This paper focuses on the ethics of constructing and using a specific evidence-based decision aid that aims to contribute to clinical shared decision-making processes. Results of this integrated empirical ethics study demonstrate how both the production and presentation of scientific information in an evidence-based decision-support contain implicit presuppositions and values, which pre-structure the moral environment of the shared decision-making process. As a consequence, the evidence-based decision support did not only support the decision-making process; it also transformed it in a morally significant way. This phenomenon undermines the assumption within much of the literature on patient autonomy and shared decision-making implying that information disclosure is a conditional requirement before patient autonomy and shared decision-making even starts. The central point of this paper is that decision aids and evidence-based medicine are not value-free and that patient autonomy and shared decision-making are already influenced during the production and presentation of scientific information. Consequences for both the development of decision-aids and the practice of shared decision-making are discussed.