I’m trying to map my medication dataset to OMOP drug_concept table as per V6 specifications. When I read through the drug_concept_id field, it has a link of accepted concepts which filters out all the vocabs that are not standard. So, NDC which seems to be non standard filters out as well… about 8.5% of my dataset is NDC type.
What are the repercussions of loading a non-standard concept_id into drug_concept_id field ?
OR do I just input drug_concept_id = 0 for NDC rows since only standard values are accepted and then may be set drug_source_concept_id = 45018605 for example for dicyclomine ?
You have to map the non-standard concepts (NDC) to the standard ones (RxNorm), using the tables provided in the Standardized Vocabularies provided by Athena. The Book of OHDSI explains it all.
To narrow your reading see section 6.3 Step 2: Create the Code Mappings in the book of OHDSI. It has the SQL necessary to map from your non standard NDC to a OHDSI standard RxNorm concept.
Ex.
SELECT concept_id_2 standard_concept_id
FROM concept_relationship
INNER JOIN concept source_concept
ON concept_id = concept_id_1
WHERE concept_code = ‘51927134700’ – DICYCLOMINE HCL POWDER
AND vocabulary_id = ‘NDC’
AND relationship_id = ‘Maps to’;