Incorporating External Data into the Analysis of Clinical Trials via Bayesian Additive Regression Trees

TL;DR

This paper introduces a Bayesian additive regression trees (BART) approach for integrating external data into clinical trial analysis, enhancing treatment effect estimation by adjusting for covariates and heterogeneity.

Contribution

The paper proposes a novel BART-based method for incorporating external data into clinical trial analysis, offering flexible adjustment for covariates and heterogeneity.

Findings

BART outperforms hierarchical linear models in simulations.

The method effectively captures heterogeneity across data sources.

Application to an acupuncture trial demonstrates practical utility.

Abstract

Most clinical trials involve the comparison of a new treatment to a control arm (e.g., the standard of care) and the estimation of a treatment effect. External data, including historical clinical trial data and real-world observational data, are commonly available for the control arm. Borrowing information from external data holds the promise of improving the estimation of relevant parameters and increasing the power of detecting a treatment effect if it exists. In this paper, we propose to use Bayesian additive regression trees (BART) for incorporating external data into the analysis of clinical trials, with a specific goal of estimating the conditional or population average treatment effect. BART naturally adjusts for patient-level covariates and captures potentially heterogeneous treatment effects across different data sources, achieving flexible borrowing. Simulation studies demonstrate that BART compares favorably to a hierarchical linear model and a normal-normal hierarchical model. We illustrate the proposed method with an acupuncture trial.