부산대학교 도서관

Background: Abstract presentation at conferences provides the opportunity to rapidly communicate research findings. The outcome and impact of publications arising from cardiac electrophysiology abstracts are not known.Objective: The purpose of this study was to examine the characteristics of abstracts presented at the annual scientific sessions of Heart Rhythm Society (HRS), their publication rate, and the indexed impact of subsequent publications.Methods: Two independent database searches (MEDLINE and EMBASE) were performed by cross-referencing authors and keywords from abstracts originally presented at HRS in 2003. ISI Web of Knowledge was accessed for impact factors and citation rates.Results: A total of 790 abstracts were presented, of which 377 (47.7%) resulted in publication of an original article. Median time to publication was 1.39 years (interquartile range [IQR] 0.88-2.30 years), and the median impact factor and citation rate of published articles was 4.14 (IQR 3.48-11.05) and 10 (IQR 4-25), respectively. Experimental research abstract category (odds ratio [OR] 2.03, P <.001), randomized study design (OR 0.53, P = .02), and positive findings (OR 0.80, P = .06) were independently predictive of publication by stepwise logistic regression. Independent predictors of higher citation rates were randomized study design (P = .03) and impact factor of the publishing journal (P <.001).Conclusion: Almost half of all abstracts presented at HRS resulted in publication in journals with a high impact factor. Experimental research abstracts, those with a randomized study design, and those demonstrating positive findings were predictors of subsequent publication. Randomized study design and greater impact factor of the publishing journal were found to predict higher citation rates.