Thanks for merging these threads.
Theoretically, the FU and MED would be linked to the earliest non-stale (30 days) OP date, because potentially this is the participant's first of several OP encounters.
However, in the absence of a reasonably straightforward solution, the closest OP to FU date is acceptable.
To address your previous questions:
1. A DATE_FU can equal or follow a DATE_OP.
2. A DATE_MED cannot precede a DATE_FU, but can be linked to the same DATE_OP, if for instance, they are all on the same date.
Generally speaking: OP <= FU <= MED
3. Answered above, but to reiterate, DATE_FU & DATE_MED are linked to the first non-stale OP instance, but the closest preceding OP is allowable.
@mkeintz wrote:
Must a DATE_FU actually follow a DATE_PO, or is a link between equal dates allowed?
Can a DATE_MED precede a DATE_FU and be linked to the same DATE_PO?
The first DATE_FU in your sample data could be linked to either the first or second DATE_PO. It appears from your sample WANT outcome, that you have a rule to match a DATE_FU to the earliest preceding DATE_PO, even if there is a closer preceding DATE_PO. Is that correct?
... View more
