Orange County

UNC groups see pattern in pharma perks, cancer-treatment prescriptions

The Old Well on the UNC-Chapel Hill campus.
The Old Well on the UNC-Chapel Hill campus. UNC-Chapel Hill

A string of papers has surfaced in the medical literature over the past few years suggesting that doctors who get perks from a drug company tend to prescribe its products more often. Now, a team at UNC-Chapel Hill says it’s seeing the same pattern in cancer treatment.

UNC School of Medicine oncology fellow Aaron Mitchell and his colleagues were scheduled to present their findings Saturday at a conference in Chicago. Their analysis of “Open Payments” data collected by the federal government suggests a doctor who’s gotten meals or lodging from a drug company is about 78 percent more likely to prescribe its offerings to treat one form of kidney cancer

They’re also about 29 percent more likely to prescribe the perk-giver’s wares to treat a form of leukemia.

That’s noteworthy because while previous studies looked at potential conflicts of interest in the deployment of anti-depressants, cholesterol inhibitors and blood-pressure controls, the pharmaceuticals used in chemotherapy are “far more expensive and far more weighty” in terms of the side effects they can trigger in patients, said Mitchell, the study’s lead author.

For tracking with the earlier studies, the team’s findings reinforce observations of a “pretty consistent correlation between getting industry money and physician practice” along with a need to “continue moving in the direction of transparency and in the direction of managing” doctor/industry relationships to minimize potential bias, he said.

Mitchell, who did his medical residency at Duke University and the Durham VA Medical Center, worked with Stacie Dusetzina, a professor in UNC’s Eshelman School of Pharmacy, and Aaron Winn, a Ph.D. student in its Gillings School of Global Public Health.

The team was extending a line of research that’s figured in several reports that have appeared in the Journal of the American Medical Association’s internal medicine edition and other publications. Like their colleagues at other universities, they’ve relied on the Open Payments database the U.S. Centers for Medicare & Medicaid Services set up to monitor payments from drug companies to physicians and teaching hospitals.

The database is one of the mandates of the Affordable Care Act — the 2010 legislation more popularly known as Obamacare.

Like its predecessors, the UNC team cautions that its methods only establish a correlation between payments and prescriptions, not a claim that one causes the other. But in two AMA commentaries, Robert Steinbrook, a physician and professor at Yale University, Mitchell’s undergraduate alma mater, has questioned the need to wait on proof of a causal link.

“Part of the privilege of caring for patients is to be mindful of their finances as well as their health,” Steinbrook said in the second of his commentaries, published on May 2.

The UNC team looked at kidney cancer and leukemia drugs because for the variants involved, there are each three drugs that seem equally effective in fighting them. They differ, however, in their side effects.

The ones for metastatic renal cell cancer “tend to cause fatigue,” some to a greater degree than others, Mitchell said. Intestinal-tract effects can follow their use, and skin rashes that “can vary from minor to debilitating.” The drugs for chronic myeloid leukemia include one that’s known for sometimes causing lung damage.

In addition to managing a patient’s side effects, doctors might favor one drug over another for having “different takes on the evidence that’s available,” because of a patient’s medical history or because an insurer might have its own preference.

But while “there are certain things that would influence the treatment choice,” those “wouldn’t be things that would explain the correlation with industry funding,” Mitchell said.

The discrepancy in odds between the kidney cancer and leukemia drugs likely is a kind of statistical artifact. The three kidney-cancer drugs the team looked at are almost equally popular treatment choices, while the three for leukemia treatment includes one “almost everyone is using,” Mitchell said. That means there’s not as much variance among leukemia-drug prescriptions for the researchers to explain.

The team also looked at whether drug company research subsidies correlate with prescription choices, but didn’t find a consistent relationship. But they caution that doesn’t rule out the possibility of one.

Ray Gronberg: 919-419-6648, @rcgronberg