Structured Expert Elicitation in HTA Submissions: A Practical Guide Using the Example of Pembrolizumab for Treating Relapsed or Refractory Classical Hodgkin Lymphoma

Speaker(s)

Lamb A1, Barton S2, Fordham R2, Asaadi A2, Engelbrecht K2, Maconachie R2
1MSD (UK) Limited, Horsham, WSX, UK, 2MSD (UK) Limited, London, England, UK

Presentation Documents

OBJECTIVES: Expert opinion can be elicited to inform health economic modelling parameters in evidence light areas. The NICE methods guide states that structured expert elicitation (SEE) methods, adhering to existing protocols are preferred to unstructured approaches as they minimize bias and provide an indication of uncertainty around estimates. These gold standard approaches can be challenging to implement due to the constraints of timelines and clinician availability.

Relatively few SEE exercises have been submitted to NICE. We conducted a SEE to inform model parameters for a NICE evaluation of pembrolizumab for UK patients with relapsed/refractory classical Hodgkin lymphoma (RRcHL). Here we present key learnings, aiming to provide practical advice to researchers conducting SEE exercises to support HTA submissions.

METHODS: We developed a bespoke SEE exercise aligned to the MRC SEE protocol by adapting publicly available STEER resources. Experts received background information and training, then completed a SEE using the variable interval method (chips and bins) in Excel. A follow-up meeting was held to discuss inter-expert variability and understand the rationale behind the expert’s chosen values. The experts’ estimates were used directly in the economic model and to validate modelled survival outcomes.

RESULTS: We found that publicly available resources for conducting SEE exercises were invaluable for constructing a robust exercise that was aligned with the MRC protocol. Discussion of inter-expert variability was a key step in the process as it allowed experts to provide clear clinical rationales for their estimates and fully contextualize any uncertainties around them. The SEE exercise successfully captured key model input parameters and these were viewed as acceptable for decision making by the NICE committee.

CONCLUSIONS: Utilizing inputs from the SEE, pembrolizumab was recommended for routine use in the NHS for treating RRcHL for patients who have already had brentuximab vedotin and are ineligible for autologous stem cell transplant (NICE TA967).

Code

HTA160

Topic

Clinical Outcomes, Economic Evaluation, Methodological & Statistical Research, Study Approaches

Topic Subcategory

Comparative Effectiveness or Efficacy, Cost-comparison, Effectiveness, Utility, Benefit Analysis, Survey Methods, Surveys & Expert Panels

Disease

Drugs, Oncology