A survival model with dependent competing events and right-hand censoring: Probation and parole as an illustration

Researchers are frequently interested in estimating the elapsed time until the occurrence of an event, such as an arrest following release from prison, dissolution of a marriage, and death following the onset of a disease. Estimation is often hindered by the occurrence of a competing event, which prevents the event of interest from being observed, and by limited follow-up periods. A parametric procedure developed in this paper is suitable for dealing with nonrepeatable competing events when the events are correlated and when independent righthand censoring also exists. Weak distributional assumptions are accommodated by using a Box-Cox transformation and by explicit modeling of heteroscedasticity. The procedure is applied to data pertaining to failure on probabation and parole.