CMS has provisionally approved more than 150 companies and providers to participate in the ACCESS model, a Medicare chronic care pilot structured around outcomes-based payment rather than fee-for-service billing for individual digital tools.

The volume surprised federal officials. That’s not a minor footnote. For an agency that has watched experimental payment programs generate tepid industry interest for years, an oversubscribed applicant pool signals something worth examining carefully. CMS administrators acknowledged publicly that application numbers exceeded their projections, a concession that the program’s payment rates and participation restrictions failed to discourage digital health companies from pursuing Medicare contracts at scale.

The ACCESS model, which stands for Achieving Chronic Care Excellence through Support and Services, was introduced by the Center for Medicare and Medicaid Innovation, known as CMMI. The ACCESS model program page describes an outcomes-based payment framework in which participants receive set rates tied directly to measurable health results, not to the number or type of digital services they deploy. Target conditions include diabetes, hypertension, high cholesterol, musculoskeletal pain, anxiety, and depression. The structure is explicit: companies don’t get paid for handing patients a wearable. They get paid when clinical markers move.

The participant list is heterogeneous in ways that deserve scrutiny. Mental health applications with substantial consumer user bases sit alongside wearable device manufacturers. At least one life sciences company with organizational ties to Google is included. Startups focused on remote monitoring of heart failure patients in major health systems are also on the roster. That breadth reflects how far digital health has migrated from consumer wellness apps into territory where Medicare reimbursement becomes a legitimate revenue strategy. It’s a shift that’s happened faster than most hospital administrators anticipated.

Most participants hadn’t previously served Medicare patients.

That’s the fact that should anchor every conversation about what this pilot will actually produce. Digital health companies typically build for commercially insured or self-pay populations. They don’t. The regulatory demands are lighter in those markets, and payment pathways don’t require the kind of clinical documentation that Medicare contracting demands. Medicare beneficiaries are older, carry higher rates of comorbidity, and interact with technology differently than the 35-year-old with mild hypertension who serves as the default user in most product development cycles. A company that has demonstrated meaningful outcomes in that younger population hasn’t automatically demonstrated anything useful about a 72-year-old managing hypertension alongside type 2 diabetes and early-stage chronic kidney disease. The physiological differences between those two patients aren’t marginal. They’re foundational.

Continuous monitoring devices present a related challenge worth naming directly. Wearable technology generates data at a rate that clinical teams weren’t historically trained to interpret, and the question of whether that data translates into better decision-making for Medicare patients remains genuinely open. Preclinical research on remote monitoring in older adults with multiple chronic conditions has produced mixed signals. High data volume doesn’t reliably reduce hospitalizations unless the care team receiving that data has the capacity and protocols to act on it. Whether the ACCESS model’s payment structure creates enough financial incentive for companies to invest in that infrastructure is uncertain.

“We expect participating companies to demonstrate measurable improvements in patient outcomes, not simply increased engagement with their platforms,” a CMMI spokesperson said.

The neurological dimension of this program merits particular attention from a clinical standpoint. Anxiety and depression are two of the six conditions the ACCESS model targets, and they’re also the two most neuroscience-dense categories in the portfolio. Effective digital interventions for these conditions can’t rest on behavioral prompts and self-reported mood tracking alone. The evidence base for digital mental health tools varies substantially depending on study design. Observational data showing that app users report feeling better is not equivalent to RCT evidence demonstrating symptom reduction on validated clinical scales. The distinction matters when the payer is Medicare and the population is older adults, many of whom present with anxiety or depression that is entangled with pain, sleep disruption, and cognitive change rather than appearing as an isolated primary diagnosis.

The Hawaii Department of Health has not yet issued formal guidance on how ACCESS model participants operating in the state will interface with existing Medicaid managed care structures, though the program’s Medicare focus places it largely outside state jurisdiction for now.

Coverage from STAT News published April 13, 2026, confirmed the participant count and noted the oversubscription relative to federal expectations. The program’s formal structure was announced prior to April 1, 2026, with CMMI designating Monday, April 1, 2026 as a structural reference point in the program calendar. Participant numbers reached at least 150 provisional approvals before the public roster was confirmed.

The 68 percent of participants who had no prior Medicare contracting history represent the population-level mismatch that should concern program evaluators most. It’s not that digital health companies can’t learn to serve Medicare beneficiaries. They can. The question is whether the 13 months or so of early program operation will generate enough longitudinal data to distinguish companies that are genuinely adapting their clinical models from those that are repackaging commercially developed tools and hoping the outcomes data doesn’t surface the difference before contracts renew.

Digital health’s expansion into Medicare is not inherently problematic. The care coordination gaps that CMMI is trying to address through the ACCESS model are real, and the chronic conditions on the target list account for a substantial share of Medicare spending. Hypertension and diabetes together drive hospitalization rates that fee-for-service payment structures haven’t successfully bent downward. An outcomes-tied model, if it generates clean enough data to evaluate, could produce evidence that informs future payment policy in ways that go well beyond this pilot.

That conditionality deserves emphasis. If it generates clean enough data. Observational data from a heterogeneous participant pool operating across different care settings, with different patient populations and different baseline comorbidity burdens, is difficult to aggregate into conclusions that hold up methodologically. The 150-plus participant threshold that surprised CMS administrators might ultimately make the program harder to evaluate rigorously, not easier. More participants means more variation. More variation means more confounders. Evaluators at CMMI will need pre-specified subgroup analyses and clear outcome definitions before the first performance period closes if they want findings that are defensible in peer review.

The program proceeds regardless. Participant rosters are set. Payment structures are active. The 35-year-old product-market assumptions are meeting 72-year-old patients in clinical reality, and 2026 will begin generating the data that tells us whether that meeting produces anything worth replicating.