Evaluation of a sampling strategy for estimation of long-term PM2.5 exposure for epidemiological studies.

As part of the European Community Respiratory Health Survey (ECRHS) PM2.5 (particles collected with an upper 50% cut point of 2.5 microm aerodynamic diameter) was measured using an EPA-WINS (Environmental Protection Agency Well Impactor Ninety-six) sampler. The monitoring schedule was restricted to 7 days per month for one year. Simultaneously, during this one year study period a collocated Harvard Impactor (HI) was run on a daily basis in Erfurt, Germany. Here we validated the reliability of annual, seasonal and monthly means estimated using the ECRHS scheme (measurements taken less than 25% of the whole study period) with the 'true' long-term averages, which were estimated using all available daily means. The daily PM2.5 means, obtained by both instruments operated in parallel, were only slightly different (the mean difference between EPA-WINS and HI was 1.8 microg m(-3) and 2.8 microg m(-3) for the winter means). The values obtained by the two instruments were highly correlated (r = 0.95). In view of that negligible difference, no additional bias was seen with respect to the annual and the winter means estimated by the two different sampling strategies (the difference was 1.7 microg m(-3) and 2.7 microg m(-3), respectively). Monthly means, however, can only be considered to be a crude estimate that may substantially under- or overestimate the true monthly mean value.