4 September 2012 | Marion K Campbell director, Gilda Piaggio honorary professor, Diana R Elbourne professor of healthcare evaluation, Douglas G Altman director, for the CONSORT Group
The CONSORT 2010 statement was updated to include guidance for reporting cluster randomised trials (CRTs), which involve randomising groups rather than individuals. CRTs are used when individual randomisation may lead to contamination or when group randomisation is the only feasible method. The updated statement includes a 25-item checklist, with items clarified and expanded, and addresses the unique reporting requirements of CRTs. The checklist now includes items related to cluster design, sample size calculation, intracluster correlation coefficients, and the analysis of data. The statement also provides guidance on the reporting of abstracts of CRTs.
Cluster randomised trials involve two levels of inference: the cluster level and the individual participant level. The effective sample size is smaller than the actual number of participants due to the correlation between participants within clusters. The intracluster correlation coefficient (ICC) is a key factor in determining the sample size and analysis of CRTs. The design of CRTs can vary, including completely randomised, matched, or stratified designs. The conduct of CRTs differs from individually randomised trials in that randomisation is done at the cluster level, and consent may be sought at the cluster level before randomisation.
The analysis of CRTs requires special considerations, such as accounting for the correlation between participants within clusters. The results of CRTs may be interpreted at the cluster level, individual participant level, or both. The quality of reporting of CRTs has been found to be poor in some cases, with deficiencies in reports remaining common. Recent reviews have shown some improvement in reporting, but the quality of reporting remains well below an acceptable level.
The updated CONSORT statement provides a revised checklist for reporting CRTs, including items related to the definition of clusters, the design of the trial, the eligibility criteria for clusters and participants, the interventions, the outcomes, the sample size calculation, the randomisation process, and the implementation of the trial. The statement also provides guidance on the reporting of abstracts of CRTs, including the number of clusters per arm and the level of inference for the primary outcome. The statement emphasizes the importance of clear and detailed reporting to ensure the validity and reliability of CRTs.The CONSORT 2010 statement was updated to include guidance for reporting cluster randomised trials (CRTs), which involve randomising groups rather than individuals. CRTs are used when individual randomisation may lead to contamination or when group randomisation is the only feasible method. The updated statement includes a 25-item checklist, with items clarified and expanded, and addresses the unique reporting requirements of CRTs. The checklist now includes items related to cluster design, sample size calculation, intracluster correlation coefficients, and the analysis of data. The statement also provides guidance on the reporting of abstracts of CRTs.
Cluster randomised trials involve two levels of inference: the cluster level and the individual participant level. The effective sample size is smaller than the actual number of participants due to the correlation between participants within clusters. The intracluster correlation coefficient (ICC) is a key factor in determining the sample size and analysis of CRTs. The design of CRTs can vary, including completely randomised, matched, or stratified designs. The conduct of CRTs differs from individually randomised trials in that randomisation is done at the cluster level, and consent may be sought at the cluster level before randomisation.
The analysis of CRTs requires special considerations, such as accounting for the correlation between participants within clusters. The results of CRTs may be interpreted at the cluster level, individual participant level, or both. The quality of reporting of CRTs has been found to be poor in some cases, with deficiencies in reports remaining common. Recent reviews have shown some improvement in reporting, but the quality of reporting remains well below an acceptable level.
The updated CONSORT statement provides a revised checklist for reporting CRTs, including items related to the definition of clusters, the design of the trial, the eligibility criteria for clusters and participants, the interventions, the outcomes, the sample size calculation, the randomisation process, and the implementation of the trial. The statement also provides guidance on the reporting of abstracts of CRTs, including the number of clusters per arm and the level of inference for the primary outcome. The statement emphasizes the importance of clear and detailed reporting to ensure the validity and reliability of CRTs.