Online Program Home
My Program

Abstract Details

Activity Number: 30
Type: Contributed
Date/Time: Sunday, July 31, 2016 : 2:00 PM to 3:50 PM
Sponsor: ENAR
Abstract #320210 View Presentation
Title: Accounting for Potential Measurement Errors in Environmental Preterm Studies
Author(s): Yinjun Zhao* and Shuangge Ma
Companies: Yale University and Yale University
Keywords: measurement error ; semiparametric model ; multiple imputation ; environmental preterm

In maternity studies, it is of critical interest to identify demographic and environmental factors that contribute to preterm. In the literature, the standard practice is to conduct the logistic regression of the binary preterm indicator on environmental exposures. Such an analysis fails to accommodate potential measurement errors. First, the pregnancy week is based on recall, which may have a substantial bias, and the bias can be "directed". In addition, the measurements on environmental exposures and imputation procedure (for locations without direct measures) may also be subject to measurement errors. In this study, we use a semiparametric model to directly model pregnancy week (as opposed to the preterm indicator). An iterative estimation procedure is proposed, which uses multiple imputation to accommodate potential measurement variations in both outcome and exposures. The proposed method is applied to data recently collected in Lanzhou, China. The data analysis results are different from that using the standard approach and provide additional insights.

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

Back to the full JSM 2016 program

Copyright © American Statistical Association