Is the spirit of Piltdown man alive and well?

The celebrated hoax was thought to be a one-off but scientists are still under pressure to break the rules, says Stephanie de Bono

Ever since the Piltdown man was shown to be a hoax about half a century ago, science has been haunted by the spectre of fraud.

By and large, most researchers have felt themselves part of an honourable tradition of being seekers after scientific objectivity. And examples of trickery and deceit have been far and few between.

However, recent studies have shaken this view and challenged it as at best complacent, at worst misleading. The major scientific crimes of fabrication, falsification and plagiarism may be only the tip of the iceberg and there is evidence of a much wider and deeper problem, not of outright fabrication of results but of distortion, omission and exaggeration.

This has become such a concern that a new body - the UK Panel for Health and Biomedical Research Integrity - has been set up to advise on the extent of malpractice, from minor to major, and how to tackle it.

"Most current published research findings are false," reports one analysis of theoretical, clinical and molecular research in the online journal Public Library of Science (PloS): Medicine last month. It began with a study of 49 research articles that had been cited by other scientists 1,000 times or more and revealed 14 of them -almost a third - were later refuted by other work.

After modelling the source of error mathematically, the analysis concluded that even a large, well-designed study with little bias has only an 85 per cent chance of being right. An underpowered, poorly performed drug trial with researcher bias has but a 17 per cent chance. Overall, more than half of all published research is probably wrong.

Another recent survey from the journal Nature reported that up to one third of several thousand National Institutes of Health scientists in America were engaged in "questionable practices". As expected, when falsifying data and plagiarism were present they were rare (0.3 per cent and 1.4 per cent of scientists respectively). However, other "minor" misdemeanours not previously investigated "are striking in their breadth and prevalence".

By their own admission five per cent of respondents had published the same results in two or more publications, six per cent had failed to present data contradicting their own research, 10 per cent had given inappropriate authorship credit, 15 per cent ignored data on the basis of a gut feeling that it was wrong, and 15.5 per cent had changed the design, methodology or results of a study in response to pressure from a funding source.

"To the extent that the environmental circumstances of science are similar here to what they are in the US, I would suspect that similar issues of misbehaviour might exist in Britain," says Dr Brian Martinson of the HealthPartners Research Foundation, Minnesota, and lead author.

In Britain there has been no equivalent study of scientific misdemeanour. However, an increasing number of these cases have been examined by the Committee on Publication Ethics' (Cope), which was set up in 1997 for editors of science journals who were struggling to deal with breaches in research and publication ethics.

The most common transgressions Cope has had to deal with are "lesser ethical problems" such as duplicate publishing of results and failure to disclose conflict of interest. But it has also dealt with rare but florid cases of apparent fraud, including two that were recently highlighted in the British Medical Journal.

"There is certainly more misconduct out there than was thought," says Fiona Godlee, chair of Cope and editor of the BMJ. "Editors have a responsibility to pursue allegations of fraud, but their resources and remit are limited. The academic establishment in Britain has not taken the problem seriously enough".

To meet these concerns, the UK Panel for Health and Biomedical Research Integrity has been set up, backed for three years by the Medical Research Council, Biotechnology and Biological Sciences Research Council and Department of Health. The project, an experiment to see what can be done to keep scientists on the straight and narrow, has been spearheaded by Professor Mike Farthing, Principal of St George's Medical School, London, and founder of Cope. Although the panel stops short of actually carrying out investigative work, it will support institutes in tracking down and dealing with a spectrum of scientific misbehaviours.

With up to a third of researchers admitting they routinely commit scientific sins, it is important for the new panel to understand how the current scientific environment fosters and permits misbehaviour, so that it can halt future outbreaks. Some of the factors are already known. Burdens on modern scientists include fierce competition for scarce resources.

Precious funding for salaries, laboratory space and equipment must be fought for. And, as ever, there is the pressure to publish or be damned.

Scientists are being pressurised to tailor their output to that primary performance indicator, and gauge of scientific success, the number of publications weighted by the impact factor - perceived importance - of the journals they publish in. This fosters behaviour such as deriving tenuous links to human disease, salami- slicing of results, spreading them over as many papers as possible, and including authors who contributed little to the research.

"It has become the aim of most scientists to get high up in these evaluations," says Dr Peter Lawrence, group leader at the Medical Research Council Laboratory of Molecular Biology, editor of the journal Development for 30 years, and past editorial board member of Cell and EMBO Journal.

Scientists are therefore "strongly tempted to oversell their work". "Most PIs [principal investigators] now spend much time networking with editors, not at the bench, and it is they who spend their remaining energy hotting up their students' work for publication, without necessarily having much detailed idea of the work itself, making it easier for them to give it more spin than is justified."

The desperation to flog a paper to the top ranking journals means that scientists bend over backwards to manipulate manuscripts and satisfy reviewers' demands, whether justified or not. In turn, the selective reporting of only the most significant and glamorous findings in the medical literature is a major cause of the larger number of reported false results.

"While currently there is unilateral emphasis on "first" discoveries, there should be as much emphasis on replication of discoveries," says PloS study author John Ioannidis of Tufts University School of Medicine in Massachusetts. However, this is easier said than done, as funding and publication are much more likely to succeed if a discovery is novel or groundbreaking rather than replicative: who wants to come second, after all?

The Royal Society is also preparing to comment on the process of peer review and issues such as conflict of research interests and appropriate authorship in its Report on Best Practice in Communicating Research Results to the Public. The initiative was launched in 2003, in the wake of public concern with inconsistent research reports on issues such as the MMR vaccine and GM crops.

"The consequences of biomedical research are much more immediate for the general public," says Bob Ward, spokesman for the Royal Society. "We need to ensure that there are proper safeguards and commercial interest never overrides the public interest."

The report is expected to examine ways in which the peer review process might be improved, as well as whether there are alternatives to peer review for checking the quality of research results. It will be published later this year.

However, there are doubts whether new committees and hand-wringing reports will be enough to stop scientists spinning their results for reasons of public relations or sponsorship.

"I am not convinced that increasing regulations will help," says Dr Lawrence. "We need a gentle revolution in the way in which scientists are evaluated."