Secretary of Health and Human Services (HHS) Alex Azar recently expressed his interest in continuing to use HHS authority granted by the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA) and the Affordable Care Act to "drive the value-based transformation of our entire health system." He also signaled the need to take a fresh look at possible approaches to health provider payment. Two recently published papers supported by the Commonwealth Fund (in JAMA Internal Medicine and Health Affairs) analyzed the effects of one payment model that has not yet received much attention at the federal level: the global budget. This post highlights key findings of those papers and considers global budgets in the context of other payment reform initiatives.

In 2014, under a Centers for Medicare and Medicaid Services waiver, Maryland initiated an all-payer global budget program for all acute-care hospitals in the state. Thirty-six hospitals joined 10 rural hospitals that had participated in an earlier pilot program. The program limits growth in per-person total spending on hospital care, across all payers in Maryland, to a predetermined percent each year. By setting limits on hospital spending, the program was designed to contain overall spending growth in Maryland and encourage hospitals to identify lower cost ways of delivering hospital-level care and dedicate the unspent resources to increasing the use of primary care.

Early results from an independent evaluation funded by the Centers for Medicare and Medicaid Innovation (CMMI) suggested that Maryland’s global budget program reduced total expenditures and total hospital expenditures "without shifting costs to other parts of the health care system outside of global budgets." Analyzed against a matched control group of hospitals outside of Maryland, there appeared to be a decline in hospital admissions—a plausible contributor to savings. Price changes may also have contributed (absent changes in utilization, hospitals were required to adjust prices to stay on budget).

In the recently published papers, researchers from Harvard Medical School and the University of Pittsburgh studied whether the program altered the use of health care services in areas of Maryland where hospitals received global budgets, also focusing on a matched control group of geographic areas not subject to the global budgets. The first study examined the impact of the global budget program’s statewide expansion on use of acute hospital care, emergency departments, outpatient care at hospitals, and primary care physicians. The second study looked at changes in acute-care hospital utilization and spending in rural hospitals that were not included in the earlier pilot of global budgets.

These latter studies call into question whether spending reductions observed under Maryland’s global budget program can be attributed to changes in hospital and primary care utilization. There were reductions over time in hospital use and changes in primary care use among areas served by the global-budget hospitals. However, the researchers found similar trends in health care use and spending in areas served by hospitals that were not subject to global budgets. In addition, the authors found that changes in hospital stays and primary care visits were already under way prior to the global budget program. Differences between these results and the findings of the federal evaluation may arise because of different assumptions about utilization trends prior to the program, the years studied, and the protocols used to identify comparison groups.

Across evaluations, the observed changes in patterns of care were subtle and varied depending on the type of service and payer examined. Should we conclude that global budget models are not a viable path for future reforms? We believe such a conclusion would be premature for several reasons:

  1. Maryland’s global budget model was limited to hospital services rather than total cost of care: Payments to physicians or other providers were not included in hospitals’ budgets. Physicians, having little financial risk or stake in the success of the program, may have done little to change established patterns of care delivery. Maryland has signaled that it will move toward a broader definition of ‘global’ by including physicians and physician spending in future refinements of its payment model.
  2. The downward pressure of the global budget was not very strong: The underlying payment approach of the program was still fee-for-service: hospitals continued to bill on a per-admission or per-service basis, raising and lower their prices to meet their budgets. The continued reliance on a fee-for-service architecture—and implicit dependence on patient volume for revenue—may have dampened the incentives.
  3. It may be too early to discern effects: Payment reforms that use financial incentives to change hospital management, operations, and clinical care can take time to fully implement. Even the three-year period evaluated by the researchers may be too short a time to see all of the changes in care delivery that hospitals may be required to make.
  4. Evaluation of local programs in the midst of other large-scale reforms is challenging: When major policy reforms are occurring nationally, local program evaluations may fail to tease out important effects. For example, other payment models mentioned below may have been driving changes in Maryland and other parts of the country, muting changes related to a single policy.

Many alternative payment models are getting under way. How do global budgets compare? Overall, programs such as physician-led accountable care organizations (ACOs), the oncology care model, and the Comprehensive Primary Care Plus (CPC+) model have produced modest savings without lowering quality. And early results from bundled payment for hip and knee replacement surgery led to larger reductions in Medicare spending per case, suggesting that more judicious use of health care services is achievable. Since no single model is likely to suffice on its own to drive value throughout the health system, comparing the global budget approach to other payment reform models may be less useful than refining each model and thinking about how the value-based payment reform models can be deployed in combination. By evaluating these models carefully and setting realistic expectations, the CMMI can play a central role in generating both the innovative programs and the evidence necessary to “drive real change in our system.”