Use of the Beta-Binomial Model for Central Statistical Monitoring of Multicenter Clinical Trials