The scientific integrity of a clinical trial and the credibility of the data produced depend primarily on the design of the trial. In the case of comparative trials, the protocol should therefore describe:

- an a priori rationale for the targeted difference between treatments that the trial is designed to detect, and the statistical power to detect that difference, taking into account clinical and scientific information and professional judgement on the clinical significance of statistical differences;

- measures taken to avoid bias, particularly with regard to the randomization, when relevant, and selection of patients.