Deconfounded Warm-Start Thompson Sampling with Applications to Precision Medicine

TL;DR

This paper introduces DWTS, a method that uses observational data and causal inference techniques to improve adaptive clinical trial efficiency by better initializing Thompson Sampling with confounded data.

Contribution

The paper presents DWTS, a novel approach combining Doubly Debiased LASSO with Thompson Sampling to utilize observational data effectively in clinical trials.

Findings

DWTS outperforms standard LinTS in synthetic environments.

DWTS reduces cumulative regret in virtual cardiovascular risk trials.

Offline causal insights enhance personalized treatment decisions.

Abstract

Randomized clinical trials often require large patient cohorts before drawing definitive conclusions, yet abundant observational data from parallel studies remains underutilized due to confounding and hidden biases. To bridge this gap, we propose Deconfounded Warm-Start Thompson Sampling (DWTS), a practical approach that leverages a Doubly Debiased LASSO (DDL) procedure to identify a sparse set of reliable measured covariates and combines them with key hidden covariates to form a reduced context. By initializing Thompson Sampling (LinTS) priors with DDL-estimated means and variances on these measured features -- while keeping uninformative priors on hidden features -- DWTS effectively harnesses confounded observational data to kick-start adaptive clinical trials. Evaluated on both a purely synthetic environment and a virtual environment created using real cardiovascular risk dataset, DWTS consistently achieves lower cumulative regret than standard LinTS, showing how offline causal insights from observational data can improve trial efficiency and support more personalized treatment decisions.