The Vocabulary Team is proud to share with you that with the Summer 2022 release, finally another major release of the OHDSI Standardized Vocabularies has been made available. @Alexdavv has carefully compiled a list of changes in the release notes which is worth studying equally carefully. There are many improvements that would also very likely affect the ETL logic in some projects, so there might be an impact when making use of the new content that would require adjustment of the local ETL logic accordingly.
Most notably, we have finally newer versions of SNOMED. As discussed previously, owing to the collection of multiple versions and the conversion as well as adjustment process to other vocabularies, we always face a bit of a delay which means the International SNOMED version this is based on is already one year old, while the other two are a bit more recent.
Due to popular demand, we have also refreshed HCPCS and CVX, as well as many condition vocabularies in the ICD family.
This time we have also added some sort of “known issues” section that explains shortcomings we found during the process but had to postpone solving them as this would have been too large an effort for the time being and would have delayed the release further.
Enjoy the new content and let us know if you find an issue in it. The first one to find a considerable bug (e.g. wrong mapping, typos do not count) in new content (old bugs also do not count) and creates a respective github issue, gets a pack of TimTam chocolate cookies delivered to their doorstep. Take my word for it!
We are also planning to collect vocabulary changes that didn’t make it into this release (HemOnc, OPS, more Cancer Modifier content) for an Indian Summer release, maybe even a smaller technical release before the Symposium.
Excellent release notes and forum posting, @mik and @Alexdavv!
Those of us with EHR data, might want to consider how the following change to the domain designations in HCPCS might affect our ETL, if at all.
Thank you @MPhilofsky!
The actual Domain change of non-Standard (from now) concepts should not affect anything, because only the target concept’s Domain matters (it defines the destination table).
On the other hand, it looks like most of the ETLs have to add the rule for HCPCS concepts to end up in the Visit/Provider-specific tables. But if you still don’t have a rule, the change of the Domain of the source concepts may be painful because the Domain is the same as for a target, so you anyway need a rule.
Maybe COALESCE (…, Observation) would help as a temporary solution.