Presentation (CTI)

OBJECTIVES: eDelphi studies are frequently used to address persisting evidence uncertainties where available evidence is limited, conflicting, or outdated, to inform clinical decision-making. Development of a robust survey instrument is critical to ensure expert consensus can be meaningfully assessed. This abstract describes the methodological challenges and lessons from developing an eDelphi survey instrument to evaluate referral candidacy and treatment eligibility for chimeric antigen receptor (CAR) T-cell therapy in immuno-oncology.
METHODS: A global eDelphi study was designed to identify and evaluate core clinical criteria assessed by referring community physicians and additional criteria assessed by cellular therapy specialists at authorized treatment centres. Survey instrument development was informed by an evidence synthesis adapted for rapidity and relevancy to the survey objectives, followed by iterative refinement through engagement with an international scientific steering committee of therapeutic area experts (TAE). Methodological priorities included translating complex clinical concepts into structured statements suitable for consensus assessment and defining threshold-based variables where required. Across three survey rounds, refinement included rewording statements and converting open-ended questions into structured categorical variables.
RESULTS: Survey instrument development revealed several methodological challenges, including differentiating referral candidacy from definitive treatment eligibility, operationalizing clinical criteria into consensus statements, and defining clinically meaningful threshold values where evidence was limited or heterogeneous. Iterative refinement improved clarity and interpretability while preserving clinical relevance. TAE involvement was integral for resolving ambiguity, refining terminology, and conversion of open-ended questions into meaningful threshold categories for consistent evaluation of complex clinical criteria.
CONCLUSIONS: eDelphi is a robust methodology for developing consensus-based recommendations in areas of clinical uncertainty. However, the quality and interpretability of findings depend heavily on rigorous survey instrument development. Combining adapted evidence synthesis methods, expert steering committee input, and iterative refinement can strengthen instrument accuracy and validity and robustly support development of actionable recommendations for complex referral and eligibility decisions in real-world practice.