Hello,
I would like to do a sample size calculation of a two sample survival analysis.
My event rate is 6% a month, the drop out rate is 2.5% a month. Hazard ratio is 0.5. The study lasts for 25 months. Power is supposed to be 0.8.
I have written the following program:
proc power;
twosamplesurvival test=logrank
hazardratio = 0.5
refsurvexphazard = 0.06
grouplossexphazards=(0.025 0.025)
accrualtime = 0
totaltime = 25
ntotal = .
power = 0.8;
run;
I am a little irritated because I have constant hazard rate and drop out rate, but SAS demands for exponential hazard rates. Does the “exp” in refsurvexphazard and grouplossexphazard demand for other hazard rates than constant or is it right like I did it?
Thank you very much!