engleski

Editorial police what say you? Analysis of duplicate publications in PubMed.

Background: Committee on Publication Ethics states that journal editors have a responsibility for pursuing cases of suspected misconduct even in submissions they do not intend to publish, and provides guidance on how to publish expressions of concern, corrections and notices of retraction. Aim: To determine how transparent are journal editors in their handling of duplicate publications. Methods: We analysed 1011 articles listed as “duplicate publication” under publication type in PubMed on 16 January 2013, checked their titles, time from the duplicate article publication to the notice of duplication/retraction and analyzed their content, including reasons for the notice and duplication, and methods of resolving them. Results: 680 (67.3%) duplicate publications identified in PubMed lacked any notice in their respective journals. 175 notices of 331 duplicate publications were marked as either Comments or Erratums, however 97 (55.4%) of these were not linked to the articles at their respective journals’ websites. Of the 175 notices, 112 (64%) reported author’s error as the reason of duplication and 63 (36%) the publisher’s error, yet only 23 (13.1%) were retracted. Of notices reporting author’s error 45 (40%) do not mention contacting the author or the authors response. In 2 (2%) of notices, both published in the same journal, editors have decided to ban authors from further publication in that journal. Conclusions: More than half of duplicate publications identified in PubMed have not been corrected by journal editors, and half of those that were are not visible on the journals website. Furthermore, almost half of notices do not provide information on contacting the authors, and only 13% are followed by retraction of articles. Journal editors seem to be unwilling to take actions in cases of article duplications. In order to preserve the integrity of the published record a more active role of all stakeholders is needed.

duplicate publications; research integrity; Medline

nije evidentirano