Stream: terminology
Topic: ConceptMap priority
Alexander Henket (Jan 26 2017 at 09:56):
Suppose I have a concept that maps to multiple other concepts e.g. in a SNOMED CT / ICD10 map. How do I give one mapping priority over the other? I could not find any text that hints at priority. I'm currently assuming that the order of ConceptMap.group and ConceptMap.group.element bears implicit relevance, but since that behavior is not defined it feels dangerous to rely on.
Jim Steel (Jan 27 2017 at 00:02):
I don't follow. $translate should return all mappings
Rob Hausam (Jan 27 2017 at 00:58):
You shouldn't assume that there is any meaning implied by the ordering of ConceptMap.group and ConceptMap.group.element. I think if you really feel that being able to indicate the mapping priority is needed, then that needs to be added as an extension.
Alexander Henket (Jan 27 2017 at 08:15):
Well I'm told none of the SNOMED/ICD mapping work without priority. That means that without it you rule out a very important part of the mapping world. @Jim Steel yes $translate returns all mappings, but if there are multiple mappings for 1 concept then what? How do I know which is preferred/has priority?
Alexander Henket (Jan 27 2017 at 08:33):
@Rob Hausam : Thanks for the confirmation that this functionality is missing. I don't agree however with the idea that this missing functionality should be solved using an extension. I've added GF#12719 to request a solution.
Michael Lawley (Jan 27 2017 at 10:04):
be careful @Alexander Henket - the snomed - icd mapping from IHTSDO is a set of rules, not a lookup table. It is guided by the coding rules that come with ICD (IIRC) and is not really appropriate for a FHIR ConceptMap (some rules require knowledge of gender, age, etc)
Ultimately, unless you know about the map itself you'll never know what the "priority" is.
Grahame Grieve (Feb 06 2017 at 02:51):
this feels like an R4 problem
Rob Hausam (Feb 06 2017 at 02:58):
agree - GF#12719 is deferred, and we can pick it up for R4
Last updated: Apr 12 2022 at 19:14 UTC
