Retrospective outcome dependent sampling (ODS) designs are an efficient class of study designs that may be implemented when resource constraints prohibit the ascertainment of an exposure on an entire cohort. One type of ODS design for longitudinal binary data stratifies individuals into three strata according to a categorization of their response vector: those who did not experience the outcome, those that only experienced the outcome, and those that exhibited response variation (Schildcrout and Heagerty, 2008). For time-varying covariate effects, it has been shown that sampling only those individuals with response variation results in nearly fully efficient estimation compared to the full cohort analysis. If inference lies in a time-invariant covariate effect, or a combined time-varying and time-invariant covariate effect, then the choice of how to allocate resources, or how to define sampling probabilities, is not obvious. We propose a class of two-stage ODS designs for longitudinal binary data. We extend standard (or single-stage) ODS designs to permit two waves of data collection. Fixed two-stage ODS designs utilize pre-specified sampling probabilities, and adaptive two-stage ODS designs use information from stage one to inform our choice of the stage two sampling probabilities. These designs are applied to data from the Lung Health Study where it is of interest to identify genetic determinants of lung function decline among individuals with mild chronic obstructive pulmonary disease.
|Advisor:||Harrell, Frank E.|
|Commitee:||Hardell, Frank E., Lipworth, Loren P., Schildcrout, Jonathan S., Shepherd, Bryan E., Shotwell, Matthew S.|
|School Location:||United States -- Tennessee|
|Source:||DAI-B 80/06(E), Dissertation Abstracts International|
|Keywords:||Efficient study designs, Electronic health records, Longitudinal data, Misclassification, Outcome dependent sampling, Survey data|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be