Multi-city studies

Table 1 reviews examples of three types of studies (1) multi-city studies, such as the Six Cities and ACS studies (2) multi-county studies and (3) studies of a particular location, in this case the Atlanta area (chosen because of the wealth of studies in this area). Health outcomes examined are restricted to all-cause and cardiopulmonary mortality, and cardiovascular disease hospital or emergency department admissions, in order to retain focus. Although additional studies could be included, these studies were chosen because, in concert, they are among the best to illustrate with the least amount of complexity how epidemiology studies with some similarities can reach different conclusions about which pollutants are harmful. In particular, the multi-county and Atlanta studies were chosen because they monitored for many types of PM2 5. Table 1 generally does not include factor analyses (such as those of Mar et al. [2(X)0, 2006]), because (as discussed in Sect. 1.4) such analyses typically come to inconsistent conclusions, because of varying inputs and methodological factors. 

Two points stand out with regard to the two multi-city studies. First, PM2 5 components such as BC (black carbon) or EC (elemental carbon, virtually the same as BC), OC (organic carbon), and metals (such as iron [Fe], nickel (Ni], and vanadium [V]) are not monitored. Secondly, these studies find health associations with sulfate, the only PM2.5 species monitored by both studies. 

The multi-county studies are different from the multi-city studies in that they (1) monitor for several more PM2.5 components (2) are done at a smaller... 

Table 1 Multi-city studies, multi-county studies, studies in Atlanta area varying conclusions regarding which particle components are associated with allcause mortality, cardiopulmonary mortality, or emergency department visits for cardio-vascular disease... 

Analysis of cohorts in a multi-site study coordinated by the EMCDDA shows substantial differences in mortality and causes of death between locations. In cities with high HIV infection among drug users, the impact of AIDS from the mid-1980s has raised mortality rates. In Barcelona (Figure 14), mortality reached over 50 per 1 000 users per year from 1 992 to 1 996 before falling markedly, reflecting a drop in AIDS deaths (probably because of new antiretroviral treatments) and, to a lesser extent, in overdose deaths. 

Multi-City, Multi-County and Site-Specific Studies... 

In epidemiology, ongoing efforts to deal with exposure misclassification should be continued. One particularly useful area of exploration might be to seek to understand the extent to which using smaller geographic areas (in terms of total size, and in terms of population density) instead of cities, in multi-locality epidemiology studies, may cause less exposure misclassification. 

---