Edinburgh Health Economics Digital Seminar: Does bias in personalised Decision Aids lead to lower use by doctors: a lab experiment Title: Does bias in personalised Decision Aids lead to lower use by doctors: a lab experiment Speaker: Alistair Irvine, Scottish Government Biography: Alastair completed his PhD in health economics at HERU, University of Aberdeen, in 2018 before spending a year as a research fellow, then joining Scottish Government in November 2019. His research interests are in time and risk preferences, patient-doctor interactions, and game theory. His Government work has been on Minimum Unit Pricing, obesity, and the response to COVID-19 in Scotland. Abstract: Asymmetric information is a barrier to doctors and patients choosing utility maximising treatments. Doctors can decide how much effort to take to overcome this asymmetry. Decision Aids (DA) can reduce the effort needed and have been promoted as part of shared decision-making. Recently, DAs have used stated-preference methods to predict patients’ ‘preferred’ treatments. These predictions may be inaccurate for some patients, leading to welfare losses. The DA may omit relevant treatment attributes, or the information presentation may induce biases in patients’ responses. Increased use of, and potential biases in, DA treatment recommendation mean it is important to know whether doctors reduce the use of DAs when the bias is sufficiently large. This is tested in a lab experiment in which doctors spend effort points to receive information about patients’ preferences. Information quality depends on patient type and doctor effort. Participants’ income is decreasing in effort: the DA effort cost is between that of short and long conversations. Payments are based on one randomly chosen patient, and we donate the patient’s utility from treatment to charity. The experiment is developed in oTree, using 98 students in the doctor role. Our results will have implications for understanding how DAs are used in practice, potentially anticipating barriers to adoption. Please email Andrew Stoddart for the joining instructions: andrew.stoddart@ed.ac.uk Aug 21 2020 14.00 - 15.00 Edinburgh Health Economics Digital Seminar: Does bias in personalised Decision Aids lead to lower use by doctors: a lab experiment Edinburgh Health Economics Digital Seminar Microsoft Teams
Edinburgh Health Economics Digital Seminar: Does bias in personalised Decision Aids lead to lower use by doctors: a lab experiment Title: Does bias in personalised Decision Aids lead to lower use by doctors: a lab experiment Speaker: Alistair Irvine, Scottish Government Biography: Alastair completed his PhD in health economics at HERU, University of Aberdeen, in 2018 before spending a year as a research fellow, then joining Scottish Government in November 2019. His research interests are in time and risk preferences, patient-doctor interactions, and game theory. His Government work has been on Minimum Unit Pricing, obesity, and the response to COVID-19 in Scotland. Abstract: Asymmetric information is a barrier to doctors and patients choosing utility maximising treatments. Doctors can decide how much effort to take to overcome this asymmetry. Decision Aids (DA) can reduce the effort needed and have been promoted as part of shared decision-making. Recently, DAs have used stated-preference methods to predict patients’ ‘preferred’ treatments. These predictions may be inaccurate for some patients, leading to welfare losses. The DA may omit relevant treatment attributes, or the information presentation may induce biases in patients’ responses. Increased use of, and potential biases in, DA treatment recommendation mean it is important to know whether doctors reduce the use of DAs when the bias is sufficiently large. This is tested in a lab experiment in which doctors spend effort points to receive information about patients’ preferences. Information quality depends on patient type and doctor effort. Participants’ income is decreasing in effort: the DA effort cost is between that of short and long conversations. Payments are based on one randomly chosen patient, and we donate the patient’s utility from treatment to charity. The experiment is developed in oTree, using 98 students in the doctor role. Our results will have implications for understanding how DAs are used in practice, potentially anticipating barriers to adoption. Please email Andrew Stoddart for the joining instructions: andrew.stoddart@ed.ac.uk Aug 21 2020 14.00 - 15.00 Edinburgh Health Economics Digital Seminar: Does bias in personalised Decision Aids lead to lower use by doctors: a lab experiment Edinburgh Health Economics Digital Seminar Microsoft Teams
Aug 21 2020 14.00 - 15.00 Edinburgh Health Economics Digital Seminar: Does bias in personalised Decision Aids lead to lower use by doctors: a lab experiment Edinburgh Health Economics Digital Seminar
