Copula Based Cox Proportional Hazards Models for Dependent Censoring

Most existing copula models for dependent censoring in the literature assume that parameter defining is known. However, prior knowledge on this dependence often unavailable. In article we propose a novel model under which does not need to be The based parametric relation between survival time (T) and (C), whereas marginal distributions of T C follow semiparametric Cox proportional hazards model, respectively. We show identified, estimators nonparametric cumulative hazard finite-dimensional parameters. It shown parameters function are consistent asymptotically normal. also investigate performance proposed method using finite-sample simulations. Finally, apply our estimation procedure follicular cell lymphoma dataset. Supplementary materials available online.