HL7 for multi-site hospitals and facility routing
A multi-site HL7 interface is not just one hospital with longer location names. Facility identity can appear in MSH-4, MSH-6, PV1-3, CX assigning authorities, provider identifiers, accession numbers, account numbers, and local routing tables.
The safe rule is to preserve the authority with the value. A patient number from North Hospital and the same text from South Hospital are not the same identifier unless the assigning authority says they are.
This is synthetic sample data for learning and testing. Open it in HL7 Soup Web before mapping it so the segment groups, repeated fields, and coded values are visible.
Routing Values Are Not Display Labels
Do not route production traffic from a free-text facility label. Use agreed facility codes, assigning authorities, sender/receiver application/facility, patient location components, and message type. The PL datatype behind PV1-3 is useful because point of care, room, bed, and facility are separate components.
Multi-site routing often needs exception handling: one lab serves several hospitals, one RIS receives for multiple sites, one billing system handles the region, and one specialty system only accepts one campus. Make the routing table visible and supportable.
- Keep MSH sender and receiver application/facility fields.
- Keep assigning authority in PID-3, PV1-19, ORC/OBR identifiers, and accession values.
- Log the facility decision and destination for every routed message.
Merges, Moves, And Shared Services
Patient merges and transfers become more sensitive when identifiers come from multiple campuses. An ADT_A40 merge at one site may need to update a regional repository, a lab system, and a radiology system that each use different patient IDs. Do not strip the source authority before routing.
A Practical Integration Soup Workflow
In Integration Soup, put facility routing in a table, variables, or clearly named filters. Use MSH-9, MSH-4, MSH-6, PV1-3 facility, patient class, and assigning authority as explicit inputs. Keep routing decisions visible in logs so support can answer why a message went to a destination.
The Integration Host getting started tutorial gives the workflow foundation, and the transformers tutorial is the next step for facility-code normalization.
The Test Pack I Would Ask For
Ask for each site, shared-service sender, inter-site transfer, site-specific patient ID, enterprise MPI ID, merge, cancelled movement, and a message that should be rejected because the facility is unknown.