Open comorbidity opened 10 months ago
then SQL would be like
create table $tablename as select
For examples of this syntax, see https://github.com/smart-on-fhir/cumulus-library-suicidality-icd10/blob/main/cumulus_library_suicidality_icd10/suicide_icd10/define_dx_attempt.sql
Hmm, the directionality here is a little interesting. Let me see if I understand what you're asking.
You want a way to easily pass around the labels from the label studio export. You basically want a trimmed down version of the label studio export.
You mention ICD10, but I don't think this ask makes sense in that specific context -- ICD10 labels would be coming from Athena - it doesn't sound useful for chart-review
to package those up as a CSV because it ingests those labels as a CSV already from a manual Athena export.
But this ask does make sense for the Label Studio export, which is a bit gross to work with and is not in Athena. You basically want to create a CSV you could easily run queries on instead of the big ol' json blob that LS gives you. Maybe ingest it back into Athena (so no original IDs then, just the anonymized ones).
Research use case example (Suicidality)
We want to know how many charts were annotated for
ideation-past
but no other label. For IC10, this will always be "zero" because ICD10 has no label forideation-past
.To perform these types of detailed analysis we need to be able to query on the level of chart specific answers. CSV file example columns
chart_id, docref_id, annotator_id 1001, aaaa-bbb-ccc-etc, andy, ideation-past 1001, aaaa-bbb-ccc-etc, rena, ideation-past 1001, aaaa-bbb-ccc-etc, icd10, None (or Null)