Sample size calculation for a non-inferiority trial with POWER procedure

[Yasu]

Dear all,

I'd like to ask you about sample size calculation for a non-inferiority trial with POWER procedure.

What is the formula for calculating samplesize used in POWER procedure?

Sample SAS code is as follows:

PROC POWER;

TWOSAMPLEFREQ

TEST = PCHI

SIDES = U

ALPHA = .025

GROUPPROPORTIONS = (.75 .80) 

NULLPROPORTIONDIFF=-.10

POWER = .8

Ntotal = .

;

RUN;

Thanks in advance,

Yasu

[AndreyMyslivets]

Hello Yasu!

One of the best books about it is "Chow, S.C.; Shao, J.; Wang, H. 2003. Sample Size Calculations in Clinical Research. Marcel Dekker. New York."

So your formula looks like:

n = (Z_α/2+Z_β)² * (p₁(1-p₁)+p₂(1-p₂)) / (p₁-p₂-NImargin)²

Z_α/2=1.96

Z_β=0.84

better late than never

Andrei