Estimating Associations Between Source-Apportioned Particulate Matter and Emergency Department Visits in Multicity Studies

• Activity Number: 603
• Type: Contributed
• Date/Time: Wednesday, August 12, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Jenna Krall* and James A. Mulholland and Armistead G. Russell and Sivaraman Balachandran and Andrea Winquist and Paige E. Tolbert and Lance Waller and Stefanie Ebelt Sarnat
• Companies: and Georgia Institute of Technology and Georgia Institute of Technology and University of Cincinnati and Emory University and Emory University and Emory University and Emory University
• Keywords: Air pollution ; Source apportionment ; Environmental epidemiology

Abstract

Particulate matter (PM) air pollution is a heterogeneous mixture of individual pollutants generated by natural and anthropogenic sources. The chemical composition of PM varies by its generating sources and associations between PM and health may also vary by source. While total PM concentrations in the air can be directly measured, we generally do not observe PM concentrations from specific sources. Source apportionment models (often based on latent variable models) can be used to estimate PM sources. Source apportionment is generally applied to data from a single monitoring site and no standard methods exist for pooling data from multiple cities. Combining source-apportioned PM from multiple cities is challenging because PM from specific sources often varies spatially in chemical composition. Development of approaches for using multicity source apportionment outputs in epidemiologic analyses is critical to facilitate comparison of health effect estimates across regions and to increase precision of estimated associations. We compared methods for combining source-apportioned PM and used these methods to estimate associations with emergency department visits in four US cities.