ABSTRACT Reference-based imputation (RBI) is a popular method for missing data. The methodology is well established for continuous end points but less well developed for repeated binary end points due… Click to show full abstract
ABSTRACT Reference-based imputation (RBI) is a popular method for missing data. The methodology is well established for continuous end points but less well developed for repeated binary end points due to the lack of natural multivariate conditional distributions for such end points. In this paper, we propose RBI methods for repeated binary end points based on a multivariate probit model and a logistic model, including jump-to-reference (J2R), copy-reference (CR) and copy-increment-in-reference (CIR). We explore the distribution of the missing binary end points under RBI and propose efficient algorithms to implement the proposed RBI methods. We evaluate the proposed methods by simulations and a data set from a clinical trial.
               
Click one of the above tabs to view related content.