Questions to ask:
- What are the main issues
- Who is involved?
- Do I need more information?
- If so, from whom?
- Who do I need to contact, and in which order?
- What possible courses of action can I take?
Case 1
A reviewer informs you that he considers a submitted intervention trial from the
UK to be unethical. This is because a possible, albeit rare, side effect could
be anaphylaxis. Given the risk, he does not believe participants would have
agreed to the treatment, so concludes that fully informed consent had not been
obtained.
You know that the senior author had unfortunately died so make enquiries of the
new corresponding author, who was his registrar when the study was undertaken.
He supplies a letter showing that his senior author had sought ethical approval
and that numerous concerns had been expressed by committee members, who asked
for further information. However, after an informal lunch meeting between the
senior researcher and the ethics committee chairman, the latter took personal
action to dispense with the need for approval.
Discussion
- There is no clear evidence of misconduct by three separate individuals. A senior
author seemed to have embarked on conducting a trial under dubious conditions of
ethical approval. The junior author took on responsibility to act as the
corresponding author, and should have been much clearer about the ethical
approval required for the trial. - Based on the limited information available, the chair of the ethics committee
himself played rather loose with ethical approval without informing his
committee. - The junior author and the ethics committee chairman should each get letters
outlining concerns, and asking for clarification, and if not satisfied with
responses, the editor is duty bound to raise the matter with the junior author's
employer and the ethics committee hierarchy. - What happens if the junior author responds, saying that the senior author told
him that ethics committee approval had been obtained? If the ethics
committee chairman is retired and the institution doesn't reply, should the
matter be referred to the regulatory body? - There is a question about how the doctors conducted the trial and whether the
corresponding author has documented evidence of permission gained. So if these
concerns were not adequately addressed, the editor would have to take it up with
the employer. - If the ethics committee chairman has retired, whoever took it on should have
given proper consideration as to whether ethical committee procedures had been
correctly followed. - The junior doctor has an individual responsibility to see the ethics approval
document as the corresponding author. - Irrespective or ethics committee approval, the editor has to decide whether or
not it's ethical to do the trial. The concern raised by the reviewer is
whether consent was properly obtained, so the question to ask is: How did
you obtain consent, and what discussion took place? A copy of the information
leaflet supplied to the patient should be available, so the editor can then
judge whether it's ethical on the very specific point raised by
the reviewer. - The editor who had submitted the case explained that the institution exonerated
the corresponding author/registrar, on the grounds that he had acted in good
faith, but they did criticise the ethics committee chairman. He removed himself
voluntarily from the medical register, so no further evidence could be taken
against him, but the trial participants could launch a civil case. - The corresponding author's and the institution's defence was that this was not
research, just an extension of clinical practice. The nub of the issue was the
consent form, which the journal requested, but on receipt found it to be "hopelessly
inadequate." For this reason, the ethics committee
chairman was criticised by the investigation. - This case took three years because the NHS had reorganised in the interim, and
ethics committees had come under different governance. It proved unbelievably
difficult to obtain responses to anything. - Technically, it's assault if you do something to someone without proper
consent, so such an offence should be reported to the police. - The police are generally not interested unless someone actually dies, and the
burden of proof for assault requires that you can show that a person was
assaulted; absence of informed consent would not be sufficient. - The consent form did include the warning of the potential for significant side
effects, and stated that consequently a senior member of staff would be present
at all times to deal with this. - When people are involved in research, and exposed to risk, there is a duty to
try and publish the study to ensure that the data are in the public domain. Is
there therefore a case for publishing the research, but with a commentary to
accompany it, stating that the research was unethical? - This is a common problem: Do you want this in the public record, so you can
alert people, or do you reject it, so that the authors will simply take it to
another journal where the standards could be lower and not feel the need to
publish a commentary? - What are the broader ethical obligations to the trial participants if evidence
that fully informed consent absent had not been obtained, but the nature of the
trial might indicate some long term damage that does not manifest for some
years? Does the editor have some obligation to ensure that patients are
informed?
Case 2
A reviewer, who happens to be an associate editor of another journal, tells you
that a paper you sent her to review has also been submitted to her journal. Both
covering letters stated that the paper had not been submitted elsewhere.
Moreover, the reviewer points out that a MEDLINE search shows two references
cited in the paper to other work by the author, which she considers were also
duplicate publications. When you write to the author, he apologises for the
error, which he states, resulted from a misunderstanding between his co-authors.
He wishes to withdraw the paper from consideration. The other editor has,
meanwhile, rejected the paper sent to him.
Discussion
- The main issue is the extent to which the editor can trust the author when he
apologises for the error, given the fact that he has done it before. - It merits a letter to him and then another to his institution to say that
although he explained this was an error, there is evidence that it has been done
before. Somebody should look at his whole publication record. - Whose responsibility is this? It's probably something the institution should
undertake, but would they do it? - The co-authors should be written to separately, because the corresponding author
alleges misunderstanding. If he is a serial offender, how do you share that
information among journals? - Should you write to the editors of all the journals, despite the fact that this
is an onerous task, outlining evidence of duplicate publication? At least
write to the other editor. - Editors cannot be responsible for the burden of proof, and this would put an
editor in a very difficult legal position. But they do have an obligation to
draw attention to suspected misconduct. - An example was cited of a suspected case of duplicate publication, which
prompted the journal editors to write to the authors, requesting an explanation
and warning them that it might result in retraction of the paper. The request
was answered with the threat of legal action, on the grounds that retraction
would damage careers. - The editors investigated and discovered that the other journal had published the
paper later. The study had been published online first. The other journal
applied to the permissions department of the first to reproduce the figures, and
published a statement saying that the article had been previously published
online. The author claimed that reporting conference proceedings made this
acceptable practice, despite the fact that it said original research
proceedings. - It is important to distinguish between redundant publication of original
research, which could skew a meta analysis and opinion pieces, and editorials or
review articles, which might be considered as secondary publications. There is
no rule against publishing your opinion more than once.
Case 3
You accept a paper, but ask for some revisions, which you
consider relatively minor. The new version takes a long time to arrive, and when
it does, the named authors have dropped from four to two. The corresponding
author explains that his co-authors were unable to agree on the changes. You
contact one of the 'disappeared' authors and suspect that she has been leant on by
her head of department, a senior official in the government's health department.
It appears that he is concerned that the conclusions are overstated and may
result in patients stopping a safe and necessary drug, if (as is likely because
of the topicality of the subject), the media quote the
conclusions without a clear description of the difference between causation and
association.
She and her head of department reject your offer that they should write an
accompanying commentary. Having reread the paper, you are sympathetic to the
government department opinion and agree that there is a chance that any ensuing
publicity might mislead patients. However, the two remaining authors refuse to
change their conclusions, adding that your last letter to them stated you would
accept the revision if they took into account the reviewer's comments,
which they have done in full.
Discussion
- The editor has already said s/he would accept with minor revisions, which
had been done. Arguably the onus is on the editor to accept this article. - As to the two disappeared authors, was it the suggested revisions, which
prompted this, or had they simply had second thoughts? We know one of them
was leant on, and it may be that this paper has very important public health
implications. The editor should find out why the authors have disappeared, and
should seek further peer review of the paper. - The editor should honour his promise to publish, but with an editorial comment
that two authors have disappeared, and perhaps ask for a commentary by another
clinician on this particular area of research. - Considering the authorship has changed substantially and ideally the authorship
contributions have also changed, arguably it is a different paper, so there is
no obligation to publish it. - But the revisions are minor, so it's not that different.
- Send it back to the original reviewer with an explanation of what has happened
as well as sending it to a new reviewer, outlining concerns about the
interpretation of the data, but not specifying why. - If the changes were what the editor asked for, surely there are no grounds for
rejecting it? - But the changes did not relate to author changes as well. Whenever authorship
changes it should always be investigated, because more often than not it means
that there is some disagreement somewhere. And perhaps the data are unreliable,
so the onus is on the editor to investigate, because there could be a serious
problem with the paper. - Don't promise acceptance after final revisions. Whether major revisions or
minor revisions, the process just continues until the editor is satisfied. - Contact the two authors who dropped out, and have them agree in writing to their
deletion from the authorship credits and that they are happy about this. - But they may have been leant on by their head of department, as is often the
case with junior authors. - Could the suggested revisions have prompted the authors to drop out? If
this is the case, the editor might be wrong. The paper might not be as good as a
result of the suggestions, and it may be that the authors are simply compliant
because they want it published. The editor needs to know that. - The message might be misunderstood by the press, and we have an obligation to
convey the correct interpretation. - If an editor is on the medical register, is there a conflict between the duty to
publish a peer review paper and his/her duty as a doctor to act in the
best interests of public health, especially if it is a safe drug and it could
prevent deaths? - That's impossible to know. Put both sides and let readers make up their
own minds. - Should the two remaining authors be attributed for all the work? If the other
two authors are happy to have their names taken off, that is their decision. But
if they are not happy, then there is something substantially wrong with the
paper. And if they are happy, but don't want their contribution
acknowledged, the editor could be publishing something fraudulent. - When there are many authors on a paper, sometimes asking what each contributed
prompts some to request acknowledgement rather than authorship. - Creating a paper with several authors entails discussions to reach a consensus.
Revisions also merit agreement among all the authors. Therefore, the paper
cannot be published until the editor finds out why they have withdrawn. - The editor who referred this case explained that all four authors agreed about
the scientific evidence, but two authors had been made to withdraw. They had
been told that the conclusions they had reached were not acceptable to the
government. - The two authors initially accepted the offer to publish their side of the story,
but subsequently refused, presumably because they were told that they couldn't.
The problem was that the paper had been accepted and all the requested revisions
had been made. - The paper was published with a long commentary, which had to adopt a cautious
tone. It could not say that the editor had been contacted by government
officials, merely that the authors had withdrawn for political reasons. The
authors were neither named nor acknowledged. - There was no audit trail, because everything had been done on the telephone, so
it would not have been possible to publish the whole tale. - Would it be unethical behaviour to force someone to take their name off
authorship? If so, do you report the senior government official to the
GMC? Would this be a case to refer to the new independent panel?
Case 4
You have accepted a paper which uses a case report of a rare adverse incident to
highlight the author's belief that a widely practised form of therapy is
not evidence based, is illogical and potentially dangerous. The author had asked
for it to be fast tracked because of the need to protect patients in future, and
you have agreed to this request.
Just as the proof copy arrives before publication, the author telephones your
technical editor. He explains that he is appearing as an expert witness on
behalf of a claimant seeking recompense from an NHS trust for alleged clinical
negligence in treating her in the manner outlined. He would like to know when
the paper will be published as he wishes to use it when giving evidence. The
technical editor did not ask whether the case report is about this claimant, the
claimant whose demographic details have been altered to preserve
confidentiality, or another case altogether.
Discussion
- The issue here is his conflict of interest as an expert witness and whether the
patient is the same one on whose behalf he is going to give evidence. Presumably
it is possible to find this out from the written consent signed by the patient
before publication. - Has he been an expert witness before in this area? Was the case fast
tracked so that it could coincide with his giving evidence? Did he know he was
giving evidence at the time it was submitted? - Is his fee the same whether he wins or loses the case? In which case, is there
really a conflict of interest, although perhaps it would have been better to
advise of his role? - Win or lose, it is a conflict of interest. Being an expert witness should be
declared, and if not declared, the editor has the right to reject the paper even
at acceptance. - Whether he acted as an expert witness is irrelevant, it is up to the court to
decide what weight to give to the evidence presented in court; that's
their job. Acting as an expert witness should not preclude acceptance of the
paper. - Not all journals ask authors to declare competing interests, so if he was asked,
and didn't declare, he is at fault. But if he was not asked, then he is
not at fault. - What are the criteria for defining conflict of interest? This is not
someone who has a contract with a pharmaceutical company, this is someone
legitimately giving evidence in a court of law, and they have obligations to the
court to be honest. - As an expert witness, your first duty is to the court, not to yourself or the
claimant on whose behalf you are giving evidence. - It is naïve to assume that expert witnesses are always honest to the court.
They can be pretty biased in what they say. It is a relevant competing interest
because it adds to his credibility and expert witnesses can be selective in what
they report. - Why do authors want papers fast tracked? Should that be an index of
suspicion? There have to be strong arguments in the public interest, and
often when that is requested, the issue fades away. - What about reviews and discussions of public data? Could these be expert
witnesses attempting to have their interpretation of the data published in peer
reviewed journals? - If it's peer reviewed and scientifically sound, we publish it, regardless of the
motives of the authors. But the question is: is this a conflict of interest and
should it be declared? - It's very hard to disentangle. If you write a lot about a subject, and
have done a lot of research, you are more likely to be invited to be an expert
witness. It depends on where your primary responsibility lies: to your
peers and other professionals or to the courts? Ideally, you would hope
that someone would include in their conflict of interest declaration that they
are a regular expert witness for that particular case. - If anything is published, it won't make any difference, because the court
applies the medical knowledge as it was at the time, not as it appears
subsequently. - This person did have a conflict of interest and wanted to
up their credibility in the eyes of the court by publishing.