I am seeking advice on the best approach to map local microbiology examinations codes to LOINC codes.
The team I’m working with proposed two potential approaches. For example, in the case of describing a “Body fluid culture result,” the options are:
1.Use the most specific LOINC code : For example, LOINC 6114-6 | Bacteria identified in Body fluid by culture.
2.Use a general LOINC code combined with a SNOMED code : For example, LOINC 6463-4 | Bacteria identified in Specimen by Culture, combined with SNOMED code 258450006 | Cerebrospinal fluid specimen (specimen).
If your LIS has specific fields for specimens (as is often the case these days), these fields should be mapped to SNOMED codes within the specimen hierarchy. SNOMED provides a more complex ontology than what is typically required in microbiology.
In the procedures field, we use LOINC codes. The difference between specimen-specific codes and specimen-agnostic codes is defined by the clinical or process relevance of the specimen. Is your procedure validated for a limited list of specific specimens? Is your post-analytical process dependent on the type of specimen used? If so, use specimen-specific codes. If not, use general codes “in specimen” and rely on the specific field for specimen identification.
Hi Ella, Ivan stated it very well.
I would like to add a note about using the SNOMED code for Specimen. This is in relation to the LOINC SNOMED collaborative work. Specimen that are mapped to a SNOMED codes will have a one to one mapping to a LOINC codes going forward(maybe not right now, but this is the strategy for the close future). I reiterate from Ivan that it depends on your ultimate use case in analytics, and reporting that will justify what you are going to map to currently.
Thank you Ivan , and Ella if you have further questions, please let us know or If you need a concept that is not in the DB, please submit.
Eza
