This paper will discuss the analysis of a cluster randomized binomial non-inferiority two-sample trial. The determination of the intra-cluster correlation coefficient (ICC) and its use in the calculation of the 97.5% upper confidence bound for delta, the true difference in binomial proportions between the active control and the experimental treatment groups, will be outlined.
McCarthy, William F., "The Calculation of the 97.5% Upper Confidence Bound: Application to Clustered Binary Data in a Binomial Non-Inferiority Two-Sample Trial." (July 2008). COBRA Preprint Series. Working Paper 37.