Online Program Home
My Program

Abstract Details

Activity Number: 179 - Statistical Methods for Measurement Error and Missing Data in Covariates/Exposures
Type: Contributed
Date/Time: Monday, July 29, 2019 : 10:30 AM to 12:20 PM
Sponsor: Section on Statistics in Epidemiology
Abstract #305291 Presentation
Title: A Semiparametric Approach to Analyzing Error-Prone Failure Time Outcomes and Exposures
Author(s): Lillian Boe* and Pamela Shaw
Companies: University of Pennsylvania and University of Pennsylvania
Keywords: EM algorithm; measurement error; misclassification; proportional hazards; regression calibration; survival analysis

In clinical research, measurement error arises commonly in settings that rely on data from electronic health records or large observational cohorts. In particular, self-reported outcomes are typical in cohort studies for chronic diseases such as diabetes in order to avoid the burden of costly diagnostic tests. Dietary intake, which is also often subject to measurement error, is a major factor linked to diabetes and other chronic diseases. These errors can bias exposure-disease associations and impact clinical decisions. We have extended an existing semiparametric likelihood-based method for handling error-prone, discrete failure time outcomes to also address covariate measurement error. We conduct a numerical study to evaluate the proposed method in terms of bias and efficiency in the estimation of the regression parameter of interest. This method is applied to data from the Women’s Health Initiative, which has information available on the error structure of the outcome and exposure of interest. We compare analyses of the effect of dietary energy and protein intake on the risk of incident diabetes mellitus using our proposed method and naïve analyses that ignore measurement error.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program