A journal accepted a paper and then discovered that a very similar paper by the same authors had been published in another journal the previous year. The latter journal carried the piece as a short report while the former planned to publish it as a full paper.
The full paper carries extended research on oral and maxillofacial surgeons while the short report looks only at attitudes of dentists to the treatment of a particular condition. The sample size, methods and results for dentists are identical in each paper, although the emphasis in the two discussions is slightly different. Both papers have been carefully worded to avoid any wholesale duplication of text.
The short report was not mentioned in the reference list of the full paper.
The co-author of a paper has contacted us about a paper he published 5 years ago together with a researcher who has now been convicted of serious professional misconduct by the GMC for research misconduct.
The co-author is worried that the paper he co-authored may also be fraudulent.
The research was in two parts. The first was analysed by a doctor not convicted of research misconduct but the convicted doctor was responsible for the interviews and original data collection. The co-author has no doubt that this part of the project was carried out properly. The second part involved a follow-up telephone questionnaire, which was undertaken by the convicted doctor without prior knowledge of the co-author. The co-author has seen no questionnaire answer sheets to provide him with any evidence that the telephone interviews took place. The original data cannot now be found, but this may not indicate much as there have been many reorganisations of the institution concerned. The convicted doctor did provide a list of individuals he said he had called, and the co-author who has written to us undertook a follow-up feasibility survey to see how many patients recalled the telephone interview. Nobody did. Should we retract the article?
COPE advised the editor to retract the article.
The editor retracted the article with a statement, which attracted a write-up in the NewYork Times.
We have received a study in which patients with healed duodenal ulcers were randomly allocated to treatment with either placebo or ranitidine. Patients were also categorised as to whether they were type A or type B personality; the hypotheses being tested was that patients who were type A might be more likely to relapse. Patients did not have their H pylori status determined.
Subjects were withdrawn from the study if their duodenal ulcer relapsed.
The question is whether it is ethical to treat such patients with ranitidine or placebo rather than with H pylori eradication treatment. A gastroenterologist has advised me that he does not think it is.
The paper makes no mention of ethics committee approval or informed consent. I have rejected the paper on methodological grounds but have also written to the authors asking them to provide information on the ethical aspects of the study.
What should I do if the authors do not respond? What should I do if they respond with an unconvincing explanation-and it’s hard to see how they could respond with a convincing one?
The employers were notifed, as was the relevant medical association of that country, and the International Human Rights Commission. So far, only the Commission has replied, saying that it would investigate the issue.
An author submitted part of his PhD thesis as a paper. The section editor of the journal asked the PhD supervisor to review the paper. This induced a very heated response from the reviewer who made various claims regarding the paper:
The author does not credit one of the tests he uses in his work
There is no proper acknowledgement of co-workers who perhaps should have been co-authors (including the reviewer himself).
The manuscript is similar to others—one published with the reviewer’s name but without his consent
The author is taking credit for work done by others—most notably the reviewer
The author has refuted many, if not all, of the allegations
What should the editor do now? He has invited advice from the university who awarded the PhD, but is not sure what standpoint the university will take, particularly if there is any allegation of misconduct at a scientific level which can be supported by hard evidence. At the moment, the editor is inclined to go ahead with publication and call the reviewer’s bluff. However, this might be expensive in litigation.
COPE advises the editor to back off this one and let the university sort it out. The editor should certainly not publish the paper or do anything further until the university reaches a judgement.
The editor informed COPE that the reviewer who had complained about the author plagiarising his work had backed down. The paper was re-reviewed in the normal way.
The paper discussed the use of drug X in condition Y, submitted to journal A. It is a double blind randomised controlled trial, presenting the 1 year result in 129 women. It finds that drug X helps in condition Y. The authors published a similar paper in journal B, 2 months before submission of this paper to journal A. The journal B paper studied the same question in 601 women with a 2 year follow-up. The only new feature in the journal A paper is that all the women have a level of index greater than 1 standard deviation below the mean. The journal B study included those between -2.2 and +2.0 standard deviations. There was therefore some overlap of the inclusion criteria in the two trials. The journal A paper does not make explicitly clear whether the women described form part of a subgroup of the cohort discussed in journal B. In fact, they make only passing reference to that paper, but do not discuss its relation to the paper they are submitting to journal A. The authors did not supply a copy of the journal B paper when submitting the journal A paper. What should we do now?
The two sets of data overlap and the authors have not been explicit about this. The editor was advised to go back to the authors for an explanation and seek independent assessment of the degree of overlap.
A manuscript submitted to journal X was remarkably similar to a paper already published in journal Y. The similarities were noticed by one of the peer reviewers for journal X. The paper has been rejected by journal X but the editor has now written to each of the authors asking for an explanation and has told them that if a reasonable explanation is not forthcoming, she will inform the dean of the medical faculty from where the paper was submitted.
Does COPE think the editor has acted appropriately?
COPE was advised that the paper has been rejected. The editor wrote to each of the six authors threatening them with being reported to the dean if their explanations were unsatisfactory. The editor received five replies with four different excuses. The editor was advised to write to the dean of the medical school and include the authors’ replies to the dean too.
The editor wrote to the dean of the relevant medical school enclosing the copies of the ?ve replies. A year later no reply has been forthcoming.
We have accepted a systematic review for publication and have commissioned an accompanying commentary. The authors of the commentary noticed that a particular randomised controlled trial was included in the systematic review while a duplicate version of the trial, published in another journal, was excluded because of inadequate randomisation. The authors of the commentary pointed this out in their commentary. We showed the commentary to the authors of the review (as is our practice) and they said that they had excluded the duplicate version of the study, not because of inadequate randomisation, but because it was a duplicate. The authors of the review have thus changed their text. The authors of the commentary find this explanation hard to believe and want to include a sentence in their commentary, making the point that the review was changed in the penultimate draft. We think that it is unsatisfactory to publish the commentary in this way, because it leaves a serious accusation hanging in the air. Either we must accept that the authors of the systematic review made an error and not mention it in the commentary, or we must raise the possibility of research misconduct and ask the host institution to investigate. What does COPE think we should do?
The authors’ explanation is adequate; the commentary can only be published without the criticism.
The commentary was published without the criticism.
This case was described to me by an author who is about to submit a paper. He has discovered that a member of his team has produced a lot of fraudulent data for other studies, and has forged consent from ethics committees. This researcher has been reported to the GMC and his case is pending. The problem with the paper about to be submitted to us is that the fraudulent researcher falsely claimed that he had gained consent from three ethics committees for patients to be x-rayed. The author has gone back to the three committees and they have all agreed to give approval after the event. Their judgement is that it would be unethical to suppress these useful data because of the consent problem. The author came to see me to ask whether we would be worried about publishing this paper. I said that I thought it would be acceptable to publish the paper, but that we should be explicit about the problems surrounding it. Does the committee agree?
COPE agreed, on the proviso that the data collection and analysis did not go through the fraudster’s hands. The author of the fraudulent data has now been struck off the medical register because of the fraud and forgery. The editor is sure that the fraudster did not collect the data. Several members of COPE said that they would not publish the paper. The editor should be advised to get further assurance regarding the data and then publish the paper with a commentary explaining the history.
The paper was published, along with an explanation of its history.
I received a letter from a reader in November 1997, pointing out that a paper published in the BMJ in 1996 was substantially the same as a paper published in another journal in 1994. We have examined both papers and discovered: (1) The papers describe the same cohort. There are the same numbers of patients, recruited in the same year; they have the same range of starting and finishing blood pressures. They are give the same drugs in the same hospital and had the same length of follow-up. The same outcomes are presented in both papers. (2) There is no lifting of text verbatim. (3) No substantially different or new material is presented in either paper in comparison with the other. A little more information on deaths and dropouts appeared in the BMJ paper in response to questions by the statistical reviewer. (3) Neither paper was referenced to the other, and the authors did not inform us of the existence of the paper in the other journal. (4) All authors signed the copyright form (5) The BMJ paper includes two authors who were not included in the other paper. This seems a very straightforward case of redundant publication. We have asked the authors for an explanation, and our expectation is that we will need to publish a notice of redundant publication. I have also written to the editor of the other journal.
The authors pleaded ignorance when challenged by the editor. This is no defence: the editor to publish a notice of redundant publication.
We have received a paper from a GP testing the hypotheses that because 24,25 cholecalciferol has a similar structure to commercially available statins, it may act as an inhibitor of HMA co-reductase. He screened 350 patients in his practice and identified 77 who had a cholesterol concentration above 6.5 mmol per litre. Thirty-three of them agreed to return for a second test 2 weeks later. They then agreed to take calcium and ergocalciferol tablets for 2 weeks and to return for a further estimate of their cholesterol. The author observed that their cholesterol concentration decreased, and he observes that the reduction in cholesterol is similar to that published in large studies on the effects of statins. This is a scientifically meaningless paper, and there is no evidence that the GP got consent from an ethics committee. Indeed, I don’t see how he could possibly have got consent for such a study. I have written to ask him if he has got such consent and I’m waiting to hear from him. The patients clearly agreed to take the tablets, but I worry that they were misled over the scientific value of the study. What should I do, if as I suspect, I discover that the GP did not have ethical approval? Should I report him to the GMC?
Reject the paper.
There is no obligation on a GP to get ethics committee approval (they are not employees).
The GP would need a DDX certificate if using a drug for an unlicensed use.
If a journal states that it requires ethics committee approval then that journal could reject the paper on those grounds.
The journal could ask him whether he obtained a DDX.
The journal could seek the GMC’s advice on GP research. Conclusion The journal should be advised in future to require ethics committee approval for all submitted research. The editor should tell the GP that it would be good practice to get ethics committee research and warn him to do so in future.
The GP did not have ethics committee approval, nor did he have informed consent from his patients.