Using clinical preferences in argumentation about evidence from clinical trials

Medical practice is increasingly based on the best available evidence, but the volume of information requires many clinicians to rely on systematic reviews rather than the primary evidence. However, these reviews are difficult to maintain, and often do not appear transparent to clinicians reading them. In a previous paper, we have proposed a general language for representing knowledge from clinical trials and a framework that allows reasoning with that knowledge in order to construct and evaluate arguments and counterarguments that aggregate that knowledge. However, clinicians need to feel that such a framework is responsive to their assessment of the strengths and weaknesses of different types of evidence. In this paper, we use a specific version of this existing framework to show how we can capture clinical preferences over types of evidence, and we evaluate this in a pilot study, comparing our system against the choices made by clinicians. This pilot study shows how individual clinicians aggregate evidence based on their preferences over the relative significance of the items of evidence, and it shows how our argumentation system can replicate this behaviour.