Ethics and errors of clinical trials and the role and place of different types of trial.

Controlled clinical trials, usually carried out in a multicentre, randomized manner, have become fashionable and popular because they contribute to the general acceptance of the results, particularly when the difference between the treatment arms is small. The present book and the present introduction attempt to discuss when and how this form of trial is appropriate. It is recognized that the discussion may be one-sided in the sense that the undoubted and well-known merits of controlled clinical trials are not reiterated. The focus of the present discussion is on the scientific, practical and ethical problems inherent in such trials and on possibilities for improving their design. In the future clinical trials should be used with more restraint than hitherto. Phase I trials should replace inter-patient by intra-patient dose escalation in order to give each patient a chance to benefit from the treatment. Since such trials frequently cause discomfort, they should be limited to one centre at a time. The number of patients subjected to phase II trials can be appreciably reduced if reasonable statistical measures are taken, as proposed here, to optimize and minimize the trial. Phase III trials must improve the relevance of the questions asked, reporting of errors, patient information and statistical interpretation. They must be based on promising phase II studies with historical controls so that their efficacy is increased.