Joint Modeling of Multivariate Longitudinal Data and Competing Risks Data

• Activity Number: 531
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Jeevanantham Rajeswaran*+
• Companies: Cleveland Clinic
• Keywords: Multiple longitudinal data ; Multiple failure types ; Joint Modeling ; Competing risks data

Abstract

In clinical studies, it is common to observe one or more longitudinal processes (LP), as well as one or more time to event outcomes. When the missing data mechanism in the LP is non-ignorable, one has to jointly model the longitudinal and time to event data to obtain valid inferences. A natural approach in the literature is the joint modeling of failure types and LP. We propose an extension where multiple LP is jointly modeled with competing-risks data (multiple failure types). Our joint model consists of two sub-models: a system of non-linear mixed effects models for the multiple LP and a system of cause-specific hazards frailty models for competing risks, with associations among multiple LP and competing- risks are modeled using latent parameters in a shared parameter approach. The model is applied to a data set of patients on a mechanical circulatory support (MCS) device awaiting heart transplant. While on the MCS, their liver and renal functions may worsen and these in turn may influence one of the two possible competing events; death before transplant or transplant. Using the joint model we assess the effect of liver and renal functions on the competing risks.