The following case was submitted anonymously by an editor and discussed at the Fourth International Congress on Peer Review in Barcelona in September 2001. The case was presented to the audience by Michael Callaham of the WAME Ethics Committee and was discussed by an expert panel, consisting of Richard Smith (BMJ), Richard Horton (Lancet), and Frank Davidoff (Annals of Internal Medicine). There was much commentary from a very informed audience. A very brief summary of the topics and key discussion points follows below.
We received a submission for publication accompanied by the standard letter signed by all authors that stated that the manuscript had not been submitted to any other journals for consideration. By chance, one of the reviewers we selected received the identical manuscript from another Journal approximately one month later. The second Journal confirmed that they too had received a covering letter stating that the manuscript had not been submitted elsewhere for consideration.
The questions with which we grappled were as follows:
Should the manuscript be automatically rejected for consideration by us?
Should we inform the other Journal that we had received the identical manuscript for consideration despite the covering letters?
How do we deal with the authors of the manuscript who clearly attempted to mislead the editors of two journals with regard to their submissions? Should this be made public by publishing the details in an editorial in the Journal? Should the offending authors be named? Should we refuse further submissions from these authors because of their dishonesty? One of the authors was the Department Chairman of a prominent medical school.
Questions and discussion on case #2:
Why do we have a ban on duplicate submission (it's not the practice in other forms of journalism). — The history of this practice is the Ingelfinger rule established by the New England Journal of Medicine. The goal is to ensure that a journal is publishing only original material not previously available elsewhere. One goal of this policy may be efficiency— it is more efficient for only one journal, rather than many, to review and publish a manuscript. However, another goal is probably to the exclusive benefit of the journal, namely, to have exclusive material to make it more desirable to its subscribers (and to advertisers). This policy is not advantageous to authors, since it requires them to limit their audience and, perhaps more importantly, to submit to longer delays owing to sequential review and consideration at journals.
What about duplicate submissions or publication in different languages? — Many clinicians read only in their own language. Discussion from the audience made it clear that this can be a valid reason for duplicate publication, provided that the authors reveal the potential duplication on submission. Often the potential audience can only be reached by several methods–eg, English publication in a large international journal, but also publication in another language in a much smaller local journal, most of whose subscribers are in one country. If the potential duplication is revealed to all parties on submission, is approved by the editors (as is regularly done), and ideally is also cited in the bibliography, there is no ethical breach and “duplicate publication” in the pejorative sense has not occurred.
Did the reviewer violate confidentiality by reporting the duplicate submission to the other journal? — This is a difficult question to answer. Generally, a review is a confidential process and the reviewer is not allowed to share the materials without at least the permission of the editor supervising the review. However, in their designated role as reviewer for the same paper at both journals, they were entitled to communicate with the editor about it. Furthermore, one may argue that faced with such irrefutable evidence of duplicate submission (apparently covert), the normal obligations of the reviewer are changed by the potentially unethical conduct of the authors. At that point, the reviewer’s primary obligation would be to the journals, not the authors.
How explicit was the letter the authors signed saying they had not submitted elsewhere? — The more explicit the written agreement with the authors, the better. Ideally the authors are required on submission to submit a signed statement that the manuscript has not been submitted or published elsewhere. If this is not specifically addressed or the wording is vague, there is more opportunity of confusion (genuine or feigned).
How should the offending authors be punished? — The issues and general approach are the same as outlined in the previous case. Initial contact with the authors is suggested, asking for their explanation in writing; if their explanation is not satisfactory and the editors conclude that there may well have been deliberate deceit, this can be reported to their university or funding agency. It may subsequently appear that "nothing has happened" (because such referrals are always slow, details are usually not released to “outsiders”, and in some cases the institution ignores the issue) but nonetheless the impact on reputation can be considerable if any action is taken. Whatever action is taken, the involved journals should try to devise a common strategy. Richard Smith stated that if the BMJ discovered such a case of duplicate submission it would not publish the details but would automatically reject the article. One of the big problems is finding solid proof that authors have behaved unethically, but if they can lie about a submission they could have lied about their results and the paper certainly should not be published.
What about blacklisting authors who do this? — As addressed in Reviewer Conflict of Interest, blacklisting does occur, but suffers from a number of limitations already discussed. In fields with many suitable journals, it would have little impact unless adopted by all the journals. Blacklisting can be of indefinite duration, or limited to a fixed amount of time (although one might argue that willful deception on the part of authors about submissions reveals something about character not likely to improve with time).
Is an editorial or published statement on this case excessive punishment? — This approach was recommended and has been occasionally used. If the authors have unequivocally violated their signed agreement, the punishment may not be excessive and it is likely to be faster and more certain than referral to their institution. In a case where there might somehow have been some genuine misunderstanding, it would be excessive. Also, depending on the litigiousness of the country in which the journal resides, such an editorial might be considered defamatory. No instances of legal consequences of such an editorial were offered or discussed by the audience.
Are there other mechanisms of enforcement other than universities or governmental agencies?
— Probably one of the better resources is a national council on publication ethics, such as the Council on Publication Ethics
of the United Kingdom. This voluntary group of editors reviews and discusses cases involving publication ethics, and then publishes a written summary of their recommendations, which is posted on the Web. However, cases are limited to those from the UK, and few other countries have such resources. The Ethics Committee of WAME strives to provide a similar service for editors in other countries. The advantages of such councils is that they probably have fewer conflicts of interest in their review than other organizations and their review is often faster. Cases can be posted anonymously on the WAME list and informal individual responses are available within days. On the other hand, all the participants are volunteers, and thus the amount of time devoted to review may be restricted.
Are there other, more reliable mechanisms for discovering duplicate submission than this chance one of having the same reviewer on the same paper for both journals? — Obviously all editors would like to be able to reliably detect such submissions, but no more effective mechanism was proposed by the discussants. A MEDLINE search of all submitted manuscripts would be time-consuming and might detect a paper already published elsewhere, but could not detect duplicate submissions.