Cluster trial overview — Cluster-randomized parallel trial

Question

Define the intervention question and why group-level assignment is appropriate.

Define clusters

Specify cluster boundaries, participant eligibility, target population and potential contamination.

Protocol and registration

Prespecify consent, recruitment timing, estimands and analysis; obtain approvals and register prospectively.

Recruit clusters and participants

Recruit clusters and participants while documenting any recruitment after allocation and its potential selection effects.

Randomize clusters

Generate concealed cluster allocation, recording stratification or restrictions and the implementation sequence.

Deliver intervention

Deliver assigned cluster interventions and monitor fidelity, contamination and deviations.

Follow

Retain cluster and participant outcomes and losses with distinct units and denominators.

Analyze clustering

Use estimand-compatible cluster-aware estimation and small-sample inference where needed.

Report

Report cluster and participant flow, intracluster correlation assumptions, weighting and design limitations.

Cluster trial overview

End flow

Protocol and decision register

Record the protocol, approvals and registration where applicable. Preserve planned decisions and amendments separately from results.

Protocol and rationale evidence

Attach the dated protocol, approvals and amendments. A clinical trial registry publicly records the trial plan, such as ClinicalTrials.gov or an eligible WHO primary registry such as ISRCTN. Record the registry name, identifier and registration date before first enrollment. WHO ICTRP is a search platform, not itself a registry. Registration does not imply randomization.