NSTAGEBINOPT: Stata module to compute admissible multi-arm multi-stage trial designs with binary outcomes

nstagebinopt searches for multi-arm multi-stage designs with binary intermediate (I) and definitive (D) outcomes which have the desired overall type I error rate and power (feasible designs). The stagewise operating characteristics of feasible designs which minimise the Bayesian loss function (1-q)E(N|H0)+qE(N|HK) for some q between 0 and 1 are then reported. Here E(N|H0) is the expected sample size under the hypothesis that all K experimental arms in the study are ineffective and E(N|HK) is the expected sample size assuming all K experimental arms are effective.