https://ccforum.biomedcentral.com/articles/10.1186
Post# of 148278
25% mortality rate is pretty accurate. Would suggest as you said 16 deaths expected. Each death = 2 (2:1) so 32 deaths in Placebo.
45-16 = 29 deaths possible in treatment group. So 32 placebo versus 29 treatment? Possible, my pure speculation is we are picking up the sickest of the sick in our trials (less exclusion criteria) and have a higher mortality rate....if this is the case have say 32% deaths in placebo we have 21/65 and achieve stat sig.