We received a paper describing the results of an analysis of pathogen gene sequences from patients who had been given an investigational drug as treatment for their infection. The study had been done in Europe. One reviewer said that the paper did not explain whether the patients had been treated in the context of a trial or not and that no information about study sponsorship, ethical approval or patient consent had been included.
The reviewer was concerned that an attempt was being made to publish data without the approval of the trial sponsors. The editors rejected the paper but recommended that the authors submit it to a sister journal at the same publishing company, but that they would need to address all of the points raised. The paper was then submitted separately to the sister journal but without any response to the reviewer’s points. The authors eventually explained that the patients in this study had been enrolled in an expanded access program for the investigational drug, which was underwritten by governmental sources. This was not regarded as a trial, and the authors explained that gene sequencing was carried out as part of routine clinical care, following national treatment guidelines.
The editors felt that there were insufficient grounds for them to follow-up further, and decided to withdraw the paper from consideration at the sister journal, recommending that the authors describe the context for the study more fully in their paper when they prepare it for publication elsewhere.
Should we have done more; is it reasonable that ethical approval and informed consent would not be needed for this study?
The Forum commented that this is probably an example of different rules for ethical approval in different countries. This would seem to be the issue here. The Forum agreed that no further action was required and that the matter should not be pursued.
The journal followed the COPE Forum’s advice and did not take further action.
We published a paper in the journal which reported on microarray expression profiling of cell lines from a specific type of cancer (not named here, to preserve anonymity). That paper suggested that a particular compound might affect the function of a protein expressed in the cell lines; this compound could therefore be a possible candidate drug for use in this type of cancer, which might be explored further in clinical trials. (This compound currently has approved uses for other cancers.)
A few months after publication, we received a short electronic letter from a researcher who said that as a result of reading the paper and other data, the researcher and his colleagues administered the candidate drug to three patients with the specific type of cancer mentioned above, who had not responded to other therapies. The researchers reported that one individual improved transiently but ultimately relapsed and that all three individuals eventually experienced a progression of their disease.
The authors did not say whether they had obtained ethical approval for this experimental therapy or whether the patients gave informed consent. We wrote to the authors to say that we could not offer to publish the electronic letter. We encouraged them to rewrite and submit this as a formal report of a clinical trial for our sister journal, but providing they also document ethical approval for the study and informed consent procedures. We never heard back from the authors.
What should we do now?
Although most of the Forum felt there was not much that could be done, some felt that this was an issue of informed consent and the editor had a duty to find out if consent was obtained. The advice was to contact the authors and ask them to provide information on whether or not consent or ethics approval had been obtained, with a deadline for an answer. If no or no satisfactory answer is received, the Forum suggested contacting the authors’ institute and asking them to investigate the issue. No allegation should be made and the authors should be informed of this course of action.
The journal wrote to the authors to ask for clarification regarding individual consent for the course of treatment, and ethical approval for the study. The authors replied to say that patients did give written consent, but no ethical approval was required or given because the course of these patients’ treatment was not carried out as a research study but instead as a course of off-label therapy, which was subsequently written up into a research paper. No further action was taken.
A paper was submitted to us describing an RCT carried out in a Far Eastern hospital. Soon after the manuscript had been sent out for review, one of the reviewers sent a letter alerting us to a “possible case of fraud”. The reviewer in question appears to have compared notes with another investigator in his institute, and together they realised that the same group had submitted two similar trials to two different journals (without enclosing the related paper in their submission to us).
There is ambiguity here, but we believe that the allegation is one of fraud (Do the patients really exist? Are the two groups of patients overlapping?) rather than duplicate publication.
We challenged the authors, asking them to send us the related manuscript and protocol. This they did very willingly, and they sent files of primary data too.
The authors then withdrew the related paper submitted elsewhere. Although this seems suspicious, it seems that the authors thought we would be less favourably inclined to the paper they had submitted to us if a related paper from them was accepted by another journal first.
Review of the paper submitted to us proceeded, although the individual who had alleged fraud did not send a report.
Sufficient technical criticisms were raised to justify rejecting the paper, in our opinion. However, as far as we know, independently, one of the other referees commented “somewhat puzzling are the exceptionally well balanced, perfect patient characteristics in table 1”.
We have now rejected the paper in question. We would like COPE’s advice on what to do further—we are open to persuasion, but are not sure that there is sufficient evidence of wrongdoing to ask for an institutional investigation. We have not yet responded to the reviewer who raised the initial allegation of fraud.
The Forum questioned the idea of requesting an institutional investigation in view of the fact that there is no clear evidence of wrongdoing. Also, would the investigation be conducted properly and would the authors receive a fair trial? Others questioned whether it would be a sensible use of editorial resources. The Forum suggested that as the authors are aware that the editor has raised suspicions regarding the paper, that may be sufficient warning for them. Other advice was to contact the reviewer explaining that it is not good practice to compare notes with other investigators as such material should be treated as confidential.
The paper was rejected and the editor decided not to try to initiate an institutional investigation owing to insufficient evidence of wrongdoing. Later the editor heard from the refereeing “whistleblower” (who had written to the editor to allege “possible fraud”) that the same paper had been accepted at another journal, and that he had been invited to write an editorial. A couple of weeks ago, the editor heard from the same “whistleblower” that the other journal rejected the paper and were considering an investigation. The journal is no longer actively involved in pursuing this matter.
A general medical journal received an RCT from a seldom-published, single-author, in an eastern European country. The results were striking, with an effect size that surpassed that of established medications for this condition, so the manuscript was sent for peer review. One reviewer commented that the results were “so highly statistically significant it is almost too good to be true. Virtually every parameter was observed to be statistically superior in the [intervention] group”.
On closer inspection, a number of concerns arose regarding recruitment, inconsistencies and reporting.
The author denied fabrication, but the editors were not satisfied with his explanations. The manuscript was rejected and the journal approached the institution’s ethics committee, and then the director of the institution, who undertook an investigation.
The internal investigation reported that the author looked after many patients with the condition, some of whom were contacted and affirmed that the intervention was beneficial. The statistical analysis was confirmed by a university in the USA. The investigators believed that the conclusions were true and that there was no evidence of misconduct.
Does COPE have any comments or suggestions regarding the journal’s actions?
The Forum felt that as there was no evidence of misconduct and the editor had pursued all of the usual channels of investigation, there was little more that could be done. Some suggested obtaining the raw data and seeking expert advice from a statistician, although that may not be possible as the journal rejected the paper. All agreed that the editor had taken all of the appropriate steps and possibly gone as far as he could with this case.
At COPE’s sensible suggestion, the editor requested a copy of the external report vindicating the analysis, but this was not available. Although not satisfactory, the editor considers the case now closed.
In October 2005, our journal commissioned a review on a specific topic from an expert in that field. The commission was accepted, and a submission date set. This was followed up and renegotiated several times over the coming months. A first draft was received in January 2007, and assessed inhouse for feedback to the authors. Some suggestions were made regarding structure and content, and the manuscript was officially submitted in August 2007. As considerable time had elapsed since the original commission, a check was made to ensure that the review was still timely. A publication was noticed in May 2006 by the author, in another journal, with the same co-author on both. This was further found to have a near-identical abstract, and a copy was then requested. When received, a side-by-side comparison found the manuscripts to be approximately 90% the same, with entire sections duplicated. Differences comprised slightly poorer language usage in the recent submission, some single sentence updates, one extra paragraph in one section and some changes to the conclusions. Some sections from the published review had been deleted to fit our shorter word limits. No reference to the earlier publication had been made at any stage. A retrospective check showed that the first draft had been even more similar to the published version.
Both authors were contacted to explain:
- why a near-identical manuscript had been submitted;
- how this situation had happened, given the timeline and date of original commission.
In reply, the authors stated that the organisation of the manuscripts was different and similarities inevitable. They stated that they had hoped for reviewer feedback on how to change the structure and wording. They asked to have a revised manuscript reconsidered.
Does this constitute an attempt at dual publication? If so, what further action should we take?
The Forum agreed that this was an attempt at duplicate publication. The authors should be contacted and told that their behaviour was unacceptable. It should be explained to them that if the editor had published the review, copyright may have been infringed as essentially the same article would have been published in another journal, thereby putting the journal in a very difficult position. Some suggested that in the future, the commissioning letter should state that there should be no significant overlap of information between any two papers, and perhaps this should also be added to the instructions to authors. There was disagreement as to whether the first journal should be contacted, with some arguing that it should, while others suggested that as this was an attempt and not a breach of ethics it was not necessary to inform them.
Further to the discussion at COPE, we checked the copyright on the initial publication, and found that copyright would indeed have been breached had we published our manuscript. The authors were contacted to be advised that we would not be considering the manuscript further, that they would have been in breach of copyright had we proceeded, and quoting from our original commissioning letter, which flagged the unacceptable nature of copying sentences and paragraphs from published papers whether the authors own or by others. We then heard back from the authors stating they understood our position and thanking us. We also decided not to contact the journal in which the article had been published.
The journal submitting this case to COPE sent a paper [paper 1] to a reviewer who wrote this in the review: “…That apart, this manuscript seems to be another report of the already published **** trial, looking at the data from a slightly different angle. I am not convinced, however, that the data is worthy of so many submissions.”
And, in a separate email to the handling editor: “Just by chance, I have already reviewed a paper [paper 2] by the same group involving the same study for xxxx journal recently. I do not know the outcome of the refereeing process at that journal, but it does seem to me that the two papers are similar in many respects, and too similar to be both published. I have taken the unusual step of attaching the paper [paper 2] I was asked to review by that journal so you can decide whether or not you really wish me to comment on the one submitted to you. If you think this is "inappropriate", just ignore the attachment, let me know, and I will review your paper tomorrow. Sorry for this convoluted message, but I thought you ought to be aware of the situation.”
The handling editor felt that paper 2 did not overlap too much and when the editorial team discussed paper 1, paper 2 was included in the pack of reading material and read by all or most of the editors, the external editorial adviser and the statistician.
At the meeting, the team discussed the ethical problem raised by this and decided that:
The reviewer should have said “I know about this other paper - would you like to see it (not “here it is, tear it up if you like”)?”
The editors should have contacted the authors and said “the reviewer’s told us you have another similar paper - you should have mentioned this in your cover letter, can you tell us about it now?”
The editors should not have read paper 2 without the author’s permission because it was being considered in confidence at another journal.
Outcome so far:
The editorial team discussed paper 1 on its merits and rejected it because the research question was only indirectly answered with an over technical analysis and because the paper did not add enough to previously published work, including the author’s own.
The editors did not mention to the authors that they had seen paper 2.
The editors agreed to ask COPE’s advice on whether to take the ethical problem further.
Questions for COPE:
(1) Should the editors tell the authors all the above now, apologising, and explaining again that there were standalone reasons for rejecting paper 1?
(2) Doing so would unblind the reviewer of paper 2: this journal uses open (signed) peer review but the other journal doesn’t. Should the editors seek the reviewer’s permission before contacting the authors?
As there were standalone reasons for rejecting paper 1, the Forum agreed that contacting the authors would serve no purpose. The advice was to contact the reviewer and explain that he should not have sent paper 2 to the editor, breaching confidentiality. The reviewer should have raised the issue with the editor stating that he had concerns regarding the paper but should not have shared confidential information. All agreed that the reviewer should be made aware of his mistake so as to prevent the occurrence of such an incident in the future.
The editor concerned has found this a useful learning experience. The reviewer was contacted (very tactfully but making it clear that he too had slipped up). No reply has been received to date.
Author A was paid to facilitate a meeting and write a meeting report for internal purposes. He was paid to do this by author C’s company. The report was posted as a PDF on author C’s company website. No authors were listed on the report.
Authors B, C and D co-authored an article that has been published in a journal supplement. It later transpired that the main substance of this journal article had been lifted directly from the internal report written by author A. This misdemeanour was reported to the journal editor by author A, who works at a different institution to authors B, C and D (author B is a junior researcher).
When authors B, C and D were alerted to this problem, it transpires that they misunderstood copyright and thought that, as the meeting report “belonged” to their company, they were entitled to re-use the content without permission. They proposed a resolution whereby author A was included as an author of the journal article via an erratum.
This erratum was published, but author A remains discontent and has made it clear that he would like the journal article to be retracted. Authors B, C and D have yet to be consulted on retraction, although they have made every effort thus far to comply with author A. Author A believes that there is still merit in a paper bringing together the data discussed in the offending article, but with correct attribution and incorporation of some data from author B’s PhD thesis. This article would be written by author B alone, with acknowledgement of the help from authors A, B and D.
If all authors concerned agree to this, is it a suitable resolution?
The view of the Forum was that there are no grounds for retraction of the journal article. Retraction of an article on the basis of incorrect ordering of the names of the authors is inappropriate. It was suggested that the editor should have considered obtaining the agreement of all of the authors as to the order of the names on the paper before publishing the erratum. The advice was to try and obtain an agreed statement from the authors and then publish an erratum of the erratum, which would resolve the issue. Alternatively, some suggested taking no further action would be appropriate. All agreed that author’s A suggestion of publishing another paper bringing together the data discussed in the offending article, but with correct attribution and incorporation of some data from author B’s PhD thesis, was not a reasonable request.
The editor retracted the article against COPE’s recommendations. The editor now considers the matter to be closed.
In February 2007, author A and a colleague submitted a paper (paper A) to our journal, which uses double-blinded peer review. We sent paper A for external review.
Four weeks later, group B submitted a paper (paper B). The editorial office sent paper B to external reviewers, one of whom was author A. Both groups of authors are known to us and well regarded within our discipline. The reviews on both papers were received and we were on the point of writing to both groups asking for revisions when author A emailed to say that he was concerned about similarities between the two papers and felt that this was more than a coincidence. He believed that the method he had used, in the context he had applied it, was entirely original and that the authors of paper B must have had prior sight of his paper.
His case was that paper A reported a study that had been presented at a conference in April 2007. He thought that one or more of the authors of paper B might have reviewed his paper for the conference the previous November, and had used it to design an identical study. He claimed that paper B showed evidence of having been “rushed into submission”.
We contacted the conference organisers to ask if they could reveal the names of the reviewers of author A’s paper, but they refused. Our editors could find no evidence of plagiarism, except they did agree that the application of the method in this particular context was, to quote one of our editors, “striking” in its similarity.
Group B responded robustly to our request for further details. They said that although the method was new within our discipline, it was widely reported in other literatures. They produced papers (including one they had published in our own journal several years previously) to show that they had a reasonable prior claim to the use of the method, since they had been using it for some time. They argued that it was only the application to this particular context, and not the method itself, that was original.
We felt that this response was reasonable and credible. Faced with this apparent deadlock we took the view that misconduct had not been proved. Furthermore, trying to establish which group had prior claim to an idea that was already largely in the public domain would be next to impossible. We were unwilling to report any further accusations without real supporting evidence.
We therefore felt that the only course of action open to us was to level the playing field, since to do otherwise would be to rule in favour of one or other group. Accordingly, we wrote to both groups of authors and told them that:
In the absence of any evidence to the contrary, we had no choice but to assume that both papers had been written in good faith and that it was an example of coincidence;
We were going to publish both papers side by side once satisfactory revisions had been received (we publish dates submitted, reviewed and accepted in the journal, which would give evidence that group A had submitted first);
As group A had had the slight advantage of seeing an early draft of paper B, we felt it only fair to let group B have sight of paper A. No further exchanges between the groups would be facilitated thereafter.
We planned to discuss the matter with COPE and would report to both groups.
Author A responded to say that he felt this was unfair. He did not want his manuscript to be sent to group B since his paper was the better of the two—it had received more positive reviews—and that he himself had not gained anything by seeing group B’s manuscript. He felt that he had lost the advantage of having submitted first and that his paper should be published first.
We have three questions, therefore:
How should we now respond to author A?
Have any members of COPE experienced anything like this before and, if so, how did they resolve the matter?
And, on a more general point, how far should editors reasonably go in exploring a case like this, when it is a matter of one reputable group’s word against that of another, and there is little or no evidence either way?
The general consensus was that the editor’s course of action was reasonable, except for point (c) above—that is, most agreed that group B should not have had sight of paper A. On a general note, the Forum emphasised the fact that it is not the job of editors to adjudicate in such matters. If there is no evidence of misconduct, the authors should be advised to sort the matter out among themselves. One viewpoint was that, in retrospect, the editor should perhaps have allowed the papers to follow the usual editorial procedures, without interference, and should not have shared the information with the authors. Some argued that paper A should be published first as it was submitted first, but others suggested that from the reader’s point of view it would be more interesting for the two papers to be published together.
Both papers are likely to be accepted and once the final decisions have been made, we will send all of the documentation to our independent Quality Standards Advisory Group for a report that will be shared with the authors. The COPE report, with your permission, will form a valuable part of the documentation
The journal received a paper on an imaging technique which reported changes in a normal healthy volunteer. No adverse events were experienced. This study demonstrated the feasibility and safety of the imaging studies in a particular condition.
The paper was rejected on methodological and priority grounds but during the review process it was noted that no specific mention was made of ethics approval, despite the fact that an invasive procedure was carried out. It was, however, noted that the control subject was a member of the department and one of the authors.
The editor raised this with the corresponding author who responded that the study was done as part of a programme of work in both animals and humans which had ethics approval. There was, however, concern that were the subject a junior member of the department there would have been implicit coercion. In view of that, the editor requested a copy of the ethics approval. The issue was discussed at that time with the BMJ Ethics Committee and the course of action was thought appropriate.
The editor has subsequently received a translated letter from the local ethics committee approving the study but this has been dated recently (that is, since the issue was raised). However, it has also been explained that the subject was in fact the head of department.
The editor is of the view that the study should have had ethics approval but many studies have been performed by investigators upon themselves and the editor intended not to take this further. Is this the right course of action?
This case raised the issue of whether ethics approval is required to perform an invasive procedure on yourself. Most thought that in reality most individuals do not seek ethics approval and there have been many anecdotal instances of such cases in the past. As it transpired that the subject in question was the head of department and not a junior member of the department, and thus the issue of coercion was ruled out, the advice was to take no further action.
A doctor who trained in country A took the licensing examinations in country B because he wished to work in country B.
After the examinations, he carried out a survey (with a very poor response rate) of other doctors who had taken the same national licensing examination. The survey asked four major questions: How representative was the examination relative to the scope of your reading? How fair do you think this method of evaluation is to foreign medical graduates? What would be your most ideal method for evaluation of foreign medical graduates? Why did you immigrate?'
There was no IRB approval for this study and no mention of written informed consent. When asked to clarify ethical issues, the author stated: “No protocol was prewritten, as the question of what and who would constitute an IRB for such a survey on issues relating to foreign medical graduates in the transition was a difficult one to answer. Thus no ethical review was done. However, all participants were adults, of sound mind, and implied consent to have their opinions published.”
The paper has been rejected, but do the editors need to take further action? The risk of harm to the survey participants (all doctors) from this study seems low. Nevertheless, the survey might conceivably have included questions of a personal nature that could have caused distress (eg, reasons for immigration). The licensing board was not informed about the survey. If we report this doctor to the licensing board, it may well affect the author's chances of employment in country B.
As the subjects were not patients, some argued that it is not necessary to have ethics approval although others believed that ethics approval should always be sought for any study on human subjects. As the study involved a survey, all agreed that written consent was not required as the subjects could not be identified. However, the general view was that the author should have at least contacted the licensing board to ask permission to carry out the survey. The advice from the Forum was to write a strong letter to the author explaining the ethics involved in carrying out any research and suggesting that he should have contacted the licensing board before carrying out the survey.
As recommended by COPE, the editor wrote a stern letter to the author. In reply, the author thanked the editor and also noted that he himself was confused as to which institution should have been approached ahead of the study. The editor considers the case now closed.