OHDSI Home | Forums | Wiki | Github

Incorrect Concept Names for NDC Codes

Vocabulary Users
#1

Hi,
I’m working with the concepts table for a study in my company and I’ve noticed a couple of potentially incorrect descriptions applied to some NDC Drug Codes:

Concept_ID = 45020745, Concept_Class_ID = “11-Digit NDC”, Concept_Code = “55887085810”, Concept_Name = “Ciprofloxacin 500 MG Oral Tablet [Cipro]”

Concept_ID = 45072043, Concept_Class_ID = “11-Digit NDC”, Concept_Code = “55887085830”, Concept_Name = “Ciprofloxacin 500 MG Oral Tablet [Cipro]”

I believe these 2 NDC Codes are actually for “Simvastatin 40 MG Oral Tablet”.
As a result, they are being incorrectly mapped across dictionaries.

#2

We’ll figure out why it happens.
@mcshama1, how did you discovered this issue?
any testing script or algorithm?

#3

I’ve developed an app that searches through codes and uses the various “relationships” to map across various dictionaries of interest.
As we are a pharma company, a lot of our clients are familiar with the international ATC Drug Codes, but not as familiar with NDC Codes (the codes used in our database). One of the drugs our client was interested in was Ciprofloxacin (ATC Code: J01MA02).
We also have an internal NDC Code Dictionary. My app cross-references with this and that’s when I noticed that Simvastatin was in my final list, along with all the other Ciprofloxacin NDC codes.

#4

First,

and

have invalid_reason =‘D’,
and actually if you search something in concept table, better to add condition “where …invalid_reason is null”

I’m trying to check these concept codes using external tools,
https://mor.nlm.nih.gov/RxNav/#ndc
and There is no result for ‘55887085830’ (as NDC)
what do you use as a source of NDC codes?

#5

@mcshama1:

This is one of the rare cases where an NDC seems to be reused. The Cipro died in 2012, as you correctly said. What I don’t understand is why the Symvastatin isn’t added. We will debug. Thanks for the hint.

FDA SPL and RxNorm distribution files.

#6

I’ve looked on another website that says 55887-0858-30 is an obsolete Simvastatin NDC code (obsolete date=25Jul2008).

When it comes to codes that have become invalid/obsolete, is there any reason to not include them if your defined study period includes a time period when the codes were still valid?
So, for example, if I was interested in looking at statin use from 2006-2008 say…I would have thought that I should include all codes that were valid during that time period?!

#7

Yes, we don’t care about end dates for source concepts. Because your data will contain the old stuff anyway, and the data usually doesn’t care about those valid dates. So, the last valid NDC code was Symvastatin, and that’s what it’s going to be.

We already found the bug. NDCs will be re-released and all those are going to be added.

t