Multivariate failure time data are commonly encountered in biomedical research since each study subject may experience multiple events or because there exists clustering of subjects such that failure times within the same cluster are correlated. There are numerous statistical methods reported for the analysis of right-censored multivariate failure time data and among these. In this paper we use the frailty approach to catch the related survival variables and assume each event is a discrete analogue as an interval of clinical examinations periodically. For estimation, an Expectation Maximization (EM) algorithm is developed and is applied to the diabetic retinopathy study (DRS).
Communications in Statistics - Simulation and Computation 43(7), pp.1825-1835