The first five posts in this series were about clocks: the 60-day baseline, the rolling 70–110-day window, the six validity windows, the collection-method rules, the 425-day end-of-period math. This one is about the other side of those clocks — what it costs when one of them bites.
We'll walk it in plain arithmetic, using only the payment amounts CMS has actually published. Where CMS has not published a number, we say so and stop, rather than inventing one. And we'll be explicit up front about the two ways a reporting problem costs you money, because they are different and people conflate them.
Medicare's ACCESS Model pays up to $420 per patient, per year for the CKM track in the Initial Period (the other tracks pay less — $360, $180, $180 — on the same structure).[^footnote] That headline number is where the sizing usually starts. It is not where the risk lives.
The risk lives in one sentence of the payment rules: half of the Medicare portion of every Outcome-Aligned Payment (OAP) is withheld until CMS reconciles your outcomes. In dollar terms, that is $72–168 per patient, per year held back at Initial Period rates, released only after CMS looks at your outcome attainment.[^footnote] That withheld band — not the headline — is the number a finance lead should have taped to their monitor. It is the part of the payment that a reporting failure can cost you.
So when we say "what a missed window costs," we mean: what happens to that withheld $72–168, per patient, when a clock is missed.
The cheapest miss to understand is the one that never enters the system. Every eligible-but-unenrolled patient is $180–$420 per year of allowed revenue you simply don't bill.[^footnote] There's no clever reconciliation math here — it's revenue that never starts. Multiply by however many eligible beneficiaries are sitting in your panel un-enrolled and you have the size of the "we didn't get to it" pile.
This one is a capacity-and-workflow problem, not a data-integrity one, and we're not going to pretend our reporting rail fixes your enrollment funnel. We flag it because it's usually the biggest number on the page and it's the one people forget to count.
This is the subtle one, and it's where the reporting rail actually matters. Here is the rule that does the damage: a beneficiary who disengages or misses a reporting window still counts against your 50% attainment threshold — an unreported measure is treated as non-attainment.[^footnote]
Read that twice, because it inverts the intuition. A missed window is not neutral. It is not "we'll catch it next quarter." In the attainment math, a measure you didn't report lands in the same bucket as a measure you reported and failed: it drags your outcome attainment rate (OAR) down. And your OAR is the dial that decides how much of that withheld $72–168 comes back.
So a slipped 70–110-day window doesn't just delay a payment — it can convert a patient who would have attained into a patient who counts as non-attainment, purely because the data didn't arrive inside the window. The clock and the dollar are the same event.
Correction, 2026-07-26. An earlier version of this post told you that CMS had not published the formula for what you're paid below the 50% threshold, and that anyone quoting you a precise below-threshold figure was quoting a number CMS had never printed. That was wrong on both counts. CMS published the formula in the model's governing document — the ACCESS Request for Applications, v1.1 (updated 02-12-2026), pages 28–31 — and works three examples of it on those pages. We had been watching for its release in a different document that hasn't come out, and never went back to check the one we cite constantly. We're leaving the correction visible rather than quietly swapping the paragraph, because the mistake is more useful to you than a clean page would be: it is exactly the failure mode this post is about, which is a number going unchecked because nobody was assigned to look at it.
Here is what CMS actually says.
Above the threshold: hit at least 50% attainment (the Outcome Attainment Threshold) and you earn 100% of the payment.
Below it, you earn (your OAR ÷ the OAT) × the full OAP amount. CMS's own example: if 40% of your aligned beneficiaries met all their measures, you earn 40 ÷ 50 = 80% of the full amount, and the Clinical Outcome Adjustment — defined as 1 − (OAR ÷ OAT) — is 20%.
And then the part worth knowing before you model any of this: the Clinical Outcome Adjustment is capped at a 50% reduction. CMS's example again — a participant at 20% attainment would arithmetically land at 40% of the payment, but the cap holds them at 50% of the gross OAP amount. There is a floor, and it is half.
Three more mechanics sit on the same pages, all of them in your favour, and all of them worth reading against the source before you model anything:
So the honest version of this section is no longer "nobody can tell you the number." It's: the number is published, the downside has a floor, and the two penalties don't stack. What we still won't do is turn that into a promise — your actual payment depends on alignment, eligibility, attainment and reconciliation, and none of those are ours to warrant.
Last tested against primary: 2026-07-25 (RFA v1.1, the current version per the document's own version table).
The worked math above uses illustrative panel counts and attainment rates — placeholders, labeled as such, because we have no customer data to substitute and we won't borrow someone else's and call it a benchmark. The right version of this calculation uses your numbers: your eligible count, your realistic enrollment, your own view of how many windows you can reliably hit.
If you want to run it against your panel, we keep a simple scenario tool that does exactly this arithmetic on the published rates — enrollment in, at-risk band out — with every assumption visible and the below-threshold curve computed on CMS's published formula rather than on a guess. It's marketing math on CMS's own numbers, not an output of any engine, and it makes no claim about how any product changes your attainment. Ask us and we'll walk your panel through it.
We're a reporting-infrastructure company, so here's the honest boundary of what we do and don't affect on this ledger.
We do not raise your outcome attainment. No tool does — attainment is clinical, and any vendor telling you their software lifts your OAR is making a claim they can't keep. What a reliable rail changes is narrower and real: it keeps the reportable from becoming unreported. The Cost #2 failure above — a patient who attained but whose data missed the window and got scored as non-attainment — is precisely the failure that on-time, reconciliation-ready submission is built to prevent. We can't make a patient attain. We can keep a patient who did attain from being counted as if they didn't, because the data arrived late.
That's the whole pitch, stated at its true size: we don't move the clinical number, we keep the reporting from quietly costing you the clinical number you already earned.
Up to $420 per patient per year, of which $72–168 is withheld pending reconciliation; unreported measures count as non-attainment, so a missed window can cost you that withheld band on a patient who would otherwise have earned it back; below the 50% threshold you earn (OAR ÷ OAT) of the full amount with the Clinical Outcome Adjustment capped at a 50% reduction, reconciled semi-annually, and only the larger of the two downward adjustments applies in any period (RFA v1.1 pp. 28–31); and the only part of this a reporting vendor legitimately affects is making sure the reportable gets reported on time.
---
Outcome Rail builds reporting infrastructure for ACCESS participants. This post is an illustration of publicly published payment mechanics, not financial advice and not a guarantee of revenue. If you want to size the at-risk math against your own panel, we're happy to walk through it: hello@outcomerail.com.
[^footnote]: Based on CMS-published ACCESS Model payment amounts (Effective Period Jul 2026–Dec 2027). Actual payments depend on enrollment, outcome attainment, and CMS payment adjustments. Not a guarantee of revenue.
Sources: CMS ACCESS Model page · ACCESS Model Payment Amounts & Performance Targets (PDF). Also: ACCESS Model Request for Applications v1.1 (PDF), pp. 28–31, for the below-threshold payout formula, the two adjustment caps and the semi-annual reconciliation cadence (last tested against primary 2026-07-25). All dollar figures are CMS-published payment amounts (Table 1 / Payment Frequency, p.3–4); the below-threshold curve is stated as CMS's published formula, and no specific below-threshold dollar figure is quoted for any reader's panel. This post states no measure-target values and no benchmark figures of its own; it is not financial or legal advice.
Device, lab, and PROM data in; compliant FHIR submissions out — validity windows, cadence clocks, and provenance rules enforced before CMS ever sees the bundle. We're onboarding a small founding cohort of design partners this quarter.