can we calculate sample size in clinical trials based on history by using the mean data (mean improvement and mean SD) for a number of studies ? or it must be obtained from one study?
Please if one can provide me with a reference also for simple sample size calculation