Posting: # 18339
Hope everything is fine with everyone here.
I am just wondering if anyone can help me in how to calculate the sample size and power for clinical trials (CROs) as I want to know the basis for calculating sample size for assessing safety and efficacy of one test drug compared to control.
Any equations, guidelines, references, software recommended would be an asset.
@ Mohamed Yehia
Posting: # 18352
» […] how to calculate the sample size and power for clinical trials (CROs) as I want to know the basis for calculating sample size for assessing safety and efficacy of one test drug compared to control.
That’s a terribly wide field where different hypotheses are tested (time to event / survival, dose-response, superiority, non-inferiority, equivalence,…) based on different types of data (discrete: nominal and ordinal scale, continuous: interval and ratio scale), different distributions (normal, lognormal, binomial,…). There’s not enough space here to list the number of statistical models/tests. All them them require different methods for sample size estimation.
» Any equations, guidelines, references, software recommended would be an asset.
See the respective sections in ICH E8 and E9.
Be warned about Chow et al.1 Although the book gives a good overview it contains a lot of typos. Always cross-check the formulas with other references. Julious2 is a nice appetizer. My favorite is the comprehensive book by Julious3 which covers practically everything. If you need something else, consult with a qualified biostatistician.
Due to the high variability and/or small effects in phase III sample sizes generally are so large that approximations by the normal distribution can be used. Makes life easier.
The quality of responses received is directly proportional to the quality of the question asked. ☼