Help using Base SAS procedures

group sequential design for survival analysis

Reply
Contributor
Posts: 50

group sequential design for survival analysis

I used proc seqdesign to produce interim analysis boundaries. In the boundary plot, the fixed-sample critical values(the small circles) are before first stage, it seems that doesn't make sense, it should be around the later stages. Any idea/experience on this?  Thanks very much!

The code:

procseqdesign plots=(boundary(hscale=samplesize)
errspend(hscale=stage))

               altref=0.45

               boundaryscale=mle

               ;

     Surv4interim: design

               info=cum(0.5 1)

               method=obf

               nstages=2

               stop=both

               alt=twosided

               alpha=0.05

               beta=0.10

               ;

   samplesize model =inputnevents(

                     d = 276

                     sample = two

                     medsurvtime = 16

                     acctime = 24

                     foltime = 18)

                     ;

   odsoutput Boundary=BND_PropGrad;

run;

Ask a Question
Discussion stats
  • 0 replies
  • 150 views
  • 0 likes
  • 1 in conversation