Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Inference on Censored Survival Data Under Competing Risks

Lim, Jeong Youn (2012) Inference on Censored Survival Data Under Competing Risks. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

[img]
Preview
PDF
Primary Text

Download (622kB) | Preview

Abstract

Competing risks is commonly encountered in survival data. While fundamental methods have been established to analyze survival data in the presence of competing risks, some of
methods still remain undeveloped. The primary goal of this study is to extend existing methods for survival analysis to the competing risks settings.
In the first study is to determine the optimal cutpoint in the presence of competing risks. A continuous variable often needs to be dichotomized to quantify the prognostic effect. The "outcome-oriented" cutpoint approach is the useful method without any prior knowledge about that variable, which is to seek an optimal cut point that provides the maximum difference in prognostic effect between the splits. The rescaled sequential method is one of
the approaches for estimating the optimal cutpoint and for adjusting its significance after the dichotomization. We adapted the concept of improper random variables from Gray's
test and modified log-rank test to apply the rescaled sequential approaches. We present simulation results of the operating characteristics of the proposed method. A real dataset from National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14 is exemplified.
In the second part, a quantile residual life regression model was developed for competing risks. Residual life analysis provides useful information when the effect of prognostic factors on the distribution of remaining lifetimes is evaluated at several years after the initial diagnosis/therapy. This model allows for meaningful interpretations of covariate effects on not only any quantile residual life but also at a specific time point. Simulation studies are performed to assess the finite sample properties of proposed method in terms of the parameter estimator, type I error and power of the test statistics at different time points. The new regression method is illustrated with a NSABP B-04 dataset.
Although competing risks have been an important issue in survival analysis research, it is often neglected by clinical researchers due to its complex nature and lack of available methodology. Development of inference procedures suitable for competing risks data would provide more accurate additional information, which has great significance in a public health perspective leading to improved patient care in clinical settings.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Lim, Jeong Younjel66@pitt.eduJEL66
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Thesis AdvisorJeong, Jong-Hyeonjeong@nsabp.pitt.eduJJEONG
Committee MemberAnderson, Stewartsja@pitt.eduSJA
Committee MemberKingsley, Lawrencekingsley@pitt.eduKINGSLEY
Committee MemberCheng, Yuyucheng@pitt.eduYUCHENG
Date: 29 June 2012
Date Type: Completion
Defense Date: 22 November 2011
Approval Date: 29 June 2012
Submission Date: 27 March 2012
Access Restriction: 1 year -- Restrict access to University of Pittsburgh for a period of 1 year.
Number of Pages: 65
Institution: University of Pittsburgh
Schools and Programs: Graduate School of Public Health > Biostatistics
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: Competing risks, Survival analysis, Residual life analysis
Date Deposited: 29 Jun 2012 21:35
Last Modified: 15 Nov 2016 13:56
URI: http://d-scholarship.pitt.edu/id/eprint/11499

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item