Time series analysis of air pollution and health accounting for covariate-dependent overdispersion Pubblico

Abstract

The time series study design is routinely used to estimate short-term associations between various adverse health outcomes and exposures to ambient air pollutants. This is accomplished by analyzing daily air pollution concentrations and aggregated counts of adverse health events over a geographical region via a Poisson log-linear model under the assumption of constant overdispersion. In this paper, we develop covariate-dependent Bayesian generalized Poisson and negative binomial models to account for potential time-varying overdispersion. The proposed models are applied to a time series study of daily emergency department visits for respiratory diseases and ozone concentration in Atlanta, Georgia during the period 1999 to 2009. Allowing for covariate-dependent overdispersion results in a reduction in ozone effect standard error, while the ozone-associated relative risk remains robust to different model specifications. Through simulation studies, we similarly found that the standard quasi-Poisson approach can result in larger standard error for the air pollution effect estimate when the constant overdispersion assumption is violated. Our findings suggest that improved characterization of overdispersion may result in more accurate and precise health effect estimates in studies of short-term environmental exposures.

Table of Contents

Table of Contents

1. Introduction…………………………………………………………………………….......................1

2. Atlanta Emergency Department Visit and Air Quality Data…………………………...2

3. Methods…………………………………………………………………………………........................4

3.1. Time Series Analysis of Air Pollution and Health……………………………......……..4

3.2. Time Series Analysis with Covariate-dependent Overdispersion…………………5

3.3. Estimation…………………………………………………………………………….......................6

4. Simulation Studies…………………………………………………………………….....................6

5. Application to Atlanta Emergency Department Visit and Ozone……….……………8

6. Discussion……………………………………….......................................................10

References……………………………………………..................................................... 12

Appendix: Tables and Figures………………………………………………………..............…...14

About this Master's Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Rollins School of Public Health
Department	Biostatistics and Bioinformatics
Subfield / Discipline	Biostatistics - MPH & MSPH
Degree	MSPH
Submission	Master's Thesis
Language	English
Research Field	Biology, Biostatistics Environmental Health
Parola chiave	negative binomial generalized poisson air pollution overdispersion health effect quasi-poisson
Committee Chair / Thesis Advisor	Chang, Howard, Emory University
Committee Members	Sarnat, Stefanie Ebelt, Emory University
Partnering Agencies	Does not apply (no collaborating organization)

Ultima modifica

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Time series analysis of air pollution and health accounting for covariate-dependent overdispersion ()	2018-08-28 13:41:19 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions