Comparison of two ethnic populations using distribution adjusted mean

摘要

In similarly designed clinical trials conducted in different regions or a multiregional trial, we sometimes observe different mean responses between trials or regions due to differences in distributions of an ethnic factor. Uesaka et al. (2003) used adjusted means for covariate, the distribution adjusted mean, to deal with such a situation. This paper examines distributional properties of the distribution adjusted mean and shows its applicability to comparison of mean responses from clinical trials where the distribution of an ethnic factor is different between regions. A real example and simulation results are shown. The results of this paper enable us to evaluate the similarity of different responses in clinical trials from different ethnic populations.