EHR integration with Rhazes

Rhazes connects to your EHR over SMART on FHIR, so any system exposing a SMART on FHIR endpoint can be integrated.

Updated

Built on SMART on FHIR

Rhazes integrates with electronic health records over SMART on FHIR, against FHIR R4.

That is the important architectural point, and it is what makes the question "do you support our EHR" usually answerable with yes. Rhazes is not written against one vendor's private API. If your system exposes a SMART on FHIR endpoint, we can integrate with it.

What is live today

Epic and Oracle Health, formerly Cerner, are supported and in use.

Anything else running SMART on FHIR is an integration conversation rather than a rebuild. The retrieval, the resource handling and the clinical reasoning on top of it are already there and vendor-neutral. What changes per system is the endpoint, the client registration and the scopes your organisation approves.

Configured once, for your deployment

The EHR connection belongs to your instance rather than to each clinician. The endpoint, the tenant and the enabled resources are set up as part of your deployment, so clinicians never configure anything. See How Rhazes Enterprise is deployed.

Clinicians sign in with their EHR account

Rather than holding a separate service account, Rhazes uses your clinicians' own EHR identities. A clinician signs in to Rhazes through your EHR, and from then on Rhazes reads the record as that clinician.

This is the design decision that matters most for governance. Rhazes does not have its own view of your patient data and cannot see anything the signed-in clinician could not see. Their permissions are the permissions. When somebody leaves and their EHR account is disabled, their access through Rhazes goes with it, without anyone having to remember to tell us.

It is read-only

Every scope Rhazes requests is a read scope. It cannot write to, amend or delete anything in your EHR.

See What Rhazes reads from your EHR.

What clinicians do with it

Find a patient by MRN, link a Rhazes record to the chart, pull the clinical data across, and use it in chats, documents and workflows.

The day-to-day of that is documented with the rest of the patient features:

Workflows can also read the patient record directly, so an automated sequence can start from live clinical data rather than from something typed in. See Choosing what a workflow can draw on.

What setting it up involves

Roughly, and your organisation will recognise most of this:

  1. A conversation about scope, covering which resources you are willing to expose and which clinicians are in the first group.
  2. Client registration in your EHR vendor's app programme.
  3. Information governance review on your side, which is the step worth starting early.
  4. A non-production connection first, so clinicians see it working against test data before it touches a real chart.
  5. A pilot with a small group, then widening.

Email support@rhazes.ai to begin.

Related articles