A primer on expert elicitation with IDEA (Investigate Discuss Estimate Aggregate) for rare disease trialists

Abstract In clinical trials for rare diseases, small patient populations constrain recruitment, limiting data collection. Expert elicitation, the process of extracting expert beliefs into quantitative expressions, can act as a source of evidence in a setting with limited data. Expert judgement can be combined with trial data through the prior distribution in a Bayesian trial. This approach makes the most use out of the available information and possible trial recruitment, unlike conventional approaches that rely on trial data alone. Previous elicitations for rare disease trial prior distributions have used ad hoc approaches, rather than an established protocol designed to limit biases in judgement. This practical guide introduces rare disease clinical investigators to expert elicitation using the IDEA (Investigate, Discuss, Estimate, Aggregate) protocol. IDEA is a structured elicitation technique where experts individually estimate quantities, discuss their answers, then revise their estimates before aggregation into a final distribution. In an elicitation for a trial, experts can either estimate the control arm response or the treatment effect. The elicitation estimates then inform the design and analysis of the trial. IDEA has previously been described in a general context, but this paper provides tailored recommendations for conducting an elicitation for a clinical trial for a rare disease. By outlining practical steps to using IDEA, we aim to promote more structured, transparent practices that will increase the uptake and acceptance of expert elicitation to improve clinical trials for rare diseases.