Why I... believe we need to do more to combat fraud in scientific research

Fraud in a scientific journal can range from having the same material published more than once, to plagiarism or the fabrication of data. We do not know how common it is. Some of us who take an interest in the matter get alarmed because, as editors, most of the research and publication misconduct we do detect is serendipitous.

For example, I received a paper from a group outside the United Kingdom. I sent it to a reviewer who wrote back to say that about 80 per cent of it was plagiarised from two other papers, both of which happened to be by him. If the paper had gone to another reviewer, it might not have been picked up. In the past three years, Cope, the Committee on Publication Ethics that I chair, has looked at more than 100 cases.

Prevention is better than cure and one of Cope's objectives is to increase awareness. It sounds ridiculous, but some people just do not seem to know the rules of the game. The accepted protocol is that you do not send your paper off to three journals at the same time. Some people do not seem to realise even that. There is a lot of education to be done, not only of researchers but of their supervisors. The rules about regular supervision meetings, checking laboratory work books and so on need to be taken seriously.

We also need to create a climate in which whistleblowers are not hung up and left out to dry. In the case of the recent Gut incident*, there were two whistleblowers. One of them has been deeply damaged by the affair. And it is not just a question of legal protection. Institutional isolation is also a problem. Most of us think that what editors and reviewers pick up is nothing by comparison with what whistleblowers could tell us. In my experience, most fraudulent laboratory workers are suspected by colleagues months or years before what has been happening comes to light. But they are reluctant to say anything and internal inquiries may just hush things up.

Some researchers say it does not matter. They are fearful of more bureaucracy impeding important research. Some say most research does not have much impact on patients' lives anyway. And in the end, someone else will prove that the finding is wrong so everything will be fine.

I do not want more bureaucracy, but I would argue against this view on a number of grounds. One is that we cannot tolerate a culture of dishonesty. The whole of scientific research and the practice of medicine is based on honesty.

But the most important thing is that harm can be done before the truth comes out. The British Medical Journal recently had to retract a paper published a number of years previously that had already begun to change funding priorities and influence the delivery of a clinical service.

We need some kind of national advisory panel. The United States and some of the Nordic countries have had bodies like this for almost ten years. As I point out in the current issue of Gut that carried the retraction, it took 20 years longer to establish research ethics committees in Britain than it did in the US.

In 1999, the Royal College of Physicians of Edinburgh organised a consensus conference on misconduct in biomedical research. It recommended that a national panel should be set up to promote good practice, help with the investigation of research misconduct, and collect data on the incidence of it. Discussions have happened, and a report is apparently being prepared, but we have not yet seen it.

Unfortunately, if we do get a body, but it is one with no teeth that simply presses for internal inquiries, then these issues will continue to be buried.

*The latest issue of Gut retracts a paper published ten years ago but now deemed to have included false information.


Michael Farthing
Executive dean of medicine at the University of Glasgow and editor of the medical journal Gut