Session (CTI)

PURPOSE: This workshop will guide participants along the trail of individual patient preference data from the moment a patient responds to a survey, through analytic methods that transform those responses into actionable insights, and finally to the clinical encounter where these insights inform shared decision-making. DESCRIPTION: Participants will begin their journey with a threshold exercise, experiencing firsthand the cognitive process patients navigate when expressing risk tolerance using this approach. Next, discussants will compare approaches for estimating patient-level preferences and risk-tolerance (probablistic threshold technique, the multidimensional threshold technique, respondent-level estimates from mixed logit models, and segmentation of individuals across latent classes that vary by risk-tolerance). Discussants will highlight practical considerations across methods including sample-size requirements, participant burden, and within-person consistency. We will conclude with a demonstration of how latent-class findings can be utilized to design predictive questions for decision-support tools, enabling clinicians to personalize benefit-risk discussions for individual patients. Attendees will conclude their journey with a greater understanding of the patient experience (i.e. what it feels like to contribute individual level preference data) and the researcher experience (i.e. weighing the methodological differences and practical applications across methods). Agenda: • Welcome and introduction (5 minutes) • Participate in threshold exercise (10 minutes) • Share/reflect break-out (5 minutes) • Probablistic Threshold and DCE Case Study (10 minutes) • Multidimensional Threshold Case Study (10 minutes) • Method Compare Q&A (5 minutes) • Individual Latent-class Predictions (10 minutes) • Questions/Group Discussion (5 minutes)