ISSUE: Preference elicitation methodologies require researchers to make assumptions about patients’ choice behavior, including their willingness to make trade-offs, the presence or nature of heterogeneity in preferences, and the role of the status-quo option. These assumptions guide the design of stated preference instruments as well as the analysis of collected data. However, common assumptions may not always adequately reflect respondents’ choice behaviour. Understanding the implications can help improve instrument design, and may guide the development of targeted validity tests and analytical approaches.

OVERVIEW: The panel will initially introduce fundamental assumptions made in health preference research. Panellists will then provide insights into how and why some of these assumptions may be violated and the potential implications that these have on instrument design, analysis, and data interpretation. Dr. Heidenreich will outline implications of non-trading behavior in probabilistic thresholding. He will give an example of how internal validity tests may be able to detect such behavior and provide advice on adequate analytical approaches. Dr. Boeri will discuss challenges resulting from protest behaviour and serial non-participation. He will outline the use of hurdle models to account for such choice behavior. Dr Ng will discuss the role of the status quo in choice experiments and outline how status-quo options have been handled in the literature and the common pitfalls observed. Dr Tervonen will discuss the role patient-reported outcomes can play in understanding preference heterogeneity using two case studies. The panel will close with an open discussion with the audience about implications of behavioral heterogeneity in patient preference research. Patient preference researchers and HEOR scientists benefit from attending this advanced panel.