Analyzing multi-center randomized trials with covariate adjustment while accounting for clustering

2025-04-17

AI Summary

This paper addresses the inefficiency in covariate adjustment and undercoverage of confidence intervals arising from ignoring center-level clustering in multicenter randomized trials. We propose a robust estimation framework that jointly incorporates center-specific random effects and covariate adjustment. Specifically, we construct unbiased estimators for the counterfactual mean and average treatment effect (ATE) tailored to randomized-center designs. Inspired by random-effects meta-analysis, we develop a novel inferential method that remains asymptotically valid under model misspecification. Our approach integrates augmented inverse probability weighting (AIPW) with G-computation, embedding generalized linear prediction models and random-effects–based inference. Simulation studies demonstrate substantial gains in estimation efficiency—particularly under strong between-center heterogeneity in treatment effects—while maintaining nominal 95% confidence interval coverage. The method’s robustness and practical utility are empirically validated using data from the WASH Benefits Bangladesh trial.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Intelligent Robots: State Estimation

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

Augmented inverse probability weighting (AIPW) and G-computation with canonical generalized linear models have become increasingly popular for estimating the average treatment effect in randomized experiments. These estimators leverage outcome prediction models to adjust for imbalances in baseline covariates across treatment arms, improving statistical power compared to unadjusted analyses, while maintaining control over Type I error rates, even when the models are misspecified. Practical application of such estimators often overlooks the clustering present in multi-center clinical trials. Even when prediction models account for center effects, this neglect can degrade the coverage of confidence intervals, reduce the efficiency of the estimators, and complicate the interpretation of the corresponding estimands. These issues are particularly pronounced for estimators of counterfactual means, though less severe for those of the average treatment effect, as demonstrated through Monte Carlo simulations and supported by theoretical insights. To address these challenges, we develop efficient estimators of counterfactual means and the average treatment effect in a random center. These extract information from baseline covariates by relying on outcome prediction models, but remain unbiased in large samples when these models are misspecified. We also introduce an accompanying inference framework inspired by random-effects meta-analysis and relevant for settings where data from many small centers are being analyzed. Adjusting for center effects yields substantial gains in efficiency, especially when treatment effect heterogeneity across centers is large. Monte Carlo simulations and application to the WASH Benefits Bangladesh study demonstrate adequate performance of the proposed methods.

Problem

• Estimating treatment effects in multi-center trials with clustering
• Addressing model misspecification in covariate-adjusted analyses
• Improving efficiency and coverage in center-adjusted estimators

Innovation

• AIPW and G-computation for covariate adjustment
• Random-effects meta-analysis inspired inference framework
• Efficient estimators for multi-center clustering effects