ISEE/ISEA 2006 Conference Abstracts Supplement: Symposium Abstracts: Abstracts
Symposium Title: Regional, Multicity Time-Series Studies of Air Pollution and Health: Progress and Prospects
Symposium Organizers: A. Cohen* and H. R. Anderson† *Health Effects Institute and † St. Georges Hospital Medical School/U London
Hundreds of time-series studies of the health effects of short-term exposure to air pollution have now been conducted worldwide. Collectively, these studies provide compelling evidence of the adverse effects of short-term exposure, but also pose problems of interpretation due to variation in analytic methods and reporting and the possibility of publication and analytic bias. Heterogeneity in effect estimates may also be due to spatial variations arising from differences in meteorologic and other environmental conditions, healthcare systems, and other spatially related factors. A number of recent meta-analyses of published results have summarized this literature, but few have focused on global patterns in the relative rates of mortality and morbidity and evidence as to the causes of their spatial variation. The Air Pollution Epidemiology Database (APED) from St. George's, University of London, is a comprehensive database of epidemiologic studies of the effects of short-term exposure to air pollution published in the peer-reviewed literature. It contains effect estimates from 374 time-series studies published worldwide (141 Europe, 148 North America, 32 Latin America, 53 Asia and the Western Pacific) and descriptive data concerning the locales in which these studies have been conducted. Data from APED will be used to investigate some of the sources of heterogeneity in the effect estimates, including spatially varying factors, analytic methods used, and so on. These data will also be used to illustrate some of the issues arising from such meta-analyses of the published literature, including publication bias and other study-specific limitations. Finally, the advantages and disadvantages of individual studies versus multicity studies are discussed.