Author

Qi Jiang

Publication Date

2015

Document Type

Dissertation/Thesis

First Advisor

Basu, Sanjib

Degree Name

Ph.D. (Doctor of Philosophy)

Department

Department of Mathematical Sciences

LCSH

Statistics||Biostatistics||Medicine--Research--Statistical methods||Terminal care--Research--Statistical methods||Terminally ill--Research--Statistical methods||Biometry

Abstract

Advancement in medical sciences has produced therapies with curative potential for some diseases regarded as terminal in the past. Yet, patients may simultaneously be subject to multiple diseases, not all of which can be cured. When there is only disease in the context and that disease has a cure possibility, a rich class of cure models can be built by taking into consideration the underlying mechanisms leading to cure and relapse. We propose a class of flexible cure models that accounts for uncertainties in such mechanisms. Our proposed model includes many existing cure models as special cases and can be expressed in simpler mathematical forms than existing models of similar objectives. We provide numerical examples to illustrate the consequences of neglecting underlying mechanisms of cure and relapse and show the ability of our proposed model to identify such mechanisms.;On the other hand, competing risks data arise if patients are exposed to more than one disease. Modeling of competing risks data based on the cumulative incidence functions approach has the advantage of providing direct inference on the survival probabilities from each risk. We propose a flexible model that directly models the cumulative incidence functions and automatically accounts for the possibility that one or more of the diseases may be cured. We further develop regression models that can be used to obtain direct inference of the effect of a covariate on the marginal failure probability of each risk. We use a Bayesian approach for the analysis and provide numerical examples to illustrate the performance of the proposed model.

Comments

Advisors: Sanjib Basu.||Committee members: Nader Ebrahimi; Balakrishna Hosmane; Zhuan Ye.

Extent

121 pages

Language

eng

Publisher

Northern Illinois University

Rights Statement

In Copyright

Rights Statement 2

NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.

Media Type

Text

Share

COinS