.