NDC codes as concept_code?

[MPhilofsky (Melanie Philofsky)]

We have 9 and 11 digit NDC codes for our administered drugs. What is the method to convert from a 9 or 11 digit NDC code to the OMOP concept_code so I can find the appropriate concept_id?

Thanks,
Melanie

[Chris_Knoll]

Use the CONCEPT_REPLATIONSHIP table, looking for the ‘MapsTo’ relationship_id, where concept_1 is the NDC concept (found by looking for for the concept_code = the NDC code) and concept_2 will be the standard concept (from RxNorm) that this NDC maps to. Use the concept_2 as the value you write to in DRUG_EXPOSURE’s drug_concept_id column.

-Chris

[Christian_Reich]

@MPhilofsky:

It’s described somewhere, but I can’t find it.

Here is the gist:

1. If you have a 11-NDC, try finding an exact match and use it
2. If you have a 9-NDC, try finding an exact match and use it
3. Take all 11-NDCs and all 9-NDCs that share the first 9 digits and check their ‘Maps to’ relationship.

• If they all map to the same RxNorm code, use those 9 digits and map your 11-NDC (first 9 digits) or 9-NDC.
• If they all map to different RxNorm codes, but these represent a single Clinical-Branded pair (have a ‘Has a tradename’ relationship between them), use the Branded Version

4. Post the NDC here in the Forum and complain about having no mapping.

[Jorge_Luna]

Hi: Trying to find NDC for continuous glucose monitoring devices:

But it seems that OMOP’s Athena doesnt have these codes

Does that mean that these NDC are not mapped in OMOP? If not, how do we get glucose devices mapped?

[Jorge_Luna]

Athena does not find this specific NDC code, does this indicate a general class of devices is missing?

[aostropolets (Anna Ostropolets)]

It means that the official NDC source (which is FDA) doesn’t provide us with those codes. Might indicate that they don’t have them or that your codes are wrong.
For example, other “verio” are there:
http://athena.ohdsi.org/search-terms/terms?vocabulary=NDC&page=1&pageSize=15&query=verio&sort=concept_code&order=desc
If you believe that your codes are correct, then submit the names, codes and start/end dates here:

Missing NDCs remote studies

Thanks, @Vojtech_Huser. However, you need to add: Names/descriptions Timing of use Can you add that? NDCs exist since the early 80ies. No. They dropped the ball in the late 80ies. The key agency is the NLM, and then there are private repositories (Red Book, FDB, Medi-Span, etc.) Because we got it from another source, most likely from Dailymed and the SPL. Because we got it from another source and cobbled the mapping together. But this really should go to the NLM. That’s what we get…

[Jorge_Luna]

Thank you, the NDC codes are correct and our organization’s NDC codes are derived from Medi-Span. Is there a general strategy to integrate Medi-Span codes? Or should I pursue a one-off as per your link?

[Christian_Reich]

@Jorge_Luna:

Medi-Span is one of the sources for NDCs. Apparently, the NLM decided against this one. Feel free, as @aostropolets said, to submit the code and we will bring it up with them.

However: non-drug NDCs are step children anyway. NDC, which stands for National Drug Code, was designed as a system of unique product identifiers used in the United States for drugs intended for human use. Over time, manufacturers started making NDCs for everything that is sold in a pharmacy, drug or not, including glucose sticks. There really is no authority that oversees these funny NDCs. So, I wouldn’t bet the farm on those kinds of codes.

[Jorge_Luna]

Thank you for this response and general background info primer. My current interest is to identify continuous glucose monitoring (CGM) device users. I will review the task of curating a comprehensive set of NDCs to identify CGM devices and related supplies. Strips were just a test case. Great community here, thanks all !!!