What is the no. of subjects required for Phase III trials of a drug/ vaccine ?
CHP: Simple answer is ‘enough”. By that it is meant be the number of subjects to answer the question statistically. This depends on the questions asked and the variability of the responses. I do not know of any hard and fast rule. I will ask my colleague Bala to reply also. Bala: I agree with Charles. For sample size estimation, one needs to consider the type of design used (parallel, cross-over, etc.), type I error rate (usually 0.05), the required power (80%, 90%, etc.) and most importantly variability estimate.
