Skip navigation

Patients' preferences

Will randomised trials eventually become impossible? The question is academic but not unrealistic as the problem of recruiting patients to trials increases. To recruit patients to randomised trials there must be what is called "equipoise" - genuine doubt on behalf of the investigator recruiting patients about which treatment is better. But how does the patient - the recruitee - feel about it? What happens when recruitment levels are low - when potential recruitees made a decision not to participate in a trial?

Particular problems arise when evaluating interventions which require active patient participation and therefore motivation, for example sticking to a diet. Some people have developed what is called "patient preference design" to help them deal with this. The subject is dealt with in an article [1] in the Lancet. A number of strategies are described and discussed for those who would wish to avoid the randomised controlled trial as the best test of effectiveness. This article points out the problems in all these other methods and emphasises that randomised controlled trials should not be seen simply as a means of testing effectiveness, but also as a means of protecting patients.

This thoughtful work arises from the work of a Cochrane Collaboration methods group, emphasising the importance of the Collaboration not simply as a systematic review production line, but as an international movement designed to improve the quality of randomised trials as well as accessibility and usefulness of the results of the trials.


  1. W Silverman, D Altman. Patients' preferences and randomised trials. Lancet 1996 347:171-4.

previous or next story in this issue