Why the Oregon CCO Experiment Could Founder

Abstract

The most recent Oregon Medicaid experiment is the boldest attempt yet to limit health care spending. Oregon’s approach using a Medicaid waiver from the Centers for Medicare and Medicaid Services utilizes global payments with two-sided risk at two levels—coordinated care organizations (CCOs) and the state. Equally important, the Oregon experiment mandates coverage of medical, behavioral, and dental health care using flexible coverage, with the locus of delivery innovation focused at the individual CCO level and with financial consequences for quality-of-care metrics. But insightful design alone is insufficient to overcome the vexing challenge of cost containment on a two- to five-year time horizon; well-tuned execution is also necessary. There are a number of reasons that the Oregon CCO model faces an uphill struggle in implementing the envisioned design.

In the Point essay, Howard et al. provide important perspective regarding reasons for optimism for the Oregon coordinated care organization (CCO) experiment; nonetheless, ample room for concern remains. For Oregon and other states, detailed consideration of the challenges facing the CCO experiment could be helpful to design programs and operational plans that maximize the odds for success.

1. The timeline to demonstrate cost savings is very ambitious

Based on agreements between the state and the Centers for Medicare and Medicaid Services (CMS), cost savings must be achieved by the end of year 1, and a full 2 percent reduction in health care inflation must be achieved by the end of year 2. Yet the state legislation authorizing CCOs was signed into law only one year prior to the beginning of the timeline, and individual CCOs were not certified by the state until six weeks before the timeline began (Oregonian, n.d.; Oregon Health Authority 2012). Not surprisingly, clinical delivery systems appear to be far from mature in many CCOs. While it is possible that participating organizations can rapidly transform systems of care or that the dramatic nationwide slowdowns in health care inflation preceding the Oregon experiment will persist (Cutler and Sahni 2013), a robust operational plan should be positioned to succeed in both opportune and adverse environments.

2. The tenets on which CCO reforms are based have not been adequately proved for statewide implementation

The tools fundamental to the program’s success (including patient-centered medical homes; physical–behavioral–dental health integration; disease management programs; and care coordination) have limited evidence of effectiveness regarding cutting costs and/or improving quality. In addition, the ability to extrapolate local successes to a broad-scale reform program in a state with mixed urban and rural populations and many different health care delivery organizations is uncertain. For example, in fifteen different national demonstration programs of care coordination in fee-for-service Medicare, none generated cost savings (Peikes et al. 2009). While focused disease management programs have achieved somewhat better results, they are not unequivocal and are unproven when implemented broadly. Great hope has been placed on the idea that patient-centered medical homes will improve care, reduce costs, and reinvigorate the field of primary care, but the evidence thus far does not negate warnings against premature dissemination of patient-centered medical homes (Berenson, Devers, and Burton 2011; Hoff 2010a, 2010b).

3. Competition between health systems could prevent the clinical integration and innovation required for success

The levels of trust and integration between hospital systems or between outpatient clinicians and local hospitals may be insufficient to achieve the CCO model’s goals. The care coordination and other delivery changes envisioned in the reform program require integration of care and avoidance of unnecessary emergency room visits and hospital admissions. But competition between health systems within some Oregon CCOs is vigorous (Coughlin and Corlette 2012; Stecker 2013), as illustrated by a firsthand anecdote. During review of a disease management program’s rollout, the oversight committee members were pleased to learn of the program’s promising start. But the nurse leading the program also related a concerning story about her experience in contacting the primary care doctor of a recently discharged patient. The physician, employed by a competing hospital system, did not endorse the patient’s participation in the program (though the two health care systems are in the same CCO and the patient seeks care in both systems). This experience raises concerns about the ability of systems of care to effectively operate between competing organizations and highlights the challenges in attributing costs and benefits of system change.

While some aspects may be unique to health care organizations with extensive geographic overlap (which are responsible for approximately 40 percent of Oregon’s Medicaid population), there may also be competing interests among different organizations within well-defined geographic regions. For instance, there could be important conflicts regarding the allocation of costs, work efforts, and incentives between hospitals and autonomous clinics for the management of overlap disorders such as heart failure, ischemic heart disease, and chronic obstructive pulmonary disease.

Not only could operational aspects suffer from excessive competition, but so too could the innovation required to improve delivery models. On a longer timeline, it could be advantageous for many different CCOs to independently develop and tailor important innovations such as information technology for identifying patients at risk for inappropriate or unnecessary admission. But on a tight timeline, pooling efforts and experience will be critical to efficiently bringing forth improvements to care. The state recognizes this and has put in place a sophisticated plan centered on its Transformation Center. But it remains to be seen whether this will be sufficient to produce quickly actionable innovations in a fragmented and competitive market in which major stakeholders are motivated to develop proprietary tools for competitive advantage.

4. Reform tools may not be implemented consistently across the state in a fashion that will achieve the cost savings and quality goals

Local successes would provide important information that could propel health reform in Oregon and elsewhere. But to meet the cost and quality targets on the required timeline, some level of consistency across the state must be demonstrated. This will not be easy; for example, in the most recent Oregon Health System Transformation report (Oregon Health Authority 2013), rates of patient visits to emergency departments varied more than twofold between top- and bottom-performing CCOs, despite consistent emphasis on the importance of reducing inappropriate emergency department visits as a critical tool for cost reduction. Most measures of quality also varied considerably (a representative example, colorectal screening, varied threefold).

5. Health systems that produce a majority of their revenue from fee for service will have limited focus on redesigning systems of care to maximize savings under a global budget arrangement

Oregon is unique in its long experience with managed care. For this reason, as well as sophisticated leadership within the state government, multiple stakeholders are deeply committed to reform. Nonetheless, intellectual disposition does not supersede fiduciary duty when individuals’ or organizations’ financial health is at stake. Even in Oregon, fee-for-service payment systems still dominate the heath care marketplace as a whole. Unless there is rapid adoption of global budgets by payers across Oregon, health care organizations will be required to operate in a dual environment (Stecker 2013). If broad changes are made to reduce inappropriate utilization of resources, any savings generated among the approximately 20 percent of patients with Medicaid global payment programs could be more than offset by lost revenue among privately insured or Medicare-insured patients with fee-for-service plans. Managed care and fee-for-service systems may coexist within the same organization, but it can be uncomfortable for physicians such as specialists who may be engaged in both simultaneously and may be asked to take different approaches to medical care based on different payment models.

The Massachusetts Alternative Quality Contract (AQC) program is cited as an example of successful transformation to global payments on a tight timeline. But there are some important differences. In Massachusetts the AQC program was implemented by a private insurance company (Song and Landon 2012); thus a substantial proportion of the more profitable private fee-for-service reimbursement was shifted to global payments. In Oregon, health care under the private fee-for-service plans remains largely unaligned with the global payment systems promoted by the CCO experiment. Another important difference is that as part of a simultaneous transformation of Massachusetts Medicaid, the state has taken an aggressive approach to the supervision of health system costs and used an explicit threat of regulatory leverage if those costs are not controlled (Solman 2013), while Oregon expressly has not (Gray 2013). In theory, Oregon’s program is superior; any spending in excess of immutable benchmarks will result in automatic financial penalties for CCOs and the state. However, until Oregon’s global payment system extends to the much larger private insurance marketplace, the absence of incentives to contain costs outside the Medicaid system could undermine the impact of global payments within Medicaid. For example, individual health care organizations could calculate that the loss of fee-for-service revenues resulting from delivery transformations would exceed Medicaid capitation losses by a large enough margin to warrant limiting their engagement in reform.

Conclusion

As with any large-scale public policy shift, the Oregon CCO experiment faces challenges that are likely the result of political and fiscal necessities rather than lack of foresight. With ongoing attention by state policy makers to preempting operational challenges, hard work by all stakeholders, and perhaps a small dose of good luck, the Oregon plan can succeed. Whatever the outcome, the experience will hold important lessons regarding government-led health care transformation that can be applied in political contexts across the nation.

Biography

Eric C. Stecker is assistant professor of medicine in cardiology and electrophysiology at Oregon Health and Science University. As director of inpatient cardiology, he delivers specialty health care and designs systems of care for patients with acute cardiac illnesses from across Oregon. He received an MPH from the Department of Health Management and Policy at the University of Michigan’s School of Public Health, and his administrative time is focused on pragmatic health care delivery improvements and health policy reforms that generate the maximum value from high-technology specialty medicine. Stecker also researches cardiac arrest prevention and treatment, with publications in the Journal of the American College of Cardiology and other cardiology journals. He has had two recent health policy essays published in the New England Journal of Medicine.