Email Post: The Effect of Cluster Size Variability on Statistical Power in Cluster-Randomized Trials