Session on retractions at the European Seminar 2019, chaired by Heather Tierney, COPE Council, with a publisher's perspectives from Catriona Fennell, Director of Publishing Services, Elsevier. During the session we also heard from Howard Browman and Thed Van Leeuwen. Links to their presentations are below:
Journal A received a number of concerns from a reader regarding a paper published in the journal. These concerns were reviewed and sent to the authors of a paper, along with additional comments from the editorial board. The concern was largely around retrospective registration, and an inconsistency between the trial registry record and the published paper. An editorial board member conducted a full comparison of the trial registry entry with the paper.
The authors have admitted honest error with full explanations. The editor-in-chief has asked for confirmation that all authors and institution are aware and outlined options for next steps. The suggested options for next steps from the editor-in-chief are: (a) retraction of the paper; (b) substantial corrections and explicit declaration of the flaws of the trial procedures and protocol violations and selective and misleading reporting; which may well render the trial invalid or at least biased, and then providing a better and corrected summary table and narrative of what can be legitimately said. This is not ideal and will regrettably give the impression of insufficient rigour in the execution of a trial and the data still being in the public domain, although a more confident statement of a negative trial is better than selective reporting of some positive findings; (c) or we invite retrospective critique and commentary on trial and trials in general when reported to be invalid or flawed; this is an important educative role, but does not remedy that the trial data are in the public domain and are misleading.
Again, the authors offered an apology claiming honest error and preference for the article not to be retracted. They have offered to publish a correspondence letter to explain the registration issues in due course or correct any inconsistent sections according to the review comments and registry information. The journal is now questioning the next course of action: retraction, corrigendum and/or an editorial outlining the issue.
Question(s) for the COPE Forum
• The Editorial Board were initially considering retraction but are now considering publishing a narrative/editorial of the issues for transparency, confirming the journal’s current/new policy of requiring prospective registration and an explanation of any changes in protocol in the methods section. Should this accompany a corrigendum?
• Ethics approval: approved in April 2011, but the protocol states study execution time is August 2010 to July 2013. The authors state that the first patient was referred in May 2011. Does this need further explanation?
• Should the editorial board consider retraction?
• Are there any other actions the board should consider?
The Forum asked if the journal had contacted the institution and if there was an investigation in progress. The editor informed the Forum that the journal has asked for confirmation that all of the authors and the institution are aware of the issues, but no response has been received to date. The authors have stated that the institution was not aware of the need for prospective registration. The editor may like to pursue the institution for more information.
The Forum suggested publishing an editorial note on the paper or, if the institution agrees to undertake an investigation, publishing an expression of concern. As there seems to be no institutional oversight, perhaps the editor should give the authors the benefit of the doubt. This could be an important educational opportunity, to educate the authors regarding trial registration; although now an international standard, many authors do not know about prospective registration. Hence a lengthy corrigendum and an editorial highlighting the issues would be appropriate.
The Council of Science Editors has a lengthy section on their website about correcting the literature with samples of actual corrections (https://www.councilscienceeditors.org/resource-library/editorial-policie...). In general, the correction options are errata, corrigenda, expressions of concern, and retractions, although some of the wording is nuanced in ways that might be helpful in this situation. The editorial note referenced above seems to fall under the category of an Editorial Expression of Concern. The National Library of Medicine has a fact sheet (https://www.nlm.nih.gov/pubs/factsheets/errata.html) detailing the types of corrections that can be linked to an article. This list includes “Comments” which could be used to link commentary from the authors as well as from editors.
Some journals ask for the full protocol to be submitted to the journal along with the article. The journal then checks the protocol against the paper before the paper is peer-reviewed. The authors are asked to explain any deviations from the protocol. The editor may wish to consider this approach to avoid similar situations in the future.
The Senior Editorial Committee reviewed the advice from the COPE Forum and agreed with the suggested actions. The authors have drafted a Corrigendum, and this is currently undergoing review and revision by a board member with expertise in the field. Members of the editorial board are currently drafting an editorial for publication alongside the Corrigendum. An Expression of Concern will be published shortly to alert authors while the Corrigendum and editorial are drafted.
Advice on follow up:
Follow-up (May 2019)
The authors requested changes to the corrigendum which could not be accommodated. On this basis the Senior Editorial Committee decided to retract the paper. The retraction notice and editorial have been drafted and will be published online shortly.
The first author of a paper rejected by our journal publicly identified one of the four peer reviewers for the paper by name. She did this during a media interview conducted after the paper was published by another journal. The first author implied in that interview and subsequently on Twitter that the paper was rejected because of that person's review and also claimed the reviewer did not reveal relevant COIs.
This complaint received a great deal of attention because the rejected paper had a direct bearing on a very bitter medical/political matter, and its results were felt to bolster the case for one faction. The authors had not lodged a formal complaint with our journal about this matter. We usually do not comment on papers that we do not publish, so when contacted by the press about this accusation our initial response was to "neither confirm nor deny" that we had considered the paper. It soon became clear, however, that the reviewer was the subject of much unpleasant comment on social media and other vindictive behaviour. A colleague of the reviewer, for example, tweeted that he was "ashamed" to be a professor in the same institution as the reviewer. The reviewer also received two freedom of information requests asking for any correspondence with the journal and anybody else concerning the rejected paper.
What the journal did: 1. We immediately contacted the authors to let them know we were disappointed in their behaviour. The authors acknowledged their mistake and had already contacted the reviewer to apologise. The reviewer accepted the apology but expressed the hope that we would make his review public, saying "I am continuing to get emails from people who are assuming that I wrote negative reviews for the paper and raising questions about conflict of interest. I believe my reviews to have been supportive of publication and not to comment on whether the journal should accept or not. So from my point of view it would be helpful if you could publish my reviews regardless of whether the other reviewers agree to this..." 2. The other reviewers and the authors agreed that we could make this matter public, so we broke with precedent and published a blog. We received mostly positive comments on Twitter and in the comments section for the blog. 3. We have also amended our instructions for authors on the journal website to say "For rejected research papers, we expect that authors will keep the identity and comments of peer reviewers confidential. They may, however, share the peer review comments (although not peer reviewer names) in confidence with other journals. Authors should contact the editor who handled their paper if they have any complaints about the peer review process or the behaviour of the peer reviewers." 4. We have also amended our rejection letters to say "Although the journal has an open peer review process, in which authors know who the peer reviewers were, we expect that you will keep the identity and comments of the peer reviewers for this paper confidential. You may, however, share the peer review comments in confidence (although not the names of the peer reviewers) with other journals to which you submit the paper. If you have any complaints about the peer review process or the conduct of the peer reviewers, please contact the editor who handled your paper. Please do not contact the peer reviewers directly." 5. We continue to follow-up periodically with the reviewer to make sure he is not suffering any additional ill effects from this incident. 6. We are submitting this case to COPE and will also be referring it to the journal's internal ethics committee. The matter now seems to have died down, but it raises many questions.
Question(s) for the COPE Forum • Should the journal have handled this differently? • Should the journal formally notify the first author's institution about her behaviour? • Are the additional instructions to authors on our website and in rejection letters adequate? Should we do anything else to prevent this from happening again? • Should peer reviewers who do research in the same field as authors of a paper declare this as a COI? If so, when? Almost all reviewers are chosen because they have expertise in the same field, and commonly their conclusions have differed in some respects from those of the authors. At present very few reviewers list this among their COIs.
The Forum agreed that the journal did a good job here, and has taken reasonable steps to change their process and avoid a similar situation in future
The Forum discussed whether the author was under pressure by a media interview and gave a comment afterwards for which they later apologised, or was it deliberate on the part of the author as the paper was published in another journal and this was an “attack” on the reviewer for the journal that rejected the paper. The Forum was ambivalent on whether the first author's institution should be contacted. It is possible that the institution is already aware of the case (because of the media coverage) but the institution could be contacted in neutral terms although it is unlikely that the journal could expect much action from them.
Regarding conflicts of interest, being in the same field is not in itself a conflict of interest—in fact it is usually a reason to pick a reviewer. Because experts in the same field have interests that are similar, they may unfairly be perceived to have a conflict of interest. However, sometimes reviewers do have conflicts of interest so it may be helpful to include instructions with some clarifying exemplars to help reviewers to identify conflicts of interest. For example, if a researcher has built a career on a particular view and are ‘famous’ for holding that view, that could be a conflict of interest. The advice was to ensure that the journal’s guidelines to reviewers regarding conflicts of interest are up to date. The COPE discussion document may be helpful in this regard (http://publicationethics.org/files/u7140/Discussion_document__on_handling_competing_interests.pdf)
As managing editor, I view all manuscripts before they are assigned to an editor. Within a 4 week period, I have detected five manuscripts where photographs of either gels or plant materials were used twice or three times in the same manuscript. These manuscripts were immediately rejected.
However, we are not convinced that these are cases of deliberate misleading of the scientific community. It rather seems to us that many laboratories consider photographs as illustrations that can be manipulated, and not as original data. Thus gels are often cleaned of impurities, bands are cut out and photographs of plant material only serve to show what the authors want to demonstrate, and the material does not necessarily originate from the experiment in question.
When the editor-in-chief rejected such a manuscript, a typical response was: “I am surprised by the question and problem you pointed out in our manuscript. I checked the pictures you mentioned and I agree that they are really identical. But please be reminded that the purpose of these gel pictures was only to show the different types of banding pattern, and the gels of a few specific types were not very clear, so my PhD student repeatedly used the clearer ones. This misleading usage does not have an influence on data statistics or the final conclusion”.
Question(s) for the COPE Forum
What can be done to ensure that all gels and all photographs originate from the experiment and that they should never be tampered with?
How can the scientific community of some particular countries be taught correct scientific publishing standards?
The advice from the Forum was for journals to strengthen their guidance on this issue. Journals should provide clear guidelines in their instructions to authors on what is acceptable. Original data, such as gels, should not be used as illustrations, without an explanation. Line drawings, for example, can be used to illustrate a point, but if original data are being used just to illustrate a point, this should be accompanied by a very clear statement in the manuscript telling the reader this and explaining what is being demonstrated.
The Forum advised looking at the guidance published in other journals. Some journals have very good guidance on this issue and editors may wish to incorporate such guidance into their instructions to authors (with full attribution and after obtaining permission).
On a poll of the Forum, less than half of the delegates said that they screen for image manipulation. COPE has an eLearning module on image manipulation. Also, Mike Rosner has written extensively on this topic (Journal of Cell Biology 2004;166:11–15 http://jcb.rupress.org/content/166/1/11.full)
We were contacted by a reader who told us that he had spotted a number of cases of image duplication and mislabelling of fluorescent tags that had occurred over the past 4 years. These involved two papers published in our journal, and two other papers published in two different journals. The two papers in our journal were both reviews, and the one that had the most occurrences involved a poster (associated with the review) that we had recently published. Although each paper had different authors, there was one author, author X, common to all the papers. The reader provided extensive evidence.
I checked all the evidence myself, looking up the original sources and concurred with the reader that there was at least image duplication (I could not verify myself whether mislabelling had occurred). I also discussed the case with our publisher. I then contacted the head of author X’s institute, copying in author X and his co-authors on the poster article. I told the institute head that I was making no assumptions about wrongdoing but presented the evidence and asked for an explanation.
The matter was referred by the institute head to Dr Y, the Associate Vice President for Research of the university, who appointed a Committee of Inquiry. This committee found reason to launch a full investigation, and so an Investigatory Committee was appointed. I was told that I would receive their report.
While I was waiting for their decision, the editors of the two other journals in question approached me to ask that I keep them informed. I told them about the Investigatory Committee and recommended that they contact Dr Y if they wanted to be kept abreast of the outcome.
Almost 2 months later, I received an email from author X, copied to Dr Y, with an explanation of what had happened. He could not provide an explanation for the error in the first paper. There was one image in that paper that claimed to show something labelled with a particular fluorescent tag, but a different, although similar, tag was in fact used. He said he spoke with his co-authors, who agreed that the error did not affect the scientific conclusions of the paper and that he could either correct the tag name or provide a new correct image.
For the poster, he said that there were a number of images mislabelled, and that this occurred, in essence, because he used ‘placeholder’ images while creating the poster but forgot to replace them with the correct images. He offered to redo the poster with the correct images and also wanted to replace some other images that were correctly labelled because he had ‘better versions’ of them. He assured me that the text did not need to be changed and again said that his co-authors agreed that the scientific conclusions of the paper were unaffected.
After again conferring with our publisher, we have come up with a plan:
I have contacted Dr Y and asked him to verify that the Investigatory Committee has accepted author X’s explanation and found no evidence of fraudulent intention.
I have asked author X to have his co-authors on both papers contact me directly to confirm that the scientific conclusions of the papers were not compromised and that they are satisfied with the replacement images author X is proposing.
I will ask author X to provide a replacement image for the first article.
I have told author X that it is not acceptable to replace correctly labelled images with ‘better versions’ and that we will only deal with those that are incorrectly labelled.
I plan to issue corrections for both papers. On the poster article, because it involved 10 images, I also plan to include, with the correction, a statement to the effect that a committee was appointed by author X’s institute to investigate the mislabelling and that they found no evidence of malicious intent (I’m wary of the wording I use here).
I also want to reprint the poster and send a copy of it to each of our print subscribers as they will have received a copy of the incorrect version. I am talking to our online hosts about how we can provide a link to the corrected version of the poster because, although I feel strongly that the original should remain online as it is, a correct version should be available.
I will ask author X to cover all costs associated with the redesign, printing and mailing of the poster, in addition to the costs of the corrections themselves.
I am not convinced by author X’s explanation and did look to see whether there were any other published corrections associated with author X’s previous publications but did not find any. If the Investigatory Committee confirms that they did not find evidence of fraudulent intent, however, then I feel I have to accept that decision and will proceed according to our plan outlined above. I would very much appreciate COPE’s advice on how we have handled the situation so far, and whether our plan of action could be/have been improved.
The Fourm agreed that the editor had done all he could and had handled the case well. The institution has investigated and found no fraudulent intent. The editor should publish corrections, stating the facts and avoiding accusing any of the authors, and then let readers draw their own conclusions. Regarding the whistleblower, the editor does not have a duty to keep him/her informed of all of the details of the case. The whistleblower cannot expect to be involved in the case. The whistleblower can contact the institution if they want. The editor should keep correspondence with the whistleblower as formal as possible, reply only to direct questions and not involve him/her in the investigation.
The editor followed the plan that he outlined, taking into consideration the advice received from the Forum. Corrections were published for both of the articles involved, with agreement from all co-authors, that stated the facts; a link to a corrected version of the poster was provided (keeping the original in place as published) and print copies were sent to all print subscribers (all costs covered by the author). The whistleblower was contacted only to confirm that corrections had been published. The editor considers the case to be resolved.
We published a randomised trial by six authors. Some years later, we received a letter from a researcher who had been looking into the trial in the context of a meta-analysis. She noted “implausibilities of serious concern”, including “a highly unusual balance in the distribution of baseline characteristics”, 95% CIs that were non-symmetrical about the effect estimate, and use of a stratification variable the value of which could not have been known in all patients if the trial was conducted in the way reported.
We asked the corresponding author to write a letter of response, which he eventually supplied a few months later. Owing to the author’s poor English and the level of statistical knowledge needed to assess the response, we sent the exchange of letters to a statistical reviewer. The reviewer said that the letter of concern was “completely correct” in everything it said and that the author’s explanation for the unusual degree of balance in the covariates was “rubbish”, and that the 95% CIs had either been “doctored” or “incompetently estimated”.
In the meantime, an exchange of to-and-fro letters between a different researcher and the same author was published in another journal, relating to a paper reporting on a subset of the same trial data. We were alerted to this by the editor of that journal.
We sent the reviewer’s remarks to the author of our trial, who then consulted two independent statisticians of his own. He soon contacted us to say that, “surprisingly and regretfully”, these statisticians agreed that there were implausibilities and inconsistencies in the data, and asked for more time to investigate more fully. During this time, the author of the letter expressed concern that we had not made the possibility of these problems know to our readership, so we published her letter.
The author has now sent us a more comprehensive response, admitting that the randomisation process was not as described, the 95% CIs were all wrong (he supplied a recalculation), and the trial report had omitted some details of the protocol necessary for understanding it properly (now supplied). Our reviewer suspects that, given his free admission of all this, the author is probably incompetent rather than fraudulent, but that the extent of the incompetence could not give us confidence in any of the data. What now?
The Forum agreed with the editor’s opinion that the author is probably incompetent rather than fraudulent and should be given the opportunity to redeem himself. It was suggested that perhaps the paper should be submitted for review again. The Forum noted that this was probably a good internal learning exercise in that the statistical errors should have been picked up when the statistical review of the data was performed by the journal. A suggestion was made for the journal to set up a “sin bin”. Some journals operate a “sin bin” or “publication review committee” where once a year papers which readers or others have expressed serious doubts about post-publication are reviewed to determine whether or not it was “a mistake” to publish the paper.
However, some members of the Forum argued for stronger action and suggested contacting the author’s institution. But most agreed that the paper should be retracted as the research may be unethical.
August 2008 We have managed to find details of whom to contact regarding informing the author's institution, and the deputy editor has written to him. We await a reply with anticipation.
May 2009 The institute has responded to say that an investigation is under way and will take another couple of months to conclude.
November 2009 Following an internal investigation by the author’s institution, a number of serious problems were encountered, including: lack of ethics approval, lack of written consent, lack of treatment-allocation concealment and an inability to verify the authenticity of the data. We therefore retracted the paper on 10 October 2009.
One of the referees of our journal has brought to our attention a potential case of plagiarism.
The referee feels that the a manuscript submitted to our journal plagiarises an article published in another journal. The authors are from an institute in a far-eastern country.
We would be grateful if COPE could provide an opinion on this issue, as well as advice on what would be the best course of action.
The Forum was informed that one paragraph in the introduction and one other sentence had obviously been cut and pasted from another (cited) paper but that the work itself was not duplicated. The advice from the Forum was to write an educative letter to the authors pointing out that if they directly quote from another paper they should either place the words between quotation marks and state who had written them or report them indirectly citing the author. The Forum did not consider it was necessary to go any further.
The editor-in-chief wrote to the authors pointing out that quoted text should be placed between quotation marks, stating who has written the text, or alternatively, authors should create their own text to describe the situation and not use other individual’s text. This letter was sent to the corresponding author together with the decision letter. The authors have responded with a letter of apology, acknowledging their mistake.
The editors received a manuscript from a Far Eastern country ready to accept. The senior author (who has spent a lot of time in the West) was in the US when the editors asked for final signatures to be sent. The senior author instructed his team to collect and fax signatures while he was away and this was sent to the editors.
When the signatures were examined by the editors, it appeared that some of the signatures had been written by the same hand. The editors challenged the senior author who investigated immediately upon his return and a fax was sent:
“…The last copy was signed in my absence and not all signatures were signed by the authors personally but by other colleagues with oral permission from the relevant authors ……Some of the signatures may still look similar in writing style and this was due to the fact that some of the colleagues were not familiar to [sic] signing their names in English, but these were truly their signatures.”
The new batch of signatures show differences from the previous signatures and now appear to be signed by separate individuals.
This highlights a potential issue applicable to anyone not familiar with using non-European scripts.
Should the editors be worried about this? Should the explanation be accepted? If not, what is the next step?
Members of the Forum agreed that the editors should accept the explanation of the authors and believed that the case has probably arisen out of cultural differences. From an educational point of view, the editors could point out to the authors that it is not good practice for anyone to sign on behalf of an author. However, in the absence of any other evidence of misconduct, the Forum believed this was a minor discretion that could be overlooked.
Following the COPE meeting, the editor was greatly reassured and was happy to accept the signatures. The journal still insists on signatures but is now happy to accept signatures in the authors’ mother-tongue script.
An article by a Far Eastern group was published in our journal in November 2005. We were later alerted by an interested reader that the same article, slightly changed, was published in an American journal. I contacted the American journal and the article will now be officially retracted from that journal. Part of the explanation could be poor communication between the authors, but I am not sure this is the whole truth. Two of the authors accept guilt but are now asking that the others not be “punished”.
I would appreciate the committee’s advice on this matter as I have no experience regarding how to react to such situations. What is “the accepted way to react/punish”?
The committee agreed with the editor that it is unlikely that this was “an honest mistake” or misunderstanding on the part of the authors. It is most likely that the two papers were submitted for publication at the same time. The advice was to contact the author’s head of department informing him of the situation and asking him to consider investigating the case. Other advice offered was to contact the American journal and ask them if the authors had stated in a letter that the paper had not been published previously.
The editor received a very profuse apology from the authors who stated that the misunderstanding arose because of lack of communication between the authors. Those authors who thought that the paper had been rejected started a process with the American journal (where the paper was published 6 months later). The editor contacted the chief editor of the American journal and informed him of the situation. The article was officially retracted from the American journal. The editor has decided not to pursue the matter any further.
Dr X submitted a paper to a journal that was assigned by a rather hung-over editorial assistant to an associate editor who was a co-author on the paper. Realising the mistake, she emailed the associate editor to reassign the paper. He expressed surprise as he did not know Dr X, had not seen the paper before submission, and knew of no reason why he should be a co-author.
Dr X was asked to explain and account for all the co-authors’ accreditations. From his reply it was clear that he did not understand the requirements for authorship as the associate editor was listed as having been the inspiration for several of his papers.
A decision was taken to contact the head of department who also appeared on the paper as the final co-author, asking him to clarify the accreditations given by Dr X. The head of department replied:
“I share your concern regarding the proper use of the authorship credit. As the Chair of the Department, it is my responsibility to provide academic and professional guidance within the department for all of our students, including our post doctoral associates. This includes developing and disseminating an accurate understanding of what “authorship” means academically.
Upon reading your email I was at first quite surprised to see my name listed as a co-author. My second thought was that the questions that you raised with respect to co-authorship are related to the relative naivety of Dr X. He has only been in this country and at this university for the past three months.
Earlier today I met with both Professor Y, who is Dr X’s faculty sponsor, and Dr X. We discussed your concerns and questions in detail, as well as the greater implications of the ethical principals of authorship. During that meeting it became clear to both Dr X and myself that the inappropriate crediting of authorship was not intentional and instead represents a cultural misunderstanding. Dr X used the author title as an honorific with some of the persons he identified. It was his belief that such a practice was expected and condoned in this country.
Professor Y did not catch this error since Dr X submitted the manuscript to your journal without his review and input. Your email and the resulting meeting allowed me to fully explain this misunderstanding to Dr X in hopes that he does not make similar mistakes in the future. It also provided a mechanism by which Professor Y and I could explain the responsibilities of an author in properly submitting a manuscript for review and publication.”
The head of department asked Dr X to:
Withdraw the manuscript from consideration for publication while all the issues related to authorship are resolved.
Make clear that the work described was conducted in his country of origin and not at this university, should he resubmit.
Work with Professor Y to ensure that any future articles for publication meet the highest ethical and professional standards.
Is this case resolved?
This issue is worth writing about, but by someone who is not already involved in the case.
There could perhaps be three related articles: one investigative journalism piece; one on the science; and another on the data analysis.