The Compensation For Few Clusters In Clustered Randomized Trials With Binary Outcomes