The results of the simulation studies show that the proposed method is feasible in handling high-dimensional covariates. It can quantify variability associated with treatment selection and serve as a reliable and uniform inferential tool in clinical follow-up studies
Danhui Yi (Tue,) studied this question.