Factors relating to the publication rate of abstracts presented from 2012 to 2015 at the Society for Vascular Surgery Annual Meeting
Additional Document Info
OBJECTIVE: The Society for Vascular Surgery's Annual Meeting acts as a means of disseminating research findings among vascular surgeons through the presentation of research abstracts. Following presentation at the meeting, research is often compiled into a full-text manuscript and submitted to a peer-reviewed journal. However, not all abstracts accepted to the Vascular Annual Meeting (VAM) eventually have a corresponding full-text publication. The objectives of this study were to establish the publication rate of abstracts presented between 2012 and 2015 to the VAM and to identify factors correlating with publication status. METHODS: Abstracts presented at the VAM were available through the Journal of Vascular Surgery. Data extracted from eligible abstracts included level of evidence according to the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence scheme, type of study (prognostic, therapeutic/harm, diagnostic), sample size, and status of outcome (positive, negative, or descriptive findings). Publication status of the abstracts was determined through a comprehensive literature review of PubMed (MEDLINE), Ovid (MEDLINE), and Embase. A multivariable logistic regression was conducted to determine factors correlating with publication status. RESULTS: The publication rate during the study period was 43.0% with a median time to publication of 9 months, with 412 of the 958 abstracts having a corresponding full-text publication in 48 journals with weighted mean impact factor of 3.40. Eleven journals collectively published 372 (90.3%) of the articles, with the Journal of Vascular Surgery publishing 280 (68.0%) of the manuscripts. Our logistic regression model demonstrated that factors positively affecting publication status were a positive status of outcome (odds ratio, 2.59; 95% confidence interval, 1.56-4.28) and a logarithmic increase in the sample size of the study (odds ratio, 1.35; 95% confidence interval, 1.13-1.60). In addition, studies with a corresponding full-text publication had a greater median sample size (250) compared with those without one (143; P < .001). CONCLUSIONS: From 2012 to 2015, 43.0% of VAM abstracts had a corresponding full-text publication, with greater sample size and a positive status of outcome positively correlating with likelihood of publication. Studies with negative findings made up a small proportion of conference abstracts (9.6%) and were the least likely to be published. Given the relatively small size of the specialty of vascular surgery, it may be particularly important to be mindful of publication bias. It may be worthwhile to give additional consideration to acceptance of abstracts or publication of studies with negative results that meaningfully contribute to the literature.