Publication bias in plastic and reconstructive surgery: A retrospective review on 128 abstracts presented to the Annual EURAPS Meeting

Background: Publication bias is defined as favoring publication of research results based on their nature and direction (e.g., positive vs. negative results). We sought to determine the presence of publication bias in the plastic surgery literature. Methods: Abstracts presented to the 2003 and 2004 Annual Meeting of European Association of Plastic Surgeons (EURAPS) were independently reviewed by two different reviewers. Abstracts were investigated for sample size, type of study, level of evidence, country of origin, presence of statistical results, and sponsorship. Publications within 6 years from the meeting were ascertained by electronic search of Medline and Embase databases. In the absence of positive searching results, authors were contacted directly. Fisher's exact test and logistic regression analysis were used to identify predictors of publication. Results: One hundred thirty-two abstracts were reviewed. Of these, 128 abstracts met our inclusion criteria. Seventy-three (60.8 %) were published within 6 years of the European Association of Plastic Surgeons' meeting. Primary authors of the remaining 55 unpublished abstracts were contacted. Fishers exact analysis demonstrated that positive studies were more likely to be published (P=0.045) with an odds ratio of 6.54. However, a logistic regression analysis did not show any statistically significant difference. Conclusions: We found no evidence of publication bias in the review of manuscripts presented at the 2003 and 2004 Annual Meeting of EURAPS. Thus, researchers can be confident that the likelihood of manuscript acceptance will not be influenced by the direction of study findings. Level of Evidence: Not ratable. © 2014 Springer-Verlag.