Beyond Pay for Performance — Emerging Models of Provider-Payment Reform
Meredith B. Rosenthal, Ph.D.
http://content.nejm.org
Escalating costs and the growing imbalance between primary and specialty care have increased the urgency of calls for fundamental reform of the health care payment system. At the core of the problem is the fact that the dominant fee-for-service model rewards volume and intensity rather than value. But although the faults in the way we currently pay for health care are obvious, it is much less clear what feasible approach would yield better results.
Earlier this decade, pay for performance took center stage as a tactic for realigning payment with value. Payers’ experiences during this period, as well as several major studies, clarified the limitations of this approach — characterized by some as putting lipstick on a pig. Both the enthusiastic adoption and somewhat lackluster early results of pay for performance have given rise to a broader payment-reform movement, with proposals and pilots emerging from a wide variety of stakeholders and policy leaders
The contours of proposed reforms of the health care payment system follow the fault lines of current reimbursement models — either undoing perverse incentives in existing payment approaches or augmenting the incentives for providing high-value care. A number of incremental payment-reform models that have gained traction over the past several years address individual issues; more ambitious reform proposals attempt to correct multiple shortcomings.
Among the incrementalist approaches embraced by many payers is enhancement of existing pay-for-performance programs through changes in scope, performance measures, and magnitude of funding. The changes appear to be focused on two perceived shortcomings of earlier efforts: too little impact on provider behavior and not enough focus on demonstrable benefit — including both health outcomes and spending — as opposed to process-of-care measures. At the same time, nonpayment for treatment of preventable complications has emerged as the mirror image of pay for performance. Early adopters of this approach, including HealthPartners in Minnesota, refuse to pay for “never events” (rare and preventable errors or complications); the Centers for Medicare and Medicaid Services (CMS) has cast a somewhat broader net, aided in part by new “present-on-admission” diagnostic codes.
The downward spiral of the primary care profession in terms of compensation, professional satisfaction, and numbers of new entrants to the field has sparked a payment-reform movement specifically focused on primary care. Prominent among these efforts has been a set of proposals wrapped around the notion of a “medical home” (sometimes called the “patient-centered” or “advanced” medical home). The medical home is a set of philosophical and structural elements designed to ensure that a physician practice (usually in primary care) takes responsibility for providing and coordinating timely and appropriate care for its patients.2 The medical-home payment model typically includes a case-management fee, tiered according to the extent and sophistication of office systems and other practice capabilities attained, and pay for performance to support the delivery of optimal preventive and chronic-disease care.
An alternative vision for primary care payment that acknowledges the functions encapsulated in the medical-home concept goes further by replacing fee-for-service payment with primary care capitation.1 This “comprehensive” payment model advocates payments computed (over a typical patient-panel size) to cover salaries for a multidisciplinary clinical team, infrastructure costs (e.g., the cost of implementing electronic health records), and other practice expenses that are deemed necessary for building a functioning medical home. Although primary care physicians would not pay for downstream costs such as referrals, the model includes substantial performance incentives for quality and cost efficiency (amounting to 15 to 25% of total payments).
Outside the primary care arena, some groups are turning to episode-based payment systems such as Prometheus Payment, developed by a panel of experts and stakeholders. Global case-payment rates for a given condition are developed on the basis of clinical standards for appropriate care rather than solely through examination of current patterns of care, which reflect high rates of underuse, misuse, and overuse. Calculation of payments includes risk adjustment and a warranty for care in the event of related complications. Performance incentives (equal to 10 to 20% of the case-payment rate) related to clinical quality, patient experience, and cost efficiency are also part of the model.
Geisinger Health System’s ProvenCare payment concept is also based on clinical quality standards as applied to a defined episode of treatment.3 For elective coronary-artery bypass surgery, for example, the ProvenCare payment includes preoperative care, all services associated with the surgery and inpatient stay, plus 90 days of follow-up care. The episode price set by the health system is based on the cost of routine services plus an amount equal to half the average cost of complications.
Meanwhile, the Medicare Physician Group Practice Demonstration program is a leading example of the shared savings model of payment reform, which resembles the soft capitation contracts of the 1990s. In this program, participating group practices agree to manage the care of a population of Medicare patients with the prospect of sharing in savings that accrue to Medicare. Savings are calculated as the difference between actual spending and the risk-adjusted spending trend in a given market. Once this difference surpasses 2 percentage points, savings are shared with the integrated physician groups involved, which can receive up to 80% of these savings by performing well on cost-efficiency and quality measures.
Similarly, in late 2004, the State of Alabama instituted a program whereby 50% of any documented savings associated with primary care physicians in the state’s primary care case-management program is shared with those physicians. Shared savings are allocated according to a point system that takes into account physicians’ scores on three risk-adjusted measures of performance (use of generic medications, emergency department use, and number of office visits) and an index of their actual-versus-expected total of allowed charges.
Although these approaches to payment reform span a wide range of models, a number of common themes emerge. The first is value-based payment: although cost control is a major goal of most reforms, clinical guidelines and quality measures play important supporting roles. For example, both the episode-based and comprehensive primary care payment models require payment levels to cover the costs of explicitly defined “best practices.”
The second theme reflects a lesson from earlier iterations of capitation-payment systems: the need to distinguish random variation in outcomes and patient mix from variation in practices and avoidable complications. The new CMS hospital payment rule is the most obvious example of an attempt to make such distinctions, but both the episode-based payment models and shared-savings approaches involve this type of accounting.
Finally, many of the payment approaches are inseparable from specific care delivery and organizational models. The medical homes are the most explicit examples of this trend, but it is also noteworthy that Medicare’s shared-savings model was piloted only in large, integrated health care systems. Policy developments in new models of accountability share this view that aligning provider incentives with payer goals will require organizational forms that can coordinate care more effectively than the fragmented current system.4
There are, fundamentally, no “new” methods of health care payment. Novel approaches such as those described here are new combinations of old ideas, with updated features such as improved risk adjustment. Economic theory, as others have long noted, suggests that such mixed payment models will function better than any single approach.5 Which recipe will yield the best balance of meaningful incentives for cost control and quality improvement, risk protection for providers, and selection incentives remains to be seen. The prospects for payment reform, however, hinge more on politics than on economics. Given that the two major goals of reform are to constrain spending growth and to move money from more intensive to less intensive settings — from doctors who carry endoscopes and scalpels to primary care physicians, for example — there will be substantial resistance to even the best-designed plans.
Dr. Rosenthal reports having an unpaid role in the design and testing of the Prometheus Payment system. No other potential conflict of interest relevant to this article was reported.
Dr. Rosenthal is an associate professor of health economics and policy at the Harvard School of Public Health, Boston.
