OHDSI Home | Forums | Wiki | Github

DQD error for CDM_SOURCE table

Hi all,

The OMOP CDM v5.4 user guide for the CDM_SOURCE table states that the CDM_RELEASE_DATE is: “The date the ETL script was completed. Typically this is after the source_release_date.”

The DQD the temporal plausibility check for this table suggests the opposite, however: “The number and percent of records with a value in the cdm_release_date field of the CDM_SOURCE that occurs after the date in the SOURCE_RELEASE_DATE. (Threshold=0%).

In our case, we set source_release_date to “2023-03-31” because that was the latest event data included in our database. We set cdm_release_date to “2024-03-31” because that is when we generated the CDM. This resulted in a FAIL of the temporal plausibility check for the CDM_SOURCE table.

Can someone clarify which date should come first?

Aren’t these the same?

Hi @Alison_Park I was noticing this recently as well. The DQD check is incorrect, the documentation on the CDM is correct. The CDM date should be after the source release date.

@Dave.Barman the source release date is the date the native data were released and the CDM release date is the CDM built off of that native data was released. These could be the same but if they are different the source release date must be first.

Thank you @clairblacketer! That’s what I suspected.

I noticed this recently too! I dug a bit and it seems it’s been around a while though :sweat_smile:

I filed a GitHub issue for the fix here: CDM_SOURCE plausible temporal after bug · Issue #539 · OHDSI/DataQualityDashboard · GitHub

Sorry I was unclear. At any rate, it does seem like a bug if its failing. I see Katy found the bug.

What I meant is the statement for the check: “The number and percent of records with a value in the cdm_release_date field of the CDM_SOURCE that occurs after the date in the SOURCE_RELEASE_DATE. (Threshold=0%). "

Looks correct as its written.

t