BMJ Ethics Committee Meeting, 17 July 2002

Present:

Richard Smith
Sandy McCall Smith
Tom Wilkie
Liz Wager
Ann Sommerville
Peter Singer (via telephone)
Jeff Tobias
Derick Wade
Rachel Fetches

 

Item 1

DW asked for a statement attributed to him in item 2 to be amended, as he had not been present. This has been altered to TW.

This item was deferred to the next meeting in October 2002. The Committee requested that an hour of time be set aside to discuss the matter. SMS suggested that a preliminary, informal discussion with the GMC about the matter would be useful. AS informed the Committee that the BMA was updating Medical Ethics Today and this was an issue they would be looking at.

Read and comment on the guidelines

RF tabled the comments from the BMJ Editorial team on the draft consent guidelines. Roger Robinson joined the discussion as he submitted many of the filler cases to the Committee.

SMS stated that the Committee's revision of the consent guidelines originated because there was the feeling that the existing guidelines were too rigid. Often difficulties arose where patients were in developing countries or extremely remote areas. In regards to comments about 3(a) "The patient who is the focus of the article is untraceable without an unduly burdensome effort", PS thought that this sentence would be better as "The patient who is the focus of the article is untraceable." RS felt that it would be difficult to be certain that any person was truly untraceable and thought that 3(a) should be tried out in practice to see if this section proved unworkable. In regards to the comment that different standards between developing and developed countries would occur, PS felt that this would probably allow for more publication from the developing countries, where previously an article would have had to be rejected based on a lack of consent. SMS asked Roger Robinson about the current system and whether awareness of a potential problem of a double standard would be sufficient to safeguard against this.

In regards to 3(d), the editorial team thought that it might be extremely difficult to minimise risk of identification completely. One member felt that the publication of photos and the further statement that "Data relating to anonymised tissues and other samples will not normally be considered personal" might conflict. The Committee were referred to a photograph that had been published in Minerva where a patient had correctly identified a photograph of body parts. Thus it was queried whether even rigorous anonymisation of data might still create a problem with a patient recognising his or her own case. TW felt that the publication of a photograph would always be a matter of judgement depending upon the nature of the image. PS and SMS both suggested that the guidelines could be altered to differentiate between data presented as anonymised tissue and a more identifiable image in a photograph.

In regards to comments about consent being obtained from a patient's next of kin in the case of incapacity, being a minor, or death of the patient, SMS stated that it is not a legal requirement but good practice to ask relatives for consent where a patient has died. TW felt that the issue would be to look at who would be harmed by publication and what the residual harm and distress to the next of kin might be. DW thought that others such as close friends or significant others might also be in the category of persons who might be harmed by publication without consent. LW thought that the concern to obtain consent is defined more by the nature of the harm rather than who the particular person is. RS thought that following the line of close friends/significant others there could be many people who might be caused distress by publication. RS pointed out that a living person could give consent to publication of personal data—for example, genetic details—which could harm other relatives. TW thought that the standard is that of a reasonable person in the patient's position, which would take into account harm to significant others. TW also pointed out that all the conditions 3 (a–e) must be fulfilled before publication without consent is permissible. SMS pointed out that 3(a) is a means of obtaining proxy consent from the person in the closest relationship. AS felt that the definition would be better as obtaining consent from the person most likely to be harmed. SMS felt that the more legal position of the closest relationship would be better. RS was happy to tweak this slightly irrespective of whether a person was incapacitated or had died. TW felt that the criteria should be more strenuous for those who were incapacitated.

In regards to comments of there being a tension between 3(a) and (c), TW felt that if as in (a) the solicitation is likely to cause undue distress or alarm then publication would likely also cause undue distress and a reasonable person would object (c). TW felt that the illustrative sentence in (a) should be deleted.

The Committee returned to the questions raised about 3(d) and whether data included photographs. PS stated that complete anonymisation was more difficult with photos and videos. RS thought that the editors exercise their judgement about this. AS and TW felt that there is a difference between data demonstrated through photos versus anonymised tissue. DW pointed out that there is a difference between a photograph of a person/parts of a person that could be viewed with the naked eye and a photograph of a histology slide. RS pointed out that whether a histology slide led to a patient identification, would depend on the story around the slide. PS queried whether, in the case raised as an example, had the patient really been untraceable? RS stated that the doctors responsible for her care submitted the Minerva piece. AS felt that the photo was a red herring and in reality the patient had recognised the story and had been distressed by the photograph after this realisation. PS and DW pointed out that as the case arose out of the doctor-patient relationship, under the new guidelines, consent would have been required for publication. AS pointed out that the Retained Organs Commission now requires that all stored tissue should be traceable and requires consent to be held, thus most patients will be traceable.

LW suggested adding in the word "genetic". PS pointed out that the difficulty with this was the problem of consent in genetic pedigrees. DW felt that a histology slide presented as an example in a series of cases would not be identifiable. PS stated that in that case, the case series would have been through the research ethics review process and the guidelines were aimed more at problems with the anecdotal filler type article. SMS summarised that in deciding on the permissibility of publishing anonymised data, photographs could be separated out into two categories: (1) a photograph that is truly anonymous, and (2) a photograph that illustrates a unique narrative. TW did not feel that providing further detail in this section was necessary, as the editors would always have to exercise their own judgement. DW reminded the Committee of the Manchester seaman who was allegedly the first AIDS case and whom the tabloids traced in a very short time. DW felt that there was a gap between what the medical community felt was anonymous and untraceable and the reality if the media wanted to trace an individual.

PS suggested that the terminology "anonymised aggregate data" would be better, however, this would not solve the problem of obtaining consent in genetic pedigrees. JT queried whose consent would be sought in a study including genetic pedigrees? TW felt that the guidelines would benefit from the change in wording to include "aggregate" and that genetic pedigrees should be discussed separately. DW agreed that the Committee should discuss the difficulties of consent with genetic pedigrees and also the problems of consent for the first series of cases of a new condition. RS agreed that the latter issue does raise ethical problems as happened with the first description of vCJD, where one of the identified initial victims was Hindu and the link with eating beef had caused considerable family distress. SMS stated that the Committee would continue to keep this aspect of the guidelines under review and would discuss the implications of genetic pedigrees and new conditions in a separate discussion.

In regards to comments about 3(e) being a potential liability, it was queried whether the BMJ did want to state that consent had not been obtained as this would flag up the piece as something unusual. PS felt that this section had been inserted as a method of being accountable. There would be a general assumption that all articles had consent to publication and that where an exception to this policy occurred, this should be highlighted. SMS pointed out that this was not a question of legal liability, the policy of signalling that consent had not been obtained was a matter of being cautious. SMS suggested that if consent had not been obtained but that this was not flagged in the publication, accountability would require that a report on all articles published without consent should be compiled. DW felt that cases published without consent should be more anonymised than normal and brought to the Committee. PS felt that there should be some mechanism of audit or accountability of cases published without consent. These should be brought to the Committee to monitor. SMS stated that a policy of highlighting that no consent had been obtained could invite attention.

As there was a division of opinion on the subject of whether a lack of consent should be highlighted, SMS asked for a vote of opinion. Those who did not feel signalling lack of consent was necessary: DW, LW, AS, JT, SMS; abstention: TW; those for signalling: PS, RS, RF. SMS stated that 3(e) would be removed but that a satisfactory audit and accountability system for papers published without consent would be put in place. RS felt that the practicalities of providing an auditing system might prove difficult. AS queried what the current system was? The current system is that the cases are photocopied for the Committee. The Committee agreed that the editors would make the decisions and feedback the cases and the decision to the Committee.

In regards to 3(b), Roger Robinson questioned whether the BMJ fillers were sufficiently "worthwhile" to be published without consent; fillers were more interesting/fascinating rather than profound. JT thought that in an environment of increasing pressure and burn out, the fillers were useful for their leavening effect as they provide support for doctors. AS thought that the fillers show doctors as more human and that this was positive from a lay perspective. DW stated that most of the fillers did contain learning points. LW stated that the editors needed to balance the risk of harm against entertaining. AS thought that this would leave 3(b) as too high a hurdle. TW felt that a high hurdle was necessary, as the payoff for publishing without consent would need to be large. DW stated that he did not feel that the fillers were trivial in any way. AS agreed stating that they often make important ethical, cultural and clinical points. JT stated that narrative in medicine was important. SMS summarised the Committee view that worthwhile, in practice, was important.

PS felt that there should be a paragraph in the guidelines stating that they were a trial at present and that the auditing system would be a trial also. SMS felt that this should be mentioned elsewhere. RS stated that the editorial committee would use the guidelines and report back. RS suggested that after this period SMS and PS should write an editorial for the BMJ.

The Committee was referred to the action points summary and accompanying correspondence. All action points from 24/04/02 had been followed up but there had been no response from any individual authors or institutions.

RS updated the Committee on the situation with the other medical web site discussed at the previous meeting. Since reaching the agreement on the publication of financial sponsors' nominations of report authors, the deal had faltered over the commissioning of commentaries on the report. As an added value item, the BMJ wanted to commission commentaries from a British perspective on the conference reports but the medical web site wished to commission the commentaries themselves. Negotiations are currently stalled.

(4.2)

Background: A piece of qualitative research was submitted that looked at the experience of families facing a particular illness. In the majority of the cases, the researcher was also the care provider. The researcher conducted and analysed all the interviews. The other authors jointly supervised the study, were involved in the study design, and helped with the analysis of the data. The study received Ethics Committee approval. The referees and the editorial committee took the view that this research used inadequate methodology and worried that the first author, who had undertaken the study as part of a PhD, had been inadequately supervised. The committee's doubts were forwarded to the editor who wrote to all the authors. Two issues were raised (1) the design flaw of the dual role, and (2) whether the first author's PhD was adequately supervised.

All three of the authors responded separately. The first author (PhD student) rebutted the criticisms, stating that the dual role was extensively addressed in the full thesis, and that the research sprang from a phenomenological approach. The student felt that the editorial committee did not have any remit to question the adequacy of the PhD supervision. One supervisor expressed concern that the editorial committee questioned the adequacy of PhD supervision without knowledge of the full research exercise, access to the full thesis, or the independent external examiner's comments. Secondly, the supervisor felt that the issue of inadequate supervision should not have been raised directly with the supervisors rather than the student. The main supervisor rebutted the methodological concerns of the reviewers. However, the supervisor expressed concern that (1) the editorial committee felt it had a remit to question the adequacy of the PhD supervision, (2) that by writing directly to the student placed the student in a difficult situation, and (3) the allegation of inadequacy extended to the supervisor(s), examiners, and host organisation. The authors requested that the allegations of inadequate supervision be withdrawn and offered to submit the full thesis for evaluation.

The three questions before the Committee were: (1) Does the editor have any rights to raise questions about the adequacy and the supervision of a PhD? (2) Whom should the editor alert if there is an anxiety about the supervision of a PhD? (3) Were there any grounds for raising questions in this particular case?

RS asked the Committee whether they felt that the editor had a duty to raise concerns about potential inadequacies in the supervision of a PhD? JT & DW both thought that there is a duty. JT felt that this is analogous to the duty to investigate suspected research misconduct, which may be uncovered by an investigation into suspected inadequacies in PhD supervision. PS agreed that where the editor has a duty to investigate research inadequacies with authors, there is also the duty to do the same in the case of PhD supervision. The Committee agreed that the editor has the right to questions about the adequacy and the supervision of a PhD.

In regards to whom the editor should contact, PS stated that it depends on the context. SMS stated that the supervisor would be the person with the full information about the PhD project. RS felt that the situation was analogous to misconduct where the editor would go to the author in the first instance and then escalate to a higher authority if necessary. PS stated that the issue here was whether the editor should go to the student or supervisor. SMS pointed out that in this instance, the supervisor was also an author of the paper. RS felt that the editor has a confidential relationship with the author and that all communication should be with the corresponding author. JT stated that when seeking clarification about issues of supervision, other organisations he had been involved in would go straight to the supervisor(s). The rationale for this is that the student may be unable to answer higher-level concerns. TW summarised that going to the author (student) about the editor's concerns about a matter of potential research misconduct—a matter for the supervisor—may lead to a conflation in the editor's roles. SMS felt that prima facie, the editor should go to the supervisor in the first instance about supervision concerns. The Committee agreed that prima facie that the supervisor should be contacted in the first instance.

In this particular case, PS queried when the particular issue of grounds for concerns about the adequacy of the PhD supervision had arisen? This is not specifically mentioned in the written correspondence from the referees or the Hanging Committee. RS stated that the paper had been discussed in the Hanging Committee, which contains experts on qualitative research, and they were highly concerned about the position of the researcher as care giver. The specific concerns about PhD supervision must have been conveyed verbally. RS stated that the concern was that the student should not have been allowed to do badly designed research. PS stated that there are a number of positions on qualitative research and one report might not capture an alternate phenomenological approach. PS felt that a different view should be considered before stating that this research should not be done.

DW queried whether if research was considered to be poor, would this automatically translate to the view that the supervision must have also been inadequate? RS felt that pointing a student in the wrong direction in terms of his or her PhD research did reflect on the supervisor. SMS distinguished the elements the Committee was discussing: the conscientiousness of the supervisor and allowing the student to pursue flawed research. DW stated that there had been no evidence of poor supervision in terms of the personal conscientiousness of the supervisor. SMS pointed out that stating that a degree had been flawed was now actionable. RS felt that where a degree was a research degree, the competence of the teaching was an important issue. TW felt that this was a serious allegation. SMS stated that the Committee was not suggesting negligence in terms of overall supervision as the editor's concerns were aimed at the nature of the research not the supervision. The authors' had already stated that critique of the research was "fair game" upon submission of a paper for publication. The Committee agreed that the editor would clarify that the aspect of supervision being queried was the adequacy of the research.

Background: A submitted paper detailing a randomised controlled trial shows that a new treatment, which is a combination of familiar compounds, is highly beneficial in a common but largely untreatable problem. The authors come from several different countries and include people from the company that manufactures the treatment. It proved to be difficult to obtain reviewers for the paper and many returned the paper stating they could not produce an opinion. The reviewer who did eventually produce an opinion did not believe the results and suggested that there were signs that the data were fraudulent. A statistical review was then commissioned. The statistical reviewer felt the results were unlikely but was less convinced that the data had been manipulated. The statistical reviewer suggested the BMJ request the raw data.

RS updated the Committee on the discussion at the COPE meeting. The COPE committee suggested that the editor should request the raw data in electronic format from the study. LW thought that the editor should be candid about the fact that the BMJ did not wish to publish the study. RS expressed concern that if he stated that the BMJ was not interested in publication the authors would not co-operate with the request for the raw data. DW queried if the pharmaceutical company sponsoring the trial existed? RF replied that they do have a web site. TW asked how the data was going to be used by the company—did they need the data for licensing purposes? LW stated that the company needed the data to make any advertising claims. RS stated that as the product was not a novel compound, marketing claims might not have to rely on the data.

SMS pointed out that in previous similar cases, the Committee had treated such cases as a public protection issue. SMS suggested that the next steps should be (a) make it clear that the BMJ was not interested in publishing the study. (b) The editor should request the raw data. (c) Look at the possibility of raising the issue with the relevant regulatory agency.

Background: Authors A, B, and C submitted a paper about the behaviour of a group of doctors. All authors come from one institution, and the behaviour studied was that of the doctors at the institution. Author A (the most junior author) did the data collection under the supervision of author B (who was responsible for the design of the study and acted as corresponding author and guarantor). Author C is an office holder of the institution’s work. This paper raises difficult and important ethical issues.

The journal accepted the paper after revision, edited it and sent proofs. All correspondence to date had been with author B. When the proofs arrived, author B was on leave, and author C raised serious concerns about the paper and stated that it needed to be rewritten. Author C, who is the most senior author, did not seem to have previously read the submitted paper in depth—author C disputes this. A few weeks later, author C sent in a revised paper. The general effect was to water down the negative aspects and increase the positive aspects. In particular, author C stated that the original version had misinterpreted one part of the results and the new paper includes a new interpretation but no further supporting evidence as to the new interpretation. In addition, a new author, author D, appeared on the paper and was now listed as both the corresponding author and guarantor. Author D had not even appeared in the acknowledgements in the original version. The editor wrote to author C asking for an explanation of the change of authorship, raising the possibility that poor authorship practices were evident. He asked for a written assurance from all authors that they were happy about the revision, and asked for a copy of the questionnaire used as well as evidence to support the new interpretation so that the journal could judge the changes for itself.

Author C has written two holding replies, refuting the suggestion that there has been any authorial misconduct, and asserting that he/she was involved in the research. Author C stated that the authors were currently debating the appropriate interpretation of the results of the study. RS reported that he had spoken with the original guarantor (author B) who had withdrawn from being guarantor on the revised version. Author B stated that he/she would withdraw as an author if the paper were to be published in the substantially revised form. However, author B is clearly the person who determined the methods of the study.

PS stated that the issue of authorship is important. Different authors may have different interpretations of the data but academic freedom is an important right. Suggested 3 options: (a) let the authors' sort out their differences. Though they may fail to reach a compromise due to personal animosity. (b) Publish both versions with commentaries. (c) The BMJ editor produces the "middle ground" objective version from the two papers. DW queried to whom would the editor go back to, the original or new corresponding author?

PS stated that in regards to the addition of the new author at the late stage, the new author was the leader of a group at the institution looking at the issues discussed in the paper. Author D had provided new information in the revised version on the group’s work so could claim authorship. However, author D becoming a guarantor after not being mentioned in the original version was strange. TW highlighted the fact that the senior authors (C & D) might be sensitive to a paper addressing such difficult ethical issues because it could potentially receive media attention. TW suggested that the editor could use his prerogative to expel the author C as an author as it was not clear that he qualified as an author.

RS queried whether the issue of misconduct was not more to do with the issue of guarantors and authors being put forward who clearly did not meet the appropriate criteria. LW stated that the issue of authorship could be symptoms of abuse of power within the institution. However, looking at the guidance to authors on who is classified as a contributor, it is difficult because the guidelines are so broad. PS felt that the shift in guarantor from B to D left the editor in a difficult position. DW felt that author C should reconsider his/her authorship.

JT stated that he was against PS' option (b). LW agreed that to publish both papers puts author C up to ridicule and censure. PS repeated option (c) where the editor stepped in. LW felt that it would be difficult for the editor to re-write someone else's paper to find a common ground. JT queried whether the first paper, which had been accepted, was acceptable to publish. DW suggested the editor go back to the authors and suggest that the BMJ publish the original version with a commentary by authors C and D. PS stated that as the original version was more incendiary, it might be fairer to request a copy of the questionnaire and determine whether the issue under discussion could be interpreted in differing ways. RS pointed out that in research by Richard Horton of the Lancet, six different authors of a peer-reviewed paper had been contacted and all had different views about the interpretation of the data. It is still possible to publish a paper and commentaries debating the issues. It would be difficult to publish a commentary by an author of the paper.

PS raised the issue that the paper impacted on much wider and very serious issues, and also that the author A, a student, is in a very difficult position. PS suggested contacting author A directly to ensure that he/she is not being treated unfairly. TW stated that it was likely that author A would have graduated, however, the editor should check that author A’s future prospects were not being adversely affected. PS was concerned that the author A would lose a publication if the paper were to be withdrawn. SMS stated that it was author A’s decision whether to publish or not. SMS agreed that the editor should contact the author A directly. SMS also felt that author C had a right to comment.

JT suggested publishing the original paper with commentaries, inviting the institution to be the first commentator. The editor should write a letter to the authors summarising the concerns with the papers. TW felt that the revised version was a "spin" on the results. A more positive way of approaching the issue would be to publish the original paper with authors C & D as commentators explaining the steps that had been put in place to address the issues raised in the paper. PS agreed that it would be better that C & D were not authors but could make recommendations in a commentary. LW voiced concern that this will detract from the issues at the heart of the paper. RS felt that A & B are the original authors of the piece. LW again pointed out that the contributor guidelines are vague. PS flagged the issue that the guarantor on the revised version (author D) may have been pressured into taking on this role. RS stated that he would go back to the authors and ask the authors to respond to the Ethics Committee's debate within a deadline.

AS declared that she had not participated in the discussion because of a conflict of interest with one of the authors. SMS also stated that he was acquainted with one of the authors.

RS to write to the authors with a deadline for sorting out the issue

This item was deferred to the next meeting.

Item 7

This item was deferred to the next meeting

Item 8

This item was deferred to the next meeting. RS asked for a volunteer to provide an introductory paper to start the debate from. DW offered to provide this discussion paper.

(a) RS asked the Committee to consider expanding to include a member from a developing country. PS made a suggestion. The Committee briefly discussed whether there should be a formal application process and decided that they would look at PS's suggested member first.

PS to send details to RS/SMS/RF

(b) LW suggested that the meetings should be longer in order to address all the items on the agenda. The Committee agreed that three hour meetings would be better.

RF to send out dates for next meeting and dates in 2003. Next meeting potentially the middle of October 2002.