Experiences of the BMJ ethics committee

Short abstract

The BMJ established its ethics committee nearly four years ago. What has it achieved and has it changed the journal's practice?

The BMJ ethics committee was established to help editors with difficult cases, to review policies that have ethical implications, and to expand the journal's coverage of ethics.¹ Unlike the Committee on Publication Ethics (which considers anonymised cases submitted by journal editors), the BMJ's committee discusses possible future policies as well as responding to specific issues and sees full details of all cases (although they are anonymised for the minutes, which are available on the BMJ's website (http://bmj.bmjjournals.com/advice/bmj_ethics.shtml).

Since our inception in November 2000 we have met 13 times and discussed 70 cases from the editorial team. Three broad themes have emerged from our discussions. These are:

• Patient confidentiality

• The journal's responsibilities to people outside the publishing process, and

• Research misconduct.

Patient confidentiality

Like many ethical issues, the confidentiality of the doctor-patient relationship sounds straightforward until you are confronted with difficult cases. As part of our review of journal policies with ethical implications, we considered the BMJ's guidelines on publishing information on individual patients. These stated that information should not be published without the patient's consent. That sounds fine, until Minerva wants to publish a picture of an Egyptian mummy²,³ or someone submits a piece with an important message describing a patient who died decades ago in a remote part of the world. The committee has therefore revised the guidelines to describe the situations in which publication of individual information without consent might be appropriate.⁴

We also considered whether rules about consent to publish individual data still applied after a patient had complained about a doctor to the media. Does the patient's public allegation constitute an implicit consent to have the case discussed? If not, is it fair that the doctor(s) concerned cannot respond to the allegations or attempt to set the record straight? Should the BMJ ask patients who write criticisms of doctors to sign a waiver permitting the doctor to respond? Discussions with the General Medical Council and the BMA's ethics committee are still underway, but we hope the outcome will be new recommendations that satisfy natural justice, publication ethics, and the GMC's guidelines (which currently curtail the doctors' right of reply).

Journal's wider duty of care

The committee has considered the journal's responsibilities to people who are not directly involved in the publication process. For example, we discussed a submitted paper that set out unorthodox theories about a serious condition and implied that these formed the basis for the author's clinical practice. This paper did not purport to be a clinical trial, so we could not invoke research ethics committees, but it did raise serious doubts about the author's competence. We recommended that the BMJ's editor should write to the medical authorities in the author's country raising the journal's concerns (so far he has not received a response). Was that a breach of confidence between an editor and an author, or the act of a responsible journal with a duty of care to patients?

We also discussed a letter from junior doctors who had witnessed cheating during their final exams and believed this might have affected their results. This raised an interesting problem: the doctors (now qualified) were to be congratulated for their honesty, but what could the BMJ do with the information without harming the authors? The committee felt it would be wrong to contact the medical school without first getting the agreement of the whistleblowers. The editor contacted the authors, and they agreed that he should write to the medical school explaining how this information had been obtained but not revealing its source. The medical school responded that academic staff had, in fact, been aware of the problems at the time and had taken measures to minimise the harm.

Another case involved the journal's duty of care to a patient whose case was reported in a filler article submitted by doctors at a tertiary referral centre. This implied that the patient had been severely disabled because of inadequate care received at another hospital. One reviewer remarked that the patient should sue for negligence: should the journal pass on this advice? We recommended that the authors should be encouraged to pursue the case. However, when the editor sent the reviewer's comments to the authors, it turned out that their report had considerably simplified the case, and, since submission, the patient had made an almost full recovery. These cases taught us an important lesson—that journals usually have only partial information, and it is therefore best to seek clarification from the authors before making any assumptions or taking further action.

Figure 1.

Publishing details about a mummy without consent is unlikely to cause problems

Credit: ALEXANDER TSIARAS/SPL

Research misconduct

The committee's longest running cases have concerned suspected research malpractice. The individual cases are still under investigation so we cannot report them in detail, but they have raised questions such as whether the editor should inform other journals when a BMJ investigation casts serious doubt about papers that they have published and, if so, when? One case has lasted more than four years and shows no signs of resolving. It has involved authorities on several continents and threats of legal action. It might seem fairer to wait until an inquiry has reached a conclusion before alerting other journals, but the committee felt that this would increase the risk that further potentially fraudulent research might be published.

Like many journals, the BMJ requires authors to show that research has been approved by appropriate bodies and conducted to relevant ethical standards. In most cases, research ethics committees or institutional review boards approve research proposals, but several grey areas have exercised our committee. For instance, we learnt that there was no provision for ethical review for research conducted in private practice in the United Kingdom. Similarly, it is not always clear how innovative surgical techniques should be studied, and whether variations from standard practice require ethical review and explicit patient consent. We have also discussed reports describing the use of dietary supplements or of licensed drugs at unusual doses that raised concerns about whether patients were fully informed that they were receiving unorthodox treatment.

Perhaps the greyest of areas, and one that the committee has yet to resolve, is the borderline between audit and research. Legislation and guidelines about the use of patient records are in a state of flux,⁵,⁶ and doctors may not realise that they need to seek approval. We plan to discuss this further at future meetings and to draw up guidelines (or at least point authors in the direction of some helpful resources), since we know readers are concerned about it.⁷

What else has the committee done?

Although much of the committee's work arises from cases referred to us by editors, we have also taken some initiatives ourselves, including a review of the BMJ's guidelines for contributors. Our aim was to make the guidelines consistent within themselves and with ethics committee decisions and to bring them up to date with other documents such as the latest version of the Declaration of Helsinki. The guidelines now include a section entitled “What we do if we suspect misconduct,” so that anyone submitting material to the BMJ is aware of what the journal might do in these circumstances.

Examples of cases discussed

Can the BMJ publish a report by a nurse who had been assaulted by a patient without the patient's consent? We decided that this case raised an important issue and therefore merited publication but should be published without the author's name or institution, and with only minimal details of the patient, so that the risk of either party being identified was minimised.

If the BMJ has agreed to publish a case anonymously (to avoid identification), can the author also publish a signed commentary of the case as if he were an impartial observer? We felt that this was deception and therefore unacceptable.

Review of the guidelines was one of the committee's original objectives.¹ We hope that the new guidelines about publishing patient information, and our discussion of individual cases, have helped editors take difficult decisions and thus fulfilled another objective. However, we have been less successful in increasing coverage of ethical issues in the BMJ. As we meet just three times a year, for about three hours, we do not always have time to cover all the items on our agenda. We have given priority to cases referred to us, but, even so, it is sometimes several months before we make a recommendation. This must be frustrating for both authors and editors, and we would like to improve our performance. Another effect of full agendas is that we rarely have time for wide ranging discussions of broader issues and have been less proactive than we might have hoped.

So, what next? We need to do more work on the question of where audit ends and research begins, and no doubt the BMJ editors will continue to provide us with dilemmas to stimulate our discussions. We hope to be more proactive in encouraging coverage of ethical issues in the journal and are currently discussing questions arising from medical genetics. Minutes of meetings will continue to be posted on bmj.com, and we also hope to use the rapid response and polling facilities to broaden the debate on important ethical issues and increase readers' involvement. Although we do not claim responsibility for any major changes at the BMJ, feedback from the editorial team has been positive, and we feel our work has been helpful. In a busy journal office it can be hard to find time to debate the broad implications of policies or to review policies that need updating. The ethics committee ensures that the BMJ editors regularly consider ethical topics, and we hope it has resulted in better, more thoughtful policies for the journal.