Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 133 - Statistical Issues in Environmental Epidemiology and Pharmacoepidemiology
Type: Contributed
Date/Time: Monday, August 9, 2021 : 1:30 PM to 3:20 PM
Sponsor: Section on Statistics in Epidemiology
Abstract #318499
Title: Meta-Analysis of Generalized Additive Models for Estimation of Environmental Exposure-Response Curves in Heterogeneous Populations: Nationwide Health Impacts of PM2.5 in Medicare
Author(s): Jenny Lee* and Rachel Nethery
Companies: Harvard T.H. Chan School of Public Health and Department of Biostatistics, Harvard T.H. Chan School of Public Health
Keywords: Meta Analysis; Generalized Additive Model; Exposure-Response Curve; Heterogeneous Population; Medicare

Recent discussions of environmental regulatory policy have increasingly emphasized scientific transparency and interpretability as necessary criteria for studies to be considered as policy-relevant evidence. At the same time, statistical methods for estimating exposure-response curves (ERCs) to address the challenges in observational data analysis are becoming more complex and often utilizes black-box approaches. When studying the health impacts of fine particulate matter at a national level, one of the key interests of policy-makers may be how different regions contribute to ERCs as there may exist regional differences in the health impacts of PM2.5 due to different sources of fine particles and heterogeneity in the regional populations. In this paper, we propose Meta-GAM, a recently proposed method for estimating ERCs in heterogeneous data setting using a meta-analysis of generalized additive models, can be adapted to address these issues and provide interpretable, environmental policy-relevant ERCs. We apply Meta-GAM to estimate meta-analyzed region-level and national-level ERCs for PM2.5 in the Medicare population and compare results to other causal inference methods.

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

Back to the full JSM 2021 program