- P Glemain: Clinique Urologique, Hôtel Dieu, Nantes, France. pglemain@chu-nantes.fr
Intent to treat methodology consists of keeping all patients included, in their initial group in the case of randomization, to perform final analysis of a study. This attitude is justified methodologically, as it preserves the validity of the statistical rationale, and clinically, as it more closely reproduces real prescribing conditions. In the light of a classical example, this review illustrates several clinical situations raising the problem of intent to treat.