Incorporating Patient Preferences Into Mathematical and Statistical Models in Health Economics: A Taxonomy of Approaches

OBJECTIVES: The application of theory-driven methods to study patient preferences is now increasingly common in health. There is a paucity of research showing how these methods can be incorporated into other approaches in health economics such as simulation modeling or cost-effectiveness analysis. This has hindered their application in decision making, especially in health technology assessment. We developed a novel taxonomy demonstrating the potential mechanisms for integrating patient preferences into modeling techniques in health economics.

METHODS: We conducted a literature review of the published papers on mathematical and statistical models in health economics up to December 2022 and assessed the utilization of patient preference data into the models. We identified key modeling approaches and conceptualized potential mechanisms that could benefit from the incorporation of patient preference data as well as opportunities and challenges for future applications

RESULTS: We identified three major modeling approaches that could be augmented with preference data: compartmental Markov modeling, agent-based microsimulation modeling, and cost-effectiveness analysis. Through review of underlying methodologies, we propose that integration of patient preferences could :(1) improve the accuracy of model estimates by accounting for the effects of preferences on healthcare service choices, uptake, and adherence; (2) improve the precision of model estimates by facilitating personalized estimates and recommendations; (3) accommodate dynamic changes in preferences over time, including through secular trends (Markov models) and the influence of social networks (microsimulation). While these pose opportunities for more patient-centered decision making, systematic collection of preference data and model integration and validation can be complex and challenging.

CONCLUSIONS: Incorporating preference-based methods could contribute to the development, design, and interpretation of models in health economics and presents an opportunity to enhance the applicability and effectiveness of decision making.