I'm quite surprised that he said that. That stat
Post# of 148168
If we have 30% of the deaths in the treatment arm, and 70% in the placebo arm that equates to 13-14 deaths in treatment and 31-32 deaths in the placebo.
This would represent a decrease in mortality of 77.5%,which would be undeniable evidence of efficacy.
With that said, I dont think its reasonable to expect 31-32 deaths in the placebo arm, as that would be 50%.
Again, not sure why he would make a statement like that.