Richard Nicholson argues that the requirement that most doctors undertake some research pushes them towards unethical behaviour

At long last the medical establishment, or parts of it, are beginning to take research fraud seriously. The Committee on Publication Ethics has just published its first annual report. In its first year, the committee, set up by concerned medical editors, examined 25 cases - some admittedly minor.

The General Medical Council has also formed a committee, chaired by George Alberti, president of the Royal College of Physicians, to consider how to respond to research fraud. The GMC has been the only medical institution taking the subject seriously. In the past nine years, it has struck 17 doctors off its register for fraud.

Some years ago, the RCP published guidance on how to handle suspected fraud. It had little effect because so few in the medical establishment would acknowledge any problem. Quite why the United Kingdom should be so different from the United States and Scandinavia, which have produced a steady trickle of research fraud since the mid-1970s, was never explained. That the UK is no different has now been demonstrated - not by the medical profession but by the pharmaceuticals industry. Most cases brought to the GMC were uncovered by clinical auditors employed by drugs firms to ensure the integrity of the data used in seeking drug licences.

³ÉÈËVRÊÓƵ

There are various forms of research fraud. The most common is redundant publication, when the same data, or even the same paper, is published more than once in different journals or is divided into many small papers, which is known as "salami slicing". COPE estimates that salami slicing accounts for 13 per cent of all published papers. Deliberate fabrication of data is harder to detect, and there are no accurate estimates of its frequency. The GMC cases are nearly all of this type: doctors, paid so much per patient entered into a trial, have thought it easier to invent data than to do the trial. Nor do they learn from other doctors' punishment: Dr Frank Wells, of MedicoLegal Investigations, reports several cases in the pipeline for disciplinary action.

Plagiarism is a third form of fraud: it tends to be discovered by accident, although sometimes it is so blatant as to be asking for detection. A friend of mine, a cardiology senior registrar, did a neat piece of research, showing that a drug licensed for one use was also effective for another cardiac problem. He submitted a paper to three journals, all of which rejected it. His boss was so surprised that he rang the three editors to protest and discovered that the same reviewer had read the paper for each journal. Nine months later the reviewer published an identical paper: the only action taken was that the journals agreed not to use him again.

³ÉÈËVRÊÓƵ

Against this background of inaction, the recent developments are welcome. However, they focus only on the detection and punishment of fraud. No one is asking what are the causes of research fraud and how it might be prevented.

Some cases seem to be the result of financial greed. In others, academic recognition seems the primary objective. But has the whole structure of medical research now become more conducive to fraud? That almost every health professional now has to undertake a research project, either for degree completion or to ensure progress up the career ladder, has moved the motives of many researchers away from a desire to find important new knowledge to help others and towards the idea that research is done just for one's own benefit.

Few research projects make any difference to patient care: even fewer, probably less than one in a thousand, produce major advances. By reducing the amount of research being done and by concentrating it in fewer centres and in the hands of those who really want to do it, we might both reduce fraud and other unethical uses of research subjects and have more productive research.

Dr Richard Nicholson is editor of the Bulletin of Medical Ethics.

³ÉÈËVRÊÓƵ