OHDSI Home | Forums | Wiki | Github

What is the proper way to map to a standard concept if your source has a RxNorm Brand Name?


(Michael Gurley) #1

For example: ‘Sensorcaine with Epinephrine’

http://athena.ohdsi.org/search-terms/terms/19080834

There are many ‘Brand name of (OMOP)’ relationships.


(Christian Reich) #2

Well, but that’s how it is. Take “Nyquil”. A dozen ingredients. It’s a one-to-many relationship.


(Michael Gurley) #3

Is there a reference SQL implementation of how you should map all non-standard RxNorm concepts to standard RxNorm concepts? This seems like something that could be easily formalized and help standardize ETL implementations.

Your embedded wisdom just revealed that we should map to the ingredients. But there are no direct relationship entries between a Branded Name and its ingredients. I know the ingredients can be derived by hopping to ‘Brand name of (OMOP)’ entries. But the relationship to ingredients seems to be present within RxNorm itself:

https://mor.nlm.nih.gov/RxNav/search?searchBy=String&searchTerm=Sensorcaine%20with%20Epinephrine


(Christian Reich) #4

Yes, we need that cook book. It’s in the plans, but the Covid thing is not really helping with us getting organized and moving down a well thought through plan.

In this situation, when people refer to a brand, what they really mean is a certain product with this brand name. Since we don’t know which you should take either the Branded Drug Component or Branded Drug Form, hoping one of them is unambiguous. If they aren’t, and the same brand is assigned to more than one of these two you can still try to write individual ingredient records for each of the ingredients. But that also can go wrong, if the ambiguous brand name is attached to different ingredient combinations. In that case you are out of options and have to go back to the EHR trying to figure out what they meant.

Medicine is not a precise science.


t