A Multi-Armed Bayesian Ordinal Outcome Utility-Based Sequential Trial with a Pairwise Null Clustering Prior

A multi-armed trial based on ordinal outcomes is proposed that leverages a flexible non-proportional odds cumulative logit model and numerical utility scores for each outcome to determine treatment optimality. This design uses Bayesian clustering prior the effects encourages pairwise null hypothesis of no differences between treatments. group sequential which treatments are clinically different with an adaptive decision boundary becomes more aggressive as sample size or clinical significance grows, number active decreases. simulation study conducted 3 5 arms, shows has superior operating characteristics (family wise error rate, generalized power, average size) compared designs do not allow clustering, frequentist proportional model, permutation test empirical mean utilities.