Risk Adjustment Coding Services (HCC & RAF)
A missed chronic condition or an unsupported HCC doesn't just change one claim -- it changes what a Medicare Advantage plan is paid for that member all year, and what CMS can claw back on audit. MedPrecision captures HCCs prospectively and retrospectively under the current CMS-HCC V28 model, validates every code to the MEAT standard, and keeps the data defensible against RADV audits, which CMS announced in 2025 it will extend to every eligible Medicare Advantage contract.
- Free audit
- No contract
- HIPAA-secure
- Cancel anytime
What Is Risk Adjustment (HCC) Coding?
Risk adjustment coding is the capture and validation of the chronic and severe diagnoses that set a Medicare Advantage plan's payment under the CMS-HCC model. CMS is required by Section 1853(a)(1)(C) of the Social Security Act to risk-adjust the monthly per-member payment it makes to Medicare Advantage Organizations based on enrollee health status; the CMS-HCC (Hierarchical Condition Category) model converts documented diagnoses into a Risk Adjustment Factor (RAF) score. Only diagnoses that map to an HCC affect the score, each must be documented to the MEAT standard (Monitor, Evaluate, Assess/Address, Treat), and each chronic condition must be recaptured every year. As of 2026 the model runs entirely on Version 28 (V28) for non-PACE organizations; V28 is the first CMS-HCC model built on ICD-10-CM-to-HCC mappings.
- Prospective and retrospective HCC capture so chronic conditions are recaptured and documented every year
- Every HCC validated to the MEAT standard -- a bare diagnosis list is not valid for risk adjustment
- Current CMS-HCC V28 (ICD-10-to-HCC) mappings, not retired V24 logic
- RADV audit support: records pulled and organized for each sampled HCC
Risk adjustment coding is not diagnosis coding under a different name -- it is a payment-model discipline where a missed or unsupported condition changes what a Medicare Advantage plan is paid for a member across an entire year. Under Section 1853(a)(1)(C) of the Social Security Act, CMS must risk-adjust the monthly per-member payments it makes to Medicare Advantage Organizations to account for enrollees' health status, and it does that through the CMS-HCC (Hierarchical Condition Category) model, which converts documented diagnoses into a Risk Adjustment Factor (RAF) score. MedPrecision's risk adjustment coding team captures and validates HCCs from the chart -- prospectively at the point of care and retrospectively after the encounter -- so the RAF score reflects the member's true burden of illness and every condition is supported to the MEAT documentation standard if CMS pulls the chart in a RADV audit. As of 2026 the model runs entirely on Version 28 for non-PACE organizations, and V28 is the first CMS-HCC model built on ICD-10-CM-to-HCC mappings, so capture built on the retired Version 24 logic quietly loses -- and sometimes overstates -- RAF. We code to what the record proves, delete unsupported HCCs as readily as we recapture supported ones, and organize the documentation behind each code for the RADV audit expansion CMS announced in 2025.
Who This Service Is For
The State of Risk-Adjustment (HCC) Coding Services in 2026
Risk adjustment now governs the majority of Medicare. Per KFF, 55% of eligible Medicare beneficiaries -- 35.2 million of 64.2 million with both Part A and Part B -- are enrolled in Medicare Advantage in 2026, up from 19% in 2007 (read 2 October 2026), and those members' payments are risk-adjusted under CMS's HCC models. CMS is required to risk-adjust those payments under Section 1853(a)(1)(C) of the Social Security Act, and the model itself just went through its biggest change in years: the 2024 CMS-HCC model (Version 28) phased in over three years and runs entirely on V28 in 2026 for non-PACE organizations. Per AAPC, V28 is the first model mapped from ICD-10-CM rather than ICD-9-era classifications, and CMS reports that 2,236 diagnosis codes that used to map to a payment HCC no longer do while 209 were added (CY 2024 Rate Announcement). At the same time, enforcement is being scaled up: CMS said in May 2025 it will audit all ~550 eligible MA contracts annually (up from ~60) in newly initiated audits and grow its coders from 40 to roughly 2,000. The legal footing for extrapolating RADV findings is contested: a federal district court vacated the 2023 RADV final rule on 25 September 2025 and CMS has appealed. That same announcement puts federal estimates of MA overbilling at roughly $17 billion a year, alongside MedPAC's estimate that the figure could be as high as $43 billion (read 2 October 2026). The model now rewards only capture that is both complete and defensible -- which is the whole job of risk adjustment coding.
What Is Breaking Right Now
Chronic HCCs that dropped off the RAF score because they were never recaptured and documented in the current year
RAF built on retired V24 mappings that no longer score -- or now overstate -- under the V28 model
Unsupported HCCs sitting in submitted data as RADV audit exposure rather than real revenue
Diagnoses documented as a bare list that fail the MEAT standard and won't survive an audit
Disease-interaction terms and hierarchy logic misapplied, understating or overstating the member's score
Common Risk-Adjustment (HCC) Coding Services Mistakes to Avoid
Coding risk-adjusted diagnoses off a problem list instead of MEAT-supported documentation
Per AAPC, a bare list of diagnoses does not meet the definition of an assessment and plan, and RADV audits check that each sampled HCC is supported in the medical record.
Validate every HCC for Monitoring, Evaluating, Assessing/Addressing, or Treating in the note, and query the provider or drop the code when the support isn't there.
Running capture on retired V24 mappings after V28 took over
V28 removed 2,236 diagnosis codes that used to map to a payment HCC and added 209, so V24-based workflows lose legitimate RAF on remapped conditions and can overstate others.
Remap on the current Version 28 ICD-10-to-HCC logic -- fully in effect for non-PACE organizations in 2026 -- and recheck conditions that only score, or no longer score, under V28.
Chasing RAF-adds while leaving unsupported HCCs in the submitted data
Risk-adjusted payments must be supported in the medical record; an unsupported HCC isn't revenue, it's audit exposure under a program CMS announced it would expand from ~60 plans a year to all ~550 eligible plans.
Work capture bidirectionally -- delete HCCs the record won't support as readily as recapture the ones it will -- so the RAF reflects the chart, not a target.
Ignoring the HCC hierarchy and disease-interaction terms
Interaction terms are additive and count independently of the hierarchy, so missing them understates the score, while assuming an old interaction still applies (V28 removed the immune-disorder/cancer pairing) overstates it.
Apply the current V28 hierarchy and interaction logic per member, validating both which HCC trumps within a group and which interactions still count.
What We Handle
Prospective (Concurrent) Chart Review
We review the chart before or at the point of care so suspected and known chronic conditions are addressed and documented in the current year. Because a diagnosis only counts toward the RAF score in the year it is documented and supported, concurrent review is how chronic HCCs get recaptured annually instead of silently dropping off the score.
Retrospective Chart Review & HCC Recapture
After the encounter, we mine completed documentation for HCCs that were supported in the record but never coded to a claim, and for chronic conditions captured last year but not recaptured this year. Only diagnoses that map to an HCC affect the score, so we target the conditions that actually move RAF -- not every ICD-10 code on the note.
HCC Gap & RAF Opportunity Analysis
We reconcile a member's coded HCCs against prior-year history, problem lists, medications, and lab results to surface gaps -- conditions with clinical evidence but no supporting documentation this year. Each gap becomes a suspected HCC routed back to the provider to confirm and document, or closed if the evidence isn't there.
RADV Audit Support & Defense
When CMS selects a contract for a Risk Adjustment Data Validation (RADV) audit, we pull and review the records that must support each sampled HCC, identify diagnoses that won't hold, and organize the documentation package. CMS's 2023 RADV final rule (CMS-4185-F2) would extrapolate sample findings across the contract from payment year 2018 with no Fee-for-Service Adjuster, but a federal court vacated that rule on 25 September 2025 and CMS has appealed -- so how far one unsupported HCC extrapolates is unsettled, though it is removed on review either way.
Provider Documentation Improvement (MEAT/CDI)
We coach providers to document each risk-adjusted condition to the MEAT standard -- Monitor, Evaluate, Assess/Address, Treat -- because per AAPC guidance a bare list of diagnoses is not a valid assessment and plan and will not support the code. Better notes at the source mean fewer queries, higher recapture, and audit-ready charts.
Second-Level Coding QA & Blind Review
Every coder's work is checked against the ICD-10-CM Official Guidelines and the current V28 ICD-10-to-HCC mappings before submission, with high-risk conditions dual-coded. This is the control that keeps accuracy up and keeps aggressive, unsupported capture out of the data that goes to the plan.
Compliance-First, Bidirectional Capture
Accurate risk adjustment coding means deleting unsupported HCCs as readily as adding supported ones. We code to what the record proves -- not to a target RAF -- which is a posture that survives the expanded RADV auditing CMS announced in 2025, where risk-adjusted payments must be supported in the enrollee's medical record.
Free Billing Audit · No obligation
See What Your RAF Is Leaving on the Table -- or Exposing
Send us a sample of your Medicare Advantage charts. We'll show you the supportable HCCs that were never captured, the conditions still coded on retired V24 logic, and the unsupported codes that would extrapolate against you in a RADV audit.
Prefer to talk? Book a 15-minute callRequest received
A billing specialist will review your practice details and reach out within 1 business day. No confirmation email is sent — if you need to reach us sooner, call +1-872-297-2815.
Our Risk-Adjustment (HCC) Coding Services Methodology
Bidirectional, Not Add-Only
We add supported HCCs and delete unsupported ones. Because risk-adjusted payments must be supported in the medical record, an unsupported HCC is audit exposure, not revenue. Coding to what the record proves is a durable posture.
MEAT-Standard Validation
Every risk-adjusted diagnosis is validated for Monitoring, Evaluating, Assessing/Addressing, and Treating in the note. Per AAPC, a bare diagnosis list is not a valid assessment and plan, so codes without MEAT support are queried to the provider or dropped before submission.
Current V28 Model Application
We map diagnoses on the 2024 CMS-HCC model (V28), the first built on ICD-10-CM-to-HCC mappings. Because V28 removed 2,236 diagnosis codes that used to map to a payment HCC and added 209, capture built on V24 logic quietly loses -- and sometimes overstates -- RAF. We remap to what actually scores now.
Prospective + Retrospective Recapture
A chronic HCC only counts in the year it is documented, so we work concurrently to get conditions addressed at the visit and retrospectively to recapture what was supported but never coded. The gap between last year's HCCs and this year's is a common source of lost, legitimate RAF.
Hierarchy & Interaction Discipline
We apply the CMS-HCC hierarchy (a severe HCC trumps a milder one in the same group) and the additive disease-interaction terms that count independently of it. Per AAPC, V28 revised a few interactions -- including removing the immune-disorders (HCC 47) and cancer (HCC 8-12) pairing -- so we validate each member against current V28 logic.
Risk-Adjustment (HCC) Coding Services: MedPrecision vs Alternatives
| Feature | MedPrecision | In-House | Other Providers |
|---|---|---|---|
| HCC Capture | Prospective and retrospective review so chronic conditions are recaptured and documented every year, with suspected-HCC gaps routed to providers | Capture limited to whatever reaches the claim; chronic conditions silently drop off the RAF score year to year | Retrospective capture only, with little concurrent or gap-closure work |
| Documentation Standard | Every HCC validated to the MEAT standard, with provider queries when the note won't support the code | Codes pulled from diagnosis lists that don't meet the assessment-and-plan requirement | Coded to the record as written, with limited MEAT validation |
| Model Currency | Coded on the current CMS-HCC V28 ICD-10-to-HCC mappings, catching conditions that only score under V28 | Legacy workflows still mapping on retired V24 logic, losing RAF on remapped conditions | Partial V28 adoption without a full remap review |
| RADV Audit Readiness | Bidirectional review removes unsupported HCCs; documentation organized so sampled HCCs hold up on RADV review | Unsupported HCCs left in the data as audit exposure | Add-focused capture with little audit-defense support |
| Quality Control | Second-level QA and dual-coding on high-risk conditions against ICD-10 guidelines and V28 mappings | Single-pass coding with no blind review | Sampling-based QA without condition-level dual coding |
How the Transition Works
How we deliver risk-adjustment (hcc) coding services for your practice.
Chart, Claims & History Intake
We ingest the current encounter documentation alongside prior-year HCCs, active problem lists, medication lists, and lab results -- everything needed to see which chronic conditions should recur this year and which suspected conditions have clinical evidence but no diagnosis captured on the record.
HCC Capture & RAF Coding on V28
Diagnoses are coded against the ICD-10-CM Official Guidelines and mapped to HCCs using the current Version 28 model, not the retired V24 logic. We capture every condition the record supports -- including those that only score under V28's revised mappings -- and calculate the RAF impact, hierarchy, and interaction terms.
MEAT Validation & Second-Level QA
Each captured HCC is checked for MEAT support in the note; anything without Monitoring, Evaluating, Assessing, or Treating documentation is queried to the provider or dropped. High-risk conditions are dual-coded, so the data that reaches the plan is both complete and supportable.
Recapture, Submission Support & RADV Readiness
We flag chronic HCCs that weren't recaptured this year, route documentation gaps back to providers, and organize the medical-record support behind each code -- so that if CMS pulls the chart in a RADV audit, the diagnosis holds.
What Reporting and Visibility Looks Like
Transparency is built into every engagement. You will always know where your revenue stands and what actions are being taken on your behalf.
Monthly KPI Dashboards
Track collection rates, denial trends, days in A/R, and payer-level performance with dashboards delivered on a fixed schedule.
Real-Time Claim Tracking
See claim status updates in real time so you never have to wonder where a payment stands or when follow-up is happening.
Quarterly Business Reviews
Detailed reviews with actionable recommendations covering denial root causes, payer trends, and revenue recovery opportunities.
Proactive Alerts
Automated alerts when key metrics shift, so issues are caught and addressed before they affect your bottom line.
Risk-Adjustment (HCC) Coding Services Key Terms
- CMS-HCC Model
- The Hierarchical Condition Category model CMS uses to risk-adjust Medicare Advantage payments. It maps documented diagnoses to condition categories, each carrying a coefficient, that sum into a member's risk score. The current version, V28, is the first built on ICD-10-CM-to-HCC mappings and is fully in effect for non-PACE organizations in 2026.
- Risk Adjustment Factor (RAF) Score
- The numeric score, built from a member's demographics and captured HCCs, that scales the payment a Medicare Advantage plan receives for that member. A higher supported RAF reflects a sicker member and a higher payment; an unsupported RAF is what a RADV audit removes.
- MEAT Documentation Standard
- The support a record must show for each risk-adjusted diagnosis: Monitoring, Evaluating, Assessing/Addressing, and Treating. Per AAPC, a diagnosis listed without MEAT does not meet the definition of an assessment and plan and is not valid for risk adjustment.
- RADV Audit
- Risk Adjustment Data Validation -- CMS's main tool for recovering improper risk-adjustment payments by verifying that sampled HCCs are supported in the medical record. The 2023 final rule (CMS-4185-F2) provided for extrapolating findings across the contract from payment year 2018 with no Fee-for-Service Adjuster; a federal district court vacated it on 25 September 2025 and CMS has appealed.
- HCC Hierarchy & Disease Interactions
- Within a disease group, a more severe HCC trumps a less severe one so only the highest counts (the hierarchy). Separately, disease-interaction terms add to the score and apply independently of the hierarchy -- an interaction still counts even if one paired HCC is trumped.
Common Questions
Common questions about risk-adjustment (hcc) coding services.
Get a Free Billing Audit
See where denials, follow-up delays, or workflow gaps may be hurting your collections.
Get a Free Billing AuditWhat is risk adjustment coding, and how does it differ from regular medical coding?
Risk adjustment coding captures the chronic and severe diagnoses that determine a Medicare Advantage plan's payment, rather than the procedure codes that drive a fee-for-service claim. Under Section 1853(a)(1)(C) of the Social Security Act, CMS risk-adjusts the monthly per-member payment it makes to Medicare Advantage Organizations based on enrollee health status, using the CMS-HCC model to turn documented diagnoses into a Risk Adjustment Factor (RAF) score. Only diagnoses that map to a Hierarchical Condition Category (HCC) affect that score, and per AAPC's risk adjustment guidance a simple list of diagnoses is not acceptable -- each condition must be documented to the MEAT standard. So the discipline is about accurate, supportable capture of the member's full burden of illness every year, not maximizing units billed.
What changed with the CMS-HCC V28 model?
CMS finalized the 2024 CMS-HCC model, Version 28, in the CY 2024 Rate Announcement and phased it in over three years -- per the CY 2025 Rate Announcement, CY 2025 blended 67% of the V28 risk score with 33% of the older 2020 model (V24), and per the CY 2026 Rate Announcement the model runs entirely on V28 in 2026 for non-PACE organizations. According to AAPC, V28 is the first CMS-HCC model built on ICD-10-CM-to-HCC mappings rather than the old ICD-9-based classifications; it renames and renumbers HCCs, remaps diagnoses, and changes the RAF coefficient values. Per CMS, the model has 115 payment HCCs versus 86 in the 2020 model, 2,236 diagnosis codes that used to map to a payment HCC no longer do, and 209 were added. It is also recalibrated on newer data -- 2018 diagnoses and 2019 expenditures, versus 2014 and 2015 for V24. The practical effect: conditions that scored under V24 may no longer score, so capture built on the old mappings quietly loses RAF.
What is the MEAT documentation standard?
MEAT is the framework used to decide whether a diagnosis is supported: the record must show Monitoring (signs, symptoms, disease progression or regression), Evaluating (test results, medication effectiveness, response to treatment), Assessing or Addressing (ordered tests, discussion, review of records, counseling), and Treating (medications, therapies, or other modalities) for each risk-adjusted condition. Per AAPC, a diagnosis listed without this support does not meet the definition of an assessment and plan and is not valid for risk adjustment. We code to MEAT and query providers when the note falls short, because an HCC that isn't MEAT-supported is exactly what a RADV auditor removes.
How aggressive are CMS's RADV audits now?
CMS has announced a much larger program, and its legal footing is contested. CMS said in May 2025 that it will audit every eligible Medicare Advantage contract for each payment year in newly initiated audits -- moving from about 60 plans a year to approximately 550 -- and increase its medical coders from 40 to approximately 2,000 by September 1, 2025, auditing between 35 and 200 records per plan by size instead of 35. The 2023 RADV final rule (CMS-4185-F2) would extrapolate findings across the contract from payment year 2018 with no Fee-for-Service Adjuster, but a federal district court vacated that rule on 25 September 2025 (Humana Inc. v. Becerra, N.D. Tex.) and CMS has appealed to the Fifth Circuit. CMS's completed audits for payment years 2011-2013 found between 5% and 8% in overpayments, which is why capture has to be limited to what the record supports.
Do you delete diagnoses, or just add them?
Both. Accurate risk adjustment means the score reflects what the record supports -- so we remove HCCs that aren't documented to standard as readily as we recapture ones that are. Because RADV audits exist to confirm that the diagnoses used for payment are supported by the medical record, an unsupported HCC left in the data isn't extra revenue; it's audit exposure. Bidirectional review -- adding supported conditions and deleting unsupported ones -- is a posture that holds up under the expanded RADV program CMS announced in 2025.
How do disease interactions and the HCC hierarchy affect the score?
The CMS-HCC model applies a hierarchy -- within a disease group, a more severe HCC 'trumps' a less severe one so only the highest counts. But disease-interaction terms are additive and apply independently of that hierarchy, so a documented interaction between two conditions still adds to the RAF even when one of the paired HCCs is trumped by a more severe condition. Per AAPC, V28 changed a few of these -- for example, removing the interaction between immune disorders (HCC 47) and cancer (HCC 8-12). Coders who don't account for both the hierarchy and the interaction terms either overstate or understate the score, which is why we validate every member against the current V28 logic rather than habit.
What is a RAF (Risk Adjustment Factor) score and how is it calculated?
A Risk Adjustment Factor (RAF) score is the numeric value assigned to a Medicare Advantage enrollee, based on demographics and diagnoses, that scales the plan's payment for that member. Per AAPC, the score resets every January 1 to a demographic base -- age, sex, socioeconomic status, disability status, Medicaid eligibility, and institutional status -- and each captured HCC then adds its coefficient. The model is additive: HCC values sum into the total, but a condition counts only once no matter how many ICD-10-CM codes map to it, and a more severe HCC trumps a milder one in the same family. CMS, not the coder, performs the official calculation, so a chronic condition not recaptured this year lowers the score and the payment.
How does the CMS-HCC V28 model change risk-adjustment coding?
It shifts risk-adjustment coding toward ICD-10-CM specificity while removing the specificity 'reward' on some conditions. Per AAPC, V28 is the first CMS-HCC model based on ICD-10-CM-to-HCC mappings; it expands and renumbers the categories -- 115 payment HCCs versus 86 under V24, per CMS -- so the code and category numbers used under V24 no longer line up. At the same time, V28 'constrains' conditions like diabetes and heart failure: per CMS, the three diabetes payment HCCs carry the same coefficient, so moving from diabetes without complication to diabetes with complications does not by itself change the score. The practical effect is that coders must document to full ICD-10 specificity, re-map every condition to V28, and stop assuming old crosswalks still score.
Related Services
Specialties We Serve
Related Resources
Available In
See What Your RAF Is Leaving on the Table -- or Exposing
Send us a sample of your Medicare Advantage charts. We'll show you the supportable HCCs that were never captured, the conditions still coded on retired V24 logic, and the unsupported codes that would extrapolate against you in a RADV audit.
- No contract
- No setup fees
- Reply within 1 business day