ABSTRACT
In designing a multiregional clinical trial (MRCT), statistical considerations usually focus on the primary objective of the trial and ignore the regional effect. This may produce surprising regional inconsistency that will defeat the intended efficiency of MRCTs. This chapter discusses how regional effects should be considered during study design and execution to mitigate trial failure rates. First, randomization stratification by region is discussed. The second is regional consistency evaluation. This chapter will review the available approaches and propose a new way for overall consistency evaluation. Finally, in designing MRCTs with adaptive features, it is important to take into consideration the proportion of trial participants from each region for the entire trial and at interim analyses. It is common in MRCT conduct that study sites at some countries are initiated months later than other countries, due to factors including differences in regulatory and Institutional Review Board approval requirements. If the composition of regions at interim analyses deviates significantly from the planned proportions, temporal heterogeneity could occur so that the results from the interim analyses may not be consistent with the final results. Some recommendations will be given at the end of the chapter.
