Air pollution and morbidity: a sensitivity analysis of alternative monitors, pollutants, and averaging times

Large cross-sectional data bases containing observations at the individual level are well suited for exploring the functional relationship between ambient air quality and health outcome. Since these analyses require the use of fixed-site air pollution monitors to measure pollution exposure, the results are subject to continued uncertainty. This study addresses three major concerns related to the use of these monitors. The results indicate that: (1) particulate matter alone is an important air pollution source of morbidity. It does not appear to mask or incorporate the effect of other air pollutants; (2) the estimated impact of air pollution on health is insensitive to actual monitor location. However, combining central city and suburban residents can lead to an underestimate of the pollution impact; and (3) the results are generally consistent when either a short or long term measure of pollution exposure is used.