Loading...
Please wait, while we are loading the content...
Invited commentary: air pollution and health-what can we learn from a hierarchical approach?
| Content Provider | Semantic Scholar |
|---|---|
| Author | Dominici, Francesca |
| Copyright Year | 2002 |
| Abstract | Received for publication June 27, 2001, and accepted for publication August 15, 2001. Abbreviations: NMMAPS, National Morbidity, Mortality, and Air Pollution Study; PM 10 , particulate matter ≤ 10 μm in diameter. From the Department of Biostatistics, Bloomberg School of Public Health, Johns Hopkins University, 615 North Wolfe Street, Room E4138, Baltimore, MD 21205 (e-mail: fdominic@jhsph.edu). (Reprint requests to Dr. Francesca Dominici at this address). The potential for air pollution to cause excess deaths at high concentrations was established in the mid-20th century by a series of air pollution “disasters” in the United States and Europe (1–3) that caused striking increases in mortality. By the early 1990’s, time-series studies, each conducted at a single location (4–7), showed that air pollution levels, even at much lower concentrations, were associated with increased rates of mortality and morbidity in cities in the United States, Europe, and other developed regions. At present, although these relative rates are small (an increase in mortality or morbidity of a few percentage points over a realistic exposure range), the burden of disease attributable to air pollution may be substantial, considering the very large populations exposed to air pollution and the large numbers of persons to whom the relative rates of mortality or morbidity apply. In the past, critics of single-site studies questioned the validity of the data used and the statistical techniques applied to them. The critics noted inconsistencies in findings among studies and even in the same city upon independent reanalysis (5, 6). They questioned the choice of particular cities and asked whether models had been selected that gave estimates of effect that were biased upwards. These criticisms have since been addressed by the use of multisite studies (8, 9) in which site-specific data on air pollution and health are assembled under a common framework. Hierarchical models, which combine information across locations, have provided a statistical approach for analyzing multisite studies (10). The work by Hwang and Chan (11), published in this issue of the Journal, is one of the latest contributions on this topic. Their study illustrates the utility of using hierarchical models to analyze data on the relation between air pollution concentrations and clinic visits for treatment of lower respiratory tract illness. Hwang and Chan analyzed such data (as well as data on temperature and dew point levels) for 50 sites in Taiwan in 1998. Here I discuss the advantages of using hierarchical models to analyze multisite time-series data on air pollution and health, provide perspective on the results of Hwang and Chan (11), and address the problem of publication bias in meta-analyses. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.biostat.jhsph.edu/~fdominic/papers/invitedcommentarychan.pdf |
| Alternate Webpage(s) | http://biosun01.biostat.jhsph.edu/~fdominic/papers/invitedcommentarychan.pdf |
| PubMed reference number | 11772778v1 |
| Volume Number | 155 |
| Issue Number | 1 |
| Journal | American journal of epidemiology |
| Language | English |
| Access Restriction | Open |
| Subject Keyword | Abbreviations Air Pollution Airborne Particulate Matter Cessation of life Email Estimated Morbidity - disease rate Particulate (substance) Published Comment Respiratory System Respiratory Tract Diseases Respiratory Tract Infections Scientific Publication |
| Content Type | Text |
| Resource Type | Article |
