Abstract:
|
For clinical trials with multiple binary co-primary endpoints, we present an efficient method in sample size calculation. The method involves simulating sufficient test statistic vectors and can be easily implemented by a statistical practitioner. Its outputs on power and sample size are accurate in comparison with outcomes from numerical integration method or from simulating correlated binary responses of individual subjects. When number of endpoints increase, the complexity of the method remains the same. We also illustrate the method for sample size calculation in clinical trials with correlated multiple binary endpoints using Holm and Hochberg procedures for multiplicity adjustment.
|