Publication bias in the anesthesiology literature.

BACKGROUND: Publication bias occurs because positive finding studies are more likely to be published. The dearth of studies of negative or equivalence findings can erroneously affect future research and potentially clinical care of patients. We hypothesized that positive studies were more likely to be published than negative studies in anesthesiology journals with a higher impact and circulation.
METHODS: A PubMed search for controlled trials in humans published in peer-reviewed anesthesiology journals during 2008 and 2009 was performed. Fourteen anesthesiology journals and 1163 studies were evaluated. The average clinical trial impact factor (average citations per article) for each journal was determined. The quartiles for the clinical trial impact factor for the journals included in the analysis were 4, 6.1, and 9.1. Studies were scored by 2 raters as positive or negative results of the primary stated outcome. Factors previously associated with publication were also extracted. The primary outcome, the proportion of positive and negative studies in the journals in the upper quartile of the clinical trial impact factor to the lower quartiles was compared using the Fisher exact test. The odds ratio for the effect of positive study results adjusted for other characteristics associated with publication was determined using binary logistic regression. A multinomial logistic regression model was fitted for the journals with an impact factor in the upper quartile with adjustment for study trial registration, origin of publication, positive study findings, reporting of treatment blinding, reporting of subject withdrawals, study sponsorship, and description of the randomization method.
RESULTS: Positive finding studies were identified in 72% (425 of 588) of articles in journals with a clinical trial impact factor >9.1 compared with 53% (308 of 575) in journals <9.1 (P < 0.001). After adjusting for factors associated with publication, positive study results had an odds ratio (95% confidence interval) of 2.28 (1.76-3.01) for publication in an anesthesiology journal in the upper quartile. Multinomial logistic regression identified positive study findings associated with an increased likelihood of publication in 3 of the 4 anesthesiology journals with a clinical trial impact factor >9.1.
CONCLUSION: This study reports the presence of publication bias in the anesthesiology literature especially in higher clinical trial impact factor journals. Publication bias can have potential implications for future research and the clinical care of patients. Authors should be encouraged to submit negative studies to high impact journals and the journals should be encouraged to evaluate the editorial process as the cause of publication bias.