I see, so you entered 0.7/0.3 as p to account for the expected mortality in placebo? Sorry, can't explain it any better.

I just think one needs to use a two independent proportions test, not a simple binomial...

https://www.statsdirect.com/help/proportions/unpaired.htm

https://www.google.com/url?sa=t&source=we..._hbSwf8eKU

Will need to wait for others to see which calculator is correct...