The HealthTech Policy Opening

For most of Medicare’s history, digital health companies existed in the policy margins. They sold to health systems, contracted through Medicare Advantage plans, or found revenue in Medicaid managed care. Original Medicare largely closed its door. There was no enrollment pathway, no fee schedule that paid for technology-enabled care at sustainable rates, and no model that let a digital-first organization stand up as a direct Medicare participant. That changed with the 2025 CMMI model announcements.

Three models now define the policy opening: ACCESS, which creates direct enrollment and payment pathways for technology-enabled chronic care organizations; WISeR, which has generated a market for AI-powered prior authorization vendors inside Original Medicare; and Geo AHEAD, which allows non-provider entities to take geographic population risk. Together they represent the most significant structural expansion of Medicare’s participation perimeter since the ACO programs began. The question for HealthTech companies is not whether a policy opening exists. It is how large the opening actually is, and what compliance and capital infrastructure it demands.

ACCESS as the Digital Health Beachhead

#

The Advancing Chronic Care with Effective, Scalable Solutions model launched its application period in late 2025 and begins July 1, 2026. It runs for ten years, which matters for any technology company that has tried and failed to build sustainable Medicare revenue on a two-year pilot timeline. The model covers four clinical tracks: an early cardio-kidney-metabolic track addressing hypertension, dyslipidemia, obesity, and prediabetes; a cardio-kidney-metabolic track for diabetes, CKD, and atherosclerotic cardiovascular disease; a musculoskeletal track for chronic pain; and a behavioral health track for depression and anxiety. These conditions affect more than two-thirds of the Original Medicare population, which means the addressable patient base is large by any measure.

What makes ACCESS structurally different from earlier CMMI models is its enrollment design. Beneficiaries can sign up directly with an ACCESS care organization without a physician referral, removing the attribution dependency that constrained prior digital health pilots. Referring physicians can also send patients and receive electronic care plan updates and a co-management fee. CMS will publish a public directory of ACCESS participants with risk-adjusted outcomes data, creating a transparency mechanism that functions both as accountability and as marketing infrastructure for participating organizations.

The eligibility requirement is the critical constraint for digital health companies. ACCESS participants must enroll in Medicare as Part B providers or suppliers. Notably excluded are durable medical equipment, prosthetics, orthotics, and supplies (DMEPOS) suppliers and laboratory suppliers. Organizations must designate a Medicare-enrolled clinical director with responsibility for care quality and compliance. For a digital health company that has operated entirely outside Medicare’s direct enrollment infrastructure, this is not a trivial undertaking. Medicare Part B enrollment requires licensure compliance in every state of operation, adherence to HIPAA as a covered entity, and submission through CMS’s FHIR-based APIs for outcomes reporting and care coordination. CMS acknowledged the lead time explicitly in its RFA guidance and encouraged organizations that are not yet enrolled to begin the process immediately.

The payment structure replaces fee-for-service billing with outcome-aligned payments (OAPs). Organizations receive prospective monthly payments calibrated to whether a patient is in an initial high-intensity care period or a follow-on maintenance period. Full reconciliation payments depend on the proportion of enrolled patients meeting condition-specific clinical targets: blood pressure reduction in hypertension, HbA1c and eGFR control in the CKM track, validated patient-reported outcome improvement in the MSK track, and PHQ-9 score improvement in behavioral health. CMS adjusts payments for rural populations, creating a fixed uplift for qualifying geographic areas.

Unlike ACO REACH or MSSP, ACCESS organizations do not assume downside financial risk. The model is designed as outcomes-based rather than total cost of care, which means participating organizations are paid for clinical results without taking on actuarial exposure for overall spending. For digital health companies with venture or growth equity backing that could not underwrite a risk corridor, this is a meaningful structural accommodation. It also creates an accessible pathway for organizations that historically participated in Medicare only as subcontractors to ACOs or MA plans.

A parallel FDA initiative, the Technology-Enabled Meaningful Patient Outcomes pilot (TEMPO), addresses the regulatory gap for devices that are used in the context of ACCESS care but have not yet obtained FDA premarket authorization. Manufacturers can request enforcement discretion for qualified devices. FDA expects to select up to ten manufacturers in each of four clinical use areas. TEMPO is not a blanket clearance mechanism, but for a digital health company whose device is designed specifically for one of the ACCESS clinical tracks, the pilot creates a pathway to market participation that did not exist before.

What ACCESS does not do is equally important. The model operates exclusively in Original Medicare and explicitly excludes Medicare Advantage. For HealthTech companies whose growth strategy has been built around MA plan partnerships, ACCESS creates a parallel FFS revenue channel rather than a replacement. The model also does not restructure the Medicare fee schedule. Organizations that bill under traditional CPT codes for RPM, chronic care management, or behavioral health outside of ACCESS remain subject to the same payment rates as before. The OAP structure is self-contained. This matters because an organization that joins ACCESS will need to build a separate billing and compliance infrastructure from its FFS operations if it continues to operate both.

WISeR and the AI Vendor Market

#

WISeR operates on a different axis entirely. The Wasteful and Inappropriate Service Reduction model launched January 1, 2026, in six states across four Medicare Administrative Contractor jurisdictions: New Jersey, Ohio, Oklahoma and Texas, Arizona and Washington. Rather than inviting technology companies to deliver care, WISeR contracts them to make coverage determinations. Vendors earn shared savings from averted wasteful care and are assessed against performance measures including provider experience scores.

CMS selected six vendors for the initial cohort: Cohere Health for Texas, Innovaccer for Ohio, Genzeon for New Jersey, Humata Health for Oklahoma, Virtix for Washington, and Zyter for Arizona. These companies deploy AI and machine learning tools combined with mandatory clinician review to conduct prior authorization and pre-payment review for a defined list of services that CMS has identified as prone to overuse and inappropriate billing. The model requires FedRAMP-certified workflows, FISMA compliance, and CMS’s Information Systems Security and Privacy Policy adherence. This is a federal contracting environment, not a commercial health plan environment, and the compliance requirements reflect that.

The AI infrastructure WISeR demands is substantive. Vendors must connect to providers’ EHR systems, retrieve and classify clinical documentation, match it against National Coverage Determinations and Local Coverage Determinations, and render a coverage determination within three days for standard requests and two days for urgent ones. The RFA estimated that AI has the potential to automate between 50 and 75 percent of the manual work involved in prior authorization processing. In practice, the model requires a human-in-the-loop architecture: all clinical denials must be reviewed by appropriately licensed clinicians, and vendors must maintain backup pathways including phone, fax, and electronic portals.

The gold card mechanism, which CMMI announced plans to pilot by mid-2026, creates a performance incentive layered on top of the basic model structure. Providers whose prior authorization requests are affirmed at a rate of 90% or higher over a defined review period would be exempted from future WISeR reviews. This is functionally a compliance reward and a volume reduction mechanism for high-performing providers. For vendors, it means the ongoing review burden will concentrate on providers with lower affirmation rates, raising the actuarial and clinical sophistication required to manage the workload.

The WISeR vendor selection is significant beyond the immediate model. CMS explicitly stated interest in using the same vendors and AI tools that MA organizations use for prior authorization. This creates potential for cross-market standardization. A vendor that builds the clinical data infrastructure for WISeR in Original Medicare may be positioned to extend the same platform to MA plan contracting, creating economies of scale that smaller vendors operating in a single market cannot match. The AHA and other provider organizations have raised concerns about AI governance and the risk that vendor physicians will rubber-stamp AI-generated recommendations rather than exercise independent clinical judgment. CMS has committed to tracking physician involvement and denial overturn rates, which means the compliance infrastructure for WISeR vendors will include detailed clinical audit trails.

Geo AHEAD and Non-Provider Geographic Risk

#

Geo AHEAD extends the AHEAD model’s global budget structure to non-state-entity participants. Under standard AHEAD, states take total cost of care risk for a defined geographic population using hospital global budgets as the accountability mechanism. Geo AHEAD opens geographic risk-taking to non-state entities, which CMS has indicated can include entities that are not traditional providers.

For HealthTech companies, this represents the most ambitious and most capital-intensive opportunity in the current CMMI portfolio. Taking geographic population risk requires population health data infrastructure capable of identifying high-risk beneficiaries, care coordination workflows that span providers across a geography, actuarial capacity to model cost trajectories, and sufficient capital reserves to withstand adverse experience. The organizations that are positioned for this are not early-stage digital health startups. They are scaled population health platforms with deep analytics capabilities, existing payer or ACO experience, and financial backing sufficient to sustain a multi-year model with upfront investment before savings materialize.

The meaningful policy question is whether Geo AHEAD’s non-provider eligibility actually translates into HealthTech participation, or whether the complexity of geographic accountability funnels participation toward existing health system or payer entities that add technology infrastructure rather than technology companies that build care delivery infrastructure. The model design does not answer this. The participation data from the first application cycle will.

MAHA ELEVATE and Lifestyle Technology

#

The Making Americans Healthy Again ELEVATE model covers lifestyle medicine domains including nutrition, sleep, stress, physical activity, and social connection. Digital platforms that deliver structured lifestyle interventions in these domains could participate as program providers, though CMS has not finalized the participation criteria or enrollment pathway with the same specificity as ACCESS.

The evidence bar for MAHA ELEVATE participation creates a meaningful filter. CMS requires evidence-based interventions, and the definition of evidence-based in a federal payment model context means randomized controlled trial evidence or established clinical guideline grounding, not the product-efficacy claims that populate investor decks. For companies that have invested in clinical research demonstrating outcomes at a population level, this bar is achievable. For those whose evidence base consists primarily of observational data or proprietary metrics, the path to model participation involves a clinical validation investment that precedes any revenue.

The measurement challenge is structural. Lifestyle outcomes across nutrition, sleep, stress, and social connection do not resolve in a 90-day episode. They accumulate over years. MAHA ELEVATE’s model period is long enough to capture meaningful outcome signals, but the interim measurement and reporting infrastructure requires technology platforms capable of continuous patient-reported data collection, validated instrument administration, and longitudinal outcome tracking. The companies best positioned to participate are those that have already built this infrastructure for other payers and can extend it into the model without starting from scratch.

The Reimbursement Gap

#

Each of these models creates a participation pathway. None of them resolves the underlying Medicare fee schedule problem for digital health services outside the model structures. The CPT code landscape for remote patient monitoring, remote therapeutic monitoring, and chronic care management has expanded incrementally since 2017, but the payment rates for these codes have not kept pace with the technology infrastructure required to deliver the services compliantly. Organizations billing RPM under CPT 99454 receive approximately $53 per month per patient. Behavioral Health Integration billing under CPT 99484 pays roughly $49 per month. Chronic care management under CPT 99490 runs approximately $62 per month for qualifying patients with two or more chronic conditions.

These rates were set at a time when CMS was trying to create financial incentives for care management activity inside primary care practices, not to sustain technology-enabled care companies with compliance, clinical, and engineering overhead. The gap between what the fee schedule pays and what a technology-enabled care organization actually costs to operate is significant, and it is the primary reason most digital health companies in Medicare have historically operated as subcontractors rather than direct participants.

ACCESS’s outcome-aligned payment structure represents CMS’s attempt to address this gap inside the model perimeter. Whether the OAP rates are set at levels that generate sustainable margin for digital health organizations will become clear only after the first reconciliation cycle. CMS has calibrated the initial period payment higher to reflect the cost intensity of onboarding and achieving initial clinical improvement, with a lower follow-on rate for maintenance. Organizations doing the business model math before applying should model conservatively on OAP levels and validate participation economics against the actual RFA payment amounts before committing compliance infrastructure.

The deeper question is whether these models represent a transition toward a fee schedule that genuinely accommodates technology-enabled care, or whether they are parallel tracks that leave the underlying payment structure unchanged. CMMI models are test environments, not permanent Medicare policy. ACCESS, WISeR, and Geo AHEAD will generate outcomes data over their model periods. If that data supports expansion, CMS has the authority under Section 1115A to scale successful models nationwide without additional legislation. If the models underperform or generate adverse beneficiary experience, the policy door can close as quickly as it opened.

Related Reading

#

MCR-01_04 ACCESS: Digital Health’s New Medicare Beachhead MCR-01_03 WISeR: Prior Authorization Comes to Traditional Medicare MCR-03_06 Telehealth at the Crossroads: Permanence, Benefit Design, and the Rural Access Divide