05-15-2016 02:58 PM
I have a large dataset where patients were measured multiple times on various variables (so each patient has multiple lines of data), with binary outcome variable (outcome not the same for each patient visit) and I had used NLMIXED for the repeated measures analysis. I am trying to figure out how to do pairwise comparisons (with Bonferroni correction of p-values, if possible) in SAS, other than to manually calculate. I looked at the information for PROC MULTTEST, but it doesn't seem to give me an option for the repeated measures aspect of this study.
Has anyone had experience and/or insight in how I might accomplish this?
Any assistance you can provide would be GREATLY appreciated! Thank you!
05-15-2016 04:29 PM
You can perform your statistical tests with whatever procedure is appropriate, output the p-values with ODS, and then use proc multtest to do the p-values correction. Look at option inpvalues=dataset.