Dispute raised over true cause of Medicare spending disparities
■ Study attributes cost variations to patients being sicker in some regions, not to the varying care efficiency of physicians and hospitals.
By Charles Fiegl — Posted June 10, 2013
Washington Medical costs to Medicare vary from region to region because of disease burdens specific to particular areas, challenging previous, long-held research blaming inefficiency and waste for spending differences, a recent study concluded.
Population health likely accounts for 75% to 85% of cost differences across the nation, authors wrote in a study posted online May 28 in Medical Care Research and Review. The authors also cautioned against arguments that large inefficiencies perceived in the U.S. health system can be reduced without harming overall patient care, a conclusion that has been made by researchers behind the Dartmouth Atlas of Healthcare and others.
The latest study found in reviewing data for 1.6 million Medicare beneficiaries that a patient's health status near death explained 84% of health costs incurred during the final year of life, wrote lead author James D. Reschovsky, PhD, a senior fellow at the Center for Studying Health System Change, a Washington think tank. That spending was dependent on the number and type of conditions in the beneficiary population.
However, Medicare still has inefficiencies, such as the structure of the current payment system and the prevalence of fraud, that can be addressed, Reschovsky said in an interview. The Affordable Care Act took aim at these inefficiencies by authorizing the testing of new pay delivery alternatives, such as accountable care organizations, patient-centered medical homes and bundled payments. The models use the current fee-for-service system as a foundation, but over time physicians and hospitals will be encouraged to become more efficient by sharing risk as well as the reward of keeping costs low while providing quality care. Such new innovations would be more effective than adjusting payments based on regional spending disparities, he said.
“In any area you will have some very efficient [physicians] and others who are inefficient,” Reschovsky said. “Altering rates by geographic area is a blunt tool.”
Geographic adjustments also could have an adverse effect on physicians, he added. For instance, penalizing a high-cost area would harm physicians practicing there with great efficiency. Bonuses provided to low-cost areas would reward those offering low-quality and inefficient care.
Dartmouth Atlas responds
Researchers behind the Dartmouth Atlas defended their work. Reschovsky used diagnostic billing codes as “explanatory” factors for spending, wrote Jonathan Skinner, PhD, a Dartmouth Atlas author and professor of economics, in a response to the report. Using these codes would explain more aggressive behavior as poor health status. For instance, an aggressive practice in Texas might refer a patient for treatment of serious heart disease while another in Colorado would send that same patient home. Billing data in this case would show the patient's health status as being better in Colorado than in Texas.
“As we have shown in previous work, this type of bias is very serious,” Skinner stated. “The authors attempt to deal with the problem, for example by consolidating five types of diabetes diagnoses into one, but that does little or nothing to address the fundamental problem that their risk-adjustment approach is tautological — even when they limit their attention to [hierarchical condition categories] in the previous year.”
The Dartmouth Atlas work has led some health policy experts and lawmakers to argue that Medicare spending can be cut by as much as 30% without having an adverse impact on care.
The latest research is more comprehensive and suggests that making such adjustments based on geography is not a good idea, said Stephen Zuckerman, PhD, a senior fellow with the Urban Institute and co-director of its health policy center. Zuckerman had worked with Reschovsky co-author Jack Hadley, PhD, a professor and researcher at George Mason University in Fairfax, Va., on similar research published in 2010. Providing financial incentives in Medicare is important, but they should focus on the health system serving beneficiaries and not on a locality, Zuckerman said.
In April, the American Medical Association commented on a House plan to reform the Medicare payment system and eliminate the sustainable growth rate formula. The House framework mentioned the idea of implementing efficiency measures accounting for geographic spending differences. The AMA said such adjustments would require more work than just calculating differences in hospital and physician costs. Risk adjusters, accounting for other factors such as social structure and risky patient behaviors, would be needed.
Comparing episodes of care rather than per capita costs provides better information on which physicians are most cost effective for a particular condition instead of who manages it the cheapest, the AMA wrote. The Association is in the process of developing tools to measure episode costs.
“In short, much work remains before efficiency measurement can be equitably applied, and we have doubts as to whether it will ever be appropriate for all physicians,” the Association wrote.