A general joint model for longitudinal measurements and competing risks survival data with heterogeneous random effects

• DOI: 10.1007/s10985-010-9169-6
• Publication Date: 2010-06-11T09:36:49Z
• Authors: Xin Huang; Gang Li; Robert M. Elashoff; Jianxin Pan
• Keywords: Censoring (clinical trials); Mixed model

This article studies a general joint model for longitudinal measurements and competing risks survival data. The consists of linear mixed effects sub-model the outcome, proportional cause-specific hazards frailty data, regression variance–covariance matrix multivariate latent random based on modified Cholesky decomposition. provides useful approach to adjust non-ignorable missing data due dropout enables analysis outcome with informative censoring intermittently measured time-dependent covariates, as well outcomes. Unlike previously studied models, our allows heterogeneous covariance matrices. It also offers framework assess homogeneous assumption existing models. A Bayesian MCMC procedure is developed parameter estimation inference. Its performances frequentist properties are investigated using simulations. real example used illustrate usefulness approach.