MA practice · 4,000 attributed lives · 18-month encounter sample
V28 recapture lifted RAF by +0.14 and pulled $5.8M PMPY back.
Outdated V24 condition maps, MEAT documentation gaps, and a 46 percent annual wellness visit rate were quietly bleeding RAF score. A V28 crosswalk, a suspect generator, and an AWV-first recapture campaign closed the gap inside 18 months.
Specialty
MA primary care
Attributed lives
4,000
Sample window
18 months
Model
V28 HCC
RAF lift
+0.14
Recovered
$5.8M PMPY
The lift, visualized+0.14 RAF. From 1.08 to 1.22.
The situationThe V28 transition. And the silence around it.
BASELINE
Every MA encounter ran against the old V24 map. Chronic conditions slid off the panel one quarter at a time.
Leak 01 · Outdated map
V24 stayed in the chart prompts.
MAP
V24
LEAK
#1
Leak 02 · MEAT gaps
Documentation missed the elements.
MISSING
31%
RADV RISK
High
Leak 03 · AWV drag
Suspects never came in.
AWV
46%
CONFIRMED
No
V24 to V28 · the weight shift that costs you money
RULE
Negative deltas drop RAF unless the panel is re-coded. Top 6 condition families that moved the most.
| CONDITION FAMILY | V24 WEIGHT | V28 WEIGHT | Δ |
|---|---|---|---|
| Diabetes with complications, specified | 0.302 | 0.166 | −45% |
| Vascular disease without complications | 0.288 | 0.000 | DROPPED |
| CKD stage 3a (split from CKD 3) | 0.069 | 0.127 | +84% |
| Major depression, recurrent severe | 0.309 | 0.330 | +7% |
| Protein-calorie malnutrition | 0.518 | 0.471 | −9% |
| Substance use disorder, moderate/severe | 0.383 | 0.317 | −17% |
SOURCE · CMS-HCC model V28 final ruling · weights normalized to community non-dual aged
The funnelFrom 18,000 encounters to 1,250 recaptured HCCs.
ENCOUNTERS
18,000
SUSPECTS
4,200
CONFIRMED
1,800
RECAPTURED
1,250
The fix · workflowCrosswalk. Suspect. Confirm.
01 CROSSWALK
Normalize the chart
02 SUSPECT
Surface the conditions
03 RECAPTURE
Bring the patient in
04 CONFIRM
AWV under MEAT
Where the dollars came fromTop 8 condition families. $ recovered.
ORDER MATTERS
Diabetes, CKD, and heart failure carry the recapture. Build the AWV outreach list in that order.
The 90-day operating planFour phases. One pipeline.
WEEK 01-03
Condition map
18-month sample crosswalked to V28. Drops, splits, reweights flagged per provider.
WEEK 04-06
Suspect generation
Model surfaces 102 suspects per panel from labs, meds, prior HCCs. Pre-visit packets built.
WEEK 07-11
Recapture campaign
AWV outreach goes live. Care managers schedule highest-suspect patients first.
WEEK 12
RAF measurement
First measured RAF movement. AWV above 60%. Suspect close rate clears 40%.
What the dashboard showsRAF by condition. Suspect to confirmed funnel.
GREEN
Healthy capture
AMBER
Recapture window
RED
Gap · AWV this quarter
RAF by HCC category · 4,000 lives
refreshed 22s ago
Diabetes with complicationsHCC 18 · 19 · 36
428 sus
71%
CKD stage 3 and 4HCC 138 · 139
182 sus
54%
Vascular diseaseHCC 263 · 264
214 sus
66%
Major depressionHCC 155
96 sus
48%
Heart failureHCC 226
147 sus
35%
Active neoplasmHCC 17 · 22 · 23
61 sus
78%
+24 more categories · 102 avg suspects per panel · 61% close rate
VBC Engagement ScorecardQuarterly partner report · Q4 redacted
CONFIDENTIAL
KPI movement before vs current.
RAF score, panel average
1.08
1.22
AWV completion rate
46%
73%
Suspect close rate
22%
61%
MEAT documentation pass
69%
94%
Chronic condition recapture
58%
89%
PMPY revenue recovered
$0
$5.8M
Q4 · 2026
ASP-RCM Senior Partner
The RAF lift came from the 4 percent of patients we were missing the most.
Chief Medical Officer · MA practice
OutcomesBefore. After. In numbers.
Pre-engagement · baseline
RAF score
1.08
AWV completion
46%
Suspect close rate
22%
Chronic recapture
58%
Annualized leak ≈ $5.8M PMPY
Steady-state · month 12+
RAF score
1.22
AWV completion
73%
Suspect close rate
61%
Chronic recapture
89%
18-month cumulative lift +0.14 RAF
V28 was costing us $5.8M PMPY and we were measuring it retrospectively. Once we moved measurement to the encounter, the documentation followed.
, CMO · MA practice · 4,000 attributed lives · anonymized
Capability stackWhat the recapture pipeline actually runs on.
NOT
A spreadsheet
IS
Service in HCC Coding AI
SCORECARD
Held in Reporting Cloud
AUDIT
One PHI access log
Layer 04 · AI
HCC Coding AI · Suspect generator · MEAT scorer
▼
Layer 03 · LLM Gateway
Single audited choke point · cost meter · prompt registry
▼
Layer 02 · Platform
Reporting Cloud · Panel scorecard · AWV scheduler
▼
Layer 01 · HIPAA-eligible AWS
AES-256-GCM PHI · row-level RBAC · PHI access log
Common questionsFrequently asked: V28 recapture.
What is V28 and why does it matter?
V28 is the CMS-HCC model that phases in across payment years 2024 to 2026, replacing V24. It drops some HCC categories, splits others, and tightens documentation rules. Practices still coding to V24 condition maps quietly lose RAF score, which translates directly into lower per-member-per-year payment from MA plans.
What is MEAT documentation?
MEAT is the standard CMS expects for every chronic condition reported on a claim: Monitor, Evaluate, Assess, Treat. The progress note has to show evidence of all four in a way an auditor can find. In the audited sample, 31 percent of chronic-condition notes were missing at least one MEAT element, which makes those HCCs vulnerable on RADV review.
What is the difference between a suspect HCC and a confirmed HCC?
A suspect HCC is a condition the model thinks the patient likely has based on prior claims, labs, medications, or specialist notes, but it has not been documented in the current payment year. A confirmed HCC is one where a face-to-face encounter this year captured the condition with MEAT documentation and the correct V28 code. Suspects only count when they are confirmed in a billable encounter.
Why anonymize the client?
The master service agreement includes reciprocal confidentiality. The numbers and timeline are real. A senior partner can walk you through methodology and host a reference call under NDA once both sides agree.
Can we get a reference call?
Yes, at late-stage diligence with client consent. The reference is typically a 30-minute call with the practice's Chief Medical Officer or VBC director.
How long until we see RAF movement?
The 90-day plan delivers crosswalk, provider education, and AWV scheduling. Visible suspect-close-rate movement starts in week six. Steady-state shows up by month nine when AWV completion clears 70 percent. The +0.14 RAF figure is the 18-month cumulative lift.
What does the free audit look like for value-based contracts?
Send 18 months of encounter data with diagnosis codes, your panel attribution file, an AWV completion report, and one quarter of progress notes for chart review. Inside 30 days you receive a 4-page written audit covering RAF gap analysis in dollars, suspect generation by HCC category, MEAT documentation scoring, and a 90-day recapture plan.
Does this only work for large attributed panels?
No. The math scales linearly. A 1,000-life panel runs the same crosswalk, the same suspect generator, and the same AWV cadence. The dollar lift is smaller in absolute terms but the percentage lift on RAF holds. The discipline does not change with panel size.
Want the same audit applied to your panel?
A free 30-day audit. Send 18 months of encounter data and your panel attribution file. We return a 4-page written audit covering your RAF gap by HCC category, suspect generation per panel, MEAT documentation scoring, and a 90-day recapture plan. Yours to keep. No SDR follow-up.