Study finds no clear winner among Medicare payment models

As Congress considers how to make Medicare spending more efficient and cost-effective, a new study has found it won’t be as easy as picking a single payment model.

The Medicare Payment Advisory Commission, which advises Congress on Medicare, has filed its annual report. Following up on similar research started last year, the group looked at the performance of the three main payment models: fee-for-service (the standard model tied to the provision of tests and procedures), Medicare Advantage (capitated payments per member), and accountable care organizations (groups of physicians and other providers that share in savings and risk).

The commission said it found that no one model performed the best across all 78 markets it studied, with fee-for-service being cheapest in 28 markets, ACOs cheapest in 31 markets, and Medicare Advantage cheapest in 19 markets. Researchers noted there are still few studies comparing the quality and patient satisfaction generated by the three models.

The report said that Medicare may need to change its payment rules to encourage recipients to choose the payment model that is cheapest for their market. For instance, Medicare could set a national base premium for Part-B services that would pay for either fee-for-service/ACOs or Medicare Advantage, whichever was cheaper in a particular market. Or it could calculate premiums at the market level, with higher spending markets having higher premiums, which would again pay for whichever model was cheapest. Note that the researchers folded ACOs in with fee-for-service for this exercise.

The commission also continued to express concerns about how Medicare measures care quality, saying it relied too much on clinical processes that didn’t necessarily lead to better health outcomes and sometimes placed a heavy reporting burden on physicians and other providers. Among the commission's alternatives is a new measure called “healthy days at home,” which measures the number of days during a specific period where a recipient was alive and didn’t interact with the health care system, other than for preventive or maintenance care.

Initial analysis indicates this measure could highlight meaningful differences in health outcomes across populations, especially when applied to patients with one or more chronic conditions. But the researchers also said the measure needed more study to determine if differences in post-acute-care were skewing the results and reflecting differences in practice patterns across geographic areas rather than differences in health status.