Outlook for Alternative Payment Models in Fee-for-Service Medicare

The warnings are now familiar: health care spending for retirees in the United States is unsustainable, the Medicare program is going bankrupt, and the program is in desperate need of reform. The financial burden the Medicare program imposes on the federal budget and the national economy is substantial. Medicare’s total obligations were 3.5% of the gross domestic product in 2013 and are projected to increase to 5.4% by 2035, primarily due to expected growth in enrollment from 52.3 million to 86.8 million beneficiaries, and will substantially strain the economy and the federal budget.¹ Alternative payment models (APMs), which are built on the existing fee-for-service foundation but include new payments linked to the effective management of a population or episode of care, have been proposed as potential solutions to restrain costs. In this Viewpoint, we describe the status of the APMs being tested in the Medicare population and the outlook for these models to reduce program expenditures.

The Department of Health and Human Services has announced an effort to replace 20% of Medicare fee-for-service payments with APMs in 2015 and has set goals of replacing 30% by the end of 2016 and 50% by 2018.² Under the Medicare Access and CHIP [Children’s Health Insurance Program] Reauthorization Act (MACRA), beginning in 2019, physicians who receive a substantial proportion of revenues through APMs will receive a global 5% bonus and opt out of the consolidated pay-for-performance program, termed the Merit-Based Incentive Payment System (MIPS). Although the legislation applies to services provided by physicians and other practitioners, physicians play an important role in all APMs, which affect a broad range of Medicare services.

Under current law, there are 2 principal avenues by which clinicians and health care entities may participate in APMs: expansion of the Medicare Shared Savings Program or expansion and development of APMs tested through the Center for Medicare & Medicaid Innovation (CMMI). The Medicare Shared Savings Program is a voluntary program in which accountable care organizations (ACOs), which are collections of health care organizations that agree to accept accountability for a population of beneficiaries, enter into a contract with the Centers for Medicare & Medicaid Services (CMS), whereby the ACO is eligible for a shared savings payment if it is able to maintain payments to attributed beneficiaries below a financial benchmark, while simultaneously demonstrating adequate performance for a set of quality measures. CMS anticipates that 404 ACOs will serve 7.3 million beneficiaries in 2015.³

Established by the Affordable Care Act in 2010, the CMMI was funded with approximately $10 billion to test novel payment and delivery models that aim to reduce costs while preserving or enhancing the quality of care. The ACA also granted the Secretary of Health and Human Services the authority to select models for testing and to expand models that successfully reduce program expenditures while improving or maintaining quality. CMMI has launched several models that are testing or will test APMs in the Medicare Fee-for-Service population. These include 2 models that will test variations of the Medicare Shared Savings Program, 3 models that test APMs for primary care practices, and 3 models that are testing APMs focused on episodes of care or condition-specific populations. The eTable in the Supplement summarizes these models and results to date.⁴

Evaluations of CMMI models have demonstrated at best modest gross savings (eTable in the Supplement), with the Pioneer ACO program receiving actuarial certification for expansion. The Medicare Shared Savings Program is not subjected to annual independent evaluations, and little is known beyond individual ACOs’ performance relative to the financial and quality benchmarks established through regulations. CMS altered the original Medicare Shared Savings Program regulation to allow selection of a payment arrangement without shared losses, which has been selected by nearly all ACOs, and recently revised regulations to allow ACOs to complete an additional 3-year savings-only track. Overall, performance in this program has been highly variable. The most comprehensive ACO evaluation to date has demonstrated per-beneficiary per-month reductions of $35.62 in the first year and $11.62 in the second year of the Pioneer ACO program.⁵ Little is known about the specific-delivery system strategies that are associated with the success or failure of these models, ACO investments, and organizational dynamics or about whether these models have begun to transform the care of Medicare beneficiaries. Identification and dissemination of successful strategies is critical to the success and expansion of these models.

CMS could use the MIPS to institute substantial incentives to reduce resource use, with up to a 9% penalty or 27% bonus for total physician payments, with resource use comprising up to 30% of the overall score. Historically, CMS has had difficulty measuring resource use by physicians. For example, early results for the Physician Value-Based Payment Modifier, which was limited to practices with at least 100 practitioners, scores most practices as average for both resource use and quality.⁶Assessing the performance of smaller or specialized practices will be even more challenging, potentially limiting the importance of these incentives, or potentially distorting these practices with incentive programs that do not fit their practice model. Physicians could simply choose not to participate in APMs without highly favorable arrangements, anticipating that a worse alternative will never materialize.

In addition to controlling expenditures, APMs have been proposed as a mechanism to improve quality of and patients’ experiences with care. This objective was less integral to prior payment system overhauls that focused on cost control, such as the conversion of hospital payments to diagnosis related groups. The desire to improve quality of care is laudable, and the demonstration of increased performance on an assortment of measures seemingly provides reassurance that efforts to reduce costs are not due to withholding necessary care. However, many of the measures in use are redundant to efforts to control costs (eg, rates of ambulatory-sensitive admissions), incentivize processes of uncertain significance where clinical guidelines increasingly emphasize patient-centered care goals (eg, lipid, glycemic, and blood pressure targets), or are limited because of score compression and ceiling effects (eg, many patients’ experience-of-care measures). Reporting may also have unintended consequences given the cost of generating data and the potential to crowd out more impactful quality-improvement efforts. Moreover, it is unclear whether these measures of quality matter to beneficiaries or referring clinicians, because few use this information when choosing clinicians or hospitals.⁷ Although opportunities to improve the quality and experience of care undoubtedly exist, beneficiaries’ opinions of their physicians are high: in a 2014 survey with more than 300 000 beneficiary responses, 90% would definitely recommend their clinician and only 2% would not.⁸

Recognizing the limitations of current performance assessment, MACRA directs the secretary of Health and Human Services to draft a plan for developing better performance measures by January 1, 2016. It also directs the secretary to solicit input from professional organizations and other stakeholders in developing measures of quality and resource use. The extent to which these stakeholders accept accountability for developing solutions that will meaningfully change practice is critical to the future of APMs and the MIPS. As evidenced through the Choosing Wisely campaign, in which many professional organizations avoided addressing services with important financial implications,⁹ these organizations could choose to wholly embrace the objectives of payment and delivery reform or protect the financial interests of their members. At the same time, several specialty societies are participating in the qualified clinical registry pathway for CMS quality reporting. This pathway allows their members to satisfy performance requirements through participation in practice-specific quality-improvement initiatives. By permitting greater flexibility in measure selection, promoting dissemination of best practices among peers, avoiding competition among peers through a relative scoring system, and using intrinsic motivation to improve performance, this mechanism could ultimately be an effective strategy for improving the quality of care.

The ability of CMS to control Medicare expenditures through delivery and payment reform will be closely scrutinized. In addition to measuring the uptake and summative effects of APMs for Medicare and other payers, indicators of APMs as a viable mechanism for controlling expenditures could include identification of underlying delivery system changes associated with success, determination of investments of time and capital returns for participants, demonstration of sustained changes in business strategy away from volume-driven fee-for-service efforts, engagement of key stakeholders in defining and measuring value, and clarification of the strategic objectives of the program by CMS and other stakeholders. Alternate payment models may prove to be a long-term solution that fundamentally transforms the delivery system, a stepping stone to capitation, including Medicare Advantage, or simply a small contributor to efforts to improve quality and control costs. The next few years will require bold leadership and accountability in both the public and private sectors by physician-leaders and from the US public. Medicare will continue to consume an increasing share of scarce public resources, and a failure to address this economic problem may result in even less palatable solutions down the road.