Another year in the life of the BMJ’s Ethics Committee

Elizabeth Wager
Publications Consultant
Sideview, Princes Risborough, HP27 9DE, UK
liz@sideview.demon.co.uk

Sharp-eyed followers of the BMJ’s Ethics Committee will have noticed that our minutes are no longer posted on bmj.com. We agreed this change of policy in late 2005 because of growing concerns that it was not possible to anonymize the minutes effectively while, at the same time, provide a meaningful or interesting account of our discussions. We therefore agreed to produce an annual report of our deliberations, and this report covers the committee’s activities in 2005. During this year we met four times and considered 11 new cases arising from submissions to the BMJ as well as some on-going cases from previous years and general ethical issues presented by the journal. Over the year (as before),¹ several themes emerged.

 What is research?

In most countries, research involving patients requires approval from a research ethics committee (or institutional review board). But it is sometimes hard to distinguish research from innovative practice or audit and to determine which activities need prospective approval. Since the committee has often been faced with cases that raised questions about when ethical approval was required, we asked one of our members, Derick Wade, to write something on this topic, and his article was published in early 2005.²

As in previous years, the committee considered several submissions to the journal that reported activities that included some features of research or experimentation involving patients yet apparently had not undergone ethical review. These cases are reviewed below.

Unapproved research and issues of patient consent

We considered two extended case series describing the use of unlicensed treatments (in one case, a drug approved for an entirely different indication in adults was given to children, in another, a plant extract was given to elderly nursing home residents). Doctors are free to use unlicensed treatments, and the committee recognized that medicine would never advance if innovation were stifled, but journal editors must be alert to unethical practice and, depending on the seriousness of the concerns, prepared to educate and sometimes reprimand individuals who have acted unethically. In both cases, the committee recommended that the Editor should write to the authors, raising concerns about the need for prior ethical review. This was done after both papers had been rejected by the journal. The authors did not respond to the letters, so further letters were sent to their institutions but, so far, no response has been received.

In another case, patients with a serious mental illness were recruited to a focus group by healthcare professionals responsible for their care. The results might have been of interest to BMJ readers but the committee was concerned that such qualitative research should have undergone ethical scrutiny since there was a risk that patients could have been harmed or upset either by the process or the resulting publication (since personal details might have been discussed in the groups).

Once again, the Editor agreed to write to the authors and the committee was heartened by their response which indicated that they now understood the potential dangers of their actions and would behave differently in future.

All three cases also raised important issues about patients’ consent to treatment. In the case of the focus group in the psychiatric hospital, we wondered whether patients were really free to consent (or opt out) of such discussions, when invited to participate by their own psychiatrist. The other two cases also involved vulnerable populations -- children and elderly nursing home residents -- who may be unable to give written consent. We agreed that written consent cannot be obtained in certain circumstances, but that healthcare professionals must make every effort to obtain consent or assent. For example, if verbal consent is given, then the date and time should be recorded, and an independent witness should sign to state that the possible harms of any experimental treatment were explained as far as possible to the patient and, in some cases, their legal guardian. This process should be described in the publication.

Another report described the use of neurosurgery for opioid addiction. Both the Editor and the committee had grave concerns about whether the potential harms and benefits of this untested procedure had been explained properly to the patients. Once again, the Editor wrote to the authors but this process, in itself, also spurred some discussion. Editors have a duty to potential authors to ensure that decisions about acceptance or rejection are not unduly prolonged or delayed. However, in the committee’s experience, authors are more likely to respond to the journal if their paper is still under consideration, whereas, once the journal has rejected a submission, authors often fail to respond. If the journal has serious ethical concerns over a paper it does not intend to publish, is it justified in delaying the rejection in order to engage the authors in a discussion?  The committee agreed that the journal has a role in educating authors but, as one member commented, it should be ‘concerned but not vindictive’. Journals have a duty of care even towards authors who may have acted unethically and should not delay decisions or string authors along unduly. The BMJ therefore has a policy of informing authors as soon as the editorial decision is made, while ethical issues are usually pursued later, but as soon as possible, after they have been discussed by the ethics committee.

Patient consent for interactive case reports

Another issue of patient consent arose around the BMJ’s interactive case reports. The coordinating editor sought the committee’s advice over the consent form for these reports and the possibility of contacting patients after publication to explore their experiences of being involved.

The journal has clear policies on the need to obtain patients’ consent before publishing any details about them (which the ethics committee helped to develop).^3,4 But we felt that a slightly different approach was needed for the interactive case series because patients were being asked, in effect, to give consent for something they had not seen, and which their doctor could not entirely foresee. Whereas in a traditional case report, the clinician can show a draft version to the patient, nobody can predict what comments BMJ readers will make on an interactive case. These cases, by their very nature, tend to be controversial or difficult (otherwise they would not be interesting). Therefore, it is possible, in fact likely, that readers will disagree about the diagnosis and the best options for treatment. This could leave patients feeling confused or anxious, and could also affect their relationship with their doctor. Another unusual feature of the interactive cases is that they are published sequentially, so the reporting doctor might wish to reveal details about the patient in response to reactions from readers as the case evolves.

The committee therefore reviewed the consent form carefully and suggested wording to make these potential problems explicit. The form also makes it clear that patients may withdraw their consent at any time (which would halt the process of publishing comments or further details of the case). We also agreed that a BMJ editor could contact patients  after publication (if they agreed) to discover whether the experience had been a positive or negative one for them.

Another question of patient consent to publication arises with the digitization of back issues. The journal’s policy on consent has evolved over the years and the latest statement is relatively recent. The requirement for explicit consent has gradually increased and is now required for virtually every occasion when individual patient data are published -- but this was not always the case. Many journals are now making their archive available on the Web. This should provide a useful resource for researchers and patients, but internet archives also increase the availability of individual patient information for which current consent standards were not applied. It would usually be impossible or unreasonable to require that authors must obtain consent retrospectively for cases they may have published decades ago. However, we were concerned that there should be a mechanism for patients, or their close relatives, to contact the journal if they were distressed by anything that appeared about them in the digitized archive of back issues. The committee therefore recommended that a note explaining the journal’s position be published in its editorial policies section.

Research misconduct

A long-standing case of research misconduct (dating back, originally to 1992), which had been extensively discussed by the committee, was publicised by the journal, at our recommendation, in 2005.^{5, 6} It involved suspicions of data fabrication by an author in India. The expression of concern provoked a flurry of rapid responses from all over the world, including several from India. Some supported the journal’s persistence in pursuing this case while others accused it of ‘editorial bullying’ and even racism.

Another case of suspected misconduct resurfaced when the accused author’s work was included in a meta-analysis published by the BMJ.^7-9 Unfortunately, the inclusion of the discredited studies was not spotted until after publication. Once the authors of the meta-analysis were alerted to the problem they promptly agreed to perform a sensitivity analysis to test the effect of including or omitting the papers in question. This was published, along with a correction, on the journal website. This case highlights the difficulties of alerting researchers to suspicions about fraudulent work that have not resulted in a full retraction (although, even then, there is evidence that retracted papers continue to be cited and included in systematic reviews).¹⁰

The Editor also brought to the committee correspondence expressing doubts about the veracity of an author’s work. The allegations were vague and unsubstantiated, and the committee suggested that the Editor should seek opinions (in confidence) from respected researchers in the field before taking any action. After seeking such assurances (which were all reassuring), no further action was taken.

Authorship

Two submissions that underwent dramatic and inexplicable changes in authorship in the course of publication were brought to the attention of the committee. One involved a collaborative project between a major US institution and researchers in Eastern Europe. The initial submission included no authors from the country where the research was done, but three authors were added in a revised version. On the committee’s suggestion the Editor therefore wrote to the original US authors expressing concern that authorship had not been handled appropriately. The authors responded, stating that they would endeavour to adhere to authorship guidelines in future.

In another case, it appeared that an author who worked at an NGO had felt uncomfortable at being named because of political considerations within the organization. However, instead of outlining this difficult position to the journal, and asking whether the paper could be published anonymously, the senior author submitted the paper without the true co-author’s name but including another person as co-author. The senior author also suggested that the omitted author might be a suitable reviewer. The article was rejected. In his letter appealing against this decision, the senior author explained the background to the case. The committee recommended that the Editor should write both to the senior author and his institution, highlighting this breach of authorship behaviour and apparent dishonesty in suggesting that the interested party might review the paper. After intervention from his institution, the senior author apologised.

Transparency policy

The journal asked the committee to review items in the instructions to contributors covering issues of transparency, such as the involvement of medical writers and research sponsors’ involvement in developing publications. This is another area in which public opinion (and understanding of the issues) has developed over the years and the journal needs to keep abreast and ensure that it encourages transparency and full disclosure. The new transparency policy can be found at

http://resources.bmj.com/bmj/authors/editorial-policies/transparency-policy

Rapid responses

The journal’s facility for rapid electronic responses, as well as letters that appear in the printed version, had proved a victim of its own success and it was felt that, although large volumes of responses were being received, the standard was not uniformly high and a few individuals seemed to be commenting on almost everything. The committee discussed the virtue of allowing total freedom of speech with the merits of a moderated discussion and agreed that the journal’s new policy of filtering out some responses was reasonable and appeared to be having the desired effect of encouraging more thoughtful and reasoned postings.¹¹

Changes to the committee

Over the course of 2005 the committee had a new secretary and the journal had a new Editor. We thank Sam Knottenbelt for his work until October 2005 and welcome John Weller who now works alongside Julia Burrell (the committee secretary) to take the minutes. We also said goodbye to Kamran Abbasi, who was Acting Editor in 2005, and welcome the current Editor, Fiona Godlee, who continues to supply us with thought-provoking issues. We also thank Asad Raja and Jeffrey Tobias, who left the committee during 2005, for their contributions to its work.

The Committee continues to meet four times each year, and a report of our 2006 meetings will be published in due course.

References

1          Wager E. Experiences of the BMJ ethics committee. BMJ 2004 doi:10.1136/bmj.329.7464.510

2          Wade DT. Ethics, audit, and research: all shades of grey. BMJ 2005 doi:10.1136/bmj.330.7489.468

3          http://resources.bmj.com/bmj/about-bmj/ethics-committee/consent-to-publication

4          http://resources.bmj.com/bmj/authors/editorial-policies/copy_of_patient-confidentiality

5          White C. Suspected research fraud: difficulties of getting at the truth. BMJ 2005 doi:10.1136/bmj.331.7511.281

6          Smith J, Godlee F. Investigating allegations of scientific misconduct. BMJ 2005 331: 245-246.

 7          Smith R. Investigating the previous studies of a fraudulent author. BMJ  2005 doi:10.1136/bmj.331.7511.288

8          El-Kadiki A, Sutton AJ. Role of multivitamins and mineral supplements in preventing infections in elderly people: systematic review and meta-analysis of randomised controlled trials. BMJ 2005 doi:10.1136/bmj.38399.495648.8F

9          Correction for El-Kadiki and Sutton.
BMJ 2005 doi:10.1136/bmj.331.7509.142

10        Sox HC, Rennie D. Research misconduct, retraction, and cleansing the medical literature: lessons from the Poehlman case. Ann Intern Med 2006;144:609-13

11        Davies S, Delamothe T. Revitalising rapid responses. BMJ 2005 doi:10.1136/bmj.330.7503.1284