Exposure Prediction Error and Spatial Confounding in Air Pollution Epidemiology

Joshua Keller, PhD, Postdoctoral Fellow

Department of Biostatistics

Johns Hopkins Bloomberg School of Public Health 

To conduct inference about air pollution exposures in large administrative cohorts such as Medicaid, several challenges must be addressed. Spatial misalignment between pollution monitors and area-level health data require spatial prediction of exposure. In this talk, I present a comparison of approaches for estimating area-level averages of air pollution, when the exposure prediction model is known to be mis-specified. Additionally, the limited amount of individual-level data can lead to unmeasured spatial confounding. I present a method for linking flexible spatial confounding adjustment to spatial scales. These methods are applied to an analysis of long-term particulate matter exposure and asthma morbidity among children in Medicaid. Joshua Keller is a postdoctoral fellow in the Department of Biostatistics at the Johns Hopkins Bloomberg School of Public Health. He received his Ph.D. in Biostatistics from the University of Washington in 2016. His research interests include spatiotemporal environmental exposures, measurement error and uncertainty quantification, and environmental trials.