Multiple imputation of censored bivariate event-times via inverse transform and nonparametric Gibbs sampling

Bivariate time-to-event data, subject to right censoring, frequently arise in medical research. This paper introduces a novel nonparametric multiple imputation (MI) procedure for analyzing censored bivariate time-to-event data. Our methodology offers a straightforward, easy-to-implement inverse transform MI method that effectively captures the joint distribution of bivariate random variables through the imputation of censored event-times.