OBJECTIVES: The value of incorporating the patient voice into drug development is well established. Patient Preference studies (including discrete choice experiments [DCEs]) continue to rise in popularity to inform product design, trial design, product approval, and post-approval decisions. Recent methodological recommendations and regulatory guidance emphasize the importance of patient-centric preference study design.

METHODS: Additional recommendations regarding the design and testing of patient preference surveys are provided based on widely-accepted and regulatory-endorsed standards for Clinical Outcome Assessment (COA) development and validation (e.g., FDA Patient-Reported Outcomes [PRO] and Patient-Focused Drug Development [PFDD] guidance series).

RESULTS: It is important that selected attributes and levels are reflective of the decision-making context. This can be achieved by reviewing existing published literature, product prescribing information, and/or patient-facing materials. Engagement with field experts can ensure attribute framing/level selection is reflective of discussions in clinical practice. Qualitative testing of surveys (attributes/levels and supporting materials) is critical for establishing content validity, especially given complexities associated with communicating the benefits/risks of treatments. Direct feedback from the target population can support relevance/importance of selected attributes and proposed level ranges/increments, identify missing attributes, and highlight preliminary treatment drivers to support refinement of selected attributes/levels. In-depth cognitive debriefing using established methods (e.g., think-aloud techniques) is valuable for evaluating the validity of attributes/levels (e.g., consistent and intended understanding/interpretation, minimizing grounding effects) and the appropriateness of survey design and layout (i.e., usability, responder burden) which may bias responses. Sample sizes for qualitative testing should ensure adequate representation of the target survey population with respect to socio-demographic (e.g., education level) and clinical characteristics.

CONCLUSIONS: Parallels between preference study design and COA/PRO development can be drawn. To maximise the utility and validity of health patient preference studies, in-depth qualitative pre-testing is recommended to ensure that preference surveys are fit-for-purpose in the specific context-of-use.