A solution to the problem of scientific misconduct?

Talk about synergy. Just a few days ago, my health and science journalism class at Brandeis was discussing ethics in science, using as a template an old case example involving scientific fraud in the University of Michigan laboratory then headed by Dr. Francis Collins. Back in 1996, Collins, who is now director of the National Institute of Health (NIH), investigated the allegations as soon as he was alerted to a tampered image in a paper he had submitted to the cancer journal Oncogene. After discovering that one of his graduate students was guilty of systematic fraud in five papers that he and Collins had co-authored, Collins alerted university and NIH authorities and fired the culpable student.

I've blogged before about this 1996 case here, and I teach this case not to cast aspersions on Collins, but to examine important questions about how well our current system of scientific publication (in which ambitious and overextended senior researchers routinely slap their names on scientific data they are not intimately familiar with) works in rooting out errors and outright fraud. I have also blogged about the issue here and here because now that Collins is director of NIH, he has the power to really make a difference. And yet I haven't heard peep from him about the pervasive problems in the research system, ranging from financial conflicts of interest among medical researchers to the widespread practice of ghostwriting, all of which are impairing public confidence in the scientific process.

Which is why I was gratified to see that a group of 96 ethicists, researchers and clinicians have signed and sent a letter to Collins asking him to fund more studies on these ethical conundrums. As the letter from Pharmed Out notes:

NIH funds a substantial portion of the generation and dissemination of evidence, but the uptake of that evidence and its translation into clinical practice is strongly affected by the complex web of relationships that exists among industry, academicians, medical educators and clinicians. There is growing evidence that each strand of this web is compromised by ethical lapses and financial conflicts of interest. The recent disclosure of ghostwritten articles, physician payoffs, and the use of academic opinion leaders to increase markets for FDA-regulated products indicate that ethical lapses may permeate biomedical research.

I am eager to hear Collins' response to this appeal, spearheaded by Dr. Adriane Fugh-Berman, an associate professor at Georgetown University Medical Center. In the meantime, I would like to tender an idea proffered by one of my students during class (we were discussing whether there is a need for a standard system of quality control for scientific research independent of the peer review process). My student, who by the way is pre-med, suggested that perhaps an auditing system should be put in place by the NIH or the FDA to randomly examine data in scientific labs around the country, much like the IRS randomly audits tax returns. These audits would be of the actual raw data that have led to published research in key medical and scientific journals and the very nature of their randomness might just propell scientists to clean up their act.

I think this is a brilliant suggestion but of course I think all my students are brilliant. What do you think, Dr. Collins?