Shrinkage Estimators for Consistency Assessment in Multi-Regional Clinical Trials
View Presentation View Presentation
*Hui Quan, Sanofi 

Keywords: sample size configuration, random effect model, between-region variability, weighted estimator.

Multi-regional clinical trials (MRCTs) have been widely used for efficient global new drug developments. Both a fixed effect model and a random effect model can be used for trial design and data analysis of a MRCT. Correspondingly, the James-Stein type shrinkage estimator and empirical shrinkage estimator can be utilized to estimate treatment effects of individual regions and be applied to consistency assessment of treatment effects across regions. In this presentation, we will compare these estimators to the regular sample estimator and provide the interpretation of these estimators in a MRCT setting. As demonstrated in our computation and simulation, the shrinkage estimators borrowing information from the other regions have smaller variability and associate with much higher probability for demonstrating consistency. A multinational trial example is used to illustrate the application of the methods.