Phase III Trials
Phase III trials are comparitive trials, they may compare effective treatments found in phase II with currently accepted treatments or a placebo. The sample size for a trial is determined using an estimate of the treatment difference and desired power of the study (preferably 80% or greater). In order to minimise any bias in treatment allocation, patients should be randomised to the treatment arms. Confidence in the findings of a trial will increase with sample size as a result, multi-centred trial are strongly encouraged.
Further details are given (among other places) in Friedman et al.
