TechGuru. Looks as if my math is upside down from
Post# of 148185
87 deaths total in s/c trial so far.
It looks like you used the 50/87 deaths for the leronlimab group.
And I used the 50/87 deaths for the placebo group.
If Leronlimab 50 deaths out of 260 patients = 19.2% died (your math)
If placebo group 37 deaths out of 130 patients = 28.5% died (your math)
Our earlier eIND death rate was about 14.5% (your math has 19.2%, seems if any change, death rate for leronlimab would go down, due to better SOC, not up as in you math)
14.5% of 260 is 38, so the 37 deaths are assumed to be in leronlimab group.
50 deaths in placebo group divided by 130 = 38.5% (does seem high)
37 deaths is Leronlimab gours divided by 260 = 14.2% (similar to prior eIND)
Reduction is from 38.5 down to 14.2
38.5 - 14.2 divided by 38.5 = 63% reduction in death,
p value approx = 0.0001
We don't expect such great numbers as above, but using our prior eIND data, I believe that is what it is. Even if the above is way off, probability of acceptable p value is quite high.