NuGilead, *Assuming the p value boundary is con
Post# of 148170
Quote:
*Assuming the p value boundary is consistent with this, and we need p<0.005 to get halted, what kind of mortality numbers are we needing LL/control (130/65)
Using your assumption of 23% deaths in placebo (15 patients). If the LL group (130 total patients) has a survival of 106 patients (24 deaths) we will have a p-value of 0.0039.
25 deaths will put us above 0.005. All of the above for a two-tailed test.