2020/10/26 by Ethan M. Alt, Alt, Ethan M., Matthew A. Psioda +3
Mathematics · Economics, Econometrics and Finance · #Statistical Methods in Clinical Trials #Advanced Causal Inference Techniques #Health Systems, Economic Evaluations, Quality of Life
paper · pdf · doi:10.48550/arxiv.2010.13774
Given the cost and duration of phase III and phase IV clinical trials, the\ndevelopment of statistical methods for go/no-go decisions is vital. In this\npaper, we introduce a Bayesian methodology to compute the probability of\nsuccess based on the current data of a treatment regimen for the multivariate\nlinear model. Our approach utilizes a Bayesian seemingly unrelated regression\nmodel, which allows for multiple endpoints to be modeled jointly even if the\ncovariates between the endpoints are different. Correlations between endpoints\nare explicitly modeled. This Bayesian joint modeling approach unifies single\nand multiple testing procedures under a single framework. We develop an\napproach to multiple testing that asymptotically guarantees strict family-wise\nerror rate control, and is more powerful than frequentist approaches to\nmultiplicity. The method effectively yields those of Ibrahim et al. and\nChuang-Stein as special cases, and, to our knowledge, is the only method that\nallows for robust sample size determination for multiple endpoints and/or\nhypotheses and the only method that provides strict family-wise type I error\ncontrol in the presence of multiplicity.\n