Key points are not available for this paper at this time.
In some longitudinal drug studies, regulatory agencies suggest baseline observation carry forward (BOCF) as a method of handling patient dropout, despite the existence of many criticisms to BOCF. The reason for using BOCF is not clear to many users who either treat BOCF as an imputation method or consider BOCF to be "conservative" in the sense that it allows treatment effect to be evaluated with a severe penalty for dropouts. In this article we address the following questions and issues: What is the reason for using BOCF? Is BOCF a conservative approach to assessing drug efficacy? Is BOCF reasonable? If not, what are the alternatives? Our discussions are based on both theoretical and practical viewpoints.
Shao et al. (Thu,) studied this question.