The following originally appeared on The Upshot (copyright 2014, The New York Times Company).
When the Food and Drug Administration creates an advisory committee to help it decide whether to approve drugs, it often asks academic physicians to serve on the committee as external experts. This is supposed to help the committee render judgments that are unbiased and scientific.
Physicians can therefore hold a fair amount of sway in how decisions are made. One would hope that they would be above reproach, coming to the committee’s deliberations with equipoise and an open mind.
A study published today by Genevieve Pham-Kanter in The Milbank Quarterly brings that assumption into question. She reviewed the voting behavior and financial interests of almost 1,400 F.D.A. advisory committee members who took part in decisions for the Center for Drug and Evaluation Research from 1997 to 2011. Over this time, 15,739 votes took place in 379 meetings.
On average, 13 percent of participants on each committee had some reported financial interest in a drug company whose product was up for a review by that committee. About half of all meetings had at least one participant with such a financial interest. About a third of these interests involved consulting of some kind, but one quarter included an ownership interest, and 14 percent involved a committee member who served as a member of an industry advisory board or steering committee.
She found that over all, committee members had a 52 percent chance of voting in favor of a sponsor of a drug. But members who had financial interests in only the company whose product was under deliberation were more likely to vote for its approval, with a probability of 63 percent.
If members served on advisory boards for only the company whose product was up for review, then the chance they would vote in favor of it shot up to 84 percent. (Members who had financial interests in, or served on advisory boards for, both the company whose product was being reviewed and at least one of its competitors were not more likely to vote in favor of any particular company’s drug, however.)
It’s hard to look at data like this and not be concerned about conflicts of interest. There’s a reason that 10-cent coupons exist; it’s because they work. Financial interests absolutely do influence our decision making. Since 2008, the F.D.A. has worked to reduce the number of committee members with financial conflicts of interest; the Pham-Kanter study indicates that this effort has met with significant success. Other conflicts of interest, like professional or ideological ones, can also influence our behavior, but these have not been as well studied. And the F.D.A. is not the only place where financial conflicts of interest are a concern.
Financial relationships between doctors and industry are not uncommon. In 2007, research showed that 94 percent of physicians in the United States had such relationships. More than 80 percent of doctors had accepted gifts, and 28 percent had received payments for consulting or research. Sixty percent of those physicians were in medical education, and 40 percent were involved in writing practice guidelines.
Physicians sometimes travel to nice resorts for education. A study published some time ago in the journal Chest followed doctors who went to two all-expenses-paid symposia on new drugs. Eighty-five percent of the physicians interviewed stated that accepting such invitations would not influence their use of the drugs. Nevertheless, their prescriptions for those drugs nearly tripled after the meetings, far above increases in the use of those drugs nationally. Other studies have shown that physicians who meet with and accept gifts from drug companies are significantly more likely to ask that their drugs be added to hospital formularies.
In the past, potential conflicts of interest were not discussed widely in public. Many patients were unaware that physicians accept gifts from pharmaceutical companies. When they learned that such exchanges occurred, they reported that it “altered their perception of the medical profession.” Patients also feel that gifts are more influential and less appropriate than doctors do.
This sentiment is one of the reasons that the Physician Payments Sunshine Act was passed with bipartisan support. The law, which went into effect last year, requires that almost all payments made by industry to physicians be reported to a public database. These include meals; travel; honorariums for speaking; grants for research; and ownership in companies. It’s hoped that by making these types of relationships public, we might mitigate their effects.
Doctors have had mixed reactions to these changes. Data like those now required by law to be reported were recently used to show that top Medicare prescribers of the expensive drug Acthar had financial ties to its maker. The doctors mentioned by name in that article probably did not appreciate their loss of anonymity. Many others have begun to limit their acceptance of gifts, knowing that they will be made public.
Some physicians, especially those opposed to the Sunshine Act, believe that they should be responsible for regulating themselves. But our thinking about conflicts of interest isn’t always rational. A study of radiation oncologistsfound that only 5 percent thought that they might be affected by gifts. But a third of them thought that other radiation oncologists would be affected.Another study asked medical residents similar questions. More than 60 percent of them said that gifts could not influence their behavior; only 16 percent believed that other residents could remain uninfluenced.
This “magical thinking” that somehow we, ourselves, are immune to what we are sure will influence others is why conflict of interest regulations exist in the first place. We simply cannot be accurate judges of what’s affecting us.
Conflicts of interest are real, and they are still influencing decisions from the level of the patient all the way up to national health policy. We will never be able to eliminate them all. But acknowledging them and talking about them openly is an important first step toward minimizing their impact.