Policy Brief: Next Generation ACO Model Package
April 26, 2022
By Travis Broome, SVP, Policy & Economics
In October 2021, Broome et al. described using the Medicare Shared Savings Program (MSSP) as a “chassis” for innovation. This chassis could test new innovations with only a fraction of the effort required to build a new model from the ground up and establish the MSSP infrastructure as the starting point for many providers, not a competitor to new Innovation Center models. To further demonstrate this idea, the Aledade policy team developed a series of blogs describing how this would work for four innovations: primary care capitation, value-based insurance design, reducing inequity in health care, and incorporating desired elements from the Next Generation ACO model.
CMS’ new strategy paper calls for getting all Medicare and Medicaid beneficiaries into an accountable care relationship by 2030. Generally, the ACO model has successfully reduced costs and improved quality for Medicare beneficiaries by providing concierge-level access to primary care services. But one ACO model, the Next Generation ACO, which concluded in 2021, has received a more mixed assessment. 
The Next Gen model has achieved gross savings but narrowly missed out on attaining net savings (savings after payments to the ACOs) according to the latest evaluation report covering the program through 2019. How narrowly? With a cumulative net spending impact of +0.4 percent, a more substantial discount could have easily turned the program into a net saver for Medicare. 
Looking forward, participants in Next Gen have requested that CMMI continue the model because they see ACO REACH (formerly Direct Contracting) as an imperfect alternative. We are proposing an alternative. This brief presents a new pathway for Next Gen participants by creating a new track within MSSP utilizing CMMI authority (though we believe some parts of our proposal could be implemented using CMS’s authority to test new payment models within MSSP).
The “Next Gen” Track within MSSP would include four elements. 
1. A discount instead of a shared savings rate.
Some organizations value being at “full-risk” to create alignment with other payers and meet market expectations for risk bearing. Today, MSSP does not offer this alternative. Our proposed Next Gen Track in MSSP would offer an option for full risk with a discount instead of shared savings. 
We recommend setting the discount rate equivalent to an eight percent savings rate multiplied by the Enhanced track’s 75 percent share rate. This formula leads to a two percent discount rate. ACOs electing to use the discount rate would be eligible for 100 percent of the savings and responsible for 100 percent of the losses. While Next Generation ACO had an 80 percent option, we do not propose to continue this option, preferring to create a truer “full-risk” track. 
2. Allow ACOs to select the right providers.  
Under MSSP, ACOs must include all providers associated with any participating Tax Identification Number (TIN). But TINs were created for federal tax purposes and often don’t reflect an appropriate mix of providers for a value-based organization. For example, if a TIN includes both a primary care practice and a cancer center, including this TIN would bring disparate populations into the ACO and create difficulties for implementing new population health strategies and properly measuring cost and quality. 
Therefore, the Next Gen Track in MSSP would allow ACOs to include only certain clinicians practicing in an organization (identified by NPI) to create a more coherent ACO. To control for potential gaming, an ACO would have to submit to CMMI a written explanation if they wish to exclude NPIs associated with family medicine, geriatricians, or internal medicine physicians. Similar explanations would be required to exclude nurse practitioners, physician assistants, and clinical nurse specialists. All other physicians could be included or not included at the discretion of the ACO as they would not drive assignment in Step 1 of the MSSP assignment methodology.
3. Building a network. 
One of the most popular aspects of the current Next Generation and Direct Contracting models allows ACOs to enter into agreements with health care providers and facilities to alter the payments they receive from Medicare. This can facilitate better care coordination across disparate settings and address Medicare pricing failure, which is a significant source of waste. Preferred providers (affiliated with the Next Generation ACO, but not driving attribution) make up one-third of the health care providers in the Next Generation ACO. 
CMS has built the infrastructure to operationalize this capability, and bringing this into a new MSSP track would help scale the ability to address pricing failure across Medicare.
4. Provide a range of payment waivers.
Today, Next Gen ACOs also utilize waivers more often than MSSP ACOs. Overall, 78 percent of Next Gen ACOs used the SNF 3-Day Rule waiver compared to 22 percent in MSSP. 
The waivers provided in our proposed Next Gen Track would include the telehealth site restriction waiver, SNF 3-day rule waiver, post-discharge home visit waiver, and care management home visit waiver. This would bring these opportunities for innovation into the country’s largest value-based program in an administratively simple way.
ACOs should be able to select one or all of the available waivers. ACOs electing prospective assignment should be able to use the waivers on all assigned beneficiaries and, more important, ACOs electing preliminary prospective alignment with retrospective reconciliation should have their assigned beneficiaries eligible for the waivers updated quarterly. Some waivers may span a quarter. No Medicare beneficiary receiving services under a waiver should have those services suspended if they subsequently fall off the ACO’s quarterly assignment list. 
Next Steps
No value program has more provider participation than MSSP, a credit to its financial incentive design. CMS should start where participation is most robust and where the existing infrastructure can support additional innovation. Using MSSP as the chassis removes most of the challenges in setting financial benchmarks. This new track would combine well-understood elements of existing models and, through its integration with MSSP, create no new model overlap issues. CMS would also encourage greater risk taking by providing more tools in a well-understood financial methodology,. 
This new model would also help CMS achieve many of the goals included in its 10-year vision for innovation, especially those around driving more accountability and supporting innovation. By offering this CMMI model effective January 1, 2023, CMS can create an opportunity for existing Next Generation ACOs to have a home in MSSP. For existing MSSP ACOs, this track would also provide a new and attractive pathway to take a step up in financial responsibility without evaluating trade-offs on resetting their benchmark and adopting new financial methodologies. Finally, CMS would not be assuming a significant new administrative burden because every aspect of this package has already been operationalized by CMS before. 
To achieve its goal of all Medicare beneficiaries in Part A and B in an accountable care relationship by 2030, CMS will need to continue refining MSSP and, just as important, leverage MSSP with innovations that will attract more participation.