Fraud in the laboratory

1/8/2006

In Pennsylvania, the University of Pittsburgh is investigating Dr. Gerald Schatten's role in the bogus stem-cell research that has tarnished the reputation of this promising new field of science. In Korea, Seoul National University is investigating Dr. Hwang Woo-suk, who headed the research team there.

Investigators seek to establish the roles of the individuals involved, and also may determine whether institutional changes are needed at Pitt and in Korea to strengthen safeguards on scientific integrity.

What about the third participant - the journal Science?

Science published the research paper that announced the supposedly stunning new breakthrough. Dr. Hwang's team claimed development of human embryonic stem cells customized to individual patients, a major advance in using this controversial technology in treating diseases. All 11 kinds of stem cells described in that paper actually were fakes, Korean investigators have concluded.

Major journals like Science are not potted plants in the process of scientific discovery. They are not passive onlookers that simply publish research. Journals are a critical checkpoint for assuring the accuracy of research.

Journals use a process in which in-house editors and outside panels of experts evaluate each manuscript. They check for mistakes and other flaws before publication, and make sure the researchers met high scientific standards. As a result, editors reject some manuscripts outright or tell scientists to go back to the lab and do better work.

Something went amiss with that review process for the stem-cell report. Had peer review worked, stem-cell research, Pitt, and Seoul National University might have been spared a black eye.

The problem is not unique to Science, perhaps the world's most prestigious general scientific journal. Faulty or fraudulent research slips into print in other journals all too often. From there it quickly reaches the public via the news media.

It is not enough for a journal to simply retract debunked research papers after publication, as Science now is doing for the stem-cell paper. Scientific journals simply need to get better at spotting fraud.