- Wasteful publication includes dividing the results in
- a single study into two or more papers (“salami sci-
- ence”); republishing the same material in successive
- papers (which need not have identical format and
- content); and blending data from one study with
- additional data to extract yet another paper that
- could not make its way on the second set of data
- alone (meat extenders).
- Edward J. Huth, MD1
Duplicate publication is the simultaneous or subsequent reporting of essentially the same information, article, or major components of an article 2 or more times in 1 or more forms of media (either print or electronic format).2-9 Duplicate reporting includes duplicate submission and may apply to both published and unpublished works (eg, 1 or more manuscripts not yet published but under consideration by another journal). Other terms used to describe this practice include redundant, prior, repetitive, overlapping, related, multiple, dual, parallel, fragmented, fractionally divided, and topically divided publication.3,8,9
Duplicate submission or publication is not necessarily unethical, but failure to disclose the existence of duplicate articles, manuscripts, or other related material to editors and readers (covert duplication) is unethical and may represent a violation of copyright law. Moreover, reports of the same data in multiple articles waste publishing resources (ie, those of editors, reviewers, and readers as well as journal pages),1 pollute the literature, result in double counting of data or inappropriate weighting of the results of a study and thereby distort the available evidence,2 cause problems for researchers and those who conduct systematic reviews and meta-analyses,10,11 and may damage the reputation of authors.12
Duplicate publication usually involves 1 or more of the same authors, but the number of authors and order of authors may differ among the duplicate reports. Duplication occurs when there is substantial overlap in 1 or more elements of an article or manuscript. For reports of research, duplicative elements may include any or all of the following: the design, materials and methods, samples or subsamples, data, outcomes, tables, graphics and illustrative material, discussion, or conclusions. Duplication also occurs in other types of articles (eg, reviews, case reports, opinion pieces, letters to the editor, and online blogs).
A widely accepted method of quantifying the amount of overlap or duplication does not exist. Authors and editors often disagree on how to define and quantify duplication and whether duplicate articles are justified.13 Researchers in 2 studies of duplicate publication classified an article as duplicative of another if 10% or more of the content was identical or highly similar.7,14 Others have described levels and patterns of duplicate publication for research articles that emanate from 1 study, such as reporting identical samples and identical outcomes, identical samples and different outcomes, increasing or decreasing sample sizes and identical outcomes, and different subsamples from the same overall large study and different outcomes.12,15 Studies have also shown that most duplicate articles are published within 1 year of the publication of the first report.12,16
A number of studies of duplicate publication in various fields have found that 1.4% to 28% of published articles could be classified as duplicative of other articles.7,8,10,14,16-20 In addition, these studies have concluded that as many as 5% to 32% of duplicative articles do not include a citation or reference to the original or primary article (covert duplication).7,10,14,18
Following the recommendations of the International Committee of Medical Journal Editors (ICMJE),2 a policy that prohibits or discourages duplicate publication does not preclude consideration of manuscripts that have been presented orally or in abstract or poster form at a professional meeting. This policy applies whether the presentation is made in person or via Web cast or an online meeting presentation. However, publication of complete manuscripts in proceedings of such meetings in print or online may preclude consideration for publication in a primary-source journal. News reports that cover presentations of data at scheduled professional meetings would not necessarily violate this policy, but authors should avoid distributing copies of their complete manuscripts, tables, and illustrations during such meetings. Preliminary release of information directly to the news media, usually through press conferences or news releases, may jeopardize an author’s chances for publication in a primary-source journal.21 However, exceptions are made when a government health agency determines that there is an immediate public need for such information8,21 (see 5.13.1, Release of Information to the Public and Journal/Author Relations With the News Media, Release of Information to the Public). See Box 1 for examples of duplicate reports that may be acceptable and necessary.