Developing economic experiments to understand doctor behaviour

Decision-Aids help doctors and patients make shared decisions. However, the design of a DA necessarily leaves out some aspects of treatment that are relevant to a proportion of the population. Recent work has moved towards decision aids that make a prediction about the patient’s most preferred treatment. Excluded attributes in the DA, or other question biases such as complex probabilities, could bias the prediction of a Decision Aid. Doctors are engaged in a screening problem, matching patients to treatments, with signals of varying biasedness from different elicitation tools. The project aims to establish whether doctors reach the second-best equilibrium when a decision aid is (or is not) the optimal tool to do this.

HERU investigators involved in this project:  Alastair Irvine, Verity Watson, Marjon van der Pol

External collaborators: Regier, D. (University of British Columbia)