Biostatistics Weekly Seminar


Improved Doubly Robust Estimation of Cox Regression Parameters Under Left Truncation

Sharon Xiangwen Xie, PhD
Professor of Biostatistics
Department of Biostatistics Epidemiology and informatics
University of Pennsylvania

Left truncation is a phenomenon that occurs in time-to-event studies and results in biased outcome and covariate distributions. Additionally, survival studies often suffer from missing covariate data, which might arise due to cost constraints or by study design. Augmented inverse probability weighting (AIPW) is a popular missing data approach that involves estimation of both the probability of missingness and the population distribution of the missing covariate. Estimation of the latter may be inaccurate with a left-truncated sample. We propose a novel AIPW estimation procedure that accounts for selection bias when modeling the missing covariate distribution. Using simulation studies, we explore the performance of our approach and existing methods in estimating Cox regression parameters under a variety of truncation and missing data scenarios. By improving estimation of the missing covariate distribution, our method is more robust to model misspecification than the existing approaches considered. We also develop asymptotic theory by proving that the proposed regression parameter estimates are consistent. Finally, we implement our approach to assess the association between cerebrospinal fluid amyloid beta and the risk of cognitive impairment among Parkinson’s disease patients.


Virtual: Zoom Link to Follow
26 April 2023
1:30pm


Speaker Itinerary

Edit | Attach | Print version | History: r2 < r1 | Backlinks | View wiki text | Edit WikiText | More topic actions...
Topic revision: r1 - 12 Apr 2023, CierraStreeter
 

This site is powered by FoswikiCopyright © 2013-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback