We are seeing these NDCs coming across in our data in 2006/2007.
None of these CONCEPT_CODES exist in our Vocabulary. But a couple years ago, someone gave a presentation to the OMOP/OHDSI community that said if you don’t get the 11-digit NDC, try for the 9-digit NDC. This logic is implemented in our CDM_BUILDER.
Our CDM builder is picking up these NDCs and translating them to the following. The only problem here is that SUVOREXANT didn’t start until 2014/2015.
000060005 (suvorexant 5mg/1 ORAL TABLET, FILM COATED [belsomra])
000060325 (suvorexant 15mg/1 ORAL TABLET, FILM COATED [belsomra])
000060335 (suvorexant 20mg/1 ORAL TABLET, FILM COATED [belsomra])
What is the best solution here?
Should we be using the VALID_START_DATE/END_DATE now?
Should we not be doing the try 9-digit when 11-digit doesn’t work logic?
So, these are the NDCs picked up by Dailymed and RxNorm. Their valid_start_date is in 2014. We refurbished the way the dates are set, so you can trust them.
However, the last two digits in your list are 04, 08, 40, 43, 44, 45, 56, 48 50 and 53. I have no answer what these are, and whether they are bogus or not. We collect NDCs now from RxNorm, FDB, Medi-Span and Dailymed. Usually, only OTC (cough syrup, multivitamins) and non-drugs (glucose strips) are slipping through. Google doesn’t know them either, which is usually a strong signal.
This changing a paradigm for us, so need to make sure we get it right . . . when should we be using VALID_START_DATE/END_DATE . . . in all SOURCE_CODE cases or for only certain SOURCE_CODES.
It really is only relevant for codes where the date is relevant because the code is reused. Which is only NDC and DRG. The others are only used for internal life cycle managment, so we know to make a deprecated code non-standard and create a “replaced by” relationship.
just pasting the examples from the OHDSI wiki here
There is a problem that NDC can update the vocabulary using the same concepts for a different drug concepts: in the 18/12/17 release we found out that: