Power analysis without historical data

0

Entering edit mode

8.1 years ago

fischer87 ▴ 10

Hello!

I need your help to compute a power analysis for survival analysis (Kaplan meier) without historical data. In my lab we are preparing a project in which we want to evaluate tumor free survival of two cohorts of mouse (one treated with A and one treated with B); because the experiment is new, we don't know the median tumor incidence time of the two groups.

We want to make sure to detect at least a difference in tumor incidence of 20% between the two groups (using less mouse as possible). I fixed alpha=0.05 and power=0.8. How can I calculate the required sample size (using R or STATA)? How can I calculate the hazard ratio??

Thank you in advance for your help! .. and sorry for my english errors

R • 1.7k views

ADD COMMENT • link updated 7.3 years ago by Biostar 20 • written 8.1 years ago by fischer87 ▴ 10

0

Entering edit mode

Erm, not sure if bioinformatics. Other mods, input please.

ADD REPLY • link 8.1 years ago by Ram 43k

2

Entering edit mode

I would call it biostatistics, then it is on topic I guess, otherwise there wouldn't be many alternative QAs to post this, maybe https://stats.stackexchange.com/ could work too, but then I think it's ok here.