Although alternative payment models (APMs) have existed for decades, in 2010 the Affordable Care Act accelerated their development and adoption. In particular, the Affordable Care Act launched the Medicare Shared Savings Program for Accountable Care Organizations (ACOs) and created the Centers for Medicare and Medicaid Innovation (CMMI). Charged with developing and testing alternative payment models, CMMI has the authority to expand models that curtail spending without diminishing quality or improve quality without increasing spending. The promotion of APMs was further encouraged in 2015 by the Medicare Access and Children's Health Insurance Program Reauthorization Act, which provides a 5% payment bonus to organizations participating in Advanced APMs,1 and by the Physician-Focused Payment Model Technical Advisory Committee, which was charged with recommending APMs to CMMI.
Now, a decade later and as we transition to a new presidential administration, is an opportune time to assess the path forward. Specifically, given its legislative charge, CMMI adopted a paradigm that combined 2 ideas: (1) test and diffuse and (2) let many flowers bloom. Test and diffuse was motivated by the belief that the evidence base supporting payment reform was not strong enough and that we needed to test new models before launching them broadly. Let many flowers bloom reflected the belief that testing more models would generate more evidence and that launching more models would better engage more stakeholders. Moreover, given its testing charge, CMMI naturally looks to develop and test new models after evaluations of previously launched models are complete.
Yet this approach, understandable in 2010, is not well-suited to promote transformation of the health care system to meet the overarching goals of eliminating low-value care, lowering spending, and improving quality. A better approach may be to recognize the overlap and interactions among models and move to a portfolio paradigm in which a set of complementary models is offered to meet policy goals.
Specifically, APMs recognize the centrality of clinicians and health care systems in guiding care. All APMs are founded on a basic belief: there is waste in the health care system, and the fee for service (FFS) model discourages elimination of the waste. For example, under FFS, health systems that eliminate wasteful or even low-value care lose out financially;2 those that improve quality are seldom substantially rewarded. In contrast, APMs allow clinicians and health care systems to benefit financially if low-value care is eliminated and if quality is improved.
Although there is a common thread across all APMs (extending payment and accountability across organizations and time), the details vary widely across models. Some, such as the ACO models, are broad, holding participants responsible for total spending and quality. Others, such as episode-based payment (EBP) models, are narrower, holding clinicians and health care systems responsible for the total spending and quality of an episode of care. Within each of these model types, there are many variations of key program parameters, including how spending targets are set, how much risk is imposed, how beneficiaries are assigned to organizations, and how quality is rewarded. Given the lack of consensus on the best APM design, the need to produce evidence on different approaches, and the desire to encourage participation, the many flowers paradigm was a reasonable initial approach.
Waste as an Asset
The key flaw with the many flowers paradigm is that it tends to treat each model as independent.3 Every model launched assigns a portion of the savings associated with waste reduction to program participants, such as medical groups. Effectively, waste is an asset. The organizations to whom the waste is assigned can profit by eliminating it. When more models are launched, the gains from waste reduction get split among more organizations. This reduces the incentives for any given organization to reduce waste (because others may capture some of the savings) and which ultimately, reduces participation incentives. For example, if EBP models are run concurrent with ACO models, savings from reduced episode-related post-acute care are split among the numerous participating organizations.4,5 Even if rules are put in place to separate the models (eg, preventing individuals assigned to ACOs from being counted in EBP models), the incentives for each model are changed because the volume in any given model would decline, making fixed costs more difficult to recover. The existence of many models not only complicates evaluation because it exacerbates challenges with nonrandom participation, but it also fundamentally changes operational decisions for program participants.
Even if a model is more successful than FFS, it may not be as successful when launched beside other models. One cannot sum the results from a series of independent evaluations to estimate aggregate effects. Even if one wants to let many flowers bloom, one would not plant them in the same hole.
A related concern is that the plethora of models requires substantial managerial attention to decide which, if any, to participate in, particularly when the models are very similar. Additionally, a final concern with the current mindset is that programs come and go. Without a sense of the broader trajectory of payment reform, it is hard to get organizations to commit to transforming their business models. They can always wait for a new model or switch among models.
A Portfolio Paradigm
The health system would be better served by adopting a portfolio paradigm in which a limited number of models are implemented that work together to achieve overarching goals. Such a portfolio may require legislative changes as well as a change in mindset. Like any good portfolio, the models that should be included, and their parameters, will depend on how they interact with one another. For example, there could be a single population-based payment program with multiple tracks, each with different levels of risk appropriate for organizations with differing capacities to bear risk. Other model types, such as EBP models, could be introduced when they augment savings potential as opposed to simply reassigning the savings to others. Of course, there are many other reasonable visions for a portfolio, and the details of the models in the portfolio will need to be continuously evaluated as they evolve over time. As more information becomes available, model offerings and designs may change to improve performance. Yet, the fundamental direction should be stable. The policy task is to develop the portfolio so that the models complement each other while competition between and among models is minimized.
The core portfolio design question is who owns the waste, and thus, who should profit from its elimination. It is reasonable to assign accountability for any type of waste to the organization most able to eliminate it. Often, the clinician or health care system most proximate to the wasteful care is best equipped to eliminate it, leading to a set of narrow models with complex interactions because many patients have many different conditions and get care from several different medical practices. However, such an approach may discourage a broader APM framework. As a result, portfolio designers must ponder how much waste could be captured by a portfolio of narrow models, how much waste broader models might eliminate, to what extent narrower models reduce the appeal of the broader models, and how other challenges with broader models (eg, risk adjustment and attribution) can be overcome.
Building the appropriate portfolio of APMs will require much more thought. Still, if we truly want to create a lasting system transformation, we must move away from the many flowers paradigm to a portfolio approach.
Published: May 6, 2021. doi:10.1001/jamahealthforum.2021.0232
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Chernew ME. JAMA Health Forum.
Corresponding Author: Michael E. Chernew, PhD, Department of Health Care Policy, Harvard Medical School, 180 Longwood Ave, Boston, MA 02115 (email@example.com).
Conflict of Interest Disclosures: Dr Chernew reports grant support from the Laura and John Arnold Foundation; research grants from Signify Health, Ballad Health, and HCSC; equity in Archway Health, Health at Scale, and Virta Health; being a partner in VBID Health; being Chairperson of MedPAC; being on the advisory boards of BCBSA, BHI, HCCI, and NIHCM; and speaking honorariums from AHIP, BCBS of Florida, Humana, the American College of Cardiology, and the American Medical Association, all outside the submitted work.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Chernew ME. A Path Forward for Alternative Payment: Build a Portfolio Not a Garden. JAMA Health Forum. 2021;2(5):e210232. doi:10.1001/jamahealthforum.2021.0232