Improving on Semiparametric Approach for Joint Modeling of Survival and Longitudinal Data
by Wen-Han Hwang
Abstract: Longitudinal covariates in survival models are generally analyzed using random effects models. By framing the estimation of these models as a functional measurement error problem, semiparametric approaches such as the conditional score or the corrected score can be used to establish consistent estimators for survival model parameters. These approaches require no distributional assumptions on the random effects and, in contrast with standard methods in the literature which only use covariate data in the risk set before each event time, can incorporate the entirety of longitudinal data available. Applying this approach, we develop an error augmentation technique that uses all available information to estimate the longitudinal covariate process. The consistency and efficiency improvement of the proposed semiparametric approach is confirmed theoretically and illustrated in a simulation study.
For More Information: contact: Mihee Lee. email: email@example.com