The CMS ACCESS model just added four tracks, and your claims need a scorecard

What the CMS ACCESS model actually changed on 15 September#

The fox does not pounce on a headline; it walks the hedgerow first and counts the gaps. Here is the count. On 15 September 2026 CMS said that from spring 2027 the CMS ACCESS model will add Medicare tracks for heart failure, chronic obstructive pulmonary disease (COPD), substance use disorders and tobacco cessation, according to its own CMS press release, and the CMS model page pins the start to 1 April 2027. For a digital therapeutics (DTx) or remote patient monitoring (RPM) company, that is a payment route with a calendar attached.

Rewind and the shape makes sense. CMS announced ACCESS on 1 December 2025, as the American Hospital Association reported, and the CMS request for applications set out a ten-year voluntary model that pays for outcomes rather than activity. Applications opened on 12 January 2026, the first deadline fell on 1 April 2026, and the first cohort began on 5 July 2026. The four launch tracks cover early cardio-kidney-metabolic (eCKM) conditions, cardio-kidney-metabolic (CKM) conditions, musculoskeletal (MSK) pain and behavioural health (BH).

Four figures carry the announcement, and the press release supplies every one of them.

Behind those numbers sits one design idea: CMS ties payments to measurable improvements in patient health instead of paying solely for individual services. Everything that follows, from what to claim to how to price a pilot, flows from it.

Read the fine print on the last figure before it lands in a slide. The 165 million is not a commercial-only count: the CMS payer pledge page describes members across Medicare Advantage, Medicaid and private health insurance, from 17 named payers who commit to offering aligned payment arrangements by 1 January 2028. A pledge is a promise to build a payment route, not a purchase order. ACCESS itself serves Original Medicare only, and the press release says Medicare Advantage enrollees are not eligible.

Searchers say Medicare ACCESS model, CMS says ACCESS#

People type the Medicare ACCESS model into a search box; the agency's own name is ACCESS, short for Advancing Chronic Care with Effective, Scalable Solutions. Use CMS's name in anything a compliance reviewer will read, and the searchers' phrase only where it helps a buyer find you.

Who is in the room: cms access model participants and what they buy#

Eligibility sets the buyer. Under the CMS request for applications, a participant is a Medicare Part B enrolled organisation identified by a single tax identification number, excluding durable medical equipment and laboratory suppliers, with a physician Medical Director and its own FDA, HIPAA and licensure duties. That sentence rewrites a go-to-market plan. A DTx or RPM vendor is not automatically a payee: it either becomes a participant or sells into one, and those are different marketing jobs.

CMS's ACCESS participants page lists more than 160 organisations and notes that most of them have not previously served Medicare beneficiaries. We counted the table on 19 September 2026 and found 162 rows. Newcomers to Medicare are a quarry worth stalking: they arrive with clinical ambition and without a Medicare marketing playbook.

The same table shows a varied field: 45 of the 162 organisations run all four launch tracks and 34 run only one, so a single-condition vendor will meet generalists wanting one stack and specialists wanting depth.

The public directory is the quiet giant in that list. The request for applications says CMS will show each organisation's tracks, the conditions it treats and its risk-adjusted outcomes, drawn from the clinical data collected for payment. Put plainly, participants will be displayed in public against the same metrics that pay them, so any creative that outruns those numbers can be checked by anyone who opens the page.

How the money moves: cms access model payment rates and the outcome cliff#

Follow the money the way a fox follows a scent, one step at a time. The CMS payment amounts and performance targets document, effective from 5 July 2026 to 31 December 2027, sets annual allowed amounts per beneficiary: $360 initial and $180 follow-on for eCKM, $420 and $210 for CKM, $180 for MSK with no follow-on period, and $180 and $90 for behavioural health, plus a $15 rural payment on the cardiometabolic tracks. Those figures include 20 per cent beneficiary coinsurance.

That document covers the four launch tracks. In the pages we read, CMS has published no rates yet for heart failure, COPD, substance use disorder or tobacco cessation, so any number quoted for them today is unverified.

Now the cliff. CMS pays in monthly instalments but caps them at 50 per cent of the Medicare portion; the other half is withheld until the 12-month care period ends and outcomes are reconciled, per the same document. Two tests decide what comes back. The Outcome Attainment Rate is the share of aligned beneficiaries who finish the period and hit every required target, measured against a 50 per cent threshold. The Substitute Spend Rate is the share who did not receive listed substitute services from other Medicare providers, measured against 90 per cent.

Miss either and payment scales down in proportion. The request for applications caps the outcome cut at 50 per cent and the substitute-spend cut at 25 per cent, and applies only the larger of the two at each semi-annual reconciliation.

This is where digital therapeutics reimbursement stops being a coding question and becomes a positioning question. A participant is paid a fixed amount per beneficiary per track, not a fee per device or module, so a vendor selling software is selling a cost line inside a budget of $180 to $420 a year.

Behavioural health has a second trap. Its substitute-spend list includes digital mental health treatment codes G0552 and G0553 and a remote therapeutic monitoring set-up code, 98975, when another Medicare provider bills them for the same diagnosis. A vendor billing Medicare separately on the same beneficiary can drag a participant's substitute spend rate down.

What outcome evidence lets a digital health brand say#

Outcome-based payment turns a marketing claim into a forecast the buyer will be graded on, so begin with what CMS actually scores. For eCKM, blood pressure is met at a final systolic below 130 mm Hg or a 15 mm Hg reduction. For behavioural health, a baseline PHQ-9 of 10 or more needs a 5-point reduction, per the CMS payment document.

Two rules make those measures unforgiving: blood pressure and weight readings can be no older than 15 days at submission, and outside weight and patient-reported measures, self-reported values are not permitted. A claim like better heart health matches no scored measure; a claim about systolic pressure under 130 mm Hg matches one.

A scored measure still needs a study behind it, so the table sets five published trials beside the plainest reading of what each lets a marketer say.

Same label, opposite answers, and the pair is the lesson. BEAT-HF tested telemonitoring with nurse coaching after discharge and found no readmission benefit, while TIM-HF2 tested structured remote management in patients with a recent heart failure admission and reported fewer days lost. Remote monitoring reduces readmissions is therefore not a claim about a category. It is a claim about one programme, one population and one endpoint, and it must name all three.

The other three trials carry small print of their own. The Spanish telemonitoring trial covers heart and lung failure together, the iCanQuit trial compared two apps and relied on self-reported quitting, and the TES trial tested an add-on to counselling in specialty programmes, where its authors say other settings need more research.

Independent commentators are watching for shortcuts. Writing in Health Affairs Scholar, Aditya Narayan and Bob Kocher warn that organisations may enrol people with higher digital readiness, and question whether measures can detect upcoding. A brand boasting a sky-high success rate invites that suspicion, so show the denominator beside the percentage.

Claims guardrails: FDA, FTC and the outreach rules#

Evidence tells you what is true; regulators tell you what you may say. Start with FDA, whose TEMPO pilot lets a manufacturer request enforcement discretion on certain requirements, including premarket authorisation, when a device is offered to or by ACCESS participants for an intended use that improves outcomes. FDA also says it expects participating manufacturers to seek marketing authorisation eventually, and its selected participants so far include Dexcom, Cadence, Limbic and SonderMind.

Discretion is a permit for a pilot, not a claim you can print. The pilot's four clinical areas mirror the launch tracks, and we saw no extension to the new ones. Our earlier piece on TEMPO and the claims line walks through where that line sits.

FDA's general wellness guidance, current as of 6 January 2026, adds a second fence: software for maintaining a healthy lifestyle and unrelated to a disease is not a device, while a claim about treating or preventing a condition pulls a product toward device regulation. Pick words as carefully as a fox picks its footing on frozen ground.

Then the FTC. Its health products compliance guidance says randomised, controlled human trials are the most reliable evidence and generally what experts would require for health benefit claims, and that a testimonial needs evidence the product works for buyers as it did for the endorser. Our telehealth rulings piece shows what happens when a footnote cannot carry the headline.

Outreach has its own rules. The request for applications lets participants contact potentially eligible beneficiaries, consistent with standard CMS marketing rules, and the payment document extends that to referring clinicians and adds the Anti-Kickback Statute and the Physician Self-Referral Law. Here is what the first wave of company claims looks like in the wild.

That post appeared on the day of the CMS announcement. Read the dollar figure as a compliance officer would. The request for applications lets a participant adopt a uniform policy to waive coinsurance, and a participant that collects it must disclose the expected patient payment before enrolment. So $0 is a fact about one organisation's policy, not about ACCESS itself. Shared creative therefore needs cost-sharing, track lists and eligibility as per-participant variables, filled in by the organisation that holds the policy.

A marketing plan for the CMS ACCESS model window#

The new tracks open in spring 2027, which leaves a moonlit season to prepare. Here is the plan a growth head can run now, in the order that keeps compliance calm.

Notice who is missing from that plan: the patient. A vendor's compliance-checked evidence pack is the product, and the participant speaks to beneficiaries under CMS marketing rules; a participant's own compliance team owns every beneficiary-facing word. Either way, take legal advice before anything reaches a person with Medicare.

The second version is longer and duller, and it is the one a participant's compliance team can sign. It also leaves room for your own outcome data, and in an outcome-paid model your customers will be measuring you, so measure yourself first.

Give the plan a home in your content engine. An evidence library built with content marketing gives participants proof they can reuse, and SEO and GEO keeps it quotable when buyers ask an answer engine who is in the CMS ACCESS model. Our healthcare marketing work starts from the same rule, paid social on professional networks reaches clinicians who refer, and our FTC pixel piece covers the tracking side. Follow the folkfox news stream as CMS publishes new-track detail.

Frequently asked questions#