Abstract

KIE: Four methodological problems in the design of clinical trials are discussed. First, since it is ethical to discontinue a trial as soon as one treatment appears clearly superior, long-term effects can be missed. Secondly, in non-drug treatments such as surgery or psychotherapy, the technical skill of the therapist can have an important effect on results. Thirdly, therapy on one group of patients can have differing "spill-over" effects on other groups because of differences in organization or professional competence. Finally, since expensive therapies are those most often studied, while poor institutions or countries may be forced to rely on cheaper undocumented therapies, costs and level of technology should be considered before conducting trials.