Journal Editors Are Urged To Demand More Evidence
Wednesday, November 29, 2006
Editors of scientific journals should beef up their level of skepticism about high-profile papers submitted to them and demand solid evidence that the work was completed as described, according to a review of the editorial procedures that led to the publication of fraudulent scientific papers by the now-disgraced South Korean stem cell scientist Hwang Woo Suk.
The review, released yesterday, was commissioned by Science, the American journal that published Hwang's two most important faked papers in 2004 and 2005 and then retracted them in January after investigations in Korea and the United States concluded that the results were almost all fabrications.
The fraudulent articles described what appeared to be the first creation of embryonic stem cells from cloned human embryos and the production of stem cells genetically matched to patients -- two holy grails of medicine.
Fraud will always plague the scientific enterprise, the review concluded, and it may even be on the rise, given the growing amount of fame and fortune that can accrue for those associated with major advances in science and medicine.
But more fraud could be caught in advance -- and its negative impacts on research and on public confidence in science could be lessened -- if journals reviewed submitted articles more carefully, the six-member expert panel reported.
"Progress in science depends on breakthroughs and in taking risks, both in research and in publishing," they wrote. "Nevertheless, it is essential to develop a process by which papers that have the likelihood of attracting attention are examined particularly closely for errors, misrepresentation, deception or outright fraud."
Foremost, the panel declared, Science should create a "risk assessment" process designed to flag papers that, because of their potential impact on science or public policy or their counterintuitive findings, deserve an extra dose of review.
In addition, the panel recommended that scientific co-authors be required to specify how they contributed to the work, and that more original data be included to help verify the authenticity of new findings.
Although the panel limited its investigation to how Science handles submissions, it suggested that editors collaborate with counterparts at other journals to harmonize the new protections, so scientific crooks do not simply submit their work to journals with fewer safeguards.
Experts in misconduct said the recommendations were fine but broke no new ground.
"These are good points and we should go forward with them, but they will have very little effect on the integrity of science," said Adil E. Shamoo, a professor at the University of Maryland and editor-in-chief of the journal Accountability in Research. Shamoo has called for required training in research ethics for scientists and full laboratory visits and audits of raw data by their peers. "It's ludicrous to think editorial boards can do all this," Shamoo said.
George Lundberg, editor-in-chief of the medical journal MedGenMed and of the online medical text eMedicine, also criticized the report as "business as usual."