Putting medical research results to the test

January 12th, 2015

A scientific startup in California is launching a project to reproduce results from medical studies.

Reports from Bayer Healthcare, Amgen, and other groups concluded that results from published medical studies cannot be reproduced. Science Exchange is beginning its Reproducibility Initiative for scientists who want to confirm that study findings are reliable before publishing.

A report published in March from former head of global cancer research at Amgen described attempts to reproduce findings from 53 cancer studies. Only six could be replicated.

"These are the studies the pharmaceutical industry relies on to identify new targets for drug development," said C. Glenn Begley, now senior vice president of TetraLogic, in a Reuters interview. "But if you're going to place a $1 million or $2 million or $5 million bet on an observation, you need to be sure it's true. As we tried to reproduce these papers we became convinced you can't take anything at face value."

Other studies have also shown that research results can be difficult, if not impossible, to replicate.

Those scientists willing to have their research put under the proverbial microscope can apply to Science Exchange, and a 10-member scientist board will match the studies up with a lab to best fit the study's needs. Results from the verification studies will be published in the journal PLoS One.

Critics of the rash of unreliable results point to the desire to be published in a top journal as incentive for scientist to stop gathering data when the desired result is produced, or to keep gathering until said result is achieved. Others point out that cancer study results can be extremely difficult to replicate. Different conditions can produce very different results, particularly with cancer cells.