I'm running a propensity score analysis for a home health program where we're matching treatment to control members on specific dates with YYYYMM format (eg, 201807), among other variables like gender and age.
1. Is there a feature in proc psmatch that limits the number of times the same control cohort member (identified by Member ID) can be matched to treatment members? The existing options in the MATCH statement: kmax, kmean, kmin, ncontrol don't prevent the same control member being matched over and over to multiple treatment members, do they?
2. If multiple records share the same Member ID will psmatch still perform matches (and can we limit # matches as in Question 1) or does each line have to have a unique ID?
