Semiparametric mixed-effects joint models are flexible for modeling complex longitudinal-competing risks data. Skew distributions are commonly observed for this type of data. Covariates in the joint models are usually measured with substantial errors. We propose a Bayesian method for semiparametric mixed-effects joint models with covariate measurement errors and skew distribution. The methods are illustrated with AIDS clinical data. Simulation results are conducted to validate the proposed methods.
Keywords: Bayesian inference; competing risks; longitudinal data; measurement error; partially linear mixed-effects models; proportional hazard models; skew distribution; survival data.