BookmarkSubscribeRSS Feed
El-Aura
Calcite | Level 5

Hi there! 

 

I'm trying to replicate a two-sample survival design using proc seqdesign. It has 2 stages (one interim at 73% of the information, one final at 100% information). One-sided alpha=0.025 power=0.835 (beta=0.165). O'Brien Fleming boundary will be implemented for efficacy and the Haybittle-Peto boundary will be implemented for futility. Assumed hazard ratio=0.58. Median survival time in the control arm is 42 months. Here're my codes:

ods graphics on;
proc seqdesign stopprob errspend boundaryscale=pvalue;
OneSidedOBFandHP: design alt=lower alpha=0.025 beta=0.165 nstages=2
method(reject)=obf method(accept)=hp(Z=0.489)
stop=both(betaboundary=nonbinding) info=cum(0.73 1);
model=twosamplesurvival (hazardratio=0.58 nullmedsurvtime=42);
ods output Boundary=Bnd_Time;
run;
ods graphics off;

But the results (stopping p-value boundaries) are not the same as the design I'm trying to replicate which has boundaries like:

 

El-Aura_0-1619470168871.png

 

A friend said that the problem is with "alt=lower" since for futility, alt should be "upper". But you can't specify two alts in seqdesign right? So is this the problem? How should I modify my codes to get the design I want to replicate (the screenshot above)? 

 

Many thanks!!!

hackathon24-white-horiz.png

The 2025 SAS Hackathon has begun!

It's finally time to hack! Remember to visit the SAS Hacker's Hub regularly for news and updates.

Latest Updates

What is ANOVA?

ANOVA, or Analysis Of Variance, is used to compare the averages or means of two or more populations to better understand how they differ. Watch this tutorial for more.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 0 replies
  • 806 views
  • 0 likes
  • 1 in conversation