NEJM Evid. 2026 Sep;5(9):EVIDctw2600148. doi: 10.1056/EVIDctw2600148. Epub 2026 Aug 25.
ABSTRACT
AbstractCluster-randomized trials are increasingly relevant to the evaluation of preventive and treatment strategies for chronic disease, especially when interventions are delivered through groups, practices, communities, or health systems rather than individuals. Using cardiovascular prevention as an illustrative example, we review key design, conduct, and analytic issues that determine when cluster randomization is appropriate and how such trials should be designed, analyzed, and interpreted. Particular attention is given to the choice of the unit of randomization, protection of internal validity, implications for statistical efficiency, and alignment of outcomes and analyses with the level and mechanism of intervention delivery. A clear understanding of these principles should strengthen the design, interpretation, and policy relevance of cluster-randomized trials.
PMID:42640171 | DOI:10.1056/EVIDctw2600148