We’re struggling with a number of System values which refer to multiple systems. We are designing a set of LOINC-codes for laboratories and a HL7-message to communicate them with. In this HL7-message each LOINC-test is sent with a single SNOMED-concept that denotes the material. Our set of LOINC-codes is enriched with SNOMED-concepts that for each test denote which type of specimen can be used; this is always equivalent with the LOINC-system or more specific (if LOINC-system = XXX).
This works fine for most LOINC-concepts. For Systems such as ‘Bld/Tiss’ we can send the material that was used in the HL7-message, and specify the two possible systems as two separate SNOMED-concepts in our set. But for Systems such as ‘Urine + Ser/Plas’ we have a problem. There is no such material in SNOMED, because it is in fact two separate materials: so it does not fit in our HL7-message. And if we map them to three separate SNOMED-concepts in our set, how do we show that the test needs two of them? That it is a conjunction rather than a disjunction?
How should we treat such concepts? Do we extend the HL7-message to allow multiple systems? Do we lobby with SNOMED for the addition of such combination concepts (which would be difficult or even impossible to define)? Or should we lobby for the removal of these Systems from LOINC? Do these concepts denote a single test, or are they really combinations of tests?
Some examples: 2164-2 Creatinine renal clearance in 24 hours and 42207-1 IgG in liquor cerebrospinalis/IgG in serum
Who has faced a similar issue or has advice to offer?
Feikje Hielkema-Raadsveld, Nictiz