Stream: netherlands
Topic: Mapping SNOMED to Zib code
Ted Vinke (Feb 19 2020 at 16:13):
For the MedMij qualification in the Netherlands I'm required for a certain advance directive to translate a SNOMED code from the IT supplier's legacy system to a Dutch Zib code.
The 'Zib Codestelsels` (https://zibs.nl/wiki/Zib_Codestelsels) defines a custom coding system for which it reserved a root OID of '2.16.840.1.113883.2.4.3.11.60.40.4'.
The challenge is to translate e.g. a SNOMED code of 373067005 ('No') to the code 'NEE' in the '2.16.840.1.113883.2.4.3.11.60.40.4' system.
How would I know the correct mapping between "any" SNOMED CT and the '2.16.840.1.113883.2.4.3.11.60.40.4' system?
(The entire idea for Zibs.nl to introduce such a custom system is the fact that some concepts are not available in SNOMED at all, but somehow our IT supplier uses SNOMED and returns values for e.g. yes (373066001), no (373067005) and undetermined (373068000) for an advanced directive)
Alexander Henket (Feb 20 2020 at 15:29):
@Niek van Galen I've flagged this as https://bits.nictiz.nl/browse/MM-961. The question appears to be more related to zibs than to FHIR implementation thereof.
The question is never to map any SNOMED CT concept to a code in the aforementioned codesystem. The system '2.16.840.1.113883.2.4.3.11.60.40.4' only exists for concepts not found in SNOMED CT or other relevant system. It could be that something is added to SNOMED CT later on in which case mapping could make sense.
Where mappings are relevant, the profiles call those out in a ConceptMap. Note that after today I'll be on holiday, hence I flagged a colleague to pick this up after today
Niek van Galen (Feb 28 2020 at 13:37):
Hi @Ted Vinke,
I forwarded this question to the zib experts at Nictiz and got the following answer:
with any mapping from a legacy system (particularly with yes/no lists) onto an exchange format like zibs one should carefully check if the context is in fact exactly the same. As I do not have the information from the legacy system I could not say if is in this case valid to map Yes onto JA and No onto NO. But if they are found to be the same there would be no problem to do this. However, check! .... the real challenge to me seems the JA_MAAR option.
(Also see here: https://bits.nictiz.nl/browse/MM-961)
Does this make any sense?
Last updated: Apr 12 2022 at 19:14 UTC
