# Structured expert elicitation for long-term survival outcomes in health technology assessment: a systematic review

## TL;DR

This paper reviews how expert opinions are used to estimate long-term survival in health technology assessments, finding that methods are inconsistent and need improvement.

## Contribution

The study identifies gaps in the structured use of expert elicitation for survival outcomes in health assessments and suggests areas for improvement.

## Key findings

- Only six studies used structured expert elicitation for long-term survival outcomes in the broader literature.
- Four NICE appraisals used elicitation between October 2023 and October 2024, but reporting was inconsistent.
- Elicited values are often used only for qualitative validation, not fully integrated into analyses.

## Abstract

Extrapolation of immature survival data is integral to health technology assessment (HTA) but is often associated with large uncertainty. Incorporation of expert judgements can help to address this uncertainty, but often these judgements are perceived as being “best guesses” due to a lack of methodological transparency. This review assesses the current implementation and reporting of structured expert elicitation for long-term survival outcomes in the broader literature and recent submissions to the National Institute for Health and Care Excellence (NICE).

Three literature databases were searched: MEDLINE via Ovid, Embase, and Web of Science. The search algorithm included terms to identify articles which obtained expert judgements for survival extrapolation. A pearl-growing approach was also employed using three seed papers to supplement the electronic searches. To identify recent NICE technology appraisals in oncology, the NICE guidance database was searched.

The search of the broader literature identified six studies which utilised structured expert elicitation for long-term survival outcomes. Four NICE technology appraisals were identified to have used structured expert elicitation between October 2023 and October 2024. The reporting and conduct of elicitation for survival quantities was variable in detail and rigour in both the broader literature and NICE submissions. Despite the requirement for significant resource investment within the process, elicited values are variably used within analyses and are predominantly used as qualitative external validation. At all points throughout the elicitation, including planning, conduct and reporting, there appears to be a considerable lack of technical detail, which in the context of NICE appraisals may hinder full consideration of the elicited values by reviewing committees.

Currently, the methods and limited reporting structures being used to elicit long-term survival outcomes are not fit for purpose. This review highlights key areas for improvement and identifies examples of good practice when conducting structured expert elicitations for long-term survival outcomes.

The online version contains supplementary material available at 10.1186/s12911-025-03221-2.