One challenge that may arise with cluster randomization is that, although a cluster’s units are typically assumed to be of equivalent size, this may not be true of CRTs in healthcare settings. Clusters such as physician practices or clinics may be of substantially different sizes, which can affect the statistical power of the study and decisions about sample size (Cook et al, 2016). To address these issues, study statisticians need an estimate of the range in potential sample sizes. For example, in a trial that randomly assigns clinics to an intervention that will be applied to patients who are newly diagnosed with diabetes, the statistician will need information about the number of such patients coming into each clinic over the past several months. The statistician will then consider the range in sample sizes when calculating the number of clinics and patients needed for the study.
Cook AJ, Delong E, Murray DM, Vollmer WM, Heagerty PJ. 2016. Statistical lessons learned for designing cluster randomized pragmatic clinical trials from the NIH Health Care Systems Collaboratory Biostatistics and Design Core. Clin Trials. 13:504-512. doi:10.1177/1740774516646578. PMID: 27179253.