Premium
Phase 3 adaptive trial design options in treatment of complicated urinary tract infection
Author(s) -
Viele Kert,
Mundy Linda M.,
Noble Robert B.,
Li Gang,
Broglio Kristine,
Wetherington Jeffrey D.
Publication year - 2018
Publication title -
pharmaceutical statistics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.421
H-Index - 38
eISSN - 1539-1612
pISSN - 1539-1604
DOI - 10.1002/pst.1892
Subject(s) - bayesian probability , clinical study design , robustness (evolution) , clinical trial , computer science , medicine , statistics , mathematics , artificial intelligence , biology , biochemistry , gene
SUMMARY New antimicrobial drugs for treatment of complicated urinary tract infection (cUTI) are generally assessed in randomized, double‐blind, noninferiority clinical trials. Robust historical data for the active comparator inform on treatment effect estimation, yet typically do not substitute for the active comparator data in the proposed trial. We report design options for a phase 3 trial of cUTI using a Bayesian hierarchical model and historical data from 2 well‐executed phase 3 registrational trials of doripenem. The methodology is directly applicable to other phase 3 noninferiority settings. In addition to the research design application, we provide a novel methodology for assessing the robustness of type I error control. The model borrows heavily from the prior data when the current active comparator parameter estimate approximated the historical estimate. In contrast, the model had restricted borrowing when the 2 estimates were very different. The alternative trial design, with or without the inclusion of futility stopping criteria, provides a framework for future cUTI phase 3 trials.