Individualized treatment regimens under correlated data with multiple outcomes

2025-09-26

AI Summary

Estimating individualized treatment rules (ITRs) under competing risks is challenging due to limitations of conventional methods—namely, reliance on a single endpoint or single robustness assumption—which fail to address outcome interdependence and uncertainty in unobserved cause-of-failure attribution. Method: We propose a doubly robust regression framework that mitigates bias from unobserved failure cause misclassification by weighting over all potential competing risks, and—novelty—extends doubly robust estimation to clustered data (e.g., patients nested within transplant centers), enabling proper covariance adjustment for correlated observations. Contribution/Results: Applied to real-world kidney transplantation data, our method shows no statistically significant reduction in overall survival for HCV-negative recipients receiving HCV-positive donor kidneys, supporting expanded use of high-risk donors. This work establishes a new paradigm for causal precision medicine under competing risks, balancing theoretical rigor with practical applicability.

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
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Data transparency and provenance

Abstract

Precision medicine involves developing individualized treatment regimes (ITRs) which allow for treatment decisions to be tailored to patient characteristics. Naturally, the identification of the optimal regime, that is, the rule which maximizes patient outcomes, is of interest. Several procedures for estimating optimal ITRs from observational data have been proposed; however, relatively few methods exist for estimating optimal ITRs in the presence of competing risks. Previous approaches either target one particular cause of failure, or rely on singly-robust estimators. We propose a novel doubly-robust regression-based method for estimating optimal ITRs which accounts for the uncertainty related to the unobserved cause of failure by averaging over all possible causes, or targeting the most likely cause. Our approach is straightforward to implement, and we demonstrate an extension to incorporate clustering, motivated by the question of for whom kidney transplantation with hepatitis C virus (HCV)-positive donors is safe, using data from the Organ Procurement and Transplantation Network. Our analysis suggests that a large portion of HCV-negative kidney recipients would see their overall survival unchanged if they were instead provided a kidney from an HCV-positive donor. The estimated treatment rules could be used to provide more efficient allocation of HCV-positive kidneys, increasing the donor pool.

Problem

• Develops individualized treatment rules for optimizing patient outcomes
• Addresses competing risks in treatment estimation using doubly-robust methods
• Improves allocation efficiency of HCV-positive kidneys for transplantation

Innovation

• Doubly-robust regression method for optimal treatment regimes
• Averaging over all possible failure causes uncertainty
• Incorporating clustering for individualized transplantation safety analysis