A study found that drug company payments to physicians led to a nearly 4% uptick in prescriptions, a smaller impact compared to other research.
The study, released Thursday from the National Bureau of Economic Research (NBER), tackles the controversial practice of pharmaceutical manufacturers paying doctors to speak at conferences, get meals or cover educational expenses. The study also examined the role payments played in the transition of patients from a brand-name drug onto a generic.
Researchers looked at the prescribing behavior of physicians who did and did not get payments from pharma companies. The study examined Medicare Part D claims from 2013 to 2015 and used data from the Open Payments database, which forces doctors and manufacturers to divulge financial relationships.
It is common for physicians who heavily prescribe Part D drugs to get payments related to the drugs they dole out, the study said.
Researchers found that shortly after doctors received a payment, drug spending began to increase.
“On average, expenditures are approximately $9 greater per month in the year following the payment,” the study said. “Relative to average monthly expenditures, $238, this is slightly less than a 4% increase.”
The study noted that the impact on spending is much smaller than other research has found. A similar study, for example, found payments to cardiologists increased prescriptions of statins by 73%.
NBER’s study also examined whether the estimated slight increase in expenditures is due to more patients being put on a drug.
“The number of patients does increase shortly after the payment, although the magnitude is relatively small,” the study said. “We note that an increase in the number of patients taking the drug could arise both from physicians deciding to start a new patient on the drug after an encounter with the drug firm or physicians differentially continuing current patients on the drug (instead of switching to another drug).”
However, the spending was higher overall for drugs in classes that had five or fewer promoted drugs.
“In classes with very few branded drugs, a payment might be more effective than in a class where there are many different drugs being promoted,” the study said.
In the first year after payment was received, spending on drugs in such low-competition classes increased by $25 per month.
“Given the $206 average monthly expenditure in these classes, this translates to a 12% increase,” the study added.
The difference could explain why other research has shown prescribing increased for drugs in certain classes, such as the statin study.
NBER also looked at how pharma payments impacted a transition to generics. Researchers looked at prescription patterns for five major drugs that lost patent protection and faced generic competition: Abilify, Namenda, Celebrex, Evista and Zyvox.
The researchers examined the generic efficiency rate, which measures generic penetration in a market, over the first six months of competition for the five products. NBER found no evidence that paid physicians transitioned their patients to generics more slowly than physicians who don’t get payments. However, paid physicians are more likely to put patients on a new formulation of a drug.
Sometimes, a drug company will introduce a new drug formulation just prior to generic competition of the original drug.
“Individuals who are taking the new formulation are not subject to automatic generic substitution at the pharmacy,” the study said.
Drugmaker Actavis employed this strategy with Alzheimer's disease drug Namenda, creating an extended-release version just before the patent on the original expired. By the end of 2015, 37% of physicians who didn’t get payments from Actavis prescribed patients on the extended-release version of Namenda. However, 64% of physicians getting payments put their patients on the extended-release Namenda.
“Overall, our results suggest that costs do increase as a result of marketing encounters between drug firms and physicians,” NBER researchers said. “At the same time, our results on drug quality are mixed; we do not find clear evidence that such payments are harmful to patients, only that they do not seem to be obviously helpful.”