Medicare spending variation across geographic areas driven by differences in use of post-acute care

A "geographic value index" that would tie Medicare payment rates to the health benefits and costs of health services in particular regions of the country should not be adopted by Congress, says a new congressionally mandated report from the Institute of Medicine. The committee that wrote the report concluded that decisions about health care generally are made at the level of the physician or organization, such as a hospital, not at the regional level. Because individual physician performance varies, sometimes even within a single practice group, an index based on regions is unlikely to encourage more efficient behavior among individual providers, and therefore, is unlikely to improve the overall value of health care. The new report reiterates the findings of the committee's interim report released in March 2013.

Variation in Medicare spending across geographic areas is driven largely by differences in the use of post-acute care, which includes home health services, skilled nursing facilities, rehabilitation facilities, long-term care hospitals, and hospices, the committee said. If regional variation in post-acute care spending did not exist, Medicare spending variation would fall by 73 percent, and it would fall by 89 percent if there was no variation in both acute and post-acute care. However, an overall explanation for geographic variation in spending remains elusive. The statistical analyses that the committee examined accounted for factors such as beneficiary health status and demographics, insurance plans, and factors related to health care markets, but much of the variation could not be explained by such factors.

The committee also examined the differences in spending at a variety of levels progressively smaller than geographic regions, such as hospital referral regions, hospital service areas, hospitals, and individual providers. Spending varies greatly across all these levels, and providers even at a small level do not practice the same way or perform similarly. Consequently, a geographic value index would reward low-value providers for practicing in areas that are on average high-value and punish high-value providers in low-value regions. As a result, area-level performance calculations would likely mischaracterize the actual value of services delivered by many providers and hospitals, resulting in unfair payments and inappropriate incentives.

The committee found that in contrast to Medicare, variations in spending in the commercial insurance market are due mainly to differences in price markups by providers rather than differences in the use of health care services. Medicare spending is weakly correlated with commercial insurance across regions, and total spending by all payers in a region is not strongly correlated with either Medicare spending or spending by commercial insurers.

To improve care, payment reforms need to create incentives for behavioral change by decision makers, whether they are at the level of individual providers, hospitals, health care systems, or stakeholder collaboratives. The committee recommended that the Centers for Medicare & Medicaid Services (CMS) continue to test Medicare payment reforms that incentivize the clinical and financial integration of health care delivery systems to encourage coordination of care among individual providers, real-time sharing of data and tracking of service use and health outcomes, receipt and distribution of provider payments, and assumption of risk managing their populations' care continuum. CMS should also evaluate the effects of test payment reforms on health care quality by measuring Medicare spending and beneficiaries' clinical health outcomes and use the results to improve the payment models. If these evaluations demonstrate increased quality, Congress should give CMS the flexibility to accelerate the adoption of the new Medicare payment models.

Source: National Academy of Sciences