These guidelines differ little from the existing standards, which are honoured more in the breach than in observance by researchers, sponsors and journals.

I was the principal cardiologist and co-principal investigator in the Migraine Intervention with STARFlex© Technology (MIST) Trial.[1] NMT Medical sponsored the trial. The resulting paper was published in Circulation, the premier journal of the American Heart Association (AHA). Circulation requires authors to sign a declaration that largely accords with the proposed guidelines. Authors are required to sign that they have approved the final version of the paper; that they have had sufficient access to the data to verify the manuscript’s scientific integrity; that they have made a significant contribution to conception and design of the research, data acquisition and analysis, drafting the manuscript, etc; that those who contributed to the research are appropriately acknowledged and that nobody whose contribution should be acknowledged is omitted; and that they assign copyright to the AHA. The copyright transfer specifically states that the authors should be the sole authors of the work and publication does not infringe the copyright, proprietary or personal rights of a third party.

I sent an email to the other members of the steering committee saying that I could not in conscience sign the declaration because:
1. Individuals involved in this research and instrumental in its conception, design, execution, analysis and writing have been omitted in favour of individuals who had less involvement.
2. The paper contained errors.
3. I could not in conscience agree to be an author of an article which stated that “THE AUTHORS HAD FULL ACCESS TO THE DATA AND TAKE RESPONSIBILITY FOR ITS INTEGRITY” because we did not have full access.

Another member of the steering committee, Dr Simon Nightingale, also refused to sign the declaration.

Circulation did not question why two authors named in an earlier version had disappeared from the final version and published the paper in March 2008. I immediately wrote to the editor expressing my concerns. I provided Circulation with hundreds of pages of documents to support a case for publishing a correction. Eventually a 700 word correction, a four page data supplement and a new version of the paper were published in September 2009. I know that the new version of the paper still has errors. I accept that discussion of these errors may be controversial and that the BMJ may be concerned about a libel action, so let us consider some unquestionable facts.
1. One of the authors, Dr Rickards, died in May 2004.[2] He was not on the trial design group and the protocol was not finished when he died. The first author of the paper, Dr Dowson, was recruited to be my co-principal investigator in August 2004. Ethics approval was obtained in November 2004 and patient recruitment started soon afterwards. Dr Rickards obviously did not sign the author declaration. It is difficult to see how he qualifies as an author. The correction still bears his name, although Circulation knew at the time that he had died well before the trial commenced. His death in May 2004 did not prevent his name appearing as an author of a response to a letter about the trial early in 2009.[3] Why did Circulation knowingly allow an individual to be an author when he did not satisfy the journals authorship requirements?
2. The published paper and the amended version each mention the roles played by the trial design group and the steering committee, but the members of those committees are not named. The members of the trial safety committee are named, though they had a less demanding role than the other two committees. I was on the trial design group, but not one of the authors was a member. Dr Nightingale and I were 2 of the 5 clinical members of the steering committee. Two other members were from the sponsoring company, and they are both acknowledged in the paper. Failure to acknowledge our roles might prevent questions about why Dr Nightingale and I were not authors. Dr Nightingale and I worked at the centre that did the most patient visits in the trial - over 600 out of 2000 - more than 30% of all patient visits. Dr Nightingale and I wrote an early draft of the paper and contributed to subsequent drafts. Some “authors”, who had no role in conception, design, management, data analysis or writing did as few as 8 patient visits to qualify as an author. Not surprisingly, some authors had no idea about the controversy within the steering committee over inaccuracies in the paper, which were subsequently confirmed by publication of the correction. When Circulation has a stated policy on acknowledgement of contributions, why has the journal failed to ensure adequate acknowledgement of the roles of Dr Nightingale and me in the corrected version?
3. Circulation has failed to respond to my communications since the correction was published. Is that an appropriate way for an ethical journal to behave?
4. I have provided Circulation with incontrovertible evidence that I also wrote part of the protocol for another trial sponsored by NMT Medical, the BEST Trial. My contribution to that trial has not been acknowledged.[4]

What value is there in guidelines when they are ignored, unless their sole purpose is to mislead the public into believing that publication of research has integrity?

References

1. Dowson A, Mullen MJ, Peatfield R, Muir K, Khan AA, Wells C, Lipscombe SL, Rees T, De Giovanni JV, Morrison WL, Hildick-Smith D, Elrington G, Hillis WS, Malik IS, Rickards A. Migraine Intervention with STARFlex Technology (MIST) Trial: a prospective, multicenter, double-blind, sham- controlled trial to evaluate the effectiveness of patent foramen ovale closure with STARFlex septal repair implant to resolve refractory migraine headache. Circulation 2008;117:1397-1404.

2. Anthony Francis Rickards – Obituary. BMJ 2004;329:234.

3. Dowson A, Mullen MJ, Khan AA, Rickards A, Peatfield R, Muir K, Wells C, Lipscombe SL, Hildick-Smith D, Rees T, De Giovanni JV, Morrison WL, Elrington G, Hillis WS, Malik IS. Response to letter regarding article, “Migraine Intervention with STARFlex Technology (MIST) Trial. Circulation 2009;119:e194.

4. Mullen MJ, Hildick-Smith D, De Giovanni JV, Duke C, Hillis WS Morrison WL, Jux C. BioSTAR Evaluation Study (BEST). A prospective, multicentre, phase 1 clinical trial to evaluate the feasibility, efficacy, and safety of the BioSTAR bioabsorbable septal repair implant for the closure of atrial-level shunts. Circulation 2006;114:1962-7.

Competing interests: I was co-principal investigator in the MIST Trial. I refused to be an author of the resulting publication in Circulation, because it was inaccurate and because important information had been omitted. Circulation subsequently published a 700 word correction, a four page data supplement and a new version of the paper. I maintain that the correction is incorrect. I am currently being sued for libel by NMT Medical Inc, the sponsor of the trial, for speaking about my concerns about the trial. I am defending the action.

We are pleased that the publication of GPP2 [1] has given further prominence to the importance of ethics in biomedical publications. We are particularly pleased that the guidelines recognise the role of professional medical writers and make clear the important distinction between professional medical writers, who have a legitimate and ethical role to play in publications, and ghostwriters, who do not. We believe that the guidelines contain much good advice that should help to promote high ethical standards.

However, there are two points in the new guidelines that give us cause for concern, and we would be grateful if the authors of GPP2 could clarify their intentions in those areas.

First, the original GPP guidelines [2] contained the recommendation “Companies should endeavour to publish the results from all of their clinical trials of marketed products.” This recommendation appears to be absent from GPP2. We believe strongly that both positive and negative results should be published to avoid the serious problem of publication bias. Do the GPP2 authors agree that it is still important to publish the results of all trials?

Second, the precise role of the publication steering committee as described in GPP2 seems unclear. The guidelines appear to allow for the possibility that employees of the sponsor company could be members of the publication steering committee without being authors of the publication. That sounds dangerously close to ghost authorship. Do the GPP2 authors agree that it is important to ensure that sponsor employees should either be named authors of a paper or, if they are not named authors, should have no influence on the content of the publication?

References

1. Graf C et al. Good publication practice for communicating company sponsored medical research: the GPP2 guidelines. BMJ 2009;339:b4330

2. Wager E, Field EA, Grossman L. Good publication practice for pharmaceutical companies. Curr Med Res Opin 2003;19:149-54

Competing interests: AJ and HB are both officers of EMWA, which represents professional medical writers throughout Europe. Their roles as press officer and president are both unpaid positions. For their day jobs, AJ is the director of Dianthus Medical Limited, and HB is the director of Scinopsis. Both Dianthus Medical Limited and Scinopsis provide professional medical writing services to pharmaceutical companies.

These guidelines seem to obfuscate who is responsible and accountable for a particular clinical trial, and for articles written reporting on such a trial.

The guidelines ignore the time-honored concept of a "principal investigator" who is responsible for the entirety of a research project. Instead they allow for each manuscript arising from a clinical trial to have both a lead author and a "guarantor." The latter is envisioned as having "overall responsibility for the integrity of a study and its report." Then who appoints the guarantor? Who is responsible for the integrity prior to his or her appointment? When a trial generates more than one manuscript, would they necessarily have the same guarantor? If not, who would ultimately be responsible, and how would disagreements be resolved?

The guidelines prevent understanding of who ought to be the authors of a manuscript, what their responsibilities should be, and how accountable they are for the manuscript. The distinctions among "contributors," "authors," "investigators," "sponsor employees," and "individuals contracted by the sponsor" are unclear.

In particular, the guidelines seem to condone ghost authorship. They would not prevent a professional medical writer from writing an initial outline, the first draft, and all subsequent drafts. including the final draft of a paper. Yet that writer would not necessarily be an author, since only if if he or she is willing to "take public responsibility for relevant portions of the content" then he or she may be in a position to meet the remaining ICMJE criteria for authorship. Of course, this would allow a professional writer to simply be unwilling to take such responsibility in order to avoid being listed as an author.

In my opinion, these guidelines do little to increase confidence in the integrity of commercially sponsored clinical research.

Note: the above was derived from a longer commentary on the blog Health Care Renewal (http://hcrenewal.blogspot.com) which is available here:

http://hcrenewal.blogspot.com/2009/11/how-industry- views-research-it -sponsors.html

Competing interests: None declared

I welcome the publication of GPP2 and congratulate the team on their work and, in particular, for achieving far wider consultation than we attempted for the original version.

When we started work on the first version of GPP (in 1999) many companies were unwilling to make a commitment to publish the results of all trials of their marketed products and many companies (and industry associations) were opposed to trial registration. Progress has been made towards ensuring comprehensive registration and results disclosure but concerns about transparency and responsible reporting remain.

In my contacts with pharmaceutical companies I often hear complaints that trust has been lost in company-sponsored publications. I suggest that one small step towards restoring that trust would be for companies to endorse GPP2 publicly. The next step would be to make sure they are following the guidelines.

Competing interests: I was an author of the original GPP guidelines