Using clinical preferences in argumentation about evidence from clinical trials