Nartis
A First-Principles Guide for Investors

HEDIS, Star Ratings,
and the AI Healthcare
Operating System

Healthcare quality is measured, scored, and paid out in billions of dollars every year. This walkthrough explains exactly how it works — from a single patient's lab result to plan-wide bonus revenue — and why AI transforms quality from a report into a control system.

Section 01
The Foundation

What is healthcare quality measurement?

Health plans are graded on whether their members actually receive good care. Not whether they paid their claims — but whether diabetic patients got their A1c tested, whether women over 50 received mammograms, whether patients filled their prescriptions. Quality is tracked as a percentage: of everyone who should have received care, how many actually did?

Live example — Diabetes A1c Control
Denominator
2,400 members
Adults with Type 2 Diabetes
→
Numerator
1,872 members
A1c below 8.0% this year
÷
0%
Performance Rate
Denominator
Who qualifies for measurement
The eligible population — members who meet the criteria for a given measure (age, diagnosis, enrollment length). This is the "who should be getting this care" group.
Numerator
Who met the standard
The subset of the denominator who actually received the care or met the clinical threshold. Numerator ÷ Denominator = your performance rate.
Every care gap is a patient who fell out of the numerator.
The entire business of healthcare quality improvement is: find those patients, close the gaps, and move the rate.
Section 02
The Standard

What is HEDIS?

HEDIS (Healthcare Effectiveness Data and Information Set) is the standardized rulebook that defines exactly how each quality measure is calculated. Every health plan in the country uses the same rules — who qualifies, what counts as a pass, and how long the measurement window is. Think of it as the official scoring system for healthcare quality.

Explore a measure
HEDIS turns clinical judgment into auditable math.
When a health plan knows exactly what counts as a pass, it can systematically identify every member who hasn't passed — and act before the measurement window closes.
Section 03
The Score

What are Star Ratings?

CMS (the federal agency that runs Medicare) aggregates dozens of HEDIS and Stars measures into a single composite score for each health plan — the Star Rating, from 1 to 5 stars. This isn't just a grade. It's a revenue multiplier. And it doesn't scale evenly — crossing specific thresholds triggers sudden, significant financial changes.

CMS Star Rating Scale — Quality Bonus Tiers
⭐
1.0 ★
No bonus
⭐⭐
2.0 ★
No bonus
⭐⭐⭐
3.0 ★
No bonus
⭐⭐⭐
3.5 ★
Baseline
⭐⭐⭐⭐
4.0 ★
+5%
⭐⭐⭐⭐
4.5 ★
+10%
⭐⭐⭐⭐⭐
5.0 ★
+15%
3.5 ★
Most plans operate here. No bonus. The baseline most plans are fighting to leave.
4.0 ★
The most important threshold in MA. Crossing it unlocks 5% bonus revenue — worth tens of millions.
4.5 ★
Double-bonus tier. Only a small fraction of plans reach this level. Sustained competitive advantage.
Going from 3.5 to 4.0 stars matters more than going from 4.0 to 4.5.
The non-linearity is the key insight. The industry isn't optimizing for better scores — it's optimizing to cross specific thresholds. Every intervention that moves a measure rate above its cut point is directly worth money.
Section 04
The Money

How Star Ratings turn into revenue

Medicare Advantage plans earn revenue per member per month (PMPM). CMS then applies a quality bonus percentage on top of that base revenue for plans rated 4 stars or above. The bonus applies to the entire plan's revenue — not just to the members affected by any individual measure.

Members
50,000
×
PMPM × 12
$14,400
=
Annual Revenue
$720M
×
Bonus %
5%
=
Bonus Revenue
+$36M
Calculate your plan's revenue lift
Total Members 50,000
Avg PMPM Revenue $1,200
Current Star Rating
Target Star Rating
Annual Revenue Base
$720.00M
50,000 members × $1,200 PMPM × 12 months
Current Bonus Revenue
$0
At 3.5 ★ — no bonus tier reached
Revenue Lift at Target Rating
+$36.00M
3.5 ★ → 4.0 ★ · 5% bonus applied to full revenue base
Lift Per Member Per Year
+$720
Bonus revenue distributed across all enrolled members
"A small change in quality can unlock tens of millions in revenue."
And that bonus applies to the entire membership — not just the patients whose care improved. That's why quality improvement isn't just a clinical goal. It's a financial lever.
Section 05
The Problem

Why current systems fail

Legacy healthcare quality platforms are built around a fundamental constraint: data is processed after care happens. By the time a care gap appears on a dashboard, the clinical window may already be closed. A human has to notice, interpret, and act — and at scale, most of those moments never happen.

The legacy quality improvement loop
🏥
Clinical Event
Patient visit
📦
Claims Batch
Days–weeks later
📊
ETL Pipeline
Overnight run
🖥️
Dashboard
Updated weekly
👤
Human Review
If someone looks
⏰
Manual Action
Months later
❌
Missed Outcome
Gap stays open
Batch Processing
Data arrives in weekly or nightly batches. By the time a gap is visible in the system, the clinical event that could close it may have already passed.
Human Dependency
Every gap requires a person to see it, prioritize it, and act on it. At 50,000+ members across 30+ measures, the math doesn't work.
No Feedback Loop
When an action is taken, the system doesn't know if it worked until the next batch. There's no mechanism to learn and improve in real time.
Section 06
The Solution

The AI-native closed loop

An AI-native quality operating system replaces the broken sequence with a continuous feedback loop. Every clinical event is evaluated in real time against policy. AI scores gaps and prioritizes action. The system learns from every outcome. The loop never stops.

⚡
Event
📋
Policy
🤖
AI
🎯
Action
📈
Outcome
🔄
Feedback
⚡
Event
A clinical event occurs — a lab result, a prescription fill, a missed appointment. The system ingests it in real time.
📋
Policy
HEDIS rules run as code. The system evaluates whether this event satisfies the measure's denominator and numerator criteria — instantly.
🤖
AI Analysis
AI scores the detected gap by risk, closure probability, Stars impact, and financial value. Patients are ranked by where intervention matters most.
🎯
Action
The right intervention is triggered automatically — provider alert, patient outreach, scheduling, pharmacy notification — without human triage.
📈
Outcome
The result is captured. Did the gap close? Did the measure rate improve? Did Stars performance move? Every outcome feeds back into the model.
🔄
Feedback
The system learns which interventions work for which patients in which contexts. Every cycle makes the next one faster, more accurate, and more impactful.
This isn't a faster dashboard. It's a fundamentally different system.
Legacy platforms are designed to report what happened. This platform is designed to change what happens next — continuously, at scale, across every patient and every measure simultaneously.
Section 07
The Difference

Why this changes everything

The shift from legacy analytics to an AI-native operating system isn't an upgrade — it's a change in what kind of system it is. One generates reports. The other generates outcomes.

Legacy Quality Analytics
Batch processing — data available days or weeks after events
Retrospective reporting — shows what already happened
Human-dependent review — gaps are found only if someone looks
Static measure logic embedded in ETL pipelines
Annual HEDIS ruleset updates require months of engineering
Dashboard alerts routed to care managers for manual follow-up
No learning — system performs the same regardless of outcomes
AI-Native Operating System
Real-time streaming — evaluated at the moment of the clinical event
Proactive intervention — acts before the window closes
AI-driven prioritization — every gap scored and ranked automatically
Policy-as-code — HEDIS rules versioned, tested, instantly updatable
Ruleset updates deploy in hours with full audit trail
Direct workflow integration — care teams receive ranked action queues
Continuous learning — outcomes improve the next intervention cycle
Section 08
The Scale

From one patient to millions in revenue

This is the chain that matters to investors. A single clinical action — one lab ordered, one prescription filled, one outreach call that works — is not just a better health outcome. It is a contribution to a measure rate that can move a Star threshold that unlocks plan-wide bonus revenue. The leverage is extraordinary.

🩺
Patient
One gap closed
A diabetic patient gets their A1c tested. An adherence gap is closed. One numerator event recorded.
→
📐
Measure Rate
Rate improves
One more compliant patient moves the plan's A1c control rate from, say, 80.2% toward the 81.0% cut point.
→
⭐
Star Rating
Threshold crossed
Aggregate weighted measure scores push the plan's composite rating from 3.5 to 4.0 stars — the critical bonus tier.
→
💰
Plan Revenue
+$36M
CMS applies the 5% quality bonus to the full plan revenue base — not just the patients affected by the measure.
"A small clinical action can scale into plan-wide financial impact."
The financial leverage is asymmetric. A single intervention costs hundreds of dollars. The Stars lift it contributes to can be worth thousands of dollars per member across the entire plan. That gap is where the operating system earns its value.
Section 09
The Proof

See the system working live

The Nartis prototype demonstrates every component of this operating system with synthetic patients, real HEDIS measure logic, AI gap detection, and a live Stars revenue simulator. You can evaluate any patient against all 34 measures, see the AI reasoning, simulate interventions, and watch the Stars impact calculate in real time.

🧬
Synthetic Patient Population
100+ synthetic patients with realistic diagnoses, labs, medications, and screenings — representing the breadth of a real Medicare Advantage panel.
📋
34 HEDIS/Stars Measures
Every major HEDIS and Stars measure evaluated in real time using policy-as-code — denominator/numerator logic, thresholds, and reasoning.
🤖
AI Gap Detection & Ranking
AI risk scores, gap severity, closure probability, and Stars lift potential calculated for every detected gap across the full patient population.
🎯
Intervention Simulation
Simulate clinical actions — lab orders, outreach, prescription fills — and watch the Stars impact update in real time as gaps close.
💰
Stars Revenue Calculator
Model the financial impact of improving Star ratings across any plan size — from members to PMPM to bonus revenue lift.
🔒
Full Audit Trail
Every AI decision, policy evaluation, and action is logged with user identity, agent identity, data accessed, and clinical rationale.
Run the Live Prototype → Download Architecture Brief
Built for
Medicare Advantage plans, HEDIS reporting organizations, and quality analytics teams operating at scale.
Nartis
Quality. Controlled.