Value-Based Care

Risk that reflects reality.

OptimIQ reads every encounter and turns it into defensible risk and quality evidence — so your RAF reflects the patients you actually treat, and every code holds up the day the audit arrives.

Value-Based Care · Risk Capture · Panel 4,812

CMS-HCC V28 · CY2025 · refreshed 14:02
0.02.5
RAF · CMS-HCC V281.847+0.214captured vs. submitted
Traced to source3 of 3 HCCs
  • E11.22Diabetes with CKDnote + labs
  • I50.32Chronic diastolic heart failurenote + echo
  • N18.31CKD stage 3anote + eGFR
Works with the stack you already run
  • Epic
  • Oracle Health
  • athenahealth
  • Veradigm
  • eClinicalWorks
  • HL7 / FHIR
  • CMS-HCC V24 + V28
  • HEDIS MY2025
  • CMS Stars
  • RADV
  • X12 837 / 835
The Operating Layer

Revenue and care, measured the same way.

Four jobs that usually live in four tools. OptimIQ runs them on one set of facts, so nothing drifts between them.

Risk01
1.84panel RAF · V28

Suspected and recapture HCCs, evidenced from the chart.

Quality02
4.5Stars composite

HEDIS and Stars gaps beside the risk picture.

Defense03
100%codes traced

Every code linked to the sentence that supports it.

Economics04
+$2.4kper member · modeled

Capitation and shared savings, modeled live.

OptimIQ · The Operating Layer
One set of facts
Progress notesLabsClaimsProblem listsEncountersADT feeds
How it works

The anatomy of a value-based signal.

Every suspected condition and open gap arrives with the evidence to act on it — and the trail to defend it later.

Panel signals12 open · 3 ready
  • Suspected HCC · E11.22Evidence ready0.91
  • Care gap · Eye examOpen612
  • RADV flag · I50.32Documented

Surfaced from the chart

Notes, labs, and feeds, read continuously.

Audit-ready by default

Every documented code carries its MEAT trail.

Only what's provable

Suspected codes wait until the chart supports them.

One click to close

Draft orders sit in the same row as the gap.

Take it even further with…
Live RAF modeling

Scores recompute the moment new evidence lands — no quarter-end surprises.

Quality in the same view

HEDIS and Stars gaps tracked right beside the risk picture.

Risk Adjustment

Code what the chart can prove.

OptimIQ reads notes, labs, problem lists, and feeds to surface suspected HCCs — then holds each one until the documentation supports it. What reaches your RAF is what an auditor can verify.

From the chart
  • Progress note03/14 · endocrinology
  • Lab resultA1c 9.2% · 03/12
  • Problem list12 active conditions
  • Echo reportEF 38% · cardiology
OptimIQ surfaces & supports
Suspected HCCsCMS-HCC V28
E11.22Diabetes w/ CKDsupported
0.91
I50.32Diastolic HFsupported
0.88
N18.31CKD stage 3aneeds note
0.62
Specificity checkedHierarchy prunedFY matched
Defensible RAF
Final score

1.847

+0.214vs. submitted

  • Every code traced to source
  • RADV-ready packet on file
The Workspace

The whole panel, on one screen.

Risk, quality, and economics in one view — every number traceable to the chart underneath, and every gap one click from closing.

Panel overview9,420 attributed lives · CY2025
Q2 · Apr–Jun
RiskQualityAttribution
Panel RAF1.84+0.21 vs. submitted
Recapture92%+6 pts YoY
Stars4.5composite
PMPM$1,043risk-adjusted
RAF recapture · monthlytarget 1.85
Open HCC opportunityRAF impact
E11.65Diabetes w/ hyperglycemiarecapturehigh
I12.9Hypertensive CKDsuspectmed
J44.1COPD exacerbationrecapturemed
OptimIQ Agentjust now

14 care gaps are ready to close before today's visits — drafts attached to each chart.

Care gaps · today8 scheduled
ROMRN 4471Eye exam · A1c control2
MDMRN 5520Statin therapy1
LKMRN 6018BP control · Breast screen2
Spend vs. benchmark · PMPMYTD
Actual$1,043Benchmark$1,118

$75 PMPM under benchmark · tracking to shared savings

Suspect, code, and defend — in one platform.

See it on your charts

OptimIQ reads the chart, suspects HCCs with the evidence already attached, computes RAF under the current CMS models, and keeps every code audit-ready — with your coders in control at every gate.

Physician reviewing a patient chart at a workstation
E11.22Diabetes w/ CKDMEAT 4/4
confidence 0.91Approve
Human-in-the-loop coding

The engine drafts codes with confidence scores; your coders approve through QA, SME, and audit gates before anything is submitted.

Clinician examining data in a medical report

“…A1c 9.2%, up from 8.1%— metformin continued, dose adjusted…”

Progress note · p.3DOS 03/14/2025E11.22
Sentence-level evidence

Every code links to the exact sentence, page, and date of service that supports it — affirmed, current, and about your patient.

CMS-HCC V24 + V28

Hierarchy pruning, interaction scoring, and ESRD and RxHCC variants — matched to the right fiscal year per date of service.

MEAT validation

Monitoring, evaluation, assessment, treatment — every code scored against the documentation before it counts.

Specificity engine

Unspecified codes upgraded when the chart supports stage, laterality, or etiology — never the other way around.

RADV-ready audit

Billability, hierarchy, and coding-risk checks on every submission — with a defensible packet per code.

Built for ACOs, MSOs, and at-risk medical groups. Every chart tracked from ready-for-coding through QA, SME review, and submission.

Get started
The Architecture

The infrastructure behind every signal.

One pipeline turns each encounter into defensible risk and quality — secure ingestion, isolated models, and an evidence trail built for the audit.

Data sources
Panel rosterClaims feedLab / HIE feed
Secure ingest
FHIRHL7
OptimIQ engine
RiskQuality
Encounter ingestedHCC suspectedCare gap detected
Models
V28HEDISStars
Evidence
RAFMEATAudit
Coordinator
Audit Defense

Audit-ready the moment it's coded.

MRN 4471Progress note

Progress note

03/14/2025 · A. Reyes, MD · Endocrinology

58-year-old established patient returns for diabetes follow-up. A1c 9.2%, up from 8.1% in December — metformin continued, dose adjusted. Retinal screen and foot exam ordered. eGFR 52, consistent with CKD stage 3a; recheck in 12 weeks.

Assessment & plan

E11.22Diabetes w/ CKDI50.32Chronic diastolic HFN18.31CKD stage 3a

Signed — A. Reyes, MD · attested 03/14/2025

FAQ

Questions, answered.

What risk-bearing teams ask before they connect a panel.

Talk to the team
  • How does patient attribution work?

    OptimIQ ingests your roster and payer attribution files, reconciles them against encounters, and counts every life once — so the panel you see is the panel you're measured on.

  • Will it fit our EHR?

    Yes. It reads from Epic, Oracle Health, athenahealth, and others over HL7/FHIR — no rip-and-replace, and it starts read-only so nothing changes on day one.

  • How is a code defended in a RADV audit?

    Every submitted HCC links to a dated, signed note with MEAT criteria met — exportable as a defensible packet for any sampled member, in one click.

  • How fast can we go live?

    Most panels are connected and surfacing signals within a few weeks, beginning read-only so your team can verify everything before a single code is submitted.

  • Does it replace our coders?

    No. The engine drafts and evidences suspected conditions; your coders approve every code through QA and SME review gates before submission — nothing goes out on the AI's word alone.

See what one partner running it looks like.

Walk us through how things run today. We'll come back with where the leakage is and what we'd take on in the first 30 days.