Dec 13, 2023

Charting a Path for the Medicare Advantage Value-Based Insurance Design Model: Innovating to Meet Person-Centered Needs


The Affordable Care Act (ACA) authorizes the Centers for Medicare & Medicaid Services (CMS) to test innovative health care payment and service delivery models that have the potential to lower Medicare, Medicaid, and Children’s Health Insurance Program spending while maintaining or improving the quality of care. The law requires an evaluation of each model to assess quality of care outcomes and changes in spending, and this evaluation is critical in determining the appropriate path for models, including expansion, modification, or termination.

One such model, the Medicare Advantage (MA) Value-Based Insurance Design (VBID) Model, tests an array of MA health plan innovations that have the potential to lower Medicare spending while improving the quality of care for people with Medicare. Through the VBID Model, MA Organizations (MAOs) may target supplemental benefits and reduced cost-sharing to certain enrollees, based on an individual’s socioeconomic status (SES) (using eligibility for the Part D Low Income Subsidy) and/or chronic health condition(s), to address that individual’s unique medical and health-related social needs. 

Since its launch in 2017, the VBID Model continues to grow in participation and evolve in its design. Regarding participation growth, in 2024, the model will include 69 participating MAOs (compared to nine in 2017) that will serve an estimated 8.7 million VBID Medicare beneficiaries (compared to 96,000 in 2017). The largest growth in MAO participation occurred between 2020 and 2022, likely due to: 

  1. The requirement in the Bipartisan Budget Act of 2018 (Public Law No. 115-123) (BBA 2018) to allow plans in all 50 states and territories to participate beginning in 2020 and, through the end of 2021, suspend the statutory requirement to modify or terminate the model if it is not expected to improve quality of care and/or reduce spending.
  2. Flexibilities added in 2020 allowing MAOs to target supplemental benefits based on a beneficiary’s SES.
  3. The opportunity for all types of MA special needs plans (SNPs) to participate in VBID. 

These changes have contributed to an evolution in the model’s interventions. For example, in 2024, 88% of available Dual Eligible special needs plans (D-SNPs) will participate in VBID across 49 states, the District of Columbia, and Puerto Rico (compared to a geographical reach of just three states in 2017 by a small proportion of non-SNPs). The most common flexibility under the model will be Part D cost-sharing reductions targeted to low-income beneficiaries (projected to reach 7.6 million beneficiaries), followed by food and nutrition-related supplemental benefits (projected to reach 7.3 million beneficiaries) (compared to 2017’s most common intervention of cost-sharing reductions for specialists for enrollees with certain chronic conditions that reached less than 60,000 beneficiaries). 

To continue to shape the VBID Model, the CMS Innovation Center uses findings from a robust, independent evaluation of the model, as well as feedback from model participants and others. In September 2023, the CMS Innovation Center released an early evaluation of the VBID Model with some initial findings from the first three years of the model under its statutory expansion (2020-2022). The results to date suggest that: 

  • VBID plans leverage the model as a tool for the direct provision of supplemental benefits that address medical and health-related social needs among underserved beneficiaries. 
  • VBID plans are associated with modest increases in quality of care. 
  • VBID’s most common intervention, eliminating Part D prescription drug cost sharing, is associated with greater adherence to treatment by reducing economic barriers to Part D prescription drugs and supports the Administration’s imperative to improve the affordability of prescription drugs.

In addition to these impacts on underserved communities and the modest increases in health care quality, the evaluation suggests that VBID plans had an increase in targeted enrollees’ risk scores[1] (which leads to higher risk adjustment payments) relative to comparable non-VBID plans, and VBID contracts were associated with increased rebates from improved Star Ratings, relative to comparable non-VBID contracts.[2] This increase in risk scores and rebates drove higher Medicare costs associated with VBID contracts/plans. 

As we continue to closely review the evaluation results—including the data and drivers of favorable and unfavorable findings—the CMS Innovation Center is taking immediate action to revise certain aspects of the VBID Model’s design to both strengthen our understanding of cost drivers, and to ensure the model has the right tools in place to be responsive to evaluation findings and statutory requirements. Beginning in CY 2025, CMS will:

  • Add new reporting requirements related to risk score trends and utilization of supplemental benefits. 
  • Incorporate additional participation eligibility requirements related to program integrity and compliance to better align with the MA Program and reinstitute a minimum quality requirement of three Stars or higher for participating MAOs. 
  • Require, as part of the model’s financial requirements, participating plans to show net savings to CMS over the course of CY 2025 and over the course of the model.[3]

In addition, CMS may consider limiting participation in the model for 2025 and making additional changes in the future. 

We welcome feedback on these changes from all voices, including beneficiary advocates, MAOs, providers, community-based organizations, researchers, and other interested parties. Please direct any comments regarding these changes to

Given that enrollment in MA is now approximately 50% of all people with Medicare and increasingly enrolling people from underserved communities, VBID has become a vital part of CMS’s overall strategy to ensure the MA program meets the health care and health-related social needs of MA enrollees while contributing to the long-term stability of the Medicare program. While the early VBID evaluation results are mixed, the CMS Innovation Center has a compelling history of objectively, and transparently, assessing each model, as it evolves, and leveraging the learnings to create opportunities to make each model and the Medicare program even more effective in meeting our goals for cost and quality, including health equity. We remain committed and confident that this same process for the VBID Model, and the changes beginning in CY 2025, will provide these same opportunities.

This blog focuses on the voluntary components of the VBID Model, except for the Hospice Benefit Component. 



[1] A risk score is a number representing an individual beneficiaries’ predicted health care expenditures, relative to the average Medicare patient, based on certain characteristics and health conditions.

[2] In the MA program, plans bid annually to enroll Medicare beneficiaries in their plan. Plans can bid below or above the benchmark rate, a rate set based on a formula in the statute that uses a percentage of traditional Medicare spending representing the maximum amount CMS will pay a plan. When a plan bids below the benchmark, then CMS and the plan share in savings known as a “rebate.” To finance the supplemental benefits and reduced cost-sharing offered through the model that aim to improve quality outcomes, MAOs generally bid below the benchmark. In doing so, savings are generated to the Medicare program as CMS shares in the savings from the reduced bid. Because rebates are higher for lower cost plans with higher quality in the form of Star Ratings, the VBID Model incentivizes these higher quality and lower cost plans to use rebates dollars efficiently by focusing supplemental benefits on those who may need them most. 

[3] Savings projections must be net of risk score trends attributable to model participation.