EMR MCP is a read-only integration for your behavioral health EMR, built on its API — V3 and V4. Your staff ask questions in plain English inside Claude — census, occupancy, scheduling, chart assessments — and answers come back live from your instance, governed by your own EMR permissions.
MCP — the Model Context Protocol — is the open standard for connecting AI assistants like Claude to the systems where real work lives. An MCP server is the secure bridge in the middle: Claude asks it questions, and it translates them into signed, scoped requests against the system of record.
EMR MCP is that bridge for your EMR. Each organization gets its own private connector at yourorg.emrmcp.com/mcp. Staff add it to Claude once, sign in with their own work accounts, and from then on "what's our census at Riverside today?" is a question, not a report request that waits in someone's queue.
Three properties define the integration:
Each tool covers a specific question treatment centers actually ask. Every call is read-only, signed per request, and scoped to what the person asking is already permitted to see.
| Tool | What it answers |
|---|---|
Locations | Facility list with location IDs — the anchor for every other query. |
Census | Admissions by location and date range, PHI-minimized by default. |
Occupancy | Beds, capacity, and availability for daily census questions. |
Recent updates | What changed this week — patients created or updated in a window. |
Patient record | One record by casefile ID, detail flags opt-in only. |
Appointments | Scheduler pulls by date range and location. |
Evaluations | Every assessment on a chart — biopsychosocial, ASAM, nursing, progress notes — with author and dates. |
Evaluation detail | A full assessment, returned complete instead of truncated — nothing clinical is silently lost. |
Detox census | Verified in-care headcount that handles pre-dated discharges correctly. Counts only — never records. |
Your EMR administrator generates credentials from your own instance — Settings → API management. If API access isn't enabled yet, your EMR vendor can turn it on. We validate credentials live during onboarding, and you can revoke them at any time.
We recommend Claude Enterprise with a BAA signed between your organization and Anthropic, since EMR answers appear in staff conversations. A workspace admin adds the connector once; each staff member signs in individually.
Setup takes about 15 minutes end to end — no new EMR modules, no workflow edits, no vendor project. Your admin generates API credentials and we handle the rest. The step-by-step is on how it works, and the full security model — BAA, audit trail, data inventory — is on security.
Independence, stated plainly: EMR MCP is an independent product built on your EMR vendor's public API. It is not affiliated with, endorsed by, or sponsored by any EMR vendor — it's built for teams that want their EMR answering questions inside Claude.
Real questions against your own EMR instance, answered live — read-only, and governed by your existing permissions.
Book a demo