Is Science a self-correcting process? I (and the Institute of Medicine) think not.

For years, many scientists and doctors have argued that it was not necessary to police conflicts of interest and other irregularities in scientific research, Money, they argued, couldn't possibly taint their scientific judgment. And even if it did, science was essentially a "self-correcting" process. If some researchers allowed bias into their work and skewed their data as a result, the thinking went, other scientists would eventually expose them because they would not be able to reproduce their results.

Unfortunately, this long-held dogma turns out not to be true (if it ever was). As science becomes more complicated and specialized with dozens of authors weighing in on a single paper, no one has time to reproduce the vast amount of results reported in scientific papers. Nor do most scientists have the inclination to prove wrongdoing, especially when the results favor their own purposes or pocketbooks. In her fascinating new book, Plastic Fantastic, science writer Eugenie Samuel Reich explores the notorious case of Hendrik Schon, a young Bell Labs researcher from Germany who forged data on a number of significant papers detailing supposed breakthroughs in high-performance computer transistors. Reich concludes that in this case, the "self-correcting process" of science turned out to be far less systematic than many scientists think it is. Schon's forgeries were eventually caught in 2002 not by researchers in his own lab but by a few skeptical outliers in other labs who transformed themselves into whistle blowers and refused to accept the assurances of Schon's supervisors that his work was genuine.

Schon, Reich reveals, began fudging data as a graduate student and just dug himself deeper and deeper into a fantasy world in large part to please his supervisors and cover up his inadequacies as an original thinker. While Schon's work (if it had been authentic) certainly had commercial implications, he wasn't necessarily motivated by commercial pressures, she concludes.

The same, however, cannot be said of clinical drug research, where there are tremendous commercial pressures -- to report favorable results about a new drug or medical device. And that is why the national Institute of Medicine (IOM) has weighed in so decisively in its report on conflicts of interest in research and what needs to be done to ensure the integrity of the scientific process. A good overview of the IOM report and its conclusions can be found in this week's New England Journal of Medicine.

Basically, the IOM report calls for Congress to pass legislation (like Grassley's Physician Payment Sunshine Act) that would require pharmaceutical, medical device and biotechnology companies to publicly report payments to physicians and other health care professionals. But the IOM recommendations go beyond the Sunshine Act and call for legislation that would also require professional societies and patient advocacy groups, like NAMI, which have their own blatant conflicts of interest -- see here -- to publicly disclose them.

The IOM report also concludes, as I have argued in Side Effects, that disclosure is not enough. "Academic medical centers and research institutions should "restrict participation of researchers with financial conflicts of interest in research with human participants," the report says. That means if a particular researcher is getting lots of money in consulting and speaking fees from a drug or medical device company, he or she should not be involved in clinical trials of products made by that company. That would rule out the participation of many of the conflicted doctors I've written about here, including Martin Keller, Charles Nemeroff, Alan Schatzberg, Karen Wagner, Graham Emslie, Robert Robinson. The list goes on and on.

The IOM report also calls on medical societies to reform their continuing medical education programs so that they are "free of industry influence..." And it urges physicians to reject gifts or other "items of material value" from pharmaceutical, medical device and biotechnology companies. It also suggests that advisory groups that issue clinical guidelines about medical treatment restrict industry influence and conflicts of interest among panel members.

In sum, the nation's most respected medical body is acknowledging that financial conflicts of interest corrupt scientific research and harm public health in the process. And even in the absence of financial conflicts, scientists have to do more to ensure the integrity of the scientific process. Trust alone just doesn't cut it.