Variation in Healthcare Spending: Regional and Local

It is becoming increasingly clear that there is wide geographic variation in healthcare spending; the variation is not driven by characteristics of the patients and is not associated with either the quality of care or patient outcomes. This emerging evidence has generated calls for such coverage constraints as targeting high-spending areas for lower Medicare payments, with a focus on the hospital referral regions (HRRs) identified in the Dartmouth Atlas of Health Care.

According to researchers, such policies are based on the concept that there is a system-level component driving some areas to have high levels of healthcare utilization and others low. The effectiveness of policies in reducing overutilization without lowering access to high-quality care is dependent on the effectiveness of the targeting. If substantial variation across local areas within HRRs is found, then focusing on high-cost HRRs may leave many high-spending locales untouched while “inadvertently penalizing some low-spending locales,” they noted.

There are 306 HRRs in the United States. To assess the local-market heterogeneity that larger units may mask, the researchers compared variation in medical spending and prescribing patterns in the HRRs with variation at the local level (the 3436 hospital service areas [HSAs] that compose the HRRs). They compared HRRs with HSAs in terms of variation in spending and use of prescription drugs and medical care. They also examined the degree to which high-spending HSAs are clustered together within HRRs. They reported results of the comparison in the New England Journal of Medicine [2012;367(18):1724-1731].

The researchers utilized data on enrollment, and pharmacy and medical claims from 2006 through 2009. The data was a 5% random sample of Medicare beneficiaries obtained from the Centers for Medicare & Medicaid Services (CMS). Because the CMS has both medical and pharmacy data only for beneficiaries enrolled in prescription drug plans (PDPs), the researchers identified all beneficiaries who had been enrolled for at least 1 month in Part A, Part B, and a stand-alone Part D PDP.

The sample consisted of 1,013,477 beneficiaries in 2007, 1,024,183 in 2008, and 1,022,662 in 2009, for a total of 3,060,322 beneficiary-year observations. Using residence zip codes, each beneficiary was assigned to 1 of the 3436 HSAs and thus to 1 of the 306 HRRs.

Following adjustment for beneficiary-level demographic characteristics, insurance status, and clinical characteristics, the researchers identified substantial local variation in healthcare (drug and nondrug) utilization and spending. In addition, many of the low-spending HSAs were located in high-spending HRRs; conversely, many of the high-spending HSAs were in low-spending HRRs.

There was substantial variation among HSAs within HRRs. One comparison cited by the researchers was Manhattan, New York, which was one of the HRRs with the highest adjusted drug spending, and Albuquerque, New Mexico, which was one of the lowest-spending HRRs; there was substantial dispersion in spending across the HSAs within those HRRs: the lowest-spending HSA in Manhattan had lower spending than approximately 25% of the HSAs within Albuquerque.

With respect to drug spending, only 50.7% of the HSAs located within the borders of the highest-spending quintile of HRRs were in the highest-spending quintile of HSAs. Only 51.5% of the highest-spending HSAs were located within the borders of the highest-spending HRRs. Patterns for nondrug-related spending were similar.

In conclusion, the researchers commented, “The effectiveness of payment reforms in reducing overutilization while maintaining access to high-quality care depends on the effectiveness of targeting, Our analysis suggests that HRR-based policies may be too crudely targeted to promote the best use of healthcare resources.”