All Times EDT
Keywords: vaccine, cluster randomization, real world
Cluster-randomization is one type of randomization where the unit of randomization is not a subject but a group of subjects, by community, hospital, health service provider, household or other unit. They are used in clinical trials when the intervention needs to be applied to clusters. It may reflect the real world clinical practice, and clinical trials using cluster randomization can also generate useful findings. However, there were often concerns of the biases that could be introduced by cluster randomization to the results. We explored some clinical trial data simulated by using cluster randomization and compared them to the standard simple randomization method. This helps us to understand better the potential biases of cluster randomization and how the data compares between traditional clinical trials and real world settings.