For additional discussion on topics related to this issue, see the following:
Ethics Case: Self-Plagiarism of Textbook Chapters
Ethics Case: Duplicate Submission
Simultaneous Submission to 2 Journals

We would like advice on the following issue.

We received a review article for our journal and put it through our refereeing process. After modifications suggested by the reviewers it was accepted for publication. As I was preparing the manuscript to go to the printers I became aware of a similar article just published this week by the authors in a refereed medical journal. There appears to be some overlap of material being presented in both articles. For explanation, the published article in the medical journal primarily covered the clinical details (with some lab methodology) while the one submitted to us covered primarily lab methodology (with some case histories). One submitted figure is identical to one in the published article.

My question is: how much overlap is allowed (25%? 50%?). Also, in any overlap allowed, does the text have to be different, or can it be word for word?

Rob Siebers
Editor, New Zealand Journal of Medical Laboratory Science

___________________________

I am very interested as well in this question given that there was a very extensive overlap of authors, objectives, and study results in two articles presented last week , one in JAMA and one in Academic Medicine. They were clearly derived from the same study, but just were 2 different methodologies of asking the same questions.

Margaret Gadon
Senior Scientist, Medicine and Public Health, American Medical Association

___________________________

An important distinction in the case to which Dr Gadon refers is that the JAMA study http://jama.ama-assn.org/cgi/content/abstract/294/9/1058 cited the Academic Medicine study http://www.academicmedicine.org/cgi/content/abstract/80/9/874 as part of the Methods: "A draft of the survey instrument was developed based on literature review, focus groups with residents in each specialty (ref to Academic Medicine study)." Therefore, from JAMA's standpoint the authors are not concealing the existence of a related manuscript. This situation sounds quite different from the one to which Dr Siebers refers.

Margaret Winker
Deputy Editor, JAMA
Director, Division of Scientific Online Resources

___________________________

Dear Rob, I think your case is more of a case of salami publication, where the authors are stretching their results as thin as possible, publishing 2 different aspects of the same study. It is difficult to give advice, as it depends on the editorial policy and you noticed it after the article was accepted and was getting ready for print. We have the policy of checking the accuracy of all references in articles, and the cases of duplicate or salami publications come out when we search Medline for the name of the authors on the article. In such cases we go back to the author and ask for clarification.

Ana Marusic

___________________________

My rule of thumb is: if it is duplicate data, that is unacceptable. If data must be repeated (to pursue a line of argument, for example), then it must be cited (in either the Background or Discussion sections)—never reresented as if fresh in the Results.

Chris Del Mar
Research Editor, Aust Fam Physician

___________________________

I am aware of an article pending simultaneous publication in 5 journals. This was disclosed on submission. What do we think about that?

Kristen King

___________________________

Duplicate publication is attractive: to authors to increase the number of papers on their CVs and to industry to scatter their data between more publications for marketing purposes. I regularly give workshops on publication of manuscripts and for the first time this year the most questions I received related to duplicate publication, indicating this has become a hot topic, questions such as "what can we get away with?" My advice has always been to inform the journal about all related publications and submissions.

There is also of course the problem that journals cannot accommodate long manuscripts that might include all the data. With my own journal and others too this problem is being increasingly resolved by publishing full articles on the Internet and a shorter paper in the printed version of the journal.

Elise Langdon-Neuner
Managing Editor, Journal of Men's Health and Gender

___________________________

I do many studies that are designed to address a series of related questions and produce a series of publications—sometimes 50 or more, if you're talking about the whole lab group. Certainly large national or epidemiological datasets fit this scenario. This rule of thumb re duplicate data is not applicable to nonexperimental studies.

I would also note the conflict between the pressure from many journals (at least in the US and including my own) on shortening manuscripts and the problem of multiple publication or salami publication. Whereas I used to commonly write and get accepted long manuscripts (40-60 pages) that addressed a number of closely related questions, journals will no longer touch anything longer than 30 pages, including cover page, abstract, tables and references. I have also noticed that if I write a very short, clear manuscript with just a single part of a larger project (20 pages of text, not including tables or references) that my odds of getting the paper accepted are much higher than if I write a more complex manuscript. This is especially true (and I see myself doing this when I review other people's work) because in each new paper I write I cite all the others and the prior publications increase the credibility of the current one and tend to make me more lenient in evaluating the methodology.

In other words, there are lots of incentives for breaking complex questions into a series of closely related publications. However, I think there is a difference between duplication and overlapping in that the the goals of the papers should be different and each should make an independent contribution to the literature. That sounds to me like the case in the case of a clinical vs lab focus of the 2 papers.

Nancy Darling

___________________________

Outrageous!

We make it quite clear in our guidelines for authors that we won't consider articles that have been submitted to other journals.

Of course, it doesn't always prevent it. I had a situation a couple of years ago where I was working with an author on a paper (I had accepted it for publication) and it suddenly appeared in a German journal. The author hadn't informed me that he had submitted it to the other journal and seemed quite surprised when I told him that this was not on and dropped it from mine.

I would tell your author to make his mind up!

Deborah Glover
Editor, JWC

___________________________

I wonder if you have noted the comment that the multiple submission 'was disclosed on submission'. Why is this outrageous? Surely it is nondisclosure that is unacceptable.

The CONSORT Statement has been published in >10 journals. I believe the paper alluded to is of a similar nature—ie, not a research paper. Surely it is good to disseminate such information widely as long as everything is done openly.

Doug Altman
___________________________

Of course I agree that nondisclosure is unacceptable. Perhaps my wording was clumsy, but I meant it to be slightly tongue in cheek, hence the exclamation mark...

I didn't get from Kirsten's e-mail that it was a paper similar to the CONSORT statement. I agree that such information should be disseminated widely.

However, even if the authors did disclose on submission that they had simultaneously submitted to other journals, surely this leaves the editor in a quandary. Do they send the paper for peer review and hope that one of the other titles doesn't publish it first? Do they tell the author that they will peer review it on the condition that they withdraw it from the other journals? Do they try to contact the other editors to see if simultaneous publication is in the wider interest of practitioners?

Deborah Glover
Editor, JWC

___________________________

I actually received a paper the other day, by e-mail, with the message, "Dear Editor: Please consider the attached manuscript for review in your journal...." Note that it was not addressed to me personally nor was the journal name included. However, the paper topic was appropriate for JHPN (Journal of Hospice and Palliative Nursing). I happened to glance at the e-mail address box and noticed that the author had sent the message to 5 other people—and I recognized some of the names/addresses as other editors. I wrote to the author and asked for clarification and explained that the guidelines for my journal (and most others in the world) stipulate that you can only submit to one journal at a time.

The author wrote back and thanked me for the information—clearly it was an error of inexperience. Still, the author was prepared at the master's level—you'd think that somewhere along the line she would have learned something about the rules and regs of professional publishing.

Leslie H. Nicoll
Maine Desk, LLC-Professional Editorial Services
Editorial Office for CIN and JHPN

___________________________

Most journals I publish in require that I sign a letter saying that the work hasn't been published elsewhere prior to publication (usually around the time page proofs come out and I sign a copyright release). With electronic submission, I'm wondering if it wouldn't make sense to ask people to indicate AT THE TIME OF SUBMISSION that they aren't submitting simultaneously and, if applicable, that their work complies to whatever standards for ethical behavior that the journal adheres to. Would this help, do you think?

Nancy Darling

___________________________

Let us also not lose sight to the fact that in some cases, duplicate publishing is desirable but there is disclosure. This reminds me of the task force on revitalization of academic medicine (or some other similar name) which agreed that it would be desirable if Tugwell's editorial would appear in many journals. This ended up being multiple publications and Tugwell was cited tens of times from one editorial. But this was agreed and there has been full disclosure. There is also this initiative where African Journals and "northern" journals are partnering in strengthening each other! I am part of that initiative. We discussed the possibility of some sort of duplicate publications. Take for instance, a crucial (however that may be defined) paper from the south published in the north and most of the people in the south would benefit from seeing the results/paper from the south. Can the southern journal be allowed to publish a modified version of that paper? At the time we had agreed among ourselves that it would be ok; of course with full disclosure that the paper had been published in a different format elsewhere. It is when people are hiding information and naturally, the tendency among Homo sapiens is that whomever is hiding something is the bad guy; after all if there is nothing to hide, why hide. Therefore whosoever hides a thing knows there is something to hide. This does not make sense even to me!

Adamson Muula
Malawi Medical Journal

___________________________

The latter is our current policy at The Journal of Pediatrics. We require that the letter of submission state certain things, such as not published/submitted elsewhere, etc. (This is detailed in our Guide for Authors.) The letter of submission is screened by our excellent Senior Editorial Assistants and, if it does not contain all required elements, they send it back to the authors with details about what they need to do for us to accept the submission of their paper. (Because we use an online submission and review system, the feature of "Send back to authors" means that the time for the authors to "fix" their letter does not count against our turnaround time *and* does not require that the authors "start from scratch" to (re)submit the paper.)

Of course, as Deborah pointed out, this does not necessarily prevent duplicate submission/publication and other problems....

Alice Landwehr
Managing Editor, The Journal of Pediatrics

___________________________

Dear Rob,
There may be appropriate times for overlap and duplicate/redundant publication, however, each situation needs to be addressed individually. Disclosure seems to be an important component of the situation that you describe.

We have recently been through a similar experience with a paper that overlapped with one published afterward in another journal. Both had identical wording by at least 50%, using the same project but resulting in 2 slightly different data analyses. The author did not disclose to either editor about the other publication, nor was the previous paper cited in the second publication. Thus, for us, disclosure was a primary issue.

As one of the actions to increase awareness of the subtlety of this topic, we e-mailed our editorial board members an update on redundant/duplicate publication and the relevant portion of our instructions for authors. Below is a portion of that email.

Also, the text Ethical Issues in Biomedical Publication, by Hudson Jones A, McLellan F, is very informative. I hope you find this information helpful.

Definition of duplicate publication:
Duplicate or redundant publication is the publication of a paper that overlaps substantially with one already published. Both papers need not be identical to be considered a duplicate/redundant publication. Other terms that have been used include dual, divided, republication, fragmented, prior, repetitive, and salami slicing. Although there are a few exceptions when duplicate publication is allowed, in general it is not condoned. It is especially inappropriate when an author does not notify the editors nor reference the similar/redundant work in the body of the paper.

Reasons why duplicate publication is inappropriate:

• Adds redundant material to an already extensive amount of literature
• Inappropriately influences meta-analysis by increasing number of results
• Wastes time for editors, peer reviewers, scientists, readers by reading and reviewing material that is redundant
• Wastes journal resources by using print or web space that should be used for original articles
• Uses resources to investigate a case of duplicate publication
• Copyright law infringement

Towards a better classification of duplicate publication.
Bailey described 5 levels of duplicate publication in 1989 at the First International Congress on Peer Review in Biomedical Publication. Davidhizer and Gigner (2002) summarize these levels:

Another taxonomy for duplicate publication was described by von Elm and summarized in Bandolier (http://www.jr2.ox.ac.uk/bandolier/band123/b123-6.html)

JMPT policy on duplicate publication:
The JMPT instructions for authors states: "The JMPT does not publish articles containing material that has been reported at length elsewhere. The corresponding author must include in the cover letter a statement to the editor about all submissions and previous materials that might be considered to be redundant or duplicate publication of similar work, including if the manuscript includes materials on which the authors have published a previous report or have submitted a related report to another publication. Copies of the related material may be requested by the editor in order to assist with the editorial decision of the paper. If redundant or duplicate publication is attempted or occurs without proper disclosure to the editor, editorial action will be taken as follows. If it is confirmed that a paper is a duplicate or redundant publication and is discovered in the prepublication phase, the paper will be rejected, even if an accept notice has been distributed previously to the authors. If duplicate or redundant publication is confirmed after publication, the paper will be retracted and the appropriate boards/institutions notified."

Other information on duplicate publication:
1. Huth EJ. Repetitive and divided publication. In: Hudson Jones A, McLellan F, editors. Ethical Issues in Biomedical Publication. Baltimore: The Johns Hopkins University Press; 2000. p. 112-136.
2. von Elm E, Poglia G, Walder B, Tramer MR. Different patterns of duplicate publication: an analysis of articles used in systematic reviews. JAMA. 2004 Feb 25;291(8):974-80.
3. Tramer MR, Reynolds DJ, Moore RA, McQuay HJ. Impact of covert duplicate publication on meta-analysis: a case study. BMJ. 1997 Sep 13;315(7109):635-40.
4. Abraham P. Duplicate and salami publications. J Postgrad Med. 2000 Apr-Jun;46(2):67-9.

___________________________

We do face this problem of simultaneous submission to different medical journals despite the fact that the authors have stated that the manuscript has been exclusively submitted to our journal. More recently, we detected such a case when our journal was in press and had to remove that manuscript.

We have blacklisted the authors and the decision was also conveyed to their parent institution as well as Pakistan Medical and Dental Council and other editors of medical journals published from Pakistan. Our current issue of Pakistan Journal of Medical Sciences (July-September) also carries an editorial on the same subject: "Simultaneous submission and duplicate publication: A curse and a menace which needs to be checked." Such authors need to be punished if we are serious about tackling this problem.

Those interested can have a look at the editorial at our website (pjms.com.pk) in the July-September 2005 issue.

Shaukat Ali Jawaid
Managing Editor, Pakistan Journal of Medical Sciences

___________________________

We have been dealing with this issue here a lot. It's our belief that anything that's already been published must be presented as such in a second publication-put in quotes-and tables and figures must have reprint permission. Even if the author retained the copyright, we believe that readers need to know that what they are reading was written and published before in another publication, even if one paragraph. It seems that self-plagiarism is alive and well and accepted in the scientific publishing community. Few people, including some editors, find it unacceptable unless its a duplicate publication. If the scientific community wants to be exempt from writing standards and ethical writing practices espoused by nonscientific writers, then I think it should explicitly say so in a policy statement. The International Academy of Nursing Editors has been grappling with this very issue. How many ways are there to write a methods section and readers often find it unacceptable to be referred to a prior article to find out the details of the methods. I'm eager to see the dialogue on this listserv, as we've been trying to craft a policy with regard to self-plagiarism and are finding it challenging.

Diana J. Mason
Editor-in-Chief, American Journal of Nursing

___________________________

The experience from Pakistan reported by Dr Jawaid is indeed sad. However, we need also to consider that in some cases, it is possible that the author has submitted to more than one journal because of sheer ignorance that this is not acceptable. If the instructions to the authors specifically requested disclosure, then probably the authors have little defense. But even then it may be possible that our understanding of what this means may be scanty. Let us however spread the word to all, that duplicate publications should only occur with the editors' prior knowledge that such is what has happened. But even when people have erred, as the editors may see fit, some may be whipped in public, others in private.

Adamson Muula

___________________________

I think the notion of "self-plagiarism" is a new one for a lot of authors. In their minds, the words are their own even after they have been published and another entity holds the copyright. I have always thought it was pretty clear, particularly after an article publishes with a journal's copyright notice, but there seems to be a communication breakdown. Is it a lack of clarity on the part of the publication? Lack of experience (or perhaps some willful ignorance?) on the part of those submitting? How can we better educate our authors?

Kristen King

___________________________

Would this include descriptions of methods, samples, etc. that may be identical from one article to the next working with the same large dataset? For example, if you use the identical depression inventory and measures of demographic information across a series of studies or papers, does this need to be in quotes after the first time you have written the description?

Nancy Darling

___________________________

I'm sorry, but I guess I'm just not understanding this.

It would be unacceptable for one author to write 2 papers using the same dataset to address two different papers but with identical datasets unless the whole method section was in quotes or virtually identical material were rewritten. Correct?

But it would be acceptable (because it is not self-plagiarism) for 2 different first authors to submit 2 papers with identical methods sections with no quotes? Would not at least one of them be plagiarizing?

And is not paraphrasing identical ideas simply another form of plagiarism?

Let me give an example from psychology of why I think this is problematic, but I could find examples from more biologically based disciplines.

Say one is describing a 5-item scale assessing parental monitoring. You describe what the scale is, cite the original authors of the scale, what the items area, and how scores were calculated. This might take 3 sentences in a 30-page manuscript.

1. There really are only a limited number of ways one can write that description and still be clear. But I might use this measure in 30 or 40 or 100 different papers.
2. If you put the description in quotes, you not only have to cite the developer of the scale, but then yourself for previously describing the same measure.
3. If you use 10 measures in one paper (not uncommon), you do this for each scale, often with at least several different citations to yourself because it is rare that you are using the same set of measures in another paper. That is specifically because this paper addresses a different question from the others, which is what makes it worthy of publication.
4. These additional citations (to yourself) don't really add to anyone's understanding of the literature or the independent contribution of this paper relative to your other published work. Just the measures are the same.

And remember, one of my "measures" might be gender or ethnicity.

It strikes me as more important to cite, at the beginning of the methods section, the source of the data, with at least one reference to the first publication using it. But I agree that this is a very complicated issue. From an ethical perspective, I believe it is critically important that the literature not be unduly cluttered with duplicate publications that don't provide additional information—for all the reasons previously discussed. And certainly plagiarism of others is stealing ideas without due citation or credit. I see the problem of self-plagiarism more as one of republishing ideas that you've already expressed—really a duplicate publication problem more than anything else. This is something I have run into when writing several papers on the same topic—usually in the literature review. I have to force myself to bring something different to the framing of each to make each an independent treatment. But from an ethics perspective I'm not sure that's true in a methods section, which serves a purely technical function.

Nancy Darling

___________________________

I agree that it is a complicated issue. Our position is that it is not OK to cut and paste anything from prior published work of your own without full and correct attribution, including quotes if you've left it verbatim. If you say that the rule doesn't apply to methods sections, then you're practice would violate most copyright agreements. I believe that the major biomedical publishing houses could agree to make methods sections an exception as long as the author provides the source for original publication.

That said, I'm looking forward to the NYU conference on Saturday on plagiarism and hoping that there will be some helpful consensus on how to proceed with these issues.

Diana J. Mason
Editor-in-Chief, American Journal of Nursing

___________________________

Dear Dr Mason and others on the WAME list who might attend the plagiarism meeting,

As a writing teacher of postgraduate students and an author's editor for groups of scientists in Spain, I'll be looking forward to hearing reports from the plagiarism symposium. If any reports or publications come out of it, I hope someone posts a notice to call our attention to them.

In particular, I'll be interested to see if people mention that at least some "plagiarism" is the result of the loss of the art of sensible, communicative citing practice—a link some of us have mentioned on the WAME list before. I am finding that, with so much more publishing going on nowadays, and probably many more authors producing manuscripts, many novices are alarmingly unaware of the finer points of interweaving information from sources that we used to take for granted. And some not-so-young and not-so-novice authors are blissfully unaware too, explaining why so much "mentoring" is going amiss and journals publish so many bad examples.

Whereas 15 or 20 years ago, I could point to examples in journals and say to young authors, "See how they've cited in this paper? That's how ...." These days, I can open practically any journal and quickly find another type of example—of misleading, confusing or simply incorrect citation. Or examples of an author's lifting a sentence from a source text and dumping it into a paragraph where it jars with all the other sentences before and after it because it has not been rewritten for the sake of cohesion and to make its relevance clear. Even though the sentence carries a proper reference number, and such 1- or 2-sentence copying/plagiarism seems to be tolerated by biomedical journals, readability is affected and young authors are confused by how much of such cut-and-paste is too much.

On an earlier WAME thread, some of us who are manuscript editors and/or translators have reported fortuitously finding instances of serious extended cut-and-paste plagiarism as we search the internet for guidance on terminology. I've found large swaths of copied (or literally translated) text with a reference at the end (with other references present in the source text left out!). The author is often surprised that one "can't" do that because s/he has seen such pasted text published before.

Editors: Beware any paragraph that is particularly well written upon first draft, especially if the reference is at the very end!

Best wishes,
M.E. Kerans
Barcelona, Spain

___________________________