Our Goal
We're revolutionizing healthcare with AI that thinks, learns, and optimizes risk adjustment for better patient outcomes and organizational success.
Medicare Advantage · Risk Adjustment Data Validation
RADV audits just became every plan's reality.
CMS is moving from a handful of audits a year to all eligible MA contracts — across six open payment years. Health Data Max keeps your submissions continuously audit-ready.
PY2020–PY2025 audits scheduled · Findings extrapolated across the contract
The Shift
~550
MA contracts audited per year — up from about 60
Records per contract35–200
Payment years in scope2018–2024
Audit volume change+900%
MRs per audited HCCup to 2
All
eligible MA contracts now audited annually
6
payment years (PY2020–PY2025) on the schedule
~70%
of high-risk codes OIG reviewed were unsupported
2× HCCs
max medical records per enrollee (up to 2 per audited HCC)
The Fundamentals
What is a RADV audit?
Risk Adjustment Data Validation is how CMS checks whether the diagnoses an MA plan submitted for payment are actually supported in the member's medical record.
CMS pays MA organizations a risk-adjusted amount for each enrollee. Diagnoses submitted by providers map to Hierarchical Condition Categories (HCCs), each carrying a factor that raises the member's risk score — and the monthly payment to the plan.
Because payment follows diagnosis, CMS validates a sample of those diagnoses against the underlying charts. In a RADV audit, the plan must produce a medical record that supports each sampled HCC. If the record doesn't support the diagnosis, the HCC is invalidated, the risk score is overstated, and the associated payment is recovered.
CMS can then extrapolate the sampled error across the contract's full sampling frame, turning a modest sample into a large recovery. The 2023 rule that authorized extrapolation (and removed the fee-for-service adjuster) was vacated by a federal court in 2025 and is under appeal — but CMS has expressly designed its PY2020 and PY2021 audits to support extrapolated recoveries and reserves the right to collect them if legally permissible. Plan for extrapolation; don't bank on the litigation.
How a diagnosis becomes a payment — and an audit target
1
Provider documents & codesAn encounter produces an ICD-10 diagnosis from the medical record.
2
Plan submits to CMSCodes flow via encounter data / RAPS for risk adjustment.
3
Diagnosis maps to an HCCThe HCC factor increases the enrollee's risk score.
4
Risk score drives paymentCMS multiplies risk score × base rate for the monthly payment.
5
RADV validates the chartNo supporting record → HCC invalidated → overpayment recovered & extrapolated.
Why This Matters Now
The audit landscape changed in 2025
On May 21, 2025, CMS announced an aggressive strategy to enhance and accelerate MA audits — clearing the backlog and auditing everyone.
// SCALE
From ~60 to ~550 plans
CMS will audit all eligible MA contracts each year in newly initiated audits — roughly a 900% increase in audit volume, reaching nearly every contract.
// DEPTH
35 to 200 records each
Sample size scales with contract size (200 / 100 / 50 / 35 by stratum), and members aren't picked at random — CMS targets the enrollees its improper-payment model predicts will lose the most risk score.
// SPEED
PY2018–PY2024 backlog
CMS expanded its coder workforce from ~40 to ~2,000 and is using AI-assisted review to clear years of open audits on a compressed timeline — though final overpayment determinations stay with human coders.
Inside the Process
How a contract-specific RADV audit works
CMS's PY2020 and PY2021 Audit Methods & Instructions lay out the full mechanics — from how members are selected to how overpayments are calculated and extrapolated.
It starts before you know it. CMS defines a sampling frame, draws a statistically valid random sample, and sends an Audit Notice to your CEO, CFO, COO, and Compliance Officer. Your team designates points of contact, registers in CMS's secure CDAT portal, and downloads the Enrollee Data List (EDL) naming the exact enrollees, HCCs, and diagnosis codes under review.
From there the clock runs hard: pull a valid medical record — legibly signed and dated, from a face-to-face visit by a credentialed provider, within the data collection year — for every audited HCC, and submit it through CDAT before the deadline. CMS coders then abstract diagnoses through up to three rounds of review and classify each HCC as Confirmed, Confirmed Higher, Discrepant, Discrepant Lower, or Administrative Exception.
The audit lifecycle
1
Frame & sampleCMS builds the sampling frame and draws 35–200 enrollees by random selection.
2
Audit Notice & CDATSent to CEO/CFO/COO/MCO; POCs register and access the secure portal.
3
Enrollee Data ListNames the sampled enrollees, audited HCCs, and the diagnosis codes you must support.
4
Gather & submit MRsUp to 2 records per audited HCC, as PDFs (≤100MB), each with a coversheet, by the deadline.
5
AbstractionCertified coders review each record across up to three rounds.
6
Payment error & extrapolationPer-enrollee error summed; average risk-score change extrapolated across the frame.
How CMS picks the members it audits
The sample isn't random across your whole book. CMS first ranks audit-eligible enrollees — broadly, those enrolled for all 12 months of the data-collection year and at least one month of the payment year — with a "predicted improper payment" model, then samples from the top quartile most likely to lose risk score under review. The exact formula isn't published, but it's expected to lean on circumstantial coding-risk signals:
- HCCs supported by only a single encounter
- HCCs added solely through chart review (no underlying encounter)
- HCCs with little corroborating medical or pharmacy utilization
// These are the exact OIG high-risk signatures below — acute Dx with no inpatient claim, cancer with no treatment, embolism with no anticoagulant. The audit target and the OIG pattern are the same thing.
Sample size is set by your stratum
CMS sorts RADV-eligible contracts by sampling-frame size into four strata. The ten largest contracts draw the deepest samples — which is exactly why our data shows audit probability rising with enrollment.
200
Stratum 1 — 10 largest contracts
100
Stratum 2 — top third of the rest
50
Stratum 3 — middle third
35
Stratum 4 — bottom third
How the dollars are calculated. CMS computes a payment error for each sampled enrollee, then — for the contract — multiplies the average change in risk score by the sum of county rates across every enrollee in the sampling frame, applying the lower bound of a 90% confidence interval. That makes county-level rates the multiplier on your entire exposure, and a single unsupported HCC in the sample a proxy for many. Telehealth note: for PY2021 dates of service Mar 6–Dec 31 2020, audio-video telehealth visits can satisfy the face-to-face requirement.
CMS Published Schedule
When CMS intends to initiate audits
The month CMS plans to begin RADV audits, by MA payment year. Six payment years are now on the calendar.
PY2020
Mar 2026
PY2021
May 2026
PY2024
Aug 2026
PY2023
Nov 2026
PY2022
Jan 2027
PY2025
Apr 2027
Source: CMS MA RADV Audit Schedule. Dates are subject to change; CMS updates the schedule as needed.
Payment Year 2020 · data collection year 2019
Audit milestones
Enrollee Data List in CDATApr 3, 2026
MR submission window opensApr 13, 2026
MR submission deadlineAug 28, 2026
Hardship exception deadlineSep 11, 2026
Payment Year 2021 · data collection year 2020
Audit milestones
Enrollee Data List in CDATJun 12, 2026
MR submission window opensJun 22, 2026
MR submission deadlineNov 6, 2026
Hardship exception deadlineNov 20, 2026
Once an audit is initiated you get roughly four months to assemble a valid record for every audited HCC. Source: CMS PY2020 & PY2021 Audit Methods & Instructions.
What the Audits Find
OIG's high-risk diagnosis findings
HHS-OIG built a public toolkit from its MA audits, identifying diagnosis codes that — when paired with other data — are at high risk of being unsupported. The error rates are striking.
90%error rate across these high-risk groups
Across its audits, OIG found that roughly 70% of the high-risk diagnosis codes it reviewed were not supported by the associated medical records — and within specific groups, error rates exceeded 90%. These aren't edge cases; they're patterns a plan can detect and correct before CMS arrives.
| High-Risk Group | Records in Scope | Errors | Error Rate |
|---|---|---|---|
| Acute stroke | 945 | 908 | |
| Acute heart attack | 791 | 751 | |
| Embolism | 754 | 593 | |
| Lung cancer | 391 | 345 | |
| Breast cancer | 390 | 373 | |
| Colon cancer | 390 | 368 | |
| Prostate cancer | 360 | 322 | |
| Potentially mis-keyed codes | 522 | 421 | |
| Totals | 4,543 | 4,081 | 90% |
Source: HHS-OIG, "Toolkit To Help Decrease Improper Payments in Medicare Advantage Through the Identification of High-Risk Diagnosis Codes" (Dec 2023, A-07-23-01213), errors as of Nov 2023.
The Coding Patterns
Where unsupported HCCs hide
Each high-risk group follows a recognizable signature: an acute diagnosis submitted without the clinical evidence you'd expect to accompany it. In most cases, a "history of" code — which doesn't map to an HCC — was what the record actually supported.
HCC 100 · Stroke
Acute stroke
One acute stroke diagnosis on a physician claim, with no matching inpatient or outpatient hospital claim.
Acute stroke→History of stroke
HCC 86 · AMI
Acute heart attack
An acute MI on a single physician/outpatient claim with no inpatient claim within 60 days before or after.
Acute MI→History of MI
HCC 107/108 · Vascular
Embolism
An embolism diagnosis with no anticoagulant medication dispensed — the treatment you'd expect for an active embolism.
Acute embolism→History of embolism
HCC 9 · Cancer
Lung cancer
A lung cancer diagnosis with no surgery, radiation, or chemotherapy within six months before or after.
Active lung cancer→History of lung cancer
HCC 12 · Cancer
Breast cancer
A breast cancer diagnosis without surgical, radiation, chemotherapy, or relevant drug treatment in the window.
Active breast cancer→History of breast cancer
HCC 11 · Cancer
Colon cancer
A colon cancer diagnosis with no surgical therapy, radiation, or chemotherapy in the six-month window.
Active colon cancer→History of colon cancer
HCC 12 · Cancer
Prostate cancer
A prostate cancer diagnosis in members 74 or younger with no treatment evidence in the surrounding window.
Active prostate cancer→History of prostate cancer
Data Entry
Potentially mis-keyed codes
A single transposed or mistyped code (e.g., I720 → I270) that mapped to an unrelated, unvalidated HCC.
Mis-keyed code→Correct diagnosis
The Health Data Max Approach
One platform. Continuous audit readiness.
HDM doesn't wait for an Audit Notice. Provider claims, encounters, CMS response files, payment and risk-score data, and chart coding all live in one database — so every submitted diagnosis stays continuously linked to the record that has to support it. When CMS initiates on its new accelerated timeline, you're already prepared instead of scrambling inside a four-month window.
01
Unify in one DB
Claims, encounters, CMS response files, risk scores, payments, and chart coding — reconciled and member-linked in a single source of truth.
→02
Flag the suspects
Mirror CMS's improper-payment signals and OIG high-risk logic — single-encounter HCCs, chart-review-only adds, thin utilization — before they're ever sampled.
→03
Match the evidence
Plans upload medical records; HDM surfaces the strongest supporting chart for each suspect member-HCC.
→04
Validate or correct
Confirm defensible HCCs, and route truly unsupported diagnoses for correction — accurate payment, clean record.
✓
One source of truthClaims, encounters, response files, risk scores, and charts in a single linked database — no scramble to reassemble data once an audit lands.
✓
CMS- & OIG-aligned flaggingThe improper-payment signals and high-risk patterns auditors use, run against your data continuously.
✓
Best-chart selection per HCCFor every flagged member, surface the record most likely to support the diagnosis — within the limit of two records per audited HCC.
✓
Encounter ↔ chart reconciliationTie every submitted code back to source documentation, exposing gaps before CMS does.
✓
Extrapolation-aware prioritizationFocus first on the sampled-frame HCCs whose error would hurt most when projected across the contract.
✓
CDAT-ready packages & audit trailValid, coversheet-matched PDF submissions, with every flag and decision logged — governed, defensible, HIPAA-aligned.
Accuracy in both directions. The goal isn't to maximize codes — it's to make every submitted HCC defensible. HDM flags potential over-coding and unsupported diagnoses for correction just as it surfaces supporting evidence, so your risk capture is accurate, compliant, and able to withstand extrapolated review.
Get Ahead of Your Payment Year
Be ready before the audit notice arrives.
With PY2020–PY2025 on the CMS schedule, the question isn't whether your plan will be audited — it's whether your records are ready. Let's review your highest-risk HCCs together.
ACA HHS-RADV · Knowledge Guide
HHS-RADV: the complete reference for your team.
Program rules, audit workflow, sampling logic, medical record standards, business rules, and deadlines for BY2024 and BY2025 — in one place.
BY2024 audits active · BY2025 EDGE locked Apr 30, 2026
Annual
audit cycle — every eligible issuer, every benefit year
200
enrollees sampled per issuer (standard IVA sample)
Both ±
outlier issuers receive RA transfer adjustments
3 layers
IVA → SVA → Error Estimation before payment impact
The Fundamentals
What is HHS-RADV?
HHS Risk Adjustment Data Validation (HHS-RADV) is CMS's annual audit program that verifies the accuracy of the diagnoses issuers submit to the EDGE server — the same diagnoses that drive risk scores and determine how much each issuer receives or pays through ACA Risk Adjustment.
Risk adjustment redistributes funds among issuers in the same state market risk pool based on relative member health risk. Issuers with sicker-than-average enrollees receive transfers; those with healthier-than-average enrollees pay charges. RADV audits the data quality behind those transfers.
Because payment follows diagnosis, CMS validates a sample of those diagnoses against the underlying medical records. An unsupported HCC fails, the issuer's Plan Liability Risk Score (PLRS) is adjusted, and RADV-adjusted RA transfers are applied.
How a diagnosis becomes an audit target
Provider documents & codes →ICD-10 diagnosis from encounter
Issuer submits to EDGE →Diagnosis accepted, maps to HCC
HCC raises risk score →PLRS determines RA transfer
RADV samples enrollees →Medical records must support HCCs
Unsupported HCC fails →Super HCC failure rate rises
Outlier issuer identified →Error rate adjustment to PLRS
Why This Matters Now
The audit cycle runs every benefit year
// SCOPE
Individual + small group markets
All non-grandfathered, non-transitional plans. Issuers in states with more than one issuer in the risk pool are subject to sampling.
// STAKES
RA transfer is at risk
An issuer with a high Super HCC failure rate in any of 3 Failure Rate Groups may qualify as an outlier — triggering an error rate adjustment to their PLRS for that benefit year.
// TIMING
Audits lag by ~18 months
EDGE data locked April 30 → RA cycle closes ~May → sample drawn ~Jun–Jul → IVA packages due ~Jan of the following year → final results ~Jun. Accuracy during submission time determines audit outcome.
Two-Layer Validation
IVA and SVA: independent audits at every level
Initial Validation Audit — IVA
Issuer-contracted, CMS-approved
The issuer selects a CMS-approved IVA Entity to review medical records for each of the 200 sampled enrollees. The IVA Entity validates D&E data, RXC categories, and every EDGE-submitted HCC against source documentation — submitting findings via the CMS RADV Audit Tool.
DeliverablesPackage 1 (D&E), Package 2 (HCC), Package 3 (IRR)
Coder requirement≥95% IRR before abstraction begins
Submission systemCMS RADV Audit Tool (HIOS/Enterprise Portal)
BY2025 deadline~Jan 2027 (est.)
Second Validation Audit — SVA
CMS-contracted, nationally uniform
A single CMS-designated SVA Entity independently re-abstracts a subsample of IVA records with no access to IVA findings. The pairwise means test determines whether IVA or SVA results feed error estimation. Subsamples expand automatically if precision is insufficient.
Initial subsample24 enrollees
Expansion sequence24 → 50 → 100 (full SVA-100 expansion if needed)
Statistical testBootstrap pairwise means (K=10,000, 90% CI)
Access to IVA resultsNone — fully independent
Markets covered
Non-grandfathered individual and small group (including merged) market risk adjustment covered plans, inside and outside the Exchange.
Regulation
45 CFR §§ 153.350, 153.630. Mandated under the ACA; implemented through annual CMS Payment Notice rulemaking.
HCC model BY2025
HHS-HCC Version 08 (V08). Mappings and coefficients published in the 2026 Payment Notice (90 FR 4424).
The Health Data Max Approach
One platform. Continuous audit readiness.
HDM doesn't wait for the RADVIVAS reports to land. EDGE submissions, accepted claims, RA outputs, and source enrollment are reconciled in one database — so every HCC is traceable to the chart that has to support it before CMS samples.
01
Reconcile EDGE data
Enrollment, claims, RA outputs, and RADV reports linked per issuer and benefit year.
→
02
Flag audit risk
Mirror Super HCC failure rate signals and D&E/RXC anomaly patterns before the sample drops.
→
03
Surface evidence
For each sampled enrollee-HCC, surface the strongest linked medical record and claim chain.
→
04
Validate or correct
Confirm defensible HCCs, route unsupported diagnoses for correction — accurate RA transfer.
✓
One source of truth
Claims, encounters, RADV reports, risk scores, and charts in a single linked database.
✓
CMS- & EDGE-aligned flagging
Super HCC failure rate signals and D&E/RXC anomaly patterns run continuously against your data.
✓
Best-chart selection per HCC
For every flagged enrollee, surface the record most likely to support the diagnosis before audit.
✓
Encounter ↔ chart reconciliation
Every submitted code tied back to source documentation — gaps found before CMS does.
✓
Pre-sample risk prioritization
Focus on the HCCs whose failure rate would drive outlier status, before the sample drops.
✓
Clean audit submissions
RADV Audit Tool–ready XML packages with every flag and decision logged and defensible.
Accuracy in both directions. The goal isn't to maximize codes — it's to make every submitted HCC defensible. HDM flags potential over-coding and unsupported diagnoses for correction just as it surfaces supporting evidence, so your risk capture is accurate, compliant, and able to withstand the SVA's independent re-review.
Get Ahead of Your Benefit Year
Be ready before the RADVIVAS lands.
With BY2024 audits active and BY2025 on the calendar, the window to get ahead of your IVA cycle is now.
ACA HHS-RADV · Knowledge Guide
HHS-RADV: the complete reference for your team.
Program rules, audit workflow, sampling logic, medical record standards, business rules, and deadlines for BY2024 and BY2025 — in one place.
BY2024 audits active · BY2025 EDGE locked Apr 30, 2026
The Program
200
enrollees sampled per issuer — standard IVA sample
Audit cycleAnnual
Validation layersIVA → SVA → Error Est.
PLRS adjustmentBoth ± directions
IRR threshold≥95%
Annual
audit cycle — every eligible issuer, every benefit year
200
enrollees sampled per issuer (standard IVA sample)
Both ±
outlier issuers receive RA transfer adjustments
3 layers
IVA → SVA → Error Estimation before payment impact
The Fundamentals
What is HHS-RADV?
CMS's annual audit program that verifies the diagnoses issuers submit to the EDGE server are actually supported by medical records.
HHS Risk Adjustment Data Validation (HHS-RADV) is CMS's annual audit program that verifies the accuracy of the diagnoses issuers submit to the EDGE server — the same diagnoses that drive risk scores and determine how much each issuer receives or pays through ACA Risk Adjustment.
Risk adjustment redistributes funds among issuers in the same state market risk pool based on relative member health risk. Issuers with sicker-than-average enrollees receive transfers; those with healthier-than-average enrollees pay charges. RADV audits the data quality behind those transfers.
Because payment follows diagnosis, CMS validates a sample of those diagnoses against the underlying medical records. An unsupported HCC fails, the issuer's Plan Liability Risk Score (PLRS) is adjusted, and RADV-adjusted RA transfers are applied.
How a diagnosis becomes an audit target
1
Provider documents & codesICD-10 diagnosis from a face-to-face or telehealth encounter.
2
Issuer submits to EDGEDiagnosis accepted on EDGE, maps to an HCC.
3
HCC raises risk scoreHCC coefficient increases enrollee risk score; PLRS determines RA transfer.
4
RADV samples enrolleesMedical records must support each sampled HCC.
5
Unsupported HCC failsSuper HCC failure rate rises for that issuer.
6
Outlier issuer identifiedError rate adjustment applied to PLRS.
Why This Matters Now
The audit cycle runs every benefit year
RADV isn't a periodic spot-check — it's a structural part of every issuer's annual RA settlement.
// SCOPE
Individual + small group markets
Non-grandfathered individual and small group (including merged) market RA-covered plans, inside and outside the Exchange, in states where HHS operates risk adjustment.
// STAKES
RA transfer is at risk
An issuer with a high Super HCC failure rate in any of 3 Failure Rate Groups may qualify as an outlier — triggering an error rate adjustment to PLRS for that benefit year.
// TIMING
Audits lag by ~18 months
EDGE data locked April 30 → RA cycle closes ~May → sample drawn ~Jun–Jul → IVA packages due ~Jan of the following year → final results ~Jun. Submission-time accuracy decides audit outcome.
Two-Layer Validation
IVA and SVA: independent audits at every level
Two separate entities review the same enrollees — one contracted by the issuer, one contracted by CMS — and their results are statistically reconciled.
Initial Validation Audit — IVA
Issuer-contracted, CMS-approved
The issuer selects a CMS-approved IVA Entity to review medical records for each of the 200 sampled enrollees. The IVA Entity validates D&E data, RXC categories, and every EDGE-submitted HCC against source documentation, submitting findings via the CMS RADV Audit Tool.
DeliverablesPackage 1, 2 & 3
Coder requirement≥95% IRR
Submission systemCMS RADV Audit Tool
BY2025 deadlineJan 7, 2027
Second Validation Audit — SVA
CMS-contracted, nationally uniform
A single CMS-designated SVA Entity independently re-abstracts a subsample of IVA records with no access to IVA findings. The pairwise means test determines whether IVA or SVA results feed error estimation.
Initial subsample24 enrollees
Expansion path24 → 50 → 100
Statistical testBootstrap K=10,000
Access to IVA resultsNone
Markets covered
Non-grandfathered individual and small group (including merged) market risk adjustment covered plans, inside and outside the Exchange.
Regulation
45 CFR §§ 153.350, 153.630. Mandated under the ACA; implemented through annual CMS Payment Notice rulemaking.
HCC model BY2025
HHS-HCC Version 08 (V08). Mappings and coefficients published in the 2026 Payment Notice (90 FR 4424).
End-to-End Process
From EDGE submission to RADV-adjusted RA transfer
The RADV lifecycle spans ~18–24 months from the benefit year's EDGE data cutoff to the final adjusted transfer. EDGE accuracy during the benefit year is what determines audit outcomes — the RADV command audits data that was locked months earlier.
Key timing insight
The RADV sample is drawn from data already locked on the EDGE server. By the time a RADVIVAS report lands, an issuer has no ability to correct the underlying submission. Continuous EDGE data quality monitoring during the benefit year is the only effective RADV preparation.
Sample Selection
How CMS selects the 200 enrollees your IVA will audit
The IVA sample is not a random draw from an issuer's full enrollment — CMS uses Neyman Allocation, a stratified sampling method.
CMS uses Neyman Allocation — a stratified sampling method that allocates sample size proportional to each stratum's share of the issuer population weighted by risk score variability.
For BY2025 and forward, Stratum 10 (enrollees with no HCCs) is excluded per 90 FR 4424. If a state risk pool has ≤200 HCC-bearing enrollees, all are selected — the full eligible population becomes the IVA sample.
Strata 1–3
Age 0–17
Low / Mid / High RS
Strata 4–6
Age 18–64
Low / Mid / High RS
Strata 7–9
Age 65+
Low / Mid / High RS
Stratum 10 — EXCLUDED BY2025+
Enrollees with no EDGE HCCs · per 90 FR 4424
SVA expansion sequence
100*
initial select
→
24
enrollees
→
50
enrollees
→
100
enrollees
*CMS first selects a 100-enrollee SVA subsample from the IVA-200, representative by stratum. SVA review then proceeds incrementally: 24 → 50 → full 100. Bootstrap pairwise means test (K=10,000, 90% CI) determines if results are accepted at each level. If SVA-100 precision is still poor, CMS expands to the full 200 (the entire IVA sample).
Standard IVA sample
200
If issuer's total eligible population (≥1 HCC) is ≤200, all are selected. Allocation is proportional to stratum size × risk score variability via the Neyman formula.
Error Rate Estimation
How RADV findings translate into payment adjustments
Five steps turn medical record findings into a PLRS adjustment for outlier issuers.
5-step error estimation process
1
Pairwise means testDetermines whether IVA or SVA findings are accepted, based on statistical agreement and precision.
2
Super HCC groupingHCCs sharing the same G-group in adult RA models are aggregated, sorted into 3 national Failure Rate Groups.
3
Outlier identificationIssuer needs ≥30 de-duplicated Super HCCs in a group to qualify as an outlier in that group.
4
EAF calculationGroup Adjustment Factors calculated for outlier issuers using accepted HCC failure rates.
5
PLRS adjustmentHccAdjRS(i,e) = HccEdgeRS(i,e) × (1 − EAF(i,e)) applied to each outlier issuer's HCC-specific risk score.
The adjustment formula
HccAdjRS(i,e)
= HccEdgeRS(i,e)
× (1 − EAF(i,e))
= HccEdgeRS(i,e)
× (1 − EAF(i,e))
i = Failure Rate Group (1–3)
e = Issuer
EAF = Error Adjustment Factor from group failure rate
HccEdgeRS = HCC-specific EDGE risk score contribution
e = Issuer
EAF = Error Adjustment Factor from group failure rate
HccEdgeRS = HCC-specific EDGE risk score contribution
Both positive outliers and negative outliers receive adjustments. Direction of the error rate determines whether PLRS increases or decreases.
Source: BY2025 HHS-RADV Protocols §10.4
Medical Record Standards
What makes a record valid — and what disqualifies it
Each HCC on the RADVDE Report must be supported by a qualifying medical record linked to an RA-eligible claim. The IVA Entity submits records through the Audit Tool; the SVA Entity re-abstracts the same records independently.
✓ Valid record requirements
✗ Cannot substantiate a diagnosis alone
Coding Standards
ICD-10 & diagnosis validation rules
ICD-10 code year
The ICD-10-CM code year applicable to the benefit year's date of service governs validity. Codes must be valid in the code set for the DOS.
HCC model version
BY2025 uses HHS-HCC V08. A diagnosis must map to an active HCC in V08 to contribute to PLRS. Mappings are in the Payment Notice.
Condition must be treated
The record must show the condition was assessed, monitored, or managed during the encounter — not merely listed in history.
Benefit year requirement
Records from outside the benefit year cannot support a RADV HCC. Encounter date must fall within benefit year dates.
Provider credentials
Diagnosing provider must hold credentials authorizing diagnosis in the state of practice. Scope-of-practice mismatches fail the HCC.
Code specificity
ICD-10 codes must be coded to highest available specificity. Unspecified codes where a specific alternative exists are subject to rejection.
Business Rule Repository
Mandatory requirements from CMS guidance
Extracted from BY2025 HHS-RADV Protocols, BY2024 Package submission guidance, EDGE Business Rules, and IRR documentation. Click any rule to expand source citation.
Master Timeline
Key dates and deadlines — BY2024 & BY2025
Apr 30
Annual cutoff
EDGE data submission cutoff — the hard deadline
April 30 is the last date for any claim submission or correction counting toward the RA calculation. After this date, the benefit year's data is locked. Enrollment files use full replacement; claims are incremental. EDGE data must be retained for 10 years (min. 3 years on active server).
BY2024 Benefit Year 2024 — Active
| Date | Milestone | Responsible Party |
|---|---|---|
| Nov 2025 | Audit Tool opens for IVA Audit Results Submission | IVA Entity |
| Jan 8, 2026 | Package 1 submission deadline (8:00 PM ET) | IVA Entity |
| Jan 15, 2026 | Package 2 submission deadline + IRR submission due | IVA Entity |
| Jan–Mar 2026 | SVA review window | SVA Entity |
| Mar 2026 | Summary of Final Pairwise Results released | CMS |
| +15 days | Discrepancy Window 1 (insufficient pairwise agreement only) | Issuers |
| Jun 2026 | Final Results Memo published | CMS |
| +30 days | Error Rate Discrepancy Window 2 | Issuers |
BY2025 Benefit Year 2025 — In Progress
| Date | Milestone | Responsible Party |
|---|---|---|
| Apr 13, 2026 | BY2025 IVA Entity Designation Web Form opens | Issuers |
| Apr 30, 2026 | EDGE data cutoff — last day for claim corrections | Issuers |
| Nov 2026 | Audit Tool opens for IVA Audit Results Submission | IVA Entity |
| Jan 7, 2027 | Package 1 submission deadline (8:00 PM ET) | IVA Entity |
| Jan 14, 2027 | Package 2 submission deadline + IRR submission due | IVA Entity |
| Jan–Mar 2027 | SVA review window | SVA Entity |
| Mar 2027 | Summary of Final Pairwise Results released | CMS |
| +15 days | Discrepancy Window 1 (insufficient pairwise agreement only) | Issuers |
| Jun 2027 (est.) | Final Results Memo (based on BY2024 pattern) | CMS |
| +30 days | Error Rate Discrepancy Window 2 | Issuers |
Est. = estimated based on BY2024 pattern. Source: 2025 BY HHS-RADV Activities Timeline; BY2025 Issuer Participation Slides (Mar 2026).
Reference Glossary
Key terms and acronyms used in HHS-RADV
All terms extracted from BY2025 HHS-RADV Protocols, EDGE Business Rules, IRR guidance, and CMS REGTAP materials.
The Health Data Max Approach
One platform. Continuous audit readiness.
HDM doesn't wait for the RADVIVAS reports to land. EDGE submissions, accepted claims, RA outputs, and source enrollment are reconciled in one database — so every HCC is traceable to the chart that has to support it before CMS samples.
01
Reconcile EDGE data
Enrollment, claims, RA outputs, and RADV reports linked per issuer and benefit year.
→02
Flag audit risk
Mirror Super HCC failure rate signals and D&E/RXC anomaly patterns before the sample drops.
→03
Surface evidence
For each sampled enrollee-HCC, surface the strongest linked medical record and claim chain.
→04
Validate or correct
Confirm defensible HCCs, route unsupported diagnoses for correction — accurate RA transfer.
✓
One source of truthClaims, encounters, RADV reports, risk scores, and charts in a single linked database.
✓
CMS- & EDGE-aligned flaggingSuper HCC failure rate signals and D&E/RXC anomaly patterns run continuously against your data.
✓
Best-chart selection per HCCFor every flagged enrollee, surface the record most likely to support the diagnosis before audit.
✓
Encounter ↔ chart reconciliationEvery submitted code tied back to source documentation — gaps found before CMS does.
✓
Pre-sample risk prioritizationFocus on the HCCs whose failure rate would drive outlier status, before the sample drops.
✓
Clean audit submissionsRADV Audit Tool–ready XML packages with every flag and decision logged and defensible.
Accuracy in both directions. The goal isn't to maximize codes — it's to make every submitted HCC defensible. HDM flags potential over-coding and unsupported diagnoses for correction just as it surfaces supporting evidence, so your risk capture is accurate, compliant, and able to withstand the SVA's independent re-review.
Get Ahead of Your Benefit Year
Be ready before the RADVIVAS lands.
With BY2024 audits active and BY2025 on the calendar, the window to get ahead of your IVA cycle is now.
AgentHDM
Each Agent Specializes in a Different Part of the Risk Adjustment Cycle
HCC Risk Analytics Agent
Risk Intelligence
Detects missing, miscoded, and unsorted HCCs
142 opportunities detected
97.2% accuracy
Open Agent >
ChartCopilot™ Retro
Agent
Retrospective Review
Chart extraction + MEAT → HCC detection
120 opportunities detected
Open Agent >
EncounterFlow™ Agent
Compliance Workflow
Validates encounter submissions end-to-end
352 encounters validated
95.5% accuracy
Open Agent >
ProspectiQ™ Agent
Prospective Analytics
Identifies high-value members & next-best actions
129 high-value members
98.5% accuracy
Open Agent >
3-Pillar Value Proposition
Powered by Agentic AI & Fine-Tuned LLMs
Every Health Data Max solution leverage Agentic AI—a multi-agent intelligence architecture designed to continuously review CMS files, member charts, and claims data. Each product, from chart review to RADV defense, benefits from these agents’ ability to reason, validate, and adapt to changing CMS models.
This ensures unmatched accuracy, auditability, and compliance across your risk adjustment lifecycle.
Chart AI Copilot
Delivering real-time clinical intelligence at the chart level
Leverages proprietary NLP and LLMs to scan structured & unstructured data for HCCs and missed coding opportunities.
Assists coders and QA teams with compliant code suggestions, source-linked references, and confidence scoring.
Supports prospective, retrospective, and concurrent chart reviews.
Tracks every action for audit-readiness and traceability (full audit trail).
Seamlessly integrates with your existing EHR or coding platform.
Smarter Chart Reviews, Faster Decisions
- Scans structured and unstructured clinical data using NLP and fine-tuned LLMs
- Identifies HCCs and missed coding opportunities
- Provides compliant code suggestions with source-linked evidence
- Supports prospective, concurrent, and retrospective reviews
- Maintains a full audit trail
- Integrates seamlessly with EHRs and coding platforms
CMS Submission Compliance
Automated validation and reconciliation across your entire submission pipeline
Ingests and analyses X12 837 I/P/D encounter files.
Monitors TA1, 999, 277CA, MAO-002, MAO-004, MMR, and MOR files for errors and rejection trends.
Real-time dashboards flag rejected or dropped encounters.
Built-in reconciliation tools link rejected claims to charts, ensuring no revenue is lost.
Keeps your plan audit-ready and aligned with CMS requirements.
CMS Submission Compliance
Automated validation and reconciliation across your entire submission pipeline
- Ingests and analyses X12 837 I/P/D encounter files
- Monitors TA1, 999, 277CA, MAO-002, MAO-004, MMR, and MOR files for errors and rejection trends
- Real-time dashboards flag rejected or dropped encounters
- Built-in reconciliation tools link rejected claims to charts
- Keeps your plan audit-ready and aligned with CMS requirements
RAF Optimization
Drive more accurate risk scoring and maximize revenue without over-coding
Identifies undocumented HCCs across clinical records using AI.
Prioritizes high-impact, high-chronic patients to close the right gaps first.
Supports prospective programs that improve performance before the look-back ends.
Reduces risk of over- or under-coding by flagging anomalies with contextual evidence.
Generates predictive RAF impact and submission completeness metrics.
RAF Optimization
Drive more accurate risk scoring and maximize revenue without over-coding
- Identifies undocumented HCCs across clinical records using AI
- Prioritizes high-impact, high-chronic patients to close the right gaps first
- Supports prospective programs before the look-back period ends
- Flags over- and under-coding risks with contextual evidence
- Generates predictive RAF impact and submission completeness metrics
Agentic AI in Healthcare
Transform Risk Adjustment with AI That Actually Works
Cut chart review time by 80% and boost RAF scores helped by AI agents that think, learn, and optimize your entire risk adjustment workflow.
Our Impact
From identifying missed HCCs to validating encounters before submission, AgentHDM agents turn fragmented workflows into a single, intelligent operating layer — reducing rework, improving accuracy, and strengthening compliance.
Each AgentHDM agent is engineered as a purpose-specific decision system — not a generic AI assistant. While agents share a common intelligence layer, their capabilities are tightly scoped to the function they perform, ensuring consistent behavior across analytics, chart review, and submission workflows.
Agents combine deterministic logic (CMS guidance, HCC hierarchies, and submission rules) with statistical models trained on historical claims, encounters, and chart data. This hybrid approach allows the system to produce explainable outputs — showing what was identified, how it was evaluated, and which evidence supports the result.
Rather than acting autonomously on source systems, agents generate structured, traceable outputs designed for downstream use by analysts, coders, and operations teams — enabling review, validation, and confident execution at scale.
AgentHDM agents connect chart review, analytics, and submission into one intelligent operating layer — reducing rework, improving accuracy, and strengthening compliance.
Each agent is purpose-built, not generic. Shared intelligence ensures consistency, while tightly scoped roles keep behavior predictable across workflows.
By combining CMS rules and HCC logic with models trained on historical data, agents deliver explainable, evidence-backed outputs — showing what was found, why it matters, and how it was validated.
Outputs are structured for human review, enabling analysts and coders to act with confidence — not blind automation.
80%
Time reduction
AI-powered chart reviews using your historical Medicare Advantage data run up to 80% faster by auto-summarizing clinical evidence, flagging suspected HCC gaps, and pre-building reviewer-ready findings. Reviewers spend less time searching across claims and encounters and more time validating what matters—driving faster throughput, recapture, and submissions without compromising compliance.
98%
Validation accuracy
AI chart reviews achieve 98% accuracy through cross-validation against Medicare Advantage history. Each diagnosis is validated with MEAT evidence and checked for consistency across prior claims, encounters, and MAO-004/MMR/MOR signals, reducing false positives and rework. Clear, traceable rationales ensure consistent decisions, cleaner coding, and stronger RADV readiness.
24/7
AI processing
Always-on AI chart review processing runs 24/7, including nights and weekends. As charts arrive, the system continuously triages, prioritizes, and routes evidence-backed opportunities with next-best actions. Teams start each day with the highest-impact work already queued—no downtime, no backlog spikes, just continuous progress toward compliant risk capture.
Our Platform
Autonomous, Agentic AI Solutions for the Entire Risk-Adjustment Lifecycle
Intelligent agents now collaborate across member targeting, chart retrieval, coding, submissions, and audits — closing gaps, ensuring compliance, and protecting revenue in real time.
360-DEGREE MEMBER DATA VIEW
VALIDATING SUBMISSIONS AGAINST YOUR CLAIMS AND EHR MEDICAL RECORDS
ADVANCED CLINICAL-AI PLATFORM
BUILD YOUR OWN CUSTOM-TRAINED LARGE LANGUAGE MODEL ON YOUR DATA
PROVEN ENCOUNTER SUBMISSIONS ENGINE
FINANCIAL RECONCILIATION TO MINIMIZE RISK AND MAXIMIZE REVENUE
PREDICTIVE RISK ADJUSTMENT ENGINE
CLOSE PREDICTED GAPS AT POINT OF CARE THROUGH EHR-INTEGRATED PHYSICIAN WORKFLOW
How Our AI Transforms Your Workflow
Our autonomous agents work together to revolutionize every step of your risk adjustment process
Intelligent Ingestion
AI automatically processes medical records, EMR data, and claims across all formats. No manual data entry required.
AI Analysis & Coding
Advanced NLP identifies HCC opportunities with 99.8% accuracy. Our AI learns from every interaction to improve continuously.
Compliance Validation
Built-in RADV audit protection ensures every code meets CMS requirements. Automated compliance checking prevents costly errors.
Optimized Submissions
Streamlined submission process with real-time tracking. Get paid faster with error-free claims and comprehensive audit trails.
Our Solutions
Complete AI-Powered Solutions
Everything you need for end-to-end risk adjustment automation
Health Data Max
Powered by Agentic AI, Health Data Max connects every function—analytics, documentation, validation, and audit—through one autonomous, governed system
Health Data Max
Powered by Agentic AI, Health Data Max connects every function—analytics, documentation, validation, and audit—through one autonomous, governed system
Health Data Max
Powered by Agentic AI, Health Data Max connects every function—analytics, documentation, validation, and audit—through one autonomous, governed system
AgentHDM
Each Agent Specializes in a Different Part of the Risk Adjustment Cycle
HCC Risk Analytics Agent
Risk Intelligence
Detects missing, miscoded, and unsorted HCCs
142 opportunities detected
97.2% accuracy
Open Agent >
ChartCopilot™ Retro
Agent
Retrospective Review
Chart extraction + MEAT → HCC detection
120 opportunities detected
Open Agent >
EncounterFlow™ Agent
Compliance Workflow
Validates encounter submissions end-to-end
352 encounters validated
95.5% accuracy
Open Agent >
ProspectiQ™ Agent
Prospective Analytics
Identifies high-value members & next-best actions
129 high-value members
98.5% accuracy
Open Agent >
AgentHDM
Each Agent Specializes in a Different Part of the Risk Adjustment Cycle
HCC Risk Analytics Agent
Risk Intelligence
Detects missing, miscoded, and unsorted HCCs
142 opportunities detected
97.2% accuracy
Open Agent >
ChartCopilot™ Retro
Agent
Retrospective Review
Chart extraction + MEAT → HCC detection
120 opportunities detected
Open Agent >
EncounterFlow™ Agent
Compliance Workflow
Validates encounter submissions end-to-end
352 encounters validated
95.5% accuracy
Open Agent >
ProspectiQ™ Agent
Prospective Analytics
Identifies high-value members & next-best actions
129 high-value members
98.5% accuracy
Open Agent >
HCC Risk Analytics Agent
Risk IntelligenceDetects missing, miscoded, and unsorted HCCs
97.2% accuracy
ProspectiQ™ Agent
Prospective AnalyticsIdentifies high-value members & next-best actions
98.5% accuracy
ChartCopilot™ Retro Agent
Retrospective ReviewChart extraction + MEAT → HCC detection
120 opportunities
EncounterFlow™ Agent
Compliance WorkflowValidates encounter submissions end-to-end
95.5% accuracy
Agentic AI in Healthcare
Transform Risk Adjustment with AI That Actually Works
Cut chart review time by 80% and boost RAF scores with autonomous AI agents that think, learn, and optimize your entire risk adjustment workflow.
How Our AI Transforms Your Workflow
Our autonomous agents work together to revolutionize every step of your risk adjustment process
Intelligent Ingestion
AI ingests records and claims—no manual work
AI Analysis & Coding
NLP identifies HCCs with 99.8% accuracy
Compliance Validation
RADV-ready compliance checks built in
Optimized Submissions
Faster, cleaner submissions with full audit trails
Live Webinar
Navigating CMS PY2027 Chart Review Linkage Requirements — April 16, 2026 · 2:00 PM ET · Zoom
Register Now →
