Table 3.

Reporting of (a) how clustering was considered during sample size estimation and analysis and (b) justification for using a cluster randomized design

	N = 31 trials (%)
Did sample size/power calculations account for the cluster design?
Not presented₳	11 (35%)
Yes, used patient-level data and accounted for clustering (e.g., random effects model)	11 (35%)
Yes, used cluster-level summaries	3 (10%)
No, used patient-level data without accounting for clustering	3 (10%)
Unclear	1 (3%)
Other¥	2 (6%)
Did the analysis for primary outcome account for clustering?
Yes, used patient-level data and accounted for clustering	17 (55%)
Yes, used cluster-level summaries	5 (16%)
No, used patient-level data without accounting for clustering ₱	7 (23%)
Unclear/other¥	2 (6%)
Justification for utilizing a cluster randomized design (categories were not mutually exclusive)
None provided	16 (52%)
Avoid contamination	15 (48%)
Logistical or administrative convenience	2 (6%)

^₳One study presented power calculation, but it was a post hoc power analysis

^¥This may have included using an inappropriate method for the proposed primary outcome, or the study accounted for clustering but not based on the primary outcome measure (e.g., they assumed a continuous outcome, but the primary endpoint was a proportion)

^₱One study accounted for repeated events within patients but did not report accounting for within-cluster correlation; another study reported using a generalized linear mixed model but did not specify whether they accounted for the effect of the cluster as random effect