Are We Making Expert Use of Experts? A Scoping Review of Expert Elicitation in Survival Extrapolation

OBJECTIVES: Extrapolation of immature survival data is important within health technology assessments (HTAs) but is often associated with large uncertainty. Incorporation of expert judgements can be informative in survival extrapolation. However, these judgements are often perceived as being “best guesses” due to the lack of transparency in the elicitation process. This scoping review assesses the use of expert elicitation within survival extrapolation in published articles.

METHODS: MEDLINE was searched for articles from inception to October 2023. The search algorithm included terms for expert opinion as well as structured expert elicitation methodologies used within survival modelling. The search was supplemented using a pearl growing approach of three “seed” papers. Articles were double screened according to whether they included expert elicitation or consultation to guide survival extrapolation. Data relating to the elicitation methodology and reporting detail were extracted from relevant articles.

RESULTS: The search yielded 165 unique articles, after the full-text review 16 articles were deemed relevant. The majority of articles (n=13, 81%) were case studies publishing cost effectiveness analyses within the oncology setting. The implementation of expert elicitation was typically described in minimal detail, with less than half of the studies making methodological reference to formal structured elicitation frameworks such as SHELF, IDEA or the MRC protocol.

CONCLUSIONS: To our knowledge this is the first review of the use of expert elicitation within the survival analysis setting specifically. We have discovered that generally, the reporting of expert elicitation approaches within long-term survival extrapolation often lacks important methodological details. There appears to be significant scope for improving the implementation of expert elicitation within survival analysis. We therefore envision that with the development of bespoke methodologies and reporting guidelines specific to survival extrapolation, the general perception of elicited survival quantities will improve and likewise the credibility of expert judgements within HTAs.