JOINT MODELING OF MULTIVARIATE LONGITUDINAL DATA AND COMPETING RISKS DATA

In many clinical studies that involve follow-up, it is common to observe one or more sequences of longitudinal measurements, as well as one or more time to event outcomes. A competing risks situation arises when the probability of occurrence of one event isaltered/hindered by another time to event. A classical example is different causes of death. When the missing data mechanism in the longitudinal process is non-ignorable due to informative dropout, one has to jointly model the longitudinal data and the timeto event outcomes in order to obtain valid inferences. Recently, there has been much attention paid to the joint analysis of single longitudinal measurements and single time to event data. A natural extension in the joint modeling literature is to the case whichinvolves a joint analysis of multiple failure types and longitudinal process(es). However, to the best of our knowledge only a few such extensions currently exist in the literature.We, in this thesis, propose one such extension where multiple longitudinal responses are jointly modeled with competing-risks time to event data (multiple failure types). Our joint model consists of two sub-models: a system of non-linear mixed effects sub-models for the multiple longitudinal responses; and a system of cause-specific hazards frailty submodels for competing risks, with associations among multiple longitudinal responses andmultiple failure types (competing risks) are modeled using latent parameters. That is, we use the shared parameter modeling approach in this joint model. The joint model is applied to a data set of patients with end-stage heart failure awaiting heart transplant.Heart failure occurs when heart loses the ability to pump enough blood through the body. Currently, most of the patients with end-stage heart failure who are eligible for heart transplant are inserted with a mechanical circulatory support (MCS) devices (suchas left ventricular assists devices) to support the failing heart while waiting for heart transplant. While on the MCS, their liver and renal functions may worsen and these in turn may influence one of the two possible outcomes: i. death before transplant; ii.transplant. These two events can be considered as competing risks. We use longitudinal measurements of bilirubin as a marker for liver function and longitudinal measurements of GFR as a marker for renal function. By using our joint model we assess the effect(association) of liver and renal function on the competing risks.

Name thesis:degree_name

Doctor of Philosophy

Level thesis:degree_level

doctoral

Discipline thesis:degree_discipline

Epidemiology and Biostatistics

Grantor dc:publisher

Case Western Reserve University School of Graduate Studies

Year dc:date

2013

Author dc:creator

• Rajeswaran, Jeevanantham

Contributors dc:contributor

• Schluchter, Mark

• Biostatistics
• Joint Modeling
• multivariate longitudinal data
• competing risks data
• non-linear mixed effect model
• multi-phase models

Statement dc:rights

• unrestricted
• This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws.

Language dc:language

English

Repository record dc:identifier

http://rave.ohiolink.edu/etdc/view?acc_num=case1354508776

OAI identifier oai:identifier

oai:etd.ohiolink.edu:case1354508776