Cheating scientists in many countries have national committees to answer to, write Peter Lennon and Lesley Cowling, but not in South Africa or Britain
THERE was a time when we stood to attention at the mention of “scientific” research. But the discovery last week of fraud by a researcher at the United States National Institutes of Health (NIH)highlights the danger of an unthinking acceptance of scientific integrity.
The NIH’s Centre for Human Genome Research, Dr Francis Collins, had to withdraw five published studies because some of the data used had been fabricated by a junior scientist working for him on a major leukemia research project.
According to newspaper reports, the researcher confessed to lying about performing experiments and making up data after an alert reviewer picked up a flaw in a paper submitted for publication at a journal. He had worked with Collins for three years, fabricating data, but Collins was quoted as saying he had never had a reason to question his honesty.
It is this kind of incident that has alerted scientific communities internationally to the issue of misbehaviour. It is now regarded with a beady eye by such institutions as the American Office of Research Integrity, which is investigating the fraud allegations in the leukemia research project.
The office has been picking off “Texas sharpshooters” for years. This is the trick of shooting at a barn door and then drawing a target around the hole so it looks like you have done very well indeed. The equivalent in scientific research is fishing through a mass of complex data until you find something unusual and then reporting it as if that is what you were looking for. The researcher is unmasked because he or she does not know how he arrived at his findings.
Professor Paul Riis, founder of the quaintly titled Danish Committee for Scientific Dishonesty, tells how three years ago, following reports of devious activity in the US, they decided “not to go to sleep on the illusion that fraud could never happen in Denmark”.
They set up a committee that looked into allegations of misconduct. When evidence is found, they then put pressure on the university, institution or hospital to take steps. In the first year, 15 serious cases were discovered, in the second five and now it is down to two.
This committee was alerted recently by a Danish scientist who was sent an unsigned report on the result of drug research by an American agency and asked if he would like to put his name to it.
When the committee’s chairman wrote warning the agency that he considered this to be a practice that would undermine the integrity of Danish science, instead of backing down, the American drug company threatened the committee with a law suit.
“Fortunately one of our members is a high court judge,” Riis says, “and he did not have any tremor of the hand. We published the affair and named the company in our annual report, but no action has been taken against us.”
Graham Baker, editor of the South African Journal of Science, says there isn’t a lot of serious fraud – generating false data – in South Africa. “But there is an awful lot of low-grade misrepresentation.”
These include publishing findings in two journals to give the impression of having written two different papers and people putting their names on research they have nothing to do with, especially the heads of university departments who have students working under them.
Baker says that situations sometimes arise where a student feels that he or she has not been given proper credit for work on a paper presented for publication. But it is difficult to work out if there has been misrepresentation, he says, because although the student may have carried out much of the work, the supervisor may have devised the experiment and feel they directed the research.
Dr Cyril Donninger, CEO of the biotechnology company Bioclones, says the problems often arise because of different perspectives on what constitutes work on a paper or research project. “In some universities, it is accepted that the head of department’s name goes on a paper as a matter of course,” he says.
Industry has the Patent Act to sort out misunderstandings. “The Act provides definitions of what constitutes invention,” he says.
But university researchers often haven’t defined intellectual ownership. Donnenger thinks that South Africa does not need a centralised organisation to investigate scientific dishonesty, unless it was a voluntary body meeting a few times a year. “It is the responsibility of what we call `the learned societies’ to provide an ethical code for their members.”
“If there is a case for anything,” says Professor Peter Lachmann, Biological Secretary of Britain’s Royal Society, “it would be to have an advisory body which could advise universities and other bodies on good practice”.
He adds: “But there is fairly general agreement in this country [Britain] that the responsibility for investigating fraud rests with the employers of the scientists concerned, largely the universities and the research councils.” Professor John Bailer, Chair of the Department of Health Studies at the University of Chicago, believes corruption comes from the tremendous pressure on scientists to achieve success.
“The whole scientific reward structure,” he says, “is based on publication. It’s not just a matter of seeing one’s name in print – the system determines salary, promotion, tenure and respect. So there is a lot of pressure to come up with positive results and new findings whether they are actually in the data or not.”
There is another advantage to a committee with responsibility for scientific integrity: it can also defend scientists who have been slandered.
Riis told of a spectacular case recently where rumour accused the rector of Copenhagen University of manufacturing his data on a major research project he had done on brain stimulation of animals while at the University of California at Berkeley.
A Copenhagen cultural weekly, Weekend Avison, published the rumours. Riis’s committee asked Berkeley if they were willing to make a formal accusation. The relevant department agreed. The Riis committee then set up not just a Danish but an international inquiry that completely cleared the rector.
“So,” says Riis, “the case really boomeranged on Berkeley.”
ENDS