On Cluster Randomized Trials with the Desirability of Outcome Ranking (DOOR) Endpoints

2026-04-27

AI Summary

This study addresses the challenge of evaluating patient-centered benefit–risk trade-offs in cluster randomized trials, where intracluster correlation and the inability to randomize at the individual level complicate analysis. It presents the first systematic extension of the Desirability of Outcome Ranking (DOOR) framework to cluster-randomized designs. Building on U-statistics and influence functions, the authors develop a unified nonparametric estimation and inference approach that accommodates single- or two-group comparisons, varying cluster sizes, and both small and large sample settings. The method incorporates the Wilcoxon–Mann–Whitney statistic to handle ordinal outcomes. Simulation studies demonstrate its robust performance across diverse clustering structures, and the approach is successfully applied to analyze a crossover cluster randomized trial comparing delayed cord clamping versus umbilical cord milking in newborns.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Calibration & Uncertainty Quantification
• Knowledge Representation and Reasoning: Preferences

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Cluster randomized trials are widely used when individual randomization is logistically infeasible or when correlations between observations cannot be ignored, especially in fields such as ophthalmology, infectious disease, vaccine research, and sociology. The desirability of outcome ranking (DOOR) framework evaluates patient-centric benefit-risk using an ordinal outcome and a Wilcoxon-Mann-Whitney statistic-based approach to compare outcome distributions between interventions. We propose a suite of new methods to extend DOOR to cluster trials based on properties of U-statistics and influence functions to estimate within-cluster and between-cluster treatment effects. These approaches can be applied in different scenarios, including mixtures of clusters with two treatment groups and clusters with only one group, and both small and large numbers of clusters. Simulations demonstrate that the proposed methods perform well under various scenarios regarding the number of clusters and cluster sizes. As an illustration, we apply the proposed methods to a cluster randomized crossover trial comparing delayed cord clamping and umbilical cord milking for newborns.

Problem

• cluster randomized trials
• desirability of outcome ranking
• DOOR
• U-statistics
• influence functions

Innovation

• DOOR
• cluster randomized trials
• U-statistics
• influence functions
• treatment effect estimation