Until about three years ago, researchers at Duke University in Durham, North Carolina, declared their financial conflicts of interest by filling in a simple form. “It basically just asked: do you have a relationship with a company that might entail a conflict of interest?” says James Siedow, Duke's vice-provost for research.

The form has since been changed every year, becoming more complex with each new iteration. Now, Duke researchers are asked to estimate how much money they receive from industry sources. If that value is more than US$25,000, they are required to delve into specifics. Although Siedow believes that most researchers filled in their forms accurately, he suspects that some were not forthcoming. “There were folks who didn't check that '$25,000 or more' box, yet we're almost certain they should have,” he says.

Pressure has been mounting on universities and hospitals to crack down on conflicts of interest. This month, Senator Charles Grassley (Republican, Iowa) informed Congress that three high-profile psychiatrists at Harvard Medical School may have failed to disclose a total of about $4 million of industry earnings over the course of seven years. Last August, Grassley disclosed that a researcher at the University of Cincinnati in Ohio had under-reported industry earnings from a company that made a drug she had tested in clinical trials. And Grassley has informed the National Institutes of Health (NIH) that he is investigating other cases at more than 20 different institutions.

Three high-profile faculty members at Harvard Medical School face allegations of not disclosing hefty industry earnings. Credit: R. FRIEDMAN/CORBIS

The NIH has made its stand on the issue clear: it is up to individual universities and hospitals to monitor their researchers, says deputy director of extramural research Norka Ruiz Bravo. The NIH may punish individual investigators by abridging or terminating their grants, but in cases where the misconduct seems to be systemic, sanctions could be levied against an entire institution.

But university administrators say they are unable to verify what their researchers disclose. “It's an honour system,” says Robert Alpern, dean of the Yale School of Medicine. “We rely on the faculty to tell us the truth. And to be honest, up until a few months ago, I think we all thought they were telling us the truth.”

Grassley has proposed legislation, called the Physician Payments Sunshine Act, that would require manufacturers of drugs and medical devices to disclose how much they pay doctors. The act would present institutions with a way to verify the sums their researchers have declared, and both the American Association of Medical Colleges and PhRMA, a lobbying group for the pharmaceutical industry, have praised the proposed act for increasing transparency. But the current draft of the legislation applies only to physicians and leaves out those researchers who do not practise medicine. Philip Pizzo, dean of Stanford Medical School, says he supports the act but acknowledges that it will come at a price. “This will not be an easy or inexpensive process,” says Pizzo, who notes that schools may need to hire additional staff to process the new data.

Meanwhile, Siedow says that Duke administrators plan to interview researchers they suspect of not disclosing their earnings. Grassley's recent findings have bolstered Duke's efforts to improve its reporting system. “We can see the writing on the wall in Congress,” says Siedow. “We don't want to find out from a pharmaceutical company that one of our researchers made half a million dollars that we didn't know about.”