A0673
Title: Mean residual life based illness-death model for semicompeting risks data
Authors: Liming Xiang - Nanyang Technological University (Singapore) [presenting]
Abstract: Semicompeting risks data are available in many biomedical studies, where some nonterminal event (e.g., disease progression) is of interest and subject to censoring by the terminal event (e.g., death). The illness-death model is commonly used for the analysis of such data. It is proposed to formulate the effects of covariates on the endpoints through semiparametric mean residual life regression models with shared frailty under the illness-death model framework. Novel estimating equations are developed based on a penalized quasi-likelihood incorporating the inverse probability of censoring weights to adjust for possible dependent censoring. Unlike the usual illness-death model assuming a gamma frailty, the proposed inference procedure requires no distributional assumption for frailty. Under some regularity conditions, it is shown that the resulting parameter estimators are consistent and asymptotic normal. Simulation results demonstrate that the method performs well in various realistic settings. The usefulness of the method is further illustrated via the analysis of a real data example.
