Joint Model of Multiple Longitudinal Processes and Survival Outcome

• Activity Number: 350
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Lili Yang*+ and Sujuan Gao
• Companies: Indiana University School of Medicine and IU School of medicine
• Keywords: joint model ; multiple longitudinal outcomes ; time-to-event ; time-dependent covariates ; EM

Abstract

Joint models of longitudinal and time-to-event data can be used to estimate the association between the change or history of the longitudinal measures over time and survival time. While Bayesian method has been used for the parameter estimation in the joint models involving multiple longitudinal processes, likelihood-based method has so far been only applied to joint models of a single longitudinal measure and a time-to-event outcome. In this paper, we develop a likelihood-based method to joint model multiple longitudinal processes and a time-to-event outcome. We will assess the performance of the proposed method in simulation studies and apply the proposed method to a data set with repeated measures of systolic and diastolic blood pressure, and time to depression.