Health Policy$ense

Why Your Doctor Shouldn’t Accept That Sandwich

Study finds free meals from pharma impacted statin prescribing

Pharmaceutical company payments can significantly influence physician prescribing behavior, according to a recent National Bureau of Economics Research (NBER) Working Paper by LDI Senior Fellows Matthew Grennan, Ashley Swanson, and colleagues. The authors, who looked at the Medicare Part D claims of 15,000 cardiologists in the market for statins from 2011-2012, found that cardiologists receiving meals from a given firm increased prescribing of that firm’s branded statin by 73 percent. Because these payments steer patients toward much more expensive drugs, they increased spending on statins by $67M (26%) in 2011 and $65M (33%) in 2012. 

And the net effect on consumers? Unless the meal-related interactions sufficiently conveyed information to cardiologists who were incorrectly under-prescribing statins, the authors estimate that they reduced consumer welfare by $190M.

The authors focused on cardiologists who received meals from AstraZeneca and Pfizer, the manufacturers of the two major on-brand statins at that time (Crestor and Lipitor). Statins, which lower “bad” cholesterol in the blood and reduce the risk of heart disease, heart attack, and stroke, are among the most prescribed drugs in the industrialized world.

Although pharmaceutical companies’ payments to physicians can take many forms, such as compensation for consulting or education, meals are the single most popular in-kind payment. Approximately 90 percent of physician-firm interactions from Pfizer and AstraZeneca consisted of meal-related payments during the study, and an estimated 59 percent of all cardiologists received a meal from one or both companies in 2011.

Grennan and colleagues found that the 73 percent increase in statin prescribing was primarily driven by cardiologists who prescribed the statin at low or moderate rates, suggesting that the companies targeted their meals to “low prescribers.” The authors controlled for a number of factors that could influence prescribing practices, including patient, physician, hospital, and regional variables such as Academic Medical Centers’ Conflict-of-Interest policies. The introduction of several generic statins during the study and the expiration of Lipitor’s patent in 2011 allowed them to differentiate the effects of meal payments from other market power effects. They also examined payments prior to the federal Physician Payments Sunshine Act of 2010 taking effect, which requires broad disclosure of drug and device manufacturers’ payments to teaching hospitals and physicians.

Although AstraZeneca spent nearly twice as much per meal as Pfizer in 2011, the authors found that receiving a meal of any value drove prescribing, and most of the meals studied were below $100. These findings suggest that policies enacting monetary meal limits may have limited efficacy. Most recently, New Jersey announced that they will be revising their $10,000 annual cap on physician payment and increasing their $15.00 dinner limit. In 2017, Maine banned all noncash gifts except those of “minimal value” and “modest meals” associated with educational events, while California limited meal payments to $250.00 per physician annually. Ultimately, as the authors note, meals function as a pathway for pharmaceutical companies to access physicians, regardless of value.

Few studies have empirically examined how interactions between firms and experts affect consumers. Grennan and colleagues addressed this gap by estimating the effect of pharmaceutical companies’ meal-related payments on physician prescribing and consumer welfare. Their results have important policy implications for states looking to encourage efficient drug markets by creating guardrails around industry relationships with physicians, particularly given recent debates around high drug prices.

Currently, eight states and the District of Columbia have laws beyond the Sunshine Act that regulate drug and device manufacturers’ marketing efforts or payment disclosures. While this study has important limitations, including its focus on a particular market during a defined time period, it suggests that states should examine their policies around pharmaceutical companies and physician interactions to minimize conflicts of interest and maximize consumer welfare.

The NBER paper, Physician-Industry Interactions: Persuasion and Welfare, was written by:
Matthew Grennan, PhD, Assistant Professor of Health Care Management, The Wharton School, University of Pennsylvania;
Kyle Myers, PhD, Postdoctoral Fellow, NBER;
Ashley Swanson, PhD, Assistant Professor of Health Care Management, The Wharton School, University of Pennsylvania;
Aaron Chatterji, PhD, Professor of Strategy, Fuqua School of Business, Duke University