OHDSI Home | Forums | Wiki | Github

RxNorm[Ext] and SNOMED


(Vojtech Huser) #1

Emtricitabine has this standard term http://athena.ohdsi.org/search-terms/terms/1703069 (an RxNorm)

It has mapping to ATC, dm+d (uk), MeSH, GGR (belgium)

There is also a mapping to SNOMED via relationship ‘SNOMED to RxNorm equivalent (RxNorm)’ - to this SNOMED term http://athena.ohdsi.org/search-terms/terms/4224101

My claims is:
Term http://athena.ohdsi.org/search-terms/terms/4224101 is incorrectly standard and should be made non-standard (since it duplicates another standard term). For Data Quality script checking validity of drug data (and proper value sets in cohort definitions, this is important)

Or, is the inconsistency a sign that at some point we may jump ship to SNOMED for standard ingredients? I like that SNOMED will maintain drug classes for the next 10-20 years. (e.g., Has disposition (SNOMED) Reverse transcriptase inhibitor 42536029 SNOMED )
Also, it may cover RxNormExtension ingredients (does not have just single country focus like RxNorm).

More on this topic is in this paper from late 2019:


My 2018 related work is here https://drive.google.com/file/d/1zzWaH78iV7prtyyBQzawmfiMrtGu_6Hx/view


(Anna Ostropolets) #2

I’d definitely agree that these should become non-standard. I can easily imagine that one wouldn’t pay attention and would select SNOMED as their drug vocabulary just because of string match. @Vojtech_Huser could you please create a GitHub issue?

In terms of SNOMED becoming a standard drug vocabulary - maybe they have a trick up their sleeve to make their drug representation better, but so far RxNorm has been doing much better job in terms of consistency, structure and reliability.


t