09-24-2013 10:38 PM
I'm looking for the right commands for a comparison of two matched proportions for a non-inferiority trial. The statistical test I need would be comparable to a McNemar's test, but I need to specify the non-inferiority margin. Anyone have any ideas?
02-10-2015 08:51 AM
Recently, I am working on the Non-inferiority test based on the difference of two proportions, I used the Proc FREQ riskdiff option in TABLES statement. YOu can specify the margin as an option, not sure if this works for your question.