Designing long-term clinical trials in affective disorders

Long-term clinical trials in affective disorders have become increasingly important, but their methodology continues to impose a major challenge. An ideal design can be developed but is difficult to implement. The advantages and disadvantages of different designs (mirror image, double-blind study controlled with placebo, and/or standard and discontinuation studies) are described and compared. Shortcuts in design usually carry a high risk of false-negative findings and result in a waste of resources. An abbreviated version was presented at the European Consensus Conference on Long-term Clinical Trials in Zurich, Switzerland, October 1992.