ABSTRACT
A controlled clinical trial is an experiment in which one or more treatments are compared with a control treatment, which can be nothing, a placebo or a standard clinical practice. It is necessary when experience alone does not provide sufficient evidence for selecting the best course of clinical action. A clinical trial protocol contains, an introduction, which includes the main references related to the study background; a precise formulation of the questions; randomisation method; statistical methods; and so on. The power of the trial is a measure of the sensitivity of the trial in detecting an actual difference. It is equal to the type II error. In crossover designs the patient acts as his or her own control. Several treatments can be compared, the order of which should be random. Randomisation is a method of assignment of subjects to either experimental or control treatments, whereby each patient has an equal chance of appearing in any treatment group.
