How can we allow for patient and clinician preferences for treatment in clinical trials?

Abstract:

We will discuss the design of clinical trials that take patient or clinician preferences into account. We will focus mainly on the two-stage randomised trial, in which some participants are randomised in the usual way, but others can choose their own treatment. This approach permits an assessment of the extent to which patient outcomes are affected by the choice of treatment that patients would make (if that were allowed), and by whether patients actually receive their preferred treatment. These two effects can sometimes be important, even when the usual treatment effect size is small. Preference-based designs provide deeper insight into the determinants of good outcomes, and recognise the individuality of patients.