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

Summary Functions for Data in the Presence of Competing Risks

Zhao, Yongyun (2008) Summary Functions for Data in the Presence of Competing Risks. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

[img]
Preview
PDF
Primary Text

Download (2MB) | Preview

Abstract

Although cumulative incidence function (CIF) estimates are commonly used to describe the failure probabilities when competing risks are present, the CIF has limitations in some scenarios. The objective of our research was to propose new summary functions or modify CIF to overcome the limitations. In observational studies or nonrandomized trials, CIF estimates can be biased if the distribution of a confounding variable differs among treatment groups. To reduce the bias, we developed an adjusted CIF (ACIF) estimator that is based on the use of inverse probability weighting. We derived the estimation and inference procedures, and then used simulation studies to evaluate the performance. To illustrate the application of ACIF, we used the example of liver transplant candidates with various types of end-stage liver disease. We also developed a series of adjusted survival functions to estimate the ``net¡± survival probability for a specific outcome (the main event), based on the degree of correlation between this event and the competing events. First, for cases in which there is a perfect negative correlation, we constructed an adjusted survival function that uses the Kaplan-Meier estimator and the inverse probability of censoring weight (IPCW) for patients who experience competing events. Second, for cases in which there are imperfect negative correlations, we constructed an adjusted survival function that uses the combination of the IPCW and the adjusted number at risk under the constraints of the lower and upper bounds. Third, for cases in which there are positive correlations, we constructed an adjusted survival function that uses the adjusted number of main events under the constraints of the lower and upper bounds. To recover the contribution that the competing events make to the net survival probability, we incorporated auxiliary variables into the adjusted number at risk or the adjusted number of main events. We derived the estimation and inference procedures. We demonstrate the use of adjusted survival functions in data derived from patients who had emphysema, a severe form of chronic obstructive pulmonary disease.The public significance of this work is to provide new approaches to summarize the survival data with competing risks.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Zhao, Yongyunyongyunzhao2001@yahoo.com
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairChang, (Joyce) Chung-Chou Hchangjh@upmc.edu
Committee CoChairWeissfeld, Lisa Alweis@pitt.eduLWEIS
Committee MemberRockette, Howard Eherbst@pitt.eduHERBST
Committee MemberJeong, Jong-HyeonJeong@nsabp.pitt.eduJJEONG
Committee MemberRoberts, Mark Srobertsm@upmc.eduMROBERTS
Date: 28 September 2008
Date Type: Completion
Defense Date: 31 July 2008
Approval Date: 28 September 2008
Submission Date: 24 July 2008
Access Restriction: 5 year -- Restrict access to University of Pittsburgh for a period of 5 years.
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: adjusted cumulative incidence function; competing risks; cumulative incidence function; inverse probability weighting; inverse probability of censoring
Other ID: http://etd.library.pitt.edu/ETD/available/etd-07242008-232833/, etd-07242008-232833
Date Deposited: 10 Nov 2011 19:53
Last Modified: 19 Dec 2016 14:36
URI: http://d-scholarship.pitt.edu/id/eprint/8571

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item