Spurred by work at the Dartmouth Institute, there has been intense interest in understanding the sources of geographic variation in Medicare spending. No one seems to dispute that this variation occurs, it is largely a matter of why and whether it is connected to better or worse quality outcomes. A new study, with some authors from Dartmouth, examines whether price differences can account for the variation. (Health Affairs Article) Medicare basically sets payment levels by fiat, although the process has in reality become intensely political. While there are factors that Medicare uses to vary payments by location, primarily for differences in local wage and other costs, those variations are not generally large.
In this study the researchers looked at the 306 Hospital Referral Regions in the United States and adjusted Medicare payment data to pay the same amount for the same service, regardless of location. The authors examined both hospital and physician spending. On a per capita spending basis, adjusting to a common price did not erase the substantial geographic variation. In fact there was a very high correlation between the high spending regions on an adjusted and unadjusted basis. The prices Medicare paid physicians and hospitals in different geographic areas accounts for only a small amount of the variation.
The follow-on to this line of research should include more complete analysis of why utilization varies so much–is it at all justified by patient need or is it due to other factors such as provider attempts to maximize income or inappropriate failure to follow best-practice guidelines. As we noted earlier this week, recent studies in the commercial marketplace show that unit price is the driving force in payment variations and health insurance cost increases. Medicare does not present the same opportunity for providers to increase prices, so perhaps some of them compensate by utilization increases.