A paper reported a radioisotope test for diagnosis of a speci?c,acute,neurological disease with 100% accuracy. Replication studies failed to con?rm the ?ndings and suggested that the test is positive in about half those affected and in a similar proportion of normal controls.Other publications by the same authors produced results at variance with their claims and misreported their ?ndings. One author admitted that the data had been altered to show a better result. An earlier publication from the same department described another isotope test for detecting an unrelated disease with 100% accuracy. It was later proved to be without value for the diagnosis of that disease.
The allegations of unethical experimentation
The study involved injection of a large dose of isotope into patients with acute neurological injury, in whom cognitive function was likely to be impaired. There was no mention of ethics approval or informed consent. The authors later stated that approval was not required because the test was used for clinical management. There was no previous or subsequent publication demonstrating clinical utility. The employing authority was therefore asked to explain how the test could have been used for clinical management. They replied that it was only a preliminary study. When it was pointed out that such a study would require ethics approval,they stated that this had been obtained,although they had not mentioned this in the paper or subsequent correspondence. When asked to provide a copy of the approval form, they threatened legal action. It is believed that the institution did not have appropriate approval to administer the isotope.
Attempts to silence the whistle blower
The whistleblower failed to replicate the observations and noted discrepancies in other papers by the same group.He contacted the patients involved in the study.They described events at variance with those of the published paper and produced documents to prove it. He challenged one of the authors who admitted that data had been altered to give a perfect result. The whistleblower approached the institution and asked for an investigation. Shortly afterwards he was told that an internal enquiry had found no cause for concern. The whistleblower asked why he had not been asked for the names of the patients who disputed the events described in the paper or asked to produce documents. He was threatened with legal action and expelled from an MRC committee on which he sat. The committee chairman was one of the authors of the disputed study.The institution blocked a request from the whistleblower to use information on a national database which is managed,but not owned,by the institution:the database is theoretically open to all investigators in the ?eld.
Having received no satisfactory response to his request from the head of the institution,the whistleblower approached the journal which published the paper, requesting that the journal publish a paper from him explaining that there had been scienti?c fraud and unethical experimentation,followed by a response from the authors. The editor felt that there was a case to answer and asked the authors of the original paper to respond. The editor copied the request to the head of the institution.
The head of the institution, instead, referred the whistleblower to the GMC for disparagement. The GMC investigated the whistleblower for eight months before he was exonerated and the focus of the investigation turned to the authors.
What should the editor do now?
The institution must produce evidence of the investigation.
The editor should refer the authors to the GMC if they are registered because there are legitimate doubts about the ethical procedures for this study.
A copy of the referring letter should be sent to the head of the institution.
A paper submitted for consideration in March 1997 was peer reviewed, successfully modified, and accepted for publication in June 1997. In January 1998 the paper was prepared for publication, and a commentary sought from an expert in the same field, scheduled for publication in the same issue. The expert drew the editor’s attention to the fact that a similar paper (albeit in shortened form) had been published in another journal in November 1997, after the paper to this journal had been accepted. The editor of the second journal who had no knowledge of the paper being submitted to to this journal. The papers were examined and the following conclusions drawn: 1 The sample size,methods,and results were identical for both papers. 2 The discussions were similar, lthough reworded slightly. 3 Additional data had been added to the paper for one journal which had been omitted for the short report in the other journal.
Could this be ignorance of process rather than bad behaviour? Both journals should make their position clear with regard to duplicate publication Both journals should look at their own processes for dealing with it.
The authors were asked to explain, especially as both papers had been submitted without either editor being advised of the other submission, and without a reference to the other journal in either paper. The authors claimed that the paper to this journal was a full report for the readers (one health profession) while the short report was to inform the readers of the other journal who comprised a different health profession. This explanation was regarded as insuf?cient grounds for the lack of information and the paper was not published. Interestingly, had the authors kept both editors informed and credited and referenced the original paper in the second, then it is unlikely that a possible case of duplicate publication would have been considered and both would probably have been published.
A paper was submitted crediting three authors. The paper was sent to one of the journal’s regular statistical reviewers without noticing that she happened to be the second author. She wrote back to say that she had not been involved in writing the manuscript, nor had she seen this paper before. She did say, however, that she had supervised the computer input of the questionnaire data and that she had provided some general advice on the simple statistical presentation of the data. Should any action be taken against the corresponding author, or should we simply explain what we mean by authorship and contributorship?
The editor should write to the author and take up the points raised by the statistical reviewer outlining the concerns about attribution of authorship.
The editor wrote to the author, enclosing a copy of an editorial on the differences between authorship and contributorship. The author responded, apologising for the mistake and attributing contributorship to the reviewer. The paper was published.
A specialist journal received a paper for review. An editorial board member was one of the authors. The paper was sent out for review and one reviewer replied quite favourably. A few days later the reviewer sent the editor a copy of a paper seen in another journal that was very similar to the one under consideration, and by the same authors. It was the same population and the same study, just a slightly different aspect of the paper. No mention of this study had been made in the paper submitted to the specialist journal. The editor wrote to the board member, who replied saying that there are some important differences between the papers,and that the one submitted to the journal is intended to be the main paper that will be cited, with further reports published in due course. The board member acknowledged that the previous paper should have been cited. What should happen about the specific paper? There clearly was a lack of transparency and it is hard to believe the board member was really so naive as to overlook the need to mention the previous publication. And what should be done about the board member?
Both the editor and the co-editor have been through the paper; there is a lot of duplication. The editors should write back to the authors and the head of institution saying that they are treating this as a case of redundant/salami publication. The board member has a further two years to serve, but he should be treated as the other authors, and the institution head informed. He should also be sacked from the editorial board
The editor received a long and informative letter from the board member. There were two papers being prepared by two groups who had been working on the population in question, he explained. One (the one they sent to this journal) was rejected by the first journal they sent it to and another was accepted rather quickly by a different journal. The editor thinks that the duplication between the submissions arose because of the different groups concerned. The board member and his team did a lot of hard thinking and have made arrangements to ensure that this sort of thing cannot happen again. The editor is convinced that this was a case of poor control and communication between teams rather than a deliberate attempt to deceive. The board member offered his resignation, but the editor invited him to stay which was accepted with gratitude. In fact, the editor suggested that he should collaborate on a piece for the journal about publication misconduct and what editors are doing about it.
A paper reported a questionnaire study of patients’ views on their preferences between minimal access and open access surgery. The questionnaires had been given to patients attending two types of clinic. The paper made no mention of ethical approval and the author was asked to clarify. He responded that he had not obtained ethical approval but that he had spoken to the chairman of the hospital ethics committee who would consider giving this retrospectively. A subsequent email from the chairman of the ethics committee to the journal expressed doubt about the value of retrospective approval, pointing out that when a study was reviewed prospectively it was possible to suggest changes to the protocol, which obviously could not be done when it was viewed retrospectively. The chairman added that “it is certainly not a foregone conclusion that we would have passed this study.” The internal editorial committee considered this case in particular, and the general issue of retrospective ethical approval, when the authors have simply forgotten or not thought of obtaining it. They were unanimous in deciding that retrospective ethical approval was not acceptable and that this paper must be rejected, explaining the reason.
The editor was advised to inform the head of the author’s institution that the study had gone ahead without ethics committee approval.
The paper was rejected, and the editor informed the head of the institution concerned.
A research letter was submitted from a team of investigators,A, B, C, and D. In their covering letter they reported that: A was involved in planning the study, collecting patient samples, and in writing the manuscript; B measured IL-10 polymorphisms and analysed the results; C was involved in supervising the measurement of polymorphisms and in writing the manuscript; D was involved in planning the study and writing the manuscript. The letter was peer reviewed and published. The corresponding author was D. Ten days later a letter was received from B and C who work at a different institution from D, inviting us to publish an erratum. Their substantive corrections were noted, together with the comment that “in addition, we wish to point out that B and C contributed equally to the content of this report”. C also enclosed a copy of a letter to D stating that he was very unhappy about the fact that the others had never seen the proofs so that the mistakes,as shown enclosed, could have been corrected. C considered it unethical not to show coauthors the proofs. Further strong comments about the breakdown of the research collaboration followed. D replied “surprised and saddened.”He argued that in the collaboration “the idea for this research was therefore entirely generated by us”. Furthermore, he said, B and C “saw and agreed to all the changes in the short manuscript and the final version that was submitted to the journal with all our signatures.” He went on: “I had to review the proofs within 24 hours and fax them back. There was no time to send this to the other authors for their approval (and we do not do this routinely in our department as it is usually the responsibility of the corresponding author). I am very concerned that you have sent off a letter to the journal without the courtesy of letting us see it beforehand. This is most unusual behaviour and can only have a damaging effect. The erratum is curious as these changes should have been made in the original manuscript.” What do we do about the alleged and apparently disputed erratum? Should journals have a clear policy about authors (all, some, the senior, or only the corresponding) seeing galley proofs? If so,what should the policy be?
There is responsibility to ascertain if there really is an error. The editor thinks that if there is, then it is an interpretive rather than a substantive error. The authors did not see the edited manuscript. It was agreed that it is the corresponding author’s job to clear changes with other authors. D removed B and C from the collaboration. This whole situation is not the fault of the journal, but the authors themselves. The editor should: either: invite B and C to write a letter to the editor and show it to A and D for comment. This way, the editor can ventilate this problem as a duty to readers; or: go back to the authors’ institution and have them resolve the dispute.
An international specialist medical journal has editors in the UK and abroad who function independently. An issue of a scientific journal in 1998 reported that the overseas editor had been dismissed from a university professorship because of scientific fraud. This had been documented in three published research papers.The report highlighted a particular paper, in which 27 references cited indicated the editor was the author or coauthor of 19 of the papers. Laboratory notebooks detailing the research had disappeared. The university committee stated that the study falsely presented data, and that these had been manipulated to produce the desired statistical results. The editor stated that there had been honest errors and that the laboratory staff had used poor research methods. The editor is attempting legally to overturn the university’s action. The UK editor wrote to the journal asking whether the incident discussed affects the editorial arrangements for the journal. Is there anything else the editor should do and does the problem affect his own position as an editor working in parallel with the overseas editor, as neither one is accountable to the other?
The overseas editor hired the staff who were subsequently criticised. The publishers are awaiting the results of an appeal by the overseas editor, and COPE feels that the editor should stand down or be suspended, pending appeal. If the overseas editor refuses to do this,the other editors should tender their resignations. The publishers must face up to their responsibilities.
The overseas editor resigned from the journal. It is understood that the overseas editor has not appealed, to date, over the dismissal by the university.
A paper was submitted to journal A which was published as a rapid communication. It was subsequently discovered that the major US journal in this specialty had published other findings from the same set of patients, and that the paper had been considered by them at the same time. The messages of the two papers are closely related but different, but either one could have been amalgamated into the other for one publication. All of this came to light when the authors submitted a further paper to journal A about the same patients. This time they used a new technique to expand on their studies which seemed perfectly reasonable; what is strange is that they did not acknowledge that the samples they analysed were taken from patients whose cases had been used and published before. Indeed, neither of the two previous papers acknowledges the other. A letter was sent to the senior author, asking for an explanation. The senior author responded by saying that her submission letter to me stating that none of the work had been published elsewhere was a proforma letter and the signature was an oversight.
There is no problem in using the same samples for different assays, but it is very important to be explicit. The head of the institution should be informed: non-disclosure is always a reason to inform the head of the author’s institution. This should have been one paper. A statement of redundancy should be published in all three journals and the authors should be blacklisted but the editors’responses should be consistent.
The editors have informed the authors that they have been blacklisted for two years. One of the authors who is head of department wishes to appeal and the editor has directed him to apply to COPE. The editor of the second journal blacklisted the authors permanently.
A study by Japanese authors was submitted to specialist journal A. The manuscript was sent to three reviewers, including expert X. After two weeks, expert X contacted the editorial office to say that an identical manuscript had been sent by the competing specialist journal B to expert Y in the same unit as expert X. Expert X and expert Y had compared and discussed both manuscripts. Expert X said that the Japanese authors were clearly attempting dual publication, were therefore completely unethical,and should be reprimanded severely. As editor of journal A,what should be done about: 1 The issue of apparently simultaneous submission to two journals? 2 The breach of con?dentiality by expert X (and also expert Y, commissioned by another journal B)?
Journal B doesn’t state that reviewers should maintain confidentiality. The editor wrote to authors and received a garbled response saying that they meant to withdraw the paper from Journal A. There had also been a letter from the head of the institution saying that the “authors were considering their response.” It seems that this may be a genuine mistake because of sickness. This story was corroborated by all the authors. As to reviewer confidentiality, journals vary in their practice. Breaches of confidence may be justified “in the public interest”.
The paper was withdrawn from both journals. The head of the institution formally apologised to both journal and gave sufficient explanation to make it apparent that a genuine mistake had obviously been made. He also added that he felt the corresponding author, as well as all the others,had learnt from this mistake. The breach of confidentiality was discussed by the editors of both journals involved. Expert X admitted that he had not read the instructions to referees, and had not been aware of this particular aspect of peer review. He undertook to reform his ways. He is still being used as a reviewer for journal A.
A reviewer detected that a paper received for review was almost identical to a paper published by the same group three years earlier in a journal of a different specialty. The paper concerned clinical and investigative aspects of a disease that crossed two specialties. Although the authors had included their previous paper in the reference list, the title of the paper had been changed from that in the other journal. References to this in the introduction and discussion were brief in the extreme and did not indicate in any way that the authors were re-studying, or re-reporting the same patients or data set.
The paper was openly deceitful in the reference list, but the work was not discussed and the title in the reference list had been changed. The editor was advised to seek clarification from the authors, and to refer the matter to the head of the author’s institution.
Further clarification was eventually forthcoming. The corresponding author accepted that there had been an error in the title of the paper in the list of references, but strongly refuted the suggestion that there was redundancy. The editor requested an independent review of the two papers by an expert reviewer who confirmed that there was major redundancy. A response from the authors was awaited. The authors still maintained that the paper contained new information and an impasse was reached. The paper was rejected and no further action was taken.