Government
Medicare use found to vary significantly by location
■ The preliminary MedPAC report partially confirms Medicare spending analyses by the Dartmouth Atlas Project, but diverges on some points.
By Doug Trapp — Posted Oct. 5, 2009
- WITH THIS STORY:
- » Medicare service use varies less than spending
- » Related content
Medicare beneficiaries around the country receive significantly different amounts of health care services, according to a preliminary analysis by the Medicare Payment Advisory Commission.
MedPAC found no clear nationwide pattern between a region's use of services and its growth in service use, according to a staff report of research based on recent Medicare beneficiary surveys and presented to the board at its Sept. 17 meeting. "This result suggests that to control future spending growth, we need to address both the level of service use and the growth rate," said MedPAC Senior Analyst Daniel Zabinski, PhD.
The MedPAC report on Medicare spending patterns, which is not yet final, comes as Congress continues its massive overhaul effort. Some lawmakers hope to wring savings from the program without harming care for beneficiaries.
MedPAC Executive Director Mark Miller, PhD, said the commission examined use of Medicare services because that has become part of an ongoing conversation between policymakers. "We think that there has been some serious misunderstandings in how to think about this."
The MedPAC findings both confirm and counter previous research on Medicare spending by the Dartmouth Atlas Project, a team of researchers whose work has been cited repeatedly by the Obama administration and many key lawmakers.
MedPAC staff found that the areas with the most intense use of services provided about 30% more care per capita than did the least-intense areas. Dartmouth researchers found a somewhat wider disparity, but they have since scaled back estimates slightly after changing risk-adjustment methods. Both studies used different methods of grouping beneficiaries and adjusting for health status.
Dartmouth Atlas researchers found that certain U.S. regions tend to have higher per-capita Medicare spending than others. For example, Upper Midwest cities, such as Minneapolis, have lower spending than certain coastal cities, such as Los Angeles. But the MedPAC report found that once the Medicare data had been adjusted for special federal hospital payments, among other variables, the coasts did not appear to be the outliers they once were.
Other experts have weighed in on the issue. In June, the Congressional Budget Office said Dartmouth Atlas estimates on the potential Medicare savings from changing practice patterns in high-spending areas are probably overstated.
But debating the exact level of Medicare service variation is not as important as acknowledging that differences exist and should be addressed, said Elliott Fisher, MD, MPH. He is director of population health and policy at the Dartmouth Institute for Health Policy and Clinical Practice and an Atlas co-founder. "We should not be surprised when we use slightly different methods we get slightly different results."
MedPAC, for instance, excluded the areas with the highest 10% and lowest 10% of Medicare service use. Hawaii's low use of hospitals and institutional care, for example, helps keep Medicare spending down. Miami, on the other hand, is infamous for being a Medicare fraud hotbed.
"There may be additional factors beyond practice patterns that drive an area ... out to the extreme, and ... fraud, bluntly, may be part of the story," Miller said.
The AMA welcomed the MedPAC report.
"Further study of the root causes of variations in care will help determine the best course of action to ensure access to a uniform level of high-quality care for all Americans," said AMA President J. James Rohack, MD. "Previous studies have identified multiple factors responsible for utilization patterns, including socioeconomic factors and unmeasured differences in severity of illness. MedPAC's analysis suggests the subject is far too complex for simple solutions like redistributing funds from low-spending to high spending areas, and that the most successful interventions will be locally based."
While some commissioners at the Sept. 17 meeting wanted to draw clear differences between the new data and the Dartmouth work, others focused on the general agreement between the two and the need to focus on why such significant Medicare service variation exists.
"You can make variation say a lot of things, but maybe we do need to be content with 'There's variation,' " said MedPAC Commissioner Karen R. Borman, MD, professor of surgery at the University of Central Florida College of Medicine.
But Bob Berenson, MD, a senior fellow at the Urban Institute, said members of Congress are fighting over the Dartmouth researchers' conclusions about high-spending areas. "To the extent that this analysis would show that some of [the conclusions] are not what one thought, I think it's useful to illustrate that."