A boarding pass can be perfectly valid and still not get you onto this flight.
It may belong to another date, another destination or another passenger. There is nothing necessarily wrong with the boarding pass itself. It simply does not belong in this context.

The same can be true for a code.

A code may exist in a CodeSystem, be active and have a clear definition, yet still not be allowed within the ValueSet, version or clinical application in which it is being used.

That is where validation becomes important. Validation does not only ask whether a code is correct. It asks whether it is correct here.

Does the code exist in the expected CodeSystem? Is it active in the relevant version? Does it belong to the required ValueSet? And is it permitted within this particular use case?

Those questions may sound technical, but they determine what people can record, exchange and interpret. A system that checks only whether a code exists may accept something that is technically recognisable but semantically inappropriate for the situation.

This is why validity is rarely an absolute property of a code. It is a relationship between the code, the rules surrounding it and the context in which it is used.

A valid code is not always a valid choice.

Sometimes the code is correct.
It just does not belong here.

Next week: Translation

What can move between vocabularies — and what gets lost along the way?

#SemanticWeekly #ConceptMap #SemanticInteroperability #HealthInformatics