The first two posts in this series were about when you submit — the 60-day baseline and 425-day final clocks, and the rolling 70–110-day quarterly window. This one is about something upstream of all of that: whether the reading you're submitting is still eligible at all.
Because a submission on time with a stale reading is not a valid submission. The ACCESS "Payment Amounts and Performance Targets" paper sets, on p.6, a collection-to-submission freshness window for each OAP measure: the reading has to have been collected within its window before you submit it. Six of these windows cover the common measures, and they are not all the same length.
From the Payment paper's OAP Measure Validity section (p.6):
| Measure | Must be collected within |
|---|---|
| Blood pressure | 15 days before submission |
| Weight / BMI | 15 days before submission |
| PROMs (MSK and BH) | 15 days before submission |
| HbA1c | 1 year (prediabetes / diabetes) · 2 years (all others) |
| LDL-C | 1 year (dyslipidemia) · 2 years (all others) |
| eGFR / uACR | 1 year |
Two things about this table are easy to miss on a first read.
The windows span two orders of magnitude. The vital-sign and PROM measures expire in 15 days; the lab measures live for a year or two. A workflow that treats "recent labs" and "recent vitals" as one freshness bucket will either chase lab draws far more often than required, or — the expensive direction — submit a blood pressure that aged out eleven days ago.
Two of the windows depend on the beneficiary's condition, not just the measure. HbA1c is valid for one year for someone with prediabetes or diabetes and two years otherwise; LDL-C is one year with dyslipidemia and two years otherwise, per the same page. The freshness of a lab value is therefore not a property of the lab alone — it's a function of the measure and the beneficiary's clinical context. A validity check that keys only off the test name will get these two wrong at the boundaries.
Passing the freshness window is necessary but not sufficient. The ACCESS RFA, Appendix C, governs how a measure may be collected: blood pressure has to come from a validated upper-arm cuff averaging at least three readings — manual entry is not permitted; lab values have to come from accredited/CLIA sources; self-reported values are disallowed except for weight and PROMs.
So a reading can fail eligibility two independent ways: it aged out of its window, or it was collected by a method the model doesn't accept. A date-only check catches the first and misses the second entirely. (The collection-method rules are their own topic — the next post in this series takes them apart.)
This is pure scheduling arithmetic on the windows as stated — no interpretation.
The 15-day measures are the ones that quietly govern your submission calendar. Because a blood pressure, weight, or PROM has to be collected within 15 days of the submission that carries it, the collection visit and the submission are effectively bolted together: you cannot draw a vital sign, sit on it for a month, and submit it inside a quarterly window that opened later. In practice the tightest 15-day measure in a beneficiary's bundle sets the real deadline for the whole submission — the submission is only as fresh as its most perishable reading.
That interacts directly with the rolling window from the last post. You can be comfortably inside your 70–110-day quarterly window and still have an invalid submission, because the blood pressure in it was collected 18 days ago. The two clocks are independent: one governs when the submission is allowed, the other governs whether each reading in it is still alive. Both have to be green at the same instant.
Being precise about the source means being precise about its edges. The Payment paper states the six windows on p.6; in our reading it frames them as collection-to-submission freshness bounds. It does not, for instance, spell out the exact handling of edge cases like a reading collected exactly on the boundary day, or how a measure is treated when a beneficiary's condition status changes between collection and submission (which could move an HbA1c between the one-year and two-year window). We're not going to guess in either direction. Build to the conservative edge of each window, and put boundary questions to CMS directly — the ACCESS Technical FAQ and the model's support channels exist for exactly this class of question.
The rule set dictates the design. A validity layer has to carry, per measure and per beneficiary: the collection date, the applicable window (which for HbA1c and LDL-C is resolved from the beneficiary's condition, not a constant), and the derived expiry date. A submission is only assemblable when every measure it contains is unexpired on the submission date — so the practical check is "what is the earliest expiry among the readings I'm about to submit, and is that date still in the future?"
If you're tracking this in a spreadsheet today, that's workable at small volume — the point is that "is this reading still valid" is a computed column with a per-measure, sometimes per-condition rule behind it, not a single global freshness constant. The measures that expire in 15 days are the ones that will bite first.
Next in the series: manual entry is not permitted — the ACCESS collection-method rules, and why "valid" starts before the reading is ever recorded.
---
Outcome Rail builds reporting infrastructure for ACCESS participants. If you're working through these windows, we're happy to compare notes: [email protected].
Sources: CMS ACCESS Model Payment Amounts and Performance Targets (PDF), p.6 (OAP Measure Validity) · ACCESS RFA (PDF), Appendix C · ACCESS Technical FAQ.
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.