The End Of The MA Value-Based Insurance Design Model: What Next?

Editor's Note

This article is the latest in the Health Affairs Forefront series, Supplemental Benefits in Medicare Advantage, featuring analysis and discussion from all stakeholder perspectives with the goal of informing the rapidly changing policy landscape on this topic. Additional articles will be published through mid-2025. Readers are encouraged to review the Call For Submissions for this series. We are grateful to The SCAN Foundation for its support of this work.

The Centers for Medicare and Medicaid Services (CMS) recently announced that the Medicare Advantage Value-Based Insurance Design (MA-VBID) Model would be discontinued at the end of 2025 due to “substantial and unmitigable costs to the Medicare Trust Funds.” First implemented in 2017, MA-VBID provided MA plans with added flexibility to reduce cost-sharing and provide other supplemental benefits (e.g., transportation to medical appointments and healthy food assistance) to enhance access to high-value services and prescription drugs. The model, focused on chronically ill and underserved populations, was implemented with the goals of improving beneficiaries’ health and lowering avoidable medical spending.

The law authorizing CMS to test innovative payment and delivery models such as MA-VBID requires CMS to terminate models that increase costs to the Medicare program. Evaluation findings report significantly increased costs associated with the model – an additional $2.3 billion in 2021 and $2.2 billion in 2022 – leaving CMS with no viable option to continue. Although the current model will be terminated given this regulatory requirement, this should be interpreted as an assessment of the specific design of this program as opposed to a conclusion about the general cost (or potential) of VBID as a concept. This action should also not be interpreted as a signal of the end of V-BID programs which reduce the financial and structural barriers to essential clinical services faced by Medicare beneficiaries.

Motivation For The MA-VBID Model

One-third of Medicare beneficiaries report that it is somewhat or very difficult to afford health care costs. It is well-established that out-of-pocket (OOP) cost-related non-adherence leads to lower utilization of preventive care, prescription drugs, and chronic disease services, potentially leading to future health complications and greater spending over time. To mitigate the negative impact of OOP costs and structural barriers on access to high-value care, Medicare implemented several V-BID programs which reduce cost-sharing or provide supplemental benefits, such as the ACA provision that eliminates cost-sharing for specific preventive services, a $35 monthly OOP cap on insulin products and a $2000 maximum OOP expense for Part D prescription drugs. A robust evidence base demonstrates that V-BID interventions increase adherence to high-value services, reduce disparities, improve patient outcomes, and in some instances decrease expenditures.

Implementation And Evolution

Responding to a series of bipartisan Congressional bills calling for the use of V-BID in the Medicare program, the MA-VBID model was launched in 2017 with nine plans in seven eligible states. Previously, all Medicare beneficiaries were required to have a uniform benefit design. Under the model, MA plans were given the flexibility to reduce cost sharing for designated clinicians and for services used by individuals diagnosed with specific clinical conditions. Although regulations require that Center for Medicare and Medicaid Innovation (CMMI) models achieve cost neutrality or savings, the primary goals of V-BID programs in general, and the MA-VBID model specifically, are to increase access to care, enhance equity, and ultimately improve the health of Medicare beneficiaries – not to save money. That said, VBID programs could be designed to be cost saving or budget neutral, but such alternatives would require a longer evaluation duration and/or include interventions that reduce spending on low value care.

In 2019, the model was expanded to 25 states, and by 2020, the model was expanded to all 50 states and included Special Needs Plans (SNPs) and regional PPOs, with cost-sharing customization allowed based on socioeconomic status, chronic conditions, and non-health-related benefits such as transportation and healthy food. By 2023, approximately one-fourth of all MA plans had adopted the V-BID model. It is important to note that, from its inception, MA-VBID received rare bipartisan political support in Congress and was expanded under both Democratic and Republican administrations.

Evaluation Results

Clinical

Evaluations of the MA-VBID program report that the primary clinical goals were achieved: adherence to preventive care such as breast cancer screening improved; use of essential services by enrollees with chronic medical conditions such as diabetes, hypertension, and cholesterol medications increased; and quality metrics improved. 

Financial

When appraising the financial implications of MA-VBID, is important to note almost all the clinical services incentivized by the model were cost-effective – requiring added expenditures to achieve improved health outcomes – and not cost saving. To temper the ambitious, but unlikely, expectations of reduced expenditures resulting from V-BID programs, our colleague Michael Chernew, PhD, of Harvard Medical School would often say, “Remember, if you buy more things that don’t save money, you won’t save money.”

Thus, any prospect that MA-VBID in its current form would lower aggregate short term spending would be unrealistic for several reasons, including: 1) the model’s focus on enhancing the use of cost-effective services, 2) a short evaluation timeframe that minimized the ability to capture future cost offsets from better chronic disease management, and 3) regulations explicitly did not allow payment or benefit design changes that would produce savings by decreasing the use of low-value services.

In their announcement of the winding down of the MA-VBID Model, CMS attributed two main drivers of the added expenditures to the Medicare Trust Funds: increased risk adjustment scoring and higher plan rebate payments.

Risk Adjustment

CMS payments to MA plans are determined in part by risk adjustment, a formula designed to predict medical spending for plan enrollees based on their health conditions. CMS spends more as risk scores increase, following the logic that patients with more health conditions will require more medical services. Thus, the potential for higher payments from CMS created a financial incentive for MA plans to enroll new patients with chronic conditions and identify chronic conditions in those enrolled. The 2023 MA-VBID RAND evaluation reported that risk scores increased substantially more for enrollees in MA plans participating in the MA-VBID Model than for similar enrollees in MA plans not participating in the model. Specifically, the analysis reported that the increases (relative to the non-VBID comparison group) were among individuals diagnosed with chronic conditions that were commonly targeted by VBID participants, such as congestive heart failure, diabetes with chronic complications, and rheumatoid arthritis, suggesting that VBID “may be contributing to increases in risk scores by enabling plans to identify new or reestablish existing diagnoses.” To the extent that these increases in risk scores indicate more patients who are identified with, and treated for, chronic conditions, the resultant added spending has clinical benefits. In the case where more accurate coding is not associated with changes in patient management that improve clinical outcomes, the spending is not beneficial. 

Rebates

The second factor driving CMS expenditures to MA-VBID plans was higher rebate payments. Rebate payments to MA health plans, including V-BID plans, increase with higher-risk (sicker) enrollees and higher-quality plan performance.

The rebate financing works like this:

MA plans submit bids cover their expected costs of providing Parts A and B services in each county. Each bid is compared against a “benchmark,” spending levels set by CMS to cover the provision of Medicare Parts A and B Traditional Medicare services. When bids are lower than benchmarks (as is usually the case), MA plans can use a portion of the difference – the rebate – to offer supplemental benefits (benefits not covered by TM such as lower premiums; reduced cost sharing; or coverage of additional services, such as vision and dental care). The lower the bid, the higher the potential rebate.

The rebate is partially determined by the MA plan’s star rating, a measure of the quality of care provided. MA plans that achieve a star rating of 4.5 or 5 keep 70 percent of the rebate amount; plans with a star rating of 3.5 or 4.0 keep 65 percent; and plans with 3.0 stars or below keep 50 percent. Therefore, as star ratings improve indicating better quality, CMS spends more.

Thus, several interacting factors influenced higher CMS spending on MA-VBID plans, including: 1) an increase in risk scores 2) and increase in benchmarks due to better performance on star ratings, and 3) an increase in rebates, possibly due to increases in quality of care, allowing plans to keep more of the difference between the plan bid and the CMS benchmark. Although it is impossible to precisely parse out the contribution of each of these factors to higher CMS costs, the greater the contribution from star ratings, the more that improved healthcare of Medicare beneficiaries contributed to the added costs.

Lessons From MA-VBID And Future V-BID Applications

The unwinding of the MA-VBID Model does not portend the end of the use of V-BID to improve access to high-value services for Medicare beneficiaries. Enrollees who remain in MA will likely be able to access many of the newly added benefits after MA-VBID’s termination because many of the supplemental benefits previously permitted only in the V-BID model can now be offered by all MA plans. Since the model launched, Congress and CMS made legislative and regulatory changes, such as CMS’s reinterpretation of uniformity flexibility (i.e., who can be in the target population(s) for certain benefits in MA plans) and the Bipartisan Budget Act of 2018 that created Special Supplemental Benefits for the Chronically Ill (SSBCI). SSBCI allows MA plans to offer interventions similar to those available under the V-BID model to MA members with one or more complex chronic conditions at high risk of hospitalization or adverse health outcomes and who require intensive care coordination.

Moving forward, there are alternative versions of the MA-VBID model that would produce health benefits and meet existing regulations of cost neutrality. One approach would include long-term savings achieved from offsets in spending from improved health outcomes that are currently not captured. For example, actuarial modeling of a V-BID program that reduced OOP costs for specific services for patients with congestive heart failure produced net plan savings, as the added spending on incentivized services were fully offset by savings from reduced hospitalizations and emergency room visits over time.

Another tactic would allow plans to implement policies that reduce unnecessary spending. The current MA-VBID model does not permit plans to implement policies that discourage the use of low-value services. The savings achieved from allowing payment programs and benefit design changes that reduce spending on low-value care could compensate for the added spending on high-value services. This reallocation of medical resources to spend more on high-value services and less on those that provide no clinical benefit follows the approach of a VBID concept proposed by the first Trump Administration. The use of this actuarially validated, cost-neutral plan design concept was strongly supported in the 2021 CMS Notice of Benefit and Payment Parameters rule.

Although the MA-VBID model will soon be discontinued, Congress and the second Trump Administration can build on V-BID policies implemented in the first Trump administration to lower out-of-pocket medical costs, such as the Health Savings Account-eligible High-Deductible Health Plans (IRS Notice 2019-45), which allowed coverage of 14 specific chronic disease services, including several medication classes, prior to beneficiaries meeting the plan deductible. The administration could also implement a version of the Medicare Two Dollar Drug List Model, proposed by CMMI in 2024, which offers low-cost, clinically important generic drugs that would reduce out of pocket costs for almost every beneficiary that uses their prescription drug benefit, leading to improved medication adherence and ultimately better health outcomes.

The winding down of the MA-VBID program should not suggest that the model did not improve the health of those enrolled – it did. Rather, the model succumbed to a specific regulation requiring that the program lower spending, which it did not do (as we predicted for reasons stated above when it was proposed). As there are practically no health programs or clinical services – including almost all health quality metrics – that save both lives and money, the merits of health-enhancing programs such as MA-VBID must not be tied to cost-neutrality. Instead, assessments of whether the gains in health achieved are worth the additional money spent should be conducted on the same level playing field as other health care interventions.