COADVISE: Covariate Adjustment with Variable Selection in Randomized Controlled Trials

2025-01-15

AI Summary

In small-sample, high-dimensional randomized controlled trials, conventional covariate adjustment suffers from low efficiency and poor robustness due to model misspecification and nonlinear covariate–outcome relationships. To address this, we propose the first unified estimation framework that jointly performs variable selection and flexible (linear or nonlinear) covariate adjustment. Our method integrates regularized variable selection, spline-based or other flexible basis function modeling, doubly robust inference, and robust variance estimation—ensuring consistency and efficiency even under outcome model misspecification. We establish its asymptotic efficiency gain theoretically. Simulation studies and analysis of the BestAIR real-world trial demonstrate that our approach reduces estimator variance by up to 32% compared to no adjustment or LASSO-based adjustment. The method is implemented in the open-source R package *Coadvise*.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Adjusting for covariates in randomized controlled trials can enhance the credibility and efficiency of treatment effect estimation. However, handling numerous covariates and their complex (non-linear) transformations poses a challenge. Motivated by the case study of the Best Apnea Interventions for Research (BestAIR) trial data from the National Sleep Research Resource (NSRR), where the number of covariates (p=114) is comparable to the sample size (N=196), we propose a principled Covariate Adjustment with Variable Selection (COADVISE) framework. COADVISE enables variable selection for covariates most relevant to the outcome while accommodating both linear and nonlinear adjustments. This framework ensures consistent estimates with improved efficiency over unadjusted estimators and provides robust variance estimation, even under outcome model misspecification. We demonstrate efficiency gains through theoretical analysis, extensive simulations, and a re-analysis of the BestAIR trial data to compare alternative variable selection strategies, offering cautionary recommendations. A user-friendly R package, Coadvise, is available to facilitate practical implementation.

Problem

• Handling numerous covariates in randomized trials efficiently
• Selecting relevant covariates for outcome adjustment
• Ensuring robust variance estimation despite model misspecification

Innovation

• Variable selection for relevant covariates
• Handles linear and nonlinear adjustments
• Robust variance estimation under misspecification