I find it incredible that Cytodyn would execute a trial which would succeed on p<.05 at 33% mortality reduction and then fail because the "power" statistic was not met.

I can't believe such a massive mistake would be made.

Sure, trial enrollment is difficult but there is no point in holding a trial if you know the math stats will fail.

And the same goes for the DSMC at the interim analysis. If they know the "power" will fail even if the mortality reduction is 33%, what the heck are they doing saying "No recommendations for changing the trial. Proceed to the endpoint as planned."

If you know the "power" number will cause a fail, shouldn't you increase the number of trial participants?

Maybe your decision was colored by trial enrollment difficulty. Well, then when the pace of enrollment picked up recently the DSMC should have spoken up to increase the number of trial participants.

What gives?