Synthetic Preference Data in Digital Twins for Personalized Healthcare Decisions

PURPOSE: To present the benefits and applications of integrating synthetic preference data into digital twins for enhancing personalized healthcare decision-making. Synthetic data may solve data scarcity, privacy concerns, and evidence transportability. Synthetic data generation methods promise to replicate real-world scenario complexity and representativeness. Researchers can simulate alternative treatment plans and explore resulting outcomes to optimize decision-making by enabling the creation of digital twins. We will explore the innovative use of synthetic data for real-time elicitation of patient preferences in a wide range of decision-making settings. The application of digital twins in the studies of patient behavior and statistical methods available to develop synthetic preference data will be presented.

DESCRIPTION: Workshop attendees will obtain a working knowledge of the fundamentals of digital twins and the application of synthetic preference data. The workshop will review a). the basics of synthetic data, b). methodological advancement needed to develop synthetic data for preference studies c). practical applications of synthetic preference data in the further development of personalized healthcare. Dr Mühlbacher will chair the session, introducing key definitions and contextualizing the applicability of synthetic data in the field of preference studies (10 min). Dr Manca will illustrate the causal inference underpinning the validity of the models that produce predictions to generate synthetic data and make generalizable statements (15 min). Dr Kolasa and Dr Zah will exemplify the benefits of incorporating synthetic patient preferences in personalized healthcare, enhancing medication adherence, and managing chronic diseases through personalized interventions (15 min). The highlight of the workshop is the audience participation, involving the identification of problems and solutions for a hypothetical case study in surgical planning and risk management (20 min). This interactive workshop will be valuable for researchers, clinicians, and industry analysts interested in integrating synthetic data with real-world data for a broader range of decision-making applications.