OHDSI Home | Forums | Wiki | Github

Missing rxcui codes (in the concept_code field)

Hello,
The data source I am using has approximately 70 rxcui codes that are missing from the concept_code field of the CONCEPT table. I have looked up a number of them using Google and other databases, and so far, they all appear to be valid rxcui codes, so I’m not understanding why they are missing from the CONCEPT table. How often is this table updated?
Two of the rxcui codes in my data that are not in the CONCEPT table are 166283 and 74169.

Thanks,
Betsey

@bgardstein:

The reason is that these concepts are fixed ingredient combinations. 74169 is Piperacillin / tazobactam and 166283 is Lidocaine / Prilocaine. We don’t use those fixed ingredient combinations in OMOP. Only single ingredients. So, the latter should be split into 2 records, one with 6387 Lidocaine and another with 8686 Prilocaine.

We can add those to the Mapping records, if that helps you. What kind of database are you using that gives you RxNorm?

@Christian_Reich I’m working with @bgardstein. The client has a private database they have built using RxNorm. If I read the vocabulary build code correctly, the process excludes tty=MIN. Do you recall the reasons why these are excluded in OMOP? Adding the records to concept is one thing; is it problematic to encode all the relationships?

@DTorok:

Well, we could have incorporated MIN as a concept_class, but we didn’t. I don’t remember the exact reason, it’s in the early OMOP days when we adopted RxNorm for our purposes, and MIN appeared not really necessary, because it was so easy to resolve it to two or many INGs.

As I told @bgardstein, we can add them as source concepts, with mapping to the respective INGs no problem. She has to want it though. :smile:

@Christian_Reich:

MIN appeared not really necessary, because it was so easy to resolve it to two or many INGs

That works nicely when your analytic use case is, ‘Did the patient encounter substance X?’, and not so well when your use case is, ‘Which drug was chosen to treat X?’. :smiley:

That said, if one can get to MIN, it may be possible to get to SCDF, which I’d argue is where you’d often like to be.

t