GLN identifier (Da: lokationsnummer) wrongly specified as an identifier on Organization

[gh-eriknielsen]

GLN cannot be an identifier for Organization. For GLN (Da: lokationsnummer) it is even more obvious, cf. possibility to inherit GLN identifier in Sundhedsvæsenets Organisationsregister (SOR).

See issue #177 for similar issue with provider identifier.

[jkiddo]

I disagree. Its a perfectly valid identifier for an organization. The GLN identifies an organization - see https://www.gs1.dk/gs1-standarder/gln. The swiss community agrees to that fact as well - https://fhir.ch/ig/ch-core/StructureDefinition-ch-core-gln-identifier.html

[esn-systematic]

Right, I'll rephrase. GLN can indeed function as an identifier for Organization - provided that no business objects/entities have the same GLN. In Danish health care context, GLN is used in a way that violates this prerequisite, for instance in SOR. You can implement a system with, say, a select subset of SOR that guarantees GLN uniquely identifying Organization. In general and in Danish context, however, using GLN as a FHIR Organization identifier is not a good choice. Not until each GLN is used for one thing only.

[jkiddo]

Suggestion from @nigtrifork - change the modelling to follow along the lines of what is done in EHMI -> https://build.fhir.org/ig/medcomdk/dk-ehmi-eer/SOR-EER-mapping.html - representing GLN as Endpoint instead. @esn-systematic does that sounds reasonable to you?

[jkiddo]

@esn-systematic will you make a PR?

[tmsMedcom]

@jkiddo and @esn-systematic should this be moved to the next release?