OHDSI Home | Forums | Wiki | Github

OSM Vocabulary?

In a recent download of vocabularies from Athena I found the OSM vocabulary as part of the distribution, however I’m not requesting the OSM vocabulary, in fact it doesn’t appear in the list of vocabularies to choose from as such, I’m confuzzled. :slight_smile:

Hi Frank,

OSM carries the internal flag “OMOP required” and as such will always be included in the download bundle. Other examples are the Type Concepts or the UCUM vocabulary.
However, why it is indispensable to have it, we should ask @Christian_Reich .

That’s interesting, I didn’t know there was a OMOP required list that implicitly included content in the vocabulary list. Could we get a list of what all those vocabularies are? A different approach could be to have those vocabularies checked off by default in Athena but still allow you to remove them?

I also have an Athena request related to this topic :slight_smile:

If a vocabulary contains standard concept_ids which are mapped from non-standard concept_ids, can we flag these, too?


We need to update this feature. We should include all standard and classification concepts as a default (including OSM), and source concepts by vocabulary. Time for an Athena revision.

1 Like

What’s the best way to track the required changes? Github repo for Athena?

Yes, Sir.

Should I be posting in:


Hi there!

I created an issue in Athena (not AthenaUI, which is about the Athena frontend), although it is maybe even debatable if that is not more a vocabulary issue. But yes, it affects the vocab download through Athena, so: Athena repository.

Thanks, Mik.

But we should do some more thinking. The problem is that vocabularies are no the right thing for the auto-included. We should have all standard and classification (unless proprietary) concepts included. They are the official OMOP standard of how to represent data. That should not show up in the selection page.

The selection page should contain the variable stuff. These are all the non-standard and non-classification concepts. And yes, it makes sense to organize them by vocabulary. I would suggest the pre-checked vocabularies are the ones that also contain standard concepts. For example, RxNorm should be pre-checked, so Brand Names are added (even though they are not needed for a valid OMOP CDM). Same is true for NDC (whose devices are standard). The user then only has to pick vocabularies that are explicitly source (e.g. Gemscript).


@Christian_Reich @mik , I’m fine with requiring certain vocabularies to be downloaded, particularly non-proprietary standard concepts, since they are expected to be available for every CDM instance. But, regardless of whether its required or not, our download portal should list all contents that are going to be downloaded. The user should be aware of what is expected to be contained within the download so they can verify it when they upload it into their own environment, and there shouldn’t be ‘surprises’ of vocabularies that were not listed on the website but appear in the download.

Definitely. We need to overhaul the download page, so it is transparent.