Fig. 3From: Seamless phase 2/3 design for trials with multiple co-primary endpoints using Bayesian predictive powerThe behavior of overall power and total sample size of BPP when the null hypothesis is falseBack to article page