sample size calculation for binary outcome with 2 year follow up

[jenim514]

He would like to compare these results to the sample size required if the proportion of patients experiencing stroke or any type of cardiovascular death is used as the primary response variable. With 2:1 allocation, alpha=0.05, 2 years followup, how many patients must be recruited to the trial to have 90% power to detect inferiority if an event rate up to 5.5% over 2 yrs is considered noninferior to the expected 4% rate?

...the experimental group is 2x  more than the reference group.

I'm not sure if the 2 year follow up is an important factor here since the outcome is binary.  Here is what I have...is this correct?

proc power;

twosamplefreq groupweights=(1 2) proportiondiff=0.015 refproportion=0.04 alpha=0.025 power=0.9

   test=FM sides=1 ntotal=.;

title "Sample Size Calculation for Comparing Two Binomial Proportions (Noninferiority)";

[Rick_SAS · Accepted solution]

The article "What sample size do you need for a binomial test of proportions?"

provides a discussion of PROC POWER for a test of proportions. It also contains a DATA step simulation that you can modify to check whether your PROC POWER computation matches the estimates from the simulation.