power calculation for cox regression

Manije72 — Wed, 09 Oct 2024 01:40:01 GMT

Hi,

I found an article examining the association between "x" and the risk of cardiovascular diseases (CVD). The study reported hazard ratios comparing "x" to non-"x" in relation to CVD. The sample size for this cohort study was 10,000 participants, with approximately 100 individuals developing CVD after 11.8 years of follow-up. I would like to utilize this data for my research.

As I proceed with power calculations, I am uncertain about the appropriate values to assign for standard deviation (stddev) and R-squared (rsquare). Could you please assist me with this? Additionally, is the following code appropriate for my objectives?

proc power;
coxreg
hazardratio = 2.1
rsquare = ?
stddev = ?
ntotal = 10000
eventprob = 0.0.5
power = .
;
run;

Re: power calculation for cox regression

Manije72 — Wed, 09 Oct 2024 01:55:39 GMT

If I want to use the below code, What should I put for (curve"control) and "refsurvival"?

proc power;
twosamplesurvival test=
curve("Control”) =
refsurvival =
hazardratio =
accrualtime =
followuptime =
groupweights =
ntotal =
eventstotal=
power = ;
run;

topic power calculation for cox regression in Statistical Procedures

power calculation for cox regression

Re: power calculation for cox regression