Hi,
Many of the archetypes in the CKM use local codes extensively (Diagnosis certainty in problem diagnosis, severity category in symptom sign etc.). SNOMED CT seems to include reasonable replacements for a large number of these already.
Will it not make sense for reducing the use of local codes in archetypes as that will improve interoperability of OpenEHR modeled data beyond the OpenEHR ecosystem and also reduce management overhead for the modelling community?
Do we have any specific reasons for not leaving such nodes as text with recommendations for use of coding terminology?
Each documented item in the EHR is following the pattern: Question=Answer
Archetype/Template nodes define the Question.
The Data Element in the Archetype/Template defines the Answer.
Between the two organisations behind LOINC and SNOMED there is an agreement to use LOINC for coding the question.
And use SNOMED to code the answer.
CIMI is using this convention.
This brings me to:
Question = Answer
use LOINC = use SNOMED
2-Local code sets versus International codes.
Interoperability (interpretability) demands that in the end we all use the same Basic Models/standards in our Semantic stacks.
Always there will be reasons to create local codes/classifications for specific purposes.
Interoperability (interpretability) demands that these local codes/classifications need to be mapped to International models/standards.