The Stoke CNEP Saga – did it need to take so long?

In 1999, the Royal College of Physicians of Edinburgh, the Royal College of Physicians and Surgeons of Glasgow, the Royal College of Physicians of London, and the Faculty of Pharmaceutical Medicine convened a consensus conference on Misconduct in Biomedical Research. The conference concluded that a national panel should be established – with public representation – to provide advice and assistance (on issues of research misconduct) on request.¹ It was suggested that the panel might develop and promote models of good practice for local implementation; provide assistance with the investigation of alleged research misconduct; and collect, collate and publish information on incidents of research misconduct. It was hoped that the report of the Conference would be given the fullest possible dissemination by the sponsoring bodies, and that they would convene at the earliest opportunity a meeting with the General Medical Council and appropriate partners to establish and consider the remit of the national panel.²

A national proposal, 2001

For over a year, nothing seemed to happen. As a result, a striking editorial in the BMJ stated that ‘the largely submerged problem of research misconduct is surfacing like a decomposing corpse’.³ Behind the scenes, however, there was some appropriate activity. In 2001, a proposed blueprint for the prevention and investigation of misconduct in biomedical research was published, under the title A National Panel for Research Integrity.⁴ The authors of this blueprint represented the Royal College of Physicians of Edinburgh, the Royal College of Physicians and Surgeons of Glasgow, and the Faculty of Pharmaceutical Medicine, with subdued support from the Royal College of Physicians of London. We concluded (for I was one of the authors) that the outcome of the 1999 Edinburgh conference had been a landmark in highlighting an agreed need – though with hindsight the word ‘agreed’ was far too optimistic – namely, that all stakeholders (in biomedical research) should collaborate in establishing a national body to promote education, standard-setting and audit of biomedical research within the UK. We had discussed with many other parties the practical developments needed, and, by publishing our ‘blueprint’, we thought we had done enough to help establish a National Panel on Research Integrity during 2002. In retrospect, we were unrealistically optimistic.

One of the pivotal recommendations within the ‘blueprint’ was the need to establish an agreed and recognized ‘rapid response process’ through which institutions could call on independent teams, with members drawn from national lists of trained external assessors, to investigate confidentially allegations of research misconduct. One such team already existed – MedicoLegal Investigations Ltd (MLI). I had helped to set it up in 1996 with a colleague, Peter Jay, who had previously been in the Metropolitan Police and had latterly been a forensic investigator of alleged cases of research misconduct referred to the solicitors employed by the General Medical Council (GMC). He and I had worked in sequence for several years before we established MLI. In my capacity as medical director of the Association of the British Pharmaceutical Industry (ABPI), I had received suspect cases from pharmaceutical company medical directors who were concerned that data had been fabricated or falsified and, above all, that patients had been exploited. I had worked up the cases as far as I could from within the ABPI, enough to warrant referral to the GMC; Peter then completed the forensic investigation. On my retirement from the ABPI, it made sense for us to work together. We were always aware that the gate was wide open for other investigatory teams to enter this field, and we would have welcomed that.

It was emphasized in the ‘blueprint’ that the external investigations should be conducted according to due process,⁵ using standard operating procedures (SOPs), as agreed by the National Panel. Additionally, it was emphasized that the principles of such confidential external investigations should include a rapid response to requests; investigation by a team of trained, impartial experts; protection of patients and volunteers in research studies; protection of whistleblowers;⁶ and protection of clinical and scientific researchers from unjustified allegations of research misconduct.⁵ Cases seldom took more than about six months to complete. MLI remains the only independent operator in this field, has appropriate SOPs, and subscribes to all the requirements just mentioned.

So what has happened since 2001? Did all the stakeholders get together in order to establish a National Panel in 2002? No. Things have moved on, though not to establish what the authors of the ‘blueprint’ recommended. The UK Research Integrity Office (UKRIO) was launched in 2006, hosted by Universities UK. It purports to offer support, both to research organizations and to individual researchers, and to promote integrity in research and good practice in addressing misconduct in research.⁷ Although its advice and guidance is available to all, by the end of 2008 it had a profile so low that seemingly very few workers in the field knew of its existence.⁸ The Lancet was dismissive. It called the UKRIO an ‘ineffective enterprise’ which ‘is at best bound and gagged by its ties – at worst … a smokescreen set up by universities themselves’.⁹ It was critical of the fact that its procedure for the investigation of misconduct in research, published in 2008,¹⁰ failed to encourage organizations faced with cases of alleged research misconduct (be they universities, research councils, pharmaceutical companies or other bodies) to involve an independent, trained, rapid response team to advise at an early stage on how the investigation should be handled. This omission is a fundamental flaw.

My own more recent experience

The way in which the CNEP case was handled (see previous articles in this series) demonstrates very clearly what went wrong, and what will continue to go wrong if current UKRIO advice is followed. The overwhelming evidence is that the GMC, the Royal Colleges, the universities and NHS Trusts are unable to fulfil this investigative function and to do this in a transparently fair and independent way. My experience, before and after the inception of MLI, confirms that this function requires an experienced, but above all, independent, approach. It also requires great sensitivity, particularly when a case, on investigation, proves not to have been dishonest, which happens not infrequently. Then, indeed as always, any whistleblower also has to be handled and advised appropriately, so that no criticism follows when the whistleblower has acted in good faith.

The first case in which I was involved was that of a consultant psychiatrist in Durham. He was suspected by the sponsoring pharmaceutical company to have fabricated biochemistry and haematology results for patients recruited into a pivotal study for a new tricyclic antidepressant.^11,12 The company did not wish to be involved in investigating the case itself and sought the advice of the ABPI. Although not fully independent, as medical director of ABPI I was sufficiently remote from the company itself to set up lines of investigation that were not readily open to the company, and thus to establish the facts of what had happened. The psychiatrist claimed that he had delegated the management of the trial to his registrar (whose name he had forgotten) and that it was her responsibility to ensure that the data were correct. His claim was incorrect: it had been accepted since 1986 that it was the principal investigator's responsibility to ensure the veracity of data submitted to a sponsoring company. After the regional medical officer of the Northern Regional Health Authority had provided the maligned registrar's name and new place of work, I invited her to comment on the accusations of her former consultant. She responded, indignantly, that she had had nothing whatsoever to do with his research projects. The company was pleased that the ABPI had been involved, as they did not feel confident enough to refer the case to the GMC by themselves, so the ABPI presented the case to the GMC on behalf of the company, by means of a statutory declaration. The case was eventually considered by the GMC's Professional Conduct Committee; the doctor was found guilty of serious professional misconduct; and his name was erased from the medical register.

Several cases followed. Some were referred to the ABPI only after a series of company procedures, others were referred just as soon as suspicion of an irregularity had been aroused. By the time I retired from the ABPI in 1996, I had handled 12 confirmed cases of research misconduct, all of which were referred to the GMC by means of statutory declarations. Subsequently, many more cases were referred direct to MLI, because we had inspired confidence within the pharmaceutical industry and between us had a track record of cost-effective, rapid and expert forensic investigation, and at least another 17 ended up being referred to the GMC. Many other referrals turned out to be examples of sloppiness or misunderstanding, not of fraud or misconduct. The companies concerned soon learnt not to use these doctors again. Referral to the GMC only occurred where data seemed to have been generated with an intent to deceive. All of the cases were handled in the strictest confidence: those doctors under suspicion, but who were not guilty of any misconduct, never knew they were being investigated and were found innocent; all the more serious cases, where the suspicion was justified, only eventually found out that they had been found out when a letter arrived from the GMC setting out the case against them. All of the cases dealt with in that way were referred to the Professional Conduct Committee or its successor, the Fitness to Practise Committee. All but one were found guilty of serious professional misconduct and dealt with appropriately. The procedure we used worked well, and, indeed, still does, though few cases have been referred recently.

Two cases involving two different universities are worth describing in detail in the context of this series of articles. The first is that of a consultant physician at the Western General Hospital in Edinburgh, a distinguished doctor who had previously served as an office holder in one of the medical royal colleges.^13,14 A clinical trial monitor working for one of the major pharmaceutical companies had noticed that several patient signatures on the study consent forms differed from their signatures in the hospital notes. The company drew the attention of the ABPI and the hospital authorities to its concerns. Neither the hospital nor the medical faculty of the university took any action, possibly because the hospital was in a state of considerable geographical flux at the time. By contrast, the ABPI called on MLI to investigate and a number of witnesses were interviewed. It transpired that several of the recruited patients were not aware that they had been put into a clinical trial; their consent had not been sought, let alone obtained. Several other irregularities were revealed and the case was referred to the GMC, which found the doctor guilty of serious professional misconduct. It is not clear what the conclusion of this case might have been if the independent forensic team (MLI) had not been available. At best, it would have taken considerably longer to reach the same outcome.

The second case concerned a professor of psychology at a university in the United Kingdom. He had devised a frequently cited model for making rats either depressed or stressed.¹⁵ The model was based on the drinking of a glucose solution by the rats: if they were stressed, they drank less. He asked a PhD student to be an investigator in an early trial of a new anxiolytic produced by a French pharmaceutical company, using the model to see if the stressed rats became less stressed after ingesting the anxiolytic. She duly began the trial, dividing the rats into subjects and controls, using the model devised by the professor, but could not make the model work, as the so-called stressed rats drank exactly the same amount of glucose solution as the control rats. The professor asked the student to repeat the study, telling her that the rats came from different breeding sources, and to switch the stressed and control groups round. There were still no differences. Undaunted, the professor instructed the student to proceed with any of the rats that had been subjected to stress, and to give them the new anxiolytic. They drank as much glucose as the control animals but, as the student had pointed out, they were never demonstrably stressed in the first place. Despite this, the professor submitted a report to the sponsoring company on the success of the new anxiolytic in eliminating stress in the rats.

The student expressed her concerns to her tutor, who shared them, and they duly challenged the professor, who took no notice. They then went to the relevant senior officer within the university, who effectively told them to go away as the professor had an international reputation and who were they to challenge him! The student and her tutor then turned to MLI. At the same time the student decided to take her PhD studies to a different university, over 200 miles away.¹⁴ I was closely involved in this case and suggested that the validity of the rat model be tested by the student for the benefit of her new professor, who confirmed that it did not work. I also visited the sponsoring pharmaceutical company in France, where there was incredulity that an eminent psychologist, whose model for stressing rats was widely cited, might be misleading them. The evidence from the second university nevertheless confirmed the likelihood that the results from the first university were flawed because the rats had never been effectively stressed before being given the anxiolytic. The company disregarded the results of the research from the first university and took their subsequent studies elsewhere.

The inconsistencies and anomalies revealed in this case could not be accounted for by chance, nor did there appear to be other innocent explanations. There appeared to be prima facie evidence of falsification of data with the intent to deceive, and thus serious professional misconduct. At that time there was no regulatory body (comparable to the GMC for doctors) for psychologists, so MLI submitted the case to the vice-chancellor of the university. The outcome of this case was far from satisfactory, but the professor eventually retired. Had there been a proper mechanism in place within the university for the consideration of such cases, much of the time and effort spent internally on trying to come to terms with an irregularity would have been saved and much useless research would have been avoided. The forensic team had nevertheless produced robust evidence, quite fast, that a report to a sponsoring pharmaceutical company had been false.

So, where do we go from here?

So I come to the CNEP case, in which MLI was involved transitorily, at an early stage. We were approached early in 1997 by one of the families alleging forgery of consent forms. We spent a day with the person who had contacted us and emphasized that these allegations were very serious and needed to be confirmed or refuted. We said that MLI could undertake the investigations needed, but that we were not able to do so philanthropically. As the family who had contacted us was not in a position to fund this exercise, and the regulatory authority investigating the allegation never asked for help, MLI's involvement ceased at that point.

Looking dispassionately at these three cases, what could have been done differently and how might outcomes have been altered if these differences had been made? The early successes of MLI were because we were recognized as a small dedicated team of experts in the handling of suspected or alleged research misconduct, in whom pharmaceutical companies could have confidence. We were always funded by the sponsor who had called us in; we were scrupulously confidential in our dealings with clients, witnesses and patients; and we were fast. The cases we submitted to the GMC were, beyond all peradventure, likely to lead to a finding of serious professional misconduct. As a result, we were told we had inspired confidence in the veracity and integrity of our submissions among members of the Professional Conduct Committees, who heard the cases we brought forward. Subsequently, however, the climate changed and the GMC increasingly insisted on its own solicitors being more involved in the work-up of cases, which added considerably to the delay that already occurred between submission and hearing.

The confidence that pharmaceutical companies had in our ability to do a complete and effective investigation on their behalf, including many cases where we concluded that there was no case to answer, was never reflected by other outside bodies, including the universities, although one NHS Family Practitioner Committee did use our services to investigate a series of false claims made by one of its GPs. The National Health Service, unlike the University Grants Committee, does at least have a Counter Fraud Service, but this freely admits that it does not deal with research fraud, only financial fraud. Many will, of course, find the distinction hard to grasp. Indeed in America it was precisely because federal money was underpinning at least some aspect of almost all research activity that central Government felt able to tell the Office of Research Integrity that they ought to investigate what was going on if no adequate investigation had been organized locally. The attitude in the UK has always been that universities, NHS Trusts and the medical research charities all have their own internal procedures, which can be invoked in time of need. Paradoxically, because that need is, fortunately, rare, the procedures become rusty and there is no recently used expertise available to activate such procedures with any degree of confidence.

So is this where UKRIO should come in? Certainly it appears that recently it is raising its profile and is able to offer good advice on what to do if it is suspected that something in the research context has gone wrong.^7,10 But UKRIO has no intention of investigating any cases itself, having encouraged the various stakeholders involved in research to have standard operating procedures in place on how to conduct such an investigation. Other countries in Europe have also adopted the same approach at present. However, in the UK at least, the evidence suggests that this is not working as it should. I reiterate that one of the pivotal recommendations of the original ‘blueprint’ was the need for a recognized rapid response to any allegation, with confidential external investigation using teams from national lists of trained expert assessors who could be called in by institutions as required.⁴ As already stated, only MLI currently fulfils this criterion of trained expert assessors, but it was never intended to be exclusive. However, it does exist, as a unit to be invoked whenever required. If UKRIO were to invoke such experts or, more usefully, to recommend that institutions should call in the experts themselves at the earliest opportunity, then the future for the management of research misconduct may become less bleak than it currently appears.

Footnotes

DECLARATIONS —

Competing interests None declared

Funding None

Ethical approval Not applicable

Guarantor FW

Contributorship FW is the sole contributor