MORE EFFICIENT ESTIMATORS FOR CASE-COHORT STUDIES WITH UNIVARIATE AND MULTIVARIATE FAILURE TIMESPublic Deposited
Add to collection
You do not have access to any existing collections. You may create a new collection.
Downloadable ContentDownload PDF
MLAKim, Soyoung. More Efficient Estimators For Case-cohort Studies With Univariate And Multivariate Failure Times. University of North Carolina at Chapel Hill, 2013. https://doi.org/10.17615/ftev-1r08
APAKim, S. (2013). MORE EFFICIENT ESTIMATORS FOR CASE-COHORT STUDIES WITH UNIVARIATE AND MULTIVARIATE FAILURE TIMES. University of North Carolina at Chapel Hill. https://doi.org/10.17615/ftev-1r08
ChicagoKim, Soyoung. 2013. More Efficient Estimators For Case-Cohort Studies With Univariate And Multivariate Failure Times. University of North Carolina at Chapel Hill. https://doi.org/10.17615/ftev-1r08
- Last Modified
- March 19, 2019
- Affiliation: Gillings School of Global Public Health, Department of Biostatistics
- Case-cohort study design is generally used to reduce cost in large cohort studies when the disease rate is low. The case-cohort design consists of a random sample of the entire cohort, named subcohort, and all the subjects with the disease of interest. When the rate of disease is not low or the number of cases are not small, the generalized case-cohort study which selects subset of all cases is used. In this dissertation, we study more efficient estimators of multiplicative hazards models and additive hazards models for the traditional case-cohort study as well as the generalized case-cohort study. We first study more efficient estimators for the traditional case-cohort studies with rare diseases. When several diseases are of interest, several case-cohort studies are usually conducted using the same subcohort. When these case-cohort data are analyzed, the common practice is to analyze each disease separately ignoring data collected in subjects with the other diseases. This is not an efficient use of the data. In this study, we propose more efficient estimators by using all available information. We consider both joint analysis of the multiple diseases and separate analysis for each disease. We propose an estimating equation approach with a new weight function. We establish that the proposed estimator is consistent and asymptotically normally distributed. Simulation studies show that the proposed methods using all available information gain efficiency. For comparing the effect of the exposure on different diseases, tests based on the joint analysis are more powerful than those based on the separate analysis assuming independence. We apply our proposed method to the data from the Busselton Health Study. We extend this approach to the stratified case-cohort design with non-rare diseases. We also consider the additive hazards regression model for the stratified case-cohort studies. Additive hazards model is more appropriate when risk difference is of interest. Risk difference is more relevant to public health because it translates directly into the number of disease cases that would be avoided by eliminating a particular exposure. We propose an estimating equation approach for parameter estimation in additive hazards regression model by making full use of available information. Asymptotic properties of the proposed estimators were developed and simulation studies were conducted. We apply our proposed methods to data from the Atherosclerosis Risk in Communities (ARIC) study.
- Date of publication
- Resource type
- Rights statement
- In Copyright
- Cai, Jianwen
- Doctor of Philosophy
- Graduation year
This work has no parents.