Open gkennos opened 8 months ago
@gkennos: If you are looking to realize a "pathologically confirmed XYZ" this would work if the path lab provided a TNM report. If it didn't, or it didn't get recorded, or it is not applicable (like for hematological conditions), you would not catch that.
I would look for a condition_type_concept_id=32835.
Still wouldn't hurt to have your hierarchical concepts.
Sure, but these aren't from path reports. They are recorded diagnoses with explicit stage, and a stage-type flag to select either clinical or pathological, so the valid type concept is 32840
Not following. Pathologically confirmed diagnoses are not from path labs? Where else?
An EHR problem list is something that the patient declares to the provider. Has nothing to do with cancer diagnoses.
And again, no problem with those hierarchical concepts.
Thanks @gkennos for putting together that list. This is definitely an upgrade and something we missed on our upcoming proposal about refining the hierarchies of tumor findings, specifically around staging & grading. Adding this to that ticket.
Folded into the overall stage/grade hierarchy improvement discussion here: https://github.com/OHDSI/OncologyWG/issues/558#issuecomment-2000164032
To support the use case of querying condition occurrences that have pathological confirmation only, we need to insert in the staging hierarchy a top-level 'AJCC/UICC pathological stage' and 'AJCC/UICC clinical stage' term.
csv attached has proposed list for existing concepts that would be direct children of these new concepts
needs_parent.csv