PUBLICATION BIAS AND PUBLIC-HEALTH POLICY ON ENVIRONMENTAL TOBACCO-SMOKE

Objective.-To examine the tobacco industry's claim that publication bi as against negative studies invalidates the risk assessment of environ mental tobacco smoke (ETS) conducted by the US Environmental Protectio n Agency and other reviews of the health effects of ETS. Design.-Deter mination of the number of published original research articles that te sted the hypothesis that ETS exposure is associated with adverse healt h effects and that reported statistically significant (''positive'') o r nonsignificant (''negative'') results; the number of articles that c oncluded that ETS is a health risk; and unpublished studies on the eff ects of ETS on health. Participants.-Articles identified by a computer ized search of the medical literature supplemented with material obtai ned from the tobacco industry and hand searching. Articles were classi fied as peer-reviewed journal articles or articles from sponsored symp osia. Main Outcome Measure.-The statistical significance of results re ported in the article and whether or not the article concluded that ET S exposure is a health risk. Results.-More symposium articles than jou rnal articles were reviews (46% vs 6%; P=.0001). More original journal articles than original symposium articles reported the use of statist ical tests (96% vs 54%; P=.0001). Of articles with statistical analyse s, similar proportions of journal articles and symposium articles repo rted statistically significant results (57% vs 47%; P=.329). The concl usions of 80% of the original journal articles were positive, compared with 51% of the original symposium articles (P=.006). Conclusions.-Th ere is no publication bias against statistically nonsignificant result s on ETS in the peer-reviewed literature. The high proportion of artic les in symposia that reach the conclusion that ETS is not harmful prim arily results from the inclusion of review articles.