For the ICBO 2019 paper, I instantiated the OMOP condition_occurrence table like I instantiated PDS MDM DIAGNOSIS: each row is the output of an encouter, is about some patient, and mentions some code.
However... condition_occurrence has a non-null required condition_start_datetime column and a null-allowed condition_end_datetime column. This seems like more of semantic commitment on OHDSI's part. Should we therefore treat these rows as processes that the patient participated in? Or stick to the position that they are codes that were recorded and that they may or may not mean anything about reality.
For the ICBO 2019 paper, I instantiated the OMOP
condition_occurrence
table like I instantiated PDS MDMDIAGNOSIS
: each row is the output of an encouter, is about some patient, and mentions some code.However...
condition_occurrence
has a non-null requiredcondition_start_datetime
column and a null-allowedcondition_end_datetime
column. This seems like more of semantic commitment on OHDSI's part. Should we therefore treat these rows as processes that the patient participated in? Or stick to the position that they are codes that were recorded and that they may or may not mean anything about reality.