Estimands and doubly robust estimation for cluster-randomized trials with survival outcomes

2025-10-09

AI Summary

This study addresses causal inference for right-censored survival outcomes in cluster-randomized trials (CRTs). We rigorously distinguish and define cluster-level versus individual-level treatment effect estimands within the potential outcomes framework—the first such formalization for CRTs with censored survival data. We propose a doubly robust estimator that simultaneously accommodates covariate-dependent censoring and intra-cluster correlation, guaranteeing consistency if either the censoring model or the survival model is correctly specified. The method integrates cluster-level inverse probability weighting with individual-level augmented estimation and employs a delete-1 jackknife procedure for variance estimation and confidence interval construction. Simulation studies demonstrate favorable finite-sample properties. Applied to a completed CRT, our approach substantially enhances interpretability and robustness of causal effect estimates for survival outcomes.

Technology Category

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

Application Category

• Responsible Web: Measurement, analysis, and circumvention of Web censorship
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Cluster-randomized trials (CRTs) are experimental designs where groups or clusters of participants, rather than the individual participants themselves, are randomized to intervention groups. Analyzing CRT requires distinguishing between treatment effects at the cluster level and the individual level, which requires a clear definition of the estimands under the potential outcomes framework. For analyzing survival outcomes, it is common to assess the treatment effect by comparing survival functions or restricted mean survival times between treatment groups. In this article, we formally characterize cluster-level and individual-level treatment effect estimands with right-censored survival outcomes in CRTs and propose doubly robust estimators for targeting such estimands. Under covariate-dependent censoring, our estimators ensure consistency when either the censoring model or the outcome model is correctly specified, but not necessarily both. We explore different modeling options for the censoring and outcome models to estimate the censoring and survival distributions, and investigate a deletion-based jackknife method for variance and interval estimation. Extensive simulations demonstrate that the proposed methods perform adequately in finite samples. Finally, we illustrate our method by analyzing a completed CRT with survival endpoints.

Problem

• Define cluster-level and individual-level treatment effects for survival outcomes
• Develop doubly robust estimators for survival data under censoring conditions
• Propose variance estimation methods for cluster-randomized trials with censoring

Innovation

• Doubly robust estimators for cluster-randomized survival trials
• Consistent estimation under covariate-dependent censoring models
• Jackknife method for variance and interval estimation