The V28 RAF optimization playbook.
CMS-HCC V28 is 100 percent of the risk score in payment year 2026. Fewer payment HCCs and remapped conditions compress RAF for the same clinical picture. There is no coding trick to recover it. There is only complete, compliant capture. This is the working playbook.
Executive summaryFive moves to protect RAF under V28.
The V28 transition is a documentation problem before it is a coding problem. CMS did not cut payment. It raised the bar on what has to be documented, at what frequency, to earn the same RAF. Groups that treat V28 as a capture discipline hold their PMPM. Groups that wait for the annual sweep lose it.
V28 did not cut your rate. It moved the goalpost on documentation.
The base rate is unchanged. What changed is the crosswalk: V28 expanded to the full ICD-10 code set but reduced the count of payment HCCs and remapped many conditions to lower coefficients. The same clinically accurate note now maps to a lower RAF unless the chronic burden is documented completely and every year. The recovery lever is capture completeness, not code selection.
The primary source for the transition timeline is CMS, which finalized the three-year V24-to-V28 phase-in and set payment year 2026 at 100 percent V28 in the annual Rate Announcement. Everything downstream of that decision is a workflow question: which conditions were captured last year, which are still open this year, and whether the note supports the code on audit.
What changedThe V28 phase-in, in three payment years.
CMS blended the outgoing V24 model and the incoming V28 model over three payment years so plans could adjust. In 2026 the blend is over. Risk scores are pure V28.
What V28 actually did to the crosswalk
| Condition family | Under V24 | Under V28 | RAF effect |
|---|---|---|---|
| Diabetes without complication | Payment HCC | Remapped / lower weight | Down |
| Diabetes with complication | Payment HCC | Retained, coefficient revised | Down |
| Peripheral / other vascular disease | Payment HCC | Several codes dropped from payment | Down |
| Major depression & mood disorders | Payment HCC | Retained, category restructured | Mixed |
| Serious chronic conditions (CHF, COPD, CKD) | Payment HCC | Retained as payment HCC | Stable |
Illustrative of the directional changes CMS made in V28. Confirm the specific payment status and coefficient for any code against the current-year CMS-HCC model files before coding. Directions shown are general and not code-level guarantees.
The mechanismWhy RAF compresses, and what it costs.
RAF is a multiplier on the base rate. CMS pays the plan or group base rate times the member RAF, per member, per month. When V28 remaps conditions to lower coefficients, the same documentation yields a lower RAF, and the lower RAF flows straight into PMPM.
Every open chronic HCC that is real and documentable is captured this year, not deferred.
Recapture happens across the year, so the fourth quarter is a cleanup, not the whole job.
Every captured HCC is backed by MEAT so it survives RADV. Unsupported codes are liabilities.
Signature frameworkThe recapture calendar.
This is the distinguishing discipline of V28 optimization. Every chronic HCC resets to zero at midnight on January 1. Its RAF contribution for the year is zero until a face-to-face encounter re-documents it. Left to the annual sweep, recapture piles into the fourth quarter and much of it never happens. The calendar pulls it forward.
An HCC coded in 2025 is worth zero in 2026 until you document it again.
This is the single most misunderstood fact in risk adjustment. Chronic conditions do not carry forward. Diabetes, CHF, CKD, COPD, and major depression must be re-documented in a qualifying face-to-face encounter every calendar year to contribute to that year's RAF. The recapture calendar exists so that a known chronic burden is not left to chance in December.
Find before the visitThe suspect-to-confirmed funnel.
Recapture handles conditions you already know about. Suspecting handles conditions the data implies but the chart has not yet coded. Labs, medications, prior claims, and problem-list history all point to conditions that are likely present and undocumented. The funnel turns those signals into confirmed, MEAT-supported codes.
A suspect is a question, never an answer.
The funnel narrows on purpose. A signal is only a prompt to investigate. The provider confirms or rules out the condition based on the clinical picture in front of them, and only a validated, MEAT-supported diagnosis becomes a coded HCC. Coding a suspect without clinical validation is exactly the pattern CMS RADV audits are built to catch. Volume at the top of the funnel is worthless if it is not clinician-owned at the bottom.
Documentation disciplineThe MEAT checklist.
MEAT is the test every documented HCC has to pass. A condition on the problem list is not enough. The note must show the condition was Monitored, Evaluated, Assessed, or Treated at the encounter. One MEAT element supports the code. Zero MEAT elements make it a RADV liability.
| Documentation pattern | MEAT? | Audit outcome |
|---|---|---|
| Condition listed in problem list only, no note | No | Unsupported, high RADV risk |
| "History of diabetes" with no current management | No | Status unclear, not codeable as active |
| Condition named with plan to monitor labs | Yes (M) | Supported |
| Condition assessed as stable with a follow-up plan | Yes (A) | Supported |
| Condition with a medication started or continued | Yes (T) | Supported and strongest |
Size the exposureThe revenue-at-risk model.
The playbook earns its keep when the RAF gap is quantified. The model is deliberately simple: measure the average RAF gap per member, multiply by the base rate to get PMPM lost, annualize, and scale across the panel. Every number below is illustrative and archetype-based, sized to show the mechanism, not to quote a client.
| Input | Archetype value | Note |
|---|---|---|
| Panel size | 12,000 lives | MA-heavy primary care group |
| Base rate | $960 PMPM | Illustrative monthly base, per member |
| Average RAF gap | 0.10 | Open, documentable HCCs not yet recaptured |
| PMPM lost per member | $96 | 0.10 × $960 |
| Annual per member | $1,152 | $96 × 12 months |
| Panel-level exposure | ~$13.8M / yr | $1,152 × 12,000, illustrative |
An HCC coding engine and RAF dashboard behind clinician-owned capture.
ASP-RCM runs an HCC coding engine and a RAF dashboard that surface suspect conditions from the data, flag open gaps against each member's prior-year HCC set, and deliver point-of-care prompts to the provider. The engine supports coders and clinicians. It does not code autonomously and it does not replace clinical judgment. Every suspect is validated at the point of care and every confirmed HCC is backed by MEAT. The tooling makes the recapture calendar and the suspect funnel run at panel scale; the people own the decisions.
V28 did not take money off the table. It moved the money behind a documentation bar. Groups that recapture across the year and back every code with MEAT hold their RAF. Groups that wait for the December sweep hand it back.
Common questionsV28, RAF, and recapture.
Is CMS-HCC V28 fully phased in?
Why does V28 compress RAF?
What is RAF and how does it drive revenue?
Why do chronic HCCs have to be recaptured every year?
What is MEAT documentation?
How does the Annual Wellness Visit fit the recapture calendar?
Does ASP-RCM code charts autonomously?
Know your RAF gap before the year gets away from you.
We will run a free RAF capture audit against your panel: open chronic HCCs, suspect conditions the data implies, and the revenue at risk under V28. You keep the findings whether or not we work together.