In the previous blog https://drpeterwilmshurst.wordpress.com/2022/09/25/is-the-lancet-complicit-in-research-fraud/ Professor Patricia Murray and I explained that the Lancet has failed to retract two articles by Paolo Macchiarini, which described surgery on a patient in Barcelona in 2008 and her follow up in 2014.1,2 There is no doubt that senior editors of the Lancet know that the publications are fraudulent. They have the evidence of fraud relating to these publications, which caused serious patient harms. The Lancet’s continued endorsement of Macchiarini’s publications is difficult to understand because a branch of Sweden’s Central Ethical Review Board found him guilty of research fraud in 2017 and requested retraction of publications that Macchiarini falsified when working at the Karolinska Institute. The Swedish investigators had no authority to consider Macchiarini’s original 2008 paper from when he worked in Barcelona. Since 2019, Macchiarini has received criminal convictions in Italy and Sweden for harming patients in those countries after he left Barcelona.
The Lancet is a member of the Committee on Publication Ethics (COPE). Members of COPE are expected “to apply COPE principles of publication ethics outlined in the Core Practices”. The Core Practices require that “Journals should have a clearly described process for handling allegations however they are brought to the journal’s or publisher’s attention. Journals must take seriously allegations of misconduct pre-publication and post-publication.” The failure of the Lancet to retract Macchiarini’s falsified publications raises serious concerns that the Lancet does not take seriously allegations of misconduct.
In this blog, I explain my role in the formation of COPE in 1997 and how I was informed that I was no longer a member of COPE two days after I complained to COPE that the Lancet was behaving improperly by refusing to retract Macchiarini’s fraudulent publications. I also describe how, in January 2019, COPE agreed to investigate my complaint that the Lancet had behaved improperly by failing to retract the falsified publication. Almost four years later, COPE has failed to tell me the outcome of that investigation.
I am concerned that as COPE evolved from a small group of individuals committed to improving the integrity of research and publication to a corporate body with considerable expenses, COPE has become dependent on a few large publishing houses that bankroll it. Five major publishing houses provide half of COPE’s income.
The Lancet is the premier journal of Elsevier and it has the second highest citation index of all journals. Elsevier, with 1861 journals listed as COPE members, contributes about 10% of COPE’s income – £48,863 this year. Elsevier is COPE’s second largest funder, after Springer Nature with 3091 journals enrolled. I believe that as a result of COPE’s financial dependence on membership fees, particularly from large publishing houses, it has financial conflicts which prevent it objectively assessing complaints about the ethical conduct of major journals owned by those publishing houses. From the information I will present here an observer might conclude that in return for membership fees, COPE now allows unprincipled journals to display its logo as a fake badge of integrity.
The formation and evolution of COPE
The Committee on Publication Ethics (COPE) was formed “to address breaches of research and publication ethics”. It was a discussion forum providing advice for editors. Its aims were to find practical ways of dealing with issues of concern and to develop good practice. At that time, the members consisted of a small number of editors of medical journals in the BMJ publishing group and the Lancet. There were two individuals who were not journal editors – Professor Ian Kennedy (subsequently Sir Ian Kennedy) and me. Kennedy was then professor of health law and ethics at University College London. He provided legal advice during discussion of cases at COPE meetings. My recollection is that Richard Horton, the editor of the Lancet, suggested the name Committee on Publication Ethics.
The formation of COPE had two stimuli. One arose because Michael Farthing, who was then the editor of Gut, had concerns about the probity of two articles that had been submitted to his journal. Farthing asked Richard Smith, the editor-in-chief of the BMJ, for advice and it was decided that a forum for editors to discuss such cases would be useful.
The other stimulus was a three hour seminar that Richard Smith invited me to give at the BMJ for about 40 editors and advisers of the BMJ publishing group and the Lancet in March 1996. After the seminar, Smith wrote “for years he (Wilmshurst) had been informing us of misdemeanours. Fear of libel stopped us from publishing.”3 In the seminar, I described 15 doctors who had committed research fraud (most of the doctors had published multiple fraudulent articles) and their misconduct was common knowledge but they had never been exposed or sanctioned. Five of the cases that I presented were corroborated by those present. Smith wrote “People said things like: I know that story and it’s even worse than you think.”3 A number of the editors that attended the seminar said that it altered their perception of the magnitude of research misconduct. Fifteen years later, Smith described the seminar as “a defining moment.”4 At the seminar, the BMJ and the Lancet agreed to publish editorials calling for action on research misconduct two weeks later.5,6 In addition, Richard Horton, the editor of the Lancet asked me to write an article describing some of the cases, which the Lancet published in February 1997.7 My article had a footnote added by Horton “Documentary evidence corroborating Dr Wilmshurst’s article was made available to The Lancet.”
My invitation to join COPE when it was formed in 1997 was because of my 1996 seminar.
The early meetings of COPE involved fewer than a dozen editors of UK medical journals discussing articles submitted to their journals or concerns sent in by other medical journal editors. Meetings were at the BMJ offices and secretarial / administrative work was by BMJ staff. The cases discussed were anonymised before publication by COPE. The published cases enabled a reader to see the facts of a case submitted to COPE and COPE’s recommendation to the editor.
Besides participating in discussion of cases, I was one of eight members who wrote the first COPE guidelines on good publication practice in 1999.8 Editors on COPE were keen to specify the responsibilities of authors. I pointed out that we needed a section on “duties of editors” and I drafted that section. In addition with Ian Kennedy, I co-chaired a COPE open meeting on whistleblowing.
Since then COPE has enlarged with membership becoming global and expanding to other types of journals besides medical journals. COPE’s committee members (who are its trustees) are distributed globally and the size of the committee is now larger than the original total number of members.
In 2007, COPE became a charitable company (registered as both a charity and a non-trading company) in the UK. Its charitable objects are “the promotion for the public benefit of ethical standards of conduct in scientific research and the publication of science journals”. At that time COPE started to charge for membership, which effectively constitutes its entire income. (Only a few hundred pounds each year came from other sources.)
In the first year after it became a charitable company, i.e. 2008, COPE’s income was £214,590 and expenditure was £70,736, giving it assets of £143,854. From 2008 until 2016, COPE’s income and expenditure increased, but income consistently exceeded its expenditure. As a result, COPE’s assets grew progressively to £425,295 by the end of 2016. That year income was £369,212 and expenditure was £295,334.
COPE’s income continued to increase after 2016, reaching £461,119 in 2021, but in the three years 2017 to 2019, expenditure exceeded income by between £57,000 and £105,000 each year. So by the end of 2019, COPE’s assets had been whittled down from £425,295 in 2016 to £204,869 at the end of 2019. Part of COPE’s expenses is the result of employing a full-time Executive Officer and a part-time freelance administrator, giving employment costs of £130,194 in 2021.
Other expenses include travel expenses of trustees / council members to attend council meetings. In 2012 “council members” received £813 in travel expenses. In the three years 2017 to 2019 “trustees” were reimbursed expenses for attendance at council meetings between £16,913 and £19,827 each year. It is unclear whether council members received additional reimbursements for attendance at COPE’s annual live forum, because there is are separate entries in the annual accounts for “council/forum expenses” which in 2018 was £78,951. In addition council members represent COPE at other meetings, such as “a significant COPE presence at the World Congress in Research Integrity” in Hong Kong in 2019.
In 2018 COPE’s income from its seminar in Australia was only £160. In 2019, COPE’s had two seminars in USA and Netherlands which produced combined income of £600. COPE’s reports do not usually specify attendance figures at its seminars, but COPE stated that at its virtual seminar in 2021 “the seminar topics were extremely well received with over 300 registrations”. I presume that the reason for providing a number that year was because it was unusually well attended. Even so, because COPE has more than 13,000 journal editors as members, if all the 300 who registered participated, it would represent less than 3% of members. This suggests poor engagement from the members who display the COPE badge.
The COVID 19 pandemic resulted in a reduction in COPE’s expenditure in 2020 and 2021 because, during those years, seminars for members and meetings of the COPE committee were held virtually, which avoided the costs of hiring conference facilities and travel expenses of committee members / trustees. Without the COVID 19 pandemic, it seems probable that COPE’s assets would have eroded further.
The majority of COPE’s income is from large publishing houses that obtain COPE membership for their entire portfolio of journals. Because some publishers have enrolled more than one thousand journals and Springer Nature has enrolled more than 3000 journals, there should be a question as to whether all the editors of the journals that are members of COPE are truly signed up to adhere to COPE principles and practices, rather than passively complying with the policy of their publishers. This arrangement also means that COPE does not know precisely how many members it has because publishing houses do not keep COPE informed about the number of journals in their stable. Individual journals can enrol for a small fee. COPE makes a selling point of the fact that COPE membership enables journals to use the COPE logo.
My expulsion from COPE
On 29 October 2018, I sent an email to Richard Horton, the editor of the Lancet which started “Three months ago, on 28 July, I sent you an email with supporting evidence asking for retraction of two papers in the Lancet – Macchiarini et al (2008) and Gonfiotti et al (2014). I expressed concern that patients were still having tracheal transplantation using so called “tissue engineered airways” based on Lancet publications by a team headed by a discredited researcher, Paolo Macchiarini.” The email then detailed the facts, which were that early in 2018 the Lancet had been informed by the Medical Director of the hospital in Barcelona that the information in the 2008 paper was false and, as a result the data in the 2014 paper were also false. Even before receipt of the information from Barcelona, the Lancet had known about Macchiarini’s fraudulent publications from the Karolinska Institute in Sweden because the Lancet had been asked to retract two of Machiarini’s falsified publications from when he worked in Sweden.
My email ended “You said that you have asked COPE to advise on whether the papers should be retracted. I do not think that is adequate. Time is passing and the Lancet must bear a share of responsibility for the deaths of patients that occur because you are allowing false research to remain on the scientific record. For goodness sake make an ethical decision and stop passing the buck to others.” I sent a copy of the email to COPE.
On 31 October 2018, I was informed that I was no longer a member of COPE because I am not a journal editor. I have never been a journal editor, but that had not prevented me being invited to be a member of COPE at its formation. My only reason for membership of COPE was because of my concern for integrity in research publications as described above.
After further enquiries from me, COPE informed me that “The Lancet asked COPE for advice on 7 September (2018). The case as presented to us was circulated to Council for their comments on 10 September and advice was sent to The Lancet on 14 September.”
Horton sent me a copy of COPE’s advice to the Lancet, but I have been unable to obtain from COPE or the Lancet the case details that the Lancet sent to COPE when requesting advice. COPE’s advice does not fit the facts in this case. So I wonder whether the Lancet’s presentation of the facts was accurate.
COPE’s advice to the Lancet said “The journal should resist being stampeded into a quick decision and should not be pressured to action by threats. Just because an author’s work has been found to be ethically compromised, that does not necessarily mean that every paper from that period is compromised. It also does not mean that earlier work was necessarily compromised. That may, or may not, be the case. The journal needs to ensure that the concerns are valid and substantiated for these papers as well. The only course of action is due process, where all the information is given to competent third parties to assess, and that takes time. Does the journal know if the institution investigated the other papers as well? Did the institution disagree with the statement from the clinic?”
It appears to me that the Lancet had not told COPE that the hospital in Barcelona had informed the Lancet that the information in the 2008 paper is entirely false as colleagues and I have described.9 I am mystified by the suggestion that the Lancet is under pressure from threats and puzzled by the final sentence in COPE’s advice “Did the institution disagree with the statement from the clinic?”
In order to understand COPE’s advice, I have tried unsuccessfully to get the details sent by the Lancet on 7 September 2018. I asked both the Lancet and COPE.
I wrote to Horton “The attached word document appears to be a summary of the advice received from COPE. Can I see the entire document or is that the entire advice? When did you receive this advice from COPE? Will you tell me what information you sent to COPE? Was it an anonymised summary, as is usual when getting advice from COPE? If so can I see it? I am concerned that from reading this summary it does not appear that those at COPE who gave this advice had full knowledge of what I sent to you……. When COPE advised that you should not be “stampeded into a quick decision” was COPE aware that patients are continuing to have this treatment and continuing to die as a result? The summary states that “the journal…should not be pressured to action by threats.” Can you explain why COPE gained the impression that threats had been made?”
I wrote to COPE “I have requested it (i.e. the information sent by the Lancet to COPE) from the Lancet and copied you in to that. However, the issue is transparency and how does one insure that there is transparency when the process is kept secret. If a journal misrepresents the facts to COPE, the journal can manipulate COPE to get the response it wishes. I am concerned that some journals use COPE membership as a sort of badge of integrity without actual being concerned about ethical publishing….. We have the problem that journals accept articles without any proof that the research took place or that it was conducted ethically, but then refuse to retract the articles even when there is proof of falsification and unethical conduct.”
As stated previously, when COPE advises on a case, it publishes on its website the information that the journal submitted and the advice given by COPE. The COPE website has no information about this case. So we do not know what “threats” the Lancet received. However, the four years of refusal to retract these false papers shows that the Lancet has taken COPE’s advice not to be “stampeded into a quick decision”.
Having failed to make further progress, I made a formal complaint to COPE about the Lancet’s failure to retract the falsified papers. On 13 January 2019, COPE informed me “I am writing regarding the concerns you raised to the attention of COPE in relation to the publication in The Lancet by Macchiarini et al. I have raised this case to the attention of a member of the COPE Facilitation and Integrity subcommittee for review and I will be in touch in due course.” It is almost 4 years later and COPE has not responded to my complaint.
Meanwhile, the finances of COPE are so precarious that COPE’s strategic plan for 2020-2023 is to expand membership to universities. In my experience of cases of research misconduct, universities are the group most likely to conceal misconduct in order to avoid damage to their reputations and evade having to refund research grants that have been misused. Universities and journals are the groups that will suffer greatest financial and reputational damage if there is public recognition of the extent of research misconduct and their unwillingness to deal with it.
As far as I am aware, of the original founders of COPE, who set out to improve publication ethics, only Richard Horton remains a member. His transition into a publishing house employee who lectures others on ethics but refuses to retract publications that are patently fraudulent and that harm patients is symbolic of the change in COPE itself.
- Macchiarini P et al. Clinical transplantation of a tissue-engineered airway. Lancet 2008;372:2023-2030.
- Gonfiotti A et al. The first tissue-engineered airway transplantation: 5-year follow-up results. Lancet 2014;383:238–244.
- Smith R. Editor’s choice: Corruption in medicine. BMJ 23 November 2002;325.
- Smith R. A successful and cheerful whistleblower. https://blogs.bmj.com/bmj/2012/10/10/richard-smith-a-successful-and-cheerful-whistleblower/
- Smith R. Time to face up to research misconduct. BMJ 1996;312:789.
- Editorial. Dealing with deception. Lancet 1996;347:843.
- Wilmshurst P. The code of silence. Lancet 1997;349:567-9.
- Committee on Publication Ethics: the COPE Report 1999 Guidelines on good publication practice. Occup Environ Med 2000;57:506–509.
- Schneider L, Murray P, Lévy R, Wilmshurst P. Time to retract Lancet paper on tissue engineered trachea transplants. BMJ 2022;376:o498.