Accountable Care Organizations: Title and subTitle BreakAccountable for What, to Whom, and How

Interest in accountable care organizations (ACOs) has increased dramatically with the passage of the Affordable Care Act, which establishes ACOs as a new payment model under Medicare and fosters pilot programs to extend the model to private payers and Medicaid. Proponents hope that ACOs will allow physicians, hospitals, and other clinicians and health care organizations to work more effectively together to both improve quality and slow spending growth.¹ Skeptics are concerned that ACOs will focus narrowly on their bottom line and either stint on needed care or use the leverage they achieve through local integration to demand unreasonable prices from payers.

Whether ACOs achieve their ambitious promise remains far from certain.² It is likely that the success of ACOs (and the many other payment-reform initiatives included in the Affordable Care Act) will depend in large part on whether the Centers for Medicare & Medicaid Services, private payers, physicians, and health system leaders can work together to establish a tightly linked performance measurement and evaluation framework that not only ensures accountability to patients and payers, but also supports rapid learning, timely correction of policy and organizational missteps, and broad dissemination of successful organizational and practice innovations.³ Because ACOs are likely to be one of the first major payment reform initiatives to be put in place, the measurement framework established for ACOs could also provide a foundation for evaluating other reforms.

The limitations of current approaches to performance measurement are well recognized.^{4 - 5} Measures too often assess individual clinicians and silos of care, focus largely on processes of questionable importance, are imposed as an add-on to current work, and require burdensome chart reviews and auditing or reliance on out-of-date administrative claims data. Poor performance is seen as a consequence of individual failure, rather than flawed systems. The result is a performance measurement system that often provides little useful information to patients or clinicians, reinforces the fragmentation that pervades the US health care system, and reinforces physicians' perception that measurement is a threat.

Advances in the science of improvement⁴ and progress in health information technology have led to an emerging national consensus that new, more promising approaches to performance measurement are within reach.^{5 - 6} Measurement of key processes and outcomes must be an integral part of the care process to support both improvement and accountability, reduce the burden of measurement, and improve accuracy and reliability. Explicit aims for improvement are essential; major stakeholders have agreed on the following: improving population health, engaging patients in making decisions and managing their care, improving safety and care coordination, guaranteeing compassionate and appropriate end-of-life care, and eliminating waste.⁷ Moreover, experts agree that achieving these aims will not be possible without longitudinal approaches to measurement that capture patient-reported health outcomes, the degree to which care was aligned with patients' well-informed preferences, and total costs of care.⁶

This is an ambitious vision of comprehensive, meaningful measures embedded within interconnected electronic health records that support clinicians' efforts to improve across the full continuum of care while ensuring accountability to payers, patients, and policy makers.

The challenge facing policy makers is therefore how to simultaneously ensure that the implementation of ACOs and other payment reforms provides an adequate level of accountability for participating clinicians and health systems constrained by current measurement approaches, while helping all to advance rapidly toward the more advanced measures and measurement systems envisioned.

One approach would be to build on the notion of different levels of ACOs based on different payment models (shared savings with no risk, symmetrical shared savings with some risk for excess costs, and partial capitation), which would require differing levels of organizational structure and accreditation.^{1 ,8} The same principle of differing levels of measurement capacity could be used to support more rapid development and implementation of advanced approaches to performance measurement.

For example, level 1 ACOs, those without electronic health records or well-established patient registries, could rely in the near term on the meaningful measures that can be ascertained from claims data (eg, cancer screening, diabetes testing). These ACOs might be expected to progress rapidly to report on a more advanced set of measures, such as selected health outcomes (eg, blood pressure control), patient-reported care experience measures (eg, after hours access), and total costs of care. The Alternative Quality Contract developed by Blue Cross Blue Shield of Massachusetts is currently using such a set of measures for its ACO-like global payment program.⁹ Level 2 ACOs (those with site-specific electronic health records and registries) might be expected to add more advanced measures (eg, patient-reported health outcomes for selected conditions). Level 3 ACOs (those with comprehensive electronic health records across all sites of care) could be required to test and implement measurement systems that support practice improvement and accountability in such difficult areas as informed patient choice and health outcomes for a broad array of conditions.

All ACOs should be expected to participate in performance measurement to the extent of their capabilities, rather than relying on a lowest common denominator approach. Higher-level ACOs would not only be held to higher standards of accountability but would also contribute to advancing the field of performance measurement by developing and testing the new measures and measurement approaches that are needed. All ACOs would be expected to move to more advanced measures as quickly as possible.

A rapidly advancing performance measurement infrastructure could also help accelerate learning and reduce the risk of unintended harms. It is important to know not simply whether an ACO worked (improved care, reduced costs) but also how it worked. For example, what aspects of the ACO (eg, organizational structure, leadership, care processes) and of the local environment (eg, market structure, state health policies) contributed to its success? Because private payers are increasingly concerned about the market consolidation stimulated by the growth in ACOs, it will be important to evaluate the influence of ACOs on the quality and costs of care for the non-Medicare population.

It may also be worth considering the adoption of a common evaluation framework across the variety of delivery system and payment reform initiatives called for under the Affordable Care Act. The history of demonstration programs within Centers for Medicare & Medicaid Services is largely one of designing highly specific evaluation approaches and ensuring that there is no overlap among the clinicians, practices, health systems, or regions participating in the demonstrations. This is not likely to be possible or even desirable as reform proceeds. It is important to know not only whether bundled payments work but also whether they are more or less effective when combined with other reforms, such as ACOs or medical home programs. If the overarching aims of reform are shared, a common measurement infrastructure and framework will offer important advantages in allowing the effects of diverse innovations to be evaluated and compared on a level playing field.

The notion of accountable care has broad appeal. But only a robust, comprehensive, and transparent performance measurement system can reassure the public, physicians, hospitals, others who deliver care, and payers that ACOs are worthy of the name.