B1323
Title: Evaluating informative cluster size in cluster randomized trials
Authors: Bryan Blette - Vanderbilt University Medical Center (United States) [presenting]
Brennan Kahan - University College London (United Kingdom)
Michael Harhay - University of Pennsylvania (United States)
Fan Li - Yale University (United States)
Abstract: In cluster-randomized trials, the average treatment effect among participants (p-ATE) may be different from the cluster average treatment effect (c-ATE) when informative cluster size is present, i.e., when treatment effects or participant outcomes depend on cluster size. In such scenarios, mixed-effects models and GEEs with exchangeable correlation structures are biased for both the p-ATE and c-ATE estimands. GEEs with an independent correlation structure or analyses of cluster-level summaries are recommended in practice. However, when cluster size is non-informative, mixed-effects models and GEEs with exchangeable correlation structures can provide unbiased estimation and notable efficiency gains over other methods. Thus, hypothesis tests for informative cluster size would be useful to assess this key assumption's validity formally. Model-assisted and randomization-based tests are developed for informative cluster size in cluster-randomized trials. Simulation studies are constructed to examine the operating characteristics of these tests, showing they have appropriate Type I error control and meaningful power, and contrast them to existing model-based tests used in the observational study setting. The proposed tests are applied to data from a recent cluster-randomized trial, and practical recommendations for using these tests are discussed.