Matched pair analysis for categorical variables

[#1]

I have a set of patients who revived two different drugs at different times. 

Based on their creatine changes, I then have proportion of patients who had their category of kidney failure changed ( n=x after drug A and n=y after drug B)

It is easy to compare the mean creatinine change with the two drugs by paired T test, but if I want to compare the proportion of patients, whose kidney failure category changed..what would be my option ?

thanks for any guidance..

[#2]

A matched pairs two-sample t-test is just a one-sample t-test of the difference between (after - before). So you might create a category for each possibility: lower, same, higher, for example. You can now use any categorical analysis method with your factors.