Every chart reviewed before clinic opens.

Pre-visit chart review, HCC coding, and custom clinical workflows. Connects to eClinicalWorks via FHIR R4. Patient data never leaves your infrastructure.

$89/seat/mo after evaluation

What MedScrub Does

Pre-Visit Chart Review

71 charts reviewed before the first visit.

Records sync from eClinicalWorks overnight. By the time the clinic opens, every scheduled patient has been reviewed. Problems, medications, allergies, HCC gaps, and care plan recommendations — summarized and ready. Works alongside ChartNote.

MedScrub architecture diagram showing data flow from eClinicalWorks through de-identification to AI processing
HCC Capture

Conditions the chart supports but the codes miss.

Surfaces suspected HCC conditions mapped to the CMS-HCC V28 model. Runs in bulk across a whole panel, not one chart at a time. Gaps flagged in each pre-visit summary. The provider decides — every suggestion shows its evidence, and nothing is coded until the provider documents it.

Suspected HCC 37

E11.22 — Type 2 diabetes mellitus with diabetic chronic kidney disease

Suspected HCC 155

F33.2 — Major depressive disorder, recurrent, severe without psychotic features

Recapture HCC 226

I50.22 — Chronic systolic (left ventricular) heart failure

Custom Workflows

Describe it. MedScrub builds it.

Champions describe the workflow the way they would tell a colleague. MedScrub shows a preview in plain terms. Nothing is sent or written to a chart until the provider approves. Build referral prep, medication reconciliation, care gap closure — anything that follows a pattern and touches the chart.

MedScrub custom workflow diagram

Patient data stays inside your building.

Swiss-cheese de-identification

Structured FHIR fields — one precise layer. A field map strips names, dates, identifiers, and addresses from every resource before it leaves the FHIR pipeline.

Unstructured notes, letters, and PDFs — five layers. Regex patterns, scispaCy, Stanza, BioBERT, and MedScrub’s own models each run independently across every piece of free text.

No single layer needs to be perfect. Each catches what the previous one misses.

  • Your choice of AI model — OpenAI, Anthropic, Google Gemini, Azure OpenAI, or Ollama fully local
  • API keys stored on your machines, never sent to MedScrub servers
  • With a BAA or without — charts go identified or de-identified, your rules
  • A MedScrub server on your infrastructure: your eCW connection, your data

Pricing

$89
per seat, per month
$24,000
one-time forward-deployed implementation
$30,000
annual partner support

No per-message fees. No data egress charges. No hidden costs.

See it on your own patients.

Free 2-week evaluation on your own data. We sign the BAA, install MedScrub, connect it to eCW, and let you see the results.