Randomized controlled clinical trials mostly focus on very specific outcome parameters. These may include symptom relief, psychosocial measures, specific safety issues or compliance, just to name a few. As they often represent early attempt to provide information on new treatments, the homogeneity of the studied population is a crucial study prerequisite. This generally calls for strict inclusion criteria and a large set of exclusion criteria. Understandably, these requirements allow only a certain selection of patients to enter such studies, which, in turn, jeopardizes the generalisability of the obtained results. Alternatives to this approach include so called “large pragmatic clinical trials” with broad inclusion criteria, designed to study a population of patients closer to real life. More comprehensive outcome criteria, such as the effectiveness or remission paradigms, have also contributed to the effort. In the end, results from various types of clinical trials will have to be evaluated in a synthetic fashion in order to enable the clinician to make a rational treatment choice for individual patients.