Medicare's blame game: Stanford scholar finds what conventional wisdom gets wrong
A SIEPR fellow gains first-ever access to data showing the inner workings of an influential committee advising Medicare and finds that bias among its members has different effects from what critics claim.
A Stanford scholar’s new research finds that even though members of an advisory committee for Medicare are biased toward physician specialties, the partiality often bridges across specialty lines and may improve the quality of its price-setting recommendations.
For the first time, David Chan, a faculty fellow at the Stanford Institute for Economic Policy Research, and his colleague Michael Dickstein from New York University, gained access to more than 4,000 fee proposals that were reviewed over a 21-year span by the committee, which is part of the American Medical Association (AMA). Their independent analysis is in a working paper, released Feb. 26 by the National Bureau of Economic Research.
The finding is a surprising insight. Until now, behind-closed-doors deliberations meant nobody has known for sure how the physician-based committee reaches its recommendations for health care service prices, which Medicare typically adopts. And longstanding criticisms of conflicts of interest have been largely based on anecdotal evidence and the assumption that tasking doctors with setting their own prices must be the equivalent of the fox guarding the henhouse.
But according to the empirical research, even if committee members were entirely neutral, only 1.9 percent of the $70 billion Medicare spends annually on physician care would be redistributed across all services.
“Though the analysis is not a complete vindication of the AMA committee, we find that committee bias has subtle implications for different medical fields and for Medicare,” said Chan, an assistant professor of medicine at Stanford.
“Primary care doctors once thought to be disadvantaged by the presence of specialty physicians on the committee actually benefit from shared interests with other types of physicians,” he says. “And overall, Medicare gets higher-quality information when the committee has connections with specialties.”
Benefits of bias
In their research, Chan and Dickstein, an assistant professor of economics at NYU, set out to uncover whether committee members exhibit bias in their recommendations and, if they do, how much it affects overall prices.
Since 1992, Medicare has tasked the AMA committee, formally known as the Relative Value Scale Update Committee (RUC), with calculating the time and effort component which, together with service costs, accounts for 96 percent of the Medicare reimbursement rate. Most private insurers also establish their payment rates based on Medicare pricing.
The lopsided composition of the committee – specialists significantly outnumber primary care physicians – has also fueled suspicions that prices for complex procedures are rising quickly because doctors on the committee are inclined to increase the cost of the procedures that either fall under or are closely related to their practice areas.
After reviewing internal deliberations on 4,423 fee proposals from 1992 to 2013, the researchers found an increased likelihood that committee members will recommend higher prices for specialties they are connected with. For example, a spinal surgeon on the committee is likely to agree with a price increase for a hand surgery procedure because both share revenue from orthopedic procedures.
The researchers then measured how closely connected a proposed price change was to the specialties represented on the committee and the effect that affiliation had on the recommended reimbursement. They found that the more connected the overall committee was to specialties representing a procedure, the more likely it was to go along with a suggested rate increase.
So why would Medicare rely on a biased industry group to determine its prices? The evidence, Chan said, suggests an explanation: The lack of impartiality on the committee is offset by the finding that the information members contribute to the price-setting process is of higher quality than input from neutral advisers.
“There is this trade-off between bias and the quality of information,” Chan explained. “An unbiased but very imprecise price may be worse than a biased price that is closer to the truth.”
Positive impact on primary care
Contrary to common perception, the researchers also suggest that primary care doctors are not always harmed by these biases. They found that services performed by primary care doctors and specialists often overlap, which means that Medicare pricing policies affect them in similar ways more often than people think. For example, primary care physicians who are internists and family medicine doctors perform some procedures that cardiologists and radiologists do. So, if the price of an electrocardiogram goes up, primary care doctors stand to gain financially from the procedure as much as cardiologists and cardiothoracic surgeons.
And because primary care specialties already benefit from affiliations with other specialties, doubling the number of internists on the committee and quadrupling the number of family medicine practitioners would increase their specialty revenues by less than 1 percent.
Further, the analysis showed that such shared interests – and the closer connection between committee members and the specialties communicating the costs of a procedure – helped boost the overall quality of information behind committee decisions.
“There are very likely several features in Medicare’s pricing structure that disadvantage primary care,” Chan said. “But our research suggests that the arrangement of the RUC is not one of them.”