Power for sequential design with non-inferiority outcome

[bryony1]

Hi,

Is it possible to use PROC SEQDESIGN to calculate the required power/sample size for a sequential design where the primary outcome is tested under non-inferiority?

For example, if the event is expected in 30% of the control & the non-inferiority margin is 5%. How would I program a sequential design over 3 stages?

Best wishes,

Bryony

[ChrisNZ]

Moved to procedures forum where the chances of a reply are higher.