CMS' new ACCESS model: What we know and what questions remain

CMS announced the launch of the ACCESS (Advancing Chronic Care with Effective, Scalable Solutions) model, a 10-year voluntary payment model offering outcome-aligned payments (OAP) for tech-enabled management of certain major chronic conditions, in early December. For health system and digital health leaders, it stands out as both a market awareness-raising opportunity and a large-scale test of potential for OAP effectiveness—a logical follow-up to value-based payments.

Fair warning: There’s a lot of information in this post, and it’s a long one. Here’s what we’ll cover, with hyperlinks to each section so that you can jump to the information you need.

1. Overview of CMMI programs: Seeking to improve quality and reduce costs since 2010, with mixed results

2. What we know about ACCESS so far

   1. Early reactions: enthusiasm about the bold payment approach tempered by operational concerns

3. How ACCESS works: paying for outcomes instead of visits

   1. Chronic conditions included in ACCESS affect more than 2/3 of Medicare beneficiaries

   2. Model participants: Medicare providers and digital health companies

   3. Potential model beneficiaries: Medicare recipients with one or more eligible chronic conditions

   4. Getting listed in the tools directory is the lowest-lift opportunity for digital health companies

   5. Applications have opened and are due by April

4. Supplementary TEMPO pilot offers opportunity for digital health vendors

5. What we're watching for and still-unanswered questions

Overview of CMMI programs: Seeking to improve quality and reduce costs since 2010, with mixed results

The ACCESS model is one of the nine new Center for Medicare and Medicaid Innovation (CMMI) models that CMS announced last year. CMMI was established in 2010 as part of the ACA and aims to improve care quality while lowering costs through designing and testing alternative payment models (APMs).

In early 2025, the U.S. Department of Health and Human Services (HHS) announced a new strategic direction for CMMI, one that dovetails with the Make America Healthy Again (MAHA) agenda. The new approach is focused on preventative care, empowering individuals to achieve their health goals (e.g., connecting people with health data and developing financial incentives for outcomes), and promoting competition in the healthcare marketplace. The launch of the new strategy coincided with the announcement that four models (two focused on primary care, one on home dialysis, and one on Maryland’s total costs of care for Medicare patients) would be ended early and two pharmaceutical-focused ones wouldn’t start (the Medicare $2 drug list and Accelerating Clinical Evidence models)—their discontinuation signals the current administration’s scrutiny of cost-effectiveness and desire to reduce program spending. (They estimate that discontinuing the four models saves $750M.) That said, they introduced five new models that more closely align with the MAHA agenda.

CMS’ goals for ACCESS are consistent with the new CMMI agenda (see the references to individual empowerment and transparency and the MAHA focus on chronic disease prevention and management). We’ve also seen these priorities reflected in the Medicare Physician Fee Schedule updates for CY 2026 and with CMS’ commitment to updating the digital health ecosystem.

While it’s not CMS’ first time paying for outcomes, ACCESS is unusual in its breadth; historically, CMS has had outcome-based payments for specific conditions (rather than condition categories) and for specific bundled clinical activities.

Source: Union Healthcare Insight State of Healthcare

What we know about ACCESS so far

ACCESS addresses a perceived gap in the traditional Medicare fee-for-service (FFS) model: FFS care doesn’t readily accommodate chronic condition management (especially through the use of technology), given its focus on billing for encounters and in-clinic activities. The ACCESS model capitalizes on the potential for digital tools to transform chronic conditions management, with an opportunity to assess and pay for effective management—promoting outcomes over specific activities. Much more to come on how the model does so later in the post.

Graphic credit: Chris Loumeau

Early reactions: Enthusiasm about the bold payment approach tempered by operational concerns

Optimism: acknowledging and enabling the increasing role of technology in care

• Forward looking and bold approach: Commenters often indicate that the outcome-aligned payments and the focus on technology are new and indicative of a potential sea change.

• Seeing the complexity of chronic disease: Some organizations have praised the model as an ambitious move that reflects the complex realities of chronic disease management. Other CMMI models have enabled digital health tools and investments in interoperability and EHRs, but this is the first major model explicitly centered on scaling up digital care management.

Pessimism: lacking the structural support to stay afloat

• CMMI programs are hit and miss: CMMI has tried many new payment models, and, overall, they haven’t led to meaningful long-term savings for the agency (1/3 led to substantial government savings, 1/3 substantial losses, and 1/3 minimal financial impact).

• Not an easy market entrance for smaller players: Health tech companies already offering the technology-enabled services of interest to ACCESS participants (like virtual coaching, remote patient monitoring, integrated clinical services) are good candidates for participating. But many health tech companies don’t have experience with Medicare and, as we know, directly enrolling in Medicare is complicated. The workaround would be to have an affiliated physician practice enroll in Medicare and apply to participate. However, the bottom line is that early adopters will likely be large platforms that already work at scale and have already been exposed to risk-based payments in other insurance markets.

• Potential for selection bias: The model is designed to focus on organizational accountability at the patient population level, which reduces but does not eliminate the chance that organizations avoid enrolling sicker patients at risk of not improving into their panel.

• New technologies come with startup costs: Health systems’ success is predicated on technology selection, sure, but they will also need to have access to the right data to manage their populations—an additional hurdle, per health system CIOs. Plus, outcomes-based payments hinge on upfront investment (buying the tools, setting up care coordination and data infrastructure, onboarding staff to the new tools), which could strain smaller systems with already-tight margins and potentially exacerbate gaps in care. Some argue, therefore, for upfront funding to subsidize the setup phase.

• Uncertain patient reactions: How will patients react to technology-enabled care? Some companies that have signed on are prominent in the digital health world already (e.g., Noom, Oura, Teladoc) but may rely on users coming to them of their own accord. What if these technologies are more challenging, less enticing, or both for Medicare beneficiaries who tend to be older and not digital natives? Will the companies continue to achieve the intended outcomes?

How ACCESS works: paying for outcomes instead of visits

ACCESS uses what CMS has termed Outcome-Aligned Payments (OAPs), a concept within the umbrella of value-based care but as a payment model rather than a philosophical stance. With OAPs, participating organizations receive regular payments for managing their patients’ chronic conditions, but they won’t receive the full amount unless they achieve certain health outcomes relative to each patient’s baseline. (The specific target improvements have yet to be announced.) The incentive, then, is to achieve the desired outcomes by whatever means are most effective.

While payment amounts remain undefined, we know from the Request for Applications (RFA) that they will vary by track and be tiered: a higher rate for initial care (first year) and outcome improvements and a lower rate for maintenance and ongoing management (optional follow-on period). The RFA clarifies that a portion of the ACCESS payments is based on individual results, but the payments can also be affected by aggregate results across the pool of participating beneficiaries. There are two possible adjustments (CMS will choose the larger of the two) based on panel performance:

1. Clinical Outcome Adjustment: CMS will look at the percentage of patients who meet the target improvements (the outcome attainment rate (OAR)). ACCESS organizations with an OAR at the target threshold or higher (it will increase yearly) will earn the full payment; the rest will earn a proportional payment.

2. Substitute Spend Adjustment: For every track, there’s a list of “substitute” services that treat the same conditions. CMS is interested in ACCESS participants’ Substitute Spend Rate—the percentage of beneficiaries participating in ACCESS who did NOT receive the equivalent services from other providers while participating—relative to a benchmark threshold (meeting or exceeding the threshold earns the full payment).

Patients can join different tracks with the same or different organizations; but, if they’re enrolled in multiple tracks in the same organization, CMS will discount the total payment amount due the administrative efficiencies introduced by coordinated care.

Graphic credit: Chris Loumeau

Chronic conditions included in ACCESS affect more than 2/3 of Medicare beneficiaries

The eligible conditions are grouped into four tracks. Note: Three of the tracks (musculoskeletal pain excepted given the goal of resolving that pain) have an initial year of care and then an optional continuation period at a lower rate.

Some more notes about the tracks:

• CMS has reserved the right to add additional tracks or conditions later, though there hasn’t been an obvious signal of if/when that might be.

• CMS named the specific measures and means by which they must be obtained, though not the improvement targets, in the RFA.

• Participants will have a limited amount of time to submit baseline data for each patient after enrolling them, and they will have to submit updates at predefined intervals (that vary by measure). They also need to submit measures at the end of the 12-month care period.

Model participants: Medicare providers and digital health companies

Participating organizations:

• Must offer “technology-enabled care” (e.g., coaching, behavioral support, patient education, medication management, remote monitoring, wearables).

  • CMS describes technology-enabled care organizations as ones that “deeply integrate technology into care delivery to deliver continuous, high-value care to prevent and manage chronic disease.” They further distinguish between two approaches to technology-enabled care: clinician-led, technology-augmented approaches and technology-led, clinician-supervised approaches (see table). The expectation is that technology-enabled care organizations employ both, though the latter reads as a starker departure from traditional clinical care.

Clinician-Led, Technology-Augmented	Technology-Led, Clinician-Supervised
Virtual visits Remote monitoring AI-assisted delivery	Digital therapeutics

• Can choose to waive cost sharing for their patients who participate (they don’t have to do this, though—it’s a recruitment incentive).

• Must be enrolled in Medicare Part B as providers or suppliers and must designate a doctor as a medical director to oversee their participation.

• Will bill monthly for ACCESS patients through regular claims but with track-specific G-codes (distinct from CPT codes). This keeps the option open for Medicare Advantage, Medicaid, and private plans to adopt similar systems.

• Will need to electronically share care plans and updates at pivotal moments, though it sounds like those moments are not fully laid out yet. (Plus, they need to integrate with a health information exchange (HIE) so that the clinicians who referred the patients can get updates.)

In addition to participating in the model directly, organizations might be indirectly involved by referring their patients to specialty practices participating in ACCESS. Those referring organizations (and any primary care providers, if different) will need to have access to any clinical information resulting from the patient’s participation in ACCESS via regular updates. They also qualify for an ACCESS co-management payment as compensation for the effort involved in reviewing those updates. Part B-enrolled clinicians get access to a new code (ACCESS co management--$30 per review and onboarding add-on of ~$10—amounting to about $100 per beneficiary per year) for reviewing access updates and doing care coordination.

There is a public reporting element to the program for accountability and transparency: CMS is going to maintain a directory of participants, tracks, tool use, and risk-adjusted clinical outcomes.

Potential model beneficiaries: Medicare patients with one or more eligible chronic conditions

Traditional Medicare members are eligible to participate in ACCESS. They can enroll directly or be referred by a care provider. Note: there is a possibility of individuals being assigned to a control group when they try to enroll in ACCESS so that CMS can assess the model’s impact. This does not prevent their receiving usual Medicare services.

Getting listed in the tools directory is the lowest-lift opportunity for digital health companies

CMS will publish a directory of tools for ACCESS participants to peruse when looking for technologies to support their efforts. Vendors interested in being included in the directory need to send in their own listings and attest that their product is compliant with federal and state requirements (including FDA requirements unless they are under the enforcement discretion offered under TEMPO—more on that later). There’s no indication yet that vendors will need to pay a fee to be listed in the directory, nor will they receive any compensation from CMS for their listing. Companies are being encouraged to consider promotional offers that can draw in providers, as those offers can be included in the directory.

Applications have opened and are due by April

Applications to participate in the first performance period (starting July 1) must be submitted by April 1. Any applications submitted after April 1 will be considered a performance period beginning January 1, 2027. More than 500 organizations have already signaled their intent to apply.

Supplementary TEMPO pilot offers opportunity for digital health vendors

Around the same time the ACCESS model was announced, the FDA unveiled a supplementary pilot: TEMPO (Technology-Enabled Meaningful Patient Outcomes) for digital health devices. Given the ACCESS model’s focus on chronic conditions management through technology, TEMPO seeks to facilitate the go-to-market process for technology aimed at managing the chronic conditions of interest. It makes an allowance for about 40 devices (10 per condition) that have not yet obtained full FDA premarket authorization to be used in the ACCESS program. Why make this allowance? Well, not only does it make promising tools available to providers quickly, but it also allows for the generation of real-world data—on how these models support individuals with chronic diseases. One thing worth noting for digital health companies with inpatient-focused tools: Only outpatient devices (used in partnership with a clinician) are eligible to participate in TEMPO. Presumably that means that only outpatient capabilities are relevant to ACCESS.

What we're watching for and still unanswered questions

We’ve shared a lot of information, and you probably still have a lot of questions. We do, too. Some crucial pieces of information are yet to be released, and there are, of course, unknowns regarding model implementation.

For digital health companies

• Are companies willing to go to the trouble of enrolling in Medicare to participate in ACCESS directly, given that it might be onerous (including potentially involving changing financial practices and not balance billing for Medicare patients)?

• Do health tech companies tend to already have access to sufficient data and infrastructure, or will there be significant retrofitting necessary? Participants who want to be successful with the model need reliable access to multiple platforms and resources (labs, imaging, medication information, medical device information, means of collecting patient reported outcome measures).

• Given the tech and security needs, will this model, rather than facilitating exposure for new technology companies treating chronic disease, privilege large organizations that can quickly build up (or have already built) the tech and data infrastructure?

For providers

• How ambitious are the outcome targets?

• How feasible will it be for resource-strapped organizations to get the needed measures? Collecting the patient data will require powerful reporting (including patient survey tools for patient-reported measures) and analytics skills along with high data security.

For the industry overall

• Will other insurers follow suit and adopt similar payment models? Medicare Advantage (MA) plans are the most logical candidates. In ACCESS materials, CMS mentions multiple times that MA plans are welcome to adopt similar approaches, but we don’t yet know whether other plans are enthusiastic about the possibility or not.

• Who will end up participating directly? Will this model ultimately privilege large companies and organizations with minimally complex patient panels? If so, will it continue to challenge chronic disease management status quo? We can predict that it will be an easier sell for large health systems that can shoulder the upfront time and resource costs of shopping for and implementing a new technology.

• What will the rural/urban participation distribution look like? On the one hand, health IT is an especially powerful tool for rural care providers who have patients traveling long distances for an appointment. That said, the digital divide and unreliable internet in some rural areas is a real concern, too.

• How complex will ACCESS participant panels ultimately be? What will happen if organizations have a disproportionate share of high-need patients or see high rates of noncompliance that are independent of the tech solutions? CMS’ risk adjustment methods (which we don’t know yet) will likely have an impact here.

We’ll be keeping an eye on the model as more details—especially measures and payment amounts—come out. Plus, we will talk more about ACCESS and all the other relevant policy changes heading into 2026 at the Jan. 29 Policy Board Briefing. Register now!