Risk Adjustment is one of the most misunderstood engines in healthcare — but it’s also one of the most important.
It keeps the system fair.
It keeps funding aligned with real medical needs.
And starting with the V28 model update — it’s smarter, cleaner, and closer to real-world medicine than ever before.
Let’s make sense of it.
Plain English. No equations. Zero headaches.
Why Risk Adjustment Exists (And Why It Will Never Go Away)
If Medicare paid every beneficiary, the same amount:
Healthy members would be overpaid
Complex members would be underfunded
Health plans would avoid the sickest patients
Risk Adjustment solves this problem by ensuring:
Every patient’s story is represented in the dollars used to care for them.
It protects access.
It protects fairness.
It protects populations often left behind.
That’s why CMS keeps enhancing the model — better coding shouldn’t inflate risk, and poor coding shouldn’t hide real disease burden.
What Exactly Is an HCC?
Think of ICD-10 codes as words.
They tell part of a story — but sometimes in confusing ways.
HCCs are like organized chapters:
They group similar diagnoses together
They help CMS pay based on clinical significance
They prevent double-paying for minor vs. major versions of a disease
Example idea (without specifics):
A diagnosis of seasonal allergies ≠ the same as chronic respiratory disease.
HCCs make sure payment reflects:
Severity
Clinical complexity
True resource needs
And because they’re hierarchical, the most serious active condition counts.
V28 — A Smarter Model for a Modern Era
CMS updates the model when:
Care patterns evolve
Coding changes
New technologies improve data integrity
Some conditions need more/less emphasis
V28 brings 4 important improvements:
More clinically accurate groupings
Updated mapping from ICD-10 → HCCs
Less “gaming” from borderline documentation
Better fairness across diverse populations
Bottom line:
V28 rewards quality and completeness instead of volume or loopholes.
How a RAF Is Built — The 5-Piece Puzzle
A finalized risk score is built from:
1. Demographics
Age, gender, Medicaid status, disability…
→ Sets a baseline of expected cost
2. Condition Categories (HCCs)
Documented & properly coded illnesses
→ Increase expected costs proportionally
3. Interactions
Certain combinations are clinically harder to manage
→ Adds nuance to reflect real-world care
4. Count + Hierarchy Logic
Avoids overpayment for multiple related diagnoses
→ Removes “double counting”
5. Adjustments for fairness
Ensures national costs remain stable
→ Keeps MA vs. FFS competitive balance intact
No one step alone is enough — accuracy depends on the entire chain.
Where Risk Adjustment Falls Apart
Even if a patient is complex and well-managed, their RAF may drop because of data issues.
Here’s where that happens most:
-> Chart → Code gap
Diagnosis documented but never coded = gone from the model
-> Code → Submission gap
Claim/encounter rejection = CMS doesn’t see it
-> Documentation → Validation gap
Code submitted but not supported by MEAT evidence = RADV risk
-> Yearly continuity gap
Chronic conditions not refreshed = wiped from current-year risk
Patients don’t suddenly get healthier…
But their data might say they do.
This creates payment leakage — and compliance exposure.
What Good Looks Like — Modern Risk Adjustment Standards
A strong program has:
Clear provider documentation standards
Automated data lineage tracking from chart → CMS
Internal RASS “mirroring” before CMS runs
Consistent refresh of chronic illness data
Proactive gap closure campaigns
Organizational visibility into RAF performance
This is where technology becomes a competitive advantage.
Enter Agentic AI — The New Teammate in Risk Adjustment
Agentic AI doesn’t replace staff — it extends them.
It can:
Scan millions of clinical notes for missing evidence
Flag documentation at risk of audit
Detect underreported chronic conditions
Compare internal risk vs. expected CMS results
Run year-round audit simulations
Provide clinical context, not just code checks
And importantly:
It doesn’t change diagnoses — it protects them.
AI keeps the data trail clean so care is accurately represented.
A Day in the Life of AI-Enhanced Risk Adjustment
Without AI:
Teams react to errors months later
Spreadsheets everywhere
Surprise payment variances
Scrambling during RADV cycles
With Agentic AI:
Daily documentation scoring
Real-time RAF visibility
Fewer provider touchpoints
No lost encounters
Confident compliance year-round
Plans stop playing catch-up and start shaping outcomes.
The Human Role: Still the Most Important
AI is powerful at:
Pattern recognition
Volume review
Rule enforcement
But humans excel at:
Clinical judgment
Final validation
Provider relationships
Complex medical nuance
The magic is in their partnership.
Think of AI as:
The detective
The human reviewer is the judge
Together → precision you can trust.
Final Takeaways (No Calculators Required)
Risk Adjustment ensures fairness in healthcare funding
V28 improves accuracy and clinical realism
Data continuity is what protects true patient complexity
AI keeps the entire chain clean, complete, and defensible
Humans remain the ultimate decision makers
Accurate data = accurate care = accurate payment
That’s risk adjustment done right.
Let’s turn accuracy into confidence.
And confidence into outcomes.
