Presentation (CTI)

OBJECTIVES: Patient preference evidence is increasingly used to inform health technology assessment (HTA) and regulatory decision-making. However, discussions often focus on methodological taxonomy rather than the decision questions these methods are best suited to answer. This study aimed to develop a practical framework for matching patient preference methods to specific evidence needs, including outcome prioritisation, benefit-risk assessment, treatment acceptability, subgroup heterogeneity, and patient-centred trade-offs.
METHODS: A narrative review was conducted of recent literature on patient preference elicitation in HTA and Health Economics and Outcomes research (HEOR), focusing on qualitative methods, ranking/rating exercises, best-worst scaling (BWS), threshold techniques, and discrete choice experiments (DCEs). Methods were compared according to five decision-relevant criteria: type of question addressed, ability to quantify trade-offs, respondent burden, feasibility in small or rare-disease populations, and interpretability for regulatory, payer, or HTA audiences.
RESULTS: Preference methods differed less in overall “strength” than in their fit to specific evidence questions. DCEs were most suitable for estimating relative attribute importance, modelling benefit-risk trade-offs, and exploring preference heterogeneity, but required larger samples and higher respondent effort. BWS was well suited to prioritising outcomes or treatment features, particularly where simple ranking may lack discrimination, but offered less insight into complex trade-offs. Ranking/rating exercises were feasible and accessible but most appropriate for exploratory or supportive evidence. Threshold techniques were identified as underused but highly relevant where the decision problem concerns acceptability of a specific risk, burden, or loss of efficacy in exchange for a patient-valued benefit. Qualitative methods remained foundational for identifying patient-relevant attributes, refining language, and ensuring quantitative tasks reflect lived experience.
CONCLUSIONS: No patient preference method is universally optimal. Method selection should begin with the decision question. A decision-question framework can support more purposeful integration of patient preference evidence into HTA and HEOR by aligning evidence generation with decision-maker needs.