Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Conroy Thesis Final.pdf (342.02 KB)

ETD Abstract Container

Abstract Header

A simulation study of bivariate Wiener process models for an observable marker and latent health status

Conroy, Sara A
http://orcid.org/0000-0001-9155-9456
http://rave.ohiolink.edu/etdc/view?acc_num=osu1452015350

Abstract Details

2016, Master of Science, Ohio State University, Public Health.
Threshold regression is a relatively new model for time to event data in which the health status of a subject is modeled using an unobservable stochastic process which fails once it reaches a threshold. In addition to being a conceptually appealing model, threshold regression does not require the proportional hazards assumption of the Cox model. In clinical trials, biomarkers of disease are commonly measured along with time to an event (e.g., disease onset or death). A few authors have proposed bivariate Wiener process models for an observable marker and the latent health status [13, 31]. We conducted a simulation study to evaluate the benefits of joint modeling in the context of a clinical trial. Specifically, we looked at the effects of joint modeling on the test of an effect of treatment on time to event. Simulations were run using different sample sizes, different correlation levels between the health status and marker processes, and different censoring rates. We found that the improvement in power attributable to joint modeling increases as correlation, sample size, and censoring rate increases. The greatest gain in power was with 75% censoring and a correlation of 0.9. When the censoring rate was below 75%, or ρ<0.75 there was little difference in power between the joint model and a univariate model.
Michael L. Pennell, Ph.D. (Advisor)
Rebecca R. Andridge, Ph.D. (Committee Member)
68 p.
Biostatistics
threshold regression; joint modeling; bivariate Wiener process

Recommended Citations

Citations

  • Conroy, S. A. (2016). A simulation study of bivariate Wiener process models for an observable marker and latent health status [Master's thesis, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1452015350

    APA Style (7th edition)

  • Conroy, Sara. A simulation study of bivariate Wiener process models for an observable marker and latent health status. 2016. Ohio State University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1452015350.

    MLA Style (8th edition)

  • Conroy, Sara. "A simulation study of bivariate Wiener process models for an observable marker and latent health status." Master's thesis, Ohio State University, 2016. http://rave.ohiolink.edu/etdc/view?acc_num=osu1452015350

    Chicago Manual of Style (17th edition)

osu1452015350
583
© 2016, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.