Summary
Underlying socioeconomic and demographic characteristics have important implications for the geographic patterns in Medicare spending. The findings contradict prior research claiming that most of the variation in Medicare spending can be attributed to geographic differences in medical ‘practice styles’ (the extent and intensity of medical interventions in a given area), thus calling into question the claim that the U.S. could save up to $700 billion in health care waste and inefficiency if all providers were to emulate the practices of low-costs states.
In “Why Geographic Variation in Health Care Spending Can’t Tell Us Much about the Efficiency or Quality of our Health Care System,” Brookings Senior Fellow and Hutchins Center on Fiscal and Monetary Policy Director Louise Sheiner contends that the geographic variation in health care spending does not provide a useful measure of inefficiency and waste in the system, in contrast to the widely accepted views from the Dartmouth Atlas of Health Care. Her paper finds that the ‘practice style’ that Dartmouth researchers point to is in fact reflective of underlying systemic differences across states. By employing a state-level approach, as opposed to Dartmouth’s individual-level approach, Sheiner shows that “states with similar demographic characteristics have similar levels of real beneficiary Medicare spending. Thus, what the Dartmouth researchers have deemed as differences in ‘practice styles’ are not randomly distributed, but are instead closely linked to population characteristics.”
In fact, states that appear to be high-cost, like Florida and Connecticut, are no longer big health care spenders once demographic and health variables are controlled for, just as states that appear to be low health care spenders like Utah and Vermont are actually relatively high spenders once those adjustments are made. These results suggest that the cross-state variation in Medicare spending is tightly associated with the characteristics of state populations, and that once these characteristics are controlled for, the variation in spending is fairly small.
For example, it is no surprise that states like Mississippi and Louisiana spend far more on Medicare given their populations’ characteristics. These states perform poorly on a wide range of indicators, including diabetes rates, obesity rates and lack of physical activity, suggesting that the higher Medicare expenditures may simply reflect the fact that the population is unhealthy. Furthermore, physicians practicing in states with sicker populations may practice a more intensive form of medicine for all their patients, making comparisons across states based on individual patients potentially misleading. Finally, states with higher Medicare expenditures tend to have more uninsured and lower reimbursements from private insurance and Medicaid. Because all of these factors are so hard to disentangle, Sheiner concludes that looking at variation in Medicare spending across geographic areas is not particularly informative. “A comparison of health spending in Mississippi with health spending in Minnesota is not likely to provide a useful metric of the ‘inefficiencies’ of the health system, nor is it likely to provide a useful guide to improve the quality of care in places where it is lacking,” she writes.
The author also finds a negative correlation between each state’s level and growth of Medicare spending – low-spending states tend to have higher growth rates than high-spending states. Thus, examining the practice patterns in low-cost states isn’t likely to provide the key to “bending the cost curve” in health spending, she concludes.