Intended for healthcare professionals


The BMJ's ethics committee is open for business

BMJ 2001; 322 doi: (Published 26 May 2001) Cite this as: BMJ 2001;322:1263

Editors face a surprising number of ethical problems

  1. Alison Tonks, assistant editor,
  2. Alexander McCall Smith, chairman of ethics committee and professor of medical law,
  3. Richard Smith, editor
  1. BMJ
  2. Faculty of Law, University of Edinburgh, Edinburgh EH8 9YL
  3. BMJ

    About a year ago the BMJ decided to form an ethics committee to help us with the increasing number of ethical problems we face.1 The chairman and members were appointed last year,2 and the committee met for the first time at the end of last year. Interested readers will find the minutes on, along with a paper on how the committee works and what it does. One gratifying feature of the first meeting was that our committee endorsed broad approaches—for example, about consulting authors over ethical worries and consent—that BMJ editors had already adopted.

    The ethics committee's business will be as transparent as possible. The manuscripts under discussion are confidential, however, so we have anonymised them to protect authors' privacy. The UK's Committee on Publication Ethics takes the same approach in its annual report.3

    We discussed eight papers at the first meeting, all referred by BMJ editors concerned about some aspect of the design, conduct, or implications of a report. All but two were case reports or case series, not formal research. These kinds of papers cause particular problems because they are never submitted for ethical review, don't count as research, contain identifiable patient details, and often conclude that things could have been managed better (often by someone else). Case series reporting clinical “experimentation” with different drugs or doses are equally problematical. When does the routine tinkering with treatments or surgical procedures become so unorthodox that it's unethical? Titrating doses of drugs to suit particular patients, or trying out different formulations of the same drug, is obviously acceptable, but what about offering “harmless” dietary supplements to people with a whole variety of symptoms because they seem to benefit?

    When doctors decide to write up their clinical activity and submit it to a journal, an editor becomes the first person outside the hospital or surgery to know that something has gone wrong in a clinical encounter; that a doctor is trying out a controversial treatment; or that a doctor bases clinical decisions on a scientifically meaningless theory. All these scenarios make editors' feel as though they should do something—or, more specifically, warn somebody.

    The notion of an editor's duty to warn came up in several of the manuscripts presented to the first meeting. The committee concluded that editors might have a duty to warn when there is a clear threat to patients' wellbeing. But this does not mean that they should pursue every instance of questionable practice, and the duty to warn must be interpreted sensitively. The cases that prompted this discussion are summarised in the box, but case 1 was the most difficult. Ophthalmologists from a tertiary referral center reported a patient who had gone blind and implied that doctors at a referring hospital had failed to make the diagnosis and delayed referral. One of our reviewers suggested that the patient had grounds to sue for negligence. Did BMJ editors have a duty to tell her?

    An editor's duty to warn

    Case 1—A woman went nearly blind after a delayed ophthalmological diagnosis. She was eventually diagnosed by doctors are a tertiary referral centre, who wrote up the case. BMJ reviewers felt she had grounds to sue for negligence. The ethics committee advised BMJ editors to ask the authors to discuss the patient's poor care and outcome with the referring hospital.

    Case 2—A doctor from outside the UK submitted his new theory of disease. It was scientifically meaningless, but he reported successfully treating himself and his patients on the basis of his theories. We referred him, with his knowledge, to the national regulatory authorities.

    Case 3—A plastic surgeon reported using a controversial breast enlargement technique and submitted photographs to show its success. When pressed for evidence the author cited journal articles, mostly from the US. None contained a robust evaluation. We did not publish the pictures, and told the author why. We wrote, with the author's permission, to the British Association of Plastic Surgeons asking for its opinion on the value of the controversial procedure.

    Case 4—A private practitioner reported a large case series of a scientifically dubious treatment. Some of his patients had also received increasing doses of a more established treatment. It was impossible to draw any meaningful conclusions from the series. We referred the case, with his knowledge, to the General Medical Council, who agreed that the study was meaningless but said that they could do little about it. There is no system of ethical review in the UK's private sector.

    After discussion (see minutes on we decided that the editors, with access to the limited information in the case report, were poorly placed to judge the rights and wrongs of the case and could make things worse by leaping in with disclosure. We decided instead to ask the authors to discuss this patient and her poor outcome with the referring hospital, making it clear that we would take things further if they didn't. The authors responded that this woman's care at the initial hospital had been good and it was never their intention to report otherwise. A brief report could not do justice to the complexities of her case, and we were misled by the wording of the submitted article. Happily the woman has now recovered moderate vision.

    We also learnt a lot from talking to the authors of a preliminary trial. They had treated people with a dangerous infectious disease by deliberately infecting them with another dangerous infectious disease. To us this treatment seemed shocking, but we were quickly educated that the treatment was not only acceptable (the trial had been scrutinised and passed by research ethics committees in two continents) but could be a potential lifeline for millions of people in the developing world caught up in a lethal pandemic and where drug treatment was unaffordable.

    The other issue dominating the first meeting was patients' consent to publish their case histories, clinical images, and pictures. Our policy is to ask for consent whenever an author discloses information gained in the doctor-patient relationship,4 and consent is usually given willingly. Problems occur when the subject of the report is dead, untraceable, or incapacitated. Case 9 in the minutes illustrates a common problem. A personal view describing a clinical encounter with a family in Bolivia was a powerful account with useful lessons about giving and receiving bad news. The family were thousands of miles away, had no access to the BMJ, and probably could not read. If no harm could come to them from publication, did we really need their consent to publish this personal view? The committee agreed, with one dissenting voice, with our existing policy that we did need consent. To publish without consent would have created a two tier system of consent, with patients from the developing world in the bottom tier.

    As the committee gains experience, a repeating pattern will probably emerge of the kind of problems facing medical editors in their day to day work, and we hope that the committee's discussions and decisions will evolve in to “case law” to help us deal with them. Readers can join in this experience by following and responding to the discussions posted on after the quarterly meetings.


    • Embedded Image A paper on the workings of the ethics committee and minutes of its first meeting appear on


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    View Abstract