The Center for Medicare and Medicaid Innovation: The Case for Reform

Abstract

Congress created the Center for Medicare and Medicaid Innovation and vested it with extraordinary authority. CMMI tests the hypothesis that policy experts and civil servants, insulated from the vicissitudes of the political process and accorded quasi-legislative authority, could conceive and implement transformative health care financing policy innovations that Congress could neither devise nor enact. Twelve years after its founding, CMMI has disappointed these expectations. Only 6 of the more than 50 models it has tested have met de minimis measures of success (no increase in federal spending or diminution of quality). The incremental savings associated with these 6 models have been more than offset by a string of failed concepts. More significantly, none of the demonstration projects deemed successful has resulted in scalable reforms that meet statutory criteria of reducing health care spending or enhancing its quality. CMMI has increased federal spending without producing meaningful reforms of federal health care entitlement programs. This paper will review CMMI’s statutory authorities and track record and show how faulty fiscal assumptions made by the Congressional Budget Office can impede congressional efforts to reform CMMI. It will also discuss how the law has inverted the relationship between the legislative and executive branches, requiring an Act of Congress to prevent CMMI from effectively amending federal statutes. It recommends that CBO modify its assumptions to reflect that CMMI has not produced federal savings. It also recommends that Congress curtail CMMI’s powers and shoulder the obligation of reforming federal entitlement programs.

What we already know about this topic?

CMMI has disappointed expectations for producing innovations that would reform health care delivery and reduce medical expenditures.

What this paper contributes?

The paper explores the reasons behind CMMI’s failures, particularly as it relates to the roles of Congress and the executive branch in reforming federal health care entitlement programs.

Policy implications:

Congress should curtail the authority of CMMI and reassert its own role in curbing federal health care entitlement spending and improving the quality of medical care.

Introduction

Congress created the Centers for Medicare and Medicaid Innovation (CMMI) and vested it with extraordinary authority. CMMI’s statutory mandate calls for it to “test innovative payment and service delivery models to reduce program expenditures . . . while preserving or enhancing the quality of care.” ¹ In pursuit of these aims, CMMI has the power to implement demonstration projects that are unlimited in duration and national in scope, effectively amending federal statutes through notice-and-comment rulemaking. ²

The Department of Health and Human Services (HHS) has long had legal authority to conduct Medicare demonstration projects. Section 402 of the Social Security Act of 1967 authorizes the HHS Secretary (through the Centers for Medicare and Medicaid Services) “to develop and engage in experiments and demonstration projects.” ³ The statute directs the agency to use this authority to “increas[e] the efficiency and economy of health services . . . without adversely affecting the quality of such services.”

This pre-existing authority was much more limited than that extended to CMMI. Nor did it achieve the statutory aims of increasing health care efficiency. ⁴

CMMI is testing the hypothesis that policy experts and civil servants, insulated from the political process and accorded virtually unprecedented power, could conceive and implement transformative health care policy reforms that Congress could neither devise nor enact.

Donald Berwick, then the Acting Administrator of the Centers for Medicare and Medicaid Services, said that CMMI could become the “jewel in the crown” of the Affordable Care Act, ⁵ a sweeping law that expanded Medicaid, created a new entitlement program and gave the federal government unprecedented authority to regulate private health insurance and to impose tax penalties on the uninsured.

Berwick’s enthusiasm for CMMI was matched by the Congressional Budget Office (CBO), which adopted unconventional assumptions about the agency’s capacity to save the federal government tens of billions of dollars.

Twelve years later, CMMI has disappointed those expectations. Despite testing more than 50 models involving 28 million beneficiaries and over 528 000 health care providers and plans, the agency’s work has fallen far short of its promise. ⁶ Only 6 of its projects have met de minimis measures of success (no increase in federal spending or diminution of quality).

The incremental savings associated with these 6 models have been more than offset by a string of failed concepts. CMMI has increased federal spending without producing meaningful reforms of federal health care entitlement programs. More importantly, CMMI has not produced scalable changes to federal health entitlement programs that have meaningfully improved their quality or efficiency.

Despite the best intentions of its authors, CMMI demonstrations have been no more successful than those undertaken under section 402 authority. Assessing the effect of the longstanding section 402 authority, CBO observed, “A small share of demonstrations resulted in savings, [but] most had little or no effect on Medicare spending, and some increased Medicare spending.” ⁷ The same assessment applies to CMMI.

Despite this well-established record, CMMI has issued an “Innovation Center Strategy Refresh,” which promises policy initiatives that will “lead the way towards broad and equitable health system transformation.” ⁸

This paper will review CMMI’s statutory authorities and track record and show how CBO’s faulty fiscal assumptions can impede congressional efforts to reform the agency. It will also discuss how the law has inverted the relationship between the legislative and executive branches, requiring an Act of Congress to prevent CMMI from effectively amending federal statutes. It will conclude by arguing that CBO should modify its assumptions based on more than a decade of CMMI experience. It also will recommend that Congress curtail CMMI’s powers and shoulder the obligation of reforming federal entitlement programs.

CMMI’s Demonstration Authority

The Affordable Care Act (ACA) created a new section 1115A of the Social Security Act. That provision established CMMI to seek new and more efficient ways to finance and deliver medical services under Medicare, Medicaid, and CHIP. ⁹ The statute directs the agency to create models that address “a defined population for which there are deficits in care . . . or potentially avoidable expenditures.” ¹⁰

These models need not be budget neutral during the initial period, but the agency must terminate or modify a model unless the CMS Actuary certifies that it improved medical care quality or reduced its cost. ¹¹

It also authorizes the agency, through rulemaking, to expand a demonstration if it neither increases federal spending nor diminishes medical care quality and if the expansion “would not deny or limit coverage or provision of benefits” for eligible individuals. ¹²

In addition, the statute allows CMMI to waive numerous requirements of the Medicare and Medicaid statutes. ¹³ It also bars administrative and judicial review of most agency actions relating to its work, including which models to launch, their “elements, parameters, scope and duration,” and decisions to terminate or expand them. ¹⁴ It also waives federal laws relating to the confidentiality of information collected by the agency concerning the testing, evaluation, and expansion of models. ¹⁵

The bill allocated $5 billion to CMMI for fiscal year 2010, $10 billion for 2011 to 2019, and $10 billion for each successive 10-year period. ¹⁶ Congress has so far provided CMMI with $25 billion.

CBO’s Assumptions About CMMI

Congress reposed great faith in CMMI’s capacity to test and deploy innovations that would produce federal savings. That faith was matched by CBO’s.

In September 2016 testimony before the House Budget Committee, then CBO Deputy Director Mark Hadley laid out CBO’s rationale for believing that CMMI would yield federal savings of $45 billion between 2017 and 2026. ¹⁷

“The savings that CBO expects to result from the center’s activities stem largely from the judgment that successful demonstrations will be expanded and achieve savings.” ¹⁸

CBO thus assumed, not only that CMMI would conceive successful innovations, but that these innovations would be “expanded and achieve savings.” This is a notable departure from the agency’s well-established analytical rigor.

Its “judgment” that CMMI will produce fiscal savings rests neither on analysis of the agency’s projects nor other quantifiable bases. To convert this “judgment” into a forecast of 10-year savings of $45 billion, CBO assumed that CMMI will reduce Medicare spending by 0.1% annually. ¹⁹

Just as CBO had no programmatic basis for that estimate, Hadley told Congress that it could never retrospectively verify its accuracy.

“Unlike the center’s spending, the reduction in spending on Medicare benefits will never be able to be observed.” ²⁰

CBO thus based its estimate of CMMI’s budgetary effect on non-falsifiable assumptions.

CBO baked those estimates into its baseline. It regards any legislative proposal that touches on a CMMI project or a project that CMMI might in the future undertake as costing the federal government money. In a July 2015 blog, 2 CBO officials explained its reasoning:

CBO examines any legislative proposals that seek to enact approaches similar to ones that CMMI is testing, to determine whether HHS would do something different under the proposal from what it would do under current law. . .. To the extent that legislative proposals overlap with initiatives that CMMI is undertaking (or is expected to undertake), the potential additional savings would be reduced. ²¹

CBO has thus built a protective shield around CMMI. It has assumed unverifiable savings, trended them at the rate of overall Medicare spending, embedded those savings in its baseline and determined in advance that any congressional action touching on an actual or potential CMMI model would reduce those savings and increase the federal deficit.

In this way, CBO has established an impediment to CMMI reform.

CMMI’s Track Record

Had CMMI achieved the savings CBO forecast, its authority still would be troubling. But CMMI’s own assessment of its performance indicates that it has fallen far short.

By its own reckoning, CMMI has launched over 50 demonstration projects, only 6 of which have achieved savings. ²² That self-assessment is generous. The Pioneer ACO demonstration, for example, is one of the agency’s half dozen “successes.” It began with 32 self-selected participants in 2012 and ended with just 9 in 2016. ²³ That so few participants remained may be due more to their particular characteristics than the model’s validity. Its successor projects—the Next Generation ACOs—increased net Medicare spending, suggesting that the savings among the 9 “pioneer” participants was not broadly scalable. ²⁴

A demonstration project requiring ambulance services to obtain prior authorization before giving Medicare beneficiaries rides for repetitive, scheduled medical visits achieved some modest savings among patients with ESRD. An evaluation found “no adverse effects on quality of care.” ²⁵ The Secretary expanded this niche project nationwide, but it is hardly the stuff of health system transformation. ²⁶

These de minimis and rare CMMI successes must be balanced against a lengthy list of failures. Former CMMI Director Brad Smith summarized the agency’s record.

“The vast majority of the Center’s models have not saved money, with several on pace to lose billions of dollars. Similarly, the majority of models do not show significant improvements in quality, although no models show a significant decrease in quality.” ²⁷

Smith’s New England Journal of Medicine piece presents the estimated cost and savings of 30 CMMI projects. ²⁸ The 25 projects that increased costs to the federal government include the “Comprehensive Primary Care Plus” (net increase in Medicare spending of $4.5 billion) and the “Bundled Payments for Care Improvement – Advanced” project (net increases of more than $2 billion). ²⁹ No CMMI project has achieved estimated savings even approaching those levels.

CBO’s assumption of substantial and scalable savings does not comport with CMMI’s real-world results. It should revise its methodology and assumptions to align them with reality.

The Future Direction of CMMI

CMMI remains undaunted by failure. Its “Innovation Center Strategy Refresh” acknowledges its disappointing record but says it has learned “important lessons about how to transition the U.S. health system to value-based care.” ³⁰

The report proffers “a bold new strategy with the goal of achieving equitable outcomes through high quality, affordable, person-centered care.” ³¹

The new strategy is wide-ranging, diffuse, and more ambitious than its predecessor. In addition to the agency’s core statutory mandate to reduce federal health care spending and improve its quality, the new plan calls on CMMI to promote health equity, ³² address the social determinants of health, ³³ finance non-medical social supports, ³⁴ and reduce Medicare beneficiaries’ out-of-pocket medical spending, including on prescription drugs. ³⁵

Such an expansion in mission and scope would scarcely be feasible for a successful organization. Over more than a decade, CMMI has consistently fallen short of the more modest and targeted statutory objectives.

The agency struggles with several operational challenges. Consider the problem of “overlap,” described in CMMI’s 2020 report to Congress.

As the number of Alternative Payment Models (APMs) has increased, so too has the likelihood that individual beneficiaries will be aligned to, and thus receive care through, more than one CMS or CMS Innovation Center APM during overlapping performance periods. This circumstance is commonly referred to as “overlap.” Such overlaps have the potential to affect beneficiary attribution, payment, and evaluation findings. They can result in double counting of beneficiaries (when beneficiaries are attributed to two APM participants) and in “duplicative payment for value-based care.” ³⁶

The Strategy Refresh acknowledges that the problem persists.

Complex payment policies and model overlap rules in CMS Innovation Center models can sometimes result in conflicting or opposing incentives for health care providers (e.g., multiple shared savings models operating in the same health system). ³⁷

CMMI thus may have overreached by establishing overlapping models whose outcomes defy evaluation.

CMMI also faces problems. Former Director Brad Smith observed:

“In the Next Generation Accountable Care Organization model, the antiquated CMS data systems did not apply the correct demonstration code to thousands of claims . . . a situation that resulted in losses of approximately $50 million.” ³⁸

This failure to implement codes that are critical to properly aligning incentives and measuring outcomes suggests that CMMI has not overcome institutional indifference to quality and efficiency. ³⁹

CMMI is ill-suited and poorly equipped to carry out its statutory mandate, much less the more far-reaching goal of health system transformation.

The Role of Congress in Setting Medicare Policy

Congress makes law. The executive branch administers it. CMMI has reversed this constitutional polarity. It has seized the policymaking initiative, relegating Congress to an ancillary role. The agency’s Radiology Oncology Model (ROM) illustrates this role reversal.

CMMI published a proposed rule to establish the model in July 2019. ⁴⁰ The agency planned to randomly select facilities that provide radiation treatment to cancer patients and conscript them into the demonstration. Medicare would pay a discounted rate for radiation treatments at these facilities. That discounted rate would be site-neutral, meaning Medicare would pay no more for treatments in a hospital outpatient department than at a non-hospital facility.

The demonstration would have reduced Medicare payments to hospital outpatient departments. ⁴¹ CMS estimated 5-year savings of $290 million. ⁴² The final rule, published in September 2020, set a launch date of January 2021. ⁴³ CMMI subsequently delayed it to July 2021, citing COVID-19 concerns. ⁴⁴

Congress then intervened, prohibiting CMMI from initiating the demonstration before 2022. ⁴⁵

A subsequent CMMI rulemaking provided that the demonstration would begin on January 1, 2022, the earliest date Congress would permit. ⁴⁶ But lawmakers stepped in a second time, delaying the model’s launch until at least 2023. ⁴⁷

The agency beat a strategic retreat. In April 2022, it published a proposed rule saying that CMMI would establish a new start and end date in “future rulemaking.” ⁴⁸

The issue of whether Medicare should pay more for services performed at hospital outpatient departments is a contentious one. Congress partially addressed it by preventing most medical practices acquired by hospitals after November 2015 from collecting higher Medicare fees. ⁴⁹

The volume of outpatient services billed at the higher Medicare rate has nevertheless continued to increase. ⁵⁰ A rulemaking by CMS to eliminate the payment differential for routine office visits, even for medical practices that hospitals acquired before November 2015, was the subject of federal litigation. ⁵¹

That litigation was already underway when CMMI first proposed the radiology oncology model. The agency seized the policy initiative, and Congress reacted. Even after Congress twice blocked the model’s implementation, CMMI has not formally abandoned plans to launch it.

This inversion of the constitutional order is encoded in CMMI’s DNA. Congress bears the constitutional obligation to set Medicare reimbursement policy. CMS administers that policy. Congress should assert its constitutional prerogatives and rein in CMMI.

That statement invites the observation that Congress itself has failed to reform the program. Many viewed CMMI as an antidote to congressional inertia. ⁵² The agency has fallen short of its promise in part because it is no more immune to interest group influence and political pressure than are lawmakers. Unlike the agency, Congress alone can rewrite federal statutes. There can be no meaningful health care entitlement reform apart from congressional action. Congress’s track record on such reform is no better than CMMI’s, but it is the only entity with the constitutional authority to enact such reform.

Reforming CMMI

Long before CMMI’s creation, CMS had the authority to conduct demonstration projects. ⁵³ Those projects were most often carried out at the behest of lawmakers, who sought additional information about a potential policy change before amending the law. ⁵⁴ When a demonstration project yielded positive results, the Secretary would ask Congress to amend the law. This is an appropriate division of labor between 2 co-equal branches and one Congress distorted by creating CMMI. Had the agency achieved substantial health care savings and quality improvements, its authority still would be troubling. But it has failed on both counts.

Congress should reassert itself and fulfill its obligation to reform federal health care entitlement programs. It should consider abolishing CMMI and restoring HHS’s pre-existing arrangement for launching demonstration projects. Short of that, lawmakers should take a more active role in program innovations by adopting legislation directing the department to test ideas Congress originates.

Congress also should curtail CMMI’s waiver authority and open more of its decisions to judicial review. It should also consider narrowing or eliminating the agency’s authority to compel medical providers to participate in its models.

Finally, and most importantly, Congress should strip CMMI of the regulatory power to extend models nationwide. Amending the Medicare statute is a congressional prerogative; lawmakers cannot outsource it to the executive branch.

Any serious CMMI reform would run headlong into CBO’s unrealistic budgetary assumptions. Under these assumptions, CBO would “score” reform legislation as increasing federal deficits. Years of real-world experience should convince CBO to amend its disproven hypotheses, removing a major obstacle to reform.

Conclusion

CMMI was born with great promise. Congress outsourced federal health program innovation to bureaucrats uninfluenced by voters. Regulatory alchemy would do what lawmakers could not: transform medical care delivery, while reducing its cost and improving its quality. Despite its undistinguished past, those assumptions animate CMMI’s future vision. Congress should reassert its policymaking authority and fulfill its obligation to reform health care entitlement programs. The executive branch is essential to that enterprise, but it is Congress’s junior partner.