Call for agency to stop medical research fraud

A UK agency to police allegations of scientific fraud and protect whistle- blowers who inform on colleagues was called for yesterday by Britain's top two medical journals.

In a unique collaboration, the Lancet and the British Medical Journal produced separate leading articles demanding action as a new book detailed more than 70 proven or suspected cases of scientific dishonesty and fraud worldwide.

Britain's medical profession "harbours a culture that prefers to sweep such problems under the carpet", Dr Richard Smith, the BMJ's editor said, while Dr Richard Horton, the editor of the Lancet, declared that "professional self-regulation has failed".

If Britain's medical and scientific community - the Royal Colleges, the General Medical Council, the Royal Society, the research councils and other bodies who fund research - do not act this time, the government will have to, Dr Smith warned. But an attempt last year to set up such a body foundered when representatives of the Royal Society, the Medical Research Council and the General Medical Council were "scornful" of the proposal, Dr Smith said.

The call has been given topical urgency by the General Medical Council's decision last week to strike off Dr Geoffrey Fairhurst, a Merseyside GP who was vice-chairman of a medical ethics committee and a former government adviser, who forged patient's consent forms for drug company- sponsored trials which earned him at least pounds 15,000.

It also follows the case last year of Malcolm Pearce, the consultant obstetrician at St George's Hospital, south-west London, who was struck off after falsely claiming to have relocated an ectopic pregnancy - one that begins in the fallopian tube, which transports eggs to the womb - into the womb, producing a successful birth. He also published false claims about a trial involving 191 women in a journal which he edited.

Fraud in medical research is now "pervasive", the Lancet said. But the response to scientific misconduct in the UK and many other countries "has been one of casual indifference".

An agency to investigate fraud, review complaints and establish the true prevalence of the problem is "urgently needed", it added.

The article went on to ask: "How many more patients will have to be mistreated and how much more dishonest research will have to be conducted and published before our complacency finally damages the reputation of the medical research community beyond repair?"

The United States, Denmark, Norway, Finland, Austria and Australia all now have "coherent systems" for tackling the problem, Dr Smith said. The US system is a Commission of Research Integrity, set up by Congress and the Department of Health because of the scientific community's inability to deal adequately with misconduct".

Allegations must be fully investigated, with both whistleblowers and those they accuse treated with "respect, fairness and openness".

A commission could both provide guidelines for institutions and individuals, offer impartial investigation, collect data and sponsor research. Although accurate data are lacking, the Lancet said, fraud seems to affect between 1 in 1000 and 4 in 1000 research studies.

Personal ambition, greed, the need to publish to gain recognition and promotion, and league tables of university performance where research income and output is divided by the number of staff all contribute to fraud.

Such frauds can range from the criminally invented, non-existent, study, to minor and sometimes unwitting plagiarism, to simple carelessness, to publishing the same results in different journals and being given authorship of papers to which the researcher has contributed little or nothing, John Howie, Professor of General Practice at Edinburgh University, says in an introduction to his book, Fraud and Misconduct in Medical Research - the work which has prompted the articles in the medical journals.

The lack of a career structure for research staff may also contribute, where those in mid-career become expensive to fund but need to continue to appear innovative.

"The temptations are obvious," Prof Howie said. The "fundamental" change needed is to shift the research climate firmly back "from quantity towards quality".

From Piltdown Man to fake drug trials

Scientific fraud dates at least to Piltdown Man who was "discovered" in 1912 and to the charges in the 1970s that Sir Cyril Burt invented data for his seminal study on twins and intelligence. Cases in the past two decades include:

t William Summerlin, an immunologist at the Sloan-Kettering Institute in New York, who in 1974 faked transplant results in white mice by using a black pen to produce darkened skin patches. If true, the finding would have had huge implications for organ transplantation and immunology.

t William McBride, an Australian scientist famous for helping expose the thalidomide scandal, who in 1988 was found to have forged data on the action in fetal rabbits of hyoscine, a drug used as a pre-med and anti-emetic before operations. His data appeared to raise a major threat to the drug Debendox.

t Mark Spector, a 24-year-old Cornell University graduate student was found in 1981 to have faked data on a viral cause of cancer that colleagues predicted would win him the Nobel Prize.

t Since 1975, 26 cases involving 16 doctors - all but two GPs - have been reported to the GMC for faking data in drug company sponsored trials. Doctors have made up data, faked patients' dates of birth, forged consent signatures, and forged signatures of other participating doctors.

Fraud and Misconduct in Medical Research; Stephen Lock and Frank Wells, BMJ Publishing Group, pounds 29.95