What is FHIR (Fast Healthcare Interoperability Resources)?

FHIR is the HL7 standard for exchanging healthcare information electronically, built from modular data units called resources (Patient, Observation, Encounter and so on) that can be read and written over a web API in JSON or XML. It is the standard US certification rules require EHRs to use for their patient and population data APIs.

HL7 describes FHIR as a standard for exchanging healthcare information electronically that draws on the lessons of its older v2, v3 and CDA standards. The basic building block is the resource: every piece of exchangeable content, from a patient to a lab result, is defined as a resource with shared datatypes, common metadata and a human-readable narrative. Resources point to each other through references, and an extension mechanism covers anything the core set leaves out.

FHIR supports several ways of moving data: a RESTful API, messaging, documents and service-based architectures. REST is the most familiar to web developers, but HL7 notes that you do not have to use REST to use resources.

Which version matters

The current published release on hl7.org is R5. US certification is pinned to an earlier one: the ONC standardized API criterion, § 170.315(g)(10), requires FHIR R4 (4.0.1) together with the US Core profiles and SMART App Launch. In practice that means a product that wants to read from certified EHRs should plan around R4. Epic's developer site, for example, lists R4 alongside the older STU3 and DSTU2 versions.

Example

A remote monitoring app that needs a patient's active medications calls the EHR's FHIR endpoint, searches MedicationRequest resources for that patient, and gets back a Bundle of JSON records. The same call works unchanged against any other server that implements the same profiles.

For a longer, non-technical walkthrough of resources, profiles and what to ask an EHR vendor, read our primer FHIR for non-engineers.