As with individually randomized trials, a number of considerations need to be addressed up front for CRTs to avoid downstream problems. Potential confounding is always an issue. For example, if older patients are more likely than younger patients to acquire nosocomial infections, it would be important to ensure that one of the arms of the trial does not consist of older patients. In this example, if the clusters are hospital wards, there should be some assurance of balance in the average ages of patients in the wards assigned to one arm compared to the other. Sometimes there are several potential confounders.

Using "restricted" or "constrained" randomization procedures can help minimize the chance of imbalances in baseline characteristics and improve statistical efficiency in parallel-arm CRTs (Ivers et al 2012). These procedures include pair matching, stratification, and covariate-constrained randomization. These methods require prior knowledge about the clusters and the baseline measures used for the restriction process.

Two popular mechanisms for achieving balance are pair matching and stratification. With pair matching, clusters are paired in terms of their potential confounders. Then, within each pair, one cluster is assigned to one of the study arms and the other cluster is assigned to the opposite arm. For example, considering age and sex as potential confounders, clusters would be matched into pairs such that the average age and the percentage of women in the cluster are approximately equal. Likewise, the sizes of the 2 clusters should be similar.

Stratification is a generalization of pair matching, in that strata are formed based on the potential confounders. Within each stratum, a randomization scheme that ensures balance is developed. For example, if there are 11 clusters in a stratum, the randomization would assign 5 clusters to one arm and 6 to the other. However, when there are several confounders, it can be difficult to use stratification or pair matching.

Another method that is increasingly being studied and implemented in CRTs is constrained randomization. Exploiting the fact that all the clusters are identified before randomization, each cluster can be characterized in terms of the levels of several potential confounders. For any possible randomization of this set of clusters, a balance metric is applied to “measure” the amount of imbalance that would exist if that randomization scheme were applied. From this “randomization space,” a single randomization scheme is selected. To learn more about covariate-constrained randomization, see the Covariate-Constrained Randomization section in this chapter of the Living Textbook.

For more information about considerations affecting study design decisions, see Designing With Implementation and Dissemination in Mind.