AI glossary · Basic · Data engineering · Healthcare & bio
EHR·FHIR
An EHR (electronic health record) is a system that stores and manages patients' medical records digitally, and FHIR (pronounced "fire") is an international standard that lets different EHRs and health apps exchange data in the same format.
About this term
- Level
- Basic · Technical terms you often meet in AI news
- Category
- Industry AI
- Fields
- Data engineering
- Industries
- Healthcare & bio
- Added
- 2026-10-10
How it works
An EHR holds a patient's basic information, diagnoses, test results, prescriptions and clinical notes. The problem is that every hospital and vendor structures its system differently, which makes moving data between them hard. FHIR (Fast Healthcare Interoperability Resources) is a standard created by the healthcare standards body HL7, and it defines information such as patients, test results and prescriptions as units called "resources." Each resource is exchanged using the API style common on the web and formats such as JSON, so ordinary web and app developers can work with it easily.
Example FHIR resources
- FHIR resources
- PatientPatient basic information
- ObservationTest results and vital signs
- ConditionDiagnoses and health problems
- MedicationRequestPrescriptions
- EncounterCare visits such as outpatient and inpatient
Why it matters
When patients change hospitals or get care at several institutions, their records easily become scattered. A common format like FHIR makes it much easier to share records between institutions, connect patient health apps and collect data for research. It matters for AI too: AI tools such as CDSS or medical LLMs can only spread widely in real clinical settings if they can connect to many hospitals' EHRs in the same way.
Things to know
Using FHIR does not automatically make every system compatible. Even within the same standard, countries and institutions often define their own detailed rules (profiles). Medical data is highly sensitive personal information, so access control, de-identification and consent procedures must be designed in from the start. Also, records in EHRs are entered in inconsistent ways, so data quality needs to be checked before using them to train AI.
Example
A smartphone health app that lets patients see test results and medication lists from several hospitals in one place works because hospital EHRs export data through FHIR APIs. When a hospital adds an AI-based CDSS, a common approach is to read a patient's lab values and prescription information via FHIR, predict risk, and display the results back on the clinicians' screens.