Patient preferences: a Trojan horse for evidence-based medicine?

The evidence-based medicine (EBM) movement has long acknowledged the relevance of patient preferences and values. According to EBM, clinicians first clarify the medical evidence about the benefits and burdens of the treatment in question and then, as a second step, elicit values and preferences from patients. Importantly, however, values are placed on patient-relevant outcomes. Surrogate endpoints are only used if their validity is proven. This article shows that some recent patient-preference studies attribute value to surrogate endpoints even when there is no improvement in patient-relevant outcomes. The article points out their foundation in neoclassical economics and discusses their clash with principles of EBM and medical ethics.