Presentation (CTI)

OBJECTIVES: National Institute of Health and Care Excellence (NICE) methods guidance indicates EQ-5D-3L as the preferred health-related quality of life measure. As EQ-5D-3L data may not always be available, mapping can be applied to other non-preference or preference-based measures to generate EQ-5D-3L utilities, with prior literature identifying mapping in TAs usage in a quarter of NICE submissions. Our review sought to evaluate the frequency of usage, as well as key characteristics, of mapping algorithm usage in NICE appraisals over a 5-year period.
METHODS: NICE appraisals were screened for relevance between 1^st March 2020 to February 28^th 2025. Terminated or superseded appraisals, along with appraisals without cost-utility analyses, were excluded in first pass review. In second pass, appraisals without mapping conducted were excluded, with remaining appraisals reviewed to assess the characteristics and justifications for mapping algorithms applied.
RESULTS: Of 626 appraisals screened, 442 (70.6%) were excluded for not including descriptions of mapping algorithm usage (261 and 181 in first and second pass, respectively). Of 184 appraisals included, 126 (68.5%) used cross-walking from EQ-5D-5L to EQ-5D-3L, with 75/126 (60%) using the van Hout 2012 crosswalk algorithm. 64/184 (35%) used other types of utility mapping, most commonly (37.5%; 24/64) from EORTC-QLQ-C30 to EQ-5D-3L. Among these, mapping algorithms were most commonly identified from prior NICE appraisals, literature review or the HERC database. More than half either did not report selection criteria or relied on precedence from prior appraisals. While some critique identification and selection criteria was provided, EAG feedback appeared highly specific rather than based on comprehensive review methods.
CONCLUSIONS: Our review found that mapping still plays an important role in NICE appraisals, although non-cross-walk related mapping usage appears relatively infrequent (~10% of appraisals screened). Mapping algorithm identification methods were somewhat inconsistent across appraisals with generally limited reporting of selection criteria, indicating an unmet need for clearer guidelines.