OHDSI Home | Forums | Wiki | Github

Starting ATC Mapping Working Group

Hi All,

You might have experienced the situation where you needed to group your drug products of interest to a higher level class, and decided to use the OMOP vocabulary relationships to ATC classification, but either the vocabulary lacks the correct relationship or the mapping does not exist. A simple join of ingredient to ATC 5th level may not also work well in some cases because ATC classes are not only about active ingredient name, but also dosage form, drug strength, route of administration, and other characterizations of the product. It has been discussed for a long time in emails, on the forum (#800, #2815, #3556) and the GitHub repo (#114, #121).

To address this issue, we would like to initiate the ATC Mapping Working Group to, once and for all, comprehensively assess and fix ATC relationships in the OMOP vocabulary. The goal is to develop the best approach to map concepts to ATC classes, and provide ATC mapping for all Drug concepts. For more information, visit the working group page on wiki.

We welcome everyone to join this effort. If you are interested, please leave a comment below, and fill out the doodle poll by April 27, 2018 to schedule the first meeting.



Thanks for this initiative. I am interested, but more from the reverse perspective (ATC to RxNorm). In most European data systems, the ATC code is given in the source data and needs to be mapped to RxNorm ingredients. I assume that this is also part of this WG.
Will complete the Doodle.


I think we are talking all about the same thing. The goal is to have for each RxNorm or RxNorm Extension the correct ATC classes connected. Both directions.

1 Like

Maxim, thanks for your interest. Yes, the ultimate goal is to provide bidirectional mappings between ATC and RxNorm/RxNorm Extension concepts. Once we get this part done, any drug source code can be mapped to ATC if the source-to-standard relationships exist.

1 Like

Thanks!! I’m interested, too.
In Korea, most of the drug codes are mapped to ATC code by National Health Insurance Review & Assessment Service for statistics. Though they are “Korean drug code to ATC relationships”, still could be helpful for “RxNorm to ATC” somehow.
I’d like to join.

Sure, thanks for joining!

We just finished wrestling with ATC code mappings for a recent analysis. There is a lot of value in ironing this mapping out. Please sign me up for this WG. There is one other person I will encourage to join as well who has performed a lot of exploratory analyses with ATC codes.

Hi Corina, you are in! :smile: Thanks for joining. Your experience will benefit the team for sure.

And meeeeee.

Only on a phone right now but our practice performance and safety work is
at the provider and practice levels. We do keep and use NPIs.

Wilson D. Pace
sent from phone

I’m interested as well. We were just this week attempting to use the ATC for these kinds of things!

I am interested as well.

Thanks everyone for your interest in joining the workgroup. I looked at the doodle list, and May 4th works for all but @Christian_Reich.

@Christian_Reich, is there any chance that you can join on May 4th at 11am-12pm or 12-1pm?

We will meet tomorrow (5/4/2018) at 12:30pm EST. Here is the Skype information to join the meeting: http://www.ohdsi.org/web/wiki/doku.php?id=projects:workgroups:atc

I look forward to talking to you all soon.