r*****y
发帖数: 199 | 1 比较treatment和control两个group，每个group有estimated event rate（可能来源于
以前的数据），hypothesis是treatment比control的event rate少B%（我是理解成
treatment有30个死亡，control有40个死亡，这就算是少25%的死亡率。），这个event
是属于survival里面的time to failure event，一共trial的时间是A年。这个sample
size per group应该怎么来算呢，比如用SAS的话。 | R*********i
发帖数: 7643 | 2 In general a treatment difference of 25% means the expected event rate in
group A - event rate in group B = 25%, not the event counts in Group A /
Event counts in Group B = 75%, unless you always have 1:1 ratio in evaluable
subjects from both groups. Have you tried proc power w. twosamplesurvival
statement? You need a little bit more information/assumptions. |
|
