I would like to ask about Entities? https://opaldoc.obiba.org/en/latest/variables-data.html#entities
By default, tables in DataSHIELD represent 'Participant’s. So each row is an individual, with some medical record. And the controls are designed around protecting those participants.
In some contexts it might be relevant to work with a treatment-level table instead of patient-level. Let’s say we have a cohort with patient history. I would like to be able to analyze patients that have retained the therapy A for 6 months, differently than those that have switched to therapy B at month 3. As it stands, this would require multiple rows per Participant and is a no-go; DataSHIELD will fire errors when attempting to analyze such a table.
I understand some projects such as dsomop are also trying to address this scenario. But I would like to ask for best-practices in dsBase to represent treatment-level, or even visit-level tables. As I have been unable to find examples for these cases in the Orchestrating privacy-protected big data analyses of data from different resources with R and DataSHIELD book.