Appropriate statistical methods were infrequently used in cluster-randomized crossover trials

There is a need for improved application of appropriate analysis and sample size methods, and reporting, in CRXO trials.