Abstract:
|
Modelling the negative human health impact of air pollution can be very challenging for a number of reasons. Misalignment may be an issue when the health data are total numbers of outcomes over associated with administrative areal units, and when air pollution data is observed at point locations. Furthermore, failure to account for spatial variation and uncertainty in the pollution exposure process, and the effects of unobserved confounding variables may contribute to poor estimation of the overall impact on the health outcome. We describe a new approach that avoids these issues using a new and rigorous, two-stage approach to modelling the health effects of pollution exposure, using cutting-edge spatio-temporal methodology. An application is presented in which the model is fitted to air pollution and respiratory hospital admission data collected across Local Health Authorities in England, UK.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.