A cluster randomised study (or cluster randomised trial, CRT) is an experimental design where groups or "clusters" of individuals are randomly assigned to different interventions or conditions. This approach is often used where individual randomisation isn't practical or feasible. Interventions are delivered to the entire group and outcomes are measured across participants in each cluster. CRTs are commonly used in areas like education or public health where treatments need to be applied at the group level.
• Entire clusters (e.g., hospitals, wards, geographical regions, schools, families) are randomised to intervention conditions.
• Differs from individually randomised trials, where individuals themselves are randomised.
Considerations
• Requires careful justification due to larger sample size needs, increased bias risk, and specific ethical concerns
• Appropriate when the intervention is applied at the cluster level with individual level outcome measurement, or individual randomisation risks high contamination
Variations in design
• Parallel Cluster Trial with Baseline:
• Cluster Randomised Cross-Over Design (CRXO):
• Stepped-Wedge Cluster Randomised Trial (SW-CRT):
Hemming, K., Copas, A., Forbes, A., & Kasza, J. (2024). What type of cluster randomized trial for which setting?. Journal of epidemiology and population health, 72(1), 202195
Connolly, P., Miller, S., Kee, F., Sloan, S., Gildea, A., McIntosh, E., & Bland, J. M. (2018). A cluster randomised controlled trial and evaluation and cost-effectiveness analysis of the Roots of Empathy schools-based programme for improving social and emotional well-being outcomes among 8-to 9-year-olds in Northern Ireland. Public Health Research, 6(4). Catalogue Link
Stepped wedge cluster RCTs
Stratified cluster trials