HL7 to FHIR

HL7 v2.x has been the most widespread interchange format in clinical IT for decades: segment-based and pipe-delimited, with segments like MSH (message header), EVN (event), PID (patient), PV1 (encounter), OBR (request) and OBX (observation) and the delimiters ^~\&. The modern successor HL7 FHIR R4 models the same information as clearly named resources (Patient, Encounter, Observation) and delivers them as JSON or XML - far easier for web APIs and modern systems.

Drop your HL7 v2 file; the tool reads the segments and builds a FHIR R4 Bundle (collection). A curated core is mapped: MSH becomes a MessageHeader, PID a Patient resource (identifier, name, gender, birth date, address, extended by PD1), PV1 an Encounter (class, visit number, location), NK1 a RelatedPerson (next of kin), DG1 a Condition (health condition), AL1 an AllergyIntolerance (allergy), PR1 a Procedure, OBX an Observation (code, value, unit) and OBR/ORC a ServiceRequest (order). You can drop several files at once, each becomes its own Bundle.

Everything runs entirely locally in the browser (pure JavaScript, no server, no foreign library from a foreign CDN) - the message with its patient data never leaves your device. Honest framing: this is a structural conversion of a common segment core (now including NK1, DG1, AL1, PR1, ORC, NTE and PD1), not a complete HL7-to-FHIR mapping (Z-segments and many detail fields are not covered) and not a clinical tool. Verify the result against the profiles and requirements of your target system.

Specifications

In 3 steps

1. Drop your HL7 v2 file(s) (.hl7, .txt).
2. The conversion runs automatically, locally in the browser.
3. Download the FHIR R4 Bundle (.json) - one per message.

Limitations: Converts an HL7 v2 message (e.g. HL7 v2.5, segment-based, pipe-delimited) into an HL7 FHIR R4 Bundle (collection, JSON): MSH -> MessageHeader, PID -> Patient (identifier, name, gender, birth date, address), PD1 extends the Patient resource (primary care provider, protection indicator), PV1 -> Encounter (class, visit number, location), NK1 -> RelatedPerson (next of kin, relationship, address, phone), DG1 -> Condition (health condition, condition type), AL1 -> AllergyIntolerance (allergy, severity, reaction), PR1 -> Procedure, OBX -> Observation (code, value, unit), OBR/ORC -> ServiceRequest (order), NTE attaches as a note onto the preceding resource. It is a curated segment core: Z-segments, EVN, repetitions beyond the first and many optional detail fields are intentionally not represented, so the mapping is not lossless. The file is read as UTF-8; the message's own MSH-18 character-set field is not evaluated - for ISO-8859-1/ANSI-encoded messages with umlauts or special characters, save the file as UTF-8 first. It is a structural format conversion for data exchange and not a clinical tool; verify the Bundle against the profiles (e.g. required fields, terminologies) of your target system.

FAQ

Related tools

Anonymize DICOM · FHIR JSON and XML · View HL7 CDA