Patient ID maps

A patient ID map translates Patient IDs between domains: a site’s MRNs and the enterprise IDs, or the enterprise IDs and an outside PACS’s. Configuration › Patient IDs.

A map

• Entries: from-ID → to-ID, compared exactly, case included.
• Issuers before and after (Issuer of Patient ID). With a from issuer, instances with another issuer are left alone.
• IDs not in the map: kept, or refused (received instances go to the quarantine, sent ones are dead-lettered, with the reason).
• Whether the original ID and issuer are kept in Other Patient IDs Sequence (default yes).

Using it

A map is used through the tag edit MapPatientId (choose the map as its value), so it runs in order with the other edits:

• on receipt, in a source’s edits (strip a prefix, then map, say);
• on send, in a destination’s edits;
• in a de-identification profile’s own edits.

Archives and senders with other IDs

• An archive with other IDs: an archive whose edits map patient IDs is searched with its own IDs (C-FIND, QIDO-RS and prior studies through the router), and its answers and the instances it sends back are mapped back to the router’s. An archive that retrieves by C-MOVE is then asked to move to the router rather than straight to the destination, so it must know the router.
• Senders with other IDs: a source whose edits map its senders’ IDs to the router’s lets them search and retrieve in their own IDs: their queries are mapped to the router’s IDs, and the answers and instances back to theirs.

Entries

Entries are kept by the central service, not in the configuration (so not in exports or the version history):

• add or change them one at a time, or import a CSV of from,to lines (comma, semicolon or tab, an optional header; add or replace all), and export them as CSV;
• the entries dialog warns when several from-IDs map to one to-ID (mapping back, the first is used);
• every change reaches the nodes, which keep the map on disk.

Until a node has a map’s entries, a source that needs it refuses instances (the sender can send again), a destination that needs it is held, and an archive that needs it is not asked.

The entries identify patients: listing and exporting them needs the history role, and changing them administrators with it. Every view, export and change is audited, without the IDs.

HL7 merges

When the RIS merges two patients (ADT^A40, A18, A34), the worklist moves their orders to the surviving ID. With a map chosen under Settings › HL7 › Merges update, the prior ID is also added to that map, so studies still sent with the prior ID can be mapped to the surviving one.