Julian Sheather

Senior Ethics Adviser

British Medical Association

jsheather@bma.org.uk

This year saw an increase in the membership of the committee and a broadening in the base of its expertise with the recruitment of Ainsley Newson, lecturer in biomedical ethics at the University of Bristol; Jacinta Tan, a research fellow and honorary consultant child and adolescent psychiatrist at the University of Oxford; and Nick Pace, a lead clinician in anaesthesia at the Western Infirmary, Glasgow.

During this year the committee met four times, considered twelve new cases arising from submissions to and articles in the BMJ, and continued our discussion of some ongoing cases from previous years. The committee also discussed a number of broader ethical and regulatory issues relating to the journal. As in previous years, several themes emerged. Of particular interest was the international governance of research, and the procedures by which a peer-reviewed journal can involve overseas researchers in meaningful dialogue relating to research ethics. Consent both to involvement in research and to the publication of potentially identifiable information figured significantly, as did the relationship between medical journals and the pharmaceutical industry. Potentially fraudulent research, and the use of unattributed material were also important themes.

The work of the committee

The committee continued to scrutinise the nature and scope of its role in relation both to the journal and to the BMJ publishing group, and to seek to identify ways in which it could offer more timely and effective support. As a result, in relation to identifying cases for the committee to discuss, it was decided that reminders should be sent to the whole of BMJ well in advance of the committee’s meetings to consider whether papers required ethical review. Editors should also be invited to attend to discuss their cases, even when an editorial decision had already been made. It was also suggested that the reviewers’ form for the BMJ and the specialist journals should be amended to include a statement asking the reviewer if he or she has any ethical concerns about the paper.

BMJ governance

The Editor was keen for the committee to comment on governance procedures in relation to the BMJ publishing group. Of particular concern was the BMJ’s editorial independence in the face of potential commercial pressures from elsewhere in the organisation. Only by being free to comment candidly on products from both inside and outside the organisation could the journal retain the trust of its readers. The committee also discussed the new BMJ complaints procedure relating to the policies, procedures or actions of  BMJ editorial staff, in particular the requirement that particularly serious complaints be referred to an ad hoc panel of three people of ‘high standing’. This would usually be the editor of the BMJ, the BMJ Chief Executive, and the BMA Chairman of Council. The committee suggested that whenever the panel’s opinion was sought, the recommendations and responses of both parties should be published in the BMJ.

Transparency policy

The committee returned to a discussion of the BMJ’s transparency policy, focussing in particular on the recurrent issue of relations with the pharmaceutical industry. Readers had complained about the use of drug adverts on the BMJ’s website, particularly as users were charged for access. It was pointed out that this needed to be balanced against the fact that the revenue they generated kept the price of access down. The influence of industry on the research it funds was also discussed, and it was noted that JAMA had introduced a policy requiring independent statistical analysis for papers arising from commercially-sponsored research. It was decided that a workshop should be held involving editors and representatives from the industry in order to develop ways of ethically acceptable working. The BMJ’s transparency policy was published. It can be accessed at: http://resources.bmj.com/bmj/authors/editorial-policies/transparency-policy.

An interesting variation of this perennial problem related to an edition of ABC that had been drafted by a medical communications company whose fees were paid by an unrestricted grant from a pharmaceutical company. Although the named authors would have retained full editorial control the committee felt that the obvious potential for conflicts of interest meant that it should not be published.

Research

Inevitably a great deal of the committee’s work this year involved looking at both publication and research ethics, looking where necessary at both the integrity of the research and at the subsequent manner in which the research findings were presented. Again the issues of what actually constitutes research were raised, as well as the question of what research activities require ethical oversight. The papers reviewed by the committee suggest that further work is required to ensure consistency of decision making by research ethics committees.

Research misconduct – is it just too good to be true?

A case of suspected misconduct in relation to an author who had published a suspiciously large number of extremely positive trials was brought to the committee’s attention by a peer reviewer. Suspicions were subsequently cast on papers published earlier by the author in other journals, one of which was part of the BMJ group. The committee agreed that the editors of these journals should be informed. In another paper a statistical reviewer again found that the results of a study were ‘too good to be true.’ The Editor wrote to the author who strongly rebutted the suggestions. The committee suggested that the paper, and the database it drew upon should be sent to a second independent statistical reviewer for another opinion. To date, neither paper has been published.

Double counting?

As with the work of many ethics committees, much of the committee’s time is spent exploring fine lines between acceptable and unacceptable practice. An example was a paper where research relating to a multi-centre trial included patients from an earlier, single-centre trial that had been published in the BMJ just weeks before its submission. The overall feeling among the committee was that the authors ‘should not have done what they appear to have done’. Even though the single centre study had been cited in the multi-centre paper, the committee felt that the failure to draw readers’ attention to the earlier study, and the fact that around half the patients in the later study had been involved in the first study, was unacceptably misleading.

Research ethics

The nature and extent of the approval required by research ethics committees was raised by a paper that made use of covert surveillance of health care workers. The committee felt that the ethics of the research was questionable, and that the informal advice from the ethics committee that it did not require approval because it was an observational study was possibly unreliable.

Conflicts of interest

A question about just what issues should be disclosed by reviewers under the rubric of conflicts of interest was raised by one of the papers discussed by the committee. After receiving a poor review, an author asked if the reviewer should have declared that he had previously sacked the author after he had made a complaint about bullying. After investigation, it appeared that there was a disagreement between the reviewer and the author about the alleged sacking and the committee decided that the case should not be pursued.

Consent for publication

Once again the committee discussed several variations on the issue of seeking consent to the publication of potentially identifiable information. It was raised in relation to individuals who were the subject of rapid responses in the BMJ. The committee felt that in all circumstances, authors should be strongly encouraged to seek consent. It was also raised in relation to consent from parents for publication of an article relating to their extremely ill toddler. The authors felt that the parents were unlikely to agree, and that seeking consent might jeopardise the ongoing clinical care of the child. The committee was clear that the article could not be published without parental consent.

The question of the requirement for ongoing consent in relation to interactive case reports was also discussed again this year.[i] The committee encouraged the development of a series of semi-structured interviews with patients who were the subject of the case reports to identify whether they felt that the consent process was adequate. A new consent form for interactive case reports was developed and the committee also suggested some amendments to the BMJ’s existing policy on consent to publication.

The use of unattributed material

Several cases of authors using material without attributing its source were discussed. In one, a high profile author made extensive use of unattributed text in a book review. Following legal advice, the Editor contacted the individual, informing him that the journal would no longer publish his articles and asking him to indicate whether any other pieces that he had published in the journal made use of unattributed material.

Another case related to an article submitted to the BMJ exposing an author whose use of unattributed material first emerged in the 1980s. At the time, the author’s institution conducted an investigation, which concluded that the author had been guilty, but asked for discretion. No action was taken at the time, but more recently the same author was found to have used a large section of another author’s thesis and published it as a book chapter. The committee decided that if the article exposing the use of the material passed the peer review process, it could be published.

As a result, the committee suggested that the instructions for peer reviewers should be amended to include advice about what they should do if they suspect a paper is fraudulent or if there is evidence of plagiarism.

The committee will continue to meet four times each year, and reports published of its activities in due course.

Committee membership during 2006

• Iona Heath, general practitioner, Caversham Group Practice (chair)
• Elizabeth Wager, publications consultant, Sideview
• Hilda Bastian, head of health information, German Institute for Quality and Efficiency in Health Care
• Charles Warlow, professor of medical neurology, University of Edinburgh
• Alistair Newton, executive director, European Dystonia Federation
• Derick Wade, consultant and professor in neurological rehabilitation, Oxford Centre for Enablement
• Ainsley Newson, lecturer in biomedical ethics, Centre for Ethics in Medicine, University of Bristol
• Jacinta Tan, research fellow and honorary consultant child and adolescent psychiatrist, University of Oxford
• Nick Pace, lead clinician in anaesthesia, Western Infirmary, Glasgow
• Fiona Godlee, editor, BMJ
• Julian Sheather, senior ethics adviser, BMA
• John Weller, freelancer, BMJ
• Julia Burrell, committee secretary