What makes DrChrono easy or hard to automate
DrChrono has been API-friendly for years. Its REST API covers patients, appointments, clinical notes, documents, problems, medications, insurance, line items and billing, and it supports webhooks for changes such as new or updated appointments. A practice can create its own API application from the account’s API settings, add a redirect URI, and authorize it with OAuth 2.0. For a single practice, that means you can be reading real data in days once the paperwork is done, which is unusual for an EHR.
There are trade-offs. There is no machine-to-machine client-credentials flow, so a person authorizes the app once and the integration must refresh tokens reliably forever after. Requests are rate limited per application, so bulk exports need pacing. The API is versioned, and you should pin a version so field changes do not surprise you. Selling an app to other practices or listing it in DrChrono’s marketplace goes through DrChrono’s API team, which is a separate approval.
The AI layer
DrChrono is the clinical and billing record. AI handles the admin work that eats staff time:
- Intake that asks follow-up questions in plain language and produces a clean summary for staff.
- Document handling, where a model reads a fax, decides what it is and whose it is, and proposes where to file it.
- Patient messaging, such as reminders and after-hours replies that sound like your front desk.
- Billing summaries that turn rejections and missing charges into a short worklist.
Every write to the chart goes through a person, and clinical judgment stays with clinicians.
Limits to plan around
- HIPAA and BAAs. Host the integration in a HIPAA-eligible environment, sign BAAs with each vendor that touches PHI, and log access.
- Token upkeep. Access tokens expire and refresh tokens must be stored securely and rotated. A silent refresh failure stops every workflow.
- Rate limits and fees. DrChrono’s API docs set a default limit of 500 calls per hour per application, resetting at the top of each hour, with increases by request. Its developer terms also allow DrChrono to charge fees for API use, so ask before you scope. Spread backfills over time and cache reference data such as offices and doctors.
- Permissions. The authorizing user’s permissions bound what the app can do, so authorize with a dedicated account set up for the integration.
How Lumeor typically connects it
For a single practice we usually run self-hosted n8n in the practice’s HIPAA-eligible cloud account, with a DrChrono OAuth credential, a webhook receiver for appointment changes, and an LLM step for summaries and drafts. Staff approve writes in Slack or a small review page. The patient intake guide and no-show reduction guide cover the most common builds, and our n8n cloud vs self-hosted comparison explains the hosting choice. The healthcare AI triage case study shows patient-facing intake in practice.