Here is a binomial p value calculator. Im not an
Post# of 148177
n= the number of patients in the treatment arm (34)
k= the number of patients who survive. If 90%, then 30-31
p= the percentage of patients who survive in the control arm (if 67% then input .67)
inputting 30 for k, nets a p value of .00333
A few scenarios to meet p<.05:
Control Arm Treatment Arm p value
13 survive (76.4%) 33 .0011 (32 survivors fails p value)
12 survive (70.5%) 28 .04976
11 survive (64.7%) 27 .0288 (28 fails barely)
10 survive (58.83%) 25 .0309
https://www.socscistatistics.com/tests/binomi...ault2.aspx
Assuming I am doing these calculations correctly (big assumption), this study really comes down to the mortality rate of the control arm.