Stat Computing > Seminars > Stata10

What's New in Stata 10: Power Analysis for Survival Models

stpower is a suite of Stata commands for sample-size, power, and effect-size determination for survival analysis. Its basic syntax looks like this:

• stpower Cox --- Cox proportional hazards models
• stpower logrank --- non-parametric models with proportionality assumption
• stpower exponential --- parametric models, comparing two exponential survival curves

Components that will complicate the matter of power analysis for survival models include censoring and patterns of accrual time and follow-up time  
 
Below are examples for some really simple-minded power analysis assuming no censoring.

stpower cox, hratio(3) sd(.75) power(.95) alpha(.001)

stpower logrank .2, hratio(.6667) power(.9)

stpower exponential .45 .67, t(2)

quietly stpower logrank 0.5, hratio(.01(.01) .99) n(100 300) saving(powplot, replace)
use powplot, clear
twoway (line power hr if n==100) (line power hr if n==300), ///
legend(label( 1 "n=100") label( 2 "n=300")) xtitle(Hazard Rate Ratio)

UCLA Researchers are invited to our Statistical Consulting Services
We recommend others to our list of Other Resources for Statistical Computing Help
These pages are Copyrighted (c) by UCLA Academic Technology Services