OHDSI Home | Forums | Wiki | Github

Building drug_eras with non-RxNorm ingredients

CDM Builders
#1

In CDMv5 (and VocabV5), when an ingredient is not in RxNorm (for example Buformin) we have decided that the corresponding ATC concept will be promoted to ‘standard concept’ and will be used to indicate the ingredient.

When populating the drug_era table, all drugs will be rolled up to the ingredient level. In the past, you could just use the concept_ancestor table and roll up to the point where you reached vocabulary_id = 'RxNorm' and concept_class_id = 'Ingredient' . However, with the new ingredients it is not so easy. How do you know the ATC code is an ingredient? And currently ATC codes (5th level) do not appear in concept_ancestor.

@Christian_Reich, @Patrick_Ryan: any suggestions how best to tackle this? We could assume that a concept from vocabulary ‘ATC’ that is a standard concept is always an ingredient, but that seems a bit dangerous (and hard to understand if you read the code).

#2

@schuemie:

Hm. We are in the middle of the the addition of DM+D. There are a lot of the missing ingredients, buformin is not amongst them.

The reason not to use ATC as the primary ingredient concepts is this: As we add additional national drug schemes, we will receive the relationship between their ingredient codes (in the UK DM+D for example) and the corresponding drug products. If we used ATC we would have to build those additional relationship. Each one of them takes time and is error-prone.

What happened to you getting permission to add the various schemes officially?

#3

Ok, I think I misunderstood. I thought the OHDSI decision was to promote ATC concepts to ingredients.

I think all partners have agreed to have their codes become part of the vocab. I just need to clean it up a bit and give it to you.

t