Calculating Bayesian sample sizes for multi-arm trials where several experimental treatments are compared to a common control, perhaps even at multiple stages.
