Open PhilAppleby opened 5 months ago
Worked examples or links to such are required - we generally agreed we need to handle this in the move away from such a narrow Cohort Discovery focus. But developers will need a better understanding of the desired logic. May also require changes to the JSON format, and/or CaRROT-CDM to handle the outputs.
@spco Re: Automated 1:n Mapping-> I have added examples. We may have to discuss this further in detail and POA one to many Mapping concepts .docx
Additional json rules file lines in in the "value_as_concept_id" section:
"FH: Cardiovascular disease 16974": {
"observation_concept_id": {
"source_table": "gs_cases_icd10.csv",
"source_field": "processed_code",
"term_mapping": {
"Z82.4": 4167217
}
},
"observation_datetime": {
"source_table": "gs_cases_icd10.csv",
"source_field": "case_date"
},
"value_as_concept_id": {
"source_table": "gs_cases_icd10.csv",
"source_field": "processed_code",
"term_mapping": {
"Z82.4": 134057
}
},
"observation_source_concept_id": {
"source_table": "gs_cases_icd10.csv",
"source_field": "processed_code",
"term_mapping": {
"Z82.4": 4167217
}
},
"observation_source_value": {
"source_table": "gs_cases_icd10.csv",
"source_field": "processed_code"
},
"person_id": {
"source_table": "gs_cases_icd10.csv",
"source_field": "id"
}
},
Ability to map concept_id, value_as_concept_id, value_as_number to a single record