Adding to the RXNORM vocabulary

Implementers
1

While going thru implementation of one of our US datasets, we noticed that the following drug: C9295 did not have a RXNorm concept id assigned to it. How long does it take for a new drug to get added to the RxNorm concept vocabulary after it is launched ? We are using Version 4.5.

Thanks,
Urvi

2

Hi @ushah

I think the issue is with J9047. It has a HCPCS to HCPCS mapping and a mapping type of “drug”. But there is no HCPCS to RxNorm mapping. C9295 has both a HCPCS to HCPCS and HCPCS to RxNorm mappings (at least on the version we are looking at).

Just clarifying this so that someone can verify that the problem still exists. And maybe @Christian_Reich can comment on the update cycle. With all these vocabs, I don’t know how he keeps everything straight.

3

Has there been any discussion about a formal bug/issue tracking system? As we are using the vocab data in validated software, we need to keep track of what issues are identified with which versions, and when the issues are resolved, etc. If a client runs an analysis using vocab version-X, then runs another analysis with vocab version-Y , I need to be able to tell all the differences between analyses runs. For my application I need to maintain a tuple of software component versions, obs-data versions, and vocab versions for each analysis; and have additional metadata about each version.

How has the group dealt with the version-itis issues ?

Thanks,
Don

4

Friends:

The HCPCS to RxNorm mapping will be updated with the next release in a few days.

By building an automated repeatable system. The HCPCS to RxNorm was just added, and it is very messy.

There is a place to send in problems: Issues · OHDSI/Vocabulary-v5.0 · GitHub. However, if people mention them here they will be also taken care of. The real bug tracking/backlog system is internal. Should be external as well, but that would take time and effort to migrate.

OMG. Don. you are brave. Why is your system validated? Usually observational systems are not because there is no control over large portions of the data generation process. Is that different in your world?

Please please bring it up in the CDM WG: http://www.ohdsi.org/web/wiki/doku.php?id=documentation:next_cdm. There, we are working on all change requests and issues that are more than bug fixes. Rimma is running it.