Share this article on:

Studies of Particulate Air Pollution and Mortality: When Do Authors Comment on the Policy Implications?

Begier, Elizabeth M.; Samet, Jonathan M.


Department of Epidemiology

Johns Hopkins Bloomberg School of Public Health

615 N. Wolfe St., Suite 6041

Baltimore, MD 21205

(address correspondence to: Jonathan M. Samet)

To the Editor:

This journal’s long-standing policy of excluding policy recommendations from original research articles, limiting them to editorials and commentaries, 1 has been debated on its pages throughout the last decade. 2–11 Inclusion of such statements has been advocated to clarify policy implications of scientific results and place findings into a broader public health context. 2–6 Rothman and others have argued, however, that policy statements should be set off from research results to avoid jeopardizing scientific objectivity and trivializing complicated policy issues. 7–8 Others concur, adding that a single study is unlikely to provide the weight of evidence needed to support policy recommendations. 9–11

But when authors comment on policy, are these comments based on their individual study results or do authors consider the entire body of literature? We investigated the influence of these and others factors, using the numerous time-series studies of particulate air pollution and mortality published in the 1990s. This body of work established the adverse effects of particulate air pollution on mortality, 12 leading to the 1997 promulgation of a new standard for fine particle air pollution in the United States 13 and consideration of new standards in other countries. 14

The PubMed 15 search engine, with its reference list of located articles, and the Environmental Protection Agency’s draft criteria on particulate matter 16 were used to locate articles for this review. We sought peer-reviewed English-language journal articles, published 1 January 1990 through 31 December 2000, that examined the association between daily mortality and air-particulate levels using time-series methodology (N = 102). We reviewed articles twice for policy statements and the study characteristics listed in Table 1. A policy statement was defined as “any statement that implies, suggests, or recommends that an action be taken to improve the public’s health, on the basis of evidence supporting such action.”14 Statements were categorized as “more research needed” or “specific action implied or suggested.” We excluded the seven studies published in Epidemiology because of the journal’s ban on policy statements; only one of these studies included such a statement, which called for more research. Frequencies of each statement type were tabulated by study characteristic.



Twenty-four articles (28%) included policy statements, with most of these implying or suggesting a specific course of action (78%). We found variability as to whether policy statements were included even among articles by the same author (Table 1), providing evidence that factors other than authorship were influencing inclusion of policy statements. Jackson et al. 17 found that 24% of 336 epidemiology journal publications included policy statements; the frequency was 27% if Epidemiology articles were excluded. In our sample of 26 epidemiology journal publications, the frequency was only slightly higher: 27% when including the seven Epidemiology articles excluded from our other analyses and 32% without the Epidemiology publications, suggesting that our sample is representative of epidemiology publications as a whole.

Publication after 1996 and positive study results were associated with a higher frequency of “specific-action” recommendations. Both study characteristics were independent predictors. Nearly all articles with specific-action statements had “positive” results (20/21), but the frequency of such statements for “positive” studies was 14% higher during and after 1997 than before 1997 (Table 2). Calendar time likely serves as a proxy for accumulating scientific evidence supporting a particulate-mortality association, suggesting that although authors’ own study results influenced them to comment, the related body of evidence did as well. The calendar-time effect is less likely to be attributable to the proposed American and European particulate standards in the late 1990s, as a similar increase is seen outside of these geographic areas. Specific-action recommendations were also more frequent in publications with government-affiliated authors and authors based outside of the United States and Europe, perhaps reflecting scientific culture differences by geography and affiliation.



Statements calling for more research were few and found nearly exclusively in studies with government funding and a university-affiliated first author, suggesting that those using and dependent on public funding are more likely to advocate continued research. However, small numbers limit inferences here.

In conclusion, we will offer no recommendation or statement.

Elizabeth M. Begier

Jonathan M. Samet

Back to Top | Article Outline


1. Our policy on policy [editorial]. Epidemiology 2001; 12: 371–372.
2. Teret S. So what? Epidemiology 1993; 4: 93–94.
3. Macdonald SC. Authors should be expected to elucidate policy implications of empiric data. Epidemiology 1993; 4: 557–558.
4. Diez-Roux AV, Benach J, Tapia JA. Should policy recommendations be excluded from epidemiologic research papers? Epidemiology 1994; 5: 637–638.
5. Coughlin SS. A ban on policy recommendations in epidemiology research papers? Surely, you jest! Epidemiology 1994; 5: 257–258.
6. Teret S. Policy and science: should epidemiologists comment on the policy implications of their research? Epidemiology 2001; 12: 374–375.
7. Rothman KJ, Poole C. Science and policy making. Am J Public Health 1985; 75: 340–341.
8. Rothman KJ. Policy recommendations in epidemiology research papers. Epidemiology 1993; 4: 94–95.
9. Greenland S. In: Taubes G. Epidemiology faces its limits. Science 1995; 269: 164–169.
10. Weiss NS. Policy emanating from epidemiologic data: what is the proper forum? Epidemiology 2001; 12: 373–374.
11. Greenbaum DS. Epidemiology on the edge. Epidemiology 2001; 12: 376–377.
12. Pope CA 3rd, Dockery DW. Epidemiology of particulate effects. In: Holgate ST, Samet JM, Koren HS, Maynard RL, eds. Air Pollution and Health. London: Academic Press, 1999; 31.673–31.706.
13. Environmental Protection Agency. National Ambient Air Quality Standards, 1997. Available at: criteria.html. Accessed 10 February 2002.
14. Council of European Union Directive, 1998. Available at: http://lhtee. meng. auth. gr/AIR-EIA/EU/99L0030_en. Pdf.
15. National Library of Medicine. PubMed. Available at: http://www. ncbi.nlm. nih. gov/entrez/query. fcgi. Accessed 15 May 2001.
16. Environmental Protection Agency. Air Quality Criteria for Particulate Matter (External Review Draft). 2001. Available at: http://cfpub1. epa. gov/ncea/cfm/partmatt.cfm?ActType= default.
17. Jackson LW, Lee NL, Samet JM. Frequency of policy recommendations in epidemiologic publications. Am J Public Health 1999; 89: 1206–1211.
© 2002 Lippincott Williams & Wilkins, Inc.