Methodology required to show clinical differences between benzodiazepines.

Clinical differences exist between the benzodiazepines but demonstration of such differences requires a more specialized methodology than that normally used in comparative trials. It is recommended that double-blind studies should be carried out in hospitalized patients with severe and chronic anxiety, selected according to precise criteria, and that the trials should be designed as crossover rather than as parallel group studies, with randomization of the stages and flexible dosage. A simple graphic method of representing the clinical profile of individual benzodiazepines is described and it is suggested that this could help clinicians adapt their prescribing to each patient's symptoms.