Premium
Optimal designs for regional bridging studies using the Bayesian power prior method
Author(s) -
Nagase Mario,
Ueda Shinya,
Higashimori Mitsuo,
Ichikawa Katsuomi,
Dunyak James,
AlHuniti Nidal
Publication year - 2019
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.1967
Subject(s) - bridging (networking) , bayesian probability , computer science , clinical trial , econometrics , actuarial science , data science , medicine , business , economics , artificial intelligence , computer network , pathology
As described in the ICH E5 guidelines, a bridging study is an additional study executed in a new geographical region or subpopulation to link or “build a bridge” from global clinical trial outcomes to the new region. The regulatory and scientific goals of a bridging study is to evaluate potential subpopulation differences while minimizing duplication of studies and meeting unmet medical needs expeditiously. Use of historical data (borrowing) from global studies is an attractive approach to meet these conflicting goals. Here, we propose a practical and relevant approach to guide the optimal borrowing rate (percent of subjects in earlier studies) and the number of subjects in the new regional bridging study. We address the limitations in global/regional exchangeability through use of a Bayesian power prior method and then optimize bridging study design with a return on investment viewpoint. The method is demonstrated using clinical data from global and Japanese trials in dapagliflozin for type 2 diabetes.