Prospective comparison of Bayesian and frequentist adaptive clinical trials : The SHADOW - SHINE project