Publication Bias in Abstracts Presented to the Annual Scientific Meeting of the American Society of Plastic Surgeons

Sir: Publication bias—a form of selection—refers to the fact that studies with positive outcomes are more likely to be published in peer-reviewed journals.1–3 In an era in which clinical decision-making is heavily influenced by evidence-based medicine and multiple meta-analyses, the presence of publication bias could influence physician decision-making ability and proper care given to patients. Until now, no study has addressed the role of publication bias in the plastic surgery literature. Therefore, we sought to determine the presence of publication bias, if any, in abstracts presented at the 2003 Annual Meeting of the American Society of Plastic Surgeons. A retrospective analysis of oral and poster abstracts presented at the 2003 Annual Meeting of the American Society of Plastic Surgeons in San Diego, California, in October of 2003 was performed. Abstract books were reviewed by two of the authors familiar with research methodology independently, and the appropriate data were extracted from each abstract. To determine the publication status, a MEDLINE and EMBASE search was conducted independently by two authors in September of 2009 using the names of first and senior authors and key words extracted from each abstract. If no results were obtained through electronic search, the authors were contacted by e-mail or facsimile to inquire about publication status of their abstracts. The follow-up time in our study was chosen to be 6 years from the date of presentation based on a pervious finding that more than 90 percent of abstracts are published within 5 years of presentation.4 Proper statistical analysis was performed using SPSS version 16.0 statistical software (SPSS, Inc., Chicago, Ill.) (Fig. 1 and Table 1).Fig. 1.: Flow chart of primarily data collection and data-based research. Award papers (n = 13) and case reports (n = 9) were excluded. Case series were included in the study.Table 1: Parameters Investigated in Relation to Publication BiasUnivariate analysis demonstrated that studies that are prospective and blinded and with sponsorship are more likely to be published. To account for each parameter independently, a multivariate analysis was performed, which showed that prospective design and blinding were the only two factors that significantly affected the publication rate (Table 1). Although the review of presented abstracts showed no evidence of positive outcome bias, this could be secondary to multiple factors. First, only 138 abstracts in this study were reviewed. Thus, our study might have been underpowered. Second, given the nature of American Society of Plastic Surgeons meetings, abstracts undergo an extensive screening process before acceptance for presentation. This in itself might pose a risk for publication bias. Therefore, to understand the presence of publication bias at this level, further studies are needed to study all abstracts submitted to scientific journals. To limit the influence of publication bias in the future, several techniques could be used. Increased use of a national clinical registry such as www.clinicaltrials.gov could reduce the influence of publication bias in ongoing research.5 Unfortunately, retrospective studies and case reports are not required to be registered online; therefore, the role of these national clinical registries remains questionable. In summary, review of published abstracts presented at the 2003 Annual Meeting of the American Society of Plastic Surgeons reveals that prospective and blinded studies are more likely to be published, and there is no evidence of positive publication bias. Thus, researchers could feel confident that the likelihood of their research acceptance for publication will not be affected by the direction of their findings. Hani Sinno, M.D., C.M., M.Eng. Ali Izadpanah, M.D., C.M. Arash Izadpanah, B.Sc. Mirko S. Gilardino, M.D., C.M., M.Sc. Division of Plastic and Reconstructive Surgery McGill University McGill University Health Center Montreal, Quebec, Canada DISCLOSURE The authors have no financial interest to declare in relation to the content of this article.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.