I am working on implementing the proposal for the NOTE_NLP extensions in my (CDM 5.4) data warehouse. I love the concept, but am struggling with what may be a basic lack of understanding.
I understand that nlp_event_field_concept_id is pointing to (for example) condition_occurrence.condition_concept_id. But how do I build a query that says, "Select all the rows where the nlp_system was 'CTAKES' and get a count of person_id and domain"
My instinct is to try to join NOTE_NLP on this polymorphic key but I just can't quite get my head around how to work backwards from the event_field_concept_id to get there. Do I have to start from the clinical tables that interest me and then join on row_id and filter on what I know to be the correct nlp_event_field_concept_id?
I'd really appreciate a little advice on how to use this kind of structure most effectively.
I am working on implementing the proposal for the NOTE_NLP extensions in my (CDM 5.4) data warehouse. I love the concept, but am struggling with what may be a basic lack of understanding.
I understand that
nlp_event_field_concept_idis pointing to (for example)condition_occurrence.condition_concept_id. But how do I build a query that says, "Select all the rows where thenlp_systemwas 'CTAKES' and get a count ofperson_idanddomain"My instinct is to try to join NOTE_NLP on this polymorphic key but I just can't quite get my head around how to work backwards from the
event_field_concept_idto get there. Do I have to start from the clinical tables that interest me and then join on row_id and filter on what I know to be the correctnlp_event_field_concept_id?I'd really appreciate a little advice on how to use this kind of structure most effectively.