The Use of Formal Expert Opinion Elicitation to Inform Model Structure and Inputs in Rare Diseases

OBJECTIVES: An economic model is composed of two main elements: structure and inputs. The structure is usually obtained through an investigation of the care pathway and the inputs from quantitative data. Sometimes it can be challenging to identify the pathway followed by a patient in the healthcare system. For example, in rare diseases when there are no accurate or non-invasive tests to identify the condition, the diagnostic pathway is usually complex and variable. Furthermore, clinical trials are expensive in these conditions because of the complexity of patient recruitment, causing a lack of robust evidence of treatment outcomes. These elements influence the robustness of the evidence base creating difficulties when interpreting and generalising the results of the economic analysis.

METHODS: We used a robust methodology to inform the structure and inputs of a cost-effectiveness model of treatments in adult-onset Still’s disease. The methodology comprised:

• A care pathway analysis where we interviewed eleven clinicians from multiple regions in Europe to establish the common elements and discrepancies in the pathway.
• Two systematic reviews to find the published inputs for quality of life, resource use, and clinical outcomes.
• A structured expert opinion elicitation (including training of experts to minimise bias, an individual elicitation exercise, a group workshop to discuss discrepancies, and mathematical aggregation) to elicit 34 missing inputs with plausible ranges from five experts across four countries.

RESULTS: In 12 months, we developed a flow-diagram of the care pathway, validated by the experts, that informed the model structure. We also identified the uncertainties linked to the evidence base and elicited 34 probability distributions of the missing inputs.

CONCLUSIONS: We identified a robust methodology that can be used by other researchers to provide information and data that are missing from the literature, especially when developing and populating cost-effectiveness models for rare diseases.