4 Replies Latest reply: Jan 10, 2018 10:14 AM by Stephen Koscumb RSS

    Using set analysis + aggregation

    Stephen Koscumb

      Hi all,


      I'm trying to perform a function on a column, but only where certain conditions are met.


      The example is a healthcare dataset -- I have ~100,000 unique patients, and I have ~700,000 unique encounters with those patients.


      I've created a variable called "VISIT_ORDER," which is simply a ranking of the patients visit sequence.  I also have a variable which specifies the patient's dialysis status on the encounter level: whether the patient is not currently on dialysis, if they currently are on dialysis, or if they've never been on dialysis during the time window.


      I'm trying to average a column with the first visit a patient has (my thought was minimizing VISIT_ORDER) after they've been on dialysis.  In my set analysis, I want to return the result of the row where VISIT_ORDER = the minimum VISIT_ORDER where DIALYSIS_DX_GROUPING = 'AFTER DIALYSIS', then to average COUNT_ADMITS_365 for those rows.


      I think what I need to do is some variant of this below, but while what I have works for one patient (when I've been troubleshooting), I'm having issues:





      The issue I'm having I believe is with the dollar-sign expansion, and also why the function appears to work with a single patient (PAT_ID is the unique patient identifier).

      Any help would be appreciated; driving myself nuts over here.